article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
this study is concerned with the inversion of fundamental stellar parameters from the analysis of high - resolution echelle spectra .
hereafter , we shall focus indeed on data collected since 2006 with the narval spectropolarimeter mounted at the 2-m aperture _
tlescope bernard lyot _
( tbl ) telescope located at the summit of the _ pic du midi de bigorre _ ( france ) .
we investigate , in particular , the capabilities of the principal component analysis ( hereafter pca ) for setting - up a fast and reliable tool for the inversion of stellar fundamental parameters from these high - resolution spectra .
the inversion of stellar fundamental parameters for each target that was observed with both narval and espadons spectropolarimeters constitutes an essential step towards : _ ( i ) _ the further post - processing of the data like e.g. , the extraction of polarimetric signals ( see e.g. , paletou 2012 ) but also _ ( ii ) _ the exploration , or data mining , of the full set of data accumulated over the last 8 years now . in section 2
, we briefly describe the actual content of such a database .
pca have been used for stellar spectral classification since deeming ( 1964 ) .
it has been in use since , and more recently for the purpose of the _ inversion _ of stellar fundamental parameters from the analysis of spectra of various resolutions .
it is however most often used _ together _ with artificial neural networks ( see e.g. , bailer - jones 2000 ; re fiorentin et al .
pca is used there for reducing the dimensionality of the spectra before attacking a multi - layer perceptron which , in turns , allows to link input data to stellar parameters .
our usage of pca for such an inversion process is strongly influenced by the one routinely made in _
spectropolarimetry during the last decade after the pionneering work of rees et al .
( 2000 ) . very briefly
, the reduction of dimensionality allowed by pca is directly used for building a specific metric from which a nearest neighbour(s ) search is done between an observed data set and a learning database .
the latter is made of synthetic data generated from input parameters properly covering the _ a priori _ range of physical parameters expected to be deduced from the observations themselves .
a quite similar use of pca was also presented for classification and redshift of galaxies estimation by cabanac et al .
fundamental elements of our method and its basic capabilities are exposed in sections 3 and 4 .
one of its originality relies on the simultaneous inversion of the effective temperature @xmath0 , the surface gravity log@xmath1 , the full metallicity [ m / h ] and the projected rotational velocity @xmath2 . in this study
we restrict ourselves , on purpose , to the spectral domain that will be observed by the rvs spectrometer on - board the gaia spacecraft ( katz et al .
the rvs will finally operate in a spectral domain of the order of 847871 nm ( f. thvenin , private communication ) . however , for our study we shall use the nominal spectral domain mentioned up to 2012 in the litterature and which covers the spectral domain 847874 nm , at the vicinity of the caii infrared triplet .
the pertinence of this choice for the further characterization of the observed stars was discussed by munari ( 1999 ) .
this allows us to anticipate reasonable stellar parameter inversions for stars from b8 to m8 , a very large range of spectral types similar to the actual content of our database .
the conditioning of observed spectra prior to their ingestion into our inversion tool is detailed in section 5 , and first tests made with solar spectra observed by reflection over the surface of the moon are discussed in section 6 .
additional tests are also presented and discussed in section 7 , mainly for fgk - dwarf stars for which fundamental parameters are already available from the so - called spocs catalogue ( valenti & fisher 2005 ) .
we then briefly discuss preliminary tests using spectra from giant stars as well as hotter and cooler than fgk stars .
we mainly used narval data available from the _ public _ database tblegacy .
narval is a state - of - the - art spectropolarimeter operating in the 0.38 - 1 @xmath3 m spectral domain , with a spectral resolution of 65000 in its polarimetric mode .
it is an improved copy , adapted to the 2-m tbl telescope , of the espadons spectropolarimeter , in operations since 2004 at the 3.6-m aperture cfht telescope .
the tblegacy database is operational since 2007 .
it is at the present time the largest on - line archive of high - resolution polarization spectra .
it hosts data that were taken at the 2-m tbl telescope since december 2006 .
so far , more than 70000 spectra have been made available , for more than 370 distinct targets all over the hertzsprung - russell diagram .
more than 13000 _ polarized _ spectra are also available , mostly for _ circular _ polarization .
linear polarization data are very seldom still and amount to a few hundreds spectra , but they are equally available . at the present time , the tblegacy database provides no more than stokes @xmath4 or @xmath5 spectra calibrated in wavelength .
stokes @xmath4 data are either normalized to the local continuum or not .
we have however plans : _ ( i ) _ to extend it to espadons data from the cfht telescope and _ ( ii ) _ to propose higher - level data , such as pseudo - profiles resulting from line addition and/or least - squares deconvolution ( see e.g. , paletou 2012 ) , activity indexes as well as stellar fundamentals parameters
. the latter s knowledge , besides being obviously interesting by itself , is also indispensable to any accurate further post - processing of these high - resolution spectra .
these spectra are also generally bearing high signal - to - noise ratios , as can be seen in fig.([fig1 ] ) .
indeed stokes @xmath4 spectra we have been using result from the combination of 4 successive exposures , each of them carrying 2 spectra of orthogonal polarities generated by a savart plate - type analyser .
this procedure of double so - called `` beam - exchange '' measurement is indeed meant for the purpose of extracting ( very ) weak polarization signals ( see e.g. , semel et al .
our pca inversion tool is strongly inspired by magnetic and velocity field inversion tools which have been developed during the last decade to complement solar spectropolarimetry ( see e.g. , rees et al .
improvements of this method have been recently exposed by casini et al .
( 2013 ) for instance .
hereafter we describe its main characteristics , in the context of our study .
let us call \{@xmath6 } our training database of _ synthetic _ stellar spectra .
each of these spectra is characterised by a limited number of physical parameters which serve as input parameters for the numerical code producing them .
usually , for so - called _ standard _ stellar models , the minimal set of parameters is the effective temperature of the star , @xmath0 , its surface gravity , log@xmath1 , its _ total _ metallicity [ m / h ] ( even though the specific iron abundance [ fe / h ] is also frequently used ) and a so - called microturbulent velocity @xmath7 which is an artificial contributor to line widths , in addition to thermal ( or doppler ) broadening .
because a significant part of our database is constituted of moderately , say @xmath8 to fast rotators , we adopted the collection of spectra already computed and made available by munari et al .
they used castelli & kurucz atlas code and they considered parameters spanning ranges from 3500k to 47500 k for @xmath0 , 0 . to 5 . for log@xmath1 , -2.5 to 0.5 for [ m / h ] , two distinct values , 0 .
and 0.4 for [ @xmath9/fe ] , microturbulent velocities @xmath7 from 1 to 4 km@xmath10 and projected rotational velocities v@xmath11 ranging from 0 to 500 km@xmath10 .
our training database contains about 34757 synthetic spectra after we limited ourselves to those spectra for which [ @xmath9/fe]=0 , which may affect our determinations of [ m / h ] for some objects , and @xmath12 which choice may also not be optimal for all of our observations . also , only the so - called `` new odf '' ( where odf stands for opacity distribution function ; see castelli & kurucz 2003 ) spectra were selected for @xmath13 k. hereafter we shall focus on the inversion of the set of four parameters \{@xmath14 ; log@xmath1 ; [ m / h ] ; @xmath2 } for each of our observed spectra .
it is also important to note that we used munari et al .
( 2005 ) spectra computed for a resolving power of @xmath15 i.e. , about a factor of 3.25 _ less _ than the one of the observed spectra we want to process .
we shall come back to this point in sections 5 and 9 .
each spectra from espadons and narval spectropolarimeters provides of the order of 250000 flux measurements vs. wavelength across a spectral range spanning from 0.38 to 1 @xmath3 m typically .
hereafter we consider only spectra obtained in the polarimetric mode at a resolvance @xmath16 . indeed ,
one of our main objective is that stellar parameters derived from stokes @xmath4 spectra can be directly used for the further post - processing of the multi - line _ polarized _ spectra which comes together ( see e.g. , paletou 2012 and references therein ) . for this assessment study , we put ourselves on purpose in a tough situation by restricting the spectral domain from which we shall invert stellar parameters to the ( initial ) one of the rvs instrument of the gaia space mission , that is for wavelengths ranging from about 847 to 874 nm ( katz et al . 2004 ) .
arguments in favor of the use of this very spectral domain can also be found in munari ( 1999 ) . considering this ,
the matrix @xmath17 representing our training database turns to be @xmath18 by @xmath19 .
next we compute the eigenvectors @xmath20 of the variance - covariance matrix defined as @xmath21 where @xmath22 is the mean of @xmath17 along the @xmath23-axis .
therefore @xmath24 is a @xmath25 matrix . in the framework of principal component analysis , reduction of dimensionality
is achieved by representing the original data by a limited set of projection coefficients @xmath26 with @xmath27 . in what follows for the processing of all observed spectra
, we shall adopt @xmath28 . the most frequent argument supporting the choice of @xmath29 relies on the accuracy achieved for the reconstruction of the original set of @xmath30 s from a limited set of eigenvectors ( see e.g. , rees et al . 2000 or ramrez vlez et al .
2010 ) . in the present case
, we display in fig.([fig2 ] ) the maximum reconstruction error @xmath31 as a function of the maximum number of eigenvectors considered for the computation of the so - called _ admixture _ coefficients @xmath32 .
it is noticeable that this reconstruction error is better than 1% for @xmath33 .
however the potentiel effect of noise present in the observations we want to process should also be taken into account ( see e.g. , socas - navarro et al . 2001 ) .
we shall discuss this specific point in the next section .
.,width=359 ] practically , we also found convenient to use , at every @xmath34 order , @xmath35 s both centered to their average @xmath36 and normalized to their standard deviation @xmath37 . it is a common practice in the field of pattern recognition , since it guarantees that those coefficients from which we shall define the specific metric used for the nearest neighbour(s ) search will have comparable effects ( on the comparison between spectra ) .
the above described reduction of dimensionality allows one to perform a fast and reliable inversion of observed spectra , once the latter have been : _
( i ) _ corrected for the wavelength shift vs. synthetic spectra because of the radial velocity of the target , _
( ii ) _ continuum - renormalized as accurately as possible , _ ( iii ) _ degraded in spectral resolution to be comparable to the @xmath38 of the synthetic spectra and , finally _
( iv ) _ resampled in wavelength as the collection of synthetic spectra .
we shall come back to these various stages in the next section .
however once these tasks have been achieved , the inversion process is the following .
let @xmath39 the observed spectrum made comparable to synthetic ones .
we now compute the reduced set of projection coefficients @xmath40 where @xmath36 and @xmath41 are , respectively , the mean of the standard deviation of the @xmath42 s , for @xmath43 .
the nearest neighbour search is therefore done by seeking the minimum of the squared _
distance @xmath44 where @xmath45 spans the number , or a limited number if any _ a priori _ is known about the target , of distinct synthetic spectra in the training database . in practice , we do not limit ourselves to the nearest neighbour search , although it already provides a relevant set of stellar parameters . because pca - distances between several neighbours may be of the same order
, we adopted a procedure which consists in considering _ all _ neighbours in a domain @xmath46 and derive stellar parameters as the ( simple ) mean of each set of parameters \{@xmath0 ; log@xmath1 ; [ m / h ] ; @xmath2 } characterising this set of nearest neighbours ( a. lpez ariste , private communication ) .
we did not notice significant changes in the results either for a smaller range of pca - distances or when adopting e.g. , distance - weighted mean parameters .
before processing any real observed spectra with our pca - based inversion tool , we need to investigate on the potential effect of noise .
to do so , we have ingested in our inversion tool multiple realisations of the content of the very learning database @xmath47 affected by controlled ( gaussian ) white noise - levels , and we estimated how it affects the inversion process by systematically comparing ( known ) input and inverted parameters .
.standard deviation of the absolute differences between input and inverted parameters for noisy spectra vs. a ( gaussian ) white - noise level characterized by a @xmath48 per pixel standard deviation .
[ cols="^,^,^,^,^",options="header " , ]
because one of the originality of our inversion tool is in providing the projected rotational velocity of stars , we also tested its capability for a couple of fgk stars with significant ( @xmath49 ) @xmath2 and belonging the so - called `` stellar properties of observed cool stars '' ( aka .
spocs ) catalogue ( valenti & fischer 2005 ) . for a first test
, we used a spectra from the pre - main sequence star hd377 and comparison between spocs parameters and ours are detailed in table [ table:2 ] . unfortunately , according to the pastel catalogue ( soubiran et al .
2010 ) there are almost no alternative values given in the litterature for this object .
its @xmath2 is well recovered and other parameters agree reasonably well with spocs values .
our second target is @xmath50 boo ( a , also hd120136 ) , a f6iv star for which the spocs catalogue provides a @xmath2 of 15 km@xmath10 while our estimate is about 19 km@xmath10 . however
a @xmath2 of about 18 km@xmath10 has been recently reported by martnez - arniz et al .
other parameters agree very well , except for log@xmath1 that we systematically overestimated for these two objects . a first excursion away from the fgk - dwarfs domain consists in exploring the luminosity class towards the giants domain . the red giant branch ( g5 ) star hd12328 have been used for that purpose .
our results are again reliable given also that recent works mention an effective temperature of 4808 k , in better agreement with ours ( massarotti et al .
2008 ) , as well as log@xmath1 value of the order of 3.3 dex and [ fe / h]=-0.04 dex ( see also jones et al . 2011 ) . outside the spocs catalogue , and again for ( sub-)giants , we also used data from the k1.5iii star arcturus .
our 4250 k determination of its effective temperature is very close , within 30 k , to that of prugniel et al .
( 2011 ) for instance .
we could also derive a surface gravity about 2.0 in agreement with alternative estimates ( e.g. , massarotti et al .
2008 ) and our [ m / h ] and @xmath2 values are also in pretty well identified ranges , as reported in table 2 .
note that further tests with pollux ( k0iii ) data also gave excellent results .
concerning these classes of stars , we are also perfectly aware that additional synthetic spectra with microturbulent velocities not restricted to 2 km@xmath10 as we did here , will have to be next considered .
note also that for ( sub-)giant stars , models computed in ( 1d ) _ spherical _ geometry such as marcs ones ( gustafsson et al . 2008 ) may also provide a more realistic description of the observed stellar fluxes and therefore improve the determination of the inverted stellar parameters .
our database of spectropolarimetric data already covers pretty well the h - r diagram .
we are therefore interested in spanning most of it , and basically with the same inversion tool . selected targets for these further tests
are indicated , together with the parameters we derived from their spectra , in table 2 . for cooler stars , we picked spectra of gl 205 , a m1.5v dwarf .
our values of @xmath0 ,
log@xmath1 and [ m / h ] agree very well with those of prugniel et al .
( 2011 ) , and our @xmath2 determination agrees well with the one of houdebine ( 2010 ) . there is also a reasonable agreement with values recently derived by neves et al .
( 2013 ) .
finally , we also tested our method with spectra of the a1v star sirius ( a ) .
our inversion of the effective temperature gives a value a bit larger than the 9870 k more recently reported ( hill & landstreet 1993 ) .
[ m / h ] is also overestimated , but log@xmath1 and @xmath2 are quite well determined .
we are also aware that , so far , we restricted ourselves to models such that [ @xmath9/fe]=0 which may not apply to such a star . possible enhancement of @xmath9-elements will be considered in a next version of our inversion tool , when performing spectral analysis at a larger scale than the one presented here .
however , besides this current restriction , the agreement between the observed spectrum and the nearest neighbour synthetic spectrum is already quite satisfactory , as shown in fig.([fig4 ] ) .
concerning the determination of @xmath2 , other methods of evaluation exist .
however , to the best of our knowledge , they always require a template spectrum or at least a list of spectral line _ a priori _ expected in the spectra , as auxilliary and `` support '' data ( see e.g. , daz et al . 2012 and references therein ) .
our final pipeline will therefore implement an additional module providing , once a preliminary stellar parameter inversion will be available from our inversion tool , an alternative and complementary @xmath2 evaluation .
note also that , with our pca method , we are mostly interested in the `` intermediate '' @xmath2 regimes , say between 10 and 100 km@xmath10 .
indeed , for slow rotators it is obvious that rotational broadening becomes of the order of other sources of broadening ( e.g. , instrumental or turbulent ) so that a more detailed line profile analysis may be required . on the other hand ,
( very ) fast rotators are affected by macroscopic effects such as gravity darkening ( see e.g. , espinosa lara & rieutord 2011 ) which are not accounted for , so far , by standard atmospheric ( radiative ) modelling .
another source of potential improvement relies on the content of our learning database .
atlas standard models may not be the best choice for cool stars or for metal - poor stars for instance .
it is also well - known that non - lte effects may take place in the formation of the infrared triplet of caii , which affects the spectral lines central depressions .
this issue was for instance mentioned in kordopatis et al .
more detailed elements , from the point of view of radiative modelling , are also discusses in merle et al . ( 2011 ) .
finally , we shall need synthetic spectra computed for a resolvance comparable to the one of our observations and , ideally , including the effects of rotational broadening .
the database of synthetic spectra , pollux ( palacios et al .
2010 ) already provides a partial answer to our needs , and we believe that its development will take into account such needs .
we have experimented a fast and reliable pca - based numerical method for the inversion of stellar fundamental parameters @xmath0 , log@xmath1 and [ m / h ] , as well as the projected rotational velocity @xmath2 , from high - spectral resolution echelle spectra taken from the narval spectropolarimeter in operations at the 2-m tbl telescope .
our method is fast and easy to implement .
first tests , mainly made for fgk - stars spectra show fairly good agreement with reference values published by valenti & fischer ( 2005 ) .
preliminary tests also suggest that the same method is applicable to fgk ( sub-)giants , as well as to cooler m stars or hotter stars up to spectral type a. we used it , so far , at the vicinity of the infrared triplet of caii and without any help from additional ( e.g. , photometric ) information , which is particularly challenging .
however we can easily , either extend the spectral domain used by our inversion method , or combine analyses from several distinct spectral domains , in order to constrain further and refine our stellar parameter determination .
it is finally important to realize from the present study that pca allows for a reduction of dimensionality of the order of 100 ( @xmath51 ) which is of great potential interest for the post - processing of high - resolution spectra covering a very large bandwidth like the ones from both espadons and narval spectropolarimeters .
we are grateful to dr .
tomaz zwitter ( university of ljubljana , slovenia ) who made available the synthetic spectra used in the work .
this research has made use of the vizier catalogue access tool , cds , strasbourg , france .
the original description of the vizier service was published in a&as 143 , 23 .
this research has made use of the simbad database , operated at cds , strasbourg , france .
narval data were provided by the ov - gso datacenter operated by cnrs / insu and the _ universit paul sabatier , toulouse_-omp ( tarbes , france ; polarbase.irap.omp.eu ) . | the general context of this study is the inversion of stellar fundamental parameters from high - resolution echelle spectra .
we aim indeed at developing a fast and reliable tool for the post - processing of spectra produced by espadons and narval spectropolarimeters .
our inversion tool relies on principal component analysis .
it allows reduction of dimensionality and the definition of a specific metric for the search of nearest neighbours between an observed spectrum and a set of synthetic spectra .
effective temperature , surface gravity , total metallicity and the projected rotational velocity are derived .
our first tests , done from the sole information coming from a spectral band very similar to the one the rvs spectrometer will observe from the gaia space observatory , and with spectra mainly taken from fgk stars are very promising .
we also tested our method with a few targets beyond this domain of the h
r diagram .
= 1 [ firstpage ] methods : data analysis stars : fundamental parameters astronomical databases : miscellaneous |
the `` problem of the @xmath6 spheres , '' so named by schtte and van der waerden @xcite and leech @xcite , asks whether there exists any configuration of @xmath6 non - overlapping unit spheres that all touch a central unit sphere .
it was raised in the time of newton by david gregory , and eventually resolved mathematically as impossible .
its resolution established that the `` kissing number '' of equal spheres in @xmath7-dimensional euclidean space is @xmath0 .
this paper is concerned with a related but different problem : _ how can @xmath0 spheres of equal radius @xmath2 touch a given central sphere of radius @xmath8 , in what patterns , and how are these patterns related ?
in other words , what is the topology of the corresponding configuration space of such spheres ? _ in this paper we review the remarkable history of this problem in several contexts , survey aspects of what is currently known about it , and present some new results and conjectures .
this problem has come up in physics and materials science .
many atoms and molecules are roughly spherical , and their local interactions are governed by how many of them can get close to a single atom .
the arrangements possible for @xmath6 nearby spheres , and allowable motions between them , are relevant to the nature of local interactions , to measuring the entropy of local configurations , and to phase changes in certain materials .
we are particularly motivated by a statement of frank ( 1952 ) made in the context of supercooling of fluids , given in section [ sec:27 ] .
insisting that exactly @xmath0 equal spheres touch a @xmath6-th central sphere , possibly of a different radius , gives a mathematical toy problem that can be subjected to careful analysis . as a mathematical problem ,
the @xmath0 spheres problem has both a metric geometry aspect and a topology aspect .
lszl fejes tth made major contributions to the metric geometry of the problem , which concerns extremal questions , formulated as densest packing problems . in connection with the tammes problem , described in section [ sec:3 ] , he found the largest radius of @xmath0 spheres that can touch a central sphere of radius @xmath8 , realized by the dodecahedral configuration @xmath9 , and found other extremal configurations of touching spheres for smaller @xmath3 .
he posed the dodecahedral conjecture concerning the minimal volume voronoi cell in a unit sphere packing , and posed another conjecture characterizing all configurations that pack space with every sphere having exactly @xmath0 neighboring spheres .
both of these conjectures are now proved .
the topological side of the @xmath0 spheres problem concerns allowable motions and rearrangements of configurations of spheres , and topological constraints on them .
a major part of this paper addresses the topological side of the problem , concerning the topology of configuration spaces , and the change of topology as the radius @xmath2 is varied .
the arrangements of the @xmath0 touching spheres are encoded in the associated _ configuration space _ of @xmath0-tuples of points on the surface of a unit sphere that remain at a suitable distance from each other .
this space has nontrivial topology and geometry . in topology
the general subject of configuration spaces started in the 1960s with the consideration of topological spaces whose points denote configurations of a fixed number @xmath3 of labeled points on a manifold .
this paper considers the _ constrained _ configuration space @xmath10 $ ] of @xmath3 non - overlapping spheres of radius @xmath2 which touch a central sphere @xmath11 of radius @xmath8 , centered at the origin .
( here `` non - overlapping '' means the spheres have disjoint interiors . )
it can also be visualized as the space of @xmath3 spherical caps on the sphere , which are obtained as the radial projection of the external spheres onto the surface of the central sphere , whose _ angular diameter _
@xmath12 is a known function of @xmath2 .
the centers of these caps define a constrained @xmath3-configuration on @xmath11 where no pair of points can approach closer than angular separation @xmath13 . for generic ( `` non - critical '' ) values of @xmath2 for a range of values
@xmath14 , this space is a compact @xmath15-dimensional manifold with boundary , not necessarily connected .
the group @xmath16 acts as global symmetries of @xmath10 $ ] by rigidly rotating the @xmath3-configuration of spheres touching the central sphere .
the _ reduced constrained configuration space _ @xmath17 = { \operatorname{conf}}(n)[r]/so(3)$ ] is obtained by identifying rotationally equivalent configurations . for generic values of @xmath2
it is a compact @xmath18-dimensional manifold with boundary ; for the case of @xmath0 spheres this is a @xmath19-dimensional manifold .
the subject of constrained configuration spaces has in part been developed for applications to fields such as robotics .
for an introduction to the robotics aspect , see generally abrams and ghrist @xcite or farber @xcite .
this paper surveys results for small @xmath3 on the metric geometry problem of determining the maximum allowable radius @xmath5 for @xmath10 $ ] ( equivalently @xmath20 $ ] ) to be nonempty ; this is a variant of the tammes problem , also treated in the literature under the name _ optimal spherical codes _ ( see section [ sec:3 ] ) .
this paper also studies the topology of configuration spaces of a fixed radius @xmath2 , and the changes in topology in such spaces as the radius @xmath2 is varied . in the latter case the configuration space changes topology at a set of _ critical radius values_. associated to these special values
are _ critical configurations _ , which are extremal in a suitable sense .
the change in topology is described by a generalization of morse theory applicable to the radius function @xmath2 , which we discuss in section [ sec:4 ] . to determine these changes
one studies the occurrence and structure of the critical configurations .
the simplest example of such topology change concerns the connectivity of the space of configurations as a function of @xmath2 , reported by the rank of the @xmath21-th homology group of the configuration space .
the @xmath0 spheres problem includes as its most important special case that of unit spheres , where the sphere radius @xmath22 .
this special case is the one relevant to sphere packing in dimension @xmath7 .
we treat the topological space @xmath23 $ ] in sections [ sec:5 ] and [ sec:6 ] , and formulate several conjectures related to it .
the radius @xmath22 is a critical radius , and two configurations @xmath24 and @xmath25 on the boundary of the space @xmath23 $ ] are critical configurations .
the topology of @xmath23 $ ] appears to be very complicated , and its cohomology groups have not been determined . in section [ sec:6 ]
we describe how it is possible to move in the space @xmath23 $ ] to deform any dodecahedral configuration @xmath9 of @xmath0 labeled spheres to any other labeled @xmath9 configuration , permuting the @xmath0 spheres arbitrarily , a result due to conway and sloane .
this suggests the ( folklore ) conjecture asserting that @xmath22 is the largest radius value for which the configuration space @xmath26 $ ] is connected , i.e. it is the largest @xmath2 for which the @xmath21-th cohomology group of @xmath26 $ ] has rank @xmath8 .
this paper establishes some new results .
it makes the observation ( in section [ sec:43a ] ) that the family of @xmath27-configurations of spheres achieving @xmath28 ( see figure [ fig : example5 ] ) is topologically complex .
it completely determines ( in section [ sec:47 ] ) the cohomology of @xmath29 $ ] for allowable @xmath2 .
it makes precise the notion of @xmath3-configurations being _
critical for maximizing _ the injectivity radius on @xmath30 , and provides a necessary and sufficient _ balancing condition _
( theorem [ thm : converse ] ) for criticality , prefatory to a `` morse theory '' for such min - type functions @xcite . and
it formulates several new conjectures in sections [ sec:65 ] and [ sec66 ] .
configuration spaces are of interest in physics and materials science .
jammed configurations are a granular materials criterion for a stable packing . according to torquato and stillinger
@xcite they are : `` particle configurations in which each particle is in contact with its nearest neighbors in such a way that mechanical stability of a specific type is conferred to the packing . '' packings of rigid disks and spheres have been studied extensively by simulation ( lubachevsky and stillinger @xcite , donev et al . @xcite ) .
it has been empirically discovered that randomly ordered hard spheres achieve in random close packing a density around @xmath31 percent @xcite , and pass through a jamming transition around @xmath32 percent @xcite .
the appearance of a jamming phase transition , signaled by a change in shear modulus , and the formation of a glass state , is relevant in studying the behavior of colloidal suspensions and granular materials .
the large rearrangement of structure required in making a phase transition is relevant in the phenomenon of supercooling of liquids ( see section [ sec:27 ] ) .
the nature of glass transitions has been called `` the deepest and most interesting unsolved problem in solid state theory '' ( anderson @xcite ) . for articles and reviews of these topics , see generally ediger et al .
@xcite , ohern et al .
@xcite , and liu and nagel @xcite . for a survey of hard sphere models , including the idea of a liquid - solid phase transition in packings ,
see generally lwen @xcite .
one may make an analogy between the configuration spaces @xmath17 $ ] treated here and a sphere packing model for jamming studied in @xcite , which treats spheres having repulsive local potential at zero density and zero applied stress , and includes hard spheres for one model parameter value . in the latter model ,
the order parameter is the packing fraction of the spheres . in the configuration space model ,
a proxy value for the packing fraction is the radius parameter @xmath2 , which determines the fraction of surface area of @xmath11 covered by the @xmath3 spherical caps .
an analogue of the jamming transition value in the configuration space model is then the maximal radius @xmath33 at which the constrained configuration space @xmath17 $ ] remains connected ; this property is detected by the @xmath21-th cohomology group .
finer topological invariants of this kind are then supplied by the various critical values @xmath34 at which the ranks of the individual cohomology groups @xmath35 , { { \mathbb q}})$ ] change .
our configuration model is simplified in being @xmath36-dimensional , with constrained configurations on the surface of a @xmath36-sphere @xmath11 , which , however , has the new feature of positive curvature , giving a compact constrained configuration space . for the jamming problem itself ,
the space of ( constrained ) configurations of hard spheres in a large @xmath7-dimensional box seems a more appropriate space .
the general direction of inquiry investigating the transition of topological invariants ( like betti numbers ) of configuration spaces as the radius parameter is varying could shed new light on the nature of jamming transitions .
for further remarks , see section [ sec:7 ] .
this paper is mainly of a survey nature , and the sections of the paper have been written to be independently readable .
section [ sec:2 ] gives a brief history of results on the @xmath0 spheres problem and sphere packing .
section [ sec:3 ] surveys results on the maximal radius @xmath37 for configurations of @xmath3 equal spheres touching a central sphere of radius @xmath8 for small @xmath3 .
this problem is equivalent to the tammes problem .
section [ sec:4 ] begins with the topology of configuration spaces of @xmath3 points in @xmath38 and on the @xmath36-sphere @xmath11 , corresponding to radius @xmath39 .
it then considers spaces of configurations of equal spheres of radius @xmath2 touching a sphere of radius @xmath8 for variable @xmath40 .
it defines a notion of critical configuration in the spirit of min - type morse theory .
section [ sec:5 ] discusses the special configuration space of @xmath0 unit spheres touching a @xmath6-th central sphere , i.e. the case of radius @xmath22 .
it focuses on properties of the @xmath24 configuration , the @xmath25 configuration and the dodecahedral configuration @xmath9 .
it shows that the @xmath24 and @xmath25 configurations are critical ( in the sense of section [ sec:42 ] ) in the reduced configuration spaces @xmath23 $ ] . and
it shows that there are continuous deformations between a dodecahedral configuration to an @xmath24 configuration and to an @xmath25 configuration in the reduced configuration space @xmath23 $ ] .
section [ sec:6 ] considers the problem of permutability of the spheres of the dodecahedral configuration for @xmath22 , conjecturing that @xmath23 $ ] is connected , and that this is the largest value of @xmath2 where connectedness holds .
it also considers the @xmath41 case and formulates several conjectures about disconnectedness .
section [ sec:7 ] makes some concluding remarks .
we begin with some historical vignettes concerning configurations of @xmath0 spheres touching a central sphere , as they have come up in physics , astronomy , biology and materials science .
johannes kepler ( 15711630 ) studied packings and crystals in his 1611 pamphlet `` the six - cornered snowflake '' @xcite . in it
he asserts that the densest sphere packing of equal spheres is the @xmath24 packing , or `` cannonball packing . ''
he states that this packing has @xmath0 unit spheres touching each central sphere : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ in the second mode , not only is every pellet touched by its four neighbors in the same plane , but also by four in the plane above and four below , so throughout one will be touched by twelve , and under pressure spherical pellets will become rhomboid .
this arrangement will be more compatible to the octahedron and the pyramid .
the packing will be the tightest possible , so that in no other arrangement could more pellets be stuffed into the same container . ] _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ he expands on the construction as follows : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ thus , let @xmath42 be a group of three balls ; set one @xmath43 , on it as apex ; let there be also another group @xmath44 , of six balls , and another @xmath45 , of ten , and another @xmath46 , of fifteen .
regularly superpose the narrower on the wider to produce the shape of a pyramid .
now , although in this construction each one in the upper layer is seated between three in the lower , yet if you turn the figure round so that not the apex but the whole side of the pyramid is uppermost , you will find , whenever you peel off one ball from the top , four lying below it in square pattern . again as before ,
one ball will be touched by twelve others , to with , by six neighbors in the same plane , and by three above and three below .
thus in the closest pack in three dimensions , the triangular pattern can not exist without the square , and vice versa .
it is therefore obvious that the loculi of the pomegranate are squeezed into the shape of a solid rhomboid .... copula trium globorum .
ei superpone @xmath43 unum pro apice ; esto et alia copula senum globorum @xmath44 , et alia denum @xmath45 et alia quindenum @xmath47 impone semper angustiorem latiori , ut fiat figura pyramidis .
etsi igitur per hanc impositionem singuli superiores sederunt into trinos inferiores : tamen iam versa figura , ut non apex sed integrum latus pyramidis sit loc superiori , quoties unum globulum deglberis e summis , infra stabunt quattuor ordine quadrato .
et rursum tangetur unus globus ut prius , et duodecim aliis , a sex nempe circumstantibus in eodem plano tribus supra et tribus infra .
ita in solida coaptatione arctissima non potest ess ordo triangularis sine quadrangulari , nec vicissim .
patet igitur , acinos punici mali , materiali necessitate concurrente cum rationaibus incrementi acinorum , exprimi in figuri rhombici corporis ... '' [ translation by colin hardie @xcite ] ] _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the cannonball packing had been studied earlier by the english mathematician thomas hariot [ harriot ] ( 15601621 ) .
hariot was mathematics tutor to sir walter raleigh , designed some of his ships , wrote a treatise on navigation , and went on an expedition to virginia in 15851587 as surveyor , reporting on it in 1590 in @xcite , his only published book .
he computed a chart in 1591 on how to most efficiently stack cannonballs using the @xmath24 packing , and computed a table of the number of cannonballs in such stacks ( see shirley @xcite ) .
hariot supported the atomic theory of matter , in which case macroscopic objects may be packed in arrangements of very tiny spherical objects , i.e. atoms ( * ? ? ? * chap .
he corresponded with kepler in 16061608 on optics , and mentioned the atomic theory in a december 1606 letter as a possible way of explaining why some light is reflected , and some refracted , at the surface of a liquid .
kepler replied in 1607 , not supporting the atomic theory .
the known correspondence of hariot with kepler does not deal directly with sphere packing .
the statement that the maximal density of a sphere packing in @xmath7-dimensional space equals @xmath48 , which is attained by the @xmath24 packing , is called the _ kepler conjecture . _
it was settled affirmatively in the period 19982004 by hales with ferguson @xcite .
a second generation proof , which is a formal proof checked entirely by computer , was recently completed in a project led by hales @xcite .
the discussion of isaac newton and david gregory in 1694 was related to preparing a second edition of newton s _ principia_. it concerned the question whether the `` fixed stars '' are subject to gravitational attraction .
what force is `` balancing '' their apparent fixed positions ?
gregory ( * ? ? ?
* vol iii , p. 317 ) summarized in a memorandum a conversation with newton on 4 may 1694 concerning the brightest stars as : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ to discover how many stars there are of a given magnitude , he [ newton ] considers how many spheres , nearest , second from them , third etc .
surround a sphere in a space of three dimensions , there will be @xmath6 of first magnitude , @xmath49 of second , @xmath50 of third .
@xmath36-dae , @xmath51 3 ae . '' ] _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ newton s own star table `` a table of ye fixed starrs for ye yeare 1671 '' records @xmath6 first magnitude stars , @xmath52 of the second magnitude , @xmath53 of third magnitude ( see ( * ? ? ?
* vol ii , p. 394 ) ) .
newton drafted a new proposition to be included in a second edition of the _ principia _ , stating [ in translation ] @xcite : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ proposition xv .
theorem xv .
the fixed stars are at rest in the heavens and are separated by enormous distances from our sun and from each other . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ in a draft proof he wrote [ in translation]@xcite _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ that the stars are at huge distances from our sun is clear enough from the absence of parallax ; and that they lie at no less distances from each other may be inferred from their differing apparent magnitudes . for there are @xmath6 stars of the first magnitude and roughly the same number of equal spheres can be arranged about a central sphere equal to them .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ and : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ for if around some sphere there are arranged more spheres of about the same size , the number of spheres which surround it closely will be @xmath0 or @xmath6 ; at the second stage about @xmath54 ; at the third about @xmath55 [ roughly @xmath56 ; at the fourth , @xmath57 [ @xmath58 , ... _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ this argument is similar to one of kepler ( * ? ? ?
* liber i , pars ii , p. 138 ) ( translation in koyr @xcite ) , with roots in the claim of giordano bruno , that all stars are suns . after further work , over several drafts ,
newton abandoned the proposition ( hoskin @xcite ) .
it was not included in the second edition of the _ principia _ when it finally came out in 1713 .
gregory continued with the geometric problem underlying the spacing of stars . in an ( unpublished ) notebook he considered the packing problem of @xmath36-dimensional disks in concentric rings and , in @xmath7 dimensions , that of equal spheres , noting that @xmath6 spheres might touch a given equal sphere ( * ? ? ? *
vol iii , letter 441 , note ( 10 ) , p. 321 ) .
he continued to consider the @xmath6 sphere question in later years , making the following memorandum in 1704 @xcite : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ oxon .
23 nov@xmath59 1704 .
kyl said that if @xmath6 equal spheres touch an equal inmost sphere , @xmath60 must touch one that include these former @xmath61 , because there is nine times as much surface to stand on .
i told him that we must reckon by the surface passing through their centers .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ we may infer that newton left the question of how many spheres might touch unresolved , and that gregory believed @xmath6 spheres might touch .
the issue of whether @xmath6 equal spheres might touch a central equal sphere was discussed in the physics literature in the period 18741875 , with contributions by c. bender @xcite , reinhold hoppe @xcite and siegmund gnther @xcite .
hoppe noted a mathematical gap in the argument of bender .
gnther offered a physical intuition , but no proof .
they all concluded that at most @xmath0 unit spheres could touch a central unit sphere . in 1994
hales @xcite noted a mathematical gap in the argument of hoppe . in another context
the crystallographer william barlow ( 18451934 ) noted another optimal sphere packing , the _ hexagonal close packing _ ( @xmath25 ) . in a paper `` probable nature of the internal symmetry of crystals '' @xcite he considered five symmetry types for crystal structure .
the third kind of symmetry he describes is the @xmath24 packing ( fig . 4 and 4a ) .
he then stated : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ a fourth kind of symmetry , which resembles the third in that each point is equidistant from the twelve nearest points , but which is of a widely different character than the three former kinds , is depicted if layers of spheres in contact arranged in the triangular pattern ( plan d ) are so placed that the sphere centers of the third layer are over those of the first , those of the fourth layer over those of the second , and so on .
the symmetry produced is hexagonal in structure and uniaxial ( figs . 5 and 5a ) . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ here `` plan d '' is the two - dimensional hexagonal packing , and figs .
5 and 5a depict the @xmath25 packing .
he suggested that the atoms in a crystal of quartz ( @xmath62 ) occur with the fourth kind of symmetry ( see figure [ fig:0 - 2 ] ) .
barlow also stated later in the paper ( * ? ? ?
* figs . 7 and 8 , p.207 )
the following about twinned crystal arrays with a connecting layer : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the peculiarities of _ crystal - grouping _ displayed in twin crystals can be shown to favour the supposition that we have in crystals symmetrical arrangement rather than symmetrical shape of atoms or small particles . thus if an octahedron be cut in half by a plane parallel to two opposite faces , and the hexagonal faces of separation , while kept in contact and their centres coincident ,
are turned one upon the other through @xmath63 , we know that we get a familiar example of a form found in some twin crystals . and a stack can be made of layers of spheres placed triangularly in contact to depict this form as readily as to depict a regular octahedron , the only modification necessary being for the layers above the centre layer to be placed as though turned bodily through @xmath63 , from the position necessary to depict an octahedron ( compare figs . 7 and 8)
the modification , as we see , involves _ no departure from the condition that each particle is equidistant from the twelve nearest particles .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ' '' '' and @xmath25 packings , title="fig : " ] and @xmath25 packings , title="fig : " ] + and @xmath25 packings , title="fig : " ] and @xmath25 packings , title="fig : " ] and @xmath25 packings , title="fig : " ] the dutch botanist pieter merkus lambertus tammes made in 1930 a study of the equidistribution of pores on pollen grains @xcite .
he asked the question : what is the maximum number of circular caps @xmath64 of angular diameter @xmath13 that can be placed without overlap on a unit sphere ? here
@xmath13 is measured from the center of the unit sphere @xmath11 in @xmath65 .
tammes ( * ? ? ?
3 ) empirically determined that @xmath66 , while @xmath67 for @xmath68 .
let @xmath69 denote the maximal value of @xmath13 having @xmath70 .
he concluded that @xmath71 .
the problem of determining various values of @xmath64 is now called the _ tammes problem_. it is related to a dual question of determining the maximal radius @xmath72 possible for @xmath3 equal spheres all touching a central sphere of radius @xmath8 .
namely , the maximal value of @xmath73 having @xmath74 , determines the maximal allowable radius @xmath72 of @xmath3 spheres touching a central unit sphere by a formula given in lemma [ lemma:31 ] below . in 1943
lszl fejes tth @xcite conjectured that the volume of any voronoi cell of any sphere packing of @xmath75 by unit spheres is minimized by the dodecahedral configuration of @xmath0 unit spheres touching a central sphere .
the voronoi cell of the central sphere is then a regular dodecahedron circumscribed about the sphere .
the packing density of the dodecahedron is approximately @xmath76 , which is larger than the density of the known fcc packing of @xmath75 .
this conjecture became known as the _ dodecahedral conjecture _ and was settled affirmatively in 2010 ( see section [ sec:53 ] ) .
the problem of molecular rearrangement in the liquid - solid phase transition is relevant in materials .
the structure of ordinary ice , the @xmath77 phase labeled ice @xmath78 , has an @xmath25 packing of its oxygen atoms , as observed in 1921 by dennison @xcite .
note that the hydrogen atoms are free to change their orientations to some extent ( pauling @xcite ) .
water exhibits a phenomenon of supercooling at standard pressure down to @xmath79 ; under special rapid cooling it can avoid freezing down to @xmath80 , and enter a glassy phase ( angell @xcite ) . in 1952
frederick charles frank @xcite argued that supercooling can occur because the common arrangements of molecules in liquids assume configurations far from what they would assume if frozen .
he wrote : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ consider the question of how many different ways one can put twelve billiard balls in simultaneous contact with another one , counting as different the arrangements which can not be transformed into each other without breaking contact with the centre ball? the answer is _ three_. two which come to the mind of any crystallographer occur in the face - centred cubic and hexagonal close packed lattices .
the third comes to the mind of any good schoolboy , and it is to put one at the centre of each face of a regular dodecahedron .
that body has five - fold axes , which are abhorrent to crystal symmetry : unlike the other two packings , this one can not be continuously extended in three dimensions .
you will find that the outer twelve in this packing do not touch each other .
if we have mutually interacting deformable spheres , like atoms , they will be a little closer to the centre in this third kind of packing ; and if one assumes they are argon atoms ( interacting in pairs with attractive and repulsive potentials proportional to @xmath81 and @xmath82 ) one may calculate that the binding energy of the group of thirteen is @xmath83 greater than for the other two packings .
this is @xmath84 of the lattice energy per atom in the crystal .
i infer that this will be a very common grouping in liquids , that most of the groups of twelve atoms around one will be of this form , that freezing involves a substantial rearrangement , and not merely an extension of the same kind of order from short distances to long ones ; a rearrangement which is quite costly of energy in small localities , and which only becomes economical when extended over a considerable volume , because unlike the other packing it can be so extended without discontinuities . _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the three local arrangements frank specifies we shall label as @xmath24 ( face - centered - cubic ) , @xmath25 ( hexagonal close packing ) and @xmath9 ( dodecahedral ) , for convenience .
the crystalline arrangements of @xmath24 and @xmath25 are `` extremal '' ( i.e. on the boundary of the configuration space ) , while the balls in @xmath9 configuration are free to move independently .
frank s assertion that there are exactly three possible arrangements is _ false _ if taken literally .
there are continuous deformations between any arrangement of types @xmath24 , @xmath25 and @xmath9 and any of the other types ( see section [ sec:53 ] ) .
there is however an important kernel of truth in frank s statement , which buttresses his argument made concerning the existence of supercooling : each of the three arrangements above is `` remarkable '' in some sense ( see section [ sec:5 ] ) . to move from a large arrangement of spheres having many @xmath9 configurations to one
frozen in the @xmath25 packing requires substantial motion of the spheres . in a paper titled `` das problem der
dreizehn kugeln '' [ `` the problem of the thirteen balls '' ] kurt schtte and bartel leendert van der waerden @xcite gave a rigorous proof that one can not have @xmath6 unit spheres touching a given central sphere .
there has been much further work on this problem . in his 1956 paper
titled `` the problem of @xmath6 spheres '' john leech @xcite gave a two page proof of the impossibility of @xmath6 unit spheres touching a unit sphere .
more accurately he stated : `` in the present paper i outline an independent proof of this impossibility , certain details which are tedious rather than difficult have been omitted . ''
various authors have written to fill in such details , which balloon the length of the proof .
these include work of maehara @xcite in 2001 , who gave in 2007 a simplified proof @xcite .
other proofs of the thirteen spheres problem were given by anstreicher @xcite in 2004 and musin @xcite in 2006 . in 1969 lszl fejes tth
@xcite discussed the problem of characterizing those sphere packings in space that have the property that every sphere in the packing touches exactly @xmath0 neighboring spheres .
the @xmath24 and @xmath25 packing both have this property , as already noted by barlow ( 1883 ) .
there are in addition uncountably many other packings , obtained by stacking plane layers of hexagonally packed spheres ( `` penny packing '' ) , where there are two choices at each level of how to pack the next level .
fejes tth conjectured that all such packings are obtained in this way .
this conjecture of fejes tth s was settled affirmatively by hales @xcite in 2013 . in their book : _ sphere packings , lattices and groups , _
john h. conway and neil j. a. sloane considered the question : _ what rearrangements of the @xmath0 unit spheres are possible using motions that maintain contact with the central unit sphere at all times ?
_ in ( * ? ?
* chap . 1 ,
appendix : planetary perturbations ) they sketch a result asserting : the configuration space of @xmath0 unit spheres touching a @xmath6-th allows arbitrary permutations of all @xmath0 touching spheres in the configuration .
that is , if the spheres are labeled and in the dod configuration , it is possible , by moving them on the surface of the central sphere , to arbitrarily permute the spheres in a dod configuration .
we will describe the motions in detail to obtain such permutations in section [ sec:6 ] .
what is the maximal radius @xmath72 possible for @xmath3 equal spheres all touching a central sphere of radius @xmath8 ?
this problem is closely related to the _ tammes problem _ discussed above , which concerns instead the maximum number of circular caps @xmath64 of angular diameter @xmath13 that can be placed without overlap on a sphere .
the latter problem is also the problem of constructing good spherical codes ( see ( * ? ? ?
* chap . 1 , sec . 2.3 ) ) .
one can convert the angular measure @xmath13 into the radius of touching spheres ; for a sphere touching a central unit sphere , its associated spherical cap on the central sphere is the radial projection of its points onto the surface of the central sphere .
related to radius @xmath2 ] [ lemma:31 ] for a fixed @xmath85 , the maximal value of @xmath86 having @xmath74 determines the maximal allowable radius @xmath5 of @xmath3 spheres touching a central unit sphere , using the formula @xmath87 conversely , given @xmath88 , we obtain @xmath89 choosing @xmath90 . from the right triangle in figure
[ fig:7 - 2 ] we have @xmath91 this relation gives a bijection of the interval @xmath92 to the interval @xmath93 the tammes problem has been solved exactly for only a few values of @xmath3 , including @xmath94 and @xmath95 .
the tammes problem was solved for @xmath97 and @xmath0 by lszl fejes tth @xcite in 1943 , where extremal configurations of touching points for @xmath98 are attained by vertices of an equilateral triangle arranged around the equator , and for @xmath99 by vertices of regular polyhedra ( tetrahedron , octahedron and icosahedron ) inscribed in the unit sphere .
fejes tth proved the following inequality : for @xmath3 points on the surface of the unit sphere , at least two points can always be found with spherical distance @xmath100 note that @xmath101 is the edge - length of a spherical equilateral triangle with the expected area for an element of an @xmath3-vertex triangulation of @xmath11 .
the inequality is sharp for @xmath102 and @xmath0 for the specified configurations above . in 1949
fejes tth @xcite gave another proof of his inequality .
his result was re - proved by habicht and van der waerden @xcite in 1951 . after converting this result to the @xmath2-parameter using lemma [ lemma:31 ] , we may re - state his result for @xmath103 as follows .
[ th72 ] ( fejes tth ( 1943 ) ) @xmath104 the maximum radius of @xmath0 equal spheres touching a central sphere of radius @xmath8 is : @xmath105 here @xmath106 is a real root of the fourth degree equation @xmath107 .
@xmath108 an extremal configuration achieving this radius is the @xmath0 vertices of an inscribed regular icosahedron ( equivalently , face - centers of a circumscribed regular dodecahedron ) .
the tammes problem was solved for @xmath110 in 1950 by van der waerden , building on work of habicht and van der waerden @xcite .
it was solved for @xmath111 by schtte .
these solutions , plus those of van der waerden for @xmath112 and schtte for @xmath113 appear in schtte and van der waerden @xcite .
they give a history of these developments in @xcite .
their paper used geometric methods , introducing and studying the allowed structure of the graphs describing the touching patterns of arrangements of @xmath3 equal circles on @xmath11 .
these graphs are now called _ contact graphs _ , and schtte and van der waerden credit their introduction to habicht .
schtte also conjectured candidates for optimal configurations for @xmath114 and van der waerden conjectured candidates for @xmath115 ( see @xcite ) .
l. fejes tth presented the work of schtte and van der waerden in his 1953 book on sphere - packing ( * ? ? ?
* chapter vi ) .
this book uses the terminology of _ maximal graph _ for the graph of a configuration achieving the maximal radius for @xmath3 . in 1959
fejes tth @xcite noted that the set of vertices of a square antiprism gave an extremal @xmath112 configuration on the @xmath36-sphere . in his 1963 habilitationsschrift @xcite ( see the 1986 english translation @xcite ) , ludwig danzer made a geometric study of the contact graph for a configuration of @xmath3 circles on the surface of a sphere .
this graph has a vertex for each circle and an edge for each pair of touching circles .
a contact graph is called _ maximal _ if it occurs for a set of circles achieving the maximal radius @xmath5 .
it is called _ optimal _ if it has the minimum number of edges among all maximal contact graphs .
a contact graph is called _ irreducible _ if the radius can not be improved by altering a single vertex . for each small @xmath3
, danzer found a complete list of irreducible contact graphs .
he used this analysis to prove the conjectures of schtte and van der waerden @xcite above for the cases @xmath117 .
[ th73 ] ( danzer ( 1963 ) ) @xmath104 for @xmath118 there is , up to isometry , a unique @xmath13-maximizing unlabeled configuration of spheres with @xmath74 .
@xmath108 for @xmath119 , the vertices of a regular icosahedron form the unique @xmath13-maximizing configuration .
the @xmath13-maximizing configuration for @xmath120 is a regular icosahedron with one vertex removed . in ( * ?
* theorem ii ) danzer classified irreducible sets for @xmath121 .
there are additional @xmath3-irreducible graphs for @xmath122 in these cases . for @xmath123
he finds one optimal set and one irreducible set with one degree of freedom .
he also finds for @xmath112 an irreducible set with two degrees of freedom .
for @xmath124 he finds one optimal set , two irreducible sets with no degrees of freedom , and five with at least one degree of freedom .
danzer states that the irreducible sets with no degrees of freedom ( presumably ) give relative optima .
an irreducible graph having a degree of freedom fails to be relatively optimal , since deforming along its degree of freedom leads to a boundary graph with an additional edge , where the extrema is reached .
danzer s work was not published in a journal until the 1980 s . in the interm ,
brczky @xcite gave another solution for @xmath120 , and hrs @xcite for @xmath124 .
very recently the tammes problem was solved for the cases @xmath126 and @xmath127 by oleg musin and alexey tarasov @xcite .
their proofs were computer - assisted , and made use of an enumeration of all irreducible configuration contact graphs ( see @xcite )
. earlier work on configurations of up to @xmath128 points was done by brczky and szab @xcite ) .
the case @xmath95 was solved in 1961 by raphael m. robinson @xcite .
he proved a 1959 conjecture of fejes tth @xcite , asserting that the extremal @xmath129 and that the extremal configuration of @xmath130 sphere centers are the vertices of a snub cube .
coxeter @xcite describes the snub cube .
table [ tab711 ] summarizes optimal angular parameters and radius parameters on the tammes problem for @xmath94 and @xmath95 ( see aste and weaire ( * ? ? ?
the configuration name given is associated to the vertices in the corresponding polyhedron being inscribed in a sphere , e.g. an icosahedron has @xmath119 vertices ( see melnyk et al .
* table 2 ) ) . in the case
@xmath110 the polyhedron is any from a family of trigonal bipyramids , including the square pyramid as a degenerate case . for @xmath120
the polyhedron is a singly capped pentagonal antiprism , i.e. the icosahedron with one vertex deleted .
the cases @xmath131 and @xmath132 are described in @xcite .
figure [ fig : contact ] shows schematically the optimal contact graphs for @xmath94 .
.the tammes problem for small @xmath3 [ cols=">,^ , < , < , < " , ] by taking the alternating sum of each row , or more directly by evaluating @xmath133 , we can compute the euler characteristic @xmath134 of @xmath135 for @xmath136 .
we also have ( * ? ? ?
* proposition 2.3 ) : [ thm:412 ] ( feichtner - ziegler ( 2000 ) ) for @xmath136 the moduli space @xmath137 is homotopy equivalent to the complement of the affine complex braid arrangement of hyperplanes @xmath138 of rank @xmath139 , since @xmath140 its integer cohomology algebra is torsion - free .
it is generated by @xmath8-dimensional classes @xmath141 with @xmath142 with @xmath143 and has a presentation as an exterior algebra @xmath144 where the ideal @xmath145 is generated by elements @xmath146 and @xmath147 here the complexified @xmath148-arrangement of hyperplanes @xmath149 of rank @xmath139 is cut out by the hyperplanes @xmath150 its complement @xmath151 is homeomorphic to @xmath152 .
the associated affine arrangement is : @xmath153 treating @xmath154 , we set @xmath155 a more refined result determines the integral cohomology ring for the configuration spaces of spheres , which includes torsion elements .
it was determined by feichtner and ziegler , who obtained in the special case of @xmath156 the following result ( see ( * ? ? ?
* theorem 2.4 ) ) .
[ f - z ] ( feichtner - ziegler ( 2000 ) ) for @xmath157 , the integer cohomology ring + @xmath158 has only @xmath36-torsion .
it is given as @xmath159 in which @xmath145 is the ideal of relations given in theorem [ thm:412 ] .
in this result the expression @xmath160 denotes a direct summand of @xmath161 in cohomology of degree @xmath162 , e.g. there is a @xmath163 direct summand in @xmath164 .
morse theory , as treated in milnor @xcite , concerns how topology changes for the _ sublevel sets _
@xmath165 of a given , sufficiently nice , real - valued function @xmath166 on a manifold @xmath167 , as the level set parameter @xmath168 varies . at the _ critical values _ of the function , where its gradient vanishes , the topology changes
this change can be described by adding up the contributions of individual _ critical points _ of the function that occur at the critical values .
more precisely , a _
morse function _ is a smooth enough function that has only isolated _ critical points _ , each of which is non - degenerate , and arranged so that only one critical point occurs at each critical level @xmath169 . here _
non - degenerate _
means that the function @xmath166 is twice - differentiable and its hessian matrix @xmath170 $ ] is nonsingular at the critical point .
the topology of a sublevel set @xmath171 is changed as @xmath168 ascends past a critical value , up to homotopy , by attaching a cell of dimension equal to the _ index _ of the critical point : the number of negative eigenvalues of the hessian .
our interest here will be in _
superlevel sets _
@xmath172 whose topology changes as @xmath173 descends past a critical value by attaching a cell of dimension equal to the _ co - index _ of the critical point : the number of positive eigenvalues of the hessian . in the 1980 s goresky and
macpherson @xcite developed morse theory on more general topological spaces than manifolds , namely _
stratified spaces _ in the sense of whitney @xcite , and applicable to a wider class of real - valued functions .
the configuration spaces such as @xmath174 studied here are in general stratified spaces in whitney s sense , because viewed using the @xmath2-parameter they are real semi - algebraic varieties . for the case at hand of @xmath175 and the injectivity radius function @xmath176 , we have a further problem that @xmath177 is not a morse function .
its critical points are degenerate and non - isolated , and even the notion of `` critical '' needs care in defining , since @xmath176 is a min - function of a finite number of smooth functions ( see definition [ def:41 ] ) .
technically , the angular distance function from @xmath178 is not smooth at the antipodal point @xmath179 , with angular distance @xmath180 on @xmath11 ; however we can treat these functions as if they were smooth using the following trick , valid for the nontrivial cases @xmath157 where @xmath181 : simply include the constant function @xmath182 among those functions over which we take the min , and smoothly cut off the other pairwise angular distance functions @xmath183 if they exceed @xmath184 .
an appropriate version of morse theory that applies in this context , called _ min - type morse theory _ , has only recently been sketched by gershkovich and rubinstein @xcite ( see also baryshnikov et al .
related work includes carlsson et al .
@xcite and alpert @xcite .
the treatment of @xcite studies a notion of topologically critical value . in what follows
we develop an alternative max - min approach to criticality and a morse theory for the injectivity radius function @xmath176 on configurations that is in the spirit of the criticality theory for maximizing _
thickness _ or normal injectivity radius ( also known as _ reach _ ) on configurations of curves subject to a length constraint ( or in a compact domain of @xmath75 , or in @xmath185 ) studied earlier in optimal ropelength and rope - packing problems by cantarella et al .
this approach provides a notion of critical configuration , refining the notion of a critical value .
the farkas lemma ( and its infinite - dimensional generalizations in the case of the ropelength problem ) is a key tool used in these works that relates criticality to the existence of a balanced system of forces on the configuration .
a more detailed treatment is planned in @xcite . to understand criticality for the injectivity radius function @xmath176 on @xmath186
, we first need to make sense of varying a configuration @xmath187 along a tangent vector @xmath188 to @xmath189 at @xmath190 ; here @xmath191 is a tangent vector to @xmath11 at @xmath192 , for @xmath193 . for sufficiently small @xmath168 we can define a nearby configuration @xmath194 by translating and projecting each factor back to @xmath11 . in particular , the @xmath195-directional derivative @xmath196 of a smooth function @xmath166 on @xmath189 at @xmath190 is simply @xmath197 , so @xmath190 is a critical point for smooth @xmath166 provided all its @xmath195-directional derivatives vanish at @xmath190 ; this means that the increment @xmath198 , where @xmath199 is a function which tends to @xmath21 faster than linearly .
the operation taking @xmath190 to @xmath200 can be thought of as the spherical analog of translating @xmath190 by @xmath201 via vector addition in the linear case , hence the suggestive sum notation .
the map taking @xmath201 to @xmath200 approximates ( to within @xmath202 ) the exponential map at @xmath190 .
now we make precise `` max - min criticality '' for the injectivity radius function @xmath176 .
[ def:46 ] a configuration @xmath203 is _
critical for maximizing _
@xmath176 provided for every @xmath188 and sufficiently small @xmath168 , we have @xmath204_+ = o(t),\ ] ] where @xmath205_+=\max\{g,0\}$ ] denotes the positive part of @xmath206 .
equivalently , a configuration @xmath190 is critical if _ no _ variation @xmath195 can _ increase _ @xmath176 to first order .
otherwise , a configuration @xmath190 is _ regular _ , that is , there exists a variation @xmath195 which _ does _ increase @xmath176 to first order , and so , by the definition of @xmath176 as a min - function , this means that for all pairs @xmath207 realizing the minimal angular distance @xmath208 , their distances increase to first order under the variation @xmath195 as well .
note that the set of regular configurations is open . if each configuration in this @xmath209-level set is regular , then this level is _ topologically regular : _ that is , there a deformation retraction from @xmath210 to @xmath211 for some @xmath212 ( see ) . [ def:47 ] for @xmath213 ,
the _ contact graph _ of @xmath190 is the graph embedded in @xmath11 with vertices given by points @xmath192 in @xmath190 and edges given by the geodesic segments @xmath214 $ ] when @xmath215 .
examples of contact graphs for extremal values of the tammes problem were given in figure [ fig : contact ] of section [ sec:3 ] .
[ def:48 ] a _ stress graph _ for @xmath213 is a contact graph with nonnegative weights @xmath216 on each geodesic edge @xmath217 $ ] .
a stress graph gives rise to a system of _ tangential forces _ associated to each geodesic edge @xmath217 $ ] of the contact graph .
these forces have magnitude @xmath216 , are tangent to @xmath11 at each point @xmath192 of @xmath190 , and are directed along the outward unit tangent vectors @xmath218 to the edge @xmath219 at its endpoints @xmath220 , respectively .
a stress graph is _ balanced _ if the vector sum of the forces in the tangent space of @xmath11 at @xmath192 is zero for all points of @xmath221 a configuration @xmath190 is _ balanced _ if its underlying contact graph has a balanced stress graph for some choice of non - negative , not - everywhere - zero weights on its edges .
[ thm : contact ] to each critical value @xmath222 for the injectivity radius @xmath176 , there exists a balanced configuration @xmath190 with @xmath223 .
the vertices of the contact graph are a subset of the points in @xmath190 and the geodesic edges of the contact graph all have length @xmath13 . as in ( * ? ? ? * corollary 3.4 and equation 2 ) , since @xmath176 is a min - function on @xmath224 , if @xmath222 is not a topologically regular value of @xmath176 , then some configuration @xmath225 is balanced . because @xmath226 , the conditions on the vertices and edge lengths are clearly met .
we now prove a converse result .
[ thm : converse ] if a configuration @xmath227 on @xmath11 is balanced , then @xmath227 is critical for maximizing the injectivity radius @xmath176 .
we will need a preliminary lemma . consider a planar graph @xmath161 embedded on the unit sphere @xmath11 via a map @xmath228 which is @xmath229 on the edges of @xmath161 .
( by slight abuse of notation , a point on its image in @xmath11 may also be denoted by @xmath178 . )
suppose each edge @xmath219 of @xmath161 is assigned a _ nonzero _
weight @xmath230 .
let @xmath231 denote the length of edge @xmath232 induced by the map @xmath178 , and let @xmath233 be the total _ weighted length _ of the embedded graph @xmath234 .
we can vary the map @xmath178 using a @xmath229 vector field @xmath235 , just as we varied a configuration : for sufficiently small @xmath168 , each point @xmath236 on the image of the graph is moved to @xmath237 .
let @xmath238 denote the first derivative at @xmath239 of weighted length for this varied graph , i.e. the _ first variation _ of @xmath240 along @xmath235 .
[ lem:412 ] the first variation @xmath238 of the weighted lengh @xmath240 for the embedded graph @xmath234 vanishes for every vector field @xmath235 on @xmath11 if and only if the following two conditions hold : @xmath104 each edge @xmath219 joining a pair of vertices @xmath241 of @xmath161 maps to a geodesic arc @xmath242=[{{\bf u}}^-,{{\bf u}}^+]$ ] in the embedded graph @xmath234 ; @xmath108 at any vertex @xmath178 of the embedded graph @xmath234 , the weighted sum @xmath243 , where the sum is taken over the subset of edges @xmath244 incident to @xmath178 , and where @xmath245 is the outer unit tangent vector of @xmath244 at @xmath178 .
this lemma is a direct consequence of the first variation of length formula @xmath246 ( see , for example , hicks ( * ? ? ?
* chapter 10 , theorem 7 , page 148 ) ) . here
@xmath247 is the unit tangent vector field of the edge @xmath232 , and @xmath248 is the geodesic curvature vector of @xmath232 ; with respect to any local arclength parameter on @xmath232 , the geodesic curvature vector is the projection to @xmath11 of the acceleration : @xmath249 , which is tangent to @xmath11 and normal to @xmath232 , and which vanishes iff @xmath232 is a geodesic arc .
now express @xmath250 as a sum of edge terms and vertex terms . the geodesic arc condition ( 1 ) that @xmath251 along every edge implies the edge terms in @xmath238 all
vanish for any variation @xmath235 of the map @xmath178 ; and the force balancing condition ( 2 ) implies all vertex terms vanish for any variation @xmath235 .
conversely , given any interior image point @xmath178 of an edge , take a variation @xmath235 supported in an arbitrarily small neighborhood of @xmath178 , and orthogonal to @xmath232 at @xmath178 : the vanishing of @xmath238 implies condition ( 1 ) that @xmath251 ; similarly , at any given vertex @xmath252 , consider a pair of variations @xmath253 supported in an arbitrarily small neighborhood of @xmath178 which approximate an orthogonal pair of translations of the tangent space to @xmath11 at @xmath178 : the vanishing of @xmath238 for both of these @xmath253 implies the forces balance ( 2 ) . in case @xmath254 , vanishing for the first variation of @xmath255 does not imply @xmath232 is a geodesic arc : instead , the edges with nonzero weights form a balanced geodesic subgraph of the original embedded graph @xmath234 .
lemma [ lem:412 ] suggests the following definition .
[ def:413 ] an embedded graph satisfying properties ( 1 ) and ( 2 ) is called a _ balanced geodesic graph . _
( note that there is no requirement here that the geodesic edge lengths are integer multiples of some basic length , as would be the case for a contact graph . )
lemma [ lem:412 ] shows that a balanced geodesic graph has vanishing first variation of weighted length @xmath255 , even if some of its edge weights @xmath216 are zero . by hypothesis , there are non - negative edge weights ( not all zero ) so that the resulting stress graph @xmath234 for the configuration * u * is balanced . by lemma [ lem:412 ]
the first variation @xmath238 of weighted length for @xmath234 vanishes for all variation vector fields @xmath235 on @xmath11 .
suppose ( to the contrary ) that * u * were _ not _
critical for maximizing the injectivity radius @xmath176 .
then there would be a variation @xmath256 of @xmath257 so that every geodesic edge of the stress graph has length increasing at least linearly in @xmath256 .
extend @xmath256 to an ambient @xmath229 variation vector field @xmath235 on @xmath11 .
since the edge weights are _ nonnegative _ , and not all zero , that implies the weighted length of the stress graph also increases at least linearly in @xmath235 , a contradiction .
a key property of balanced configurations is that for each @xmath136 the set of radii @xmath2 such that @xmath135 contains a balanced configuration @xmath190 of injectivity radius @xmath258 is finite .
it follows that _ the set of critical radius values for @xmath135 is finite_. this finiteness result can be proved using the structure of the spaces @xmath17 $ ] as real semi - algebraic sets , which we consider in @xcite .
we will assume this finiteness result holds in the discussions in [ sec:43a ] ; it can be directly verified for small @xmath3 . for small radii , it is convenient to state results for @xmath259 in terms of the angle parameter @xmath13 . for sufficiently small angles , the superlevel sets
@xmath260 will have the same homotopy type as the full configuration space @xmath189 . in terms of the radius function
, the conclusion of this result applies for @xmath261 , where @xmath262 is the smallest critical value for @xmath10 $ ] .
[ thm:43 ] suppose @xmath136 .
the smallest critical value for maximizing @xmath176 on @xmath135 is @xmath263 , achieved uniquely by the @xmath3-ring configuration of equally spaced points along a great circle .
moreover , for angular diameter @xmath264 the following hold .
@xmath104 the space @xmath260 is a strong deformation retract of the full configuration space @xmath265 .
@xmath108 the reduced space @xmath266 is a strong deformation retract of the full reduced configuration space @xmath135 . consequently each has , respectively , the same homotopy type and cohomology groups as the corresponding full configuration space .
this result corresponds to ( * ? ? ?
* theorem 5.1 ) .
first note that by using equal weights on each of its edges , the @xmath3-ring is balanced and hence a critical configuration by theorem [ thm : converse ] .
the balanced contact graph on @xmath11 of a @xmath267-critical @xmath3-configuration has geodesic edges with angular length @xmath268 . in order to balance
, its total angular length must be at least @xmath269 , the length of a complete great circle .
thus if @xmath270 , then the total length @xmath271 and there is no balanced @xmath3-configuration in @xmath260 and @xmath222 is not a critical value for @xmath176 . in this case
, a weighted @xmath176-subgradient - flow provides the strong deformation retraction of @xmath272 to @xmath260 .
[ cor:48 ] for @xmath273 and @xmath274 the configuration spaces @xmath260 and @xmath275 are path - connected , but not simply - connected .
these spaces have the same homotopy type as @xmath189 ( resp .
@xmath135 ) , which is connected since @xmath276 ( resp .
@xmath277 ) .
they each are closures of open manifolds and are connected , so are path - connected .
we have @xmath278 for some @xmath279 , using the formula applied for @xmath274 , so @xmath135 is not simply - connected .
finally , @xmath189 is not simply connected via the product decomposition in theorem [ thm:411 ] .
we consider reduced configuration spaces @xmath17 $ ] having radius parameter @xmath2 sufficiently close to @xmath5 , depending on @xmath3 . using the finiteness of the set of critical values , there is an @xmath280 such that the upward `` gradient flow '' of the injectivity radius function @xmath176 ( or of the corresponding touching - sphere radius function @xmath2 ) defines a deformation retraction from @xmath17 $ ] to @xmath281 $ ] for the range @xmath282 .
the simplest topology that may occur at @xmath5 is where @xmath283 $ ] has all its connected components contractible ; the property holds for most small @xmath3 in fact , for all @xmath284 except @xmath110 .
when it holds , the cohomology groups for @xmath17 $ ] in this range of @xmath2 will have the following very simple form : for @xmath110 the cohomology does _ not _ have the purity property . the reduced configuration space @xmath289 $ ] is @xmath290-dimensional for @xmath291 but becomes @xmath36-dimensional at @xmath292 .
some optimal maximum radius configurations at @xmath293 have room for an extra sphere ( giving @xmath294 ) : the sphere centers form five vertices of an octahedron , and either vertex in an antipodal pair of vertices can freely and independently move towards the unoccupied sixth vertex of the octahedron .
the resulting reduced configuration space @xmath295= { \operatorname{bconf}}(5 ; \frac{\pi}{2})$ ] is a simplicial @xmath36-complex which is not contractible ; it is pictured schematically in figure [ fig : example5 ] .
it has a single connected component having euler characteristic @xmath296 . for further discussion of this space as a critical stratified set ,
see section [ sec:46 ] .
-maximal stratified set for @xmath110,width=192 ] does the purity property hold for all or most large @xmath3 ?
we do not know .
one might expect that extremal configurations for high values of @xmath3 at @xmath297 will have most spheres are held in a rigid structure , and for @xmath2 near it all individual spheres will only be able to move in a tiny area around them , each contributing a connected component to the reduced configuration space . against this expectation ,
computer experiments packing @xmath3 equal - radius two - dimensional disks confined to a unit disk suggest the possibility for some @xmath3 that extremal configurations could have _ rattlers _ , which are loose disks that have motion permitted even at @xmath298 ( lubachevsky and graham @xcite ) . however , even with rattlers one could still have contractibility of individual connected components .
the hypothesis of extremal configurations being rigid ( and unique ) is known to hold for @xmath299 .
when the purity property holds one can ( in principle ) determine the number of connected components for the set of near - maximal configurations ; call it @xmath300 .
this value depends on the symmetries of each maximal configuration under the @xmath16 action . denoting the isomorphism types of the connected components of maximal rigid ( labeled ) configurations of @xmath3 points at @xmath297 by @xmath301 for @xmath302
, one would have @xmath303 for @xmath304 , excluding @xmath305 , the extremal configurations for the tammes problem are known to be unique up to isometry ; call them @xmath306 .
the analysis of danzer given in theorem [ th73 ] covers the cases @xmath307 . for the case
@xmath119 , the unique extremal configuration @xmath9 of vertices of an icosahedron has @xmath308 , the alternating group , of order @xmath309 , whence @xmath310 connected components of critical strata necessarily have dimension at least three from the @xmath16-action . in what follows
we consider reduced critical strata that quotient out by this action . at a critical value @xmath177
there can be several disconnected reduced critical strata , and such strata can have positive dimension .
we give examples of each .
for @xmath110 a positive dimensional reduced critical stratified set occurs at the maximal radius value @xmath311 .
the set of ( reduced ) critical configurations forms a family , which is two - dimensional , containing multiple strata .
a generic contact graph at the maximal injectivity radius @xmath312 is a @xmath313-graph having @xmath36 polar vertices and @xmath7 equatorial vertices .
this contact graph , depicted in figure [ fig : contact ] , has @xmath7 faces and @xmath314 edges and is optimal .
the three angles between equatorial vertices can range between @xmath315 and @xmath180 , with the condition that their sum is @xmath269 , defining a @xmath36-simplex . as long as none of the equatorial angles
is @xmath180 , criticality is achieved using weights that are non - zero on all the edges . when an equatorial angle is @xmath180 , corresponding to a corner of the @xmath36-simplex , these equatorial vertices may be regarded as a new pair of polar vertices . in this configuration , as the angles between equatorial angles go to @xmath180 , some weights of the stress graph can go to @xmath21 and the support of the weights degenerates to a @xmath316-ring .
the limit contact graph consists of the edges of a square pyramid whose base is that @xmath316-ring .
this gives a non - optimal contact graph with @xmath27 faces and @xmath317 edges .
for @xmath119 there are several distinct reduced critical strata at the critical value @xmath318 , two of which correspond to the @xmath24-configuration and @xmath25-configurations , singled out in frank s discussion in section [ sec:27 ] .
these configurations are defined in section [ sec:52 ] below , and their criticality is shown in theorem [ th913bb ] .
for very small @xmath3 it is possible to completely work out all the critical points and the changes in topology .
we illustrate such an analysis on the simplest nontrivial example @xmath319 ( see figure [ fig : morse4 ] , explained below ) . ]
we consider the reduced superlevel sets @xmath320 $ ] . since @xmath321 is @xmath317-dimensional , away from the critical values these spaces are @xmath27-dimensional manifolds with boundary .
if we ignore the labelling of points and classify the contact graphs for four vertices , there are exactly two geometrically distinct @xmath176-critical @xmath316-configurations in @xmath322 : 1 .
the @xmath316-ring of four equally spaced points around a great circle on @xmath11 with @xmath323 which is a saddle configuration for @xmath176 .
there is a @xmath8-dimensional subspace of the tangent space to @xmath322 at the @xmath316-ring along which @xmath176 increases to second order , i.e. the co - index is @xmath324 .
the critical value for @xmath2 for the @xmath316-ring is @xmath325 .
2 . the vertices of the regular tetrahedron with @xmath326 , which is the maximizing configuration for @xmath176 on @xmath322 , i.e the co - index @xmath327 .
the critical value for @xmath2 for tet is @xmath328 there are two intervals @xmath329 and @xmath330 $ ] on which the topology of @xmath331 $ ] remains constant . from theorem [ thm:411 ]
, it can be seen that on the interval @xmath332 , @xmath320 $ ] is homeomorphic to @xmath322 .
this has the homotopy type of @xmath38 punctured at two points , hence @xmath333 , { \mathbb z } ) = { \mathbb z } , \quad h^{1}({\operatorname{bconf}}(4)[r ] , { \mathbb z } ) = { \mathbb z}^2,\quad h^{k}({\operatorname{bconf}}(4)[r ] , { \mathbb z } ) = 0 , \ , \mbox{for}\,\ , k \ge 2.\ ] ] on the open interval @xmath334 , the manifold @xmath320 $ ] has two connected components , each diffeomorphic to a @xmath27-ball , which can be seen from the strong deformation retraction to @xmath335 $ ] consisting of the two points associated to the orientated labelings of @xmath336 configurations , @xmath337 and @xmath338 .
hence @xmath333 , { \mathbb z } ) = { \mathbb z}^2 , \quad h^{k}({\operatorname{bconf}}(4)[r ] , { \mathbb z } ) = 0 , \ , \mbox{for}\,\ , k \ge 1.\ ] ] figure [ fig : morse4 ] above is only a schematic picture , since we can not draw a @xmath27-dimensional manifold .
it compresses four of the dimensions .
the visible points take @xmath2-values with @xmath339 .
the value @xmath340 is a circular vertical ring in the middle , and the values of @xmath2 increase as one moves to the left or right , reaching a maximum at @xmath337 and at @xmath338 . from table
[ table4 ] , we can easily compute the euler characteristic @xmath341 . the indexed sum of critical points of the function @xmath342
gives an alternative computation of the euler characteristic as @xmath343 we count the _ labeled _ configurations in @xmath322 : since the @xmath316-ring has symmetry group @xmath344 of order @xmath317 in @xmath16 , there are @xmath345 critical points of this type with co - index @xmath8 ; and since tet has symmetry group @xmath346 of order @xmath0 in @xmath16 , there are really @xmath347 critical points of this type with co - index @xmath21 ; and so we obtain @xmath348 as predicted .
in fact , the morse complex for @xmath176 captures the fact that @xmath322 itself has the homotopy type of the @xmath313-graph : there are @xmath36 vertices ( @xmath21-cells ) in the complex corresponding to the maxima ( co - index @xmath21 ) @xmath337 and @xmath349 configurations ; there are @xmath7 edges ( @xmath8-cells ) corresponding to the @xmath7 saddle ( co - index @xmath8 ) @xmath316-ring configurations .
the complexity of the changes in topology of the configuration space grows rapidly with @xmath3 . for larger values of @xmath3
there are many @xmath176-critical configurations which are not maximal .
the value @xmath119 is large enough to be extremely challenging to obtain a complete analysis of the critical configurations of the configuration space , and to analyze the variation of the topology as a function of the radius @xmath2 .
the betti numbers for @xmath119 for radius @xmath39 given in table [ table4 ] differ greatly from those at @xmath351 where the cohomology of @xmath352 $ ] is entirely in dimension @xmath21 , according to the purity property , which holds for @xmath119 by results in section [ sec:3 ] .
this topology change involves millions of ( labeled ) critical points .
its full investigation remains a task for the future .
in this section , we discuss @xmath353 $ ] and @xmath23 $ ] , the configuration spaces of @xmath0 unit spheres touching a central unit sphere @xmath11 .
these configuration spaces are remarkable and have some special properties .
the value @xmath354 is a critical value and has ( at least ) two geometrically distinct critical points , the @xmath24 and @xmath25 configurations .
we believe @xmath8 is the maximal radius @xmath2 where the spheres @xmath26 $ ] are arbitrarily permutable with motions remaining on @xmath11 ( see section [ sec:65 ] ) .
the case where all spheres are unit spheres has been extensively studied in connection with sphere packing .
the value @xmath22 is a critical value of the radius function @xmath2 , and we will see that the associated configuration spaces @xmath353 $ ] and @xmath355 $ ] are not manifolds . to better understand their topology , it is useful to consider the spaces @xmath356 $ ] and @xmath26 $ ] for @xmath2 in a neighborhood of @xmath8 .
these are stratified spaces naturally embedded in @xmath357 and filtered by @xmath2 . for noncritical values of @xmath2 , the spaces @xmath356 $ ] and @xmath26 $ ] are submanifolds with boundary . for all @xmath358
, the space @xmath356 $ ] has top dimension @xmath130 .
after factoring out the ambient @xmath16-action , the space @xmath26 $ ] has top dimension @xmath19 .
we now consider the three configurations of @xmath0 touching spheres singled out by frank ( 1952 ) @xcite . in figure
[ fig : configs ] , the three polyhedra have vertices located at the @xmath0 touching sphere centers of these configurations and centroids located at the center of the central sphere .
the edges of these polyhedra specify the contact graphs of these configurations , also pictured schematically in figure [ fig : configs ] .
the @xmath9 configuration realizes the optimal contact graph for @xmath119 given in figure [ fig : contact ] in section [ sec:33 ] . *
the @xmath9 configuration is obtained by placing @xmath0 spheres touching a central @xmath6-th sphere at the vertices of an inscribed icosahedron ; such touching points are also the centers of the faces of a circumscribed dodecahedron .
it has oriented symmetry group @xmath359 , the icosahedral group , of order @xmath309 and in @xmath23 $ ] there are @xmath360 of these . * the @xmath24 configuration is obtained by stacking three layers of the hexagonal lattice , with the third layer not lying over the first layer .
the inscribed polyhedron formed by the convex hull of the @xmath0 points of the @xmath24 configuration where the spheres touch the central sphere is a cuboctahedron .
the circumscribed dual polyhedron which has the @xmath0 points as the center of its faces is a rhombic dodecahedron .
it has oriented symmetry group @xmath361 , the octahedral group , of order @xmath130 and in @xmath23 $ ] there are @xmath362 of these . * the @xmath25 configuration is obtained by stacking three layers of the hexagonal lattice , with the third layer lying directly over the first layer .
the inscribed polyhedron formed by the convex hull of the @xmath0 points of @xmath25 where the spheres touch the central sphere is a triangular orthobicupola .
this polyhedron is the johnson solid @xmath363 .
the circumscribed dual polyhedron which has the @xmath0 points as the center of its faces is a trapezoidal rhombic dodecahedron .
it has oriented symmetry group @xmath364 , the dihedral group of order @xmath314 , and in @xmath23 $ ] there are @xmath365 of these .
the @xmath24 and @xmath25 configurations are elements of @xmath366 $ ] , while the @xmath9 configuration is an interior point of @xmath23 $ ] .
we first consider rigidity properties of these packings . in the following definition
we identify a sphere tangent to the @xmath36-sphere with the circular disk ( i.e. spherical cap ) on @xmath11 that it produces by radial projection .
connelly @xcite ) a packing of disks on @xmath11 is _ locally jammed _ if each disk is held fixed by its neighbors .
that is , no disk in the packing can be moved if all the other disks are held fixed .
we say a configuration of disks is _ jammed _ if it can only be moved by rigid motions .
we call it _ completely unjammed _ if each disk can be moved slightly while holding all the other disks fixed . [ th913 ] the @xmath9 configuration in @xmath353 $ ] is completely unjammed . its space of ( infinitesimal ) deformations has dimension @xmath130 .
the deformation space is @xmath19 dimensional if viewed in @xmath23 $ ] .
the maximal radius for @xmath0 spheres is @xmath367 and is achieved in the @xmath9 configuration .
therefore the deformation space at @xmath9 is full dimensional .
in contrast , both the @xmath24 and @xmath25 configurations are _ locally jammed _
, i.e. they are rigid against motion of any one disk while holding all the other disks fixed ; each of their infinitesimal deformation spaces has codimension at least @xmath36 .
later , in section [ sec:53 ] , we describe a deformation of the @xmath9 packing to the @xmath24 packing .
this deformation , properly adjusted , has @xmath314 moving balls during its final phase arriving at @xmath24 .
( the @xmath314 fixed balls form an antipodal pair of `` triangles '' . )
we believe this value @xmath314 to be the smallest number of moving balls needed to unjam the @xmath24 configuration . for a manual on how to unlock @xmath24 , see section [ manual ] .
a deformation of the @xmath9 configuration to the @xmath25 configuration , also described in section [ sec:53 ] , requires @xmath368 moving balls at the instant of arrival at @xmath25 .
we believe this value @xmath368 to be the smallest number of moving balls needed to unjam the @xmath25 configuration . for a manual on how to unlock @xmath25 ,
see section [ manual ] . a possible reason for the larger number of moving balls needed to unjam
the @xmath25 configuration compared with that of the @xmath24 configuration is that the @xmath25 configuration has fewer local symmetries .
the @xmath24 and @xmath25 configurations are critical configurations for @xmath22 . according to theorem [ thm : converse ]
it suffices to show that these configurations carry balanced contact graph structures .
[ th913bb ] the @xmath24 configuration and @xmath25 configuration for @xmath22 carry balanced contact graph structures .
consequently , @xmath22 is a critical value of the radius function on @xmath369 . by theorem [ thm :
converse ] , a sufficient condition for the criticality of a configuration for maximizing injectivity radius is that its contact graph can be balanced .
that is , a set of positive weights may be assigned to the edges of the contact graph so that at each vertex the weighted vector sum , defined by the outward tangent vectors to the incident edges , vanishes . and @xmath25 configurations , title="fig:",width=230 ] ( a ) fcc configuration + and @xmath25 configurations , title="fig:",width=259 ] ( b ) hcp configuration + we now indicate weight values for the @xmath24 and @xmath25 configurations ( see figure [ fig : stress - graph ] ) .
\(1 ) at radius @xmath22 for the @xmath24 configuration , the stress graph is balanced when all the weights are equal .
this can be seen from the cubic- or @xmath361-symmetry of the contact graph .
\(2 ) at radius @xmath22 for the @xmath25 configuration , consider a weight @xmath370 on edges between triangular faces and square faces , a weight @xmath371 on edges between pairs of square faces , and a weight @xmath372 on edges between pairs of triangular faces . from the structure of the contact graph , it is possible to choose a constant @xmath373 and find a weight @xmath374 which balances the associated stress graph .
this suffices to balance this configuration with some zero weights .
however , it is also possible to add a uniform constant weight @xmath375 to the equatorial great circle , giving a balanced stress graph with positive weights on all edges of @xmath376 .
the @xmath9 configuration is not a critical configuration for @xmath22 ; instead it is a critical configuration at the maximal radius @xmath377 .
as noted in section [ sec:2 ] , fejes tth @xcite conjectured that this configuration does have a certain extremality property for local packing by equal spheres , that it gives a minimizer for a single voronoi cell of a unit sphere packing .
this statement , the dodecahedral conjecture , was proved in 2010 by hales and mclaughlin @xcite .
[ th514a ] ( hales and mclaughlin ( 2010 ) ) a @xmath9 configuration of unit spheres minimizes the volume of a voronoi cell of a unit sphere with center at the origin of @xmath75 over all sphere packing configurations of unit spheres containing that sphere .
the volume of this voronoi cell gives a local sphere packing density of approximately @xmath76 , which exceeds the sphere packing density @xmath48 in @xmath7-dimensional space .
we now show that the @xmath9 configuration can be continuously deformed inside @xmath23 $ ] to the @xmath24 configuration and to the @xmath25 configuration .
[ th55 ] @xmath104 on the space @xmath353 $ ] there is a continuous deformation of the @xmath9 configuration to the @xmath24 configuration that remains in the interior @xmath378 $ ] of @xmath353 $ ] till the final instant .
@xmath108 there is also a continuous deformation of the @xmath9 configuration to the @xmath25 configuration that remains in the interior @xmath378 $ ] of @xmath353 $ ] till the final instant .
the motions of these two deformations , measured from the touching points of the @xmath0 spheres to the central sphere , can be given by piecewise analytic functions on the @xmath36-sphere .
the proof of theorem [ th55 ] is given in sections [ sec:531 ] - [ sec:546 ] .
an unlocking manual for doing it is given in the appendix ( section [ manual ] ) .
to describe the deformations we will need coordinates . in sections [ sec:531 ] - [ sec:545 ]
we suppose a ball with radius @xmath8 centered at @xmath21 touches all the @xmath0 balls of the same given radius @xmath2 .
we initially allow all values of @xmath1 $ ] , but in the move @xmath379 described in later subsections we will necessarily restrict to @xmath380 .
we take @xmath9 to consist of @xmath0 equal balls of radius @xmath2 , touching a central unit sphere at @xmath0 vertices of an inscribed icosahedron @xmath381 .
we view this icosahedron @xmath381 as embedded in @xmath65 with cartesian coordinates so that we have : * its centroid is at the center @xmath382 of the unit sphere .
* it has two opposite faces parallel to the @xmath383-plane . in other words , for some @xmath384 , the intersections @xmath385 are triangular faces of @xmath386 here @xmath387 where @xmath388 is the golden ratio .
-move @xmath379,title="fig : " ] + ( a ) phase 1 -move @xmath379,title="fig : " ] + ( b ) phase 2 the _ north balls _ are those which have their centers in the plane @xmath389 while the _ south balls _ have their centers in the plane @xmath390 the three north balls form a _ north triangle _ which is centrally symmetric to the _ south triangle _ formed by the three south balls , as in @xmath24 .
together with north and south triangles , we have @xmath314 remaining balls , which will be called _ equatorial _ , even though in the initial configuration they do not have their centers on the equator .
the equator lies in the plane @xmath391 . to fix their positions ,
let the _ greenwich meridian _ be defined as @xmath392 and the longitude @xmath393 be measured from it in the counterclockwise direction viewed from the north pole @xmath394 .
we require : * the center of one of the north balls is in the half - plane of the greenwich meridian , i.e. this ball touches the greenwich meridian .
let us call this ball @xmath395 * the center of one of the equatorial balls is in the half - plane of the greenwich meridian .
let us call it @xmath396 it will necessarily be in the southern hemisphere .
this fixes the location of all @xmath0 balls . with this orientation of the icosahedron ,
the meridians of the north triangle are spaced by @xmath397 furthermore the meridians of the other three balls in the northern hemisphere are also spaced by @xmath184 and the meridians combined are spaced by @xmath182 as in @xmath24 .
the same holds for the six balls in the southern hemisphere .
we now define the `` @xmath314-move '' deformation @xmath379 , which has two variants , one leading from @xmath9 to the @xmath24 configuration , and the other leading to the @xmath25 configuration .
this move proceeds in two phases .
the first phase is the same for both variants .
it moves the @xmath314 balls that are not equatorial at constant speed along meridians towards the poles , until they form north and south triangles of three mutually touching balls .
the @xmath314 equatorial balls do not move . in the second phase
, all @xmath0 balls are moving .
in both variants , the @xmath314 equatorial balls , initially not on the equator , move towards the equator along their meridians at constant speed , to arrive on the equator at the end of the move , forming a ring of six balls on the equator .
this ring is an allowed configuration only if @xmath399 they do not touch during this move , until the last moment , and then all touch if @xmath22 . at the same time
, the north and south triangles will rotate about the polar axis at a variable speed , the same for all six , in such a way as to avoid the equatorial balls .
they will rotate by @xmath182 to their final position . for the @xmath24 move ,
the north triangle and south triangle rotate in the same direction , while for the @xmath25 move they rotate in opposite directions .
a key issue is to suitably specify the variable speed of rotation .
denote by @xmath400 the parallel @xmath401 where the three centers of the north triangle stop .
let @xmath402 be the parallel @xmath403 where the south triangle stops .
each of the six centers of the equatorial balls , initially not on the equator , will move at constant speed along their respective meridians towards their final positions on the equator .
parametrize this motion so that at @xmath239 , the @xmath314 balls are at their initial positions , while at @xmath404 , the @xmath314 balls are at their final positions on the equator .
the centers of the north triangle are on @xmath400 at @xmath239 and will remain on @xmath400 throughout the move . similarly ,
the centers of the south triangle are on @xmath402 at @xmath239 and will remain on @xmath402 throughout the move . the triangles simply rotate .
it now suffices to specify functions @xmath405 , which describe the _ increment _ of the longitude of the north triangle during the time @xmath406 $ ] and @xmath407 , the _ increment _ of the longitude of the south triangle during the time @xmath406 $ ] .
we will take @xmath405 to be a continuous , non - decreasing function , with @xmath408 @xmath409 we get two different moves to @xmath24 and @xmath25 depending on which direction the south triangle rotates .
the motion @xmath410 will take us to @xmath24 , and choosing the opposite rotation @xmath411 will take us to @xmath25 . the function @xmath413 defined so that no ball from the two triangles hits any equatorial one , is certainly not unique .
here is a minimal definition of @xmath413 beginning at the second phase .
recall that our balls are open , and that : * the center of one of the three north balls , @xmath414 , is on the half - plane of the greenwich meridian . * the center of one of the equatorial balls , @xmath415 , is also on the half - plane of the greenwich meridian .
* there is an equatorial ball with the longitude @xmath416 call it @xmath417 note that the center of @xmath415 is south of the plane @xmath418 while that of @xmath419 is north of the plane @xmath391 . throughout the second phase of @xmath379 the ball @xmath415
will move north while @xmath419 moves south .
let @xmath420 @xmath421 denote their positions , @xmath422.$ ] then for every @xmath423 define @xmath424 to be the ball with the center at @xmath400 and with the longitude @xmath425 for example , @xmath426 now let us define the function @xmath412 as follows : @xmath427 clearly , @xmath428 for all @xmath168 small enough .
the only thing one needs to check is that @xmath429 holds for all @xmath430 $ ] .
+ [ lem:56 ] an increment function @xmath431 exists : @xmath432 as defined by satisfies .
we use euler coordinates on the sphere .
the latitude @xmath433 of the parallel @xmath400 is @xmath434 where @xmath13 satisfies @xmath435 by symmetry , it is enough to consider the movement of three balls : * @xmath414 on the parallel @xmath400 .
its initial angle @xmath436 the latitude @xmath433 of @xmath437 is constant .
* @xmath415 on the greenwich meridian @xmath438 its initial latitude is @xmath439 and final latitude is @xmath440 on the interval , its latitude is given by @xmath441 * @xmath419 on the meridian @xmath442 its initial latitude is @xmath443 and final latitude is @xmath444 on the interval , its latitude is given by @xmath445 the function @xmath446 is uniquely defined by : * @xmath447 * at every time @xmath448 and after initial contact , the ball @xmath449 touches the ball @xmath450 .
to complete the proof of lemma [ lem:56 ] it remains to check that the balls @xmath449 and @xmath451 of radius @xmath380 are disjoint for @xmath452 .
this fact will follow from the next lemma .
[ lem:57 ] let @xmath453 consider an isosceles spherical triangle @xmath454 where @xmath43 has @xmath455 @xmath456 @xmath42 has @xmath457 @xmath458 and @xmath459 is defined by @xmath460 and the touching condition .
then @xmath461 let @xmath45 be the middle point of the arc @xmath462 it does not depend on @xmath168 and is given by @xmath463 @xmath464 let @xmath465 be the arc perpendicular to the arc @xmath466 at @xmath467 then @xmath468 is simply the intersection of @xmath469 and the parallel @xmath470 the triangle @xmath471 is a right triangle .
evidently , the legs @xmath472 and @xmath473 become shorter as @xmath168 increases .
hence the hypotenuse @xmath474 becomes shorter as well . since @xmath475 the proof follows .
lemmas [ lem:56 ] and [ lem:57 ] complete a proof that there exists a deformation path from the @xmath9 configuration to a @xmath24 configuration and to a @xmath25 configuration , respectively .
however , the deformation path obtained does not satisfy one required condition of the theorem : remaining in the interior of the configuration space .
it exits from the interior of @xmath353 $ ] at the end of the first phase and remains on the boundary during the second phase : the three north balls are touching and the three south balls are touching .
we can modify the construction above so that no balls touch throughout the deformation until the final instant . to do this we halt the first phase just short of the three balls touching , at @xmath476 .
then in the second phase , we allow @xmath477 to increase monotonically in the north triangle at some variable speed @xmath478 as the rotation proceeds , in such a way as to avoid contact between the three north balls and the equatorial balls .
the south triangle @xmath477 variable is to decrease monotonically in the reflected motion of @xmath479 at the same time .
lemma [ lem:57 ] implies that if @xmath477 approaches @xmath8 rapidly enough in the motion that we can again avoid contact ; this is an open condition at each point @xmath168 , so by compactness of the motion interval we have a finite subcover to attain it .
\(1 ) this motion process can be continued by concatenation with an inverse @xmath379 using @xmath480 , in such a way as to arrive back at a @xmath9 configuration , differently labeled .
this is possible because there are two exit directions ( tangent vectors ) from the @xmath24 configuration and two exit directions from the @xmath25 configuration in @xmath23 $ ] .
section [ sec:551 ] studies the group of permutations of the @xmath0 labels obtainable by such deformations .
\(2 ) starting from the @xmath24 or @xmath25 configuration , there is a reference frame in which the north triangle remains fixed . the inverse of the second phase of @xmath379 describes a move which unlocks the @xmath24 configuration with @xmath314 moving balls and @xmath314 fixed balls , and which unlocks the @xmath25 configuration with @xmath368 moving balls and @xmath7 fixed balls ( see section [ manual ] ) . according to his recollection , on 25 april 1948 buckminster fuller found a `` jitterbug '' construction given by a jointed framework motion that , among other things , permits an @xmath24 configuration , given as the vertices of a cuboctahedron , to be continuously deformed into a @xmath9 configuration , given as the vertices of an icosahedron ( see @xcite ) .
in buckminster fuller s construction , the joint distances remain constant during the motion , so that they can be rigid bars , while the radii of the associated touching spheres continuously contract during the deformation . at each instant during the motion the central sphere and the @xmath0 touching spheres can all have equal radii without overlapping , and this radius varies monotonically in time .
in retrospect one may see that it is possible to rescale space during the motion via homotheties varying in time such that all spheres retain the fixed radius @xmath8 throughout the deformation . in this case
the joint lengths will change continuously in the motion .
the rescaled motion no longer corresponds to a physical object with rigid bars , but it does give a continuous motion in the configuration space of @xmath0 equal spheres touching a 13-th central sphere that continuously deforms the @xmath24 configuration to the @xmath9 configuration .
the work of buckminster fuller on the `` jitterbug '' movable jointed framework is described in schwabe @xcite .
fuller described it in his book synergetics ( * ? ? ? * sec .
460.00 - 463.00 ) .
the construction is also described in edmondson ( * ? ? ? * chap .
11 ) , with a detailed analysis in verheyen @xcite .
the `` jitterbug '' motion immediately enters the interior @xmath481 $ ] after the initial instant , in contrast to the `` unlockings '' described in appendix 8 , which adhere to its boundary .
the number of connected components of the configuration space @xmath356 $ ] is related to the ability to permute labeled spheres by deformations within @xmath356 $ ] .
the possible permutability of the ( labeled ) spheres in the @xmath9 configuration in @xmath356 $ ] depends on the radius @xmath2 of the touching spheres . conway and sloane (
* chap . 1 ,
appendix , pp .
2930 ) give a terse proof that for radius @xmath8 the labels on labeled spheres in @xmath9 configurations can be arbitrarily permuted using continuous deformations inside the space @xmath353 $ ] .
[ thm:58](permutability at radius @xmath22 ) for the radius parameter @xmath22 , each labeled @xmath9 configuration can be continuously deformed in the configuration space @xmath23 $ ] to a @xmath9 configuration at the same @xmath0 touching points with any permutation of the labeling .
-move @xmath482,title="fig : " ] -move @xmath482,title="fig : " ] we follow the outline in conway and sloane ( * ? ? ?
* chapter 1 , appendix , pp .
a main ingredient is an additional set of permutation moves which we call @xmath482-moves , detailed next .
beginning from the @xmath9 configuration centered at the origin , we rotate it so that two opposite balls have their centers on the @xmath477 axis .
call these balls @xmath484 and @xmath485 .
note that the centers of @xmath27 of the @xmath132 remaining balls are in the northern half - space , while the remaining @xmath27 centers are in the southern half - space .
call these balls @xmath486 and @xmath487 , respectively
. * first phase .
* move the @xmath27 northern @xmath488 balls towards @xmath484 , in such a way that their centers remain on their corresponding meridians , until each of them touches @xmath484 .
note that these @xmath27 balls do not touch each other , only @xmath484 .
indeed , because their centers are located at the latitude @xmath489 when viewed from the @xmath477-axis each of the @xmath27 balls subtends the dihedral angle @xmath490 of a regular tetrahedron ; but the longitude difference between the neighboring ball centers is @xmath491 , so there remains a tiny longitude gap @xmath492 the @xmath27 southern @xmath493 balls may be moved into the southern hemisphere in the same manner . * second phase . * note that for @xmath22 , the @xmath314 northern balls fit into the northern half - space , while the @xmath314 southern balls fit into the southern half - space .
the union of all the @xmath314 balls in the northern hemisphere may be rotated by @xmath491 as a rigid body , keeping the remaining balls fixed .
* third phase .
* reverse the first phase .
the net result of an @xmath482-move is a cyclic permutation @xmath494 of @xmath9 of length @xmath27 , which is an even permutation .
the halfway point of a @xmath482 as illustrated in the right side of figure [ fig:5-move ] is itself a `` remarkable '' critical configuration ; it is found at the center of a 4-simplex of critical configurations for @xmath22 .
the family of such configurations is in many ways similar to the maximal stratified set that occurs for @xmath110 .
the first step is to show that there is a continuous deformation of @xmath9 to itself , which permutes the labels by an odd permutation . to exhibit it
, we use the move @xmath495 defined before , and deform @xmath9 into an @xmath24 configuration ( note that we can do it for all @xmath496 but not for @xmath41 ) . note that the @xmath24 configuration has three axes of 4-fold symmetry passing through the opposite squares of four balls . by rotating @xmath497 around any such an axis and then deforming our configuration back to @xmath9 via @xmath498
we induce a a permutation @xmath499 of 12 balls , which is a product of three ( disjoint ) cyclic permutations , each of length 4 .
every such cycle is an odd permutation , hence their product @xmath499 is also odd .
the second step uses @xmath483-moves .
each such move gives a cyclic permutation of order @xmath27 .
since there are @xmath0 options for choosing @xmath484 , we get @xmath0 such @xmath27-cycles @xmath500 .
it is shown in conway and sloane ( using an elegant argument about the mathieu group @xmath501 , see @xcite ) that all such @xmath502 generate the alternating group @xmath503 , the subgroup of even permutations of @xmath504 combined with any odd permutation @xmath499 , the full permutation group @xmath505 is generated .
the move @xmath483 can be modified in such a way that it continues to work for all values @xmath506 , for some @xmath507 slightly bigger than @xmath8 .
we first explain the modification and then propose the value of @xmath508 the modification deals only with the second phase of @xmath482 . in order to explain it ,
it is enough to follow the @xmath132 longitude values of the touching points of our balls , which may be considered as points on the equator . for @xmath22 ,
the northern @xmath27 balls @xmath509 correspond to the longitude values @xmath510 for @xmath511 and we can suppose that at the initial moment these values are @xmath512 the longitude values @xmath513 are defined similarly , corresponding to the southern balls @xmath493 , and @xmath514 our initial move looks now as follows : @xmath515 of course there is no need for all the @xmath516 to move with the same speed ; the only constraint is that the difference between consecutive @xmath516 should equal or exceed @xmath490 at all times . in particular
, we can modify the speeds in such a way that at any time @xmath168 , we have @xmath517 for at most one value of @xmath518 now let the radius @xmath2 be slightly bigger than @xmath8 .
then , at the moment @xmath168 when @xmath519 the corresponding balls @xmath520 @xmath493 will overlap .
this , however , can be remedied by making the following small deformation of our @xmath0-configuration : * the ball @xmath488 moves up along its meridian , by the distance @xmath521 * the ball n moves along the same meridian in the same direction by the distance @xmath522 * the ball @xmath493 moves down along its meridian , by the distance @xmath521 * the ball s moves along the same meridian in the same direction by the distance @xmath522 * other balls may be rearranged in such a way that they do not intersect
. the non - overlap condition can be satisfied when @xmath523 is small enough , since there were no other collisions .
below we will show there will be @xmath27 _ bottleneck configurations _ that one encounters on the way to perform the modified @xmath483 move .
each one defines a value @xmath524 , for @xmath525 which is the maximal radius for which this configuration is allowed .
we set @xmath526 [ thm:62 ] for every @xmath527 the move @xmath483 can be modified in such a way that one can reach from an initial labeled @xmath9 configuration any labeled @xmath9 configuration whose labels are an even permutation of the initial labels .
that is , the alternating group @xmath528 is generated by the compositions of different @xmath483 moves .
there will occur @xmath27 bottleneck @xmath0-configurations of the @xmath2-balls touching the unit central ball , described by certain touching patterns that correspond to the configurations appearing during the move @xmath483 at the moment when the ball @xmath488 passes due north of the ball @xmath529 the @xmath27 bottleneck configurations have a common pattern : @xmath316 touching balls centered on the same meridian , two in the northern half - space , and the remaining two in the southern half - space .
we denote them by @xmath484 , @xmath520 @xmath530 @xmath485 .
this set of @xmath316 balls is symmetric with respect to the plane @xmath531 strictly speaking , as @xmath2 is slightly bigger than @xmath532 the balls @xmath484 and @xmath485 are now centered on the meridian _
opposite _ the one containing @xmath488 and @xmath493 .
the eight other balls are the remaining ones from @xmath533 each pair @xmath534 , @xmath535 touches , as do the pairs @xmath534 , @xmath484 } , as well as the pairs \{@xmath536 , @xmath537 .
the @xmath27 bottleneck configurations differ in how the additional pairs of balls touch .
each of the rows in the following list completes a different touching pattern that occurs as the move @xmath482 is performed : @xmath538 @xmath539 @xmath540 @xmath541 @xmath542 observe that for any @xmath41 , and for any of the @xmath27 touching patterns , such a configuration is unique if it exists , and that it _ does _ exist for @xmath543 sufficiently small .
we define @xmath544 as the maximal values for which the above configurations exist . to @xmath545,title="fig:",width=163 ] to @xmath545,title="fig:",width=163 ] to @xmath545,title="fig:",width=163 ] to @xmath545,title="fig:",width=163 ] to @xmath545,title="fig:",width=163 ] we are not asserting that the value @xmath546 defined above is the true critical value above which the ( small perturbation of the ) move @xmath482 can not be performed .
indeed , we imposed some a priori constraints in making our construction of the modified @xmath482 , and did not rule out the possibility of a more `` optimal '' modification of @xmath482 .
[ defi:63 ] _ let @xmath547 be the maximal value of the radius @xmath2 for which there exists some modified move @xmath482 .
we call it the _ upper critical radius . _ _ from the previous theorem we know that @xmath548 .
we expect @xmath547 to be a critical value for maximizing the radius function .
based on theorems [ th55 ] and [ thm:58 ] about deformation and permutability of labeled @xmath9 configurations , it is natural to propose the following statement .
( connectedness conjecture)[conj95 ] the configuration space @xmath353 $ ] is connected .
that is , every set of @xmath0 distinct labeled points on the @xmath36-sphere pairwise separated by spherical angle at least @xmath182 can be deformed into @xmath0 other distinct labeled points , with all points maintaining a spherical angle at least @xmath182 apart during the deformation .
this problem appears to be approachable but difficult to prove , despite the supporting evidence of permutability in theorem [ thm:58 ] .
one may approach it by cutting the space @xmath549 $ ] into many small path - connected convex
pieces and gluing them together in some fashion .
the computational size of the problem , since the dimension of the space is @xmath19 , and has a complicated boundary , is daunting .
we also propose a stronger statement .
( strong connectedness conjecture)[conj95a ] the radius @xmath22 is the largest radius value at which configuration space @xmath356 $ ] is connected . in support of conjecture [ conj95a
] , the @xmath398-move appears to be possible only when @xmath550 . at one time instant it has @xmath314 spheres fitting in a ring around the equator , a condition which is allowed only for @xmath380 .
we also know that the @xmath22 satisfies the necessary condition of being a critical value for the @xmath2-parameter .
we formulate one further conjecture concerning the connectivity structure of @xmath23 $ ] .
it is based on a further analysis not included here ( see @xcite ) , which indicates that @xmath23 $ ] has at each of the @xmath551 @xmath24-configurations , and at each of the @xmath552 @xmath25-configurations , a unique tangent line along which it can be approached from the interior @xmath481 $ ] .
in addition , this analysis shows that each of these is a _ local cut point _
, a point of a space which when removed , disconnects a small open neighborhood of the point .
that is , these configurations are points at which the space @xmath26 $ ] locally disconnects as @xmath2 increases past @xmath8 .
the following conjecture asserts that these local cut points form unavoidable bottlenecks in @xmath23 $ ] in making certain rearrangements of spheres in the configuration space .
( @xmath24 and @xmath25 bottlenecks)[conj95b ] any piecewise smooth continuous curve in the reduced configuration space @xmath23 $ ] which starts at a labeled @xmath9 configuration and ends at another labeled @xmath9 configuration with the labels permuted by an odd permutation must necessarily pass through either an @xmath24 configuration or a @xmath25 configuration .
this conjecture asserts a specific way in which the @xmath24 and @xmath25 configurations may play a remarkable role in rearrangements of @xmath0-configurations , illuminating the assertion of frank ( 1952 ) in section [ sec:27 ] . for the region
@xmath553 we propose the following conjecture .
[ conj:66 ] ( two connected components ) let @xmath547 be the upper critical radius defined after theorem [ thm:62 ] .
then the space @xmath356 $ ] for @xmath554 has exactly two connected components .
two labeled configurations , @xmath9 and @xmath555 , where @xmath556 is a permutation of twelve labels , belong to different connected components of @xmath356 $ ] if and only if the permutation @xmath557 is odd . in view of the existence of the @xmath482-move , the argument in theorem [ thm:58 ] indicates that there can be at most @xmath36 connected components containing @xmath9 configurations in this region . conjecture [ conj:66 ] asserts there are exactly these two , and no other , connected components .
we next note that the five `` bottlenecks '' in the @xmath27-move lead to the possibility of connected components not containing any @xmath9 configuration , for certain ranges of @xmath2 . during the @xmath483 move joining two @xmath9 configurations , there are @xmath27 bottlenecks all through which one can pass at least up to a radius @xmath558 .
there is however room for configurations of spheres of larger radius occurring between the bottlenecks .
if we increase the radius above the smallest two of the bottleneck radii , it may be possible for a sphere to get stuck in the middle of one of these regions , so it can neither go backwards nor forwards via the @xmath482-move to a @xmath9 configuration . assuming that `` trapped '' configurations ( from blocking of the @xmath483 move ) exist containing no @xmath9 configuration , eventually as we increase @xmath2 some `` trapped '' configuration must become a critical configuration
it would then be a local maximum in the configuration space , an isolated point in some @xmath26 $ ] .
the critical value at which this occurs would necessarily be strictly smaller than @xmath106 , using the result of danzer ( in theorem [ th73 ] ) that the extremal configuration for @xmath119 is unique .
[ conj:65 ] ( non-@xmath9 components ) there is a nonempty interval of values of @xmath41 such that the reduced configuration space @xmath26 $ ] has connected components that do not contain any copy of a @xmath9 configuration .
a positive answer to this question raises the possibility of a value of @xmath2 for which the number of connected components of @xmath26 $ ] exceeds the number of labeled @xmath9 configurations , which is @xmath559 . to obtain the latter number ,
let us take the @xmath9 configuration and label its @xmath0 balls .
there are @xmath560 such labelings .
two labeled configurations are equivalent ( i.e. the same in @xmath26 $ ] ) iff one can be obtained from the other by the @xmath16 rotation action .
clearly , there are @xmath561 labelings in every equivalence class .
finally , for the region of @xmath2-values very close to @xmath106 , we assert that each of the spaces @xmath356 $ ] and @xmath26 $ ] has exactly @xmath562 connected components , with each component containing a @xmath9 configuration . this fact follows assuming the finiteness of the set of critical radius values @xmath2 for @xmath369 , since at the point @xmath106 only the @xmath9 configurations survive , according to the uniqueness result of danzer ( theorem [ th73 ] ) , and the topology of @xmath26 $ ] does not change above the next largest critical value of @xmath2 below @xmath106 . ' '' ''
this paper treats configuration spaces of touching spheres for very small values of @xmath3 .
we have shown that the configuration space of @xmath0 equal spheres touching a central @xmath6-th sphere is already large enough to exhibit interesting behavior in its critical points .
concerning @xmath0-sphere configurations in the equal radius case @xmath22 we have made the following observations .
* we have clarified an assertion of frank ( 1952 ) given in section [ sec:27 ] , showing that in the space @xmath23 $ ] there are deformations interconnecting all @xmath24 , @xmath25 and @xmath9 configurations .
* we have given evidence suggesting that @xmath23 $ ] is a connected space , and conjectured that @xmath22 is the largest parameter value where @xmath26 $ ] is connected .
* we have shown that all elements of the finite set of @xmath24 and @xmath25 configurations lie on the boundary of the topological space @xmath23 $ ] and are critical points for maximizing the radius parameter .
* we have conjectured that when a deformation of 12 spheres in a @xmath9 configuration yields an odd permutation of elements , the ( finite set of ) @xmath24 and @xmath25 configurations in @xmath23 $ ] are `` unavoidable '' points .
many challenging and computationally difficult problems remain to better understand the constrained configuration space @xmath23 $ ] . as mentioned in the introduction ,
configuration spaces are of interest in physics and materials science , particularly in connection with jamming in materials .
hard sphere models which view spheres packed inside a box have been extensively studied for jamming .
materials scientists have studied configuration spaces of small numbers of hard spheres by simulation in connection with nanomaterials .
recently , holmes - cerfon @xcite developed an algorithm that enumerates rigid sphere clusters and has determined those with up to @xmath563 spheres .
the cases of small numbers ( but larger than the @xmath3 treated here ) of spheres were studied in phillips et al .
@xcite and glotzer et al .
@xcite , giving estimates for extremal configurations at values of @xmath3 larger than can be currently treated mathematically .
we note that simulations of phase space can sample only a small part of it . in the simulation experiments reported in @xcite for @xmath119 equal spheres ,
the experimenters were unable to detect that the radii at which the @xmath482-move and the @xmath379-move permutation cease being feasible are in fact different ( as discussed in section [ sec:64 ] ) .
study of the jamming problem leads the sub - problem concerning what is a good notion of rigidity for such configurations .
there is a notion of `` locally jammed configuration '' in which no particle can move if its neighbors are fixed .
the tammes problem or ( extremal ) spherical codes problem , of determining @xmath5 is analogous to determining maximally dense jammed configurations of spheres in a box .
various notions of rigidity for spherical codes were formulated in tarnai and gspr @xcite .
more recently cohn et al .
@xcite give a mathematical treatment of rigidity of extremal @xmath564-dimensional spherical codes .
in configuration theory models like @xmath17 $ ] of this paper , certain critical configurations at critical values of the radius parameter might serve as a proxy for jammed configurations , with the balancing condition in theorem [ thm : converse ] capturing the locally jammed condition .
perhaps only a subclass of critical configurations should be interpreted as jammed , for example those that are local maxima of the radius function .
to unlock the @xmath24 configuration , a good way is to do it with the help of a friend , hereafter called charles .
please follow these steps : 1 .
ask charles to hold the @xmath7 north balls and the @xmath7 south balls firmly in their positions .
these @xmath314 polar balls remain fixed during the whole process . as a result , the @xmath6-th central ball stays fixed as well .
2 . roll the remaining @xmath314 equatorial balls in a direction roughly parallel to the equator .
if properly lubricated , this does not require a big effort .
the equatorial balls can all be pushed either to the east or to the west , in a coordinated way .
4 . at all times
you must ensure the @xmath314 rolling balls touch the central ball .
this requires some practice , but it is possible and not terribly hard .
observe that the @xmath314 balls roll around the central ball along the equatorial `` valley '' between the polar balls kept fixed by charles .
these rolling balls can not always move equatorially , but instead move north and south slightly , in an alternating manner , as you roll them .
6 . because the @xmath314 rolling balls move north and south , some of them do not touch each other any more : free space may appear between them . also , some space can be created between them and the @xmath314 balls kept fixed by charles
this is normal .
as you proceed by @xmath565 the @xmath314 rolling balls realign in the equatorial plane , touching each other and the polar balls .
note that at this moment the configuration is locked back into @xmath24 .
each of the @xmath314 rolled balls is touching two of its equatorial neighbors , one ball to the north and one ball to the south .
unlocking the @xmath25 configuration is similar to the @xmath24 configuration , except that charles has somewhat more to do
. please follow these steps : 1 .
ask charles to hold firmly the three north balls and the three south balls .
the three south balls will remain fixed during the whole process .
but the north triangle has to be rotated as a whole in its plane , at some constant speed , which can be either eastward or westward ( there are two choices ) .
it will move through an angle @xmath566 .
the @xmath6-th central ball stays fixed as before .
roll all the remaining @xmath314 balls in the ( roughly same ) equatorial direction as the north triangle is rotating .
this movement direction is forced on all six equatorial balls by the motion of the north triangle .
the rest of the process goes basically in the same way as for the @xmath24 configuration .
as charles proceeds to rotate the north triangle by @xmath184 , you proceed by @xmath565 the six middle balls align back into the equatorial plane , touching each other and the six polar balls .
note that at this moment the configuration is locked back into the @xmath25 configuration .
each of the @xmath314 rolled balls is touching two of its equatorial neighbors , one ball to the north and one ball to the south .
note that for @xmath24 , the equatorial balls underwent cyclic permutation of length @xmath314 . for @xmath25 , the equatorial balls underwent a cyclic permutation of length @xmath314 and the @xmath7 northern balls a cyclic permutation of length @xmath7 .
these give odd permutations of @xmath24 and @xmath25 .
the authors were each supported by icerm in the spring 2015 program on `` phase transitions and emergent properties . ''
r. k. was also supported by the university of pennsylvania mathematics department sabbatical visitor fund and by msri via nsf grant dms-1440140 .
w. k. was also supported by austrian science fund ( fwf ) project 5503 .
j. l. was supported by nsf grant dms-1401224 and by a clay senior fellowship at icerm .
part of the work of s. s. has been carried out in the framework of the labex archimede ( anr-11-labx-0033 ) and of the a*midex project ( anr-11-idex-0001 - 02 ) , funded by the `` investissements davenir '' french government programme managed by the french national research agency ( anr ) .
part of the work of s. s. has been carried out at iitp ras . the support of russian foundation for sciences ( project no .
14 - 50 - 00150 ) is gratefully acknowledged .
the authors thank bob connelly , sharon glotzer , mark goresky , tom hales and oleg musin for helpful comments .
parts of section [ sec:41 ] are adapted from unpublished notes by r. k. and john sullivan ( msri , 1994 ) about critical configurations of `` electrons '' on the sphere . c. bender , _ bestimmung der grssten anzahl gleich grosser kugeln , welche sich auf eine kugel von demselben radius , wie die brigen , auflegen lassen _ , acrhiv der mathematik und physik * 56 * ( 1874 ) , 302306 .
l. danzer , _ finite point - sets on @xmath569 with minimum distance as large as possible _ , discrete math .
* 60*(1986 ) , 366 .
[ english translation of danzer habilitationsschrift , with extra references added . ]
l. fejes tth , _ ber die abschtzung des krzesten abstandes zweier punkte eines auf einer kugelfches liegenden punktsystems _ , jber .
* 53 * ( 1943 ) , 6668 .
l. fejes tth , _ ber die dichteste kugellagerung _ , math . z.
* 48 * ( 1943 ) , 676684 .
t. hales , _ the strong dodecahedral conjecture and fejes tth s conjecture on sphere packings with kissing number twelve _ , pp .
121132 in : _ discrete geometry and optimization _ , ( k. bezdek , a. deza and y. ye , eds . ) fields institute communications * 69 * : fields institute , toronto , on 2013 .
t. hales , et .
m. adams , g. bauer , dat tat dang , t. harrison , truong le hoang , c. kaliszk , v. magron , s. mclaughlin , thang tat nguyen , truong quang nguyen , t. nipkow , s. obua , j. pleso , j. rute , a. solovyev , hoai thi ta , trung nam tran , diep thi trieu , j. urban , ky khac vu , r. zumkeller , _ a formal proof of the kepler conjecture _ , arxiv:1501.02155 .
r. hoppe , _ bemerkung der redaktion _ , archiv der mathematik und physik ( grunert ) * 56 * ( 1874 ) , 307312 . m. a. hoskin , _ newton , providence and the universe of stars _ ,
journal for the history of astronomy ( jha ) * 8 * ( 1977 ) , 77101 .
j. kepler , _ epitome astronomiae copernicae , usitat form quaestionum & responsionum conscripta , inq ; vii .
libros digesta , quorum tres hi priores sunt de doctrina sphaeric _ , lentijs ad danubium , excudebat johannes plancus , mdcxviii .
r. kusner , w. kusner , j. c. lagarias , and s. shlosman , max - min morse theory for configurations on the @xmath36-sphere , paper in preparation . _
the kepler conjecture : the hales - ferguson proof , by thomas c. hales , samuel p. ferguson _
( j. c. lagarias , ed . ) , springer - verlag : new york 2011 .
b. lubachevsky and r. l. graham , _ dense packings of @xmath571 equal disks in a circle for @xmath572 and @xmath27 _ , pp .
302311 in : ( ding - zhu du and ming li , eds . ) computing and combinatorics , first annual conference , cocoon 95 , lecture notes in comp .
, vol . 959 .
springer : new york 1995 .
i. newton , _ the correspondence of isaac newton _ ( 9 volumes , h. w. turnbull , f. r. s. , ed . ) , cambridge university press 1961 .
l. pauling , _ the structure and entropy of ice and other crystals with some randomness of atomic arrangement _ , j. amer .
chemical soc .
* 57 * ( 1935 ) , 26802684 . c. l. phillips , e. jankowski , m. marval and s. c. glotzer , _ self - assembled clusters of spheres related to spherical codes _ , phys .
e * 86 * , ( 2012 ) , 041124 .
15 october 2012 . c. l. phillips , e. jankowski , b. j. krishnatreya , k. v. edmond , s. sacanna , d. g. grier , d. j. pine and s. c. glotzer , _ digital colloids : reconfigurable clusters as high information density elements _ , soft matter * 10 * ( 2014 ) , 74687479 . | the problem of @xmath0 spheres is to understand , as a function of @xmath1 $ ] , the configuration space of @xmath0 non - overlapping equal spheres of radius @xmath2 touching a central unit sphere .
it considers to what extent , and in what fashion , touching spheres can be moved around on the unit sphere , subject to the constraint of always touching the central sphere .
such constrained motion problems are of interest in physics and materials science , and the problem involves topology and geometry .
this paper reviews the history of work on this problem , presents some new results , and formulates some conjectures .
it also addresses results on configuration spaces of @xmath3 spheres of radius @xmath2 touching a central unit sphere , for @xmath4 .
the problem of determining the maximal radius @xmath5 is equivalent to the tammes problem , to which lszl fejes tth made significant contributions . |
over the past decade , observations have demonstrated that strong ly@xmath0 forest absorbers ( lyfas ) generally show associated metal line absorption ( meyer & york 1987 ; lu 1991 ; cowie et al .
1995 ; womble et al . 1995 ; songaila & cowie 1996 , hereafter sc ) .
sc find civ absorption in nearly all lyfas at @xmath2 with neutral column densities @xmath10 and in @xmath11 of systems with @xmath12 .
the observed civ column densities are consistent with the absorbers having a mean metallicity @xmath13 } \sim -2.5 $ ] and an intrinsic scatter in metallicity of about an order of magnitude ( rauch et al .
1996 ; hellsten et al .
an important question is whether or not a chemically pristine or extremely metal poor population of lyfas exists , and if so , what the characteristic hi column densities of this population are . for @xmath14
the associated civ lines are close to the detection limits of sc , and it can not yet be determined if these absorbers have the same metallicity distribution as those with higher @xmath1 .
some theoretical models of early metal enrichment predict that the mean metallicity declines steadily with decreasing @xmath1 ( gnedin & ostriker 1997 ) .
considerable progress has been made in the theoretical understanding of lyfas . from the work of several groups ( e.g. cen et al .
1994 ; zhang et al . 1995 ; petitjean et al .
1995 ; hernquist et al .
1996 ; bi & davidsen 1997 ) a picture has emerged in which the lyfas are interpreted in terms of `` gunn - peterson '' absorption in an inhomogenous igm pervaded by a background ionizing radiation field , presumed to originate from qsos and perhaps young , star - forming galaxies .
this picture links the ly@xmath0 forest directly to cosmological structure formation .
the analysis of cosmological models of the lyfas consists of producing artificial spectra by evaluating the absorption properties of the baryonic igm along lines of sight through n - body and hydrodynamical simulations of structure formation .
these spectra can then be analyzed in much the same way as observed qso spectra , and the resulting distribution functions in @xmath1 and linewidths are found to agree quantitatively with those observed , lending considerable support to the cosmological lyfa picture ( e.g. miralda - escud et al .
1996 ; dav et al . 1997 ; zhang et al .
1997 ) .
the interpretation of the metal line data within the framework of these models provides a strong test of this picture .
if a metal enrichment pattern of the baryonic igm is specified , the properties of selected absorption lines are readily calculated from the knowledge of densities , temperatures , and uv radiation field along the lines of sight ( haehnelt et al . 1996 ; rauch et al .
1996 ; hellsten et al .
1997 ) . until now
, the metal lines considered have been limited to ovi(1032,1038 ) , nv(1239,1243 ) , civ(1548,1551 ) , siiv(1394,1403 ) , and cii(1335 ) .
these lines have been chosen because they have already been identified with lyfas in real qso spectra . with the exception of the ovi doublet they all have rest wavelengths @xmath15 , which means that for lyfas at sufficiently high redshifts , they appear redward of the ly@xmath0 forest and hence are relatively easy to detect . instead of making an a priori selection of a few metal lines to include in a cosmological model for lyfas , a more satisfactory approach would be to make the models _ predict _ which metal lines , out of thousands of candidates , should be observable ( in spectra of a given resolution and s / n ) in lyfas of different column densities .
such an approach would allow a comprehensive and swift selection of lines that deserve a more detailed treatment , depending on the specific purpose of the modeling .
it would provide a sharper test of the model and place more stringent constraints on the metallicities and uv radiation spectrum , by making it possible to compare the complete list of metal lines predicted to be observable to the list of lines actually detected in qso spectra , as well as the observed relative strengths of these lines .
this paper presents and applies a technique for implementing this comprehensive , a priori approach to metal line absorption predictions in cosmological models . in section 2
we define a line observability index , which can easily be evaluated for a database of metal lines , within a particular model for lyfas .
this index is a measure of column density and hence allows the lines to be sorted by strength .
it also predicts which lines should be detectable in lyfas for a wide range of column densities and metallicities . in section 3 we apply this technique to two specific cosmological models of the ly@xmath0 forest , based on hydrodynamic simulations of structure formation . in section 4 we discuss the dependence of the lox on model parameters such as the mean baryon density and radiation field and estimate how it changes with redshift , and we argue that the lox has only a very weak dependence on the underlying cosmology .
we discuss the results and summarize the conclusions in section 5 .
detection limits for absorption lines can be conveniently expressed in terms of the equivalent width @xmath16 where @xmath17 denotes the received energy flux per wavelength interval , normalized to the continuum level . for absorption lines that are weak enough to be optically thin , the rest frame equivalent width
@xmath18 is proportional to the column density @xmath19 of absorbing atoms along the line of sight @xmath20 where @xmath21 and @xmath22 are the oscillator strength and rest frame wavelength of the line considered .
the other symbols have their usual meaning .
lines that satisfy equation ( [ lincog ] ) are said to be on the linear part of their curve - of - growth .
( for a derivation of this relation see , e.g. , spitzer 1978 . )
we will label an absorption line produced by an element @xmath23 in ionization stage @xmath24 with rest transition wavelength @xmath25 and oscillator strength @xmath21 as @xmath26 . for a metal absorption line of metallicity @xmath27 \
, \equiv \ , \log{(n_z / n_{\rm h } ) } \ , - \ , \log{{(n_z / n_{\rm h})}_{\odot}}$ ] associated with a lyfa of hydrogen column density @xmath1 we now define the following quantity : @xmath28,n_{\rm hi } ) \ ,
\equiv \ , -17.05 \ ,
+ \ , \log{n_{\rm hi } } \ , + \ , \left[\frac{z}{\rm h } \right ] \ , + \log{(f\ , { \lambda}^2 ) } \ , + \log{{\left(\frac{n_z}{n_{\rm h}}\right)}_{\odot } } \ , + \ , \log{\left(\frac{x_z^i}{x_h^0}\right ) } , \label{lox}\ ] ] which we will call the _ line observability index_. the inspiration for this definition comes from rewriting equation ( [ lincog ] ) with the column density of the absorbing metal ions expressed in terms of @xmath1 , [ z / h ] , and the solar ratio of the number densities of metal and h atoms @xmath29 .
the model - dependent ionization corrections are contained in the last term , which is a mean value of the ratio of the ionization fractions @xmath30 of @xmath31 and @xmath32 in the region where the ly@xmath0 absorption arises .
equation ( [ lox ] ) assumes units of @xmath33 for @xmath1 and for @xmath25 ; the choice of additive constant then implies @xmath34 for weak lines .
the lox can thus be used to rank metal lines in terms of strength , and it can be directly compared to the detection limit in spectra of a given quality .
computation of the lox as a function of @xmath35 $ ] and @xmath1 requires a model of the densities and temperatures within ly@xmath0 forest absorbers , which together with the ionizing background radiation field @xmath36 determine the ionization term @xmath37 as a function of @xmath1 .
we have applied the lox technique to two different cosmological simulations .
one is the standard cdm simulation described in dav et al .
a box of length 22.222 comoving mpc containing baryonic and dark matter is evolved from @xmath38 to @xmath39 using treesph ( hernquist & katz 1989 ) , assuming a cdm spectrum of density fluctuations with the parameters @xmath40 , @xmath41 , @xmath42 , @xmath43 , and @xmath44 , where @xmath45km / s / mpc and @xmath46 is the slope of the initial perturbation spectrum .
the other simulation , a flat , nonzero-@xmath47 model , also assumes a cdm spectrum of density fluctuations , with @xmath48 , @xmath49 , @xmath50 , @xmath51 , @xmath43 , and @xmath52 .
the standard model is inconsistent with the amplitude of cmb anisotropy measured by cobe ( smoot et al .
1992 ) , but the normalization of the @xmath47 model ( computed using the cmbfast program of seljak & zaldarriaga 1996 ) is matched to the 4-year cobe data .
for further details regarding the simulation method , see katz , weinberg , & hernquist ( 1996a ) .
the analysis of the two cosmological models yielded very similar results .
the lox is a local property , depending on the conditions within individual density peaks along the line of sight , which are insensitive to the underlying cosmology .
hence , in the following we will only describe the calculations for one of the models , the @xmath47cdm model , and we will focus on results at @xmath53 . a spatially uniform photoionizing background radiation field is imposed , with the shape computed by haardt & madau ( 1997 , in preparation ) .
they calculate the modification due to absorption and re - emission by the igm of an input radiation field with @xmath54 for @xmath55 , @xmath56 for @xmath57 , and @xmath58 for @xmath59 .
this radiation field , at @xmath53 , has a larger drop above the @xmath60 absorption edge at 4 rd than the previously published haardt & madau spectrum ( 1996 ) , by roughly a factor of three , and it is therefore in better agreement with the relatively large number of siiv systems observed at that redshift ( see the discussion in hellsten et al .
1997 ) . to evaluate the ionization correction term in expression ( [ lox ] ) we need to know the typical volume densities and temperatures in the absorbers for different values of @xmath1 . to find the relation between these quantities , we generated artificial ly@xmath0 absorption spectra along 480 lines of sight through the simulation box at @xmath53 , using the method described in hellsten et al .
the lines of sight were selected to include a relatively large number of the high - column density absorbers in order to get good statistics for these rather uncommon systems .
we use artificial spectra along random lines of sight to fix the overall amplitude of the background radiation field , scaling @xmath36 by a constant factor so that the mean ly@xmath0 flux decrement in the simulated spectra matches the value @xmath61 found by press , rybicki , & schneider ( 1993 ; see further discussion by rauch et al .
1997 ) .
figure [ figcolnh ] shows results from an analysis of 1456 hi absorption features detected in these artificial spectra .
lines are identified by a threshold criterion in the following way .
first , we define the velocity interval from a position where the hi optical depth @xmath62 increases above a critical value @xmath63 until it drops below that value again as belonging to an hi absorption feature .
next , we augment this interval on both sides until either @xmath64 or a local maximum in @xmath62 is encountered .
this procedure separates weakly blended `` w - shaped '' absorption features , if @xmath65 at the central local minimum .
the hi column density of the absorber is found by integration of @xmath62 over the velocity interval , i.e. @xmath66 where @xmath67 is measured in km / s in the rightmost expression .
the crucial question for our purposes is the relation between @xmath1 and the typical volume density of gas in the absorbing region . because of peculiar motions and doppler broadening , the absorbing hi atoms in a given velocity bin originate at slightly different spatial locations .
the code we developed to generate the artificial spectra also determines the _ mean _ total hydrogen volume density @xmath68 in the regions from which the hi atoms originate . in figure [ figcolnh ]
we plot @xmath69 vs. @xmath1 , where @xmath69 is the maximum value of @xmath68 in the velocity interval @xmath70 of that absorption feature . in the spectra ,
the location of this density peak is close to the location of the maximum of @xmath71 , as one would expect . as can be seen in figure [ figcolnh ]
, there is a well - defined correlation between the hi column density of a lyfa and the typical total volume hydrogen density of the absorbing region .
the thick solid line is the power - law fit @xmath72 which we will use to evaluate the lox .
relation ( [ nnhfit ] ) depends on the value of @xmath73 , i.e. , the total mean baryonic density .
increasing the value of @xmath73 will shift the curve upward if @xmath74 is held constant ( to satisfy the mean flux decrement constraint ) by adjusting the intensity of the uv radiation field .
the indicated scaling with @xmath73 in ( [ nnhfit ] ) is only approximate , because it neglects effects of the temperature dependence on @xmath73 ( croft et .
al . 1997 ) .
we will assume @xmath75 throughout this paper , except for the discussion in section 4 .
while a similar correlation between @xmath69 and @xmath1 holds at other redshifts , the constant of proportionality is different because of the evolution in the radiation field and the igm structure , in particular the dilution of the cosmic mean density by the expansion of the universe .
the @xmath69-@xmath1 correlation for the scdm model is very similar , with a slope of about 0.73 instead of 0.7 , and the resulting line observability indices are nearly identical for the two relations . also shown in figure [ figcolnh ] is the mean total hydrogen density at @xmath53 in this model .
it is seen that the model predicts that lyfas with @xmath76 at @xmath53 arise from underdense regions ( similar to the result of zhang et al .
1997 ) .
the relation between density and temperature is slightly more complex .
figure [ figtnh ] shows a plot of @xmath69 versus mean temperature for the same density peaks as those shown in figure [ figcolnh ] . for low densities a large fraction of the peaks
follow the relation @xmath77 . at densities
@xmath78 , the density - temperature correlation arises mainly from the interplay between photoionization heating and adiabiatic cooling due to the expansion of the universe .
low density regions have lower neutral fractions , hence lower photoionization heating rates , and they expand more rapidly , thus experiencing stronger adiabatic cooling .
some of the peaks lie above this temperature - density relation , at @xmath79 , because of moderate shock heating .
peaks with @xmath80 have large enough densities for radiative cooling to become significant , and the temperature is roughly determined by the equilibrium between radiative cooling and photoionization heating .
( for a more detailed discussion of the density - temperature relation in cosmological simulations , see , e.g. , croft et al . 1997 ; hui & gnedin 1997 ; zhang et al .
1997 ) . for our lox calculations we will adopt the relation defined by the solid line in figure [ figtnh ] .
this relation is again very similar to that in the scdm model , although the latter shows a slightly larger scatter at lower densities .
because photoionization is the main source of ionization at such low densities , the lox is insensitive to the exact value of the temperature . for an arbitrary metal line @xmath81 associated with a lyfa with neutral column density @xmath1 and metallicity [ z / h ] , we can now evaluate the ionization correction term in ( [ lox ] ) using the photoionization code cloudy 90 ( ferland 1996 ) .
the inputs to cloudy are the hydrogen volume density and the temperatures from the relations discussed above , as well as the uv radiation field . for systems with @xmath82 , continuum heii absorption affects the radiation field in the central regions of the absorbers .
we include this effect in an approximate way , by solving self - consistently for the radiative transfer through a plane parallel slab with hi column density @xmath83 , illuminated from both sides .
we then use the volume averaged values of the ionization fractions , which should roughly be equal to those in the central region of an absorber with hi column density @xmath1 . including this heii - shielding effect only changes the predicted lox values slightly .
the results in table 1 are produced by evaluating the lox for the 2249 transitions in the compilation by verner , barthel , and tytler ( 1994 ) and then sorting the lines according to the observability index .
the strongest lines are shown for five different values of the hi column density , @xmath84 , 14 , 15 , 16 , and 17 .
we show , for @xmath85 , only the ten strongest lines with @xmath86 .
these extreme ultraviolet lines are numerous , but in spectra of the faint qsos at @xmath2 they are extremely difficult to detect , even with state of the art instruments such as the space telecope imaging spectrograph , installed on the hubble space telescope ( hst ) in february 1997 ( e.g. danks et al . 1996 ) . at somewhat lower redshifts ,
the prospects should be better for observing these lines in stis spectra of a few very bright qsos , such as he 2347 + 4342 ( @xmath87=2.88 , reimers et al .
1997 ) , but the observations have yet to be made . with the cosmic origins spectrograph ( cos ) ,
scheduled to be installed on the hst in 2002 ( green et al .
1997 ) , detection of these important lines may finally become possible for a larger number of qsos at high redshift . a metallicity of @xmath88}=-2.5 $ ] has been assumed in table 1 , and we use the solar metallicity values from grevesse & anders ( 1989 ) .
we evaluate the lox for solar _ relative _ abundances , i.e. [ z / o]=0 .
it can then be scaled to any other relative abundance pattern a posteriori .
for example , if [ o / c]=0.5 , as observed in some low metallicity systems ( e.g. wheeler et al . 1989 ) , we have to add 0.5 to the lox values for o lines , ( i.e. use [ o / h]=-2 ) in order to compare them to the c lines for [ c / h]=-2.5 . before discussing the results in table 1 ,
let us estimate detection limits in present day , high signal - to - noise ratio ( s / n ) hires spectra .
for a line with rest frame wavelength @xmath89 , the rest equivalent width threshold for an @xmath90-sigma detection is @xmath91 where @xmath92 is the number of pixels used for the equivalent width detection limit and @xmath93 ( @xmath94 ) is the velocity ( wavelength ) spread per pixel ( churchill 1997 ) .
for the spectrum of q1422 + 231 obtained by sc , @xmath95 and @xmath96 . hence lines with @xmath97 should be detectable at the 5-sigma level in that spectrum .
( recall from 2 that @xmath98 for weak lines . )
the s / n per pixel is roughly 110 outside the ly@xmath0 forest and 45 within the forest , which translates into a minimum detectable lox of approximately 0.26 outside the forest and 0.65 within the forest .
the integration time for this spectrum was 8.2 hrs .
if the qso is observed for 3 - 4 nights the s / n ratio could be roughly doubled , so a practical absolute lower limit for observable lines would be lox @xmath99 . looking at table 1 , the ovi(1032,1038 ) doublet is predicted to be a dominant metal line feature associated with lyfas with @xmath84 , 14 , and 15 . for [ o / h]=-2 , ovi(1032 ) becomes observable for @xmath100 and tops the list for more than two orders of magnitude in @xmath1 . because this ovi doublet has rest wavelengths less than 1216 ,
the absorption features will always be located in the ly@xmath0 forest where they are subject to heavy blending and blanketing by hi lines , making them difficult to identify , especially in high redshift spectra .
a family of strong absorption lines from lower ionization states of oxygen is seen to be present for the higher column density lyfas .
these lines all have @xmath101 , and can not ( for @xmath102 ) be observed in the optical .
they present a challenge for the next generation of ultraviolet spectrographs . for lines with @xmath9 , the civ(1548,1551 )
doublet is seen to be dominant over the entire range of hi column densities , a remarkable result , given that this range in @xmath1 spans three orders of magnitude in volume density and hence photoionization parameter . only for hi column densities close to that of lyman limit systems ( @xmath103 )
do other metal lines , such as the siiv(1394,1403 ) doublet , attain strengths comparable to that of civ .
this prediction is consistent with the fact that the first metal lines found to be associated with lyfas were civ lines , and that civ systems are the best studied and most numerously observed metal line systems to date .
the lines with @xmath104 reaching the largest lox values are ciii(977 ) and siiii(1206 ) . in the high column density
lyfas these lines are predicted to be so strong that the prospects of detecting them in the ly@xmath0 forest should be reasonable , and they have indeed been observed in a few lyfas in the q1422 spectrum ( sc 96 ) .
these lines deserve a more thorough investigation with artificial absorption spectra . a good way of testing simulation predictions for high column density lyfas would be to compare ciii absorption in real spectra to the ciii features in a model with [ c / h ] determined observationally from the corresponding civ features . figures 3 and 4 show graphical representations of the @xmath1-lox relations for 12 lines selected from table 1 . in low density regions , the lox is insensitive to the scatter in gas temperature , because photoionization is much more important than collisional ionization and because the recombination coefficients are not very temperature sensitive .
a more complicated issue is what would happen if the _
mean _ temperature of the low density gas were different in the model , for example because of a different assumption about the heat input during the reionization period ( hui & gnedin 1997 ; rauch et al .
the effect on the lox is small in this case also , as we will argue in detail in section 4 .
figure 3 contains lines from three different ionization states of carbon and silicon .
the civ(1548 ) , ciii(997 ) , and cii(1335 ) lines are expected to be observable in absorbers with @xmath105 , 14.9 , and 15.8 , respectively .
these values are obtained assuming a resolution and s / n similar to that of sc s spectrum of q1422 + 231 , and a metallicity of [ o / h]=-2.5 .
it is straightforward to predict these values for spectra of other resolutions and signal - to - noise ratios by means of equation ( [ declimit ] ) , and for other metallicities by shifting the lox - curves upwards by an amount @xmath106(lox ) = [ o / h]+2.5 ( or alternatively sliding the horizontal detection limit lines downwards by the same amount ) .
for example , for civ to be detectable in a lyfa with @xmath107 in the sc spectrum , one can estimate from figure 3 that [ o / h ] would have to be @xmath108 .
the diagonal arrows on figure 3 indicate how the curves will shift if @xmath109 is multiplied by a factor of 2 ( solid arrow ) or if @xmath110 is changed from 0.0125 to 0.02 and @xmath109 is adjusted to maintain the observed @xmath111 ( dashed arrow ) . the direction and magnitude of these arrows can be derived from approximate scaling laws , as will be discussed in section 4 .
the modeling of the ionization correction term in the lox is based on a single - density model , assuming that most of the absorption takes place at or near the maximum of the local density peak associated with a lyfa .
this is certainly true for hi .
as long as the volume density of a species @xmath31 increases with total density ( i.e. the lox increases with @xmath1 ) this assumption holds . in situations where @xmath31 decreases with total density
this approximation becomes worse .
for example , in our artificial spectra we sometimes see that the cii(1335 ) feature associated with a high column density absorber ( @xmath112 ) is centered on the same velocity as hi , whereas the corresponding civ ( or ovi ) absorption is seen as shoulders that are offset relative to the hi feature .
this occurs because the civ density profile does not follow that of the total density .
therefore , our maximum density assumption causes us to underestimate the lox values on the decreasing parts of the curves in figures 3 - 4 .
they are lower limits to the true lox values , and it is necessary to do a full generation and analysis of artificial spectra if one wants to use , say , observed cii / civ ratios to infer ionization properties of the igm
. direct comparison to the lox values in figure 3 would overestimate the photoionization parameter in the central region of the lyfa .
figure 3 also shows the silicon lines siiv(1394 ) , siiii(1206 ) , and siii(1260 ) . here , the siiv and siiii features are observable only for @xmath113 , and the siii feature for @xmath114 .
in addition , the siii line falls redward of the ly@xmath0 forest only for lyfas close to the ly@xmath0 emission peak . clearly , for most lyfas these silicon lines are not as promising for probing ionization conditions as the carbon lines discussed above .
lines from the family of very strong o transitions oiii(833 ) , oiv(788 ) , ov(630 ) , and ovi(1032 ) , as well as nv(1239 ) and svi(1063 ) , are displayed in figure 4 .
ovi is by far the strongest high ionization line with @xmath104 , and it is observable in the entire interval @xmath115 . for systems with @xmath116 , ovi should be detectable in the sc spectrum , provided @xmath88 { \mathrel{\copy\simgreatbox}}-3}$ ] . in our model , the ovi is produced mainly from photoionization , and the lox is not very sensitive to the assumed temperature . with present day spectra , ovi is the best and only candidate for probing the metallicity of the igm in regions close to the mean density .
as we already mentioned , observations of the ovi lines are hampered by the high line density within the ly@xmath0 forest .
low column density lines in cosmological simulations are usually broadened by hubble flow and peculiar velocities rather than thermal motions , so the ovi lines may not necessarily be distinguishable by low @xmath117-parameters . in individual cases , however , it should at least be possible to place upper limits to the metallicity of the systems that happen not to have strong hi contamination at the expected positions of the ovi features .
the nv(1239,1243 ) is predicted to be weaker than the ovi doublet and probe a somewhat denser part of the igm . if [ n / c ] is as low as -0.7 as suggested by pettini et al .
( 1995 ) the nv doublet will be observable only in lyfas with relatively high metallicity , roughly @xmath118 { \mathrel{\copy\simgreatbox}}-1.8 $ ] .
this seems consistent with the fact that sc observed nv in only one out of six candidate absorbers in their ly@xmath119 selected sample . towards lower redshifts , where the ly@xmath0 forest begins to thin out
, it should become increasingly easy to identify ovi absorption lines by their coincidence with corresponding ly@xmath0 and ly@xmath119 absorption , at least in a statistical way .
though somewhat arduous , this type of investigation with ovi is probably the most promising route to testing the trend between metallicity and hi column density predicted by some models of early igm enrichment ( e.g. , gnedin & ostriker 1997 ) .
figures 3 and 4 present lox results for a single redshift and a single set of model parameters : @xmath47cdm at @xmath53 , with @xmath120 and the haardt - madau ( 1997 ) uv background spectrum scaled in intensity to give an hi photoionization rate @xmath121 such that @xmath61 in the simulated spectra .
however , we can estimate the effects of changing some of these parameters by simple scaling arguments because the dynamics of the gas in the density regime relevant to lyfas is driven by the gravity of the dark matter , making the spatial distribution of the gas insensitive to changes in @xmath122 , @xmath36 , or the igm temperature .
we can therefore compute the effects of such changes on the @xmath123 vs. @xmath69 correlation ( figure [ figcolnh ] ) , and we can compute the effect on lox values through equation ( [ lox ] ) if we make the approximation that the ionization fractions @xmath124 depend only on the ionization parameter @xmath125 . because recombination rates vary only slowly with temperature and the range of temperatures in lyfas is much smaller than the range of densities , this is a good , but not a perfect , approximation .
consider first the effect of changing @xmath36 by a constant factor @xmath126 .
this change in isolation would alter the mean flux decrement @xmath111 , which we have thus far adopted as an observational constraint on the model , but current observational uncertainties in @xmath111 allow some range in @xmath36 ( compare , e.g. , zuo & lu 1993 to press et al . 1993 and rauch et al .
1997 ) , and comparison of simulated and observed hi column density distributions favors a somewhat higher value of @xmath36 than we have adopted here ( dav et al . 1997 ; gnedin 1997a ) .
we assume that the shape of @xmath36 does not change , so that the photoionization rates @xmath127 for all species change by the same factor .
an absorber that previously had column density @xmath128 now has a column density @xmath129 raising @xmath36 lowers the hydrogen neutral fraction and therefore lowers @xmath123 .
the density @xmath69 in this absorber is the same as before , so the ionization parameters @xmath130 are the same as those in an absorber that previously had column density @xmath131 .
we have used the correlation @xmath132 from figure [ figcolnh ] to convert a factor @xmath133 change in @xmath69 to a factor @xmath134 change in @xmath123 . assuming that @xmath135 depends only on @xmath130 , we see from equation ( [ lox ] ) that @xmath136 where the second term accounts for the difference @xmath137 .
for example , the effect of doubling @xmath36 while keeping other parameters fixed is to shift the lox@xmath138 curves diagonally by the vector ( 0.13,0.13 ) , indicated by the solid arrow in figure 3 .
the magnitude of this shift is small , and the direction is nearly parallel to the lox curve itself .
the same shift would apply to figure 4 . if we keep @xmath36 fixed but change @xmath122 by a factor @xmath126 , then absorber densities @xmath69 change by @xmath126 and column densities by @xmath139 ( because the hydrogen neutral fraction is itself proportional to @xmath69 ) .
an absorber that previously had column density @xmath128 now has @xmath140 .
the ionization parameters are the same as those in an absorber that previously had column density @xmath141 with the same reasoning as before , we find @xmath142 since the observational constraints on @xmath111 are better than those on the ionizing background intensity , a factor @xmath126 change in @xmath122 should really be accompanied by a factor @xmath139 change in @xmath36 , so that @xmath111 and the hi column densities of absorbers remain the same .
the net change is to increase @xmath130 by a factor of @xmath126 at fixed @xmath123 .
an absorber with column density @xmath123 has the same ionization parameters as an absorber that previously had column density @xmath143 .
thus , @xmath144 for example , the dashed arrow in figure 3 shows the effect of increasing @xmath122 from the value @xmath145 used in our simulation to the value @xmath146 favored ( for @xmath147 ) by the low [ d / h ] value of tytler et al .
( 1996 ) .
plausible variations in the assumed gas reionization history can alter the temperature of the low density igm by factors @xmath148 ( miralda - escud & rees 1994 ; hui & gnedin 1997 ) . because the hi recombination coefficients are @xmath149 for @xmath150k , a factor @xmath126 change in gas temperatures requires a factor @xmath151 change in @xmath36 to keep @xmath111 fixed . from the scaling relation ( [ scal1 ] ) , we see that the impact of such a temperature change would be small . to estimate the evolution of the lox with redshift
, we can appeal to the argument of hernquist et al .
( 1996 ) and miralda - escud et al .
( 1996 ) that the primary driver of evolution in the ly@xmath0 forest at high redshift is the expansion of the universe , which changes the cosmic mean density in proportion to @xmath152 .
non - linear gravitational evolution changes the overdensity field @xmath153 , but this evolution has a smaller impact on the forest than the overall expansion , at least for low column density absorbers . matching the observed evolution of @xmath111 requires an hi photoionization rate @xmath121 that is roughly constant between @xmath154 and @xmath39 ( hernquist et al .
1996 ; miralda - escud et al .
1996 ; rauch et al .
1997 ) , so if we ignore changes to the shape of the spectrum we can take @xmath36 to be constant over this interval .
a factor of two _ decrease _ in @xmath152 roughly the change between @xmath154 and @xmath53 or between @xmath53 and @xmath39 has the same effect as a factor of two _ increase _ in @xmath36 at a given redshift , illustrated by the solid arrow in figure 3 .
while this argument is only approximate , it is enough to show that the lox should not vary strongly with redshift over the range @xmath155 , except in the case of ionic species that are affected by the change in the spectral shape of @xmath36 .
structure formation models that have a power spectrum amplitude similar to that of @xmath47cdm on mpc scales at @xmath53 will predict similar spatial and density structure in the diffuse igm .
( models with very different power spectra will of course predict different structure , but such models may soon be ruled out by the ly@xmath0 forest data themselves [ see , e.g. , discussions by rauch et al . 1997 ;
gnedin 1997b ; croft et al .
if the baryon density @xmath73 is held fixed , then the most important effect of changing the cosmological model is to alter the expansion rate @xmath156 .
matching @xmath111 requires @xmath157 ( see , e.g. , equation [ 2 ] of rauch et al .
1997 ) , so halving @xmath156 has the same impact as doubling @xmath36 . in practice
the difference between @xmath156 values in currently popular models at @xmath158 is less than a factor of two , so the expected impact is small .
we conclude that our quantitative predictions for the lox are likely to hold rather generally in the cosmological picture of lyfas , at least for column densities @xmath159 , where the scaling arguments used in this section are most secure .
our comparison of two numerical simulation models , a standard cdm model and a @xmath47cdm model , supports this claim , and it does not show any major changes to the lox predictions even for higher column densities . of course , the lox is always directly affected by the metallicity of the absorbing gas ( equation [ [ lox ] ] ) , and the insensitivity to cosmology implies that absorption observations can therefore yield robust estimates of the metallicity in different regions of the igm .
ionization corrections are essential in the interpretation of such measurements , but the corrections are not strongly dependent on the assumed cosmological model .
in this paper we have introduced the lox , a measure that allows a comprehensive selection of the metal lines that are useful for testing the detailed predictions of models for lyfas .
we have applied this method to a particular model , a cosmological @xmath47cdm simulation , and we have quantitatively discussed the dependence of the results on the model parameters @xmath109 , @xmath160 , and @xmath87 . we have also performed this analysis for a different cosmological simulation , a standard cdm model , but the results differ negligibly from those of the @xmath47cdm model .
we find that very few detectable metal lines are associated with the low to moderate density regions of the igm that give rise to absorbers with @xmath161 .
most of the baryons in the universe at @xmath53 are believed to reside in this part of the igm ( e.g. , zhang et al .
1997 ) , and the ovi(1032,1038 ) doublet is found to be the best probe of the metallicity in these regions .
other strong o lines , such as ov(630 ) , oiv(788 ) , and oiii(833 ) are technically difficult to observe , and have yet to be detected .
the ciii(977 ) and siiii(1206 ) lines are the strongest metal lines in high column density lyfas .
the civ(1548 ) line is the strongest metal line with @xmath162 for all lyfas with @xmath163 . while the specific values of the lox as a function of column density will depend on redshift and on the adopted cosmological model and radiation field ,
we have argued that the qualitative conclusions presented above are insensitive to the choice of cosmological model of high redshift lyfas , and they should hold generally within this scenario .
the relatively uncommon absorbers with @xmath82 should offer the strongest test of cosmological simulations .
these systems should have more than a dozen observable metal lines associated with them ( for @xmath88 } \sim -2.5 $ ] ) .
the relative strengths of these lines will depend on the assumed radiation field and abundance pattern , so a first check would be whether it is possible to make the line strengths in the model spectra match those of lines associated with high column density lyfas for reasonable values of model parameters . then one should make a detailed examination of the relative locations in velocity space between absorption components of individual species . a detailed comparison to observations will ultimately determine if the resolution of the simulations is adequate in these high - density regions , and to what extent local dynamical effects associated with the production of metals can be neglected .
there are some hints of discrepancies between models and observations for these high column density systems .
for example , the models in some cases predict the velocity components of civ and cii absorption lines to differ , whereas there seems to be some new observational evidence that cii and civ components are found at the same positions ( a. boksenberg , private communication ) .
cosmological simulations depict the typical low column density lyfas as simple structures : low density , smooth , and governed by the straightforward physics of gravity , cosmic expansion , and photoionization .
studies of absorption along parallel lines of sight provide direct observational support for this point of view , especially recent hires observations of gravitationally lensed qsos that set stringent lower limits to the scale of any substructure within low column density absorbers ( rauch 1997 ; see also smette et al .
1992 , 1995 ) .
the simulations do not predict a sudden change in lyfa properties at any particular column density , but they do predict a steady increase of gas density with @xmath1 ( fig . 1 ) . for @xmath164 ,
the typical absorbers are partially collapsed structures , and gas within them has begun to experience radiative cooling . in this regime ,
the finite mass and spatial resolution of the simulations may limit the accuracy of the results , and local astrophysical processes such as star formation and supernova explosions might influence the structure of these systems .
the standard cdm model studied here produces only about 1/3 of the observed number of lyman limit systems ( @xmath165 ) , even though it matches the abundance of much stronger , damped ly@xmath0 absorbers quite well ( katz et al .
1996b ; gardner et al .
1997 ) . while further work is needed to assess the sensitivity of this result to cosmological parameters
especially the baryon density parameter @xmath122 the discrepancy of numbers suggests that a population of absorbers unresolved by the simulations may become important at column densities @xmath166 .
the high column density lyfas in the cosmological simulations represent the transition between the weak igm fluctuations traced by low column density lines and the cold gas concentrations in virialized dark halos traced by damped ly@xmath0 absorption .
the high - lox lines listed in table 1 can reveal many details of the structure and physical conditions in these absorbers , testing the accuracy of the simulations in this regime , providing clues to the nature of any additional absorber populations , and capturing snapshots of gas as it makes its way from the igm into high redshift galaxies .
we thank the referee for drawing our attention to the importance of the absorption lines in the extreme uv .
uh acknowledges support by a postdoctoral research grant from the danish natural science research council .
this work was supported by the nsf under grant asc93 - 18185 and the presidential faculty fellows program and by nasa under grants nag5 - 3111 , nag5 - 3525 , and nag5 - 3922 .
computing support was provided by the san diego supercomputer center .
bi , h.g . , & davidsen , a. 1997 , , 479 , 523 cen , r. , miralda - escud , j. , ostriker , j.p . , & rauch m. 1994 , , 437 , l9 churchill , c. 1997 , ph.d .
thesis , univ . of california , santa cruz .
cowie , l.l . ,
songaila , a. , kim , t .- s . , &
1995 , aj . , 109 , 1522 croft , r.a.c . ,
weinberg , d.h . , katz , n. , & hernquist , l. 1997 , , in press croft , r.a.c . ,
weinberg , d.h . ,
katz , n. , & hernquist , l. 1998 , , submitted danks , a. , woodgate , b. , kimble , r. , bowers , c. , grady , j. , kraemer , s. , kaiser , m.e . ,
meyer , w. , hood , d. & van houten , c. 1996 , in science with the hst - ii , space telescope science institute , eds .
benvenuti , macchetto & schreier .
dav , r. , hernquist , l. , weinberg , d.h . , & katz , n. 1997 , , 477 , 21 ferland , g.j .
, 1996 , university of kentucky , department of astronomy , internal report gardner , j.p . , katz , n. , hernquist , l. , & weinberg , d.h .
1997 , , in press ( astro - ph/9609072 ) gnedin , n. y. 1997a , , submitted ( astro - ph/9707257 ) gnedin , n. y. 1997b , , submitted ( astro - ph/9706286 ) gnedin , n. y. & ostriker , j. p. , , submitted , astro - ph/9612127 green , j. , shull , j.m . , morse , j. et al .
1997 , proposal for `` cosmic origins spectrograph '' , selected by nasa for hubble space telescope 2002 refurbishment mission .
grevesse , n. & anders , e. 1989 , cosmic abundances of matter , aip conf .
proceedings 183,1 .
gunn , j.e .
& peterson , b. a. 1965 , , 142 , 1633 haardt , f. & madau , p. 1996
, , 461 , 20 haardt , f. & madau , p. 1997
, in prep .
haehnelt , m.g . , steinmetz , m. , & rauch , m. 1996 , , 465 , l65 hernquist , l. & katz , n. 1989 , , 70 , 419 hellsten , u. , dav , r. , hernquist , l. , weinberg , d. & katz , n. 1997 , , vol .
487 hernquist , l. , katz , n. , weinberg , d.h .
, & miralda - escud , j. 1996 , , 457 , l51 hui , l. , gnedin , n. , 1997 , , submitted , astro - ph/9612232 katz , n. , weinberg , d.h . , & hernquist , l. 1996a , apjs , 105 , 19 katz , n. , weinberg d.h . ,
hernquist , l. , & miralda - escud , j. 1996b , , 457 , l57 lu , l. 1991 , , 379 , 99 meyer , d.m . & york , d.g .
1987 , , 315 , l5 miralda - escud , j. , cen , r. , ostriker , j.p . &
rauch , m. 1996 , , 471 , 582 miralda - escud j. , & rees , m. j. 1994 , , 266 , 343 petitjean , p. , mcket , j.p . & kates , r.e .
1995 , a&a , 295 , l9 pettini , m. , lipman , k. , & hunstead , r.w .
1995 , , 451 , 100 press , w.h . ,
rybicki , g.b . , & schneider , d.p .
1993 , , 414 , 64 rauch , m. 1997 , in proc . of the 13th iap colloquium , structure and evolution of the igm from qso absorption line systems , eds .
p. petitjean & s. charlot , ( paris : nouvelles frontires ) , astro - ph/9709129 rauch , m. , haehnelt , m.g . , & steinmetz , m. 1996 , , 481 , 601 .
rauch , m. , miralda - escud , j. , sargent , w.l.w . ,
barlow , t.a . , weinberg , d.h .
, hernquist , l. , katz , n.s .
& ostriker , j.p .
1997 , , in press , astro - ph/9612245 .
reimers , d. , khler , s. , wisotzki , l. , groote , d. , rodriguez - pascual , p. , & wamsteker , w. 1997 , a & a , 327 , 890 .
seljak , u. & zaldarriaga , m. , 1996 , , 469 , 437 smette , a. , surdej , j. , shaver , p. a. , foltz , c. b. , chaffee , f. h. , weymann , r. j. , williams , r. e. , magain , p. 1992
, , 389 , 39 smette , a. , robertson , j. g. , shaver , p a. , reimers , d. , wisotzki , l. , koehler , t. 1995 , a&as , 113 , 199 smoot , g. f. , et al .
1992 , , 396 , l1 songaila , a. & cowie , l. l. 1996 , aj , 112 , 335 [ sc ] spitzer , l. jr . , 1978 , physical processes in the interstellar medium , wiley - interscience .
tytler , d. , fan , x .-
m . , and burles , s. 1996 , nature , 381 , 207 verner , d.a . , barthel , p.d . ,
& tytler , d. 1994 , a&as , 108 , 287 wheeler , j.c . ,
sneden , c. , & truran , j.w .
1989 , ara & a , 27 , 279 womble , d.s . , sargent , w.l.w . , &
lyons , r.s .
1995 , in cold gas at high redshift , eds .
m. bremer et al . , kluwer 1996 , ( astro - ph/9511035 ) .
zhang , y. , anninos , p. , & norman , m.l .
1995 , , 453 , l57 zhang , y. , meiksin , a. , anninos , p. , norman , m.l . , 1997 ,
apj , in press ( astro - ph 9706087 ) .
zuo , l. , & lu , l. 1993 , , 418 , 601 | we develop a prescription for characterizing the strengths of metal lines associated with ly@xmath0 forest absorbers ( lyfas ) of a given neutral hydrogen column density @xmath1 and metallicity [ o / h ] .
this _ line observability index _ ( lox ) is line - specific and translates , for weak lines , into a measure of the equivalent width .
it can be evaluated quickly for thousands of transitions within the framework of a given model of the ly@xmath0 forest , providing a ranking of the absorption lines in terms of their strengths and enabling model builders to select the lines that deserve more detailed consideration , i.e. those that should be detectable in observed spectra of a given resolution and signal - to - noise ratio .
we compute the lox for a large number of elements and transitions in two cosmological models of the ly@xmath0 forest at @xmath2 derived from hydrodynamic simulations of structure formation .
we present results for a cold dark matter universe with a cosmological constant ; an @xmath3 cold dark matter model yields nearly identical results , and we argue more generally that the lox predictions are insensitive to the specific choice of cosmology .
we also discuss how the lox depends on redshift and on model parameters such as the mean baryonic density and radiation field .
we find that the ovi ( 1032 , 1038 ) doublet is the best probe of the metallicity in low column density lyfas @xmath4 .
metallicities down to [ o / h ] @xmath5 -3 yield ovi absorption features that should be detectable in current high - quality spectra , provided that the expected position of the ovi feature is not contaminated by hi absorption .
the strongest transitions in lower ionization states of oxygen are ov(630 ) , oiv(788 ) , and oiii(833 ) .
these absorption lines are all predicted to be stronger than the ovi feature , but even at redshifts @xmath6 they will have to be observed in the ultraviolet , and they are extremely difficult to detect with present uv instruments , such as the space telescope imaging spectrograph ( stis ) . at lower redshifts ,
detection of these lines may be possible in stis spectra of the very brightest qsos , while one may have to wait for next - generation instruments such as the cosmic origins spectrograph ( cos ) to detect such lines in a number of high - redshift qsos .
the strongest metal lines with restframe wavelength larger than @xmath7 associated with higher column density lyfas at @xmath2 are ciii ( 977 ) and siiii ( 1206.5 ) , which peak at @xmath8 . of the lines with rest wavelengths @xmath9 , which can potentially be observed redwards of the ly@xmath0 forest
, the civ(1548,1551 ) doublet is expected to dominate in all lyfas , regardless of the value of @xmath1 .
we argue that civ and cii absorption may peak in different spatial regions , and that comparison of single - phase models of the civ / cii ionization ratios with observed civ / cii column density ratios can lead to an overestimate of the ionization parameter in the central parts of the absorbers .
= = = 1=1=0pt = 2=2=0pt |
the barely covered story of rising foreclosures among the condominiums of florida or california in early 2007 was a harbinger of a much larger collapse in the worldwide financial system .
the increase of foreclosures over the priced in foreclosure risk in mortgage - backed securities , otherwise deemed high - grade assets , began the confusion of the value of collateral assets and subsequent seizing up of credit markets around the globe .
the collapse of several institutions , such as bear stearns , lehman , and fortis , has accentuated the level of crisis now facing the world markets .
previously , loosely regulated titans of finance , such as hedge funds and private equity groups , have been hit by waves of unprecedented losses and demands by investors for redemptions , causing them to sell even more assets or close positions and creating a positive feedback death spiral .
though the hardest hit markets are lesser - known markets , such as commercial paper , the equity markets have become the most widely known indicators of the ongoing meltdown .
in fact , most non - experts likely use the movements of the equity markets , fallaciously , as a key gauge of the severity or progress of the crisis .
the equity markets , however , did not originate the crisis nor are they the key force perpetuating it . in this short paper ,
the spread of the credit crisis will be discussed by referring to a correlation network of stocks in the s&p 500 and the nasdaq-100 indices .
the fact that the spread resembles a contagion or cascade , however , may be mainly superficial given the underlying dynamics are completely different .
in this paper , a stock correlation network , similar to the one in refs .
@xcite , is created .
we start by defining a correlation matrix of returns between two stocks , where the correlation between stocks @xmath0 and @xmath1 , @xmath2 is defined as @xmath3 with @xmath4 and @xmath5 being the log - returns of stocks @xmath0 and @xmath1 at a given time , @xmath6 and @xmath7 being the mean value of the stock log - returns over the measured time period , and @xmath8 and@xmath9 being the standard deviations of @xmath0 and @xmath1 over the measured time period . the correlation is taken over the time period august 1 , 2007 , to october 10 , 2008 , where each daily value of @xmath10 is the log - return of the closing price from the previous day . as refs .
@xcite demonstrate , however , correlation is not a distance metric ; therefore , we create an adjacency matrix with weights on the edges matching the distance metric between stocks , @xmath0 and @xmath1 . that matrix is defined as @xmath11 using these distances
we finally create a graphical minimal spanning tree by using the python - graph module , pydot , and graphviz .
because over 500 stocks are included , the ticker labels are relatively small but the central part of the component is dominated ( though not exclusively ) by certain finance and service sector stocks , which are heavily cross - correlated and thus tightly linked with each other , while the outer branches are more industry specific , including utilities , basic materials , technology , and some less - central financial stocks .
these are the stocks later impacted by the credit crisis ( see fig . [ sector ] ) .
the average correlations among stocks both within each category and between stocks of each category are given in table [ corrtable ] .
, green at distance @xmath12 , blue at distance @xmath13 , purple at distance @xmath14 , and black at distance @xmath15 ( the maximum allowed by the metric in eq .
[ distanceeq]).,width=491 ] the stocks in fig .
[ spread ] , represented as nodes , are colored according to the following methodology based on the stock return since august 1 , 2007 .
events in the figures are taken from the timeline at ref .
the fall in stock valuations flows outward in the correlation network from stocks with relatively high centrality in the center to those on the periphery , which are more industry specific or otherwise uncorrelated to the core sectors of the stock market . in fig .
[ distanceplot ] , this spread is emphasized by showing the average return among stocks at a distance @xmath16 from the stock with the highest betweeenness centrality ( here cbs , a major s&p 500 stock , and here classified under the services industry ) , where @xmath16 is defined by eq .
[ distanceeq ] . here , we see that the greater the distance from the central part of the network , the more delayed the decline in valuation
. therefore , the credit crisis spreads among affected stocks from more centralized nodes to more outer ones as the news of the extent of the damage to the global economy spreads .
using methods of statistical physics and complex networks to investigate phenomena in stock markets is increasingly common @xcite . the increasing complexity and globalization of financial markets has led to many large and sometimes unpredictable effects . in ref .
@xcite , the effects of globalization upon the korea stock exchange were demonstrated by showing the increasing grouping coefficient of stocks from 1980 - 2003 .
the credit crisis , however , presents a challenge of a whole new magnitude .
as viewed by the wider market , the collapse in stock price returns began in the financial and services sector of the economy .
soon it moved across more mainline banks and firms , and more recently has affected stocks across the board .
though the spread of the collapse in stocks down the tree resembles an infection or cascade on a network , such ideas are more appropriately viewed as analogies or metaphors than explanations . unlike a disease or cascading collapse , the stock crash is not being transmitted from one stock to another .
what the collapse reveals is a complex and collective systemic collapse of the financial system , which spreads as its extent becomes more recognized and affects the credit or demand for sectors across the economy .
the spread is carried both by the news of the extent of the crisis and the fact that similar risky asset bases make the co - movement of certain stocks more likely and thus more highly correlated .
in addition , as credit becomes restricted , capital flows formerly relied on as a given begin to disappear , causing financial difficulties in companies and selling of equities ( among other assets ) to raise capital . as panic and the extent of the devastation spread
, stocks are punished accordingly . in normal times , the failure of a company and its stock is not a cause for a systemic crisis . also , since the correlation was calculated over an entire year s activity , the stock prices are correlated because they tend to fall similarly over time .
the correlation shown in this network does not cause the transmission chain of collapse , but is inextricably tied to it . in addition , the correlation generally increases with volatility ( for example , see ref .
@xcite ) and negative returns affect volatility more than positive returns of the same magnitude @xcite , so over time , the correlation has been increasing among stocks , and the network will likely be more dense and structured differently due to the steadily increasing market volatility . finally , one should note that this is not an example of the widely cited ` financial contagion ' in the press .
financial contagion refers to the coupling of financial panic across national borders and not among stocks in an exchange .
however , these do illustrate the spread of the credit crisis and how what was once a problem among home builders and mortgage finance companies has engulfed the entire economy .
gower , biometrika , * 53 * , 325 ( 1966 ) .
mantegna , eur.phys .
j. b * 11 * , 193 ( 1999 ) .
r.n . mantegna and h.e .
stanley , an introduction to econophysics : correlations and complexity in finance , ( cambridge university press , cambridge , 1999 ) .
kim , i.m .
kim , y. lee , and b. kahng , j. korean phys .
, * 40 * , 1105 ( 2002 )
kim , y.k .
lee , b. kahng , and i.m .
kim , j. phys .
japan , * 71 * , 2133 , 2002 .
j. cox,``credit crisis timeline '' , the university of iowa center international finance and development , retrieved october 17 , 2008 ( 2008 ) .
zhuang , z.f .
min , and s.y .
chen , journal of northeastern university(natural science ) , * 28 * , 1053 ( 2007 ) ( in chinese ) .
l. ping and b.h .
wang , systems engineering , * 24 * , 73 ( 2006 ) ( in chinese ) .
oh and s.h .
kim , j. korean phys .
* 48 * , 197 ( 2006 ) .
jo and s.y .
kim , j. korean phys .
* 49 * , 1691 ( 2006 ) .
jung , o. kwon , j.s .
yang , and h.t .
moon , j. korean phys .
* 48 * , 135 ( 2006 ) .
andersen , t. bollerslev , f.x .
diebold , and h. ebens , j. of financial economics , * 61 * , 43 ( 2001 ) . f. black , proceedings of the american statistical association , business and economic statistics section , 177 ( 1976 ) .
j.y . campbell and l. hentschel , j. of financial economics , * 31 * , 281 ( 1992 ) .
hoovers , http://www.hoovers.com | the credit crisis roiling the world s financial markets will likely take years and entire careers to fully understand and analyze .
a short empirical investigation of the current trends , however , demonstrates that the losses in certain markets , in this case the us equity markets , follow a cascade or epidemic flow - like model along the correlations of various stocks .
this phenomenon will be shown by the graphical display of stock returns across the network and by the dependence of the stock return on topological measures . |
radio emission as a property of x - ray binary systems ( xrbs ) has developed in recent years from being a rare and unusual facet of their broadband spectrum to being recognised as a ubiquitous property of several classes of xrbs . in the early 1990s
, it was still ( almost ) possible to discuss individually radio - emitting x - ray binaries ( e.g. hjellming & han 1995 ) , and numbers of systems in which collimated outflows , or jets , were directly observed could be counted on the fingers of one hand . as the 1990s drew on , more and more of these systems became resolved into such jets , and in the middle of the decade the apparent superluminal motions observed from grs 1915 + 105 ( mirabel & rodriguez 1994 ; see also fender et al .
1999 ) and gro j1655 - 40 ( tingay et al . 1995 ; hjellming & rupen 1995 ) established beyond doubt that xrbs could accelerate powerful flows to bulk velocities in excess of @xmath0 . by 2000
we are beginning to realise that jets may be rather more ubiquitous than previously imagined ( e.g. mirabel & rodriguez 1999 ; fender & hendry 2000 ; fender 2000 ) , and in figs 1(a),(b ) i summarise our ( or at least my ) current understanding of the empirical relation of radio emission to x - ray ` state ' of the neutron star ( ns ) and black hole candidate ( bhc ) systems respectively .
in addition to the relations indicated in these figures , it seems that discrete , radio - emitting ejection events ( in which the radio spectrum rapidly becomes optically thin ) are associated with state transitions ( including outbursts of transients ) .
while some areas of the coupling between accretion flow and jet are still empirically uncertain ( the atoll sources , the intermediate / very high state for bhcs ) , it seems that all xrbs except the high - field x - ray pulsars will , under the right conditions , produce a synchrotron - emitting jet .
it is therefore an important question to address the significance of this jet , energetically and dynamically , for the process of accretion onto compact objects as a whole .
one thing obvious from inspection of figs 1(a),(b ) is the apparent anti - correlation between the mass accretion rate , @xmath1 , traditionally estimated from x - ray studies alone , and the presence of a jet , in both ns and bhc systems ( see also belloni , migliari & fender 2000 and homan et al .
2001 for further discussion ) .
in cases where it is possible to associate a given synchrotron luminosity with a given volume , the minimum energy associated with the event can be estimated assuming approximate equipartition . in this situtation ,
the energy associated with the magnetic field is approximately equal to that associated with the electrons , and there is a corresponding equipartition magnetic field , b@xmath2 . below the equipartition field ,
the energy in particles ( @xmath3 ) dominates ; above equipartition , the energy in the field ( @xmath4 ) dominates . while there is little obvious physical justification for assuming equipartition , it is useful in that it represents the case of maximal radiative efficiency for a synchrotron - emitting plasma . one obvious way to associate luminosity and volume
is by observing discrete ejection events sufficiently well that the rise time ( and therefore maximum volume ) and peak flux can be accurately measured . in fig 2
we see an example of a sequence of repeated discrete ejection events , observed from radio through the mm band to the near - infrared band .
full details of the observations , and the application of equipartition estimations to these data , are found in fender & pooley ( 2000 ) .
the results of the equipartition calculations are presented in table i. a more basic introduction to minimum energy requirements via equipartition can be found in e.g. longair ( 1994 ) . from the estimates in table
i , it is clear that even in the minimum energy case , ie . a pair plasma with no significant bulk kinetic energy ,
the power required for the jet is @xmath5 erg s@xmath6 .
the integrated x - ray luminosity ( presumably _ not _ coming from the jet ? ) at this time was @xmath7 erg s@xmath6 , so we find that even in this case the ratio of jet power to total x - ray luminosity , @xmath8 .
if we assume that the ejections share the bulk relativistic motion of the major ejections , then it seems @xmath9 .
spencer ( 1996 ) presents a summary of energy estimates based on the same application of equipartition to radio varibility , for five xrb systems . in all cases
the equipartition magnetic field lies in the range 0.011 gauss , lower than that derived for the events from grs 1915 + 105 considered above ; however this may not be unreasonable considering that the peak emission from the longer events considered by spencer would have originated in larger volumes .
in all the five cases considered by spencer ( 1996 ) , the apparent jet power was at least equal to the integrated x - ray luminosity of the system . in cases where the radio structure can be clearly resolved ,
the volume responsible for a certain luminosity can be estimated directly from images , and again equipartition applied .
spencer et al .
( these proceedings ) apply this method directly to the steady , resolved jets in cygnus x-1 , and paredes et al .
( 2000 ) apply it to jets from the unusual system ls 5039 , finding equipartition magnetic fields of 0.3 and 0.2 gauss respectively .
both works further conclude that the energy associated with the resolved outflows is large , @xmath10 erg ; however a major uncertainty in estimating the power in the outflow is associated with how continuously it must be generated .
.calculation of ( synchrotron ) radiative luminosity , equipartition magnetic field , total energy , jet power and mass - flow rate for the oscillations reported here , given different physical assumptions .
@xmath11 is the bulk motion lorentz factor ; a filling factor of unity is assumed , otherwise the equipartition magnetic field would be high enough that radiative losses would be significant in the infrared . in these calculations
a distance of 11 kpc and doppler factors for relativistic bulk motion which are the same as those reported in fender et al .
( 1999 ) are all assumed . mass flow rate @xmath12 and jet power p are based upon one ejection every 20 min . for more details , see fender & pooley ( 2000 ) . [ cols=">,<,<,<,<,<,<,<,<",options="header " , ]
without large amplitude variability , or directly resolved jets , it is no longer possible to associate a given luminosity with a certain volume , and we must apply other arguments in order to estimate the total jet power .
this is the situation we face in trying to estimate the power in jets from bhcs in the low / hard x - ray state , which produce a flat spectrum in the radio band with no large - amplitude variability ( fender 2000 ) . in this case
we may estimate the total jet power by ( a ) carefully measuring the extent of the synchrotron spectrum which it produces , and ( b ) introducing a radiative efficiency , @xmath13 , which is the ratio of total to radiated power ( in the jet s rest frame ) . from this
we can estimate the jet power as @xmath14 where @xmath15 is the total radiative luminosity of the jet , @xmath13 is the radiative efficiency , and @xmath16 is a correction factor for bulk relativistic motion with lorentz factor @xmath11 and doppler factor @xmath17 , ( @xmath18 see fender 2000 ) .
starting from the reasonable assumption that all the emission observed in the radio band is synchrotron in origin , we can try to see how far this spectrum extends to other wavelengths .
firstly , it should be made clear that most systems have not been observed at @xmath19 ghz , and while some low - frequency turnovers have occasionally been observed , there are no reported cases of a complete cut - off to the synchrotron emission at low radio frequencies . in any case , while a low - frequency cut - off is important for estimating the mass of the ejecta in the ( by no means certain ) case that there is a proton for each emitting electron , the radiative luminosity is dominated by the high - frequency extent of the synchrotron spectrum . since the spectrum is unlikely to have a spectral index @xmath20 ( where @xmath21 ) , then the most important observation is simply determination of the highest frequency at which synchrotron emission is observed . in fig 3 the broadband radio optical spectra of eight xrb systems in ( low/)hard x - ray states
is presented ; in all cases the spectra hint at an extension of the synchrotron spectrum to the near - infrared or even optical bands .
there is a large amount of additional evidence for such a high - frequency extent to the spectrum , each piece of which may be considered ` circumstantial ' but when gathered together presents a compelling case ( a lot of this evidence is presented in fender 2000 ) .
note also that observations of grs 1915 + 105 ( admittedly not a ` typical ' system ! ) , such as those shown in fig 2 , unambiguously establish that the synchrotron spectrum can extend to the near - infrared band .
our best recent case is the low / hard state transient xte j1118 + 480 , which clearly shows excess emission at near - infrared and probably also optical wavelengths ( hynes et al .
2000 ) and whose radio spectrum smoothly connects to a sub - mm detection at 850 @xmath22 m ( fender et al .
2000 ) . in fender et al .
( 2000 ) it is argued that in this case the synchrotron radiative luminosity is already @xmath23% of the bolometric x - ray luminosity .
how important the total jet power is then depends on our estimates for the radiative efficiency , @xmath13 . in fender & pooley ( 2000 ) a detailed study
was made of the power required to maintain the repeated ejection events shown in fig 2 . in table
i , we also tabulate this radiative efficiency for the different cases considered ; its maximum value is @xmath24 . in one of the original works on the power in steady jets , blandford & knigl ( 1979 ) determined that @xmath25 , where @xmath26 and @xmath27 , @xmath28 being the upper and lower bounds of the electron energy spectrum respectively . since it seems likely that @xmath29 is in the range @xmath30@xmath31 , then @xmath32 .
celotti & ghisellini ( 2001 ) find a comparable value for the radiative efficiency of agn jets .
the best - studied cases of clearly resolved ( in time ) ejection events , such as those presented in fig 2 , seem to require that the ratio of jet power to x - ray luminosity , @xmath33 .
a more general study of large outbursts from five different systems shows that , via the simple equipartition arguments outlined above , @xmath34 ( spencer 1996 ) .
other applications of equipartition to xrb ejection events exist , and in all cases it seems that they imply the jet power is extremely significant . using a different , but not altogther unrelated , method of measuring the high - frequency extent of the synchrotron spectrum
we find that for the bhc transient xte j1118 + 480 , @xmath35 , where @xmath32 for bhc systems in the canonical low / hard x - ray state .
since it seems that the relation of x - ray to radio flux in this state is comparable for all the systems ( ie .
the similarities in the broadband spectra of the sources presented in fig 3 extend to at least the x - ray band ) , then it seems reasonable to conclude that @xmath36 for all bhc xrbs in the low / hard x - ray state .
again it seems that the jet must be highly significant for the process of accretion .
two different ( but not entirely independent ) approaches have revealed that a large fraction of the entire luminosity ( radiative and mechanical ) of some _ classes _ ( this is important we are no longer talking about individual ` oddballs ' ) of x - ray binaries may be carried by the jet .
this fraction is likely to be @xmath37 and may be @xmath38 , ie
. the jet may even be the _
dominant _ power output channel . since it is clear from studies of e.g. grs 1915 + 105 that the disc jet coupling is occurring in the inner few 100 km of the accretion flow , the extraction of such a large amount of energy ( and presumably angular momentum ) must have a dramatic effect on the accretion flow close to the compact object .
based on this , it seems that models for accretion in x - ray binaries which at least parameterise such an outflow are needed .
they are not entirely non - existent for grs 1915 + 105 ( again ! ) both nayakshin , rappaport & melia ( 2000 ) and janiuk , czerny & siemiginowska ( 2000 ) have begun to consider the effect of a moderately powerful outflow on the accretion process . a more radical step has been taken by markoff , falcke & fender ( 2000 ; see also these proceedings ) in which almost the entire broadband spectrum of the bhc low state ( excepting thermal emission in the optical uv range ) is explained by jet emission , and in which it is suggested that the x - ray power - law may even be _ synchrotron _ emission .
this would be contrary to the generally accepted view that the x - ray power law in the low state arises in comptonisation of ` seed ' photons in a hot , thermal , corona .
it is further worth noting that once mass - loss via jets is considered , and the empirical connections between x - ray ` state ' and jet formation , as outlined in figs 1(a),(b ) recognised , then it is no longer clear that the global mass transfer rate can be inferred from x - ray observations alone .
i would like to thank many of the participants at the granada workshop for stimulating conversations , and annalisa celotti and kieran obrien for reading through a draft of this work .
in particular i would like to acknowledge many relevant conversations , over several years , with ralph spencer , guy pooley and the late bob hjellming .
hjellming , r. m. , han , x. , 1995 , _ radio properties of x - ray binaries_. in : lewin , w. h. g. , van paradijs , j. , van der heuvel , e. p. j. ( eds . ) , x - ray binaries , camridge university press , cambridge , 308330 | i discuss the energetics of synchrotron - emitting outflows , increasingly found to be present in many different classes of x - ray binary systems .
it is shown that the outflow is likely to be comparable in power to the integrated x - ray luminosity , traditionally taken to be an indicator of the global mass - transfer rate .
this is especially found to be the case in the ( low/)hard states of black hole candidate systems .
i conclude that jets are extremely important , energetically and dynamically , for the accretion process in the majority of known x - ray binary systems . |
merging ( or inspiralling ) dns systems are one of the prime targets for ground - based gravitational - wave ( gw ) interferometers such as geo600 , tama , virgo , and ligo ( the laser interferometer gravitational - wave observatory ) @xcite .
event rates of the dns inspiral searches by these detectors can be inferred using the rate estimates with an extrapolation out to the maximum detection distances for any detector under consideration . before 2003 , the galactic dns merger rate had been estimated between @xmath4 yr@xmath2 ( see kalogera et al .
2001 and references therein ) . at that time
, there were only two systems available for empirical studies , psrs b1913 + 16 @xcite and b1534 + 12 @xcite .
we calculated the pdf of galactic dns merger rates , @xmath5 , based on these two systems ( see kim et al .
2003 , hereafter kkl , for further discussions ) .
soon after the discovery of the highly relativistic system psr j0737@xmath33039 , we were able to revise @xmath5 including psr j0737@xmath33039 in collaboration with the observation team@xcite .
the discovery of j0737@xmath33039 , resulted in a significant increase in the estimated galactic dns by a factor of @xmath06 .
this implies a boost in event rates for dns searches for gw interferometers . in this paper
, we present our recent results on : ( i ) the pdf of galactic dns merger rate estimates with updated observations ; ( ii ) the _ global _ pdf of rate estimates considering the systematic uncertainties ; ( iii ) constraints on upper limits for rate estimates based on the observed supernova rate incorporated with our theoretical understanding of the sn - dns relation ; ( iv ) the approximate contribution of j1756@xmath32251 to the galactic dns merger rates and uncertainties in our assumptions on the efficiency of the acceleration searches for parkes multibeam pulsar survey ( pmps ) .
finally , we discuss implications of the most recently discovered pulsar binary j1906 + 0746 @xcite .
here , we describe the main components of the calculation of the combined @xmath6 considering the three observed dns systems in the galactic disk . the merger rate of a given population _
i _ can be defined by @xmath7 where @xmath8 represents the number of pulsars in our galaxy with pulse and orbital characteristics _
similar _ to an observed sample _
( e.g. psr j0739@xmath33039 ) and @xmath9 is a correction factor to take into account pulsar beaming ( typically @xmath10 ) .
@xmath11 is the lifetime of system _ i _ based on its observed properties . in order to calculate @xmath8 , we distribute a large population of pulsars _ all _ similar to the system _ i _ , in a model galaxy assuming spatial and luminosity distributions .
since pulsar luminosities are drawn from a distribution , the observed flux and estimated distance for the dns system are not relevant in our calculation .
moreover , the selection effects for faint pulsars are implicitly taken into account .
once we generate a pulsar population with a size @xmath12 , we can then calculate the number of pulsars detected ( @xmath13 ) by large - scale pulsar surveys .
we repeat the survey simulations with a detailed modeling of selection effects for observed dns systems . for a fixed @xmath12
, we find that @xmath13 follows a poisson distribution , p(@xmath14 ) , where @xmath15 is a mean value of @xmath13 for a given population size ( @xmath12 ) .
we require @xmath16 , i.e. , we consider only one observed system , and calculate the best - fit value of @xmath15 . with a wide range of @xmath12 , we find @xmath17 as expected where c is a constant . applying bayes theorem to these results ,
we calculate a p(@xmath12 ) , a pdf for the population size of a given system _
i _ knowing that there is one observation .
the calculation @xmath5 from @xmath18 is straightforwad , and the full derivation and formula can be found in appendix a of kim et al .
( 2004 ) .
the lifetime of a dns system is defined by @xmath19 , where @xmath20 is the spin - down age of a recycled pulsar @xcite and @xmath21 is the remaining lifetime until the two neutron stars merge @xcite .
based on the most recent observations , we estimate the lifetime of j0737@xmath33039 to be @xmath0230myr @xcite .
this is the shortest among the three observed systems .
the beaming correction factor @xmath22 is defined as the inverse of the fractional solid angle subtended by the pulsar beam .
its calculation requires detailed geometrical information on the beam .
following kalogera et al .
( 2001 ) and updating the values with recent observations , we calcualte @[email protected] for psr b1913 + 16 @xcite and @xmath06.0 for psr b1534 + 12 @xcite without good knowledge of the geometry of j0737@xmath33039 , we assume @xmath23 to be the average value of the other two systems ( @xmath24 ) .
is shown on a log scale .
the thick solid line is the galactic rate estimate overlapped with results for individual observed systems ( dashed lines ) .
dotted lines indicate confidence intervals for the rate estimates .
the same results are shown on a linear scale in the small inset .
all results are from our reference model.,width=275 ] in fig . 1 , we show @xmath6 for a reference model ( model 6 in kkl ) .
we obtain the most likely value of @xmath25 , larger by a factor of @xmath26 than the rate estimated before the discovery of j0737@xmath33039 .
the increase factor is found similar for all pulsar population models we examined .
the increased merger rates imply a boost in the inferred detection rate of gws from dns inspirals for ground - based gw interferometers such as ligo . in order to calculate the detection rate ( @xmath27 ) , we assume a homogeneous distribution of galaxies in nearby universe and a spherical symmetry in detector sensitivity .
then , we can write @xmath28 , where @xmath29 is the number of galaxies in the detection volumne ( @xmath30 ) .
we calculate the number density of galaxies derived by the observed blue luminosity density , ( @xmath31myr@xmath32 ( see phinney ( 1991 ) and kalogera et al . (
2001 ) for more details ) .
the detection volume of ligo can then be defined as a sphere for a given detection distance ( 20 mpc and 350 mpc for the inital and advanced ligo , respectively ) , and the number of galaxies within @xmath30 is simply @xmath33 . for our reference model
, we find that the most probable event rates are about 1 per 30yrs and 1 per 2 days , for initial and advanced ligo , respectively . at the 95% confidence interval ,
the most optimistic predictions for the reference model are 1 event per 9yrs and 1.6 events per day for initial and advanced ligo , respectively .
more details can be found in kalogera et al .
( 2004 ) . as shown in the fig . 1 ,
the galactic dns merger rate is dominated by psr j0737@xmath33039 .
recently , lorimer et al . ( 2005a ) suggested that the current age of j0737@xmath33039 is @xmath34 myr .
this implies a even shorter lifetime ( @xmath35 myr ) knowing that the estimated merger timescale of this system is @xmath085 myr .
based on their results , we find the most likely value of @xmath36 for the reference model is @xmath37myr@xmath2 .
the beaming correction for j0737@xmath33039 is not yet constrained and we assume msps discovered in dns systems are not very different . as a conservative lower limit , without any beaming corrections for all observed systems for a reference model , we obtain @xmath38 myr@xmath2 with a 95% confidence interval .
the corresponding detection rates for initial and advanced ligo are @xmath39 yr@xmath2 and @xmath40 yr@xmath2 , respectively .
only when the axis geometry of j0737 - 3039 becomes available , we will be able to constrain the uncertainties of the beaming fraction , and in turn , the rate estimates .
in kkl , we showed that empirical dns merger rates are strongly dependent on the assumed luminosity distribution function for pulsars , but not on the pulsar spatial distribution .
therefore , we can consider only the rate dependence on the pulsar luminosity function for simplicity . here , we describe how we can incorporate the systematic uncertainties from these models and calculate , @xmath41 , a _ global _ pdf of rate estimates .
note that the results available on prior functions for pulsar luminosity distribution are currently out of date .
specific quantitative results could change when constraints on the luminosity function are derived from the current pulsar sample .
we assume a power - law luminosity distribution for a radio pulsar luminosity function @xmath42 .
this function is defined by two parameters : the cut - off luminosity @xmath43 and power - index @xmath44 .
we assume prior distributions for these two parameters and calculate @xmath45 .
we fit the marginalised likelihood of @xmath43 and @xmath44 presented by @xcite and obtain the following analytic formulae for prior functions , i.e. @xmath46 and @xmath47 : @xmath48 and @xmath49 , where @xmath50 and @xmath51 are coefficients we obtain from the least - square fits and the functions are defined over the intervals @xmath52 $ ] mjy kpc@xmath53 and @xmath54 $ ] .
we note that , although cordes and chernoff ( 1997 ) obtained @xmath46 over @xmath55 $ ] mjy kpc@xmath53 centered at 1.1 mjy kpc@xmath53 , we consider @xmath46 with a peak at @xmath56mjy kpc@xmath53 considering the discoveries of faint pulsars with l@xmath57 below 1 mjy kpc@xmath53 @xcite .
next we calculate @xmath45 as follows : @xmath45 + @xmath58 . on a linear scale ( lower panel ) and the assumed intrinsic distributions for @xmath43 and @xmath44 ( upper panels ) .
dotted lines represent the lower ( @xmath59 ) and upper ( @xmath60 ) bounds on the observed sn ib / c rate scaled by 5% of the observed sn ib / c rates , @xmath61 myr@xmath2 ( see text).,width=275 ] in fig . 2
, we show the distributions of @xmath43 and @xmath44 adopted ( top panels ) and the resulting global distribution of galactic dns merger rate estimates ( bottom panel ) .
we find the peak value of @xmath45 at _ only _ around 13myr@xmath2 .
we note that this is a factor @xmath62 smaller than the result from the reference model ( @xmath6371myr@xmath2 ) . at the 95% confidence interval , we find that the _ global _ galactic dns merger rates lie in the range @xmath0 1145myr@xmath2 .
these imply ligo event rates in the range @xmath64yr@xmath2 ( initial ) and @xmath65yr@xmath2 ( advanced ) .
since 1997 , the number of known millisecond pulsars has more than doubled , and therefore , constraints on @xmath43 and @xmath44 and their pdfs based on the most up - to - date pulsar sample are urgently needed .
according to the standard binary evolution scenario , the progenitor of the second neutron star is expected to form during a type ib / c supernova .
therefore , the empirical estimates for the type ib / c sn rate in our galaxy can be used to set upper limits on the dns merger rate estimates . based on cappellaro et al .
( 1999 ) , we adopt @xmath66myr@xmath2 considering sbc sd galaxies , a hubble constant @xmath67km / s / mpc and the blue luminosity of our galaxy @xmath68 .
we calculate the fraction of sn ib / c actually involved in the formation of dns with a binary evolution code startrack @xcite and estimate the rate ratio : @xmath69/@xmath70 .
motivated by this result , we adopt the empirical @xmath71 assuming @xmath72 and compare the value with the global pdf ( fig .
[ cols= " < , > , > , < , < , < " , ] @xmath73 pulsar spin period + @xmath74 orbital period + @xmath75 eccentricity + @xmath76 lifetime of a system + @xmath77 population size for a given dns system + @xmath78 approximate @xmath79 .
results depend on the survey modeling ( see text ) .
+ @xmath80 the nature of the companion is not yet clear .
@xmath80 @xmath81 where @xmath82 is a death - time of the pulsar .
+ we note that our most optimistic dns merger rate is @xmath83 myr@xmath2 at a 95% confidence interval ( model 15 in kkl ) .
we obtain @xmath84 with respect to the center value of the empirical sn type ib / c rate ( @xmath85 myr@xmath2 ) and the upper limit of @xmath86 at the 95% confidence interval ( @xmath87 myr@xmath2 ) .
this corresponds to @xmath88 for a sn type ii rate , which is factor 6.1 larger than that of sn type ib / c . in both cases ,
the most optimistic model is lower than the current empirical supernova rate estimates , but not really consistent with the results of population synthesis calculations .
if we consider a upper limit of @xmath86 at the 95% confidence interval from the global pdf , we obtain @xmath89 13% and 2% for the center value of sn type ib / c ( 1100 myr@xmath2 ) and ii , respectively .
recently , faulkner et al .
( 2005 ) discovered psr j1756@xmath32251 , the 4th merging dns in the galactic disk from the parkes multibeam pulsar survey ( pmps ) .
the standard fourier method failed to find this pulsar and they reanalysed the pmps data with an acceleration search ( or a ` stack search ' as described in their paper ) . in order to calculate the merger rate including psr j1756@xmath32251 ,
a detailed simulation is necessary to calculate the effect of the acceleration search with the pmps .
however , the approximate contribution of psr j1756@xmath32251 to the galactic merger rate can be easily obtained .
we find the total rate increases by only @xmath90 due to the new discovery .
this is expected because psr j1756@xmath32251 can be identified as a member of the b1913 + 16-like population , which has already been taken into account in the calculation .
only future detections of pulsars from a significantly different population ( compared to the known systems ) , or from the most relativistic systems , will result in a non - trivial contribution to the rate estimates .
finally , we note the implications of j1906 + 0746 on the pulsar binary merger rates .
this system has drawn attention due to the extremely young age of the pulsar ( characteristic age of @xmath0 112 kyr ; lorimer et al .
if the companion is another neutron star , j1906 + 0746 would be the first discovery of a non - recycled component in a dns system .
currently , the nature of the companion is totally unknown , and it can be either a light neutron star , or heavy ( o - mg - ne ) white dwarf . assuming j1906 + 0746 is a dns system , we calculate its contribution to the galactic merger rate .
because of its short lifetime ( @xmath082myr ) , j1906 + 0746 can increase the galactic dns merger rate by about a factor 2 .
this implies that the current estimated dns merger rate including j0737@xmath33039 can still be doubled !
if j1906 + 0746 is an eccentric ns wd system , such as j1141 - 6545 @xcite or b2303 + 46 @xcite , it will be as important as j1141 - 6545 , which is currently dominate the birthrate of eccentric ns - wd binaries . in table 1
, we summarize some important properties of observed merging dns systems , which used in the calculation of @xmath91 in this paper .
we summarize our results on the galactic merger rate of double neutron stars ( dns ) in view of the recent discovery of psr j0737 - 3039 .
we update the results based on the most recent observations .
global _ pdf of rate estimates incorporates the already known systematics due to the radio pulsar luminosity function .
the most likely value obtained from the global distribution is at only @xmath0 13 myr@xmath2 , and a reanalysis of the _ current _ pulsar sample and radio luminosities is needed for a reliable assessment of the best fitting distribution .
the dns merger rate can be constrained by empirical type ib / c supernova rates , and we compare the supernova rate with the global pdf .
recent discoveries of relativistic systems , such as j1756@xmath32251 and j1906 + 0746 are important for the reduction of the uncertainties associated with the empirical rate estimates . as for j1756@xmath32251
, we expect their contribution to the galactic dns merger rate is negligible , and this pulsar can be seen another sample of b1913 + 16-like population .
it is not clear whether the companion of j1906 + 0746 is a neutron star or a white dwarf .
this system may increase the estimated merger rate of the relevant population by a factor of 2 whether it is a dns or another eccentric ns - wd binary .
this research is partially supported by nsf grant 0353111 , and a packard foundation fellowship in science and engineering to vk .
drl is a university research fellow supported by the royal society . | we present the current estimates of the galactic merger rate of double - neutron - star ( dns ) systems . using a statistical analysis method
, we calculate the probability distribution function ( pdf ) of the rate estimates , which allows us to assign confidence intervals to the rate estimates .
we calculate the galactic dns merger rate based on the three known systems b1913 + 16 , b1534 + 12 , and j0737 - 3039 .
the discovery of j0737 - 3039 increases the estimated dns merger rate by a factor @xmath06 than what is previously known .
the most likely values of dns merger rate lie in the range @xmath1 myr@xmath2 depending on different pulsar models .
motivated by a strong correlation between the peak rate estimates and a pulsar luminosity function , we calculate a _
global _
probability distribution as a single representation of the parameter space covered by different pulsar population models .
we compare the global pdf with the observed supernova ib / c rate , which sets an upper limit on the dns merger rate . finally
, we remark on implications of new discoveries such as of j1756@xmath32251 , the 4th dns in the galactic disk , and j1906 + 0746 , a possible dns system .
method : statistical ; stars : binary ; pulsars : individual ( psr j0737 - 3039 ) ; gravitational waves |
the score detector is designed to study @xmath0-rays in the energy regime from a few @xmath1{tev}$ ] up to some @xmath1{pev}$ ] .
the detector consists of a large array of light sensitive detectors with a half opening angle of about @xmath2 and is used to sample the cherenkov light shower front of extensive air showers ( eas ) .
each detector contains a few photomultiplier tubes ( pmts ) which are equipped with winston cones , facing towards the sky . with a light sensitive area in the order of @xmath3{m^2}$ ] in the individual detector stations it will be possible to sample the shower front up to high distances from the shower core , thus making it possible to place the stations about @xmath4{m}$ ] apart from each other .
the signals from the pmts will be sampled by fast flash - adcs , making it possible to use not only the photon density but also timing features for event reconstruction .
a comprehensive overview of the physical motivation and the detector concept can be found in another contribution on this conference @xcite . in this paper a scheme for event reconstruction and results of simulations testing this reconstruction algorithm are presented .
the simulation chain consists of air shower simulations in cosiksa @xcite ( version 6.75 ) , the detector simulation _ sim_score and a reconstruction framework .
showers induced by @xmath0-rays and protons in the energy range from @xmath1{tev}$ ] up to @xmath5{pev}$ ] have been simulated , and the results are analysed using different detector geometries .
the two detector alternatives discussed here are : _ 1 .
@xmath4{m}$ ] spacing and @xmath6{m^2}$ ] entrance window 2 .
@xmath7{m}$ ] spacing and @xmath8{m^2}$ ] entrance window in each energy band showers with zenith angles of @xmath9 , @xmath10 and @xmath11 have been simulated , and all results given here are averaged values over the different angles ( the same number of showers has been simulated for each angle ) .
the detector simulation includes atmospheric absorption of cherenkov light and the anticipated night sky background .
the detector modules are modelled with a realistic pmt pulse shape function ( taken from @xcite ) and a wavelength dependent quantum efficiency ( taken from @xcite ) .
the angular acceptance of the winston cones has been studied with ray tracing simulations and incorporated into the detector simulation .
all resolutions given in here are @xmath12 values of the difference between mc input and the reconstructed quantity .
firstly , the lateral distribution of the photon density is recorded .
it can be seen that the intensity falls off exponentially up to about @xmath13{m}$ ] from the shower core . from @xmath13{m}$ ] to @xmath14{m}$ ] the light distribution follows a power law , until it changes to an exponential function again ( see fig . [ lat_dist ] ) : @xmath15{m } \\
q ( \frac{r}{\unit[220]{m}})^k & \mbox{for } \unit[120]{m } < r <
\unit[400]{m } \\ p_2 \exp(d_2 r ) & \mbox{for } r > \unit[400]{m } \\
\end{cases } \label{ldf}\ ] ] { pev}$ ] @xmath0-particle . for demonstration purposes a @xmath16{m}$ ] sampling
is assumed.,width=268 ] the part up to @xmath13{m}$ ] can only be used if a spacing of @xmath4{m}$ ] is assumed , otherwise there are not enough data points in this range .
the data points beyond @xmath14{m}$ ] usually have a quite poor signal to noise ratio , making the power law part the most important for score .
if continuity at the connection points ( @xmath13{m}$ ] and @xmath14{m}$ ] ) is required in the fits , the number of free parameters can be reduced to four : @xmath17 , @xmath18 , @xmath19 and @xmath20 .
typical values for the power law index @xmath18 are in the range of -2 to -2.6 . secondly , the peak ( highest entry in signal histogram ) and the width of each signal ( fwhm ) are determined .
plotting the peak times versus the distance from the shower core one finds that a parabolic function describes the data very well ( fig .
[ peaks ] ) .
the width of the timing signal increases linearly at distances greater than about @xmath7{m}$ ] , while at lower distances the width is roughly constant due to the time resolution of the pmts ( fig .
[ widths ] ) .
{ pev}$ ] @xmath0-particle , recorded by a detector array with a spacing of @xmath7{m}$].,width=268 ] with a linear fit from @xmath7{m}$ ] to @xmath21{m}$].,width=268 ]
the simplest approach for the reconstruction of the shower core is to calculate the centre of gravity of the measured light intensities ( used for example in airobicc , see @xcite ) . while being robust and fast , this method can achieve only mediocre resolutions as it does not take into account the non - linearity of the lateral photon density function ( ldf ) shown in fig .
[ lat_dist ] . a better resolution can be achieved if the known ldf is fitted two - dimensionally to the recorded light intensities . using the power law part of the ldf
[ ldf ] ) , a resolution of better than @xmath22{m}$ ] can be achieved at higher energies . towards lower energies
both methods become less accurate when the signal approaches the noise level . while at higher energies the spacing of the detector stations makes no difference , the smaller spacing significantly decreases the threshold for a good reconstruction ( see fig .
[ core_resolution ] ) .
-ray energy for two different detector layouts with @xmath4{m}$ ] ( blue ) and @xmath7{m}$ ] ( red ) spacing ( see section [ simulations ] ) and two reconstruction methods : centre of gravity method ( dashed lines ) and ldf fit ( solid lines).,width=268 ]
simulations show that the photon density at any point of the cherenkov shower front is proportional to the energy of the primary particle , which makes the energy reconstruction a straightforward process .
however , at small distances from the shower core the photon density is also affected by the depth of the shower , while at large distances the signal to noise ratio gets worse .
the best trade - off has been found at a distance of @xmath23{m}$ ] , independent of the detector geometry . taking the functional value at @xmath23{m}$ ] of the fitted ldf ( eq . [ ldf ] ) , an energy resolution of better than 10% can be achieved at high energies . at lower energies , the detector variant with @xmath4{m}$ ] spacing performs significantly better due to the four times larger number of data points in the same distance interval , providing a better signal to noise ratio ( see fig .
[ energy_resolution ] ) .
-ray energy for two different detector layouts with @xmath4{m}$ ] ( blue ) and @xmath7{m}$ ] ( red ) spacing of stations and different sensitive areas per station ( see section [ simulations]).,width=268 ] -ray energy for two different detector layouts with @xmath4{m}$ ] ( blue ) and @xmath7{m}$ ] ( red ) spacing of stations and different sensitive areas per station ( see section [ simulations]).,width=268 ]
it can be seen from fig .
[ peaks ] that the peak arrival times are distributed according to a parabolic function for a vertical shower . for non - vertical showers ,
this parabola is tilted by the angle of the incoming particle . for finding the direction of the particle , the superposition of a plane and a three - dimensional parabola is fitted to the data points : @xmath24 @xmath25 and @xmath26
can then be used to calculate zenith and azimuth angle .
both detector layouts achieve an angular resolution of better than @xmath27{^\circ}$ ] at higher energies , while the detector with the smaller spacing has a better performance at lower energies again ( fig .
several methods for the calculation of the shower depth are discussed in the literature . the tunka collaboration has shown that both the steepness of the inner part of the ldf and the widths of the timing signals ( see sec .
[ data ] ) can be used for this purpose @xcite .
these methods are called intensity method and width method from here on .
simulations for score show that also the time delay of the peak arrival times , which is fitted by a parabola ( fig . [ peaks ] ) can be used , if an accurate inter - station time calibration can be implemented ( called timing method ) .
combining the different methods improves the depth resolution significantly . in order to improve the performance of both the width and the timing method , all signals within a certain distance interval from the shower core can be summed up before determining peak position and signal width . as the number of detector stations available in a certain distance interval is proportional to the distance to the shower core
, this procedure makes it possible to sample the width and the peak times up to high distances from the shower core .
the size of the distance intervals is a critical parameter of the reconstruction and has to be optimized for each detector layout and energy range .
if the signals are still too noisy for a good determination of the width , a lognormal function can be fitted to the signals ( see @xcite ) .
especially the determination of the width of the signal is improved significantly by that .
figure [ depth_resolution ] shows the depth resolution for the two different detector layouts .
the @xmath4{m}$ ] detector variant uses all three described methods , while the @xmath7{m}$ ] detector can use only the width and the timing method , as there are not enough intensity data points at the inner part of the ldf .
both detectors achieve a resolution of better than about @xmath28{g / cm^2}$ ] at energies above @xmath29{tev}$ ] , but the detector with the smaller spacing performs better at all energies as it can combine all three methods for determining the shower depth .
-ray energy for two different detector layouts with @xmath4{m}$ ] ( blue ) and @xmath7{m}$ ] ( red ) spacing of stations and different sensitive areas per station ( see section [ simulations]).,width=268 ]
the power to distinguish air showers of @xmath0-photons and hadrons is an important contribution to the sensitivity of a @xmath0-ray detector .
the usual measure to quantify this power is the quality factor , which is defined as @xmath30 where @xmath31 is the fraction of particles of the type @xmath32 which survive the applied cuts . obviously , it is desirable to have a high chance to identify a @xmath0-photon as signal ( @xmath33 close to one ) and a low chance to falsely identify a hadron as signal ( @xmath34 as small as possible ) , thus a high quality factor .
two approaches for @xmath0 / hadron separation are investigated : it has been noted in section [ section_depth ] that the shower depth can be obtained in different ways . while they all work well with both @xmath0 and hadron showers
, it seems that the timing method and the width method produce slightly different results for different particles .
this is believed to be due to the differences in the longitudinal development of showers of photonic or hadronic origin .
while photonic showers tend to emit most light near the shower maximum , the light from hadronic showers is emitted from a wider range of altitudes .
this makes the signals of hadronic showers slightly longer and leads to an overestimation of the shower depth for hadronic showers in the widths method . comparing the two depth values
can therefore be used as an indication for the nature of the particle . at higher energies ( where both methods have good resolutions ) this can be used for a basic @xmath0 / hadron separation .
figure [ gamma_hadron_manuell ] shows histograms of the ratio of the depth values obtained by the two methods for @xmath0 and hadron showers
. a quality factor of about 1.2 can be reached for energies above @xmath21{tev}$ ] .
-rays ( blue ) with energies between @xmath21{tev}$ ] and @xmath35{tev}$ ] .
the quality factor for a separation at 1 is about 1.2.,width=268 ] in order to find a structure in the data that may improve the power for @xmath0 / hadron separation , self - learning algorithms can be used .
so far , a boosted decision tree algorithm with @xmath36-boosting ( for reference , see e.g. @xcite ) has been implemented and first tests look promising .
further tests with different input variables , algorithm parameters and algorithms are under way .
the simulations conducted so far have shown that a shower front detector like score can achieve angular and energy resolutions of the same level as current imaging cherenkov telescopes ( better than @xmath37 and @xmath38 respectively ) . also , a satisfactory shower depth resolution of better than @xmath28{g / cm^2}$ ] has been found . currently , the quality factor for background rejection is only at about 1.2 , but this has to be considered as a first preliminary result .
optimizations of the shower depth calculation algorithm are in progress ( e.g. better fit routines and functions , cuts on bad data points ) , and it is anticipated that a better shower depth resolution will have a positive impact on the @xmath0 / hadron separation .
furthermore , other analytic and self - learning algorithms for background rejection are being explored currently . of the two detector layouts presented here the one with the smaller spacing and
the smaller sensitive area per detector station has been found to be superior in terms of reconstruction power at lower energies .
however , it uses about 25% more pmts and four times more detector stations .
therefore , a reasonable approach might be to combine the two layouts , i.e. use a part of the detector section with smaller spacing for detection of lower energy particles and employ a high energy extension for sensitivity at highest energies .
currently , further simulations with different detector layouts and spacings are under way .
the aim of this part of the project is to find the optimal detector configuration for the desired application and calculate a sensitivity curve based on the simulated reconstruction power of this instrument .
daniel hampf acknowledges the support of the german ministry for education and research ( contract number 05a08gu1 ) .
99 n. m. budnev : _ cosmic ray energy spectrum and mass composition from @xmath39 to @xmath40{ev}$ ] by data of the tunka eas cherenkov array _ , 29th icrc ( 2005 ) , pune , * 00 * , 101 - 106 v. r. chitnis and p. n. bhat : _ simulation studies on arrival time distributions of erenkov photons in extensive air showers _ , 1999 ,
astroparticle physics , volume 12 , issue 1 - 2 , p. 45 - 64
et enterprises limited : _ 9352 kb series data sheet _ , 2008 ,
uxbridge , uk d. heck , j. knapp , j. n. capdevielle , g. schatz , t. thouw : _ report fzka 6019 _ , ( 1998 ) v. henke : _ studie zur auswertung der ankunftszeitverteilung des erenkov - lichts ausgedehnter luftschauer _ , diploma thesis , hamburg , 1994 a. karle : _ entwicklung eines neuartigen atmosphrischen tscherenkovdetektors und messungen an hochenergetischer kosmischer strahlung zwischen @xmath41{tev}$ ] und @xmath42{tev}$ ] _ , phd thesis , mnchen , 1994 byron p. roe , hai - jun yang , ji zhu , yong liu , ion stancu , gordon mcgregor : _ boosted decision trees as an alternative to artificial neural networks for particle identification _ , 2005 , nuclear instruments and methods in physics research section a , volume 543 , issue 2 - 3 , p. 577 - 584 .
martin tluczykont , tanja kneiske , daniel hampf , dieter horns ; these proceedings | the proposed score detector consists of a large array of light collecting modules designed to sample the cherenkov light front of extensive air showers in order to detect high energy @xmath0-rays .
a large spacing of the detector stations makes it possible to cover a huge area with a reasonable effort , thus achieving a good sensitivity up to energies of about a few 10 pev . in this paper
the event reconstruction algorithm for score is presented and used to obtain the anticipated performance of the detector in terms of angular resolution , energy resolution , shower depth resolution and @xmath0 / hadron separation .
shower - front - sampling detector event - reconstruction |
over about the past decade , the parameter space of traditional stellar astrophysics has been greatly expanded with the discovery and characterization of brown dwarfs , objects that for most of their lifetimes are colder and less luminous than main - sequence stars . despite ample progress in finding and characterizing brown dwarfs , very few direct measurements of their physical properties
have been made so far .
in particular , dynamical masses for brown dwarfs are sorely needed to test the theoretical models over a wide range of parameter space . in comparison to the @xmath15100 binary stars with direct mass determinations , dynamical masses
have been measured for only a handful of objects clearly below the stellar / substellar boundary : 1 .
the secondary component of the m8.5+m9 visual binary gl 569bab , which itself is a companion to a young ( @xmath14100300 myr ) field m2.5 dwarf @xcite ; 2 .
the very young ( @xmath1myr ) eclipsing m6.5+m6.5 binary brown dwarf 2mass j05352184@xmath00546085 @xcite ; 3 .
the secondary component of the gj 802ab system , where the secondary has an estimated spectral type of l5 and the primary is a m5.5+m5.5 spectroscopic binary @xcite .
in addition , gj 569ba itself may be a unresolved binary brown dwarf @xcite , and the secondary component of the l0+l1.5 binary 2massw j0746425 + 2000321ab , which appears to be an old ( @xmath161 gyr ) field system , has a mass near the stellar / substellar boundary .
about 100 ultracool visual binaries are known , found with high angular resolution imaging surveys conducted by _ hubble space telescope _
( _ hst _ ; e.g. , @xcite ) and ground - based adaptive optics ( ao ) imaging ( e.g. , @xcite ; liu et al . , in prep ) .
only @xmath1410% of these are binaries composed of the coldest class of brown dwarf , the t dwarfs .
t dwarfs are distinguished by their very red optical colors arising from pressure - broadened alkali resonance lines and very blue near - infrared
( ir ) colors from strong , , and collision - induced absorption ( e.g. * ? ? ? * ; * ? ? ?
* ; * ? ? ?
these are the lowest luminosity and coolest objects directly detected outside of our solar system , with bolometric luminosities ( ) of @xmath17 and effective temperatures ( ) of @xmath18 k . as such , analyzing their physical properties is an important pathway to understanding the properties of gas - giant extrasolar planets .
the subject of this paper is the t dwarf 2mass j15344984@xmath02952274ab , hereinafter , which has an integrated - light infrared spectral type of t5 @xcite . ) . ]
this object was first resolved as a 0.065 binary in august 2000 in _
imaging @xcite . among known visual ultracool binaries in the field , this system has the shortest estimated orbital period , only 4 years ( see compilation in @xcite ) . in combination with its very high quality parallax measurement of @xmath19 mas @xcite and the fact that its two components are nearly equal - magnitude ( indicating nearly equal masses ) , this system is a very appealing object for astrometric monitoring .
laser guide star ( lgs ) ao provides a powerful tool for high angular resolution studies of brown dwarf binaries . through resonant scattering off the sodium layer at @xmath190 km altitude in the earth s atmosphere , sodium lgs systems create an artificial star bright enough to serve as a wavefront reference for ao correction .
thus , most of the sky can be made accessible to near diffraction - limited ir imaging from the largest existing ground - based telescopes .
we have previously used keck lgs ao to discover that the nearby l dwarf kelu-1 is a binary system @xcite and to identify the novel l+t binary sdss j1534 + 1615ab @xcite . in regards to dynamical mass determinations , ground - based telescopes equipped with lgs ao
can provide the necessary long - term platforms for synoptic monitoring of visual binaries , especially where the required amount of observing time at each epoch is relatively modest but many epochs are needed , in contrast to _ hst _ where target acquistion can be slow and monitoring a populous sample over many epochs is quite telescope time - intensive .
we present here the results of multi - epoch imaging of , observed as part of our ongoing high angular resolution study of ultracool binaries using lgs ao . 2
presents our keck lgs ao observations and ( re-)analysis of archival _ hst _
3 presents the resolved photometric properties of the binary and fitting of the orbit using a markov chain monte carlo method . 4 compares the resulting total mass against evolutionary models , and
5 summarizes our findings .
those readers interested solely in the results can focus on 4 and 5 .
we imaged from 20052008 using the sodium lgs ao system of the 10-meter keck ii telescope on mauna kea , hawaii @xcite .
conditions were photometric for all the runs .
we used the facility ir camera nirc2 with its narrow field - of - view camera , which produces a @xmath20 field of view .
setup times for the telescope to slew to the science targets and for the lgs ao system to be fully operational ranged from 720 min , with an average of 12 min ( e.g. * ? ? ?
the lgs provided the wavefront reference source for ao correction , with the exception of tip - tilt motion .
the lgs brightness , as measured by the flux incident on the ao wavefront sensor , was equivalent to a @xmath21 mag star .
tip - tilt aberrations and quasi - static changes in the image of the lgs as seen by the wavefront sensor were measured contemporaneously with a second , lower - bandwidth wavefront sensor monitoring the @xmath22 mag field star usno - b1.0 0601 - 0344964 @xcite , located 31away from . at each epoch , was imaged in filters covering the standard 2.2 atmospheric window from the the mauna kea observatories ( mko ) filter consortium @xcite .
our initial observations in april 2005 were carried out with the @xmath23 ( 2.12 ) filter to minimize the thermal background from the ao system , which is kept at ambient temperature .
subsequent runs employed the @xmath24 ( 2.20 ) or @xmath25 ( 2.15 ) filters .
hereinafter , for brevity we refer to all these data simply as @xmath24-band observations . on each observing run
, we typically obtained a series of dithered @xmath24-band images , offsetting the telescope by a few arcseconds between each 12 images .
the sodium laser beam was pointed at the center of the nirc2 field - of - view for all observations . in april 2005
, we also obtained images with the mko @xmath26 ( 1.25 ) and @xmath27 ( 1.64 ) filters . in april 2008 , we also obtained images with the @xmath28 filter , which has a central wavelength of 1.592 and a width of 0.126 ; this filter is positioned around the @xmath27-band flux peak in the spectra of mid / late - t dwarfs .
the images were reduced in a standard fashion .
we constructed flat fields from the differences of images of the telescope dome interior with and without continuum lamp illumination .
then we created a master sky frame from the median average of the bias - subtracted , flat - fielded images and subtracted it from the individual images .
images were registered and stacked to form a final mosaic , though all the results described here were based on analysis of the individual images .
outlier images with much poorer fwhm and/or strehl ratios were excluded from the analysis .
instrumental optical distortion was corrected based on analysis by b. cameron ( priv .
comm . ) of images of a precisely machined pinhole grid located at the first focal plane of nirc2 .
the 1@xmath29 residuals of the pinhole images after applying this distortion correction are at the 0.6 mas level over the detector field of view . since the binary separation and the imaging dither steps are small , the effect of the distortion correction is minor , smaller than our final measurement errors .
table [ table : keck ] compiles the details of our observations , and figure [ fig : images ] presents our keck lgs data .
full widths at half maxima ( fwhm ) and strehl ratios were determined from two field stars located @xmath1456 from .
the tabulated errors on the fwhm and strehl ratios are the standard deviation of these quantities as measured from the individual images .
to measure the flux ratios and relative positions of s two components , we mostly used the two aforementioned nearby field stars , observed simultaneously with on nirc2 .
these stars provided an excellent measurement of the instantaneous psf .
we empirically modeled the psf using the starfinder software package , which is designed for analysis of blended ao images . for the jan 2008 data
, we employed a different procedure , fitting analytic psfs comprising multiple elliptical gaussians to model the binary images .
these data were taken at much higher airmass than all the other data . because of the larger atmospheric dispersion and the different colors of the field stars relative to , psf fitting produced less accurate results than the analytic approach , as determined by the artificial binary tests described below .
for every image , we fitted for the fluxes and positions of the two components and then computed the flux ratio , separation , and position angle ( pa ) of the binary .
the averages of the results were adopted as the final measurements .
overall , our psf fitting produced very high quality relative measurements , with errors of order 1% for the flux ratios , 0.05 pixels for the binary separation , and 0.2 for the pa .
note that these latter two values account only for the internal instrumental measurements and do not include the errors on the astrometric calibration of nirc2 , which we include below . in order to gauge the accuracy of our measurements , we created myriad artificial binary stars from images of the two psf stars .
one psf star was used to create artificial binaries , and the other was used as the single psf for fitting the components . for data at each epoch ,
starfinder was applied to the artificial binaries with similar separations and flux ratios as .
these simulations showed that any systematic offsets in our fitting code are very small , well below the random errors , and that the random errors are accurate . in cases where the rms measurement errors from the artificial binaries were larger than those from the measurements , we conservatively adopted the larger errors .- band data , and the multi - gaussian fitting was better for the @xmath26 and @xmath27-band data .
overall , fitting the binary images by themselves produced very good astrometry with internal errors ( i.e. , without the uncertainty in the absolute nirc2 astrometric calibration ) of about 3% for the flux ratio , 0.1 pixels for the separation , and 0.3 for the pa but slightly worse than our fits using field stars as psfs .
the exception was the jan 2008 data , as described in the text . ]
to convert the instrumental measurements of the binary separation and pa into celestial units , we used a weighted average of the calibration from @xcite , with a pixel scale of @xmath30 mas / pixel and an orientation for the detector s + y axis of @xmath31 east of north .
these values agree well with keck observatory s notional calibration of @xmath32 mas / pixel and @xmath33 , as well as the @xmath34 mas / pixel and @xmath35 reported by @xcite .
also , comparison of nirc2 images of m92 to astrometrically calibrated _
hst_/acs wide - field camera images gives a pixel scale for nirc2 that agrees to better than 1 part in 10@xmath36 with our values ( j. anderson , priv .
comm . ) .
finally , we must consider the effect of atmospheric refraction .
because of the southern declination of , all of our keck observations were necessarily undertaken at significant airmass ( @xmath151.55 ) . because the two components of the binary do not have exactly the same spectral types ( [ sec : phot ] ) , the observed positions on the sky are subject to slightly different amounts of differential chromatic refraction ( dcr ) .
we computed the expected shift in the relative astrometry at each epoch using the prescriptions of @xcite for the dcr offset and for the refractive index of dry air .
we assumed a fiducial temperature of 275 k and pressure of 608 millibars for conditions on mauna kea @xcite .
we computed the effective wavelengths for spectral types of t5.0 and t5.5 for the two components using using all available spectra of these subclasses contained in the spex prism spectral library ( from @xcite , and @xcite ) and the appropriate filter response curve .
note that because of the unusual spectra of t dwarfs , the effective wavelengths of the secondary component is actually bluer in @xmath27- and @xmath24-bands and redder only in @xmath26-band compared to the primary component .
given the fact that the secondary is mostly north of the primary at all of our keck epochs , dcr causes the separation of the binary to appear slightly smaller at @xmath27 and @xmath24-bands and slightly larger at @xmath26-band compared to the true position as would be observed at zenith .
the amplitude of the dcr effect is about 0.3 mas , much smaller than the measurement errors at most ( but not all ) of the keck epochs .
however , the effect is a systematic one so we correct the relative astrometry of the two components based on our calculations .
table 1 presents the final resulting measurements from our keck lgs data .
for the april 2005 dataset , all three filters give astrometry consistent within the measurement errors ; we use only the @xmath27-band results in the orbit fitting discussed below . in the table and in our orbit
fitting ( [ sec : orbit - fitting ] ) , we take care to discriminate between the instrumental errors ( namely those that arise solely from fitting the binary images ) and the overall astrometric calibration of nirc2 , and thus any future refinements in the latter can be readily applied to our measurements .
the two components of are only barely resolved in the _
hst_/wfpc2 @xmath37 discovery images from august 2000 .
therefore , to determine their relative positions and fluxes , we must model the images using the sum of two blended psfs .
the psf of wfpc2 s planetary camera ( pc ) is undersampled ( fwhm = 1.7 pix for @xmath37 ) ; this makes any empirical determination of the psf difficult without psfs sampled at many subpixel locations .
moreover , anderson & king ( 2003 ) found that the wfpc2 psf varies significantly over the detector due to geometric distortion , making it impossible to construct a reliable empirical psf from other stars in the same image , even if there are enough to sample many subpixel locations .
the original analysis by @xcite employed a hybrid gaussian / empirical psf to fit for the binary parameters with resulting uncertainties of @xmath387 mas in separation and @xmath389@xmath39 in pa .
the astrometry from the wfpc2 discovery epoch is obviously very important to the orbit determination , so we undertook our own analysis with a more accurate psf model to improve the precision of the binary parameters .
we used the tinytim software package ( krist 1995 ) to create model psfs for the wfpc2 images .
we generated 5@xmath40 super - sampled psfs that included the effects of ( 1 ) variation with position on the detector ; ( 2 ) broad - band wavelength dependence , by taking into account the filter response curve and the spectrum of the source ( using the keck / lris optical spectrum of the t4.5 dwarf 2mass j05591914@xmath01404488 from burgasser et al .
( 2003 ) as the template for the individual components of ) , ( 3 ) telescope jitter ( 0 to 20 mas of gaussian jitter ) , and ( 4 ) telescope defocus ( @xmath3810 @xmath41 m ) to account for _ hst _ breathing effects . because the geometrical distortion is location - dependent , we used tinytim model psfs generated for the nearest integer pixel location to the centroid of the binary or single t dwarf .
also , we used the template spectrum closest to the spectral type of the t dwarf with sufficient wavelength coverage ( 0.700.96 @xmath41 m ) from s. k. leggett s spectral library.1404488 for t4.5 and t5.5 objects @xcite ; sdssp [email protected] for t6 objects @xcite ; sdssp [email protected] for t6.5 objects @xcite ; and gl 570d for t7.5 objects @xcite . ]
these tinytim model psfs were used to fit simultaneously for ( 1 ) the location of the primary , ( 2 ) the location of the secondary , ( 3 ) the normalization of the model psf to the primary , and ( 4 ) the flux ratio of the two components .
when fitting positions , the super - sampled tinytim psf was interpolated using cubic convolution to the appropriate subpixel location .
the best fit values were found using the ` amoeba ` alogorithm ( e.g. , * ? ? ?
* ) to find the minimum @xmath43 value of a [email protected] subimage centered on the binary .
the image was cleaned using the idl routine ` cr_reject ` in the goddard idl library to identify and mask the numerous cosmic rays in the undithered wfpc2 image pair .
masked pixels were excluded from the computation of the @xmath43 value .
the noise in each pixel was determined from the bias - subtracted raw wfpc2 image , assuming a read noise of 5.3 e@xmath44/pix and a gain given by the header keyword ` atodgain ` .
a grid of psfs in telescope jitter and defocus were tried , and the fit corresponding to the jitter and defocus combination yielding the lowest @xmath43 was chosen . for images of we found that our psf - fitting routine yielded residuals of @xmath452% of the peak value in 90% of pixels with @xmath46 . due to the optical distortion and `` 34-th row '' defect present in the wfpc2 ( anderson & king 1999 , 2003 )
, the best - fit pixel locations of each binary component do not exactly correspond to their locations in an undistorted celestial reference frame . to remove these effects
, we applied the corrections of anderson & king ( 2003 ) , using a pixel scale of @xmath47 mas / pix .
the 34-th row effect could change , for example , the binary separation by as much as 0.7 mas ( a systematic shift of @xmath00.6@xmath29 ) , but because the binary components do not straddle a defective row and are separated by a mere 1.4 pixels , the application of the 34-th row and distortion corrections have a negligible effect on the astrometry .
with only two undithered wfpc2 images of , it is challenging to quantify the measurement uncertainties , and it is impossible to completely quantify the systematic errors , which arise from an imperfect psf model and also probably depend on the subpixel positions of the two components given the undersampled nature of the data . using only the rms scatter of the two measurements ,
the inferred random errors in separation , pa , and flux ratio are 0.9 mas , 0.07@xmath39 , and 0.03 mag , respectively . to derive more robust random errors and to investigate the systematic errors , we conducted an extensive monte carlo simulation of our fitting routine .
we used wfpc2 @xmath37 images of seven other t dwarfs from the same _ hst _ program , all apparently single , to create artificial binaries which we then modeled using our psf - fitting routine.1404488 ( t4.5 ) , 2massi j0937347@xmath42293142 ( t6.0 ) , 2massi j1217110@xmath0031113 ( t7.5 ) , 2mass j12373919@xmath426526148 ( t6.5 ) , gl 570d ( t7.5 ) , 2massi j1546291@xmath0332511 ( t5.5 ) , and 2massi j2356547@xmath0155310 ( t5.5 ) , using spectral types from @xcite . the remaining object from this program , 2massi j1047538@xmath42212423 ( t6.5 ) , was unusable for our purposes because it landed on a bad column . ] because the wfpc2 psf is severely undersampled , we only created artificial binaries with integer - shifted positions .
it turns out that the wfpc2 locations of the two components of are at a very nearly integer - shifted separation of 1.4 pixels ( @xmath48 pix , @xmath49 pix ) .
therefore , in determining the uncertainties and systematic offsets we used only the configuration most nearly matching that of with a separation of @xmath50 pix and instrumental pa of 225@xmath39 . we found that using any or all of the other three @xmath50 configurations gave consistent uncertainties .
after subtracting the best - fit model , artificial binary images yielded residual images in which 90% of pixels with @xmath46 were below 13% of the peak flux , comparable to the residual images of . the images of the single t dwarfs span a range in @xmath51 from about 1.5 mag brighter to 1.3 mag fainter than the primary component of .
we used these images at their native @xmath51 when simulating the primary component . to simulate the secondary component , we degraded the @xmath51 of the images assuming a flux ratio of 0.30 mag .
we also tried flux ratios of 0.25 mag and 0.35 mag to explore the possibility that the uncertainties depend on the assumed flux ratio , but we found that this had an insignificant effect on our predicted uncertainties ( @xmath52 , where @xmath53 and @xmath54 is the number of simulations ) .
signal - to - noise degradation of an image was done by a multiplicative scaling followed by the addition of normally distributed random noise to each pixel , according to the same wfpc2 noise model we used to determine @xmath43 in the psf - fitting procedure .
in fact , by running simulations where the primary images were degraded to much lower @xmath51 , we found that all of the single t dwarfs are in a high @xmath51 regime in which systematic errors ( psf model imperfections ) dominate , not random errors ( photon noise ) our simulations showed no dependence between the @xmath51 of the t dwarf used to construct artificial binaries and the resulting astrometric uncertainties .
therefore , we used the rms of the results from all simulated binaries in order to determine the final uncertainties for .
as expected , the uncertainties in separation , pa , and flux ratio from our simulations were somewhat larger than those derived from the standard deviation of the measurements from the two images , since both random and systematic errors have been evaluated in the simulations .
in fact , the larger uncertainties are not simply due to averaging over , for example , the many subpixel locations of the single t dwarfs used in the monte carlo , because the simulated measurements for each single t dwarf show scatter consistent with the final derived uncertainties .
table [ table : hst ] presents our final results for the wfpc2 images , with systematic offsets from the monte carlo simulations applied .
our astrometry agrees well ( better than 1@xmath29 ) with the original results of @xcite , though our measurement errors are a factor of 8 smaller .
part of this improvement comes from our use of tinytim - computed psfs , as opposed to the simpler burgasser et al .
psf model of a gaussian plus empirical residuals .
we also used all possible single psfs in our artificial - binary simulations , whereas burgasser et al .
used only wfpc2 images of 2mass j0559@xmath01404 , a source that is suspected to be an unresolved binary.1404 is roughly twice as luminous as objects of similar spectral type @xcite .
high angular resolution observations have not detected any multiplicity ( @xcite ; liu et al .
, in prep . ) , but the source could be a very tight system , e.g. , with a 0.5 pixel separation .
if so , its multiplicity could confuse any attempts to fit only two single psfs to the tightest artificial binaries constructed from its image .
interestingly , we found that at the smallest ( 1.0 pixel ) separations , artificial binaries constructed from images of 2mass j0559@xmath01404 yielded extremely large uncertainties ( 5 mas and 15 ) .
we did not observe such behavior for any of the other six apparently single t dwarfs , nor were the uncertainties for larger separation binaries made from 2mass j0559@xmath01404 abnormally large .
one natural explanation would be that the source is just marginally resolved in the wfpc2 imaging . ] and part of the improved uncertainties is somewhat illusional , as it arises from the different parameter space explored in monte carlo simulations by us and burgaser et al . as a check , we ran a suite of simulations more comparable to that of burgasser et al . ,
in which the ranges of artificial binary parameters were 1.03.0 pixels in separation , the full range of pas , and 0.01.0 mag in flux ratio .
these yielded similar astrometric uncertainties to the published errors , suggesting that the smaller uncertainties we derive are due to our more restricted choice of artificial binary configurations ( @xmath55 pix , @xmath56 pix ) and/or averaging over many single psfs to reduce systematics associated with any one specific object .
our improvement to the wfpc2 astrometry was essential in our early attempts to fit the orbit based on keck data obtained in 20052007 .
however , with the addition of data in 2008 , the final _ hst_+keck dataset has sufficent time baseline and astrometric quality that the choice of wfpc2 astrometry does not highly impact the final orbit fitting results ( [ sec : altorbits ] ) .
the system was observed on 2006 january 19 and 2006 april 11 ( ut ) with the high resolution camera ( hrc ) of _ hst _ s advanced camera for surveys ( acs ) by program go-10559 ( pi h. bouy ) .
the binary is much more widely separated at these epochs than in the wfpc2 observations , but the psfs of the two components are still blended .
we have therefore applied the same tinytim psf - fitting technique described in the previous section to derive the relative astrometry from the acs images .
the primary differences between the wfpc2 and acs datasets are : ( 1 ) acs has much more severe geometric distortion than wfpc2 , which changes the shape of the psf and complicates astrometry because the pixels projected on the sky are not square , and ( 2 ) the acs data are of much lower @xmath51 , with a total exposure time of only 50 sec for each cosmic - ray rejected , combined dithered image ( cf .
, 1300 sec for a single wfpc2 image ) .
because of the lower @xmath51 , we found it unwarranted to fit the acs images for telescope jitter and defocus as adding these free parameters did not improve the quality of the fits ( as verified in the monte carlo simulations discussed below ) .
also , we found that the @xmath1425@xmath40 lower @xmath51 of these data almost exactly negates any improvement to the astrometry that might be expected given the larger binary separation at these epochs .
we used distorted model psfs generated by tinytim to fit for the position and flux of each binary component in images that had been cosmic - ray cleaned ( ` crsplit=4 ` ) by the latest _ hst _ pipeline .
best - fit pixel locations were corrected for geometric distortion using the solution of anderson & king ( 2004 ; instrument science report 04 - 15 ) , and we used their measured acs pixel scale , which was derived by comparing commanded ( ` postarg ` ) offsets of _ hst _ in arcseconds to the resulting pixel offsets .
they derived two such pixel scales for two epochs of observations of 47 tuc , and we adopt the mean and standard deviation of these two : @xmath57 mas / pix . which is well below the other errors in the measurements ( j. anderson , priv .
comm . ) . ] for each epoch , we adopted the mean of the measurements from all four dithered images for the binary parameters of ( table [ table : hst ] ) . again ,
to investigate the measurement errors thoroughly , we performed monte carlo simulations of our fitting routine .
we used images of single brown dwarfs to construct artificial binaries in configurations resembling . at both
acs epochs , the binary is well represented by integer - pixel shifts on a grid where @xmath58 and @xmath59 .
there have been no _ hst_/acs science programs dedicated to studying single brown dwarfs ; however , acs images of the single brown dwarfs 2mass j00361617@xmath421821104 ( l3.5 ; @xcite ) and 2mass j05591914@xmath01404488 ( t4.5 ; @xcite ) were obtained for calibration purposes and are available in the _ hst _ archive ( cal / acs-10374 , pi giavalisco ) .
we found that despite the large difference in spectral types , any corresponding difference in the psf does not alter the results of the monte carlo simulations .
the @xmath51 of each of these single objects is much higher than that of , so we degraded the @xmath51 of the single brown dwarfs for the artificial binary simulations .
in fact , by varying the @xmath51 of the simulations , we found that the acs data for are well in the @xmath51 regime dominated by random photon noise , while the images of the single objects are in a high @xmath51 regime dominated by systematic errors ( akin to the wfpc2 images of ) .
therefore , given that we have four images , we divide the rms of the monte carlo results by @xmath60 to represent the final uncertainties .
table [ table : hst ] contains our final acs results .
note that the acs data are contemporaneous with our keck lgs data , and observations by the two telescopes in 2006 separated by less than one month show excellent agreement .
however , the relatively low @xmath51 of the acs data means it has the larger astrometric errors .
our simulations confirm that if the acs data were of higher @xmath51 ( i.e. , longer exposure times than 50 s ) , the resulting astrometric precision would be much better than , instead of comparable to , the wfpc2 astrometric uncertainties .
we obtained low - resolution ( @xmath61150 ) integrated - light spectra of on 2008 may 16 ut from nasa s infrared telescope facility ( irtf ) located on mauna kea , hawaii .
conditions were photometric with seeing of about 0.7 fwhm near the target .
we used the facility near - ir spectrograph spex @xcite in prism mode , obtaining 0.82.5 spectra in a single order .
we used the 0.5 wide slit , oriented at the parallactic angle to minimize the effect of atmospheric dispersion .
was nodded along the slit in an abba pattern , with individual exposure times of 180 sec , and observed over an airmass range of 1.641.60 as it rose .
the telescope was guided during the exposures using images obtained with the near - ir slit - viewing camera .
the total on - source exposure time was 720 sec .
we observed the a0 v star hd 142851 contemporaneously for flux and telluric calibration .
all spectra were reduced using version 3.4 of the spextool software package @xcite .
the reduced irtf / spex spectrum is plotted in figure [ fig : spectra ] and compared to t dwarf spectral standards from @xcite .
visual examination shows an excellent match to the t5 spectral standard 2mass j1503 + 2525 , as does measurement of the @xcite spectral indicies for : @xmath26 = 0.271 ( t4.8 ) , @xmath26 = 0.420 ( t4.8 ) , @xmath27 = 0.345 ( t5.0 ) , @xmath27 = 0.430 ( t5.0 ) , and @xmath24 = 0.224 ( t5.1 ) , with spectral type estimates based on the polynomial fits to the indices from @xcite .
we use our measured flux ratios and the published @xmath62 photometry from @xcite to derive resolved ir colors and magnitudes for on the mko system .
we use the _
hst_photometry from @xcite in determining the resolved @xmath37 magnitudes .
then to infer spectral types for the individual components , we compare these to magnitudes and colors of ultracool dwarfs from @xcite and @xcite , excluding known binaries .
we use near - ir spectral classfications from the @xcite scheme .
we assume that the components of are themselves single , not unresolved binaries .
figure [ fig : colorcolor ] shows that component a has ir colors most typical of t4.5t5 dwarfs , and component b is most similar to t5t6 dwarfs .
the individual absolute ir magnitudes ( given in table [ table : resolved ] ) give similar results .
the `` faint '' polynomial fits for absolute magnitude as a function of spectral type from @xcite give @xmath63 ) = \{14.5 , 14.6 , 14.7 , 14.9 , 15.2 } mag , @xmath64 = \{14.6 , 14.8 , 15.0 , 15.2 , 15.5 } mag , and @xmath65 = \{14.7 , 14.9 , 15.1 , 15.4 , 15.7 } mag for near - ir spectral types of t4.5 , t5 , t5.5 , t6 , and t6.5 , respectively .
averaging the same data for each individual subclass gives @xmath66 = \{[email protected] , 14.1 , [email protected] , [email protected] , [email protected] } mag , @xmath64 = \{[email protected] , 14.2 , [email protected] , [email protected] , [email protected] } mag , and @xmath65 = \{[email protected] , 14.3 , [email protected] , [email protected] , [email protected] } mag , where the uncertainties are the rms of the photometry for each subclass ( and no listed uncertainties for subclasses with only one object ) .
altogether , the absolute magnitudes suggest types of t5t5.5 for component a and t5.5t6 for component b. the resolved @xmath67 colors provides a third means to estimate the spectral types , as these track the @xmath27-band methane absorption , which correlates well with overall near - ir spectral type ( e.g. , figure 2 of @xcite ) . first
, we compute individual @xmath28 magnitudes for using the @xmath28 flux ratio from our lgs images , the integrated - light photometry of @xmath68 mag from @xcite , and an integrated - light color of @xmath69 mag synthesized from the near - ir spectrum of @xcite . including the measurement errors in the flux ratios and @xmath27-band photometry , we find @xmath70 and @xmath71 mag and @xmath72 and @xmath73 mag for components a and b , respectively .
note that the _ relative _ @xmath67 color of the two components is known to higher precision , since the above computed colors for the two components contain the same 0.03 mag error that originates from the integrated - light @xmath27-band photometry .
( in other words , the 0.05 mag uncertainties in the colors of the two components are not independent , but correlated . )
removing this effect gives a relative color of @xmath74 mag between a and b , i.e. , greater methane absorption in component b is detected . to determine the behavior of @xmath67 with near - ir spectral type , we synthesized colors from the spex prism spectral library collection , which contains low - resolution spectra of 68 t dwarfs after removing spectrally peculiar objects and known binaries . figure [ fig : ch4short ] plots the results .
we fit a 2nd - order polynomial for the dependence of color on near - ir spectral type and vice - versa : @xmath75 where @xmath76 for t0 , = 21 for t1 , etc .
the rms scatter about the fits are 0.02 mag and 0.3 subclasses , respectively . using these polynomial relations , the observed @xmath77 )
colors give spectral types of t4.5 @xmath38 0.7 and t5.6 @xmath38 0.6 for components a and b , respectively , where the spectral type uncertainties come from formal propagation of the uncertainty in the colors . in addition , just as the relative @xmath67 color of the two components are known more accurately than the absolute colors , we compute a relative spectral type of @xmath78 subclasses between components a and b. combining the inferences from the @xmath62 colors , the absolute magnitudes , and the @xmath67 colors , we adopt spectral type estimates of t5 @xmath38 0.5 and t5.5 @xmath38 0.5 for the two components .
the relative @xmath79 color favors a slightly larger spectral type difference than the absolute magnitudes but consistent with the adopted uncertainty .
( also the difference of the two components computed in
[ sec : temp ] favors a 0.5 subclass difference . )
higher - order multiple systems are very rare among ultracool binaries , with an estimated frequency of @xmath80 @xcite , and thus a priori we do not expect to fall into this category .
the colors and magnitudes are consistent with the system being composed of only two components , and not being a partially resolved higher order multiple system .
if component b was actually an equal - mass binary , the absolute magnitudes of its components would be 0.75 mag fainter than the integrated - light of b , meaning @xmath81 mags .
this would suggest a spectral type around t7 , based on the polynomial relations in @xcite , which is clearly too late - type compared to the integrated - light spectrum and the observed near - ir colors of b. to measure the for the system , we combine our spex 0.92.4 spectrum with the published integrated - light @xmath37 , -band , and _ spitzer_/irac thermal - ir photometry and uncertainties @xcite.-{\mbox{${l^\prime}$}})$ ] color is @xmath10.2 mag redder than any other t dwarf and @xmath10.3 mag redder than objects of similar spectral type .
both of these bandpasses are affected by the 3.3 fundamental band of , with the spitzer [ 3.5 ] data being subjected to greater absorption , and thus the redder color might point to anomolously strong .
however , the near - ir spectrum and the @xmath62 and spitzer 3.57.9 colors are consistent with other mid - t dwarfs , and thus perhaps the -band photometry is incorrect .
either way , this has a negligible effect on the computed .
] we flux - calibrated the spex data using the published @xmath27-band mko photometry from knapp et al .
spitzer_/irac data , we adopted the photometric calibration and an overall 2% absolute uncertainty based on reach et al .
we extended the binary s spectral energy distribution ( sed ) to short wavelengths by linearly interpolating from the @xmath37 datum to zero flux at zero wavelength and to long wavelengths by assuming a rayleigh - jeans spectrum beyond the reddest spitzer bandpass ( 7.87 ) ; this extrapolation increases the total flux by 2% .
we then integrated the sed , using a monte carlo approach to account for all the measurement errors .
we find @xmath82 dex for the system , with the uncertainty increasing to 0.018 dex after including the uncertainty in the distance .
( as discussed in 4 , we keep track of these independent uncertainties in our calculations . ) the largest uncertainty in the integration arises from the 0.03 mag uncertainty in the integrated - light photometry used to normalize the spex spectrum .
we cross - checked our method using the same data for the t4.5 dwarf 2mass j0559@xmath01404 and found excellent agreement with the measured by @xcite using absolutely flux - calibrated spectra from 0.615 .
the computed total agrees well with that inferred from using the @xmath24-band bolometric corrections ( @xmath83 ) from golimowski et al.(2004 ) , namely using the resolved @xmath24-band absolute magnitudes and the estimated spectral types , which would give @xmath84 and @xmath85 dex for the individual components and thus @xmath86 dex for the total system .
however , the uncertainties are larger when using @xmath83 to derive , since this incorporates the uncertainties arising from the 0.5 subclass uncertainty ( 0.06 mag in bolometric magnitude ) and the intrinsic scatter in the golimowski et al . @xmath83 relation ( 0.13 mag ) . in short , direct integration of the observed sed
is more accurate . to apportion the observed total into the individual components ,
we assume the observed @xmath24-band flux ratio of the system represents the luminosity ratio .
this would be exactly correct if the two components had identical spectral types ( and neglecting photometric variability ) . to account for the difference in spectral types , we generate a monte carlo distribution of @xmath83 values for each component using the golimowski et al .
polynomial fit as a function of spectral type subject to the following rules : the spectral type of component a is uniformly distributed from t4.5t5.5 ; the spectral type of component b is no later than t6 ; and the difference in their spectral types is at least 0.5 subclasses .
this produces an average difference in @xmath83 between the two components of 0.09 mags and an rms of 0.03 mags .
thus , we find @xmath87 and @xmath88 dex for the two components , including the uncertainty in the distance .
we have data at 9 independent epochs , which is formally sufficient to determine the 7 parameters of a visual binary orbit given our measurements ( 9 positions + 9 times ) .
however , two pairs of measurements are separated by only one month ( april / may 2006 and march / april 2007 ) , and the cadence of the orbital phase covered is limited , with the _ hst_/wfpc2 datum being taken almost 5 years before the next epoch .
while standard gradient - descent ( levenberg - marquardt ) least - squares techniques would be sufficient to derive an orbit ( [ sec : fitting ] ) , we also would like to accurately determine the probability distribution of the orbit parameters ( which may not be normally distributed ) and the associated degeneracies .
for epochs with keck data taken in multiple filters , we choose the filter with the smallest astrometry errors .
thus , we first used a combination of gradient - descent techniques from random starting points and simulated - annealing techniques to isolate the class of potential orbital solutions near a reduced chi - squared ( ) of 1 .
then , to fully explore this class of solutions , we used a markov chain monte carlo ( mcmc ) approach ( e.g. * ? ? ? * ) .
mcmc provides a means to explore the multi - dimensional parameter space inherent in fitting visual orbits that is computationally efficient , able to discern the degeneracies and non - gaussian uncertainites in the fit , and allows for incorporation of a priori knowledge . in short ,
the mcmc approach is distinct from ordinary monte carlo methods in that instead of a completely random steps through the model parameter space , the steps are chosen such that the resulting number of samples ( the `` chain '' ) is asymptotically equivalent to the posterior probability distribution of the parameters being sought .
( see @xcite , @xcite , and @xcite for explications of applying mcmc to astronomical data . )
we parameterized the binary s orbit using the standard 7 parameters : period ( @xmath89 ) , semi - major axis ( @xmath90 ) , inclination ( @xmath91 ) , epoch of periastron ( @xmath92 ) , pa of the ascending node ( @xmath93 ) , which can only be resolved with radial velocity information . ] , eccentricity ( @xmath94 ) , and the argument of periastron ( @xmath95 ) .
we used an mcmc chain length of @xmath96 , with the parameters stored every hundredth iteration .
we started the chain at the global minimum found by the simulated - annealing and gradient - descent algorithms .
we used the metropolis - hastings algorithm to sample the joint probability distribution with a variant of the usual gibbs sampler . instead of chosing one of the 7 parameters to increment or decrement , we chose instead to move randomly either forwards or backwards along one of 7 orthogonal directions in the parameter space .
these directions were initially chosen to be along the individual coordinate axes of our 7 parameters . then , every @xmath97 iterations , the covariance matrix of the 7 parameters was calculated and new directions chosen along the directions corresponding to the eigenvectors of the covariance matrix .
after covariance matrix calculation , the trial step size was set to be the square root of the covariance matrix eigenvalues .
this enabled the long , thin , curved minima in our parameter space to be sampled much more efficiently .
each set of @xmath97 iterations can thus be thought of as its own sub - chain with fixed directions .
as the first @xmath97 iterations uses a less efficient sampling ( sampling directions aligned with the 7 parameters themselves ) , we treated this part of the chain as the `` burn - in '' time , neglecting it in the final analysis .
as the chain was running , the trial step size in each direction was continuously scaled over a timescale of 700 steps so that the success rate of jumps averaged to 0.25 .
the correlation length of our most correlated chain , as defined in @xcite , was @xmath98 for the orbital period ( @xmath89 ) , with equal or smaller correlation lengths for other parameters .
this gives an effective length of the chain of @xmath99 , which in turn gives statistical uncertainties in the parameter errors of about @xmath100% , i.e. , negligible .
these uncertainties in the errors are consistent with the results from running multiple test chains . by making steps of the same size in the positive and negative directions for these parameters in constructing the markov chain
, we would implicitly assume that our prior knowledge of these parameters is a uniform distribution .
this is not an accurate representation of our prior knowledge since , for example , binaries with periods between @xmath101 and @xmath102 years are not 10 times more common than binaries with periods between 10 and @xmath101 years .
we therefore applied a prior to the likelihood function in the mcmc fitting where @xmath89 and @xmath90 are distributed evenly in logarithm and that the parameters @xmath103 and @xmath104 are uniformly distributed , rather than @xmath94 and @xmath95 , to save the algorithm from unnaturally preferring circular solutions .
this is equivalent to the @xmath105 distribution as discussed by , e.g. , .
the very small effect of the choice of prior is discussed below .
larger errors .
( in fact , the mcmc fitting gives basically the same results if the acs data are excluded . )
thus simply applying the nirc2 astrometric calibration error of 0.1% in pixel scale and 0.07 in pa to all six keck epochs would incorrectly treat this as a random error , when in fact the uncertainty in the nirc2 calibration globally impacts the overall solution of , e.g. , the semi - major axis and the pa of the line of nodes . therefore , we do the following : ( 1 ) we apply the errors in the nirc2 calibration in quadrature to the _ hst_/acs and _ hst_/wfpc2 measurements before orbit fitting , and ( 2 ) once the semi - major axis distribution has been determined , we apply the uncertainty in the nirc2 pixel scale in determining the error on the total mass .
the net effect is negligible , given the much larger acs errors compared to the keck lgs data and the fact that the error in the total mass is dominated by the parallax error ( which is @xmath106 larger than the nirc2 pixel scale uncertainty ) . ] as a consistency check , we also ran our mcmc fitting code on the astrometric data for the binary l dwarf 2massw j0746 + 2000ab from .
we found excellent agreement between the orbital parameters derived by us ( using mcmc ) and by bouy et al.(using a variety of chi - square minimization approaches ) .
not only do the results agree to within the quoted errors , there is better than @xmath141% agreement on the best - fit results and better than @xmath1420% agreement on the 95% confidence intervals .
figures [ fig : orbit - summary ] show the resulting probability distributions for the orbital parameters from the mcmc chain .
the probability distributions are clearly not gaussian . for a given parameter
, we adopt the median as the result and describe a confidence limit of @xmath107% as simply the bounds of the sorted sample . at 68(95)% confidence
, we find a modest eccentricity of 0.25@xmath108 , an orbital period of 15.1@xmath109 yr and a semi - major axis of 171@xmath110 mas ( 2.3@xmath111 au including the uncertainties in the plate scale and parallax ) .
two of the orbital angles are very well - constrained , the inclination @xmath112 deg ( nearly edge - on ) and the pa of the ascending node @xmath113 deg .
the final results are summarized in table [ table : orbits ] .
figure [ fig : orbit - p_e ] shows the strong correlation between the determination of the orbital period and the eccentricity .
it illustrates that there are two classes of possible orbits : one branch having shorter periods and smaller semi - major axes and the other branch having longer periods and larger semi - major axes .
figure [ fig : orbit - p_omg ] shows that the short-@xmath89 branch orbits have just passed apoastron ( @xmath114 , @xmath115 ) .
this is the favored solution , with 98% of the steps in the mcmc chain residing in this branch ( using @xmath116 as the dividing criteria in the @xmath117 plane ) .
however , a nearly circular orbit means it can be difficult to clearly distinguish whether the system has just passed apoastron or periastron , and thus a minority of the mcmc steps ( 2% ) fall into the long - period branch . larger than the projected separation .
the large discrepancy with our orbital period determination is not surprising , since a single epoch of imaging provides a highly uncertain period estimate the likely true periods can span a factor of several greater or smaller ( e.g. * ? ? ?
] the mcmc fitting provides probability distributions for the orbital parameters , but does not provide a single best - fitting orbit per se , since a range of possible orbits fit the data with similar values .
one illustration of this is the result for @xmath95 = 179@xmath118 deg , where the 95% confidence limits are broad enough to span both the short - period and long - period solution branches .
thus to plot orbits on the sky , we employ gradient - descent methods to find the best - fitting orbit with the mcmc - derived values as the starting point .
figures [ fig : orbit - orbit ] and [ fig : orbit - sep ] shows the resulting orbit , which has a period of 15.2 yr , a total mass of 0.0556 and @xmath119 . to illustrate how the uncertainty in the orbital period impacts the orbit
, we also show the best - fitting orbits found when fixing the period to 12 and 20 yr , which have total masses of 0.0523 and 0.0590 and @xmath120 of 1.1 and 1.0 , respectively .
all three orbits reside in the short - period branch and show that the projected separation is now rapidly decreasing .
the system is expected to be well - resolved again in the year 2011 .
applying kepler s third law to the period and semi - major axis distributions gives the posterior probability distribution for the total mass of the binary , with median of 0.0556 , a standard deviation of 0.0018 ( 3.2% ) , and a 68(95)% confidence range of about @xmath380.0018(0.0037 ) ( figure [ fig : orbit - masses ] ) .
however , the mcmc probability distribution does not include the uncertainties in the parallax ( 1.6% ) and the nirc2 pixel scale ( 0.11% ) . by kepler s third law ,
the quadrature sum of these errors amounts to an additional 4.9% uncertainty on the derived total mass .
since the mcmc - derived mass distribution is asymmetric , we account for this additional error in a monte carlo fashion ; for each step in the chain , we draw a value for the pixel scale and parallax from a normal distribution and then compute the total mass .
the resulting mass distribution is essentially gaussian ( figure [ fig : orbit - masses ] ) .
our final determination of the total mass is @xmath121 at 68(95)% confidence .
thus , the total mass of this system is well - measured , with the parallax error being the dominant uncertainty .
this is the coolest and lowest mass binary with a dynamical mass determination to date .
the wpfc2 discovery epoch in 2000 is obviously a key component to fitting the orbit . as described in
[ sec : wfpc2 ] , we independently analyzed this dataset and greatly reduced the measurement errors compared those reported by @xcite . to examine the impact of this improvement
, we also tried fitting the orbit using the original burgasser et al .
astrometry . without the 2008 keck data ,
our improved wfpc2 astrometry is essential for a well - constrained fit .
however , with the complete dataset , the fitted orbital parameters and the total mass are insensitive to the specific choice of wfpc2 astrometry , changing by less than 1@xmath29 . also , to check the effect of our assumed prior on the mcmc fitting ( flat in @xmath122 and @xmath123 ) , we tried three alternative priors from the literature : 1 .
_ solar - type stars : _ analyzed a well - defined sample of 164 nearby solar - type stars ( spectal type f7 to g9 ) and found a log - normal distribution in orbital period : @xmath124\ ] ] where @xmath89 is the period in days and @xmath125 .
2 . _ ultracool visual binaries : _
@xcite conducted a detailed analysis of published imaging surveys of 361 ultracool field objects to model the separation distribution as a log - normal distribution : @xmath126\ ] ] where @xmath90 is the semi - major axis in au and @xmath127 .
events ) , perhaps at variance with spectroscopic binary studies @xcite .
the young eclipsing m6.5+m6.5 binary 2mass j0535@xmath005 @xcite is also highly anomalous with this distribution ( @xmath128 event ) . finally , the allen analysis does not separately consider objects of different spectral types , whereas @xcite have suggested the separation distribution for t dwarfs may be tighter than for the l dwarfs .
thus our default prior of flat in @xmath123 is a conservative choice .
note that longer period of we find relative to the original estimate of @xmath14 years ( @xmath129 au ) is in accord with the distribution proposed by allen , which would indicate that such a @xmath11.1 au system would be uncommonly rare ( @xmath130 ) relative to the @xmath1100 known ultracool binaries . ]
_ ultracool visual and spectroscopic binaries : _
@xcite analyzed a sample of 47 ultracool binaries with multiple radial velocity measurements and adopted a log - normal distribution truncated at large separations to match the known visual ultracool binaries : @xmath131 \qquad \hbox{\rm for $ a < 10$~au}\ ] ] where @xmath90 is the semi - major axis in au , @xmath132 , and @xmath133 for @xmath134 au .
figure [ fig : priors ] shows the posterior probability distributions for the total mass from the different priors .
overall , the choice of prior has very little effect on the mass determination .
the @xcite distribution favors slightly higher masses , but its results are consistent with the other priors . as a final independent check
, we also fit our astrometry using the linearized least - squares fitting routine orbit , using the mcmc - derived parameters as the starting guess .
the orbit results are given in table [ table : orbits ] , with a resulting @xmath119 , and show excellent agreement with the mcmc results .
a primary goal of measuring fundamental properties for ultracool binaries is to compare the measurements against theoretical models of their physical properties .
a number of studies have been published for the previous ultracool visual binaries with dynamical masses ( 1 ) , with subtle and/or overt differences in the ways that observations are compared to models . in the analysis that follows
, we strive to clearly elucidate the comparison of our observations to the models , both in terms of its approach and limitations . from the standpoint of the observations
, we have high quality measurements of ( 1 ) the total mass of the system and ( 2 ) the individual absolute magnitudes , with ( 3 ) the individual bolometric luminosities only slightly less reliable .
( we have not measured the complete spectral energy distribution but have accounted for this uncertainty in computing in [ sec : lbol ] ) .
we now examine what can be learned from these data in concert with evolutionary models and theoretical atmospheres . the most immediate result from our measurement
is that is a bona fide brown dwarf binary .
the total mass of @xmath135 is below the solar - metallicity stellar / substellar boundary of @xmath140.0700.074 ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , with the boundary increasing to higher masses for lower metallicities @xcite therefore , the individual components are clearly substellar .
this is the second binary where both components are directly confirmed to be brown dwarfs , after the young eclipsing m6.5+m6.5 binary 2mass j0535@xmath005 @xcite , and is the first such field binary in this category .
brown dwarfs follow a mass - luminosity - age relation .
we have measured two of these quantities , the ( total ) mass and the luminosity , and by using evolutionary tracks we can determine the third quantity , the age of the system .
we use models from the tucson group @xcite , which provide predictions for , and the `` cond '' models from the lyon group , which predict both and absolute magnitudes .
we conservatively assume the system is coeval and that the system is a true binary , not a partially ( un)resolved higher order multiple system ( [ sec : phot ] ) . for each tabulated model age
, we use the individual absolute magnitudes and/or bolometric luminosities to calculate the mass of the components and then sum the masses .
we then apply the observed total mass to determine the age range of the system .
all measurement errors in and the total mass are accounted for in a monte carlo fashion , namely we repeat the model calculations over multiple realizations for the and the total mass values .
we take great care to account for the covariance between the relevant quantities in the calculation . for instance , the total mass of the system and the luminosity both depend on the parallax , and thus their errors are positively correlated ; we therefore draw the parallax values from a normal distribution and incorporate these in determining the monte carlo distribution of total masses and luminosities , which themselveed are then propagated in the model - based calculations .
this approach results in a probability distribution for the system s age ( as well as the other resulting parameters discussed below ) , which we summarize with the median value and confidence limits .
figure [ fig : models - totalmass ] shows the results of these calculations to determine the age of the system .
for a consistent comparison between the lyon and tucson models , we use only the results derived from the measurements .
however , the figure also shows that using the absolute magnitudes predicted by the lyon models would give similar results .
we determine an age of @xmath136 gyr from the burrows models and @xmath137 gyr from the baraffe models at 68(95)% confidence . to construct a representatve `` average '' of the model results , we merge the results of the individual monte carlo calculations into a single distribution and compute its confidence limits .
thus , we assign an age of ( table [ table : evolmodels ] ) .
this is relatively youthful compared to the main - sequence stars in the solar neighborhood , e.g. , @xmath1495% of nearby solar - type stars have age estimates of @xmath161 gyr .
however , the mean age of t dwarfs is expected to be younger than for field stars , since the known census is magnitude - limited and younger objects are brighter .
the age distribution of field ultracool dwarfs has been modeled by @xcite and @xcite ; they generally find that field t dwarfs can span younger ages than for low - mass stars , though the predicted age distributions for both types of objects have large spreads .
kinematics provide an independent ( albeit indirect ) indicator of age , as older objects are expected to generally show larger space motions due to their accumulated history of dynamical interactions .
the tangential velocity of ( @xmath138 ) is the second smallest measured for t dwarfs , with only the t5.5 dwarf 2mass 1546@xmath03325 being smaller ( @xmath139 ; @xcite ) .
this is generally in accord with the @xmath140 gyr inferred from the evolutionary models , namely that is among the youngest members of the nearby field population .
however , since the measured @xmath141 distribution of field t dwarfs is quite broad , with an unweighted average of 38.4 and a standard deviation of 20.4 among the 21 unique objects in the @xcite and @xcite parallax samples , does not appear to be anomalously young for a field object .
a radial velocity measurement is needed to determine the binary s space motion and thus better constrain its kinematics . with the age of the system determined above ,
the combination of the observations and the evolutionary models provide highly precise values for the remaining physical parameters .
the results derived from the two sets of evolutionary tracks are given in table [ table : evolmodels ] and are computed from the same monte carlo approach that accounts for the covariance in the measurements .
the tucson and lyon models give consistent values , with the tucson models giving slightly larger radii and thus slightly cooler temperatures .
again , to compute a representative `` average '' for each parameter , we merge the monte carlo distributions computed from each set of models and compute confidence limits for the aggregate .
this is not intended to be physically meaningful , but rather is a quantitative representation of the results that accounts for non - gaussian and/or inconsistent distributions from the two sets of models .
we thus find radii of and , effective temperatures of @xmath142 k and @xmath7 k , and surface gravities of and @xmath143 for components a and b , respectively .
to re - iterate , these properties are derived using only the measured _ total _ mass and resolved magnitudes / luminosities , along with the assumption that the system is coeval and composed of two components .
no additional assumptions have been made to determine the individual masses .
we also have avoided using spectral types and/or effective temperatures in this aspect of our analysis ( contrary to some previous studies ) , as these quantities can introduce additional systematic errors and/or circular reasoning .
for instance , it would be incorrect to employ the relations between spectral type and from @xcite or @xcite to determine for the two components and then compare to evolutionary tracks , as the vrba et al . and golimowski et al
. relations are derived from the radii of field brown dwarfs predicted by the evolutionary tracks themselves .
likewise , it is not necessary to use determinations from atmospheric models to determine the age of the system or the physical properties of the individual components in our approach .
( see also [ sec : tests ] . ) as already noted above , our monte carlo calculations account for the covariance in the measurements .
one important effect is that the uncertainties derived from the resolved magnitudes and luminosities of the two components are correlated , since they all depend on the measurement uncertainties in the integrated - light photometry of the system . as a consequence ,
the _ relative _ temperture difference between the two components ( @xmath144 k ) can be calculated to higher precision than would be indicated by the uncertainties in the individual determinations ( @xmath145 k = 24 k ) .
this @xmath146 agrees with the @xmath1470 k difference expected from the @xcite polynomial fits for the 0.5 subclass difference between the two components .
( the @xmath146 from the golimowski et al .
fits would be about twice as large for a 1 subclass difference . )
the physical parameters for 2mass j1534@xmath02952a and b are in general agreement with previous determinations for the properties of field t dwarfs .
however , our values have much higher precision , because the accurate total mass measurement leads to a small age range , which leads to strong constraints on the radii and thus small uncertainties on the derived and values .
( we discuss this further in
[ sec : tests ] . )
t dwarf surface gravities have been inferred to be @xmath147 by comparing theoretical model atmospheres to optical spectra @xcite , near - ir colors and line strengths @xcite , and low - resolution near - ir / mid - ir spectra @xcite .
this range encompasses our determinations for . on the other hand ,
our very precise temperatures for the two components of are discrepant with previous studies of t dwarfs : the s of appear to be cooler than determined previously for mid - t dwarfs .
these discrepancies occur for two _ separate _ comparisons .
temperatures for field t dwarfs have been inferred by combining accurate determinations with radius predictions from evolutionary tracks .
this approach is expected to be reasonably accurate , since the radii of brown dwarfs older than @xmath1100 myr are predicted to vary by @xmath4530% .
@xcite adopt a typical age of 3 gyr and a plausible range of 0.110 gyr in computing from , and @xcite adopt a radius range of @xmath148 predicted by @xcite simulations of the solar neighborhood assuming a constant star formation history .
( both studies use the tucson evolutionary tracks . )
there are 8 t4.5t6.5 dwarfs in these studies ( after updating the spectral types to the latest classification by @xcite ) , and all of these appear to be single based on high angular resolution imaging ( @xcite ; liu et al .
, in prep ) .
we compute the average and standard deviation of both studies to obtain @xmath149 k for t4.5 ( 2 objects ) , 1146 k for t5.0 ( interpolated ) , 1077 k for t5.5 ( 1 object ) , @xmath150 k for t6.0 ( 3 objects ) , and @xmath151 k for t6.5 ( 3 objects ) .
we have excluded the t6+t8 binary 2mass j1225@xmath02739ab and the peculiar t6 dwarf 2mass j0937 + 2931 , and we have assumed the t4.5 dwarf
2mass j0559@xmath01404 is an equal - magnitude unresolved binary based on its pronounced overluminosity ( e.g. , figure 3 of @xcite ) .
neither sample contains any t5.0 objects , but the 4 other subclasses almost exactly follow a straight line , so we linearly interpolate to find for t5.0 . in comparison to the field objects , the components of appear to have @xmath14100 k cooler temperatures relative to their spectral subclass ( @xmath152 k for the primary and @xmath153 k for the secondary , where we have adopted a 30 k uncertainty for the t5 and t5.5 field objects based on the other subclasses with more than one object ) .
the disagreement is modest , and a more definitive comparison is hampered by the few determinations ( i.e. , parallaxes ) for t4.5t6.5 dwarfs . nevertheless , the result is potentially intriguing .
in particular , @xcite have noted perhaps a similar effect for the three known late - l ( l7l8 ) dwarf companions to field stars .
more precise estimates can be obtained from evolutionary models for these companions than for field objects by incorporating the age estimates of their primary stars .
( see [ sec : tests ] for details . )
metchev & hillenbrand find that the l dwarf companions appear to be @xmath14100200 k cooler than single field late - l dwarfs .
they raise the possibility that the model - derived radii are at fault , either due to incorrect cooling rates or systematic overestimate of the field dwarf ages .
however , they prefer the hypothesis that the discrepancy is a manifestation of an unanticipated surface gravity dependence of the l / t transition , causing younger l / t transition objects to have cooler temperatures than older ones .
this is motivated by their analysis of the young ( 0.10.4 gyr ) l7.5 companion hd 203030b and apparently supported by the young ( 0.10.5 gyr ) t2.5 companion hn peg b , which also appears to be @xmath14200 k cooler than field objects of the same spectral type @xcite .
we find that the t5.0 and t5.5 components of may also be @xmath14100 k cooler than comparable field objects .
since these two components are later - type than the l / t transition ( e.g. , their positions in ir color - magnitude diagrams is coincident with the locus of mid / late - t dwarfs with blue near - ir colors ) , this may suggest that the discrepancy might not be solely associated with the l / t transition . instead , and
the aforementioned l / t companions may indicate that a systematic error in the estimated ages and radii of field late - l and t dwarfs is the culprit .
the @xmath1410% temperature discrepancy for amounts a @xmath1420% underestimate of the radii . for a fixed value of ( which is the appropriate constraint here ) , the @xcite scaling relations give @xmath154 where @xmath155 is the age and @xmath156 is the radius
this agrees well with the exponent value of 8.28.4 extracted from the tucson models for sources of @xmath157 to @xmath158 . propagating the @xmath159 uncertainties in the disagreement ,
the implied age overestimate is a factor of @xmath160 , meaning implied ages of @xmath140.31.0 gyr for the field population .
the same discrepancy can be seen in an alternate fashion , namely by comparing the luminosities for the same t4.5t6.5 field objects : @xmath161 for t4.5 , @xmath162 for t5.0 ( interpolated ) , @xmath163 for t5.5 , @xmath164 for t6.0 , and @xmath165 for t6.5 .
the luminosities of are comparable to the field objects of similar type ( @xmath166 for t5.0 component a and @xmath88 for t5.5 component b ) .
thus , in order for all the objects to have similar temperatures and , they must have about the same radius and thus about the same age as .
in other words , the measured total mass of is too small ( by a factor of @xmath142 ) compared to the mass expected from the evolutionary models for 3 gyr objects with @xmath167 . a representative age of @xmath10.5 gyr for the field population
is not ruled out given the state of the observations .
though golimowksi et al .
did consider the range of 0.110 gyr , they adopted a nominal age of 3 gyr in determining for field dwarfs , based on the 24 gyr age estimate from the tangential velocities of ultracool dwarfs by @xcite .
a younger age could be accomodated , since the tangential velocity of a population is only an approximate statistical estimate of its age .
indeed , kinematic analysis of the space motions of l and t dwarfs suggests a younger age of @xmath140.52 gyr @xcite .
similarly , the radii of @xmath148 adopted by vrba et al . is based on a mass function where @xmath168 ; a somewhat steeper mass function would lead to younger typical ages ( e.g. , figure 8 of @xcite , though @xcite suggest @xmath169 based on a small sample of t dwarfs ) .
thus , the discrepancy of evolutionary model - derived temperatures between objects of known mass / age and the field population can be plausibly explained by a modest overestimate of the ages of the field population . a larger sample of ultracool dwarfs with known masses and/or ages
is needed to better explore this issue ( [ sec : tests ] ) .
the spectrum of has not yet been fitted with model atmospheres due to its composite nature .
in fact , the t dwarf class as a whole has not been extensively subjected to such comparisons .
@xcite determined for a sample of sixteen t5.5t8 dwarfs by comparing near - ir spectral indices to condensate(dust)-free atmosphere models from the tucson group .
they determined @xmath170 k for one t5.5 dwarf , and an unweighted linear fit of atmosphere - derived versus spectral type for their sample ( excluding the peculiar t6 dwarf 2mass j0937 + 2931 ) gives @xmath171 where @xmath172 for t5.5 , @xmath173 for t6 , etc .
the rms about the linear fit is 50 k , which we adopt as the uncertainty ( a value somewhat larger than the @xmath3810 k to @xmath3840 k range computed for individual objects in their sample ) .
extrapolating the linear fit gives @xmath174 k for t5 .
this is obviously approximate , e.g. , given the potential systematic effects in the models and the spectral classification scheme , though this value agrees with the @xmath175 k found by by fitting model atmospheres to the spectrum of the t4.5 dwarf 2mass j0559@xmath01404 @xcite .
thus , model atmospheres indicate @xmath176 k and @xmath177 k for 2mass j1534@xmath02952a and b , respectively .
the temperatures we find using evolutionary tracks appear to be cooler by @xmath14100 k at modest significance ( @xmath178 k for component a and @xmath179 k for component b , if we assume that the errors add in quadrature ) .
we can not objectively discern if the problem lies in the evolutionary tracks , the model atmospheres , or both .
however , the evolutionary models are thought to be robust to the principal input uncertainties @xcite . on the other hand ,
the model atmospheres are quite uncertain . even though the spectral appearance of mid- and late - t dwarfs is relatively simple
dominated by collision - induced , , and in the near - ir and the wings of the 0.77 resonance line in the far - red the line lists for and are known to be incomplete , and the input physics to the atmosphere models are complex .
current atmospheres generally match the observed spectra of late - t ( t6t8 ) dwarfs , but not exactly so ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
therefore , while a larger sample of objects is needed both for dynamical mass determinations and model atmosphere fitting , the plausible hypothesis is that the observed discrepancy arises from an overprediction of by current model atmospheres .
we have directly measured the total mass of the system .
however , using the evolutionary tracks to determine the physical properties also implicitly determines the mass ratio , since the model - derived age and observed individual luminosities translate into individual masses ( again with the assumption that the system is composed of only two components ) .
we infer the mass ratio of the system from the ratio of the bolometric luminosities , as this is very robust . to illustrate this , consider the analytic scaling relation for solar - metallicity substellar objects from @xcite : @xmath180 where @xmath181 is the mass , @xmath155 is the age and @xmath182 is the rosseland mean opacity .
we measure a @xmath183 dex difference in between the two components would give an uncertainty of @xmath184 dex in the difference , but this would include the uncertainties in the distance modulus ( 0.04 mag ) and the integrated - light photometry ( 0.03 mag ) , which are common to both components . ] which leads to a mass ratio @xmath185 .
the uncertainty in the mass ratio is small due to the weak dependence of mass on luminosity at fixed age . using the actual tabulated tucson and lyon models and again keeping careful track of the covariance in the calculations , we compute a final value of , where the error includes the uncertainties in the model - inferred age and the observed difference .
( the tucson and lyon models give basically identical results for @xmath186 . )
this gives individual masses of @xmath187 ( @xmath11 ) and @xmath188 ( @xmath13 ) for components a and b , respectively .
we first compare the individual components against the cond evolutionary models of the lyon group , which provide predictions for the absolute magnitudes and colors .
the model predictions are generated for the cit photometric system , so we transform our resolved mko photometry for to this system using the results of @xcite .
figure [ fig : cmd ] shows that the models are somewhat too red compared to the data .
this is not surprising , as model atmospheres for t dwarfs are known to be deficient in the and opacities relevant at these wavelengths ( e.g. * ? ? ?
the plotted cond models are also computed only for solar - metallicity , and a non - solar metallicity for would impact the colors and magnitudes ( e.g. * ? ? ?
* ; * ? ? ?
* ) . indeed
, current models do not exactly match the observed color - magnitude loci for field t dwarfs ( e.g. , figure 8 of @xcite and figure 7 of @xcite ) .
nevertheless , will provide a strong test to for future models , since the components magnitudes , colors , and masses are very well - measured . with the individual mass estimates and an independent determination of
, it is possible to directly test different evolutionary tracks using the hertzsprung - russell ( h - r ) diagram .
we use the values of @xmath176 k and @xmath177 k for 2mass j1534@xmath02952a and b , derived in [ sec : temp ] from model atmosphere studies .
figure [ fig : hrd - masses ] shows the individual components on the h - r diagram and compares these to the tucson and lyon evolutionary tracks .
the locations of the two components disagree with both sets of models ( which agree very well between themselves ) .
the likely interpretation is that the temperatures from the model atmospheres place the components to the left of the evolutionary tracks , i.e. , too warm . as discussed in
[ sec : temp ] , the model atmospheres are a significant source of the uncertainty in placing the components on figure [ fig : hrd - masses ] . a possible systematic error of only @xmath14100 k would be sufficient to resolve the discrepancy with the data .
therefore , while acquiring resolved spectra of the two components could help refine the temperature determination , the systematic uncertainties in the atmosphere models will still hamper accurate placement on the h - r diagram .
we discuss this further in the next section .
the opposite interpretation is that the evolutionary models are incorrect , leading to a @xmath1450% overprediction of the luminosities and a @xmath1420% overprediction of the radii , given the component masses .
equivalently , if one were simply to assume the h - r diagram positions are accurate , the evolutionary models would suggest individual masses of around 0.05 and 0.06 from the lyon and tucson models , respectively , i.e. , nearly a factor of two overestimate in the masses . while it may be that the evolutionary models are so substantially incorrect , such a conclusion is not compelling at this point , given the plausible errors in the determinations
. would move one evolutionary track closer to one component , while the other track would move farther away from the other component .
the evolutionary models could be brought into agreement with the data for component a for @xmath189 , but this is implausible given the nearly equal magnitudes of the two components .
such a small @xmath186 would also exacerbate the disagreement between the observations and the models for component b. ] direct mass determinations for the individual components from radial velocity monitoring and/or absolute orbital astrometry will help to further characterize the system .
such data will directly test the @xmath190 determined from the evolutionary tracks .
the expected maximum radial velocity difference of the two components is only 4.6 .
since the two components are nearly equal mass and brightness , the orbital motion will be very difficult to detect in the integrated - light spectrum .
resolved ao spectroscopy will be required , and the small amplitude will make it a challenging measurement given the few accuracy that has been achieved for t dwarfs on the largest existing telescopes @xcite .
individual mass measurements can in principle also test the evolutionary tracks directly .
one can estimate the age of each component from its mass and luminosity ( as we have done using the total mass ) and see if the ages indicate coevality for the system .
however , given the near - equal flux ratio of this system ( and most ultracool binaries ) , this coevality test is unlikely to be very discriminating .
moreover , individual masses can not resolve the discrepancy seen in the h - r diagram ( figure [ fig : hrd - masses ] ) , which largely arises from the uncertainties in the model atmospheres . with the advent of lgs ao on the largest ground - based telescopes
, we can expect an increasing number of dynamical masses for ultracool field dwarfs in the near - future .
the most useful systems for testing theory will be those with both independent mass and age determinations , namely binaries that are associated with open clusters / groups and/or field stars of known age .
the former will present a significant technical challenge , e.g. , ultracool binaries in the hyades ( @xmath191 pc ; ) and pleiades ( @xmath192 pc ; @xcite ) with suitably short orbital periods are unresolvable with current technology and thus none are currently known .
ultracool binary companions to field stars are extremely rare and thus while very valuable systems , these will only probe a very limited range of spectral type , age , and mass : only four systems are known with suitably short orbital periods ( @xmath193 yr ) the t1+t6 binary bab , the l4+l4 binary hd 130948bc @xcite , the l4.5+l6 binary gj 417bc @xcite , and the l4.5+l4.5 binary gj 1001bc @xcite .
therefore , there is significant motivation to develop analyses that employ masses derived from the much more numerous field binaries . in this regard and as illustrated by our analysis for , one can identify two orthogonal pathways to confront theory : ( 1 ) comparison to evolutionary tracks and ( 2 ) comparison to atmospheric models .
direct measurements of , , and mass ( or age ) for brown dwarfs enable use of the h - r diagram , by comparing the observations to evolutionary tracks that correspond to the measured masses of the objects .
as illustrated by figure [ fig : hrd - masses ] , the lyon and tucson tracks differ at the 510% level in mass , and thus mass determinations of 23% accuracy could discriminate between the two models , if and can be well - measured .
( improvements in the parallaxes of many ultracool binaries will also be needed to achieve such accurate masses . )
accurate measurements for are largely straight - forward , as good as a few percent ( e.g. , [ sec : lbol ] ) .
however , direct determinations are extremely challenging , since radius measurements are needed .
brown dwarfs are too small and faint to be resolved with current or planned interferometers , and no eclipsing ultracool field binaries are yet known .
thus , must be derived from modeling the observed colors , magnitudes , and/or spectra ; the approach currently suffers from uncertainties at the level of a few to several hundred kelvin and systematic errors that are difficult to quantify ( e.g. * ? ? ?
* ) . in comparison ,
figure [ fig : hrd - masses ] shows that determinations good to @xmath4530 k are needed .
therefore , decisive tests of evolutionary tracks using field binaries will be challenged by this uncertainty in , in the absence of direct radius measurements .
brown dwarfs obey a mass - luminosity - age ( @xmath194 ) relation , and for most field objects neither the mass nor the age is known . a commonly used approach to circumvent this limitation
is to study brown dwarfs that are companions to main - sequence stars , where ( indirect ) age estimates are available from the primary star .
this approach can also be applied to members of coeval clusters / groups and companions to post - main - sequence stars of known age ( e.g. * ? ? ?
* ; * ? ? ?
. in these situations , and @xmath155 are known , and combined with evolutionary models , one can derive @xmath181 and consequently and . then the observed colors , magnitudes , and spectra can test the accuracy of atmospheric models with the same and .
examination of the known `` age benchmark '' t dwarfs in this fashion finds that the properties deduced from atmospheric models are in good agreement with those from the evolutionary models , within the uncertainties in the ages and metallicities of the primary stars @xcite .
we suggest that , in an analogous fashion , field binaries with known masses can also serve as `` benchmark '' objects . in this case ,
@xmath181 and are known , and combined with evolutionary models , one can derive @xmath155 , as we have done in 4.1 .
this provides and and thereby allows tests of atmospheric models .
the chain of analysis is identical to brown dwarf companions of known age : given independent knowledge of two quantities out of @xmath195 , use evolutionary models to derive the third . in both cases
, these benchmarks can also serve as anchor points for direct empirical calibration of spectroscopic diagnostics of and ( e.g. * ? ? ?
* ; * ? ? ?
* ) . at face value , using objects that are age benchmarks or mass benchmarks is less fundamental than direct tests of the evolutionary models using the h - r diagram
however , in practice the benchmark test is much more feasible to implement and subject to much smaller systematic errors . in the absence of direct radius measurements ,
the h - r diagram test is held hostage to the systematic errors in the determination of from atmospheric models . in contrast , the benchmark test relies on the evolutionary models , which are thought to be more robust ( e.g. * ? ? ?
* ) . in short ,
given the choice of relying on atmospheric models ( h - r diagram test ) or evolutionary models ( benchmark test ) , the evolutionary models are likely to be preferred . to assess
the relative utility of `` age benchmarks '' ( brown dwarf companions to stars ) compared to `` mass benchmarks '' ( brown dwarfs with dynamical masses ) , we turn to equation [ eqn : lbol ] . for an object with a measured and ignoring the weak dependence on @xmath182 , given a measurement of @xmath181 or @xmath155 with accompanying uncertainty of @xmath196 or @xmath197 , the fractional error in the remaining quantity
is related by : @xmath198 typical uncertainties in the ages of main - sequence field stars are about 50100% ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , and thus age - benchmark objects would have a 2550% uncertainty in the mass inferred from evolutionary models .
in contrast , dynamical masses of @xmath14510% accuracy will be possible in the next few years for many ultracool dwarfs , as we have already achieved with , and hence mass - benchmark objects will have only a @xmath141020% error in the age inferred from evolutionary models .
so overall , age benchmarks can be expected to have @xmath199 \{50100% , 2550% } while mass benchmarks would have uncertainties of order @xmath199 \{1020% , 510%}. we can use the analytic fits to evolutionary models from @xcite to gauge the relative accuracy on and derived from both types of benchmarks . using standard error propagation and assuming uncorrelated errors , we find for : @xmath200 @xmath201 where @xmath202 and @xmath203 are the uncertainties in the temperture , mass , luminosity , and surface gravity , respectively .
and then for , we find : @xmath204 @xmath205 thus with representative values for the fractional errors in age ( 50% ) , luminosity(10% ) , and mass ( 5% ) , we see that and are better constrained by a factor of @xmath145 using mass benchmarks than age benchmarks . figure [ fig : benchmarks ] plots the derived analytic estimates for both types of benchmarks .
these contour plots provide a convenient means to gauge the expected errors in and determinations from benchmarks .
the morphology of the contours also illustrates whether the observational errors in the age , mass , and/or dominate the uncertainties in and . for the specific case of ,
there is good agreement between the analytic estimates and the values derived directly from the actual evolutionary models ( table [ table : evolmodels ] ) .
we have determined the first dynamical mass for a binary t dwarf , the t5.0+t5.5 system , by combining six epochs of keck lgs ao imaging from 20052008 with three epochs of _ hst _ imaging obtained in 2000 and 2006 .
both datasets achieve milliarcsecond accuracy or better for the relative astrometry of the two components and are validated through extensive testing with images of simulated binaries .
we employ a markov chain monte carlo analysis to determine the orbital parameters and their uncertainties .
the time baseline of our complete dataset covers about half of the total period .
we find that the orbital motion of the binary is viewed in an almost edge - on orientation and has a modest eccentricity .
our determination of a @xmath206 yr orbital period is significantly longer than the original 4-year estimate , as by chance the binary was at a very small projected separation when discovered in 2000 .
the total mass of the system is @xmath207 ( @xmath208 ) , including the uncertainty in the parallax .
this is the second brown dwarf binary directly confirmed , the first among the field population .
it is also the coolest and lowest mass binary with a dynamical mass determination to date . with very accurate measurements of the total mass and the bolometric luminosity ( @xmath209 )
, we use the tucson and lyon evolutionary tracks to determine the remaining physical properties for the system .
the two sets of models give largely consistent results , which highlights the difficult of distinguishing between them even with such precise observational data .
we average the model results to represent the final determinations .
we find a relatively youthful age for the system of [email protected] gyr ( 1@xmath29 ) , consistent with its low tangential velocity relative to other field t dwarfs .
the remaining physical parameters of the individual components are then fully determined : radii of and , effective temperatures of @xmath210 k and @xmath211 k , surface gravities of and @xmath143 , and masses of @xmath10 ( @xmath212 ) and @xmath12 ( @xmath213 ) for components a and b , respectively .
we take care to account for the covariances inherent in the measurement uncertainties , by using a monte carlo approach to derive these physical quantities from the evolutionary models .
our approach also assumes that the system is coeval and composed of only two components .
these precise determinations for 2mass j1534@xmath02952a and b are in general accord with the and values found previously for field t dwarfs based on model atmospheres and with the ages of t dwarfs predicted by monte carlo simulations of the solar neighborhood .
however , upon closer scrutiny , there are two potential discrepancies with past studies .
both suggest that the temperatures of field t dwarfs may be overestimated by @xmath14100 k , though we stress that the two discrepancies must arise from independent effects .
( 1 ) the temperatures of 2mass j1534@xmath02952a and b appear to be cooler than field objects of comparable spectral type .
this resembles discrepancies previously noted by @xcite and @xcite for late - l / early - t dwarfs that are companions to young main - sequence stars .
they have hypothesized that the effect is due to the gravity sensitivity of the l / t transition .
the fact that this discrepancy also occurs for suggests instead that the problem may arise from a factor of @xmath214 overestimate in the adopted ages of field objects when determining their temperatures using evolutionary tracks .
ages of @xmath10.31.0 gyr are preferred based on this binary .
( 2 ) the temperatures of 2mass j1534@xmath02952a and b are slightly cooler than inferred for other mid - t dwarfs from model atmospheres .
detailed analysis of the system s integrated - light and resolved spectra with model atmospheres is needed to directly assess the and of the two components and to refine the comparison with the values derived from evolutionary models .
the positions of the two components on the h - r diagram are discrepant with theoretical evolutionary tracks corresponding to their individual masses .
in fact , taken at face value , using the h - r diagram positions to infer masses from the evolutionary tracks would lead to masses of @xmath140.050.06 , about a factor of two larger than the actual measured masses . while this discrepancy could stem from large systematic errors in the luminosities ( @xmath150% errors ) and/or radii ( @xmath120% errors ) predicted by evolutionary models ,
the likely cause is that temperatures from model atmospheres are too warm by @xmath14100 k for mid - t dwarfs .
this highlights the need for continued improvements to the model atmospheres .
future monitoring of will help to refine its orbit and its dynamical mass .
the orbital separation of the system is now rapidly decreasing and will not be readily resolvable again until around 2011 . at the same time , an improved parallax for the system will be required the uncertainty in the total mass from the orbit fitting is 3% , compared to the 5% that arises from the uncertainty in the parallax .
radial velocity monitoring and/or absolute astrometry will directly determine the individual masses and test if the system is a higher order multiple .
however , given the very similar fluxes of the two components ( implying nearly equal mass ) , individual mass measurements are unlikely to resolve the discordant h - r diagram position of the two components relative to evolutionary tracks .
this problem is likely driven by the systematic uncertainties in current model atmospheres for t dwarfs .
the fundamental characteristic of the field population is that it spans a range of ( largely unknown ) ages .
however despite this uncertainty , field brown dwarf binaries can strongly test theoretical models , if analyzed appropriately .
these systems will be especially valuable in light of the current paucity of eclipsing field ultracool binaries and resolvable , short - period ultracool binaries in open clusters / groups.600
l and t dwarfs known . ] specifically , attempts to directly test different evolutionary tracks by placing ultracool objects on the h - r diagram ( the `` h - r diagram test '' ) will be challenging , given the similarity between the tracks and the difficulty in independently determining with model atmospheres . instead
, atmosphere models can be confronted against and values for ultracool objects as derived from the evolutionary models , which can be exceptionally precise ( the `` benchmark test '' ) .
this approach has previously been applied to single brown dwarfs that are companions to stars of known age ( `` age benchmarks '' ) .
we suggest that in an analogous fashion , field ultracool binaries with dynamical mass determinations ( `` mass benchmarks '' ) can test the model atmospheres .
in fact , given the plausible observational uncertainties , mass benchmarks are likely to provide stronger constraints ( by a factor of @xmath145 ) on and than age benchmarks , since dynamical masses can be determined far more accurately than ages for main - sequence field stars . with the widespread advent of lgs ao on the largest ground - based telescopes , we can look forward to a rapid increase in dynamical mass determinations for low - mass field binaries and thus substantial advances in our understanding of the properties and evolution of substellar objects .
we gratefully acknowledge the keck lgs ao team for their exceptional efforts in bringing the lgs ao system to fruition .
it is a pleasure to thank antonin bouchez , david lemignant , marcos van dam , randy campbell , al conrad , jim lyke , hien tran , robert lafon , kenny graves , cindy wilburn , joel aycock , terry stickel , gary punawai , and the keck observatory staff for assistance with the observations .
we thank alan stockton for a fortuitous swap of observing nights in spring 2005 ; brian cameron for sharing his nirc2 instrumental distortion analysis ; hai fu for idl plotting assistance ; michael cushing and mark pitts for asssistance with the irtf / spex spectroscopy ; adam burrows and isabelle baraffe for finely gridded evolutionary tracks ; and thierry forveille and adam burgasser for careful readings of the manuscript . we have benefitted from enlightening discussions with michael cushing , adam burrows , and isabelle baraffe about substellar models ; thierry forveille about low - mass binaries ; brian cameron about astrometry with nirc2 ; and jay anderson about astrometry with _
hst_. our research
has employed the 2mass data products ; nasa s astrophysical data system ; the simbad database operated at cds , strasbourg , france ; the m , l , and t dwarf compendium housed at dwarfarchives.org and maintained by chris gelino , davy kirkpatrick , and adam burgasser @xcite ; and the spex prism spectral libraries maintained by adam burgasser at http://www.browndwarfs.org/spexprism .
mcl and tjd acknowledge support for this work from nsf grant ast-0507833 and an alfred p. sloan research fellowship .
mji acknowledges michelson fellowship support from the michelson science center and the nasa navigator program .
finally , 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 .
rayner , j. t. , toomey , d. w. , onaka , p. m. , denault , a. j. , stahlberger , w. e. , watanabe , d. y. , & wang , s. 1998 , in proc .
spie : infrared astronomical instrumentation , ed .
a. m. fowler , vol .
3354 , 468479 color versus near - ir spectral type for t dwarfs .
the plotting symbols are the synthesized colors from published low - resolution near - ir spectra of t dwarfs , excluding objects that are spectrally peculiar and known binaries .
( see [ sec : phot ] for references . )
the lines represent 2nd - order polynomial fits to the data , with the solid line being the spectral type as a function of color ( i.e. , the inverse fit ) and the dotted line being the color as a function of spectral type.[fig : ch4short],width=480 ] ) , orbital period ( @xmath89 ) , eccentricity ( @xmath94 ) , inclination ( @xmath91 ) , epoch of periastron ( @xmath92 ) , pa of the ascending node ( @xmath93 ) , and argument of periastron ( @xmath95 ) .
each histogram is shaded to indicate the 68.3% and 95.5% confidence regions , which correspond to 1 and 2@xmath29 for a normal distribution , and the solid vertical line represents the median value.[fig : orbit - summary],height=720 ] ) , orbital period ( @xmath89 ) , eccentricity ( @xmath94 ) , inclination ( @xmath91 ) , epoch of periastron ( @xmath92 ) , pa of the ascending node ( @xmath93 ) , and argument of periastron ( @xmath95 ) .
each histogram is shaded to indicate the 68.3% and 95.5% confidence regions , which correspond to 1 and 2@xmath29 for a normal distribution , and the solid vertical line represents the median value.[fig : orbit - summary],height=720 ] yr ) branch and a long - period ( @xmath215 yr ) one .
about 98% of the mcmc chain steps are in the short - period branch .
the top and side plots show the resulting probability distributions of @xmath89 and @xmath94 .
each histogram is shaded to indicate the 68.3% and 95.5% confidence limits , which correspond to 1@xmath29 and 2@xmath29 for a normal distribution , and the dashed vertical lines represent the median values .
[ fig : orbit - p_e],height=720 ] -1 in -0.8 in .
the measurement errors are comparable to or smaller than the plotting symbols , except for the three data points shown in the pa plot .
the bottom panel of each plot shows the difference of the observed astrometry and the best - fitting ( 15.2-year period ) orbit .
[ fig : orbit - sep],title="fig:",height=480 ] -1.5 in .
the measurement errors are comparable to or smaller than the plotting symbols , except for the three data points shown in the pa plot .
the bottom panel of each plot shows the difference of the observed astrometry and the best - fitting ( 15.2-year period ) orbit .
[ fig : orbit - sep],title="fig:",height=480 ] -2ex , 2@xmath29 , and 3@xmath29 for a normal distribution .
the solid vertical line represents the median value of 0.0556 .
the standard deviation of the distribution is 0.0018 .
the wider , unshaded histogram shows the final mass distribution , after accounting for the additional 4.9% error due primarily to the uncertainty in the binary s parallax the result is essentially gaussian with a standard deviation of 0.003 .
the confidence limits for both distributions are given in table [ table : orbits].[fig : orbit - masses],height=720 ] and 2@xmath29 for a normal distribution .
the dashed curves at the bottom show the resulting mass distributions after accounting for the uncertainty in the parallax .
while the @xcite prior favors slightly higher masses , all four priors give very consistent results .
[ fig : priors],height=720 ] constraints on the total mass , and the vertical lines show the corresponding ages derived from the burrows et al
. models .
both sets of models produce consistent results , given in table [ table : evolmodels ] .
the cond models provide predictions for both the absolute magnitudes and bolometric luminosity , so all of these are shown .
( the two sets of models predict very similar results , so the red and black model lines are indistinguishable on this plot . )
note that these plotted curves are computed for an object with the fluxes of and are not generally applicable to other binaries . see
[ sec : ages ] for details .
[ fig : models - totalmass],height=720 ] -2 in 0.3 in 1@xmath29 range are plotted as solid lines and dotted lines , respectively .
the numbered boxes indicate the logarithm of the model age in gyr .
the error bars on the absolute @xmath26-band magnitude is smaller than the plotting symbol .
the models do not match the data very well , which can be ascribed to the incomplete opacities in the model atmospheres .
[ fig : cmd],title="fig:",height=528 ] -1.5 in 0.3 in 1@xmath29 range are plotted as solid lines and dotted lines , respectively .
the numbered boxes indicate the logarithm of the model age in gyr .
the error bars on the absolute @xmath26-band magnitude is smaller than the plotting symbol .
the models do not match the data very well , which can be ascribed to the incomplete opacities in the model atmospheres .
[ fig : cmd],title="fig:",height=528 ] -0.7 in ) .
the uncertainity in the measurements are comparable to the vertical extent of the plotting symbols .
the solid line represents the median mass values determined for the individual components , and the dotted lines show the @xmath216 ranges .
the embedded small squares demarcate logarithms of a set of ages ; the ages for the burrows models are labelled , and the ( unlabeled ) ages for the lyon models have very similar positions along the plotted line .
there is a modest systematic disagreement between the evolutionary tracks and observations.[fig : hrd - masses],height=672 ] 2005-may-01 & @xmath26 & 1.66 & 102 @xmath38 12 & 0.020 @xmath38 0.002 & 211.3 @xmath38 1.5 ( 1.5 ) & 14.1 @xmath38 0.3 ( 0.3 ) & 0.163 @xmath38 0.014 + & @xmath27 & 1.63 & 86 @xmath38 6 & 0.047 @xmath38 0.006 & 211.7 @xmath38 0.8 ( 0.8 ) & 13.86 @xmath38 0.15 ( 0.13 ) & 0.286 @xmath38 0.011 + & & 1.61 & 88 @xmath38 6 & 0.101 @xmath38 0.012 & 212.4 @xmath38 1.1 ( 1.0 ) & 14.0 @xmath38 0.2 ( 0.2 ) & 0.278 @xmath38 0.021 + 2006-may-05 & & 1.56 & 64 @xmath38 3 & 0.210 @xmath38 0.014 & 190.6 @xmath38 0.3 ( 0.2 ) & 15.43 @xmath38 0.12 ( 0.09 ) & 0.282 @xmath38 0.010 + 2007-mar-26 & @xmath24 & 1.56 & 82 @xmath38 3 & 0.151 @xmath38 0.016 & 158.0 @xmath38 0.6 ( 0.6 ) & 17.5 @xmath38 0.2 ( 0.19 ) & 0.287 @xmath38 0.012 + 2007-apr-22 & & 1.57 & 67 @xmath38 5 & 0.20 @xmath38 0.03 & 153.7 @xmath38 0.4 ( 0.3 ) & 17.53 @xmath38 0.13 ( 0.10 ) & 0.269 @xmath38
0.010 + 2008-jan-15 & & 2.05 & 100 @xmath38 3 & 0.074 @xmath38 0.002 & 114.4 @xmath38 1.1 ( 1.1 ) & 21.5 @xmath38 0.9 ( 0.9 ) & 0.27 @xmath38 0.06 + 2008-apr-01 & & 1.55 & 87 @xmath38 4 & 0.095 @xmath38 0.018 & 102.5 @xmath38 0.7 ( 0.7 ) & 21.1 @xmath38 0.7 ( 0.7 ) & 0.25 @xmath38 0.04 + & @xmath28 & 1.58 & 78 @xmath38 7 & 0.048 @xmath38 0.018 & 102.0 @xmath38 0.4 ( 0.4 ) & 20.4 @xmath38 1.5 ( 1.5 ) & 0.21 @xmath38 0.04 + lccccccc 2000-aug-18 & wfpc2 & @xmath37 & 62.8 @xmath38 1.2 & 357.1 @xmath38 0.8 & 0.30 @xmath38 0.05 + 2006-jan-19 & acs & @xmath37 & 199.0 @xmath38 1.1 & 14.5 @xmath38 0.6 & 0.28 @xmath38 0.06 + 2006-apr-11 & acs & @xmath37 & 191.2 @xmath38 1.1 & 15.5 @xmath38 0.4 & 0.30 @xmath38 0.04 + @xmath217 ( mags ) & 4.95 @xmath38 0.04 & 5.10@xmath38 0.04 + @xmath218 ( mags ) & @xmath00.08 @xmath38 0.04 & @xmath00.21 @xmath38 0.04 + @xmath219 ( mags ) & @xmath00.30 @xmath38 0.05 & @xmath00.37 @xmath38 0.05 + @xmath220 ( mags ) & @xmath00.17 @xmath38 0.04 & @xmath00.17 @xmath38 0.04 + @xmath221 ( mags ) & @xmath00.25 @xmath38 0.04 & @xmath00.38 @xmath38 0.04 + @xmath222 ( mags ) & 19.57 @xmath38 0.04 & 19.87 @xmath38 0.05 + @xmath223 ( mags ) & 14.61 @xmath38 0.05 & 14.77 @xmath38 0.05 + @xmath224 ( mags ) & 14.69 @xmath38 0.05 & 14.98 @xmath38 0.05 + @xmath225 ( mags ) & 14.86 @xmath38 0.05 & 15.15 @xmath38 0.05 + estimated spectral type & t5.0 @xmath38 0.5 & t5.5 @xmath38 0.5 + @xmath226 & @xmath227 & @xmath228 + time of periastron @xmath229 ( mjd ) & 55960 & @xmath0240 , 210 & @xmath0740 , 450 & 56024 @xmath38 347 + orbital period @xmath89 ( yr ) & 15.1 & @xmath01.6 , 2.3 & @xmath03.1 , 5.1 & 15.2 @xmath38 2.6 + semi - major axis @xmath90 ( mas ) & 171 & @xmath013 , 19 & @xmath027 , 41 & 172 @xmath38 22 + semi - major axis @xmath90 ( au ) & 2.3 & @xmath00.2 , 0.3 & @xmath00.4 , 0.6 & 2.3 @xmath38 0.3 + inclination @xmath91 ( ) & 84.3 & @xmath00.6 , 0.6 & @xmath01.7 , 1.0 & 84.3 @xmath38 0.8 + eccentricity @xmath94 & 0.25 & @xmath00.13 , 0.11 & @xmath00.20 , 0.25 & 0.24 @xmath38 0.16 + pa of the ascending node @xmath93 ( ) & 13.0 & @xmath00.3 , 0.3 & @xmath00.9 , 0.5 & 13.0 @xmath38 0.4 + argument of periastron @xmath95 ( ) & 179 & @xmath014 , 6 & @xmath083 , 11 & 178 @xmath38 10 + total mass ( ): fitted & 0.0556 & @xmath00.0017 , 0.0019 & @xmath00.004 , 0.004 & 0.056 @xmath38 0.004 + total mass ( ): final & 0.056 & @xmath00.003 , 0.003 & @xmath00.006 , 0.007 & 0.056 @xmath38 0.005 + log(age ) & & & + radius ( ) & @xmath230 & @xmath231 & @xmath232 & @xmath233 & @xmath234 & @xmath235 + ( k ) & @xmath236 & @xmath237 & @xmath238 & @xmath239 & @xmath240 & @xmath241 + ( cgs ) & @xmath242 & @xmath243 & @xmath244 & @xmath245 & @xmath246 & @xmath247 + @xmath248 ( k ) & & & + mass ratio @xmath249 & & & + mass ( ) & @xmath250 & @xmath251 & @xmath252 & @xmath253 & @xmath254 & @xmath255 + | we present multi - epoch , near - infrared imaging of the binary t5.0+t5.5 dwarf 2mass j15344984@xmath02952274ab obtained with the keck laser guide star adaptive optics system .
our keck data achieve sub - milliarcsecond relative astrometry and combined with an extensive ( re-)analysis of archival _ hst _ imaging , the total dataset spans @xmath150% the orbital period .
we use a markov chain monte carlo analysis to determine an orbital period of 15.1@xmath2 yr and a semi - major axis of 2.3@xmath3 au .
we measure a total mass of @xmath4 ( @xmath5 ) , where the largest uncertainty arises from the parallax .
this is the first field binary for which both components are confirmed to be substellar .
this is also the coolest and lowest mass binary with a dynamical mass determination to date . using evolutionary models and accounting for the measurement covariances , we derive an age of and a mass ratio of @xmath6 .
the relatively youthful age is consistent with the low tangential velocity of this system . for the individual components , we find and @xmath7 k , @xmath8 and @xmath9 ( cgs ) , and masses of @xmath10 ( @xmath11 ) and @xmath12 ( @xmath13 ) .
these precise values generally agree with previous studies of t dwarfs and affirm current theoretical models .
however , ( 1 ) the temperatures are about 100 k cooler than derived for similar field objects and suggest that the representative ages of field brown dwarfs may be overestimated .
similarly , ( 2 ) the h - r diagram positions are discrepant with current model predictions and taken at face value would overestimate the masses . while this may arise from large errors in the luminosities and/or radii predicted by evolutionary models , the likely cause is a modest ( @xmath14100 k ) overestimate in temperature of t dwarfs determined from model atmospheres .
we elucidate future tests of theory as the sample of substellar dynamical masses increases .
in particular , we suggest that field brown dwarf binaries with dynamical masses ( `` mass benchmarks '' ) can serve as reference points for and and thereby constrain ultracool atmosphere models , as good as or even better than single brown dwarfs with age estimates ( `` age benchmarks '' ) . |
recently , the emergence and evolution of language attracts a growing interest . in this
interdisciplinary field problems like differentiation of languages , development of the speech apparatus or formation of linguistically connected social groups require joint efforts of many specialists such as linguists , neuroscientists , or anthropologists . however , there are also more general questions concerning this almost exclusively human trait .
why do we use words and then combine them into sentences ?
why all languages have grammar ?
to what extent is our brain adapted for acquisition of language ?
can learning direct the evolution ?
this sample questions justify an increasing involvement of researchers also from other disciplines such as artificial intelligence , computer sciences , evolutionary biology or physics @xcite .
computer modeling is a frequently used tool in the studies of language evolution . in this technique
two main approaches can be distinguished . in the first one , known as an iterated learning model , one
is mainly concerned with the transmission of language between successive generations of agents @xcite .
the important issue that the iterated learning model has successfully addressed is the transition from holistic to compositional language .
however , since the number of communicating agents is typically very small , the problem of the emergence of linguistic coherence must be neglected in this approach . to tackle this problem steels introduced a naming game model @xcite . in this approach
one examines a population of agents trying to establish a common vocabulary for a certain number of objects present in their environment .
the change of generations is not required in the naming game - model since the emergence of a common vocabulary is a consequence of the communication processes between agents .
it seems that the iterated learning model and the naming - game model are at two extremes : the first one emphasizes the generational turnover while the latter concentrates on the single - generation ( cultural ) interactions . since in the language evolution both aspects are present , it is desirable to examine models that combine evolutionary and cultural processes . in the present paper
we introduce such a model .
agents in our model try to establish a common vocabulary like in the naming - game model , but in addition they can breed , mutate , and die .
moreover , they are equipped with an evolutionary trait : learning ability . as a result
evolutionary and cultural ( learning from peers ) processes mutually influence each other .
when communication between agents is sufficiently frequent , cultural processes create a favourable niche in which a larger learning ability becomes advantageous .
but gradually increasing learning abilities in turn speed up the cultural processes . as a result
the model undergoes an abrupt bio - linguistic transition .
one can speculate that the proposed model suggests that linguistic and biological processes at a certain point of human history after crossing a certain threshold started to have a strong influence on each other and that resulted in an explosive development of our species . that learning in our model modifies the fitness landscape of a given agent and facilitates the genetic accommodation of learning ability is actually a manifestation of the much debated baldwin effect @xcite .
in our model we consider a set of agents located at sites of the square lattice of the linear size @xmath0 .
agents are trying to establish a common vocabulary on a single object present in their environment .
an assumption that agents communicate only on a single object does not seem to restrict the generality of our considerations and was already used in some other studies of naming - game @xcite or language - change @xcite models .
a randomly selected agent takes the role of a speaker that communicates a word chosen from its inventory to a hearer that is randomly selected among nearest neighbours of the speaker .
the hearer tries to recognize the communicated word , namely it checks whether it has it in its inventory .
a positive or negative result translates into communicative success or failure , respectively . in some versions of the naming - game model
@xcite success means that both agents retain in their inventories only the chosen word while in the case of failure the hearer adds the communicated word to its inventory . to implement the learning ability we modified this rule and assigned weights @xmath1 ( @xmath2 ) to each @xmath3-th word in the inventory .
the speaker selects then the @xmath3-th word with the probability @xmath4 where summation is over all words in its inventory ( if its inventory is empty , it creates a word randomly ) . if the hearer has the word in its inventory , it is recognized .
in addition , each agent @xmath5 is characterized by its learning ability @xmath6 ( @xmath7 ) that is used to modify weights .
namely , in the case of success both speaker and hearer increase the weights of the communicated word by learning abilities of the speaker and hearer , respectively . in the case of failure
the speaker subtracts its learning ability from the weight of the communicated word .
if after such a subtraction a weight becomes negative the corresponding word is removed from the repository .
the hearer in the case of failure , i.e. , when it does not have the word in its inventory , adds the communicated word to its inventory with a unit weight .
in addition to communication , agents in our model evolve according to the population dynamics : they can breed , mutate , and eventually die . to specify intensity of these processes
we introduce the communication probability @xmath8 . with probability
@xmath8 the chosen agent becomes a speaker and with probability @xmath9 we will attempt a population update .
during such a move the agent dies with the probability @xmath10 , where @xmath11 $ ] and @xmath12 and @xmath13 are certain parameters whose role is to ensure a certain speed of population turnover .
moreover , @xmath14 is the age of an agent and @xmath15 is the average ( over agents ) sum of weights .
such a formula takes into account both its linguistic performance ( the bigger @xmath16 the larger @xmath17 ) and its age .
if the agent survives ( it happens with the probability @xmath18 ) it breeds , provided that there is an empty site on one of the neighbouring sites .
the offspring typically inherits parent s learning ability and the word from its inventory that has the highest weight . in the offspring inventory
the weight assigned initially to this word equals one .
with the small probability @xmath19 a mutation takes place and the learning ability of an offspring is selected randomly anew . with the same probability an independent check is made whether to mutate the inherited word . a diagram illustrating the dynamics of our model
is given in the appendix @xcite .
let us also notice that the behaviour of our model , that is described below , is to some extent robust with respect to some modifications of its rules .
for example , qualitatively the same behaviour is observed for modified parameters @xmath20 and @xmath21 , different form of the survival probability @xmath17 ( provided it is a decreasing function of @xmath14 and an increasing function of @xmath16 ) , or different breeding and/or mutation rules .
to examine the properties of the model we used numerical simulations .
most of the results are obtained for @xmath22 and @xmath23 but simulations for @xmath24 and @xmath25 lead to the similar behaviour .
simulations start from all sites occupied by agents and having a single , randomly chosen for each agent word in their inventories with a unit weight .
the learning ability of each agent is also chosen randomly .
an important parameter of the model is the communication probability @xmath8 that specifies intensity of communication attempts in comparison with populational changes . in general , for small @xmath8 the model remains in the phase of linguistic disorder with only small clusters of agents using the same language .
we define the language of an agent as the largest - weight word in its inventory .
such a definition means that agents using the same language usually ( but not always ) use a recognizable word and it ensures a relatively large rate of communication successes of such agents .
a typical distribution of languages in this disordered small-@xmath8 phase is shown in the left panel of fig .
[ conf ] where agents using the same language are drawn with the same shade of grey . upon increasing the communication probability @xmath8 the clusters of agents only slightly increase , but after reaching a certain threshold an abrupt transition takes place and the model enters the phase of linguistic coherence with almost all agents belonging to the same cluster ( fig .
[ conf ] , right panel ) . to examine the nature of this transition we measured the communication success rate @xmath26 defined as an average over agents and simulation time of the fraction of successes with respect to all communication attempts .
moreover , we measured the average learning ability @xmath27 .
the measured values of @xmath26 and @xmath27 as a function of @xmath8 are shown in fig .
[ steady ] .
one can notice that the abrupt transition around @xmath28 manifests not only through the jump of the communicative properties ( @xmath26 ) but also through the jump of the biological endowment ( jump of @xmath27 ) .
such a coincidence is by no means obvious and in principle one can imagine these two transitions being separated . before examining the mechanism responsible for such an agreement let us mention about the behaviour of the model with the learning ability kept fixed during entire simulations . in this case
there is also a phase transition between disordered and linguistically coherent phases but this time the transition is much smoother ( fig .
[ steady ] ) .
large fluctuations of the success rate @xmath26 in the vicinity of the transition and absence of the jump suggest that this might be a continuous transition . in the last section we will return to this point .
let us also notice that sudden transitions in linguistic models were also reported in some other models @xcite .
the agreement of the transitions as seen in fig .
[ steady ] shows that communicative and biological ingredients in our model strongly influence each other and that leads to the single and abrupt transition . in our model successful communication requires learning .
a new - born agent communicating with some mature agents who already worked out a certain ( common in this group ) language will increase the weight of a corresponding word . as a result , in its future communications the agent will use mainly this word . in what way such a learning
might get coupled with evolutionary traits ?
the explanation of this phenomenon is known as a baldwin effect .
although at first sight it looks like a discredited lamarckian phenomenon , baldwin effect is actually purely darwinian@xcite .
there are usually some benefits related with the task a given species has to learn and there is a cost of learning this task .
one can argue that in such a case there is some kind of evolutionary pressure that favours individuals for which the benefit is larger or the cost is smaller .
then , evolution will lead to the formation of species where the learned behaviour becomes an innate ability .
it should be emphasized that the acquired characteristics are not inherited .
what is inherited is the ability to acquire the characteristics ( the ability to learn ) @xcite . in the context of language evolution the importance of the baldwin effect
was suggested by pinker and bloom @xcite .
perhaps this effect is also at least partially responsible for the formation of the language acquisition device - the hypothetical structure in our brain whose existence was suggested by chomsky .
however , many details concerning the role of the baldwin effect in the evolution of language remain unclear @xcite
. in our model the baldwin effect is also at work .
let us consider a population of agents with the communication probability @xmath8 below the threshold value ( @xmath29 ) .
in such a case learning ability remains at a rather low level ( since clusters of agents using the same language are small , it does not pay off to be good at learning the language of your neighbours ) .
now , let us increase the value of @xmath8 above the threshold value ( fig .
[ time ] ) .
more frequent communication changes the behaviour dramatically .
apparently , clusters of agents using the same language are now sufficiently large and it pays off to have a large learning ability because that increases the success rate and thus the survival probability @xmath17 .
let us notice that @xmath17 of an agent depends on its linguistic performance ( @xmath16 ) rather than its learning ability .
thus clusters of agents of good linguistic performance ( learned behaviour ) can be considered as niches that direct the evolution by favouring agents with large learning abilities , which is precisely the baldwin effect .
it should be noticed that linguistic interactions between agents ( whose rate is set by the probability @xmath8 ) are typically much faster than evolutionary changes ( set by @xmath19 ) . to observe such a difference in our simulations shown in fig .
[ time ] we increased @xmath8 up to the value 0.98 . and indeed , the linguistic changes ( success rate ) , that might be considered as niches - forming processes , are ahead of the evolutionary adaptations ( learning ability ) . as a result of a positive feedback ( large learning ability enhances communication that enlarges clusters that favours
even more the increased learning ability ) a discontinuous transition takes place both with respect to the success rate and learning ability ( fig .
[ steady ] ) .
an interesting question is whether such a behaviour is of any relevance in the context of human evolution .
it is obvious that development of language , which probably took place somewhere around @xmath30 years ago , was accompanied by important anatomical changes such as fixation of the so - called speech gene ( foxp2 ) , descended larynx or enlargement of brain @xcite .
linguistic and other cultural interactions that were already emerging in early hominid populations were certainly shaping the fitness landscape and that could direct the evolution of our ancestors via the baldwin effect .
our model predicts that when intensity of linguistic ( or cultural ) processes was large enough , a transition took place where both linguistic performance and biological endowment of our species experienced an abrupt change that perhaps lead to the rapid expansion of human civilization .
but further research would be needed to claim that such a transition did take place and explain it within the framework suggested in our paper .
in the present paper we examined an evolutionary naming - game model .
simulations show that coupling of linguistic and evolutionary ingredients produces a discontinuous transition and learning can direct the evolution towards better linguistic abilities ( baldwin effect ) .
the present model computationally is not very demanding .
it seems to be possible to consider agents talking on more than one object , or to examine statistical properties of simulated languages such as for example , distributions of their lifetimes or of the number of users .
one can also study effects like diffusion of languages , the role of geographical barriers , or formation of language families .
there is already an extensive literature documenting linguistic data as well as various computational approaches modeling for example competition between already existing natural languages @xcite .
the dynamics of the present model , that is based on an act of elementary communication , offers perhaps more natural description of dynamics of languages than some other approaches that often use some kind of coarse - grained dynamics .
there are also more physical aspects of the proposed model that might be worth further studies .
as we have already mentioned , when the learning ability is kept fixed , the transition between disordered and linguistically coherent phases seems to be continuous . on the other hand
, such a transition resembles the symmetry breaking transition in the @xmath31-state potts model , where at sufficiently low temperature the model collapses on one of the @xmath31 ground states .
however , in the two dimensional case and for large @xmath31 ( @xmath31 in our case corresponds to the number of all languages used by agents ) such a transition is known to be discontinuous .
of course the dynamics of our model is much different from glauber or metropolis dynamics that reproduce the equilibrium potts model , but very often such differences are irrelevant as long as , for example , the symmetry of the model is preserved ( which is the case for our model ) .
another possibility that would explain a continuous nature of the transition in our case might be a different nature of ( effective ) domain walls between clusters . in our model these domain walls in some cases
might be much softer and that would shift the behaviour of our model toward models with continuous - like symmetry ( as e.g. , xy model ) . to clarify this issue
further work is , however , needed .
+ * acknowledgments : * we gratefully acknowledge access to the computing facilities at pozna supercomputing and networking center .
m. a. nowak , n. l. komarova , and p. niyogi , nature * 417 * , 417 ( 2002 ) .
s. kirby and j. hurford , _ the emergence of linguistic structure ; an overview of the iterated learning model _ , in _ simulating the evolution of language _ , a. cangelosi and d. parisi ( eds . ) ( springer - verlag , berlin , 2001 ) .
h. brighton , artif .
life * 8 * , 25 ( 2002 ) .
l. steels , artif .
life * 2 * , 319 ( 1995 ) .
h. yamauchi , _ baldwinian accounts of language evolution _ ,
phd thesis , the university of edinburgh , edinburgh , scotland ( 2004 ) .
p. turney , _ myths and legends of the baldwin effect_. in t. fogarty and g. venturini ( eds . ) , proceedings of the icml-96 ( 13th international conference on machine learning , bari , italy ) a. baronchelli , m. felici , v. loreto , e. caglioti , and l. steels , j. stat .
* p06014 * ( 2006 ) . l. dallasta , a. baronchelli , a. barrat , and v. loreto , phys .
e * 74 * , 036105 ( 2006 ) .
d. nettle , lingua * 108 * , 95 ( 1999 ) .
* ibid . , * * 108 * , 119 ( 1999 ) .
java applet that illustrates dynamics of our model is available at : http://spin.amu.edu.pl/~lipowski/biolin.html or http://www.amu.edu.pl/~lipowski/biolin.html g. hinton and s. nowlan , complex systems * 1 * , 495 ( 1987 ) .
s. pinker and p. bloom , behav .
brain sci .
* 13 * , 707 ( 1990 ) . s. munroe and a. cangelosi , artif .
life * 8 * , 311 ( 2002 ) . c. holden ,
science * 303 * , 1316 ( 2004 ) d. abrams and s. h. strogatz , nature * 424 * 900 ( 2003 ) . c. schulze , d. stauffer , and s. wichmann , commun .
phys . * 3 * , 271 ( 2008 ) .
p. m. c. de oliveira , d. stauffer , s. wichmann , and s. m. de oliveira , _ a computer simulation of language families _ , e - print : arxiv:0709.0868 . | we examine an evolutionary naming - game model where communicating agents are equipped with an evolutionarily selected learning ability .
such a coupling of biological and linguistic ingredients results in an abrupt transition : upon a small change of a model control parameter a poorly communicating group of linguistically unskilled agents transforms into almost perfectly communicating group with large learning abilities .
when learning ability is kept fixed , the transition appears to be continuous .
genetic imprinting of the learning abilities proceeds via baldwin effect : initially unskilled communicating agents learn a language and that creates a niche in which there is an evolutionary pressure for the increase of learning ability.our model suggests that when linguistic ( or cultural ) processes became intensive enough , a transition took place where both linguistic performance and biological endowment of our species experienced an abrupt change that perhaps triggered the rapid expansion of human civilization . |
it is well known from direct observations that the solar chromosphere and transition region are far from homogeneous ( see , for example , the uv images shown by reeves 1976 and gallagher et al .
1998 ) . in spatially resolved observations of the sun
, transition region line emission tends to be dominated by contributions from the network boundaries .
the solar - stellar analogy tells us that a similar effect must be expected on the surfaces of other main - sequence stars .
therefore , it is of interest to determine the filling factor of emitting material in solar - like stars and look for correlations with stellar parameters , such as has been done for surface ( photospheric ) magnetic filling factors by montesinos & jordan ( 1993 ) .
such correlations may help to identify the physical processes which control coronal / chromospheric activity .
the current generation of high resolution spectrographic instruments are providing spectra of unprecedented quality in the ultraviolet ( uv ) and x - ray wavelength ranges . unfortunately
, such data can not provide direct information about the spatial distribution of emitting material in main - sequence star atmospheres since it is beyond the scope of current instruments to resolve dwarf star surfaces , with the exception of the sun .
there is little prospect of overcoming this limitation in the near future . in the absence of spatial information , semi - empirical modelling of main - sequence star spectra
is effectively limited to the creation of `` mean '' 1-d atmospheric models whose parameters are those of a hypothetical homogeneous atmosphere which reproduces the observed spectrum ( e.g. the @xmath0 eri models constructed by kelch 1978 or sim 2002 ) .
although such models are useful , tighter constraints on the true atmospheric parameters , and therefore the nature of the physical processes which operate , could be achieved if the distribution of emitting material were known .
( information on the structure within the atmospheres of cool _ giant _ stars has been obtained through studies of molecular fluorescence , see mcmurry & jordan 2000 . )
the approach which could be taken is illustrated by the work of cuntz et al .
they have explored two - component chromospheres of k2 v stars , in which one component is heated by longitudinal tube waves and the other by acoustic waves .
they simulate the supergranulation network using a flux tube model based on empirical correlations between observed surface magnetic fluxes and stellar rotation rates ( cuntz , ulmschneider & musielak , 1998 ) and calculate the flux in the ca ii h and k lines expected from magnetoacoustic and acoustic waves . in the absence of observational constraints they use an average chromospheric network filling factor of 0.305 .
thus a method which gives the filling factor in the transition region between the chromosphere and corona would provide valuable constraints on the above type of theoretical modelling .
kopp ( 1972 ) discussed the energy balance in the solar transition region , in the light of early observations of the solar supergranulation structure .
he introduced an area filling factor and by comparing emission measure distributions based on spatially integrated observations with the predictions of the energy balance equation he proposed that the transition region emission is restricted to an area of about 20 per cent of the solar surface area . in this case , the conductive flux in the mid - transition region is reduced and the lower transition region requires an additional source of heating .
this type of structure was also explored by gabriel ( 1976 ) .
the influence of the cross - section area and magnetic field expansion factor has also been recognized in the context of loop models of solar and stellar active regions ( see underwood , antiochos & vesecky 1981 ) . a method of deriving area factors
has also been discussed by jordan ( 2000 ) , in which the total pressure is not kept constant but is allowed to vary according to hydrostatic equilibrium .
although philippides ( 1996 ) used this approach to estimate the filling factor at @xmath2 k in several main - sequence stars , the observational data available were not sufficient to find the variation of the filling factor with @xmath3 .
here we apply the method to spatially unresolved stellar observations to derive , for the first time , the area filling factor as a function of @xmath3 in the upper transition region of @xmath0 eri .
the star selected for this analysis , @xmath0 eri ( k2v ) , is a nearby dwarf that has moderately high levels of chromospheric / coronal activity and for which high quality spectroscopic data are available .
this paper forms part of our study of the outer atmosphere of @xmath0 eri : see jordan et al .
( 2001a ) , jordan et al .
( 2001b ) , sim & jordan ( 2003 ) and sim ( 2003 , in preparation ) for details . in the next section
the theory required for the analysis is discussed .
section 3 discusses the observations that are used to constrain the emission measure distribution ( section 4 ) .
section 5 presents `` first - cut '' area filling factors deduced from the data and discusses possible sources of error . in section 6
, a self - consistent determination of the filling factors is presented .
the results are discussed in section 7 and conclusions are drawn in section 8 .
spatially unresolved stellar observations can be used to constrain the apparent mean emission measure ( see pan & jordan 1995 ) distribution defined by @xmath4 where @xmath5 is the electron temperature at radius @xmath6 , @xmath7 and @xmath8 are the electron and hydrogen number densities respectively , @xmath9 ( @xmath10 is the photospheric stellar radius and @xmath11 at the stellar surface ) .
the dilution factor @xmath12 specifies the fraction of photons ( emitted in an optically thin line at radial coordinate @xmath6 ) that escape the star uninterrupted by the stellar surface . for spherical geometry , @xmath13^{-1 } } } \right ) \ ; \ ; .\ ] ] the @xmath14 included in the notation for the emission measure in equation ( 1 ) is to remind the reader that the emission measure adopted is an integral over a logarithmic temperature range @xmath15 dex .
this temperature range is chosen since it is the typical temperature range across which individual spectral lines form in the transition regions of cool stars .
the area filling factor @xmath16 specifies the fraction of the total stellar surface area at radius @xmath6 ( @xmath17 ) which is occupied by emitting material .
the purpose of this paper is to show how values for @xmath16 can be extracted from spatially unresolved data . in the transition region where the pressure varies slowly as a function of the temperature , equation ( 1 ) can be written approximately as @xmath18 where @xmath19 and @xmath20 are the electron and hydrogen pressures respectively , and are approximately constant over the region where an individual line is formed .
the approximation that @xmath21 is constant in the region of line formation is less good , but is improved upon when finding final models by iteration .
as discussed by jordan ( 2000 ) , it is possible to explain the complete radiation losses from the upper transition region in terms of the divergence of the net conductive energy flux from the overlying corona .
the justification for this is that the emission lines are formed over a pressure - squared scale height @xmath22 km , whereas the main heating of the quiet corona occurs at much greater heights .
it can not be ruled out that there is an additional source of heating in this part of the atmosphere , but it is certain that any additional heating must be small . the conductive flux can be expressed as @xmath23 where @xmath24 ergs @xmath25 s@xmath26 k@xmath27 at @xmath28 k ( spitzer 1956 ) . @xmath29 increases at lower temperature ( by about @xmath30 per cent by @xmath31 k ) , but it is taken as constant here for simplicity .
building on the simpler plane parallel approximation used by jordan & brown ( 1981 ) , the logarithm of equation ( 3 ) can be differentiated and the temperature gradient replaced by the conductive flux using equation ( 4 ) . by assuming an energy balance between conduction and radiation , the net conductive flux can be replaced by the radiation losses and the following equation obtained : @xmath32 = \frac{\mbox{d } \log g(r)}{\mbox{d } \log t_{e } } + 2\frac{\mbox{d } \log f(r)}{\mbox{d } \log t_{e } } \\
+ \frac{\mbox{d } \log a(r)/a_{*}(r)}{\mbox{d } \log t_{e } } + 2\frac{\mbox{d } \log p_{g}}{\mbox{d } \log t_{e } } + \frac{3}{2 } \\ - \frac{2 p_{\mbox{\scriptsize rad}}(t_{e})}{\kappa p_{e } p_{h } t_{e}^{3/2 } g(r)^2 f(r)^2 } \left [ { \frac{a_{*}(r ) em^{0.3}}{a(r ) } } \right]^2\ ; .\end{gathered}\ ] ] @xmath33 is the radiative power loss function which describes the emissivity of a plasma of fixed composition at temperature @xmath5 .
@xmath34 is the total gas pressure .
this equation can be solved numerically to obtain the distribution of @xmath35 as a function of @xmath5 . by comparison with values of @xmath36 ( deduced from observations ) this allows the area filling factor @xmath16 to be derived .
the mean emission measure distribution can be constrained using observed emission line fluxes .
the details of this procedure are well known and have been discussed previously by various authors ( e.g. jordan & brown 1981 , jordan et al .
1987 ; griffiths & jordan 1998 ) . in this paper , the mean emission measure distribution in the upper transition region of @xmath0 eri is constrained using the observed fluxes of emission lines of iron . for the most part ,
the observed fluxes used are from the analysis performed by schmitt et al .
( 1996 ) using data recorded by the _ extreme ultraviolet explorer _ ( _ euve _ ) satellite .
the data and line flux measurements are discussed by schmitt et al .
( 1996 ) and laming , drake & widing ( 1996 ) .
schmitt et al . ( 1996 ) give fluxes for fe lines from ionisation stages ix xvi and xviii xxi .
for the current work , only one line from each ionisation stage detected in the _ euve _ data is used . for
fe xii the three lines at around 193 have been added to give the total flux in the @xmath37s @xmath37p multiplet . in other cases where there are several lines from the same ion reported by schmitt et al .
( 1996 ) , the strongest line has generally been used .
the lines used are listed in table 1 together with the percentage errors in the measured fluxes given by schmitt et al .
( 1996 ) . [ cols="<,^,^,^",options="header " , ] @xmath38 rest wavelength .
given that there are several sources of systematic error , the agreement between the form of @xmath39 used in the modelling and the values of @xmath39 predicted by the model is sufficiently good that the higher pressure model can be regarded as self - consistent if the arnaud & raymond ( 1992 ) ionisation balance is adopted .
it is noteworthy that the largest discrepancies are for fe ix , fe xv and , in particular , fe xvi which are the ions that show the largest uncertainties in the ionisation balance .
the area factors deduced from the model with lower pressure are slightly less self - consistent than those obtained from the higher pressure model , being systematically larger than those assumed in the modelling . given the scale of the other sources of error , however , the agreement is adequate . comparing tables 2 and 4 , it is clear that the main effect of including the term describing the variation of the area factor with temperature is to reduce the implied area factors for most of the high temperature lines , leading to fewer unphysical ( @xmath40 ) values .
this is the result of the steeper increase of emission measure with temperature .
the only exception is fe xvi for which the implied area factor is _ larger _ when the variation of the area factor is included in the model if the arnaud & raymond ( 1992 ) ionisation fractions are used .
this is because fe xvi forms at temperatures above the peak of the emission measure distribution , if the arnaud & raymond ( 1992 ) ionisation balance calculations are adopted . in summary
, the self - consistent treatment of the variation of the area factor presented in this section indicates that neglecting this variation in the modelling may cause the area factors in the upper transition region to be systematically overestimated to a small , but significant , extent .
accounting for the variation of the area factors does not , however , alter the apparent trend that the area factor increases across the upper transition region .
for now it is concluded that the area factors presented in table 4 ( or equivalently described by equation 10 ) are the best current estimates of the area filling factors in the upper transition region of @xmath0 eri and that most of them are accurate to within a factor of two .
in sections 5 and 6 it has been shown that the theory described in section 2 can be used to provide estimates of the area filling factors of emitting material at various temperatures in stellar transition regions .
the results show ( fig .
5 ) that in @xmath0 eri the area filling factor increases with temperature markedly in the upper transition region , from @xmath41 at @xmath42 to @xmath43 in the corona .
there is no evidence for significant variations across the mid - transition region @xmath44 .
the variation of the area factor derived for @xmath0 eri is similar to that found from observations of the solar supergranulation network made with the _ solar and heliospheric observatory _ ( _ soho _ ) ( gallagher et al .
they find ( see their figs . 5 and 8 , in particular ) that at the resolution of the coronal diagnostic spectrometer ( cds ) ( @xmath45 2 arcsec by 2 arcsec ) the area occupied by the emitting material increases dramatically in the upper transition , but does not vary significantly across the mid - transition region
. these results are in general agreement with earlier observations and the model by gabriel ( 1976 ) . when the method used here is applied to cds observations of the average network , a filling factor of @xmath45 0.2 at @xmath46 k
is deduced _ within _ the network ( smith & jordan 2002 ) , which typically occupies @xmath45 50 per cent of the solar area at this temperature ( gallagher et al .
the filling factors which we have derived are based on the assumption that the observed emission is dominated by the regions occupying the small areas
. these could refer to structures within the supergranulation network of @xmath0 eri and may not be comparable to the total solar network area observed with the resolution of cds . in this case ,
the area factors are _ larger _ than those found for the sun . using the solar analogy , the area factor found at @xmath46 k can be applied to the lower transition region .
this increases the intrinsic radiation losses and since thermal conduction now plays no role below @xmath47 k , an additional source of heating is required .
as found by sim & jordan ( 2003 ) , _ if _ the observed non - thermal emission line widths are associated with an alfvn wave ( or fast - mode wave ) flux , there is ample energy to heat both the corona and the lower transition region , although uncertainties concerning wave propagation and dissipation remain to be resolved .
the filling factor of @xmath48 found for the mid - transition region of epsilon eri can be compared with the value of 0.3 assumed for the chromosphere by cuntz et al .
since the filling factor of magnetic flux tubes should increase with height , our results suggest that a smaller chromospheric filling factor of @xmath49 would be more appropriate ; this could be explored in future two - component models .
the accuracy of the results presented in this paper is limited by the various systematic sources of error which were discussed in section 5.3 .
in particular the analysis of larger sets of lines , ideally covering a wide range of elements and ionisation species would help .
this would reduce the systematic errors associated with uncertainties in the ionisation fractions for iron and possibly help to identify possible errors arising from the adopted elemental abundance .
suitable data for such analysis are gradually becoming available for a range of stars thanks to the current generation of x - ray satellite missions .
x - ray data are also advantageous since their analysis is less sensitive to the assumed interstellar hydrogen column density .
improved calculations of atomic data for iron , and the radiation loss function @xmath50 would also make the modelling more quantitatively reliable . in future work
this technique should be applied to a range of late - type stars with differing levels of activity . by doing
so it should be possible to look for trends in the behaviour of the filling factors of emitting material with various stellar parameters . the ability to reproduce these trends
will then be a new constraint on theoretical studies of the structure and heating of late - type stars , including the magnetic field geometry .
it has been shown that by assuming an energy balance between radiation and thermal conduction in the upper transition region , it is possible to use spatially unresolved stellar observations to estimate the fractional area filling factor of emitting material in stellar atmospheres .
the technique has been applied to @xmath0 eri and it has been found that the area filling factor is @xmath41 in the mid transition region and that it increases to @xmath43 in the corona .
these filling factors are similar to those found from observations of the solar transition region , and from models in which the transition region consists of material in magnetic funnels whose radii are greatest is in the upper transition region / corona .
several potential sources of uncertainty in the calculations have been discussed and it is concluded that , although they are currently significant , they do not prevent a useful application of the methods adopted .
further observations and calculations of atomic data should be able to reduce the current uncertainties .
this work was carried out while sas was a research student in the physics department ( theoretical physics ) at the university of oxford , supported by a pparc studentship ( ppa / s / s/1999/02862 ) and was completed while he was a pparc funded pdra in the astrophysics group at imperial college london ( ppa / g / s/2000/00032 ) .
arnaud m. , raymond j. , 1992 , apj , 398 , 394 + arnaud m. , rothenflug r. , 1985 , a&as , 60 , 425 + cuntz m. , ulmschneider p. , musielak z. e. , 1998 , apj , 493 , + l117 + cuntz m. , rammacher w. , ulmschneider p. , musielak z.e . ,
+ saar s.h . , 1999 ,
apj , 522 , 1053 + crawford h. j. , price p. b. , sullivan j. d. , 1972 , apjl , + 175 , l149 + dere k. p. , landi e. , mason h. e. , monsignori fossi b. c. , + young p. r. , 1997 , a&as , 125 , 149 + drake j. j. , smith g. , 1993 , apj , 412 , 797 + dring a. r. , linsky j. , murthy j. , henry r. c. , moos w. , + vidal - madjar a. , audouze j. , landsman w. , 1997 , + apj , 488 , 760 + gabriel a. h. , 1976 , phil . trans .
london , a291 , + 339 + gallagher p. t. , phillips k. j. h. , harra - murnion l. k. , + keenan f. p. , 1998 ,
a&a , 335 , 733 + grevesse n. , sauval a. j. , 1998 , space science reviews , + 85 , 161 + griffiths n. w. , jordan c. , 1998 , apj , 497 , 883 + jordan c. , 2000 , plasma phys .
fusion , 42 , 415 + jordan c. , ayres t. r. , brown a. , linsky j. l. , simon t. , + 1987 , mnras , 225 , 903 + jordan c. , brown a. , 1981 , in r. m. bonnet & a. k. + dupree eds . , solar phenomena in stars and stellar + systems , p. 199
dordrecht , holland , nato + asic , 68 + jordan c. , doschek g. a. , drake j. j. , galvin a. b. , + raymond j. c. , 1998 , in asp conf . ser . 154 : cool + stars , stellar systems and the sun , vol . 10 , p. 91
+ jordan c. , mcmurry a. d. , sim s. a. , arulvel m. , 2001a , + mnras , 322 , l5 + jordan c. , sim s. a. , mcmurry a. d. , arulvel m. , 2001b , + mnras , 326 , 303 + kelch w. l. , 1978 , apj , 222 , 931 + kopp r. a. , 1972 , solar physics , 27 , 373 + laming j. m. , drake j. j. , widing k. g. , 1995 , apj , 443 , + 416 + laming j. m. , drake j. j. , widing k. g. , 1996 , apj , 462 , + 948 + macpherson k. p. , jordan c. , 1999 , mnras , 308 , 510 + mcmurry a. d. , jordan c. , 2000 , mnras , 313 , 423 + montesinos b. , jordan c. , 1993 , mnras , 264 , 900 + pan h. c. , jordan c. , 1995 , mnras , 272 , 11 + philippides d. , 1996 , d.phil .
thesis , university of oxford + reeves e.m .
, 1976 , solar physics .
, 46 , 53 + schmitt j. h. m. m. , drake j. j. , stern r. a. , haisch b. + m. , 1996 , apj , 457 , 882 + sim s. a. , 2002 , d.phil .
thesis , university of oxford + sim s. a. , jordan c. , 2003 , mnras , 341 , 517 + smith g. r. , jordan c. , 2002 , mnras , 337 , 666 + spitzer l. , 1956 , physics of ionized gases , interscience , new + york + underwood j. h. , antiochos s. k. , vesecky j. f. , 1981 , + in r. m. bonnet & a. k. dupree eds .
solar phenomena + in stars and stellar systems , p. 227
dordrecht , + holland , nato asic , 68 + young p. r. , delzanna g. , landi e. , dere k. p. , mason h. + e. , landini m. , 2003 , a&as , 144 , 135 + [ lastpage ] | the emission measure distribution in the upper transition region and corona of @xmath0 eri is derived from observed emission line fluxes .
theoretical emission measure distributions are calculated assuming that the radiation losses are balanced by the net conductive flux .
we discuss how the area factor of the emitting regions as a function of temperature can be derived from a comparison between these emission measure distributions .
it is found that the filling factor varies from @xmath1 0.2 in the mid transition region to @xmath1 1.0 in the inner corona .
the sensitivity of these results to the adopted ion fractions , the iron abundance and other parameters is discussed .
the area factors found are qualitatively similar to the observed structure of the solar atmosphere , and can be used to constrain two - component models of the chromosphere . given further observations , the method could be applied to investigate the trends in filling factors with indicators of stellar activity .
[ firstpage ] stars : individual ( @xmath0 eridani ) - stars : late - type - stars : coronae |
two - particle correlations provide unique access to the spatial extension of the source of particles emitted in the course of heavy ion collisions at relativistic energy ( for a recent review see @xcite ) .
the relation between the experimental correlation function , defined as the relative momentum distribution of pairs , normalized to the analogous distribution obtained via event mixing , and the size of the fireball is especially simple in case of identical pions where the main source of correlations is the bose - einstein statistics .
in fact , in the pure boson case the correlation function is @xmath0 with the source function @xmath1 describing the single particle density in 8-dimensional position - momentum space at freeze - out .
the correlation function @xmath2 depends on the momentum difference * q * = * p@xmath3 * - * p@xmath4 * and the mean pion momentum * k * = ( * p@xmath4*+*p@xmath3*)/2 .
the width of the peak at * q * = 0 is inversely proportional to the source radius .
a particularly exciting prospect is to look for a source asymmetry possibly reflecting the initial asymmetry of the fireball created in collisions with finite impact parameter .
indeed , significant dependence of the source radii on the azimuthal emission angle with respect to the reaction plane , defined by the beam axis and the impact parameter vector , was observed in au+au collisions at 2 - 6 gev @xcite and at @xmath5=200 gev @xcite . in this paper
we present results of the first analysis of the azimuthal dependence of pion source radii at sps energies .
the ceres / na45 experiment at the cern sps is described in detail in @xcite .
the upgrade by a radial time projection chamber ( tpc ) in 1997 - 1998 , in addition to improving the dilepton mass resolution , enhanced the experiment s capability of studying hadronic observables .
the cylindrical symmetry of the experiment was preserved during the upgrade , making the setup ideally suited to address azimuthal anisotropies .
about 30 million pb+au collision events at 158 were collected in the fall of 2000 , most of them with centrality within the top 7% of the geometrical cross section .
small samples of @xmath6/=20% and minimum bias collisions , as well as a short run at 80 , were recorded in addition .
the results presented here are based on a correlation analysis of the high - statistics pb+au data set from the year 2000 @xcite .
the results of an azimuthal - angle averaged analysis in @xcite are consistent with the previously published ceres data @xcite and with a recent analysis by the na49 collaboration @xcite .
the main steps of the azimuthal - angle dependent analysis are described below .
the collision centrality was determined via the charged particle multiplicity around midrapidity @xmath7/2=2.91 .
two variables , the amplitude of the multiplicity counter ( mc ) ( single scintillator covering [email protected] ) and the track multiplicity in the tpc ( [email protected] ) , were alternatively used as the centrality measure . knowing the data acquisition dead time factor and the target thickness , and assuming that all beam particles were hitting the target , the event counts were translated to the cross section for collisions with a given multiplicity .
the centrality variable used in this paper is defined as the integrated cross section divided by the geometrical cross section = 6.94 barn . the fireball created in a collision with finite impact parameter
is elongated in the direction perpendicular to the reaction plane . in the course of expansion , with the pressure gradient larger
in - plane than out - of - plane , the initial asymmetry should get reduced or even reversed .
experimentally , the source eccentricity at the time of decoupling can be determined from an analysis of two - pion correlations as function of @xmath9 , the pair emission angle with respect to the reaction plane .
the azimuthal angle of the reaction plane @xmath10 was estimated by the preferred direction of the particle emission ( elliptic flow ) .
for this , in each event the flow vector @xmath11 was constructed out of the measured particles , characterized by transverse momenta @xmath12 and azimuthal emission angles @xmath13 @xcite : @xmath14 the @xmath15 and @xmath16 components were calibrated ( shifted and scaled ) such that the peak in the @xmath17 distribution was centered at ( 0,0 ) , and its widths in the directions of @xmath18 and @xmath19 were equal .
the reaction plane angle was calculated ( modulo @xmath20 ) from the calibrated @xmath11 via @xmath21 the resolution of the so determined reaction plane angle , estimated via the subevent method @xcite , was 30 - 34 .
the event mixing , needed to obtain the denominator of the correlation functions , was performed in bins of centrality ( 2% of ) , event plane ( 15 ) , and vertex ( same target disk ) .
this ensures that the shape of the background is identical to that of the signal in all respects except for the femtoscopic correlations .
only tpc tracks with at least 12 ( out of maximally 20 ) hits and a reasonably good @xmath22 , falling into the fiducial acceptance @xmath23 , were used in the analysis .
a momentum dependent d@xmath24/d@xmath25 cut was applied to reduce the contamination of the pion sample .
pions from @xmath26 and @xmath27 decays were suppressed by a 2.5 @xmath6 matching cut between the silicon vertex detectors and the tpc .
the two - track resolution cuts applied to the true pairs and to the pairs from event mixing were different for the two possible track pair topologies : in the case of the magnetic field bringing the tracks apart from each other ( sailor ) the required two - track separation was @xmath2838 - 45 mrad , depending on the transverse momentum ; for the opposite ( cowboy ) case , @xmath2890 - 140 mrad was used .
the polar angle cut was the same for both topologies ( @xmath298 - 9 mrad ) .
it should be noted that the required two - track cuts depended somewhat on the quality cuts applied to single tracks : the higher the number of hits required for single tracks , the more pairs were lost because of the finite two - track resolution .
the two - pion analysis was performed in the longitudinally co - moving frame ( lcms ) defined by the vanishing @xmath30-component of the pair momentum . the momentum difference in this frame , * q * = * p*@xmath3-*p*@xmath4 , was decomposed into the `` out '' , `` side '' , and `` long '' components following the bertsch - pratt convention , with @xmath31 pointing along the pair transverse momentum and @xmath32 along the beam @xcite .
the particles were numbered such that was always positive .
the and correlation functions were fitted by @xmath33 with the indices @xmath34,@xmath35 = \{out , side , long}. the normalization factor @xmath36 was needed because the number of pairs from event mixing was four times higher than of signal pairs .
the correlation strength @xmath37 reflects the contribution of pions from long - lived resonances , the contamination of the pion sample by other particle species , and the finite @xmath38-resolution .
the resulting source radii @xmath39 describe the size of the source emitting pions of a given momentum @xcite .
the @xmath40 factor , @xmath41 , accounts for the mutual coulomb interaction between the pions and was calculated by averaging the nonrelativistic coulomb wave function squared over a realistic source size .
the coulomb factor was attenuated by @xmath42 similarly as the rest of the correlation function peak ; the importance of this approach was demonstrated in @xcite .
the fits were performed by the minimum negative log - likelihood method , assuming that the number of true pairs in a given bin is distributed around the expected mean value ( equal to the number of mixed pairs times the fit function ) according to the poisson statistics .
the source radii obtained from the fit were corrected for the finite momentum resolution @xmath43 the correction was determined by monte carlo and is rather insignificant for and ; for it gets as large as @xmath44 + 20% for the highest pion momenta .
correlation functions were generated separately for pion pairs with different azimuthal angles with respect to the reaction plane @xmath9 .
for this , the pions were sorted into eight bins covering @xmath45 .
the eight correlation functions were fitted s discussed in this paper are weighted averages of the @xmath46 s obtained for different s . ] and the resulting squared source radii were parametrized by @xcite @xmath47 while the @xmath48 , @xmath49 , and @xmath50 obtained represent the phi - averaged squared pion hbt radii , the second fourier coefficients @xmath51 describe the eccentricity of the observed pion source . since the reaction plane is known modulo @xmath20 odd fourier components should vanish . if the pion source were to reflect the initial collision geometry ( almond shape out - of - plane ) a positive @xmath52 and @xmath53 and a negative @xmath54 should be expected . for
symmetry reasons all anisotropies should disappear in the limit of central collisions .
the second fourier coefficients @xmath51 have been corrected for the reaction plane resolution by dividing them by the mean cosine of twice the difference between the reconstructed and the true reaction plane angles @xmath55 \rangle$ ] , estimated via the measured mean cosine of the difference between two subevents : @xmath56\rangle } } , \label{eq : rpres}\ ] ] similarly as it is done for flow measurements @xcite .
the correction factor was between 4.8 and 2.4 ( centralities of 0 - 2.5% and 25 - 70% , respectively ) .
the appropriateness of the flow correction for two - pion correlation radii has been verified with a monte carlo simulation of pion pairs emitted from an elliptical source ; the observed reduction of the second fourier coefficients of squared radii was , within the statistical errors of the simulation ( about 10% ) , in agreement with eq .
( [ eq : rpres ] ) .
the fact that separate analyses of positive and negative pions give consistent results indicates that the limited particle identification is not causing any bias .
the systematic error related to the reaction plane resolution correction , based on the comparison between eq .
( [ eq : rpres ] ) and a numerical simulation , is estimated to be not larger than 10% .
autocorrelations , i.e. using the same particles to determine the event plane and the hbt radii , might be another source of systematic errors ; unfortunately , an attempt to exclude the two particles from the event plane construction led to a strong distortion of the correlation functions as it was not clear how to do the exclusion consistently for the true and mixed pairs . an independent way to estimate the overall systematic errors
is by inspecting ( see fig .
[ fig : r2 ] in the next section ) the coefficients @xmath57 and @xmath57 which should be zero because the reaction plane is known only modulo @xmath20 . averaged over centralities
, they are 0.40 ( 16 ) and 0.22 ( 6 ) @xmath58 , respectively . attributing part ( one sigma ) of the deviation to statistical fluctuation we are left with a discrepancy of about 0.3 @xmath58 which we thus assume to be the systematic uncertainty of the results presented .
the extracted fourier coefficients are shown in table [ table : centrality ] and fig .
[ fig : r2 ] . from normalizing the coefficients to the corresponding mean source radii
in fact , since the second fourier coefficients measured at rhic @xcite seem to be , within the measurement errors , independent of transverse momentum while the mean radii vary by about 40% , normalization to mean radii could wash out the weak oscillation we are after . ] [ cols="^,^,^,^,^,^,^,^,^,^ " , ] , measured in pb+au collisions as a function of centrality .
positive and negative pion pairs have been combined .
the mean pion transverse momentum is 0.23 gev / c .
for comparison , analogous measurements at the ags ( open green symbols ) and rhic ( blue diamonds and stars for 130 and 200 gev , respectively ) are shown .
the last two panels show the first fourier coefficients @xmath59 and @xmath60 which should be zero by construction .
the estimated systematic error is 0.3 @xmath58 . ]
the anisotropies in the out and side directions indicate a pion source elongated out - of - plane . for comparison , the ags @xcite and rhic @xcite values , obtained by performing the fits using eq .
( [ eq : rfourier ] ) on their published radii ( the ags results were subsequently corrected for their reaction plane resolution ) , are represented in fig .
[ fig : r2 ] by open symbols and stars .
it appears that the anisotropies in the out and side directions at sps energy are rather similar to those observed at rhic .
the geometrical pion source eccentricity can be quantified via @xmath61 and has , for a centrality of 15 - 20% , a value of 0.043(27 ) , significantly less than the initial fireball eccentricity @xmath62 @xcite . the fact that the magnitude of the anisotropy seems to be weaker than that of indicates that not only the source geometry but also azimuthal dependence of the emission time might play a role . on the other hand ,
@xmath63 + 2@xmath53 , averaged over centralities , is 0.06(18 ) @xmath58 , consistent with zero .
this implies that the sum rule from @xcite , which is supposed to be valid for systems with emission time independent on the azimuthal angle , works rather well .
the anisotropy is negative and roughly independent of centrality .
series of checks were performed to make sure that this is not an artefact of the analysis .
this result might indicate that is sensitive to fluctuations of the azimuthal particle density .
hydrodynamic calculation of central pb+au collisions at this energy yields a similar amount of anisotropy in @xmath64 which , however , is centrality dependent @xcite .
this , and the fact that the hydrodynamic calculation overpredicts the overall magnitude of , indicates that the knowledge about the mechanism leading to oscillations of is still incomplete .
the source radius anisotropies for centrality 15 - 20% are shown as a function of the collision energy in fig .
[ fig : vssqrts ] .
the sps result fits rather well into the beam energy systematics .
the apparent ( albeit not statistically significant ) fast change of @xmath52 between ags and sps and , even more , the negative @xmath65 developing when going down in energy from rhic to sps , make the perspective of a systematic study in the course of the low - energy scan at rhic especially attractive . with the statistical errors of the present ags data even significant structures in the energy dependence of the pion source anisotropy
can not be excluded .
we have analyzed the azimuthal angle dependence of the pion hbt radii in pb+au collisions at the top sps energy .
the source anisotropy in the out and side directions has the same sign and similar magnitude as the one measured at the ags and at rhic , and indicates a pion source elongated out - of - plane .
the side anisotropy is somewhat smaller than the other which suggests that finite emission times may play a role .
the source anisotropy in the long - direction is negative for all centralities indicating that might be sensitive to particle density fluctuations .
the ceres collaboration acknowledges the good performance of the cern ps and sps accelerators as well as the support from the est division .
we would like to thank r. campagnolo , l. musa , a. przybyla , w. seipp and b. windelband for their contribution during construction and commissioning of the tpc and during data taking .
we are grateful for excellent support by the cern it division for the central data recording and data processing .
this work was supported by gsi , darmstadt , the german bmbf , the german vh - vi 146 , the us doe , the israeli science foundation , the check science foundation contract no . 202/03/0879 and the minerva foundation .
lisa , s. pratt , r. soltz , u. wiedemann , ann.rev.nucl.part.sci.55(2005)357 m.a .
lisa et al . , phys .
b496 , 1 ( 2000 ) .
j. adams et al .
93 , 012301 ( 2004 ) .
d. adamov et al . , arxiv:0802.1443v1[nucl - ex ] , accepted for publication in nucl .
d. antoczyk , ph.d .
thesis , technical university darmstadt 2006 .
d. adamov et al .
a714 , 124 ( 2003 ) .
s. kniege , ph .
d. thesis , frankfurt university 2005 ; c. alt et al , arxiv:0709.4507v2 [ nucl - ex ] .
p. danielewicz and g. odyniec , phys .
157b , 146 ( 1985 ) .
j. barrette et al , phys .
73 , 2532 ( 1994 ) ; s. voloshin and y. zhang , z.phys.c70 , 665 ( 1996 ) .
a.m. poskanzer and s.a .
voloshin , phys . rev .
c58 , 1671 ( 1998 ) .
bertsch , nucl .
a498 , 173c ( 1989 ) ; s. pratt , phys .
d33 , 1314 ( 1986 ) .
u. heinz , a. hummel , m.a .
lisa , u.a .
wiedemann , phys .
c66 , 044903 ( 2002 ) .
wells , ph.d .
thesis , ohio state university 2002 .
calculation by p. huovinen .
the paper is in preparation . | we present results of a two - pion correlation analysis performed with the au+pb collision data collected by the upgraded ceres experiment in the fall of 2000 .
the analysis was done in bins of the reaction centrality and the pion azimuthal emission angle with respect to the reaction plane .
the pion source , deduced from the data , is slightly elongated in the direction perpendicular to the reaction plane , similarly as was observed at the ags and at rhic . |
the zeldovich approximation ( za ) ( @xcite ) describes structure formation in the form of a deceptively simple equation @xmath2 rather than describing just ballistic motion , this equation hides a formalism of lagrangian collision - free fluid mechanics . in the context of emerging interest in phase - space folding descriptions of structure formation ( @xcite ) , it becomes all the more relevant to understand this expression at a much deeper level .
we can see why there is more to the za than inertial motion , if we compute particle densities from the above expression .
density increases or decreases locally as a fluid element contracts or expands . taking a fluid element from lagrangian location @xmath3
, we can quantify its deformation in terms of the deformation tensor @xmath4 .
this tensor is best studied locally in the eigenvector frame of reference @xmath5 , where @xmath6 becomes diagonal .
the density is then @xmath7 where @xmath8 is the set of lagrangian locations solving for @xmath9 in equation [ eq : zeld ] , and @xmath10 are the eigenvalues of the deformation tensor @xmath6 .
note that this expression for the density has singularities whenever for one of the eigenvalues we have @xmath11 this notion gives us the traditional interpretation of za , namely one where gravitational collapse occurs in three possible stages .
first objects form along the major axis of collapse , making pancakes ; then along the second eigenvector filaments form ; and finally if and when all three eigenvalues have passed a singularity , a cluster forms .
this corresponds to the _ morse theory _ view of nodes and saddles exploited in many structure finders .
we show that , even in the narrow confines of the za , this interpretation is not complete ; even that it is wrong on the account of the formation of the first filaments ( @xcite ) . taking the mathematics of morse theory a step further , we arrive at lagrangian _ singularity theory _ ( @xcite ) .
this theory shows how to predict the evolution of _ folds _ , _ cusps _ , _ swallow - tails _
, _ butterflies _ and _ umbilics _ directly from the initial velocity potential @xmath12 .
due to the relative complexity of this method we are forced to restrict our further discussion to the two - dimensional case . + -axis and the two remaining axes in eulerian space . on the right
is the projected view , with the fold and cusp locations marked . between the folds we find a three - stream region . ]
a _ fold _ is the simplest kind of singularity we have .
it is the caustic that separates single - stream from multi - stream regions and is also found under the cryptic name @xmath13 .
refers to the ade classification of singularities introduced by arnold . in this paper
we also deal with @xmath0 for cusp , @xmath14 for swallow - tail , and @xmath15 for umbilic singularities .
] at any moment in time , pancakes can be identified as the locations of @xmath13 folds . at a fold the phase - space sheet ( see fig .
[ fig : cusp ] ) is tangent to line of projection . in the case of za ,
this happens when @xmath16 , identifying the level - sets of @xmath17 as the lagrangian progenitors of folds .
two folds may connect at a _
cusp_. a fold being a line of tangency on the phase - space sheet , there exists points where the fold line itself is tangent to the projection , these points are the cusps ( see fig . [
fig : cusp ] ) . in the tensor field @xmath6
, a cusp is found where a level - set of @xmath17 is tangent to the corresponding eigenvector @xmath18 , or @xmath19 finding all @xmath0-points for each level - set of @xmath17 traces an @xmath0-line .
the set of @xmath0-lines trace the entire network of filaments formed in the za , throughout time .
it can be shown that all maxima and saddle points of the function @xmath20 also lie on an @xmath0-line . lowering a level - plane down on the function @xmath20
, we can see that at the maxima of @xmath17 ( @xmath21-points ) two cusps are created , while at the saddle points ( @xmath22-points ) they annihilate , merging two pancakes .
@xmath0-lines terminate only in @xmath15 _ umbilic _ points , but we choose to also truncate them where @xmath23 , beyond which they loose their physical significance .
+ -point .
the eulerian counterpart of the dashed contour is shown on the right .
it shows the pancake at its prime , highly elongated with a cusp on each end . ]
-lines in red , and triangles showing @xmath15 points . on the right : the density contrast resulting from a 2d pm code , with the eulerian displaced @xmath0-lines over - plotted . ]
a pancake may branch by creating two new cusps at a fold .
typically one of these cusps remains within the confines of the present pancake and the other dashes out to create a subsidiary pancake , often to merge later with another pancake at an @xmath22-point , creating a three - legged structure .
the point at which a pancake branches is called a swallow - tail , denoted @xmath14 .
an @xmath14 singularity is found in lagrangian space at points where an @xmath0-line is tangent to the corresponding eigenvector , or equivalently at local maxima of @xmath17 limited to the @xmath0-line .
important to note here , is that we do nt need to involve the second eigenvalue to create a node in the network of caustics . moving this discussion to the three - dimensional case
, we do nt strictly need to collapse along the second eigenvector to create a filament - like structure .
this also suggests the existence of different possible late - time morphologies for filaments .
+ we computed the @xmath0-lines for a set of initial conditions and compared the result with those of a 2d n - body code .
a non - linear time evolution may be approximated by truncating power of the initial conditions at scales smaller than the scale of non - linearity .
it is well known that observable filaments have a density contrast around unity , so this method should give realistic results .
we find good agreement for all relevant @xmath24 power - spectra , in an eye - ball comparison of filaments predicted by za with density fields from a 2d pm code .
an example is given in fig .
[ fig : nbody ] .
we showed how the emergence of caustics in the za allows us to trace the formation of cosmic structures in a formal , yet physically meaningful way .
however , the za suffers from a major flaw in that it does nt allow for gravitational interaction , and therefore hierarchical structure formation .
this is because the za is solely based on local analysis of the velocity potential and its chain of derivatives .
the adhesion model moves beyond local considerations , which makes it computationally more intensive than the za ( though still much faster than n - body ) .
still , results are computed from initial conditions directly and with complete accountability .
the adhesion model is arrived at by taking the source - free ( hence collision - free ) euler equation , and adding an artificial viscosity term to _ emulate _ the effects of gravity ( @xcite ) .
the resulting equation @xmath25 is known as burgers equation ; in the limit where @xmath26 , it has the exact solution @xmath27 the _ global maximum _ in this solution can be computed efficiently using either a legendre transform , convex hull ( @xcite ) or a weighted voronoi diagram ( @xcite ) .
one condition for reaching an extremum is that the first derivative of the maximised quantity should vanish .
performing this test reduces above equation to the za as presented in equation [ eq : zeld ] .
the global maximisation guarantees that the resulting map from lagrangian to eulerian space stays _ monotonic _ always . where and whenever shell - crossing occurs in the za , adhesion creates a solid structure .
matter inside these structures is confined to stay inside , which is the reason people may refer to the adhesion model as having `` sticky particles '' . outside collapsed structures the results from the za and adhesion are identical ; caustics from za are compressed to infinitesimally thin structures ( see fig .
[ fig : adhesion ] ) .
this unifies zeldovich idea of collapsed structures in terms of shell crossing with a hierarchical formation model .
+ the solution to burgers equation given in expression [ eq : burgsol ] is identical to the definition of the _ weighted voronoi tessellation _
, weighted by the potential . the voronoi cell of a lagrangian point @xmath28 occupies an area in eulerian space @xmath29 given by @xmath30 taking @xmath31 as the weights in this expression reduces it to the given solution in equation [ eq : burgsol ] . where there is a voronoi tessellation , there is its dual : the _ delaunay triangulation_. it is the latter that gives us the origin and mass of matter residing in the nodes , edges and faces of the voronoi tessellation ( see fig .
[ fig : dual ] ) .
one application of the adhesion model is the detection of walls , filaments and nodes in cases where some form of an initial potential is available .
we ran our adhesion code on a set of 25 constrained initial conditions _ reconstructed _
( @xcite ) to produce structures in our local universe ( @xmath32 ) ( @xcite ) .
this reconstruction is based on the 2mass redshift catalog ( @xcite ) , which covers the full sky except for galactic lattitudes @xmath33 .
figure [ fig : localvoids ] provides a remarkably detailed reconstruction of the cosmic web in the 2mrs volume .
it shows the ( surface ) density of the weblike structures in the local universe .
these are the result of adhesion simulations by @xcite and @xcite , based on the the constrained bayesian kigen reconstruction by @xcite of the initial conditions in the local volume traced by the 2mrs redshift survey . for a given gaussian primordial field , the adhesion formalism allows the accurate reconstruction of the rich pattern of weblike features that emerge in the same region as a result of gravitational evolution .
the adhesion formalism was applied to 25 constrained realizations of the 2mrs based primordial density field ( @xcite ) .
the mean of these realizations gives a reasonably accurate representation of the significant filamentary and wall - like features in the local universe .
most outstanding is the clear outline of the void population in the local universe .
the reconstruction also includes the velocity flow in the same cosmic region .
it reveals the prominent nature of the outflow from the underdense voids , clearly forming a key aspect of the dynamics of the megaparsec scale universe .
the local universe structure in figure [ fig : localvoids ] presents a telling image of a void dominated large scale universe .
many of the voids in the adhesion reconstruction can be identified with the void nomenclature proposed by fairall ( @xcite ) , who mainly identified these voids by eye from the 6dfgrs survey .
it is interesting to see that the socalled tully void appears to be a richly structured underdense region , containing at least the microscopium void , the local void and the `` trans tully void '' . in the same reconstruction
, we are studying the intricate filamentary network in and around the pisces - perseus supercluster . the image in figure[fig : perseus ]
provides a nice impresssion of the complex 3-dimensional structure and connectivity along the main ridge of the pisces - perseus supercluster .
it also shows how the main ridge connects to several neighbouring filaments , connecting near massive clusters along the ridge , and how these filaments surround a lower density planar structure .
interesting is to note the clustering and alignment of the small ( whitish ) filamentary tendrils in and around the main arteries of the cosmic web .
analysis of this weblike structures region is under progress and will be first reported in @xcite and @xcite .
we wish to acknowledge francisco kitaura and steffen he for the collaboration on the local universe reconstruction project , and for allowing the use of figure 7 and 8 in this report in advance of publication in @xcite .
2012 , _ mnras _ , 427 , 61 - 76 1986 , _ journal of soviet mathematics _
, 32:03 , 229 - 258 1982 , _ geophysical and astrophysical fluid dynamics _ , 20 , 111 - 130 a.p .
1998 , _ large - scale structures in the universe _
( wiley ) 2012 , _ apj _ , 754 , 126 1984 , _ radiophysics and quantum electronics _
, 27:4 , 303 - 313 2013 , _ mnras _ 435 , 2065 - 2076 2012 , _ proc . of the 28th socg _ , doi:10.1145/2261250.2261316 2014 , _ mnras _ , 437 , 3442 - 3472 2016 , _ in preparation _ , phd thesis , univ .
groningen 2016 , _ in preparation .
_ 2016 , _ in preparation .
_ 2012 , _ apjs _ , 199 , 26 2013 , _ mnras _ , 429 , l84-l88 1989 , _ rev .
_ , 61 , 185 - 220 2012 , _ phys . rev .
, 85 , 083005 1994 , _ a&a _ , 289 , 325 - 356 1970 , _
a&a _ , 5 , 84 - 89 | the zeldovich approximation ( za ) predicts the formation of a web of singularities . while these singularities may only exist in the most formal interpretation of the za , they provide a powerful tool for the analysis of initial conditions .
we present a novel method to find the skeleton of the resulting cosmic web based on singularities in the primordial deformation tensor and its higher order derivatives .
we show that the @xmath0-lines predict the formation of filaments in a two - dimensional model .
we continue with applications of the adhesion model to visualise structures in the local ( @xmath1 ) universe . |
an unusual type of two - dimensional conductor , the chiral metal , may be formed at the surface of a layered , three - dimensional system in the presence of a magnetic field which has a component perpendicular to the layers @xcite .
it occurs in a system of weakly - coupled layers at magnetic field - strengths for which an isolated layer would have an integer quantised hall conductance . under these circumstances ,
bulk states are localised and the extended , chiral edge states associated with each layer hybridise to form a two - dimensional conductor at the surface of the sample .
properties of this chiral metal may be probed by conductance measurements with current flow normal to the layers
. early experimental work on semiconductor multilayer samples @xcite established that the integer quantum hall effect of a single layer is indeed robust against weak interlayer tunneling , and suggested that , within a quantum hall plateau , interlayer conductivity in the bulk vanishes in the low - temperature limit .
subsequent theoretical discussions @xcite drew attention to the unusual nature of surface states in multilayer quantum hall systems , and some aspects of the theory of the chiral metal have since been explored in considerable detail @xcite .
recent experiments by druist _
et al _ have isolated the contribution of surface states to vertical transport @xcite and demonstrated that mesoscopic conductance fluctuations offer a way to study inelastic scattering in the chiral metal @xcite .
conduction by surface states has also been invoked in the interpretation of further experiments on semiconductor multilayer samples @xcite , and on the bulk quantum hall effect in both organic conductors @xcite and an inorganic quasi - two dimensional conductor @xcite . in this paper
we extend the theoretical treatment of mesoscopic chiral conductors in two respects .
first , we develop some aspects of the theory of conductance fluctuations which we hope will be useful for the interpretation of experiments .
second , we study electron - electron interactions in the dirty chiral metal .
the existing understanding of conductance fluctuations in the chiral metal is based on calculations @xcite for the variance of sample - to - sample conductance fluctuations in systems that are fully phase - coherent . from the results of those calculations
, estimates have been made @xcite for the variance of conductance fluctuations measured as a function of magnetic field in samples that are only partially phase - coherent , and also for the range in magnetic field of conductance correlations .
these estimates determine the relevant scales but do not yield numerical coefficients or functional forms .
moreover , they depend on a continuum treatment of the system of coupled edge states , and apply if the inelastic scattering rate , @xmath3 is smaller than the inter - edge tunneling rate , @xmath5 , while existing experiments appear to be in the opposite regime @xcite .
we supplement this past work by calculating in full the conductance correlation function @xmath2 , both for @xmath9 and for @xmath10 . separately
, we consider conductance as a function of fermi energy . for conventional disordered conductors , changes in fermi energy provide a way to sample the conductance distribution .
in contrast , we find that the conductance of a partially - coherent chiral metal sample , though dependent on the disorder realisation , is unchanged by small variations in fermi energy .
one consequence is that thermal smearing of the electron distribution function does not decrease the amplitude of conductance fluctuations in the chiral metal . turning to electron - electron interactions
, we establish the form of the dynamically screened interaction and examine some of its consequences .
most importantly , we calculate the inelastic scattering rate due to electron - electron interactions , obtaining @xmath6 for @xmath11 .
we also examine other interaction effects which are known to be of interest in conventional disordered conductors@xcite .
interactions in diffusive metals are responsible for a zero - bias anomaly in the tunneling density of states @xcite , but as noted previously @xcite , and as we confirm this is absent from the chiral metal .
similarly , the leading interaction contribution to the mean conductance has a singular temperature dependence in conventional disordered conductors @xcite , but we find that it cancels in the chiral metal .
the remaining sections of this paper are organised as follows .
we review briefly past work and collect our most important new results in sec .
we describe calculations on conductance fluctuations in sec .
iii , and treat electron - electron interactions in sec .
we summarise and discuss recent experiments in sec . v.
it is useful first to introduce some notation . consider a layered sample which exhibits the bulk quantum hall effect , with one landau level below the fermi energy .
a chiral metal is formed at the sample surface , from hybridisation of one edge state per layer .
we denote the interlayer spacing by @xmath12 , the chiral velocity by @xmath13 and the interlayer diffusion constant by @xmath14 .
the electron density of states per unit area is @xmath15 and the surface conductivity in the transverse direction , as given by the einstein relation , is @xmath16 , in units of @xmath17 , which we use throughout this paper .
phase coherent mesoscopic conductors are known to have three distinct mesoscopic regimes , according to relative sample dimensions @xcite . for definiteness ,
suppose that the sample is a cylinder of height @xmath18 and circumference @xmath19 , with its axis perpendicular to the layers .
take the number of layers to be @xmath20 , so that @xmath21 .
two intrinsic length scales arise , with which @xmath18 should be compared .
the first of these involves the mean conductance of the sample , which if ohm s law applies is @xmath22 : the quasi one - dimensional localisation length , @xmath23 , is the value of @xmath18 at which @xmath24 is of order the quantum unit of conductance . samples with
@xmath25 have states localised in the vertical direction and are quasi one - dimensional insulators .
samples with @xmath26 are metallic and have ohmic dependence of @xmath24 on sample dimensions .
a second length scale influences the amplitude of conductance fluctuations .
it is the distance @xmath27 that an electron diffuses in the transverse direction during the time taken to circumnavigate the sample : @xmath28 .
samples with @xmath29 are quasi one - dimensional metals without time - reversal symmetry and , in common with other examples of such systems , have a variance for the conductance of @xmath30
. samples with @xmath31 are two - dimensional chiral metals . if contacts are attached to these samples for conductance measurements , electrons will escape by diffusion to the contacts before completing a circuit of the sample in the chiral direction . as a result
, multiple scattering of an electron from a given impurity is completely supressed .
such samples present a number of essentially independent parallel strips for conduction in the transverse direction , each having a width in the chiral direction given by the distance an electron propagates before reaching a contact .
this width is @xmath32 .
the conductance of the two - dimensional chiral metal may be thought of @xcite as a sum of contributions from each such strip , and its variance has the value @xcite @xmath33 .
our concern in the following is with samples that are not completely phase coherent .
they have two different metallic regimes , analogous to those described above for phase coherent conductors .
we restrict our attention to one of these , the incoherent two - dimensional metal , in which @xmath34 so that dephasing occurs before circumnavigation : this is the simpler regime theoretically and seems the one relevant for existing experiments .
cho , balents and fisher @xcite have estimated the size of conductance fluctuations in the incoherent metal by viewing each phase - coherent region as a classical resistor and the whole sample as a random resistor network .
we summarise their argument here for convenience .
let the number of regions into which the sample is divided be @xmath35 in the chiral direction , and @xmath36 in the transverse direction .
let the conductance of a single region have mean value @xmath37 , with fluctuations of magnitude @xmath38 ; provided @xmath39 , the corresponding resistances are @xmath40 and @xmath41 .
a strip of @xmath36 such regions forms a single conduction path in the transverse direction , having mean resistance @xmath42 with fluctuations @xmath43 ; the corresponding conductances are @xmath44 and @xmath45 .
since the sample consists of @xmath35 such paths in parallel , its conductance has mean value @xmath46 and fluctuations @xmath47 .
the implications of this final result depend on the size of a patch in the transverse direction , compared to the layer spacing , @xmath12 .
if @xmath7 so that the phase coherence length in the transverse direction is much larger than the layer spacing , which is the case considered previously @xcite , then one expects a continuum treatment of the system to be adequate . in this case ,
@xmath48 and @xmath49 ; in addition , @xmath50 . combining these expressions and introducing the inelastic scattering length in the chiral direction , @xmath51 ,
one arrives at the conclusion ( eq.(4.1 ) of ref . ) that @xmath52 experimental studies use a magnetic field with a component normal to the sample surface to generate conductance fluctuations in the chiral metal @xcite ; the field scale of conductance correlations is set by the flux quantum , @xmath53 , and the area of a coherent region , being ( eq.(4.4 ) of ref . ) @xmath54 the results we obtain in sec.iii substitute for eqns.([fluc1 ] ) and ( [ fluc2 ] ) the correlation function @xmath55 where the function @xmath56 is @xmath57 with the scaling variable @xmath58 the opposite limit of weakly coupled edges , @xmath59 , has not previously been examined theoretically but seems to be important experimentally @xcite . adapting the argument outlined above , we take , as elements of a classical resistor network , regions again of length @xmath60 in the chiral direction but now of width @xmath12 in the transverse direction
. then @xmath61 and @xmath62 .
the mean conductance of one such patch is small in this limit .
we expect fluctuations to be of the same order , and take @xmath63 . in this way we obtain in place of eq.([fluc1 ] ) @xmath64 the field scale of conductance correlations is again set by the flux quantum and the area of a coherent region , but eq.[fluc2 ] is replaced in this limit by @xmath65 we calculate in sec.iii the conductance correlation function for weakly coupled edges , finding in agreement with these estimates @xmath66 where @xmath67 .
the inelastic scattering length , @xmath60 , appears as a phenomenological parameter in the expressions given above for the correlation function of conductance fluctuations .
the contribution to inelastic scattering from electron - electron interactions is known in non - chiral dirty metals to have a characteristic temperature dependence , reflecting the form of the dynamically screened coulomb interaction in a diffusive system @xcite , and it is interesting to study inelastic scattering microscopically for the chiral metal .
we do this in sec.iv , examining dynamical screening and using the results to calculate the inelastic scattering rate , @xmath3 .
we obtain in the regime @xmath7 @xmath68 where @xmath69 is a dimensionless coefficient .
a simple interpretation of this result can be given , following similar arguments established in the theory of diffusive metals @xcite .
as a starting point , note that ( in both chiral and non - chiral diffusive metals ) the dynamically screened interaction strength at long wavelengths is independent of the bare interaction strength . in consequence
, the inelastic scattering rate depends only on properties of the non - interacting system and temperature : @xmath70 is of order the single - particle energy level spacing in a system with dimensions determined by the distance an electron moves in the time @xmath71 .
these dimensions are @xmath72 and @xmath73 in the chiral and transverse directions respectively , and the estimate @xmath74 is consistent with eq.([t_in ] ) . an equivalent argument applied to non - chiral , diffusive conductors in @xmath75-dimensions yields @xmath76 , which is corrected in two dimensions to @xmath77 by detailed calculations @xcite .
the factor of @xmath78 in this last expression reflects divergent contributions to scattering in the diffusive two - dimensional metal , from processes with small energy transfer .
the same scattering processes are responsible for a difference , in diffusive two - dimensional metals , between the temperature dependence of the inelastic ( or out - scattering ) rate , which acts as a cut - off for conductance fluctuations , and the dephasing rate , which acts as a cut - off for the weak localisation effects that result in negative magnetoresistance . in the chiral metal , as for conventional , diffusive metals in more than two dimensions , small energy transfer processes are not dominant and there is a single relaxation rate .
inelastic scattering in weakly coupled edge states has been discussed in ref . , where it is noted that one expects @xmath79 from perturbation theory in the interaction stength , applied to a single edge .
we study conductance fluctuations in the chiral metal using a single - particle description of the system , supplemented by a relaxation rate to represent inelastic scattering .
we treat a sample with layers lying in the @xmath80 plane and consider a surface in the @xmath81 plane .
electrons on the surface then have a continuous coordinate , @xmath82 , in the chiral direction parallel to the edges of layers , and an integer coordinate , @xmath83 , labeling layers in the transverse ( @xmath84 ) direction , which we combine as @xmath85 .
we take the interlayer tunneling energy to be @xmath86 and represent a magnetic field component , @xmath87 , normal to the surface using the vector potential @xmath88 .
the hamiltonian @xmath89 acts on a wavefunction @xmath90 according to @xcite @xmath91 + v_{n}(x ) \psi_{n}(x)\ , , \label{schro } \\\end{aligned}\ ] ] where @xmath92 is a random potential arising from impurities and surface roughness .
we choose @xmath92 to be gaussian distributed with short - range correlations , so that @xmath93 and @xmath94 . denoting the green s function for eq.([schro ] ) by @xmath95
, we require its disorder average @xmath96 the diffusion propagator @xmath97 and its fourier tranform @xmath98 these quantities are known from calculations @xcite using the usual expansion in powers of the green s function for the disorder - free system . without disorder or a normal magnetic field component ,
eigenstates are plane waves and the electron dispersion relation is @xmath99 .
this dispersion along with the chiral motion leads to an open fermi surface on which all electrons have the same @xmath82-component of velocity . in the presence of a normal magnetic field component of dimensionless strength @xmath100 , the eigenfunctions @xmath101 in the pure system
are @xmath102 where @xmath103 is an integer and @xmath104 , the bessel function of order @xmath105 .
the associated energy eigenvalues are @xmath106 with @xmath107 .
the treatment of disorder is simple in some respects because chiral motion prevents multiple scattering of an electron from any particular impurity . as a result
, the average one - particle green s function is given exactly by the born approximation@xcite . in this way , one arrives at @xcite @xmath108 where we use the standard notation , @xmath109 and @xmath110 . from the value of the self - energy ,
the elastic scattering time is @xmath111 , and the elastic scattering length is @xmath112 . in the absence of a normal magnetic field component , @xmath113 is translationally invariant and has the fourier transform @xmath114^{-1}\,.\ ] ] for non - zero @xmath87 , @xmath113 acquires a phase under translations , which is central to calculations of conductance fluctuations with magnetic field : @xmath115 where @xmath116 , @xmath85 and @xmath117 .
the diffusion propagator is similarly given by a a sum of ladder diagrams , and has a form that reflects the combination of chiral motion in @xmath82 and diffusion in @xmath84 .
at small wavevectors it is @xmath118^{-1}\ , .
\label{diffuson}\ ] ] the diffusion constant has the value @xmath119 in the absence of a normal magnetic field component , and has the field dependence @xcite @xmath120 $ ] , where the field scale for magnetoresistance is @xmath121 . because of the simplifications in the treatment of disorder that follow from chiral motion , the results above hold both at weak disorder ( @xmath122 ) and at strong disorder ( @xmath123 ) . in a semiclassical description , electron motion follows a random walk in the transverse direction . at weak disorder ,
steps of the walk have speed @xmath124 and duration @xmath125 ; the rate for inter - edge tunneling is @xmath126 . at strong disorder , steps are of length @xmath12 and duration @xmath127 ; from the value of @xmath14 , we deduce that the rate for inter - edge tunneling is then @xmath128 .
the component of the current density operator in the transverse direction is @xmath129\,.\ ] ] the kubo formula for the transverse conductivity in a system of area @xmath130 is @xmath131\,.\ ] ] evaluating the disorder average of this expression , one finds @xmath132 as expected from the einstein relation .
conductance fluctuations have been studied previously @xcite for the two - dimensional chiral metal in the absence of inelastic scattering , in most detail by gruzberg , read and sachdev @xcite , who discussed the variance and its dependence on sample geometry . here
we present calculations which include inelastic scattering , obtaining the correlation function @xmath133 as a result of chiral motion , and provided that electrons do not circumnavigate the sample phase coherently ( @xmath134 or @xmath135 ) , the only one - loop diagram contributing to @xmath136 is the one shown in fig . [
fig : diagram ] .
it is most conveniently evaluated in real space , and represents the expression @xmath137 here , @xmath138 stands for the zero - frequency diffusion propagator , generalised to the situation of interest , in which the two green s functions entering it are evaluated for different normal magnetic field strengths , @xmath139 and in which inelastic scattering is included .
the other factors arise from a combination of single - particle green s functions and current operators , and are short range : @xmath140 , with @xmath141 our approach to evaluating @xmath136 is different in the two regimes , @xmath142 and @xmath143 , which we consider separately . in this regime
, the discreteness of the system in the transverse direction may be neglected .
it is sufficient to approximate the short - range terms in eq.([fluc ] ) by @xmath145 where @xmath146 the diffusion propagator , generalised to include the magnetic field difference @xmath4 and inelastic scattering , is most easily calculated within a continuum treatment of the transverse direction by solving , for @xmath147 , the differential equation @xmath148k_{\delta b}({\bf 0},{\bf r } ) \label{diff}\ ] ] with the initial condition @xmath149 at @xmath150 . here
, the inelastic scattering rate , @xmath3 , has been introduced , and the magnetic field difference enters through @xmath151 . to describe a sample connected to contacts , absorbing boundary conditions ( @xmath152 ) should be imposed at @xmath153 and @xmath154 .
however , provided the size of a phase - coherent region is much smaller than the sample size , so that @xmath155 , it is sufficient to use the solution of eq.([diff ] ) for a system of infinite extent in the @xmath84-direction when evaluating the right - hand side of eq.([fluc ] ) .
this solution is @xmath156 where @xmath157 combining eqns.([fluc]),([j1 ] ) , and ( [ k1 ] ) , we obtain @xmath158 ) the form of the scaling function @xmath56 is shown in fig .
[ fig : result ] .
it is normalised so that @xmath159 , and decreases as @xmath160 for @xmath161 , so that the conductance correlation function decays as @xmath162 for large @xmath163 .
if adjacent edges are sufficiently weakly coupled by tunneling , there is mesoscopic regime in which the inelastic scattering rate is small , in the sense that @xmath165 , but the interlayer tunneling rate is even smaller .
it is sufficient in this regime to calculate the conductance correlation function at leading order in the interlayer coupling , @xmath86 .
doing so , we retain @xmath86 in the current operators but use green s functions for a system of isolated edges . the disorder - averaged single - particle green s function in this limit
is @xmath166 [ x_2 - x_1])\,,\end{aligned}\ ] ] where @xmath167 is the step function .
the short - range terms associated with current operators in eq.([fluc ] ) may then be approximated as @xmath168 \label{j2}\,.\ ] ] the diffusion propagator reduces for uncoupled edges to a function which has a phase determined by @xmath4 and which decays in amplitude with distance only as a result of inelastic scattering .
we take it to be @xmath169[x_2 - x_1])\ , .
\label{k2}\end{aligned}\ ] ] combining eqns.([fluc]),([j2 ] ) , and ( [ k2 ] ) , we obtain @xmath170\end{aligned}\ ] ] and hence arrive at the correlation function for conductance fluctuations given in eq.([fluc-1d ] ) .
it is also of interest to consider the dependence of conductance fluctuations on fermi energy .
in particular , the range of conductance correlations in energy for a system at zero temperature will determine the extent to which the amplitude of fluctuations measured at finite temperature is reduced by thermal smearing of the electron distribution . we find that conductance fluctuations are perfectly correlated in energy .
to show this , we return to eq.([fluc ] ) but replace the magnetic field difference @xmath4 with an energy difference @xmath171 .
the diffusion propagator at finite energy difference simply acquires a phase : @xmath172 this phase cancels when @xmath173 and @xmath174 are combined , since the factors @xmath175 associated with the current operators are short - range , so that @xmath176 and @xmath177 in eq.([fluc ] ) . as a consequence , the correlation of conductance fluctuations is independent of @xmath171 .
the physical reason for this behaviour is that , if @xmath90 is an eigenfunction of eq.([schro ] ) for some realisation of disorder , with energy @xmath178 , then @xmath179 is also an eigenfunction , with energy @xmath180 , provided @xmath181 is an integer multiple of @xmath182 , so that periodic boundary conditions in the chiral direction are satisfied .
hence , apart from the phase factor , which does not affect the conductance , states are perfectly correlated in energy . as a result ,
the only consequence of a change in temperature is a change in the inelastic scattering rate .
in this section we study several aspects of electron - electron interactions in the chiral metal . as a first step ,
it is necessary to examine screening of the coulomb interaction between a pair of electrons , by other electrons in the fermi sea .
one can expect the frequency and wavevector dependence of the screened interaction to be significant . in conventional conductors ,
the consequence of disorder and the resulting diffusive motion of charge ( with diffusion constant @xmath14 ) is that screening of a potential having wavevector @xmath183 is suppressed @xcite at frequencies higher than @xmath184 . and
in a one - dimensional chiral metal at the edge of a two - dimensional integer quantum hall system , electron - electron interactions modify the dispersion relation for edge magneto - plasmons@xcite .
we find in the two - dimensional chiral metal that dynamical screening at finite wavevector interpolates , as a function of the direction of the wavector , between these two types of behaviour .
our results may be obtained either by using the random phase approximation within a diagrammatic calculation , or more transparently from a hydrodynamic approach , as follows .
consider the response of the chiral metal to a space- and time - dependent external potential .
let @xmath185 be the screening charge density induced in the presence of an electric field that has components @xmath186 within the surface .
we require an equation of motion for @xmath185 , which we derive in the usual way , by considering @xmath187 , the deviation from equilibrium of the current density , and using the continuity equation . treating the transverse direction as continuous , one has @xmath188 and @xmath189 so that the effect of the transverse component of the electric field is simply to generate a transverse current density .
the effect of the component of the electric field in the chiral direction is rather different : referring to the three - dimensional sample as a whole , @xmath190 generates hall currents within each layer , which transport charge between the bulk and the surface . the resulting change in surface charge density
may be represented by adding a term to the continuity equation , so that it becomes @xmath191 where @xmath192 is the quantised hall conductance of a layer , and ( as in preceding sections ) @xmath12 is the layer spacing .
it is helpful to express the conductivities in eqns.([jz ] ) and ( [ cont ] ) as @xmath193 and @xmath194 , and to recall that the density of states is @xmath195 .
then , combining eqns.([jx ] ) , ( [ jz ] ) and ( [ cont ] ) , the charge density evolves as @xmath196 .
\label{ev}\end{aligned}\ ] ] the electric field appearing in these equations arises from a scalar potential @xmath197 which is the sum of an externally imposed potential , @xmath198 , and the potential @xmath199 due to the screening charge , @xmath185 .
these last two quantities are related in the standard fashion @xcite by the poisson equation for a three - dimensional electrostatic problem , with coordinates @xmath200 and @xmath201 , in which the charge density is @xmath202 and @xmath199 is found by evaluating the three - dimensional potential at @xmath203 .
taking fourier transforms , defined according to @xmath204 and similarly for other quantities , one finds @xmath205 where @xmath206 is the fourier transform of the unscreened interaction potential betwen a pair of electrons .
expressing the electric field in terms of @xmath197 and solving the fourier transform of eq.([ev ] ) , we find @xmath207 where @xmath208 this result simplifies in several limits .
static screening ( @xmath209 ) is isotropic and as given by thomas - fermi theory , with an inverse screening length @xmath210 .
dynamical screening of a potential with fourier components only in the transverse direction ( @xmath211 ) is exactly as in a non - chiral disordered , two - dimensional metal .
response to a potential with fourier components only in the chiral direction ( @xmath212 ) is undamped , and excitations have the dispersion relation : @xmath213 . in the following , we shall treat interactions in the chiral metal using the matsubara formalism . within this approach
, we find for the polarisation operator@xmath214 with @xmath215 , which has , as its analytical continuation , eq.([pi ] ) .
similarly , the screened coulomb interaction potential is @xmath216\,.\ ] ] the enhanced interaction between slowly relaxing density fluctuations in a conventional , diffusive metal is responsible for a zero - bias anomaly in the tunneling density of states @xcite .
balents and fisher @xcite have argued that such a zero - bias anomaly should not be expected in the chiral metal .
physically , charge that tunnels into the system is swept away at the chiral velocity ; mathematically , the terms in perturbation theory associated with this anomaly yield only smooth contributions for the chiral metal . in the present subsection
we examine in more detail the reasons for this .
it is sufficient for the purpose to consider a static interaction , @xmath217 , with finite range , between electrons separated by a distance @xmath218 , and to calculate at first order in @xmath217 the disorder - averaged exchange energy @xcite .
let @xmath219 be the disorder - average of this quantity for an electron at energy @xmath178 interacting with a filled fermi sea @xcite .
the change in the density of states due to interactions is @xmath220 viewed in this way , the zero - bias anomaly in conventional , diffusive metals arises because states lying close in energy have wavefunctions which are highly correlated in space , enhancing interactions .
moverover , the wavefunction correlations depend strongly on the energy separation of the states , so that @xmath219 varies rapidly with @xmath178 if @xmath178 is close to the fermi energy , @xmath221 .
let @xmath222 and @xmath223 be single - particle eigenfunctions and eigenenergies in a disordered chiral metal .
the disorder - averaged exchange energy is @xmath224 where the two - particle spectral function , @xmath225 , is defined by @xmath226 this spectral function can be calculated from the real part of the diffusion propagator , eq.([diff - prop ] ) , or by considering the time - fourier tranform of the density associated with a spreading wavepacket .
it has the form @xmath227 combining eqns.([delta - n ] ) , ( [ sigma ] ) and ( [ a ] ) , one sees that , if the potential has range @xmath228 , then @xmath229 is independent of @xmath178 for @xmath230 . inspecting eq.([a ] ) , it is clear that eigenfunctions in the chiral metal do in fact have strong correlations : @xmath225 with @xmath231 diverges as @xmath232 for @xmath153 .
these correlations , however , depend only weakly on energy , and so do not produce a large change in the tunneling density of states .
the following subsection is devoted to a calculation of the contribution from electron - electron interactions to the inelastic scattering rate , @xmath3 , which acts as a cut - off for conductance fluctuations .
we compute @xmath3 by finding at leading order the interaction contribution to the irreducible vertex that appears in the ladder sum for the diffusion propagator used in calculations of conductance fluctuations @xcite . the two single - particle ( advanced and retarded ) green s functions from which this diffusion propagator is constructed are associated with distinct measurements of the conductance . in consequence
, one should not include in the irreducible vertex diagrams in which the interaction line connects the two single - particle propagators .
the relevant diagrams @xcite are shown in fig .
[ fig : inscrate2 ] , in which the full lines represent single - particle propagators , wavy lines are the screened coulomb interaction , single dashed lines are impurity averages , and double dashed lines are diffusion ladders .
we evaluate these diagrams treating the transverse direction as continuous , so that our results are restricted to the regime in which edges are strongly coupled : @xmath233 .
we set the energy and wavevector differences , @xmath234 and @xmath235 , at the vertex to zero , express sums on matsubara frequencies as contour integrals , and combine terms .
we expect that the dominant contribution to the resulting integral will be from small wavevectors and energies , and use the asymptotic form of the integrand appropriate for this regime , obtaining @xmath236 \times \\ \frac{x(\hbar d { q_{z}}^2)^2}{[(x+\hbar v q_{x})^2 + ( \hbar d { q_{z}}^2)^2 ] [ ( \hbar vq_{x})^2 + ( \hbar d { q_{z}}^2)^2 ] } \,.\end{aligned}\ ] ] the dependence of this expression on @xmath8 , @xmath14 and @xmath13 is made apparent by introducing the scaled variables : @xmath237 , @xmath238 and @xmath239 .
the integral can be evaluated numerically , yielding the expression for @xmath240 given in eq.([t_in ] ) .
the physical interpretation of this result has been discussed in sec.ii .
2 electron - electron interactions are known in diffusive conductors to give rise to temperature - dependent contributions to the conductivity that are singular in the low - temperature limit @xcite the origin of these interaction corrections to the conductivity is similar to that of the zero - bias anomaly in the tunneling density of states .
since the chiral metal has no zero - bias anomaly , one expects it to have no singular interaction contribution to the conductivity . in the following subsection
we outline calculations that confirm this at leading order .
the presentation follows closely that of altshuler , khmelnitskii , larkin , and lee @xcite for the diffusive metal . to compute the interaction correction to the conductivity of the chiral metal using the matsubara technique
, it is convenient to use the kubo formula in the form @xmath241 where @xmath242 is the analytic continuation in @xmath234 of @xmath243 ) j_z g(i\epsilon_m)]\rangle\,.\ ] ] the diagrams that contribute to @xmath244 at leading order in the screened interaction strength can be classified according to the number of diffusion propagators each one contains .
those with the most diffusion propagators are most singular in the low - temperature limit .
they can be separated into two categories .
diagrams in the first category are illustrated in fig .
[ fig : types ] , corresponding to figs.5(a ) , 5(b ) and 5(c ) of ref .. it has been shown for a diffusive conductor in ref . that these cancel , and we find the same to be true in the chiral metal .
diagrams in the second category are illustrated in fig.[fig : twodiag ] , corresponding to figs.5(d ) and 5(e ) of ref .. for the two - dimensional diffusive metal , these diagrams make a contribution to the conductivity which is logarithmic in temperature . for the chiral metal , their contribution is formally proportional to @xmath245 , but with a numerical coefficient which we find to be zero after integrating over the component in the chiral direction of the momentum transferred by the interaction . 2
our results , as set out in sec .
ii , are a contribution to the understanding of transport in the chiral metal . in many senses , it is a particularly simple example of a disordered conductor .
quantum interference effects are suppressed by chiral motion , and interaction effects are less singular than in diffusive two - dimensional conductors .
we have provided a rather detailed treatment of conductance fluctations , which we hope will be tested experimentally .
we have also identified a distinct mesoscopic regime , for weakly coupled edges , in which the inter - edge tunneling rate , @xmath5 , is smaller than the inelastic scattering rate , @xmath3 . in this regime
, transport is controlled by the properties of an isolated , one - dimensional chiral metal .
recent experiments @xcite appear , in fact , to probe behaviour of weakly coupled edges , though their analysis has necessarily so far been based on theory for the opposite regime . according to that analysis , the inelastic scattering length in the transverse direction , estimated from the variance of conductance fluctuations using eq.([fluc1 ] ) ,
is shorter than the inter - layer spacing , even at the lowest temperatures investigated .
moreover , this estimate of the inelastic scattering length is inconsistent with the value deduced from the correlation field for condcutance fluctuations , using eq.([fluc2 ] ) , as one might expect if the theory used is , in fact , inappropriate . it will be interesting to find whether these discrepancies are resolved by using the theory we have presented to analyse the experiments .
if so , such measurements provide a way to investigate interaction effects in a one - dimensional system of chiral fermions .
we are particularly grateful for discussions and correspondence with d. p. druist and e. g. gwinn .
we also thank l. balents , f. h. l. essler , i. a. gruzberg , r. a. smith , and a. m. tsvelik for useful discussions .
j.b . acknowledges partial financial support from the european union through the marie curie grant ; the work was also supported in part by epsrc under grant no gr / j8327 .
99 h. l. strmer , j. p. eisenstein , a. c. gossard , w. wiegmann , and k. baldwin , phys .
lett . * 56 * , 85 ( 1985 ) ; h. l. strmer @xmath246 @xmath247 _ proceedings of the 18th international conference on the physics of semiconductors , stockholm , 1986 _ , edited by o. engstrom ( world scientific , singapore , 1987 ) , p. 385 .
s. hill , p. s. sandu , j. s. qualls , j. s. brooks , m. tokumoto , n. kinoshita , t. kinoshita , and y. tanaka , phys .
b * 55 * , r4891 ( 1997 ) ; m. m. honold , n. harrison , j. singleton , h. yaguchi , c. mielke , d. rickel , i. dekkers , p. h. p. reinders , f. herlach , m. kurmoo , and p. day , j. phys . cond
* 9 * , l533 ( 1997 ) ; s. uji , c. terakura , m. takashita , t. terashima , h. aoki , j. s. brooks , s. tanaka , s. maki , j. yamada , and s. nakatsuji , phys .
rev . * 60 * , 1650 , ( 1999 ) . for reviews , see : b. l. altshuler and a. g. aronov , in _ electron - electron interactions in disordered systems _ , edited by a. l. efros and m. pollak ( north- holland , amsterdam , 1985 ) , p. 1 ; and i. l. aleiner , b. l. altshuler , and m. e. gershenson , waves in random media , * 9 * , 201 ( 1999 ) , and references therein .
b. l. altshuler , d. khmelnitzkii , a. i. larkin , and p. a. lee phys .
b * 22 * , 5142 ( 1980 ) . b. l. altshuler and a. g. aronov , zh .
pisma red .
* 30 * 514 ( 1979 ) [ jetp lett * 30 * , 482 ( 1979 ) ] . | we study the two - dimensional chiral metal , which is formed at the surface of a layered three - dimensional system exhibiting the integer quantum hall effect by hybridization of the edge states associated with each layer of the sample .
we investigate mesoscopic fluctuations , dynamical screening and inelastic scattering in the chiral metal , focussing particularly on fluctuations of conductance , @xmath0 , with magnetic field , @xmath1 .
the correlation function @xmath2 provides information on the inelastic scattering rate , @xmath3 , through both the variance of fluctuations and the range of correlations in @xmath4 .
we calculate this correlation function for samples which are not fully phase coherent .
two regimes of behaviour exist , according to whether @xmath3 is smaller or larger than @xmath5 , the rate for inter - edge tunneling , and we give results in both regimes .
we also investigate dynamical screening of coulomb interactions in the chiral metal and calculate the contribution to @xmath3 from electron - electron scattering , finding @xmath6 for @xmath7 at temperature @xmath8 . 2 |
classical novae are the result of thermonuclear explosions in interacting binary stars .
the system consists of a hot white dwarf primary with a main sequence secondary companion .
the secondary is thought to fill its roche lobe and transfers hydrogen - rich material via an accretion disc onto the surface of the white dwarf .
when the pressure in the degenerate material at the base of the accreted layer reaches a critical value a thermonuclear runaway takes place ( starrfield 1989 ) , ejecting @xmath110@xmath210@xmath3 m@xmath4 of material into the interstellar medium at velocities of up to several thousand kilometres per second .
this causes an increase in the bolometric luminosity of the system of a few orders of magnitude over a timescale of a few days .
radio emission from the nova starts to develop at a later time than the optical .
the white dwarf star photoionises the ejecta giving rise to free - free emission at temperatures of approximately 10@xmath5 k. the evolution of the ejecta is generally rapid , allowing radio interferometric observations to track the expanding shell as it brightens and then fades over favourable timescales ( e.g. taylor et al , 1987 ; eyres et al , 1996 , 2000 , 2005 ; heywood et al , 2002 , 2005 ) .
both radio and optical images of classical novae often show non - homogeneous , albeit spherical or ellipsoidal structures .
a feature common to all merlin observations of novae is apparent rotation between epochs . in this paper
we investigate the structure seen in merlin observations of v723 cas ( nova cas 1995 ) using models fitted to the radio light curve ( heywood et al 2005 - hereafter paper i ) .
the simulations presented in this paper provide an insight into how the radio images that result from interferometer observations relate to the true distribution of the emitting material .
radio images are often cited during discussions of the morphology of nova shells .
although it is common knowledge that interferometer images are an approximation of the true sky brightness distribution , it has not previously been determined just how many of the features which are common to radio observations of novae are due to the interferometer and the aperture synthesis process .
the goal for this investigation is to generate synthetic emission maps based on shell models to be subsequently convolved with an accurate @xmath0 coverage model of the merlin interferometer .
descriptions of the shell models used are presented in section [ sec : simulations ] .
a radiative transfer solution is implemented for both a spherically symmetric and an ellipsoidal shell based on the models described in paper i. this process results in simulated emission maps which are then convolved with a @xmath0 model . in all cases this process , described in section [ sec : merlin ] , is consistent with the genuine merlin observations of v723 cas presented in paper i and reproduced in figure [ fig : merlin ] .
this facilitates direct comparisons between the real and simulated radio images .
( 10,21 ) ( -4.0,2.2 )
model images of the sky brightness distribution are generated by solving the line of sight equations of radiative transfer through a model emitting region .
descriptions of the radiative transfer code and the model shell geometries are presented in this section , together with images showing the predicted brightness distributions .
synthetic interferometer maps are derived by masking the fourier transform of these images with the @xmath0 coverage of the interferometer and then inverting the transform , this process is described in section [ sec : merlin ] .
the radiative transfer code uses cylindrical polar coordinates to define the emitting region as used by , for example , dougherty et al ( 2003 ) .
the coordinate system is shown in figure [ fig : radcodefig ] .
the shell model is constructed by creating a density function in the plane defined by @xmath6 and @xmath7 .
a 3d shell is created by mirroring the distribution about the plane defined by @xmath8 and @xmath9 and rotating through 360 degrees about the @xmath10 axis .
thus the quarter circles indicated by the solid lines in figure [ fig : radcodefig ] would define full spherical regions .
the configuration shown in figure [ fig : radcodefig ] is representative of the hubble flow model described in paper i ( and previously by seaquist & palimaka , 1977 , and by hjellming et al , 1979 ) and is discussed further below .
the density and temperature are defined in an array of dimensions 256 @xmath11 512 in @xmath8 and @xmath10 respectively and when the coordinate transform is applied to define the 3d emitting region the result is mapped on to a 256 @xmath11 256 pixel image .
the arrays defining the image are chosen to be 2@xmath12 in size , where @xmath13 is an integer , so that the fast fourier transform ( fft ) technique can be applied when convolving these images in the @xmath0 plane .
the frequency of the radiation is specified and the radiative transfer equations are solved along gridded lines of sight through the shell .
the code then calls a hierarchical data system ( hds ) library which outputs images of intensity , optical depth and brightness temperature in the starlink n - dimensional data format ( ndf ) .
an inclination angle for the orientation of the model can also be specified .
this is denoted by @xmath14 and is the angle between the @xmath15 axis in the plane of the sky and the @xmath10 axis of the coordinate system .
the hubble flow model applied to v723 cas in paper i is used as the basis for the first run of the axisymmetric radiative transfer code .
the values of mass and inner and outer ejection velocity returned by the best fit routine are used to define the density and the shell geometry for this model .
the inner and outer shell boundaries , and consequently the shell density , are time dependent and emission maps are created with epochs corresponding to the 6 cm merlin observations presented in paper i. the density function of the hubble flow model is given by @xmath16 which is a function of @xmath17 , the radial distance from the centre of the shell , and also of @xmath18 and @xmath19 , the radii of the inner and outer shell boundaries respectively .
hence when the shell is defined in the radiative transfer code the density array is filled according to @xmath20 for @xmath21 .
this is a function of @xmath8 and @xmath10 and the parameters @xmath19 and @xmath18 which are fixed for a given epoch .
similarly for @xmath21 the temperature is fixed at 17000 k ( see paper i ) . outside the shell boundaries the density
is assumed to be zero and the temperature is set at 100 k. the inclination angle is a redundant parameter in the spherical case .
the nine model intensity maps created by this process are presented in figure [ fig : sph_no_uv ] where it can be seen how initially the emission is entirely optically thick . as the material expands the outer layers start to become optically thin and eventually the inner boundary of the shell is revealed as a bright ring .
( 10,21 ) ( -3.9,2.8 ) some observations of classical novae show prolate ellipsoidal shells ( e.g. fh ser , gill & obrien , 2000 ) and the intention here is to determine whether it is possible to distinguish between a spherical shell or an ellipsoidal shell by studying the radio images .
the shell model is based on the spherical hubble flow model as described in section [ synhubflow ] however some additional parameters are introduced . a polar and equatorial direction
are defined which , with reference to figure [ fig : radcodefig ] , are along the @xmath10 and @xmath8 axes respectively .
the polar velocity , @xmath22 , is greater than the equatorial velocity , @xmath23 , and the two are related by the factor @xmath24 such that @xmath23 = @xmath24 @xmath22 .
the velocity in these regions is a function of the angle @xmath25 which is defined as @xmath26 such that @xmath27 = @xmath23 at @xmath25 = 0 and @xmath27 = @xmath22 at @xmath25 = @xmath28/2 , see figure [ ellshell ] .
the velocity as a function of @xmath25 is governed by the ellipse equations . in the general case , for an ellipse of semi major axis @xmath29 , semi minor axis @xmath30 and eccentricity @xmath31 , the radius @xmath17 as a function of @xmath14 , the angle between the semi
minor and semi major axes , is given by @xmath32 where @xmath33 by subtituting equation [ eccentricity ] into equation [ ellipseradius ] it can be shown that @xmath34 and the above formula is used to govern how the velocity varies as a function of @xmath25 in the shell , i.e. @xmath35 the elliptical shell models are again constructed based on the best fit parameters of v723 cas as returned by the light curve fitting in paper i. the best fitting inner and outer shell velocities are taken to be the velocities of the inner and outer shell boundaries at the pole .
the shell mass is taken to be the best fitting shell mass from the spherical model .
the value of @xmath24 is arbitrarily chosen to be 0.6 and the inclination of the shell to the plane of the sky is taken to be zero degrees i.e. the shell is viewed ` side - on ' .
as with the spherically symmetric model , nine simulated maps are calculated to correspond to the nine epochs of 6 cm merlin observations .
equation [ ellvel ] is used to determine @xmath36 and @xmath37 , the velocities of the inner and outer shell boundaries as a function of @xmath25 , and these velocities are then used to calculate @xmath38 and @xmath39 , the radii of the inner and outer shell boundaries for the specific epoch .
the density function @xmath20 is then applied to populate the density array in @xmath8 and @xmath10 .
once again the temperature is set at 17000 k within the shell boundaries and is assumed to be 100 k elsewhere .
the nine model intensity distribution maps are presented in figure [ fig : ell_no_uv ] . as with the spherical case
the emission is initially optically thick before expansion causes the optical depth to decrease and the inner shell boundary is revealed .
an interesting feature to note is the ` waist ' of brighter emission in the equatorial region .
this is due to the inverse dependence on the quantity @xmath40 in the density equation .
this is smaller for small values of @xmath25 than it is for values of @xmath25 close to @xmath41 radians giving rise to a density enhancement near the equator of the shell .
( 10,21 ) ( -3.9,2.79 )
once the model images have been calculated the image files have to be correctly formatted before they can be imported into the national radio astronomy observatory s astronomical image processing system ( aips ; http://www.aoc.nrao.edu/aips/ ) which is used to generate the synthetic merlin images .
a world coordinate system , frequency and stokes parameter axes are added .
the aips task uvcon[multiblock footnote omitted ] reads the properties of the array from an antenna file and determines the @xmath0 coverage of the interferometer .
the visibilities are then calculated from a gridded fft of a model image .
since the models were defined by fits to the light curve , the flux density is very close to that measured in the actual observations .
gaussian noise is added to the data , the level of which is calculated by uvcon based on integration times and the total length of the track .
the @xmath0 data produced by uvcon are deconvolved , cleaned and imaged using imagr . in all cases where possible
, the imaging process has been consistent with the methods used to produce the 6 cm merlin maps shown in figure [ fig : merlin ] .
the cleaning process was terminated at the first negative clean component and a circular restoring beam of 50 mas was used .
figure [ fig : sphnine ] shows the simulated merlin images based on the spherically symmetric hubble flow model .
these images are based on a 12 hour track which is of similar length to the original merlin observations .
the images in figure [ fig:24nine ] take the same model intensity distributions as those used to generate figure [ fig : sphnine ] except the simulated @xmath0 coverage is based on a 24 hour track instead of a 12 hour track .
figure [ fig : ellnine ] shows the model merlin observations of the elliptical shell model based on a 12 hour track , and finally figure [ fig:24ell ] shows the same model with a 24 hour track . in each 3 @xmath11 3 grid ,
left to right , top to bottom , the images correspond to the december 1996 , january 1997 , march 1998 , december 1998 , february 2000 , april 2000 , january 2001 , june 2001 and october 2001 merlin observations of v723 cas .
contour levels are ( -3 , 3 , 3@xmath42 , 6 , 6@xmath42 , 12 , 12@xmath42 ... ) @xmath11 0.15 mjy for each image , as quoted in the captions .
comparison of figure [ fig : sphnine ] to the merlin images presented in figure [ fig : merlin ] reveals that even for this simple spherical shell model the similarities between the simulated radio observations and the real merlin observations are striking .
initially the source is point - like , not much bigger than the restoring beam .
the source expands and as the turn over into optically thin emission begins apparent structure in the shell becomes evident . the source becomes ring - like then divides into the double - peaked structure which is common to all multi - epoch merlin observations of classical novae .
( 10,21 ) ( -4.4,-1.5 ) ( 10,21 ) ( -4.2,-1.93 ) ( 10,21 ) ( -3.2,3.2 ) ( 10,21 ) ( -3.2,2.7 ) the most intriguing feature of these simulated maps is the switching of the alignment of the dominant peaks in the map , giving the impression of rotation . in paper
i we speculated whether this was due to optical depth effects , with higher density regions of the shell remaining optically thick for longer and appearing as peaks of radio emission as the photosphere recedes . here
this explanation can be ruled out as the mass distribution supplied by the model is spherical with a smooth radial density profile . comparing figure [ fig:24nine ] to figure [ fig : sphnine ]
, shows that for a 24 hour observation the radio emission in the synthetic maps is more circularly symmetric and consquently more faithful to the model images .
the effect whereby the alignment of the dominant peaks in emission appears to rotate between epochs is also almost completely eradicated .
the increase in @xmath0 coverage provided by the 24 hour observation must be the reason for this , providing both an improved signal to noise ratio and a drastically altered sidelobe structure in the beam .
the beam shape and @xmath0 coverage are discussed in section [ sec : beamsanduv ] .
table [ tab : rms ] lists the root mean square values of the background noise for both 12 and 24 hour tracks of the simulated spherical hubble flow model .
.comparison of the background rms values in mjy / beam for both the 12 and 24 hour simulations of the spherical hubble flow model.[tab : rms ] [ cols="<,<,<",options="header " , ] the simulated radio maps for the ellipsoidal shell model reveal a north south bias with a double peaked source being present for the first five epochs .
this gives way to a ` partial shell ' structure which in turn is replaced by the east west aligned peaks in the final map .
although a stronger north south alignment of the emission would be expected from this ellipsoidal shell model the evolution of the radio emission over time displays many similiarities to the spherical case .
both initially develop a north south bias which gives way to an east west alignment and although this process occurs later for the ellipsoidal shell it is unlikely that a 12 hour track can be used to distinguish between an ellipsoidal or a spherical shell for objects of this nature . in order to determine whether a 24 hour track would produce a distinction between ellipsoidal and spherical shells the simulations presented in figure [ fig:24ell ] can be examined . for the optically thick stage
the merlin images accurately represent the simulated brightness distributions but for subsequent epochs the images detect only the brighter east - west aligned emission around the inner shell boundary .
the faint , extended structure along the major axis of the shell is below detectability . with a greater signal to noise ratio , or alternatively an object brighter than v723 cas ,
it is likely that observations based on a 24 hour track could be used to examine the eccentricity of an ellipsoidal shell .
figure [ fig:12_24_uv ] shows the improved @xmath0 coverage of the 24 hour track compared to the 12 hour track .
plots of the beam shape for both 12 and 24 hour observations are presented in figure [ fig:12_24_beam ] .
immediately apparent is the effect that the increased @xmath0 coverage has on the sidelobes of the beam .
the sidelobe structure in the 24 hour observation completely surrounds the central maximum and only deviates from circularity due to the declination of the source .
the beam of the 12 hour simulation has @xmath43-shaped sidelobe structure . as the source expands through the beam different parts of the shell
will be buried in the sidelobes .
the non - circular beam shape associated with the 12 hour track could be a possible cause for the apparent rotation of the nova shell between different epochs .
since this effect is generally not an issue with observations involving strong sources on larger scales ( e.g. radio galaxies ) it is likely that these effects will only be present when observing objects with low surface brightness objects and small angular sizes , and would probably pass unnoticed when observing systems with less favourable dynamic timescales .
additionally a 24 hour track provides higher fidelity images due to more thorough sampling of the @xmath0 plane and an improved signal to noise ratio .
the fourier transform process can give rise to unexpected artefacts in the final image .
in this section we briefly investigate the possibility that the transition from a disc shaped structure to something that is ring - like could be responsible for unusual effects in the final radio images .
examination of figure [ fig : sph_no_uv ] shows how the model emitting region develops with time .
initially the source is optically thick and resembles a disc on the sky , with the flux density at this stage governed only by the angular size and the brightness temperature .
the highest contour level also drifts around the field as the shell expands .
later a transition to optically thin emission occurs and the ring - like inner shell boundary is revealed . in order to assess the outcome of this transition on the final radio images , simulated merlin observations of an expanding ring are presented for comparison with those of an expanding disc .
these are presented in figures [ fig : rings ] and [ fig : discs ] .
the model emitting region in each case has angular size scales and flux densities comparable with that of the shell of v723 cas .
as the intention here is to solely investigate the effect of the deconvolution process the final simulated data do not have gaussian noise added .
note the differing contour levels for the ring and disc maps .
the disc simulation is plotted with levels of ( -3 , 3 , 3@xmath42 , 6 , 6@xmath42 , 12 , 12@xmath42 ... ) @xmath11 0.15 mjy as with the real merlin maps and the simulations presented in this paper .
since the simulated ring structure is much fainter the contour levels are reduced to ( -3 , 3 , 3@xmath42 , 6 , 6@xmath42 , 12 , 12@xmath42 ... ) @xmath11 0.03 mjy .
the disc simulation is brighter simply because there is more emitting material .
examination of figures [ fig : rings ] and [ fig : discs ] reveals that even these simple zero noise models exhibit features such as apparent rotation and peaks dividing as the emitting region expands .
these effects are seen in both the ring and the disc model meaning that it is unlikely that they appear in the real observations due to the nova undergoing the transition from disc - like to ring - like .
this is further evidence that these effects are due to the limited @xmath0 coverage and are not related to the changing optical depth or other underlying properties of the source .
these results have shown that extreme care must be taken when using radio images alone to interpret the morphology of classical nova shells such as that of v723 cas .
the ( comparatively ) low radio surface brightness of a nova shell coupled with its low angular size gives makes it prone to instrumental effects which may be mistaken for unusual shell morphology and motion ( e.g. clumping , apparent rotation ) .
however these effects can be mitigated by using observations with longer tracks which both improves the @xmath0 coverage and increases the signal to noise ratio .
when using radio imaging to track structure in novae and similar objects , coupling of simulations such as those presented here with observations would provide a more thorough insight into the underlying source structure .
however , optical imaging of novae reveal that the emission is far from homogeneous and exhibits equatorial and polar enhancements ( e.g. fh ser , gill & obrien , 2000 ) , clumping of material and even filamentary structures ( e.g. gk per , slavin et al , 1995 ) .
thus there is need for spherical models to be refined and it will be interesting to see how novae are observed by the next generation of interferometers such as @xmath31-merlin . with the vastly improved @xmath0 plane sampling of @xmath31-merlin exploiting multifrequency synthesis , figure [ fig : emerlin_uv ] , one would not expect the instrumental effects documented in this paper to be present .
preliminary simulations shown in figure [ fig : emerlin_sims ] show three epochs of the elliptical shell model described in section [ sec : ellipticals ] .
immediately apparent is the north - south aligned , extended , low brightness emission which was below detectability levels in the simulations presented in section [ sec : merlin ] .
the root mean square background noise levels in the @xmath31-merlin simulations are @xmath15 @xmath44jy .
the @xmath0 plane sampling is consistent with that of @xmath31-merlin observing using a 5 ghz contiguous band , i.e. 1000 frequency channels between 2 ghz and 5 ghz .
the drawback here is that these simulations treat the source as being spectrally flat , which is obviously a poor assumption over such a wide frequency range .
nonetheless they show that the @xmath31-merlin view of sources of this nature will be much more faithful to the true source distribution due to its vastly improved @xmath0 coverage and sensitivity .
ih wishes to thank the particle physics and astronomy research council ( pparc ) .
merlin is operated as a uk national facility by the university of manchester , jodrell bank observatory , on behalf of pparc .
ih also thanks simon garrington for his valuable input .
this research has made use of nasa s astrophysics data system .
the authors wish to thank the reviewer , amy mioduszewski , for her constructive criticisms . | we compare the predictions of simple models for the radio emission from classical novae with the merlin radio observations of nova v723 cas ( nova cas 1995 ) .
spherically symmetric and ellipsoidal radiative transfer models are implemented in order to generate synthetic emission maps .
these are then convolved with an accurate representation of the @xmath0 coverage of merlin .
the parameters and geometry of the shell model are based on those returned by fitting models to the observed light curve .
this allows direct comparison of the model images with the nine 6 cm merlin images of v723 cas .
it is found that the seemingly complex structure ( clumping , apparent rotation ) evident in the observations can actually be reproduced with a simple spherical emission model .
the simulations show that a 24 hour track greatly reduces the instrumental effects and the synthetic radio map is a closer representation of the true ( model ) sky brightness distribution .
it is clear that interferometric arrays with sparse @xmath0 coverage ( e.g. merlin , vlba ) will be more prone to these instrumental effects especially when imaging ring - like objects with time dependent structure variations .
a modelling approach such as that adopted here is essential when interpreting observations .
= = = = = = = = # 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 latexl-.36em.3ex-.15em t-.1667em.7ex-.125emx [ section ] [ firstpage ] stars : novae , cataclysmic variables stars : novae : individual ( v723 cas ) radio continuum : stars techniques : interferometric |
a @xmath0 surface defined over the complex number field whose picard number equals to maximum possible number @xmath1 is called a _ singular @xmath0 surface_. shioda and
inose @xcite showed that the map a singular @xmath0 surface @xmath2 corresponds to its transcendental lattice @xmath3 is a bijective correspondence from the set of singular @xmath0 surfaces onto the set of equivalence classes of positive - definite even integral lattice of rank two with respect to @xmath4 ) .
the discriminant of a singular @xmath0 surface @xmath2 is the determinant of the gram matrix of the transcendental lattice @xmath3 . in this paper
we study jacobian fibrations , i.e. , elliptic fibrations with a section , on the singular @xmath0 surface @xmath5 of discriminant @xmath6 , which corresponds to the lattice defined by @xmath7 and is uniquely determined up to isomorphism .
jacobian fibrations on @xmath5 were classified by nishiyama @xcite .
he classified all configurations of singular fibers of jacobian fibrations on @xmath5 into @xmath8 classes and determined their mordell - weil groups .
then , we give for each fibration a weierstrass model .
more precisely , we state our main theorem .
let @xmath5 be the singular @xmath0 surface of discriminant @xmath6 . for each jacobian fibration in nishiyama s list (
* table 1.1 ) , an elliptic parameter @xmath9 , a weierstrass equation and the generators of the mordell - weil group are given by table [ classes ] .
.classification of jacobian fibrations on @xmath5 [ cols="^,^,^,^,^",options="header " , ] we explain about table [ classes ] .
the fist column shows the name of each jacobian fibrations following nishiyama s notaion .
the second column shows the configuration of singular fibers . here , for example , by @xmath10 meanes that the surface has two singular fibers of type @xmath11 and a singular fiber of ot type @xmath12 ( kodaira s notation @xcite ) .
the third column shows the mordell - weil group ( mwg ) of the fibration .
the fourth column shows an elliptic parameter @xmath9 of the fibration under the singular affine modell ( [ model ] ) of @xmath5 . the index @xmath13 is the name of the fibration .
the last column shows a weierstrass equation and rational points corresponding to mordell - weil generator of the fibration , where @xmath14 is the rational point corresponding to the zero of mwg .
recently , braun , kimura and watari @xcite showed that nishiyama s list also gives the classification of jacobian fibrations on @xmath5 modulo isomorphism .
thus , our and their results answer completely a question of kuwata and shioda @xcite .
the singular @xmath0 surface @xmath5 is known as _ a generalized kummer surface _ constructed by the following .
let @xmath15 be the complex elliptic curve with the fundamental periods @xmath16 and @xmath17 .
let @xmath18 be an automorphism of @xmath19 defined by @xmath20 .
then the minimal resolution of the quotient @xmath21 is isomorphic to the singular @xmath0 surface @xmath5 ( see ( * ? ? ?
* lemma 5.1 ) ) .
the automorphism @xmath18 has the @xmath22 fixed points @xmath23 , where @xmath24 are the fixed points of the automorphism @xmath25 of @xmath15 defined by @xmath26 .
these @xmath22 points @xmath27 correspond to the singular points @xmath28 of the quotient @xmath29 .
the minimal resolution @xmath5 of @xmath30 is obtained by replacing each @xmath28 by @xmath31 non - singular rational curves @xmath32 and @xmath33 with @xmath34 .
moreover , @xmath5 contains @xmath8 non - singular rational curves , i.e. the image @xmath35 ( or @xmath36 ) of @xmath37 ( or @xmath38 ) in @xmath5 .
we have the following intersection numbers .
@xmath39 these @xmath40 curves on @xmath5 form the configuration of figure [ config ] .
-curves , width=377 ] it is well known that the elliptic curve @xmath15 has the following weierstrass form @xmath41 we denote each factor of @xmath19 by @xmath42 then the automorphism @xmath18 is written by @xmath43 the function field @xmath44 is equal to the invariant subfield of the function field @xmath45 under the automorphism @xmath18
. then we have @xmath46 where @xmath47 and @xmath48 are naturally regarded as functions on @xmath5 with the relation @xmath49 this gives a singular affine model of @xmath5 .
we start from the equation to obtain a weierstrass form for each jacobian fibration on @xmath5 . under the above notation
, we see that the divisor of typical functions are as follows .
an elliptic parameter for fibration 1 is given by @xmath51 the divisor of @xmath52 is given by @xmath53 the zero divisor @xmath54 ( the bold lines in figure [ no1 ] ) and the polar divisor @xmath55 ( the thin lines in figure [ no1 ] ) are the singular fibers both of type @xmath11 .
eliminating the variable @xmath56 from ( [ model ] ) and ( [ par1 ] ) , we obtain the following equation @xmath57 which defines a plane curve over @xmath58 with a singularity at @xmath59 . blowing up by @xmath60
, we have the following equation @xmath61 which defines a nonsingular plane cubic curve over @xmath58 with a rational point @xmath62
. then we can convert it into a weierstrass form ( see @xcite or @xcite ) . since the rational point @xmath62 corresponds to the divisor @xmath63 ( the dotted line in figure [ no1 ] ) , choosing it as the zero section of the group structure
, we obtain the weierstrass equation for fibration 1 @xmath64 where the change of variables is given by @xmath65{4}(u_1 - 1 ) u_1 t}{(y_1 ^ 2 - 1 ) } , \quad y = -\frac{u_1 ^ 2 ( u_1 - 1 ) ( u_1 y_1 -u_1 + 2)}{y_1 + 1}.\ ] ] besides the two singular fibers of type @xmath11 at @xmath66 and @xmath67 , there is one singular fiber of type @xmath12 at @xmath68 .
it is the divisor @xmath69 ( the long dashed dotted lines in figure [ no1 ] ) , where @xmath70 is a @xmath71-curve on @xmath5 arising from a curve on @xmath72 below .
let @xmath73 be the projection given by @xmath74 then the map @xmath75 factors through @xmath76 .
let @xmath77 be the morphism of degree three from @xmath5 to @xmath72 that makes the following diagram commutative : @xmath78 \ar[dr]^{\pi } & \\
c_{\omega}^1 \times c_{\omega}^2 \ar[r ] & c_{\omega}^1 \times c_{\omega}^2 / \sigma \ar[r]_{\bar{\pi } } & \mathbb{p}^1 \times \mathbb{p}^1 } \ ] ] it is easy to verify that the equation @xmath68 means @xmath79 from ( [ par1 ] ) .
this equation defines a curve on @xmath80 .
then it lifts to the @xmath71-curve @xmath70 on @xmath5 via the map @xmath77 .
an elliptic parameter for fibration 3 is given by @xmath81 the divisor of @xmath82 is given by @xmath83 which is indicated in figure [ no3 ] .
the zero divisor @xmath84 is the singular fiber of type @xmath85 ( the bold lines ) and the polar divisor @xmath86 is the singular fiber of type @xmath87 ( the thin lines )
. the curves @xmath88 and @xmath89 ( the dotted lines ) are all the sections .
eliminating the variable @xmath48 from ( [ model ] ) and ( [ par3 ] ) , we have the following equation @xmath90 which has a rational point @xmath91 corresponding to the curve @xmath92 .
thus , choosing @xmath92 as the zero section of the group structure , we obtain the weierstrass equation for fibration 3 @xmath93 where the change of variables is given by @xmath94 besides the above two singular fibers of types @xmath85 and @xmath87 , the fibration has three @xmath95 fibers at @xmath96 and @xmath97 .
the 2-torsion rational point @xmath98 corresponds to the curve @xmath89 .
the rational point @xmath99 corresponds to the curve @xmath100 of height @xmath101 , which is a generator of the mordell - weil lattice of the fibration .
the curve @xmath102 is another free section corresponding to the rational point @xmath103 with the relation @xmath104 in the mordell - weil group .
an elliptic parameter for fibration 5 is given by @xmath105 it is clear that this elliptic parameter defines a fibration having three singular fibers all of types @xmath106 at @xmath107 and @xmath67 ( the bold lines in figure [ no5 ] ) from ( [ func ] ) .
furthermore the fibration is induced by the composition of the first projection @xmath108 and the covering map of degree three @xmath109 in ( [ proj ] ) .
the following simple coordinate change @xmath110 converts the equation ( [ model ] ) into the weierstrass equation for fibration 5 @xmath111 the curve @xmath112 and @xmath113 correspond to the zero section , 3-torsion rational points @xmath114 and @xmath115 , respectively ( the dotted lines in figure [ no5 ] ) .
an elliptic parameter for fibration 6 is given by @xmath116 since we gave the divisor of @xmath48 in ( [ func ] ) , we know that the zero divisor @xmath117 is the singular fiber of type @xmath118 ( the bold lines in figure [ no6 ] ) and the polar divisor @xmath119 is the singular fiber of @xmath120 ( the thin lines in figure [ no6 ] ) .
the curves @xmath121 and @xmath122 ( the dotted lines in figure [ no6 ] ) are all the sections . choosing @xmath123 as the zero section of the group structure
, we obtain the weierstrass equation for fibration 6 @xmath124 where the change of variables is given by @xmath125 besides the two singular fibers of type @xmath118 at @xmath126 and of type @xmath120 at @xmath127 , there are three @xmath95 fibers at @xmath128 and @xmath129 .
the mordell - weil group of the fibration is isomorphic to @xmath130 .
the curve @xmath131 corresponds to the rational point @xmath132 of order two , and remaining curves @xmath133 and @xmath122 correspond to the rational points @xmath134 of order four , respectively .
to obtain the weierstrass equation for fibration 4 , we use a 2-neighbor step from fibration 3 . for more detail about _
2-neighbor step _ , we refer to @xcite .
we compute explicitly the elements of @xmath135 where @xmath136 is the class of the fiber of type @xmath137 we are considering .
the linear space @xmath135 is 2-dimensional , and the ratio of two linearly independent elements is an elliptic parameter for @xmath5 .
since @xmath16 is an element of @xmath138 , we may find a non - constant element of @xmath135
. then it will be an elliptic parameter of fibration 4 .
let us @xmath139 be a non - constant .
the function @xmath140 has a simple pole along @xmath92 and @xmath89 , which are the zero section and 2-torsion of fibration 3 . also , it has a simple pole along @xmath141 , the identity component of the fiber at @xmath142 , a simple pole along @xmath143 , the identity component of the fiber at @xmath144 .
therefore we can put @xmath145 where the variables @xmath146 are given by ( [ par3 ] ) and ( [ change3 ] ) .
assume @xmath147 , since @xmath16 is an element of @xmath135 . to obtain the coefficients @xmath148 and @xmath149 , we look at the order of vanishing along the non - identity components of fibers at @xmath144 .
the function @xmath140 does not have any pole along @xmath150 , which intersects with the section @xmath100 of the fibration 3 at @xmath151 .
hence @xmath140 has no pole at @xmath152 , and that gives us @xmath153 .
similarly , the component @xmath154 , which intersects with the section @xmath102 , gives us @xmath155 .
consequently , we have a new elliptic parameter @xmath156 where the variables @xmath146 are given by ( [ par3 ] ) and ( [ change3 ] ) . solving for @xmath157 and substituting into the weierstrass equation ( [ eq - no3 ] ) , after suitable coordinate changes we have the following @xmath158
although this is a weierstrass equation for fibration 4 , for latter calculations , we put @xmath159 and obtain another weierstrass equation for fibration 4 @xmath160 the change of variables is given by @xmath161 the fibration has singular fibers of type @xmath137 at @xmath162 and of type @xmath95 at the zeros of @xmath163 .
the zero section corresponds to the divisor @xmath164 .
the 3-torsion rational points @xmath165 and @xmath166 correspond to the divisors @xmath167 and @xmath168 , respectively .
the free rational points @xmath169 and @xmath170 correspond to the divisors @xmath100 and @xmath102 , respectively with the relation @xmath171 in the mordell - weil group .
since the height of @xmath102 is equal to @xmath172 , @xmath102 generates the mordell - weil lattice of the fibration .
we obtain the following elliptic parameter @xmath173 for fibration 2 by a 2-neighbor step from fibration 4 ( see figure [ no4 - 2 ] ) . the variables @xmath175 are given by ( [ change4 ] ) .
then we get the following weierstrass equation for fibration 2 .
@xmath176 we put @xmath177 and obtain another weierstrass equation for fibration 4 .
@xmath178 the change of variables is given by @xmath179 the zero divisor @xmath180 is the singular fiber of type @xmath181 ( the bold lines in figure [ no2 ] ) .
the polar divisor @xmath182 is the singular fiber of type @xmath183 ( the thin lines in figure [ no2 ] ) , where the divisor @xmath184 is the lifting of the curve @xmath185 on @xmath72 by the map @xmath77 in [ fib1 ] . besides these two singular fibers , there are three @xmath95 fibers at @xmath186 and @xmath129 .
the zero section corresponds to the divisor @xmath131 .
the 2-torsion rational point @xmath132 corresponds to the divisor @xmath89 .
we give a weierstrass equation for fibration 6 in [ fib6 ] . comparing the equations ( [ eq - no2 ] ) and ( [ eq - no6 ] ) , we know easily that fibration 2 is a quadratic twist of fibration 6 .
this is the reason why we adopt the equation ( [ eq - no2 ] ) as the weierstrass equation for fibration 2 rather than the equation ( [ eq - no2 - 0 ] ) .
the computer algebra system maple and maple library `` elliptic surface calculator '' written by professor masato kuwata @xcite were used in the calculation for this paper .
the author would like to thank the developers of these programs .
m. kuwata and t. shioda , `` elliptic parameters and defining equations for elliptic fibrations on a kummer surface '' , algebraic geometry in east asia - hanoi ( 2005 ) , 177 - 215 , adv . stud .
pure math .
, * 50 * , math .
japan , tokyo , 2008 . | in this paper we give the weierstrass equations and the generators of mordell - weil groups for jacobian fibrations on the singular @xmath0 surface of discriminant 3 . |
collisional ring galaxies ( crgs ) consist of one or more sharply defined star - forming rings , generally ( but not always ) with a nucleus inside @xcite .
theoretical models , both analytical ( e.g. @xcite ; @xcite ; @xcite ; @xcite ) and numerical ( e.g. @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite @xcite , @xcite ; @xcite ; @xcite ; @xcite ; @xcite @xcite , @xcite ; @xcite ; @xcite ) , suggest that crgs are the result of nearly axisymmetric high - speed encounters between a disk galaxy ( the ` target ' or ` primary ' galaxy ) and an intruder galaxy . after the interaction , radially expanding density waves form because of the crowding of star orbits in the target disk .
this causes gas compression and the triggering of a starburst episode along the expanding ring ( e.g. @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ) .
crgs are unique laboratories for the study of galaxy collisions because of their simple interaction geometry . for example , the kinematics of the ring provides information about the dynamics of the interaction ( e.g. @xcite ; @xcite ; @xcite ; @xcite ; @xcite ) .
this information is used to constrain @xmath2-body / hydrodynamical models reproducing the formation process of well studied crgs such as the cartwheel galaxy ( e.g. @xcite ; @xcite ; @xcite ; @xcite ) and arp 147 ( e.g. @xcite ; @xcite ) .
numerical simulations are helpful tools to explore the parameter space of the collision and to disentangle the effect of each parameter on the evolution of crgs .
@xcite study the properties of crgs formed in axisymmetric encounters with different intruder - to - target mass ratios .
they show that , as the intruder mass increases , the ring becomes sharper , expands more rapidly and the secondary ring is less developed .
@xcite explore the connection between gas fraction and star formation ( sf ) in simulations of empty crgs ( i.e. crgs without a central nucleus , such as arp 147 ) , and find that the gas fraction strongly influences the peak of sf rate ( sfr ) after the starburst .
@xcite perform a large parameter study with the aim of reproducing the dynamical properties of the auriga s wheel @xcite , but they do not take into account the effects of the interaction parameters on the sf . in this paper ,
we investigate how the geometry of the impact affects the morphology and sf history of crgs .
we employ a suite of adaptive mesh refinement ( amr ) numerical simulations to study systematically the properties of crgs as a function of the interaction parameters that define the geometry of the encounter .
this is the first time that the amr technique is applied to a wide grid of simulations studying the formation of crgs .
our choice is particularly important , as the amr technique ensures a better treatment of hydrodynamical instabilities and shocks with respect to particle - based schemes ( e.g. @xcite ; @xcite ) .
the paper is organized as follows : in section [ section2 ] we describe the main characteristics of our suite of simulations and briefly discuss the parameters of interest .
in section [ section3 ] we present the results , and focus on the morphology and sf history of our models .
we also extract synthetic photometric information about the colour of the ring and we compare it with observations . in section [ section4 ] we summarize our main findings .
our simulations of galaxy interactions were performed with the public amr code ramses @xcite .
ramses solves the fluid equations on an adaptive cartesian grid by means of a second - order godunov method coupled with a riemann solver .
this method ensures a better treatment of hydrodynamical instabilities and shocks with respect to particle - based schemes ( e.g. @xcite ; @xcite ) .
gravitational interactions are calculated using particle mesh ( pm ) techniques for @xmath2-body collisionless components ( stars and dark matter particles ) , solving the poisson equation on the amr grid with the multigrid strategy of @xcite .
all the simulations presented in this paper adopt a cubic box of 960 kpc per edge , with a static coarse grid composed of @xmath3 cells .
we used 7 additional levels of refinement to reach the maximum resolution of @xmath4 pc .
grid refinement is triggered following three criteria , one for @xmath2-body particles , the other two for the gaseous component .
the particle criterion requires that at least 24 particles reside in a cell before the refinement , so as to maintain a roughly constant number of particles per cell . the first criterion for gas
is based on a ` quasi - lagrangian ' approach according to which refining is triggered if the gas mass inside a cell exceeds a threshold of @xmath5 .
the second gas criterion is based on the jeans length ( @xcite ) : a cell at level @xmath6 is split if the local jeans length does not exceed 4 times the cell dimension @xmath7 . in the limit of our maximum resolution ,
this choice allows us to properly follow the evolution of features such as dynamically unstable gaseous clumps in the expanding ring .
gas dynamics is followed assuming a polytropic equation of state ( eos ) with adiabatic index @xmath8 , corresponding to a perfect mono - atomic gas .
the interacting galaxies are embedded in a hot and thin medium with numerical density @xmath9 , and temperature @xmath10 .
the ramses code includes physically - motivated sub - grid models for gas cooling , sf and supernova ( sn ) feedback .
the code adopts the cooling function described in @xcite for temperatures down to @xmath11 .
we include a sub - grid gaseous eos @xmath12 to model the behavior of the inter - stellar medium ( ism ) , where @xmath13 , @xmath14 and @xmath15 is the sf density threshold ( see @xcite ) .
sf is implemented in ramses following the prescriptions by @xcite .
the conversion of gas density @xmath16 into star density @xmath17 is based on a schmidt - law @xcite of the form : @xmath18 where @xmath19 is the sf efficiency , @xmath20 is the local dynamical time and @xmath15 is a density threshold .
we chose @xmath21 and @xmath15 corresponding to a numerical density @xmath22 .
such threshold density was selected to resolve the jeans length ( @xcite ) for a temperature @xmath23 k ( i.e. consistent with the minimum gas temperature in our simulations ) with our best spatial resolution ( see @xcite for details ) .
thus , the adopted threshold density is the highest that can be safely used with our spatial resolution and matches the one adopted in recent cosmological simulations that reproduce realistic milky way - like ( mw - like ) galaxies @xcite , and in simulations of the formation of off - centred crgs @xcite .
most of cosmological simulations adopt a much lower density threshold ( @xmath24 , e.g. @xcite and references therein ) , that might unphysically enhance the sf ( see e.g. the discussion in @xcite ) . @xcite
adopt @xmath25 ( a factor of 20 higher than our threshold ) in their cosmological simulations , but their spatial resolution is better than ours and their cooling function goes down to @xmath26 .
our adopted value of the sf efficiency ( @xmath21 ) is a factor of @xmath27 higher than adopted in recent simulations ( e.g. @xcite ) , which follow the observational hints by @xcite . on the other hand
, we stress that ( i ) the @xmath19 parameter in our simulations is simply phenomenological and does not have a direct relation to the true sf efficiency within molecular clouds ( which we can not simulate realistically , see the discussion below ) , and that ( ii ) @xcite show that the simulated sf is much more affected by @xmath15 than by @xmath19 . although the sf recipe used by @xcite is more elaborate than ours , we find similar results about the global sf history for qualitatively similar runs ( see the discussion below ) .
star particles are then spawned following a stochastic approach @xcite , with a mass that is an integer multiple of @xmath28 .
we stress that our best grid resolution ( 117 pc ) and our adopted cooling function @xcite are not suitable to follow the formation of low - temperature ( @xmath29 k ) high - density ( @xmath30 ) gas clouds , whose typical scale is expected to be @xmath31 pc and that were studied in previous simulations of multiphase interstellar medium ( e.g. @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ) . therefore , our simulations do not model sf in cold clumps , which is physically the main channel for sf . accounting for this would be computationally prohibitive to do for a wide set of galaxy interaction runs
sn feedback is implemented as thermal feedback due to type ii sn explosions characterized by a specific energy @xmath32 @xcite . a total energy amount @xmath33 is deposited in the 27 cells around the star particle of mass @xmath34 during @xmath35 after the star particle creation .
we assume a sn efficiency @xmath36 .
we model the target galaxy as a mw - like disc galaxy composed of three @xmath2-body components and a gridded one : a dark matter ( dm ) halo , a stellar disc , a stellar bulge and a gaseous disc , respectively . for the dm halo
we choose the density profile by @xcite ( @xcite , nfw ) : @xmath37 where @xmath38 is the critical density of the universe , where we assume a hubble constant @xmath39 @xcite .
] , @xmath40 is the halo scale radius and @xmath41 is defined as : @xmath42 where @xmath43 is the nfw concentration , @xmath44 , where @xmath45 is the virial radius of the dm halo so that the virial mass of the nfw halo is @xmath46 .
we assume an axisymmetric exponential profile to model both the gaseous and stellar discs : @xmath47 where @xmath48 is the total disc mass , and @xmath49 and @xmath50 are the radial and vertical scale length of both the gaseous and stellar discs , respectively .
the gaseous disc has an initial constant temperature @xmath51 and we assume that it is composed of a mixture of ionized hydrogen and helium with a mean molecular weight @xmath52 .
the density profile of the stellar bulge follows the spherical @xcite model : @xmath53 where @xmath54 is the total mass and @xmath55 the radial scale length of the bulge .
the intruder galaxy is a dwarf , spherical early - type galaxy composed of a nfw dm halo and a central stellar bulge devoid of any gaseous component .
we chose not to simulate a gas - rich intruder because , in this paper , we are interested in studying the evolution of the gas of the target separately . the parameters for the initial conditions ( ics ) of the two galaxies are shown in table [ table1 ] .
[ table1 ] .adopted values of the model parameters for the target and intruder galaxy .
these parameters do not change among runs . [
cols="<,^,^",options="header " , ] the set up of the encounters ranges from axisymmetric ( @xmath56 , @xmath57 ) to highly off - centre and inclined ( @xmath58 , @xmath59 ) interactions . the smaller value for @xmath60 corresponds to a marginally bound interaction , consistent with observations @xcite .
the higher value @xmath61 represents a very fast interaction within a massive galaxy groups ( @xcite , @xcite ) . , @xmath62 and @xmath63 .
see section 2.3 for details.,width=302 ]
fig . [ figure_xy ] shows the face - on view of the simulated target galaxies at time @xmath64 after the interaction .
images show the mass - weighted gas density maps .
regardless of the interaction parameters , the ring - like features of the runs have a typical radius @xmath65 ( @xmath66 ) .
this value slightly decreases for fast encounters , but the effect is small because the ring expansion velocity is mainly influenced by the mass of the perturber , in agreement with the results of @xcite . simulations with @xmath56 show circular or nearly elliptical rings , depending on the value of the inclination angle .
the gaseous component of the ring is quite broad ( with a width @xmath67 ) and clumpy , due to the combined effects of cooling , sf and sn feedback . inside the ring
we observe radial gaseous spokes that are thicker and slightly less developed for the encounter with @xmath60 , and the presence of a gaseous and star - forming secondary ring that recently started propagating .
observations of the cartwheel galaxy do not support our finding that spokes are mainly composed of gas : no hi or h@xmath68 emission is found in the spokes and in the inner ring ( @xcite , @xcite ) .
however , recent observations of ngc 922 @xcite show h@xmath68 emission and the presence of young star clusters along the incomplete ring as well as in the central bar and in correspondence to possible spokes .
also the inner ring of arp 10 has h@xmath68 emission and seems to be a site of sf ( @xcite , @xcite ) .
therefore , the amount of gas in the spokes and in the central regions of crgs may depend on the pre - encounter distribution of gas @xcite .
off - centre collisions with @xmath69 and @xmath58 produce results similar to those reported by @xcite .
the asymmetry of the interaction induces the displacement of the nucleus in the direction of the impact .
the ring is also asymmetric and composed of two main structures originating from the displaced nucleus and connected at the end : a wide arc and a straighter arm .
the wide arc is generally clumpy ( e.g. run a30b5v25 or a30b10v25 ) and resembles the gravitationally unstable tidal tail produced by galaxy interactions in @xcite . for simulations with @xmath58 , the ring is loosely connected for @xmath60 , and not connected in the case of @xmath61 .
this trend is enhanced when the inclination angle @xmath63 is increased .
our results suggest an approximate upper limit @xmath70 to produce a connected ring from an off - centre high - speed collision .
another common feature of all the runs with @xmath71 is the development of a stellar and gaseous bar - like feature in correspondence to the displaced nucleus .
[ figure_yz ] shows edge - on views of all the gaseous target discs at time @xmath64 after the interaction .
images show the effects of the interaction on the vertical structure of the target disc . in all cases ,
the target is displaced from its original position in the direction of the intruder by @xmath72 and @xmath73 for slower and faster interactions , respectively .
another common feature is the development of thin gaseous tails in the direction of the intruder passage , especially for inclined interactions with @xmath74 .
these tails can be sometimes observed as hi plumes departing from the target disc @xcite .
off - centre and low - speed interactions produce more disc vertical heating than faster and symmetric encounters .
moreover , the inclination angle induces a net torque on the target disc along the @xmath75 axis . therefore , after the interaction , the ring develops on a plane that is tilted with respect to the original plane of the disc . for the case with @xmath76 and @xmath77 , we can estimate the warp analytically .
assuming that the intruder moves on a straight line in the @xmath78 plane with an inclination angle @xmath63 defined as above , only the @xmath75 component of the torque @xmath79 does not vanish on the entire target disc .
it induces an angular momentum variation that can be estimated as ( see appendix [ appa ] for details ) : @xmath80 where @xmath81 is the module of the intruder velocity , @xmath82 is the radial scale of the intruder ( for which we assume a plummer potential , see appendix [ appa ] ) , @xmath83 is the intruder baryonic mass and @xmath84 is the exponential surface density of the target disc .
the unperturbed component of the angular momentum can be estimated as : @xmath85 where @xmath86 is the circular velocity of the unperturbed exponential disc .
then , we can parametrize the warping as : @xmath87 @xmath88 is the average warping of the target disc at each radial position , due to the gravitational torque exerted by the encounter .
this simple treatment does not take into account the perturbation by the primary disc on the intruder motion and the propagation of the ring ) is acceptable , as the total mass of the intruder far exceeds the fluctuations induced in the stellar density of the target . ] .
thus , it can only predict the overall inclination of the disc , represented by the nearly constant asymptotic value that @xmath88 assumes for radii @xmath89 ( @xmath90 ) , as shown in the left - hand panel of fig . [ figure_warping ] .
this plot shows the radial profile of @xmath88 deduced from eq .
( [ eq : angle_warp ] ) at two characteristic times ( 0 and 50 myr after the interaction ) for runs a30b0v25 and a30b0v65
. for @xmath81 we use the relative velocity between the cms of the baryonic components of the two galaxies at the moment of the interaction , as measured from simulations ( @xmath91 for run a30b0v25 and @xmath92 for run a30b0v65 ) .
we assume the radial scale length of the intruder is @xmath93 ( see appendix [ appa ] ) . for run
a30b0v25 , @xmath88 approaches the asymptotic value @xmath94 , while for run a30b0v65 @xmath95 .
these values are in fair agreement with simulations ( as shown in the right - hand panel of fig . [ figure_warping ] ) and represent the overall inclination of the target disc with respect to the @xmath96 plane . off - centre interactions with @xmath77 produce rings completely warped in the vertical direction , especially for slow encounters ( see fig .
[ figure_yz ] ) . in these cases ,
the loosely connected ring develops with a nearly helicoidal shape because of the coupled role played by the inclination angle , which induce the vertical torque discussed above , and by the asymmetric propagation of the expanding ring , due to @xmath97 @xcite .
we study the effects of the interaction parameters on the global sf histories of our simulations . in fig .
[ figure_sfra ] , we show the sfr as a function of time for all the runs with different inclination angle @xmath63 , while in fig .
[ figure_sfrv ] we compare the sf histories for runs with different initial relative velocity @xmath98 . before interaction
, the target has a low @xmath99 , consistent with the observational kennicutt - schmidt relation @xcite . for all models ,
the target experiences an enhancement of sf when the intruder crosses the target disc .
the sfr rapidly increases from @xmath100 in the pre - encounter phase to @xmath101 during the starburst , in agreement with observations ( @xcite , @xcite , @xcite ) . figs .
[ figure_sfra ] and [ figure_sfrv ] suggest that neither @xmath63 nor @xmath102 influence significantly the sf history of a typical crg . by contrast , the impact parameter @xmath62 influences significantly the sf history in our simulations , as shown in fig .
[ figure_sfrb ] . we systematically observe a double - peaked sf history for axisymmetric encounters . after the interaction ,
the sfr steeply increases , reaches a maximum and then decreases on a typical time scale @xmath103 .
this is probably because the ring travels rapidly across the disc during the first @xmath104 after the interaction ; when it reaches the outskirts of the disc , the lower gas density quenches the sf . the first sf bump is followed by a second episode of sf due to the propagation of the secondary ring
. instead ,
after the initial starburst , all off - centre encounters show a nearly constant or slowly rising sfr that lasts more than @xmath104 .
the sf histories of runs with @xmath105 and @xmath106 represent a transition between truly axisymmetric and strongly off - centre interactions , as they still show a less pronounced double - peaked shape .
these results are in agreement with @xcite .
in particular , the sfr in their c and d models ( with @xmath107 and @xmath108 , respectively ) reaches @xmath109 and @xmath110 , respectively , during the first @xmath111 after the interaction .
our run a0b10v65 is qualitatively comparable to theirs and we find similar values of @xmath112 despite the different implementation of the sf in the adopted codes .
none of the considered interaction parameters appreciably affects the peak of sfr .
this feature is more sensitive to other parameters such as @xmath113 , as pointed out by @xcite .
photometric properties can be inferred from our simulations , by using synthetic stellar population models .
we use mass - to - light ratio tables based on the isochrones and synthetic stellar populations of @xcite and @xcite .
these tables span the stellar age interval from @xmath114 to @xmath115 and give the magnitude in the @xmath116 bands for a stellar system with total mass @xmath117 and a chabrier initial mass function ( imf ; @xcite ) .
we assume a constant sub - solar metallicity @xmath118 , in agreement with observations of crgs ( @xcite ; @xcite ) . in fig .
[ figure_b - v ] the synthetic @xmath0 colour maps obtained for all our simulations at time @xmath64 are shown .
we calculated these maps after assigning a random age ( between @xmath119 and @xmath120 ) to stellar particles that were in the ics ; in this process we do not take into account any pre - interaction age / metallicity gradient . for purpose of visualization ,
maps are smoothed with a gaussian filter with @xmath121 .
all models show a blue ring - like feature with typical colour @xmath122 .
the rings are sprinkled with bluer knots corresponding to regions of recent sf , with colour @xmath123 .
the underlying disc is redder with an average colour @xmath124 .
the ring is characterized by a blue colour bump with respect to the disc , @xmath125 , which is consistent with @xmath0 profiles derived by @xcite for a sample of 15 crgs .
we repeated the same analysis assuming a metallicity @xmath126 , and we found similar results , even if with a redder ( @xmath127 ) overall colour .
simulations with @xmath76 show a gaseous secondary ring that just started to expand ( see fig .
[ figure_xy ] ) and that is expected to form stars , but colour maps do not show an excess of blue colour in the inner regions of the ring .
this is in partial agreement with observations : for example , the cartwheel galaxy shows a red but gas - poor secondary ring ( @xcite , @xcite ) . in our models ,
the secondary ring starts forming stars @xmath128 after , in correspondence to the second peak of the sfr at @xmath104 after the interaction .
off - centre collisions show very blue knots mainly located in the wide arc and at the extremity of the bar - like nuclear region .
this suggests that the asymmetry of the encounter can funnel gas to the nucleus and favor sf near it , as observed , for example , in ngc 922 @xcite .
our simulations likely underestimate the clumpiness of the blue knots in the ring , as we can not resolve the giant molecular clouds and thus we can not produce realistic sf agglomerates , such as young dense star clusters ( see section [ subsec21 ] for a discussion about the limitations of the model ) .
in this paper , we investigate the properties of crgs as a function of the interaction parameters by means of amr hydrodynamical simulations .
we focus on three parameters : the impact parameter , the initial relative velocity and the inclination angle .
all the simulations show a few common features , such as the radius of the ring ( @xmath65 ) at time @xmath64 after the interaction , which is weakly influenced by the three interaction parameters described above . on the other hand ,
axisymmetric encounters produce crgs noticeably different with respect to those formed by off - centre interactions .
axisymmetric interactions result in a circular primary ring followed by a smaller secondary ring , whereas crgs born from off - centre collisions show asymmetric rings and a displaced nucleus . in the most asymmetric cases ( @xmath129 and @xmath74 ) the ring
is not connected .
the maximum impact parameter to form a connected ring is @xmath70 , where @xmath49 is the scale length of the target disc .
interactions with @xmath77 induce a torque on the target disc producing a vertical warp . in the special case of encounters with @xmath56 and @xmath77 , the ring develops on a plane tilted with respect to the original target disc plane and
we propose a simple analytical treatment to predict the overall warping of the target disc after the interaction . after the interaction ,
the ring expansion enhances the sf in the target disc , reaching a typical sfr @xmath101 , in agreement with observations ( @xcite , @xcite ) .
we show that the initial velocity and the inclination angle have weak effects on the peak of the starburst and on the evolution with time of the global sfr .
instead , the impact parameter has a crucial effect on sf : symmetric encounters cause a short - lived starburst episode , whereas off - centre interactions produce long - lasting ( @xmath130 ) sf events .
the expanding ring is the region mainly affected by sf , as suggested by observations ( e.g. @xcite , @xcite ) .
this is confirmed by the synthetic colour maps ( based on the models by @xcite and @xcite ) that we compute from the simulations .
we find an average @xmath0 colour @xmath131 for the ring , with clumps of massive sf with @xmath132 .
the ring is typically @xmath133 bluer than the rest of the galaxy , in agreement with the @xmath0 colour profiles measured by @xcite for a sample of 15 crgs .
a number of issues can not be explained by our simulations .
for example , it is still unclear which mechanism regulates sf in the inner regions of the ring , since hi / h@xmath68 observations suggest that secondary rings are gas - poor ( e.g. , the cartwheel , @xcite ; @xcite ; but see also @xcite about arp 10 ) , while our simulations predict the opposite .
this may be due to several reasons , such as the initial distribution of gas inside the disc , or the idealized sub - grid sf process that does not take into account the chemical abundances and the role of @xmath134 .
thus , future works will require more realistic initial gas distributions and/or more detailed recipes of sf ( e.g. @xmath134 regulated sf , @xcite ; @xcite ) to better constrain the sf in crgs .
another interesting issue we did not explore ( but plan to investigate in the future ) is the effect of a gas - rich intruder . in this respect ,
we remind that the sph simulations of @xcite suggest that the main differences between encounters with ( disc - like ) gas - rich and gas - poor intruders are related to the sf history , whereas the effects on the ring morphology are quite limited , even if some features ( such as plumes ) appear to be more developed in the gas - rich case .
we thank the anonymous referee for comments that significantly improved the paper .
we thank r. teyssier for useful advice about the use of ramses .
we thank l. widrow for providing us the code to generate the initial conditions .
we thank l. girardi for maintaining his public website about stellar isochrones .
we thank the authors of pymses , which we used for analysing the simulations .
all the simulations were performed with the sp6 cluster at the consorzio interuniversitario dellitalia nord - est per il calcolo automatico ( cineca ) .
we acknowledge the cineca award n. hp10cli3bx and hp10b3bjew , 2011 for the availability of high performance computing resources and support .
we acknowledge financial support from inaf through grant prin-2011 - 1 .
we thank l. mayer for useful discussions .
99 agertz , o. , et al .
2007 , , 380 , 963 agertz , o. , lake , g. , teyssier , r. , moore , b. , mayer , l. , romeo , a. b. 2009 , , 392 , 294 agertz , o. , teyssier , r. , & moore , b. 2011 , , 410 , 1391 appleton , p. n. , & james , r. a. 1990 , dynamics and interactions of galaxies , 200 appleton , p. n. , & struck - marcell , c. 1987 , , 318 , 103 appleton , p. n. , & struck - marcell , c. 1996 , , 16 , 111 binney , j. , & tremaine , s. 2008 , galactic dynamics : second edition , by james binney and scott tremaine .
isbn 978 - 0 - 691 - 13026 - 2 ( hb )
. published by princeton university press , princeton , nj usa , 2008 bizyaev , d. v. , moiseev , a. v. , & vorobyov , e. i. 2007 , , 662 , 304 bottema , r. 2003 , , 344 , 358 bransford , m. a. , appleton , p. n. , marston , a. p. , & charmandaris , v. 1998 , , 116 , 2757 carlberg , r. g. , yee , h. k. c. , morris , s. l. , lin , h. , hall , p. b. , patton , d. r. , sawicki , m. , shepherd , c. w. 2001 , , 552 , 427 chabrier , g. 2001 , , 554 , 1274 charmandaris , v. , & appleton , p. n. 1994 , bulletin of the american astronomical society , 26 , 877 charmandaris , v. , appleton , p. n. , & marston , a. p. 1993
, , 414 , 154 conn , b. c. , pasquali , a. , pompei , e. , lane , r. r. , chen , a .-
smith , r. & lewis , g. f. 2011 , , 741 , 80 donghia , e. , mapelli , m. , & moore , b. 2008 , , 389 , 1275 evoli , c. , salucci , p. , lapi , a. , & danese , l. 2011 , , 743 , 45 faltenbacher , a. , & mathews , w. g. 2007 , , 375 , 313 few , j. m. a. , arp , h. c. , & madore , b. f. 1982 , , 199 , 633 fogarty , l. , et al .
2011 , , 417 , 835 fosbury , r. a. e. , & hawarden , t. g. 1977 , , 178 , 473 gerber , r. a. , lamb , s. a. , & balsara , d. s. 1992 , , 399 , l51 gerber , r. a. , lamb , s. a. , & balsara , d. s. 1994 , bulletin of the american astronomical society , 26 , 911 gerber , r. a. , & lamb , s. a. 1994 , , 431 , 604 gerber , r. a. , lamb , s. a. , & balsara , d. s. 1996 , , 278 , 345 gerritsen , j. p. e. , & icke , v. 1997 , a&a , 325 , 972 ghosh , k. k. , & mapelli , m. 2008 , , 386 , l38 girardi , l. , et al .
2010 , , 724 , 1030 governato , f. , et al .
2010 , nature , 463 guedes , j. , callegari , s. , madau , p. , & mayer , l. 2011 , , 742 , 76 guillet , t. , & teyssier , r. 2011 , journal of computational physics , 230 , 4756 hernquist , l. 1990 , , 356 , 359 hernquist , l. , & weil , m. l. 1993 , , 261 , 804 higdon , j. l. 1995 , , 455 , 524 higdon , j. l. 1996 , , 467 , 241 higdon , j. l. , & wallin , j. f. 1997 , , 474 , 686 horellou , c. , & combes , f. 2001 , , 276 , 1141 jeans , j. h. 1919 , _ problems of cosmogony and stellar dynamics _
, cambridge , university press katz , n. 1992 , , 391 , 502 kennicutt , r. c. , jr .
1998 , , 498 , 541 krumholz , m. r. , tan , j. c. 2007 , , 654 , 304 kuhlen , m. , krumholz , m. r. , madau , p. , smith , b. d. , & wise , j. 2012 , , 749 , 36 krumholz , m. r. , mckee , c. f. , & tumlinson , j. 2009 , , 693 , 216 kuijken , k. , & dubinski , j. 1995 , , 277 , 1341 larson , d. , et al .
2011 , , 192 , 16 lynds , r. , & toomre , a. 1976 , , 209 , 382 madore , b. f. , nelson , e. , & petrillo , k. 2009 , , 181 , 572 mapelli , m. , & mayer , l. 2012 , , 420 , 1158 mapelli , m. , moore , b. , giordano , l. , mayer , l. , colpi , m. , ripamonti , e. & callegari , s. 2008 , , 383 , 230 mapelli , m. , moore , b. , ripamonti , e. , mayer , l. , colpi , m. , & giordano , l. 2008 , , 383 , 1223 marigo , p. , girardi , l. , bressan , a. , groenewegen , m. a. t. , silva , l. & granato , g. l. 2008 , , 482 , 883 mayya , y. d. , bizyaev , d. , romano , r. , garcia - barreto , j. a. , & vorobyov , e. i. 2005 , , 620 , l35 navarro , j. f. , frenk , c. s. , & white , s. d. m. 1996 , , 462 , 563 pellerin , a. , meurer , g. r. , bekki , k. , elmegreen , d. m. , wong , o. i. , knezek , p. m. 2010 , , 139 , 1369 price , d. j. 2008 , journal of computational physics , 227 , 10040 rappaport , s. , levine , a. , pooley , d. , & steinhorn , b. 2010 , , 721 , 1348 rasera , y. , & teyssier , r. 2006 , , 445 , 1 romano , r. , mayya , y. d. , & vorobyov , e. i. 2008 , , 136 , 1259 schmidt , m. 1959 , , 129 , 243 smith , r. , lane , r. r. , conn , b. c. , & fellhauer , m. 2012 , arxiv:1203.2610 struck - marcell , c. 1990 , , 99 , 71 struck , c. 1997 , , 113 , 269 struck - marcell , c. , & lotan , p. 1990
, , 358 , 99 struck - marcell , c. , & higdon , j. l. 1993 , , 411 , 108 sutherland , r. s. , & dopita , m. a. 1993 , , 88 , 253 tasker , e. j. , & bryan , g. l. 2006 , , 641 , 878 teyssier , r. 2002 , , 385 , 337 theys , j. c. , & spiegel , e. a. 1977 , , 212 , 616 wada , k. , & norman , c. a. 2007 , , 660 , 276 wada , k. , meurer , g. , & norman , c. a. 2002 , , 577 , 197 wetzstein , m. , naab , t. , & burkert , a. 2007 , , 375 , 805 widrow , l. m. , & dubinski , j. 2005 , , 631 , 838 widrow , l. m. , pym , b. , & dubinski , j. 2008 , , 679 , 1239 zhang , w. , li , c. , kauffmann , g. , zou , h. , catinella , b. , shen , s. , guo , q. , chang , r. 2009 , , 397 , 1243
we model the intruder galaxy as a plummer sphere with potential : @xmath135 where @xmath82 is the scale radius of the potential and @xmath83 is the intruder baryonic mass .
we assume that the intruder moves on a straight line in the @xmath78 plane , according to @xmath136 , where @xmath63 is the inclination angle and @xmath81 is intruder velocity . for simplicity , we assume that @xmath81 is constant in time .
the target disc initially lies on the @xmath96 plane and has a surface density profile @xmath137 .
the force exerted on an infinitesimal mass portion @xmath138 located at @xmath139 of the target disc is : @xmath140 we can now derive the torque exerted on the same mass element : @xmath141 where : @xmath142 integrating eq .
( [ eq : dn_a ] ) over @xmath143 , we obtain the total torque exerted on an infinitesimally thin annulus of radius @xmath144 .
we point out that the @xmath145 and @xmath146 components are proportional to : @xmath147 therefore , the resulting torque per unit length is : @xmath148 integrating this equation with the boundary condition @xmath149 , we obtain the total torque : @xmath150 the angular momentum @xmath151 is related to the torque @xmath79 by : @xmath152 we can integrate in time eq . ( [ eq : n_a ] ) to obtain the variation of @xmath151 due to the torque exerted by the intruder , @xmath153 , with the condition @xmath154 ( i.e. we assume that for @xmath155 @xmath156 and the torque is zero ) .
we finally obtain : @xmath157 this expression can be simplified by assuming @xmath158 . in this case
we have : @xmath159 in this treatment , the @xmath146 component of the angular momentum @xmath151 is not perturbed by the passage of the intruder
. the unperturbed component of the angular momentum is given by : @xmath85 where the circular velocity @xmath160 for an exponential disc is given by @xcite : @xmath161^{1/2},\ ] ] where @xmath162 and @xmath163,@xmath164,@xmath165,@xmath166 are modified bessel functions .
the warping of the disc can be parametrized by the warping angle @xmath88 , defined as : @xmath167 this angle represents the inclination of the disc with respect to the @xmath96 plane around the @xmath145 axis at each radius @xmath168 at time @xmath169 due to the gravitational torque exerted by the encounter .
this treatment of the torque induced by the intruder galaxy on the target disc is very approximated , as it does not take into account the gravitational effects of the target galaxy on the motion of the intruder and the propagation of the ring as a consequence of the interaction .
moreover , the intruder is modeled as a plummer sphere for the sake of simplicity .
however , we choose the radial scale @xmath82 such that the mass enclosed inside @xmath82 by the plummer profile is the same as enclosed by the hernquist profile : @xmath170 where @xmath55 is the scale radius of the hernquist profile . despite the approximations
, this treatment roughly highlights the effects of a non - axisymmetric interaction with inclination angle @xmath77 ( see [ subsub : warping ] ) . | collisional ring galaxies are the outcome of nearly axisymmetric high - speed encounters between a disc and an intruder galaxy .
we investigate the properties of collisional ring galaxies as a function of the impact parameter , the initial relative velocity and the inclination angle .
we employ new adaptive mesh refinement simulations to trace the evolution with time of both stars and gas , taking into account star formation and supernova feedback .
axisymmetric encounters produce circular primary rings followed by smaller secondary rings , while off - centre interactions produce asymmetric rings with displaced nuclei .
we propose an analytical treatment of the disc warping induced by an inclination angle greater then zero .
the star formation history of our models is mainly influenced by the impact parameter : axisymmetric collisions induce impulsive short - lived starburst episodes , whereas off - centre encounters produce long - lived star formation .
we compute synthetic colour maps of our models and we find that rings have a @xmath0 colour typically @xmath1 bluer than the inner and outer disc , in agreement with observations .
[ firstpage ] galaxies : interactions galaxies : peculiar galaxies : star formation methods : _ n_-body simulations . |
a vast majority of bacteria and archaea can grow in diverse environmental conditions .
the range of those conditions include high pressures @xcite , high temperature @xcite , low temperature @xcite , high salinity , low @xcite and high ph @xcite etc .
since these conditions are not hospitable for other life forms hence these organisms are named extremophiles @xcite .
one of the first isolated extremophile _ thermus aquaticus _
, a thermophilic bacteria can survive at near - boiling temperatures @xcite .
adaptation of these organisms to such harsh conditions raises many interesting questions how do they adapt to these conditions ? does the adaptation occur at single component level such as mutations in proteins leading to their barostability and thermostability , or the adaptation to these conditions has a collective nature whereas more than one cellular components act in compliance to preserve the functionality of each other .
recent studies on various aspects - such as taxonomy , ecology , enzymology of these microorganisms have provided insights on the adaptation of these organism to their environmental conditions @xcite .
for example , the bacterial cytoplasmic membrane must maintain its liquid - crystalline structure and semipermeability with changing conditions @xcite .
it was shown that bacterial membrane adapt to the temperature changes by changing their lipid composition @xcite .
adaptation of a protein to non - ambient conditions requires that it maintains its catalytic activity as well as its structure @xcite .
most globular proteins denature both at high as well as low temperatures .
moreover , even if a protein does not denature at low temperature , small thermal fluctuations will lead to decreased catalytic activity at low temperatures .
indeed in few studies on proteins from psychrophilic organisms , it was found that proteins are more flexible @xcite .
however , increase of flexibility also leads to high propensity of unfolding of the protein .
hence a fine balance between the structural flexibility and stability is required @xcite .
recent comparative study of an essential recombination protein reca from mesophilic and thermophilic bacteria suggests that its function of binding to single stranded dna is adapted to the conditions in which organisms grow @xcite .
a study of ssb , a single - stranded dna binding protein , from mesophilic and piezophilic bacteria show similar adaptation @xcite . while there is a large body of work on the stability and kinetics of proteins and adaptation of different components of prokaryotes obtained from extremophiles , the growth of bacteria
is only approached using conventional methods such as plate counting .
such studies have provided killing curves of saturated bacterial solutions upon increasing pressures and hence a pressure - temperature phase diagram of the bacterial survival is obtained @xcite . in order to understand the adaptation of bacterial cells to extreme pressure and temperature ,
an understanding of growth bottlenecks and physical changes of bacterial growing at ambient conditions induced by different thermodynamic conditions is important . in this paper
, we study the pressure - temperature dependence of growth and phenotypic changes of a mesophilic bacterium , _
e. coli _ , using an optical method which allows us to measure the growth of bacterial in real time at different pressures and temperatures .
we have investigated the growth and morphological changes in a wide range of pressure and temperature . in `` methods '' ,
we describe our experimental setup and protocol to measure growth of bacteria . in `` results '' ,
we summarize the results of the pressure - temperature dependence of growth followed by a stochastic model to account for the morphological changes induced by high pressure .
finally , we summarize and discuss our results in `` summary and discussions '' .
measurement of growth at normal conditions is rather easy as there are many commercial photometers available .
high pressure and temperature require that a photometer optics is built around a high pressure cell in order to obtain the growth curve .
below we describe our experimental setup to measure the growth .
bacteria absorb and scatter light with intensity which depends on the scattering angle and absorption coefficient @xcite .
the most common method of measuring bacterial concentration in a solution is turbidity method where the extinction of the light is measured at a fixed angle , usually in forward direction .
the method relies on many assumptions including ( i ) each bacterial cell is an independent scatterer ( ii ) the shape of bacterial cell is uniform ( iii ) multiple scattering is negligible .
the extinction cross section @xmath7 is a sum of cross section due to scattering @xmath8 and due to absorption @xmath9 , and can be written as @xmath10 then the coefficient of extinction @xmath11 is @xmath12 where @xmath13 is the number density of bacterial cells . the intensity @xmath14 detected by a light detector after the incident light traverses a distance @xmath15 in the scattering medium is given by @xmath16 where @xmath17 is the intensity of the light incident on the medium .
hence the difference of the logarithm of the intensities of incident and the transmitted light is proportional to the concentration of scatterers in the medium . the optical density ( od ) of the medium
is thus defined as @xmath18 a schematic of our experimental setup to measure pressure temperature dependence of growth is shown in fig .
[ fig : fig1 ] .
a sample of bacteria with lb medium is contained in a rectangular cuvette ( spectrocell ; volume : 400@xmath19 ) made up of fused silica and having a square cross - section ( 6mmx6 mm ) .
the cuvette with a flexible movable teflon cap is loaded into the high pressure cell ( iss , illinois , usa ) .
a piston is used to pressurize the water inside the pressure cell , and the pressure is measured by a pressure gauge .
the growth of the bacteria is measured in real time by shining a white light ( xenon lamp ) which passes through an excitation bandpass filter ( semrock , ff01 - 586/15 - 25 ) and is focused onto the cuvette holding the sample .
we chose light with @xmath20 nm wavelength to keep our measurements consistent with measurements done with most of the commercially available photometers .
the transmitted light is focused on a light detector on the other side of the cuvette which measures the intensity of the transmitted light .
the light detector is built around photo sensor chip tsl230r ( taos ) .
tsl230r photo sensor consists of a silicon photodiode with a current to frequency converter built into it .
the nonlinearity error is typically 0.2% at 100khz .
the frequency of the ttl signal from the detector proportional to the incident light intensity is measured using mio16 frequency counter chip interfaced to labview software ( national instruments ) .
we maintain the intensity of the light source such that all our experiments fall into the linear regime of the sensor .
the distance between the light sensor and the cuvette was @xmath21 cm .
distance between the detector and the sample dictatates the angular integration of the scattered light incident on the sensor .
the temperature of the high pressure cell is regulated using a circular water bath thermostat .
the time for growth measurement ranged between @xmath22 and @xmath23 mins , depending on the pressure - temperature dependent growth rate of bacteria .
the entirely closed structure of our experimental setup imposes a major limitation on the regulation of oxygen in our experiments .
the growth measurements were done in oxygen limited conditions .
the partial pressure of oxygen in lb medium was @xmath24kpa .
for all the experiments reported here , dh@xmath25 strain of _
e. coli _ was used .
while other wild type strains of e. coli such as mg1655 ( k-12 ) or mc1000 are common for studying the physiology of bacterial cells due to least amount of genetic mutations , dh5@xmath11 offers certain advantages for our studies .
earlier studies have shown that a major effect of the pressure on the morphology is elongation of cells .
due to cell elongation , sos system is implicated in the change of morphology at high pressures .
the reca1 mutation in dh5@xmath11 causes the elimination of the homologous recombination , an initiatiator process for sos pathway upon uv irradiation @xcite .
lack of reca mediated recombination in dh@xmath11 removes the effect of pressure on the sos pathway .
hence our experimental results would be able to distinguish the high pressure effects where recombination system is not involved ( discussed in the summary and dicussion section ) .
the drawback of using dh@xmath25 is that since it lacks the homologous recombination system , the cell are sickly and the growth is slower compared to other wild type strains . due to its slow growth ,
cells were grown in standard luria broth ( lb ) medium @xcite , which is a rich medium for bacterial growth .
the ph of the growth medium was kept to @xmath26 by adding naoh to the solution . for the consistency of the experiments ,
cells were first grown on a lb plate for about 10 hours and then subsequently used for experiments as described below . * _ growth conditions and measurements .
_ * bacterial cells picked from lb plate were first grown in lb medium at atmospheric pressure and @xmath27c in an incubator until the optical density ( od ) of the solution is about @xmath28 .
a small amount of freshly grown bacterial cells was then added to a cuvette containing @xmath29 l of medium to bring the initial od to @xmath30 and was used as the starting point for all the pressure temperature measurements . the final bacterial solution with lb medium
was then transfered to a high pressure cuvette at room temperature and pressure and was closed with a teflon cap .
the cuvette with the bacterial solution was then put into the high pressure chamber ( see the experimental setup ) equilibrated at the temperature of interest and the piston of the high pressure setup was slowly increased until the pressure gauge reading reaches the desired value of the pressure .
the growth of the bacterial cells then was assessed by measuring extinction of light as described in our experimental setup .
growth measurements were done in a sealed high pressure cell .
images were taken using sensicam cooled ccd camera connected to zeiss axiovert 35 microscope with a 40x olympus objective .
image analysis of bacterial cells was done using imagej software @xcite .
in fig . [ fig : fig2 ] , we show growth curve obtained in our experiments at @xmath31 atm and @xmath27 c. atm and @xmath27 c. the growth shows a typical lag , exponential , and saturation regimes , width=302 ] the growth curve shows a typical lag growth regime at small times followed by an exponential growth phase and finally a saturation regime .
the value of the saturation od @xmath32 is smaller compared to the saturation od ( typically 1.0 ) reached when oxygen is not a growth limiting agent . in the oxygen limited environment
both the division time and saturation od are affected . in figs .
[ fig : fig3 ] ( a ) and ( b ) , we show the growth curve of _
e. coli _ for various pressures at @xmath33c and @xmath34 c respectively .
we find that where the saturation is reached within the time scale of our experiments , the time profile of the growth curves show the typical characteristics of growth at @xmath31 atm and @xmath27c .
the number of bacterial cells at time @xmath35 in the exponential regime can be written as : @xmath36 where @xmath37 is the number of bacterial cell at the beginning of the exponential phase and @xmath38 is the division time .
@xmath39 corresponds to the number of generation in a given time @xmath35 . in fig .
[ fig : fig3 ] ( c ) and ( d ) , we show @xmath40 extracted from figs .
[ fig : fig3 ] ( a ) and ( b ) for various pressures at @xmath33c and @xmath34c respectively .
we find that @xmath41 increases , and hence the rate of growth decreases , upon increasing pressure .
we further find that the @xmath42 corresponding to the saturation regime decreases upon increasing pressure .
earlier studies on the effect of pressure on the total biomass production of different bacteria have found a similar decrease in total mass as a function of pressure @xcite . c. ( b ) growth curves at different pressures for @xmath34c .
( c ) division ( doubling ) time @xmath41 extracted from fig .
[ fig : fig3](a ) .
( d ) division time @xmath43 extracted from fig .
[ fig : fig3 ] ( b ) .
pressure dependence of @xmath41 is marked by a sharp increase at high pressures where the cells still grow but the growth is extremely slow.,height=340 ] the division time @xmath41 at a given temperature increases with pressure but shows a discontinous jump at high pressures .
we find that the discontinuous jumps in @xmath38 ocurrs between @xmath44 atm for all the temperatures studied in our experiments .
to further characterize the low pressure regime of @xmath38 , in fig .
[ fig : fig4 ] , we show the division time @xmath41 as a function of pressure for two different temperature on a linear - log plot .
we find that the low pressure regime of increase of division time with pressure can be fit by an exponential function where the exponent increases with decreasing temperature . the discontinuous change in @xmath41 coincides with departure from exponential behavior .
pressure and temperature do not only affect the structural stability of biomolecules but can also affect the thermodynamic force driving different biochemical processes inside the cell .
in general , the time scale of a given chemical reaction is proportional to @xmath45 , where @xmath46 is the boltzmann constant and @xmath47 is the volume change across the chemical reaction .
it is easy to see that any chemical reaction accompanied by a positive volume change will exponentially slowdown with pressure . in this context ,
the exponential dependence of @xmath41 with pressure ( fig.[fig : fig4 ] ) is not a surprise .
note that it is a very simple consideration as most of the biochemical processes are not invidividual but usually involve a cascade of chemical reacations corresponding to any cellular module .
nonetheless , the exponential dependence of the division time with pressure does suggest an overall positive volume change .
furthermore , @xmath47 itself is a function of pressure and temperature . at moderate pressures and temperature
one may assume it to be a constant .
it is hard to speculate the mechanisms responsible for slow growth and further experiments must be carried out to precisely figure out the decrease of growth rate at high pressure .
the other remarkable feature of the pressure dependence of division time is the abrupt increase of @xmath41 in the range of pressures 200 - 400 atm for all the temperatures studied here .
where does this discontinuity in the pressure dependence of growth come from ?
discontinuity in the growth as a function of pressure suggest that something abrupt must happen in these range of pressure .
the range of pressures where we see a discontinous jump in the division time can not be attributed to protein denaturation as the pressure is not high enough to denature the proteins .
while the proteins stability is rather unaffected , the functionality of proteins may show a large variability in this range of pressures @xcite .
we hypothesize that the discontinuous jump in the division time stem from functional changes in biomolecules . for two different temperatures @xmath33 c and @xmath27 c on a linear - log plot .
the low pressure linear depedence on a linear - log plot suggests that @xmath41 follows an exponential behavior .
the discontinous jump in @xmath41 at a given temperature is marked by its departure from the initial exponential behavior.,height=302 ]
in fig . [ fig : fig5 ] , we show the surface plot of pressure - temperature dependence of @xmath48 . it shows
smooth change as a function of pressure and temperature but high pressures as well as low temperatures growth are marked by sharp change in @xmath38 .
we further find that the slope of the locus of the points in the ( p , t ) plane where @xmath48 shows sharp transition with respect to pressure resembles the functional phase diagram of a typical protein ( shown as dotted white curve in fig .
[ fig : fig5 ] ) @xcite .
a careful observation of the @xmath49 data reveals that at low @xmath1 , there is a region where @xmath38 exhibits a non - monotonic behavior with pressure . in this narrow region of pressure and temperature , @xmath50 first decreases and then increases further upon increasing pressure . the yellow dotted line with @xmath51 marks the boundary between this anomalous behavior and normal behavior of increasing division time with increasing pressure .
we hypothesize that this anomalous behavior of division time as a function pressure results from structural transition in the phospholipids present in the cell membrane at low temperatures @xcite . .
solid red circles are the experimental data points .
white dotted line marks the loci of the points where @xmath38 changes abruptly .
yellow dotted line with @xmath52 is region separating anomalous pressure dependence of the division time.,width=340 ]
besides the slow growth of the population of bacterial cells at high pressures , the other interesting features of the response to high pressure is found in the morphological changes in bacterial cells @xcite .
we find that average bacterial cell length of e. coli increases as a function of pressure ( see fig .
[ fig : fig6 ] ) . c. ( a ) 1 atm ( b ) 50 atm ( c ) 100 atm ( d ) 200 atm ( e ) 250 atm ( f ) 300 atm .
the images were taken and analysized at the end of experiments for all the pressures.,width=340 ] to further characterize the bacterial elongation , we looked at the distribution of bacterial cell length at various pressures at a given temperature . in fig .
[ fig : fig7 ] , we show the distribution of bacterial cell length @xmath53 at the end of our experiments for pressures @xmath54 atm respectively for @xmath33 c. the distribution @xmath55 of @xmath53 at @xmath31 atm follows a gaussian distribution .
as the pressure is increased , @xmath55 starts developing a non - gaussian tail suggesting a growing bacterial cell length heterogeneities .
a major fraction of the cells still retain the same morphology as @xmath31 atm , but there is an increase in the population of elongated cells upon increasing pressure .
c. ( a ) @xmath56 atm ( b ) @xmath57 atm ( c ) @xmath58 atm ( d ) @xmath59 atm.,width=340 ] the average value of the cell length @xmath60 increases upon increasing pressure and shows a sharp increase at the same pressure where the division time also shows a sharp increase ( see fig .
[ fig : fig8 ] ) . while the bacterial cell elongation at high pressure is known , the sharp transition at high pressure is new . furthermore , we find that the pressure and temperature where the growth is marked by a sharp increase in division time correlates well with sharp changes in bacterial cell length . the exponential increase of division time with pressure as we saw in the earlier section can be interpreted as exponential decrease of overall kinetics leading to slow growth due to cell elongation at high pressure . while the increased cell length upon increasing pressure would explain the decreased rate of growth , it is not clear if the elongated cells would grow slower than the cells with normal morphology .
for example , if the elongated cells just lack the ability of cell division but replicate their genome normally then one would expect the growth rate per unit cell size not to change unless other kinetic processes also get affected by the increase of pressure . as a function of pressure at @xmath33 c.
average length of bacterial cells shows a sharp transition between @xmath61 and @xmath59 atm.,width=302 ] the elongation of bacterial cells at high pressure has been a subject of intense research and to our knowledge no consensus on the molecular mechanism responsible for it is reached @xcite . to account for the bacterial cell length heterogeneities and elongation upon increased pressure
, we propose a stochastic model in the next section .
a quick overview of figs . [ fig : fig6 ] and [ fig : fig7 ] suggests that while the average length of the bacterial cells undergoes a sharp transition at high pressures , a major fraction of bacterial still retain the morphology of a normal cell . or switch to a filamenting phenotype with a probability @xmath62 .
once a bacterial cell s fate changes to filamentation , it just grows without dividing.,width=302,height=151 ] the change of morphology can be thought of as an irreversible switch of bacterial morphology during the course of growth of the bacterial cell where the rate of switching will depend on the pressure .
let us assume that we start with @xmath63 cells at time @xmath64 .
let s further assume that the cells either decide to divide into two identical cells with probability @xmath11 or grow irreversibly to a filamenting bacteria with probability @xmath65 .
let s further assume that the internal growth rate of both the normal cells and the filamenting cells is the same and hence whenever bacterial do not divide from one generation @xmath66 to a generation @xmath67 , the cell length just doubles ( see fig .
[ fig : fig9 ] ) .
hence at the end of the @xmath66 generations of division , the system will have different distribution of bacterial cell length @xmath68 .
we can assume the initial distribution of bacterial length , as suggested by our experimental data , to be a gaussian given by : @xmath69 it can be shown easily that the above scheme of irreversible stochastic switching leads to number of various lengths @xmath53 of bacterial cells at the end of @xmath66 generations given by : @xmath70 hence in the extreme cases ( i ) no switching or @xmath71 will lead to no changes in the bacterial length ( ii ) @xmath72 , all the cells will elongate to the maximum limited by growth and number of divisions . in general
, the number of bacterial cells of length @xmath73 is given by : @xmath74 the total number of bacterial cells at the end of @xmath66 generation can be given by : @xmath75 now the probability @xmath76 of a bacterial cell with length @xmath77 is given by : @xmath78 which in terms of the switching probability @xmath62 can be written as : @xmath79}\ ] ] hence the expectation value of length @xmath80 at the end of @xmath66 generation is given by : @xmath81}\sum_{a=0}^{n}(\frac{\beta}{1-\beta})^a \\ & = & \frac{(1-\frac{3}{2}\beta)}{(1 - 2\beta)}\frac{[1-(\frac{\beta}{1-\beta})^{n+1}]}{[1-(\frac{\beta}{2(1-\beta)})^{n+1}]}\end{aligned}\ ] ] since the distribution of @xmath53 at @xmath64 is a gaussian , the distribution of length at the end of @xmath66 generations can be written as : @xmath82 } \\ & & \frac{1}{\sqrt{2\pi\sigma_l^2}2^{2a}}e^{-(l-2^a.l_0)^2/(2^{2a}\sigma_0 ^ 2 ) } \label{eq : pnofl}\end{aligned}\ ] ] .
( b ) model prediction of the average length as a function of switching probability @xmath62.,width=340 ] in fig .
[ fig : lengthvsbeta](a ) , we show the evolution of the distribution of @xmath53 for a fixed value of @xmath83 .
as the time progresses the distribution develops a long tail .
note that for @xmath84 , the system would undergo an irreversible fate where after few generations the population will be dominated by elongated cells and cells with normal length will vanish from the population in the limit of long time .
we further show the expectation value of length @xmath60 in fig .
[ fig : lengthvsbeta ] ( b ) as a function of @xmath62 for a fixed number of generations .
@xmath85 increases slowly for small @xmath62 values and growth sharply with increasing @xmath62 . to further test our model
, we compare the data of @xmath55 at @xmath86 atm with our model in fig .
[ fig : distexpmodel ] .
we find that the model can reasonably reproduce the length distribution . with experimental data at @xmath86 atm and @xmath33 c with model parameters @xmath83 and @xmath87 .
experimental data is shown in solid black line while the model prediction is shown in dashed red line.,width=340 ] to further characterize the heterogeneities in the population of bacterial cell length , we calculate a non - gaussian measure @xmath88 @xcite of the distribution @xmath89defined by : @xmath90 where @xmath91 and @xmath92 are the second and fourth central moments of the distribution @xmath89 respectively .
@xmath93 corresponds to a gaussian distribution , while a deviation of @xmath88 from zero corresponds to the degree of deviation from a gaussian distribution . in fig .
[ fig : phivsbeta ] , we show the dependence of @xmath88 on switching probability @xmath62 for @xmath87 .
we find that @xmath88 grows slowly first but increases sharply with @xmath62 . in fig .
[ fig : phivsbeta ] , we also show the non - gaussianity parameter @xmath88 measured from the experimental distribution of cell lengths at pressures @xmath94 and @xmath58 atm and temperature @xmath33 c as solid red circles .
note that model assumes a transition but is not able to say much about the physical origin of such phenotypic transition .
what are the biophysical mechanisms responsible for the cell elongation ? where does the stochasticity come from ?
the clue to the latter comes from the measured transition in the cell length observed here and the polymerization of one of the cytoskeletal proteins responsible for cell division , ftsz .
recent experiments on ftsz _ in vivo _ and _ in vitro _
suggests that ftsz protein depolymerizes at high pressures leading to delocalization of ftsz in the cell @xcite .
furthermore , it was shown that ftsz is not able to form the z - ring which is considered responsible for the mechanical forces required for the cell division .
could ftsz be responsible for the sharp transition in the growth and the cell division observed in our experiments ? or a set of other processes , including the formation of z - ring by ftsz , lead to the observed transition ?
is the cell elongation phenomenon due to only the depolymerization of ftsz at high pressures or more than one cellular processes are responsible for it
? the answers to all these questions can only come from further experiments that we are performing . on the switching probability @xmath62 .
model data is shown in solid black curve .
we have also plotted the values of @xmath88 extracted from experimental distribution of cell length at @xmath33 c for pressures
@xmath94 and @xmath58 atm as solid red circles.,width=302 ]
we have investigated the growth of _
e. coli _ in real time as a function of pressure and temperature .
we find that _ e. coli _ can grow and divide in a wide range of pressures ( @xmath2 atm ) and temperatures @xmath95 .
the division time of bacteria increases upon increasing pressure at a given temperature .
furthermore , division time at a constant temperature exhibits an exponential dependence on pressure for moderate values of pressure .
moreover , we find that for all the temperatures studied , division time shows an abrupt increase at pressures between @xmath5 atm . while at high @xmath1 this sharp increase in division time with pressure is very large where @xmath96 can be larger than @xmath22 min , at low @xmath1 , @xmath96 increases by few folds .
furthermore , we find that the division time shows an anomalous decrease and then increase with pressure at low temperature .
we hypothesize that this anomalous behavior of division time is a manifestation of the structural changes in phospholipids in the membrane .
further experiments on a veriety of cell types where the lipid composition is known would be required to answer this question .
we next looked at the bacterial cell morphology after application of pressure till the time of saturation in the cases where we could reach the saturation or few generation times where the saturation was hard to reach over the time scale of our experiments .
we find that average bacterial length increases upon pressure . while the bacterial elongation at high pressures @xcite is known , we find that e.coli
. shows a behavior of morphology very similar to growth rate or division time , whereas the average cell length also displays a sharp increase at pressures between @xmath97 and @xmath98 atm .
moreover , the heterogeneities in the cell length of bacteria increases upon increasing pressure . to explain the heterogeneities in the cell length with pressure
, we propose a simple stochastic irreversible switch model of bacterial phenotypes ( normal and filamenting ) .
we find that the model fits well the experimental data of distribution of bacterial cell length at different pressure .
moreover , the model allows us to extract the switching probability of e. coli .
bacteria to filamenting phenotype , which increases upon increasing pressure .
while the model captures the cell elongation phenomenon and explains the cell length distribution , it leaves us with many questions such as what biophysical processes give rise to the stochasticity in the phenotypic transitions as a function of pressure ?
a clue to this comes from the measured transition in the cell elongation observed here and depolymerization of ftsz protein responsible for cell division .
note that since the bacterial strain used in our experiments ( dh@xmath25 ) lacks the homologous recombination system , the cell elongation can no be interpreted as the conventional sos response of the system .
hence ftsz depolymerization and delocalization leading to non - formation of a z - ring is a potential biophysical process that may lead to phenotypic transitions proposed here .
further experiments are required where the polymerization of cytoskeletal proteins such as mreb and ftsz can be visualized along with cell division at various pressures and temperatures .
since growth is coupled to various other processes the bottlenecks could be either the structural integrity ( such as protein denaturation or membrane structural changes ) or the time integrity of various processes .
there is a large body of literature on the behavior of different biomolecules at varied physical conditions .
these studies indicate that at high pressures and temperatures the essential components that make up a cell may become unstable .
proteins can unfold and membranes can undergo structural transitions at high pressures leading to death of a cell @xcite .
the other issue which has been rather overlooked in past is the variation in the time scales of various processes . since pressure and temperature not only change the stability but they also modify the thermodynamic driving force of a chemical reaction and hence lead to changes in time scales of various processes . how the time integrity of various processes is maintained by a cell is an interesting question .
a better understanding can only come from a systematic study of the mutations in the protein / enzymes or regulatory circuits involved in various processes .
kato , c. , li l. , nogi , y. , nakamura , y , tamaoka , j. , and horikoshi , k. , extremely barophilic bacteria isolated from the mariana trench , challenger deep , at a depth of 11,000 meters._applied environmental microbiology _ * 64 * , 1510 - 1513 ( 1998 ) .
bakermans , c. , tsapin , a.i . , souza - egipsy , v. , gilichinsky , d.a . , and nealson , k.h . ,
reproduction and metabolism at @xmath99c of bacteria isolated from siberian permafrost .
_ environ .
microbiol . _
* 5 * , 321 - 326 ( 2003 ) .
oshima , t. , & imahori , k. , description of thermus thermophilus ( yoshida and oshima ) comb . nov .
, a nonsporulating thermophilic bacterium from a japanese thermal spa._int j syst bacteriol _ * 24 * , 102 - 112 ( 1974 ) . chilukuri , l.n . ,
bartlett , d.h . & fortes , p.a . ,
comparison of high pressure induced dissociation of single - stranded dna - binding protein ( ssb ) from high pressure - sensitive and high pressure - adapted marine shewanella specis .
_ extremophiles _ * 6 * 377 - 383 ( 2002 ) .
we thank a. buguin , y. t. maeda and j. merrin , and h. e. stanley for helpful discussions .
al acknowledges florence gould fellowship from the institute for advanced study , princeton , nj usa and nsf grant no .
phy-0848815 for support . | we have investigated the growth of escherichia coli _
e.coli_ , a mesophilic bacterium , as a function of pressure @xmath0 and temperature @xmath1 .
_ e.coli _ can grow and divide in a wide range of pressure ( @xmath2atm ) and temperature ( @xmath3c ) . for @xmath4 c , the division time of
_ e.coli
_ increases exponentially with pressure and exhibit a departure from exponential behavior at pressures between @xmath5 atm for all the temperatures studied in our experiments . for @xmath6 c
, the division time shows an anomalous dependence on pressure first decreases with increasing pressure and then increases upon further increase of pressure . the sharp change in division time
is followed by a sharp change in phenotypic transition of e. coli at high pressures where bacterial cells switch to an elongating cell type .
we propose a model that this phenotypic changes in bacteria at high pressures is an irreversible stochastic process whereas the switching probability to elongating cell type increases with increasing pressure .
the model fits well the experimental data .
we discuss our experimental results in the light of structural and thus functional changes in proteins and membranes .
center for studies in physics and biology , the rockefeller university , 1230 york avenue , new york , ny 10021 usa .
institue for advanced study , einstein drive , princeton , new jersey 08540 usa . |
correlation functions are essential to describe many - body systems . in particular , the first - order correlation function , or equivalently its fourier transform , the momentum distribution , is an important observable , since it witnesses various phenomena such as the bose - einstein condensation transition , the berezinskii - kosterlitz - thouless transition @xcite or the mott transition @xcite . in the case of one - dimensional ( 1d ) homogeneous gases , although one does not expect phase transitions , the correlation functions contain valuable information about the system . while the equation of state of 1d homogeneous bose gases has been extensively studied in various regimes @xcite and compared with exact theories @xcite , the momentum distribution has remained largely unexplored .
it was measured for quasi-1d quasi - condensates using bragg spectroscopy @xcite , and more recently it was investigated using the focusing technique for quasi-1d bose gases in the crossover from the ideal bose gas to quasi - condensate regimes @xcite .
density ripples that appear at the near field of a freely expanding quasi-1d quasi - condensate have also been studied @xcite . although related to the second - order correlation function , they also provide information about the first - order correlation function within the quasi - condensate theory .
the momentum distribution of an array of strongly correlated 1d bose gases @xcite constitutes the only measurement on a truly 1d system to our knowledge . from a theoretical point of view
, the momentum distribution of 1d bose gases with repulsive contact interactions is not known exactly over the entire phase diagram , but a few results have been established : the mean kinetic energy can be extracted from the exact yang - yang thermodynamics @xcite , and the short - range correlations are responsible for a momentum tail scaling as @xmath0 @xcite . the momentum distribution is known in the asymptotic regimes such as the ideal bose gas and the quasi - condensate regimes , and a classical field approximation has been proposed to account for the crossover between them @xcite . in this work
we investigate the momentum distribution of weakly interacting , purely or almost purely 1d bose gases in various regimes around the quasi - condensation crossover . on the theory side
, we find a criterion for the classical field theory @xcite to be accurate at the quasi - condensation crossover .
interactions are however too strong in our experiment to fulfill this criterion , and we implement exact quantum monte carlo ( qmc ) simulations for the equilibrium behavior of the lieb - liniger model .
we find that the measured momentum distribution agrees very well with qmc calculations .
moreover , the temperature extracted from a fit of the momentum distribution to the qmc calculations is in agreement with that obtained from _ in situ _ density fluctuation measurements @xcite .
this paper is organised as follows . in sec .
[ sec.theory ] we recall the classical field approximation and investigate the conditions under which it is valid to describe the quasi - condensation crossover .
we show that , in most experimental situations , the gas is too strongly interacting for this description to be accurate .
we then present our qmc calculations , and compare them with the classical field prediction in the asymptotic limit . in sec .
[ sec.technique ] , we describe the experimental setup and the focusing technique we use to measure momentum distribution .
we discuss how we improve the resolution of this technique by means of a guiding scheme . in sec .
[ sec.results ] , we present our experimental results that span both the quasi - condensate regime ( purely 1d ) and the degenerate ideal bose gas regime ( almost 1d ) , and compare them to qmc results .
we conclude and discuss the prospective work in sec .
[ sec.conclusions ] .
we consider a 1d homogeneous bose gas with repulsive contact interactions , described by the lieb - liniger hamiltonian , whose grand - canonical expression is @xmath1,\ ] ] where @xmath2 is the position , @xmath3 is the field operator , @xmath4 is the coupling constant , @xmath5 is the mass of one boson and @xmath6 is the chemical potential .
the thermal equilibrium state of the system is completely characterised by the density @xmath7 and the temperature @xmath8 , or by the dimensionless interaction parameter , @xmath9 , and the reduced temperature , @xmath10 .
the phase diagram @xcite in the parameter space ( @xmath11 , @xmath12 ) is given in fig .
[ fig.validity_c_field ] , where lines should be understood as crossovers . in this article , we focus on the quasi - condensation crossover , which occurs , for @xmath13 , around the line @xmath14 ( solid line ) .
the lieb - liniger model has the remarkable property of being integrable , giving access to exact results valid over the entire phase diagram .
for example , the equation of state is known exactly through the yang - yang theory @xcite . yet , the calculation of the first order correlation function @xmath15 from the bethe - ansatz solution for finite @xmath11 and at finite temperature is still a subject of active research .
approximate analytic theories for the correlation functions exist in the limiting cases of ideal bose gas regime ( where the ideal bose gas theory applies ) and quasi - condensate regime ( where the bogoliubov theory applies @xcite ) .
however , the behavior of @xmath16 at the crossover between those two regimes lacks a quantitative description . .
the ideal bose gas to quasi - condensation crossover occurs for @xmath17 ( solid line ) . within the ideal bose gas regime ( light gray area )
, the crossover from the classical to the degenerate sub - regime occurs for @xmath18 ( dashed line ) .
the region @xmath19 ( dark gray area ) is the fermionised regime .
the segments correspond to the regions explored by the two sets of experimental data ( see sec . [ sec.results ] ) .
note that the effect of quantum fluctuations on the @xmath20 function in the quasi - condensate regime are noticeable only for extremely low temperature @xcite , which is experimentally irrelevant . ]
the full quantum many - body problem of eq .
( [ eqn.ham ] ) is notoriously complex .
however , since thermal fluctuations are expected to dominate at the quasi - condensation crossover , it seems appropriate to simplify the problem using a classical field approximation , where the quantum field operators @xmath21 and @xmath22 are replaced by the complex fields @xmath23 and @xmath24 @xcite .
this strategy has been pursued in @xcite .
such an approach is particularly attractive for 1d homogeneous gases : contrary to higher dimensions , no ultraviolet divergence occurs when calculating @xmath16 , so that no energy cut - off is required for this particular quantity .
this problem can then be mapped onto a time - independent 2d schrdinger problem , describing a single particle evolving in imaginary time @xcite , thus leading to fast and easy - to - implement calculations .
the classical field model is parametrised by a single dimensionless variable @xmath25 , provided that lengths are rescaled by the correlation length @xmath26 @xcite .
simple analytic formulas are found in the two limits of the ideal bose gas regime ( @xmath27 ) and the quasi - condensate regime ( @xmath28 ) : the momentum distribution is lorentzian in both limits , with its full width at half maximum ( fwhm ) being @xmath29 in the ideal gas limit and @xmath30 in the quasi - condensate limit .
one recovers here the expressions derived from the highly degenerate ideal bose gas model and the high - temperature bogoliubov theory @xcite respectively .
results of numerical calculations spanning the crossover between those two regimes are reported in fig .
[ fig.compqmc ] , where one assumes a fixed value @xmath31 and parametrises the system with @xmath11 instead of @xmath32 .
we define the crossover region as the domain for which @xmath33 ( shown in fig .
[ fig.compqmc](a ) as the dashed line ) , differs by more than 10% from both asymptotic limits ( @xmath34 and 2@xmath34 ) , and we find that the crossover extends over about one order of magnitude in @xmath32 , or equivalently in @xmath11 at fixed @xmath12 .
looking at the shape of the momentum distribution , we recover the lorentzian distribution in both asymptotic regimes and , for any value of @xmath32 , we find momentum tails decreasing as @xmath35 , as expected from the analog single - particle problem @xcite .
the classical field theory is valid only if the population of the relevant modes is large . within the ideal bose gas regime , this requires @xmath36 , _
i.e. _ a highly degenerate gas
. this condition writes equivalently @xmath37 where @xmath38 is the interaction parameter at degeneracy .
thus , the classical field theory only correctly describes the quasi - condensation crossover ( which occurs at @xmath39 ) provided that @xmath40 .
the last condition translates into @xmath41 .
since crossovers span typically about one order of magnitude in @xmath11 , one requires that @xmath42 .
this value is very difficult to achieve experimentally on cold - atoms experiments while maintaining a temperature sufficiently low to ensure the 1d condition , @xmath43 , unless extremely weak atomic interactions are reached , using for instance a feshbach resonance @xcite . in the experiment presented here
, we have @xmath44 . in this case , according to the above argument , the classical field approach is expected to be inaccurate at the quasi - condensation crossover .
this can be seen in fig .
[ fig.compqmc](a ) , which shows the fwhm of the momentum distribution as a function of @xmath11 according to the ideal bose gas theory and the classical field prediction for @xmath31 .
the ideal bose gas theory is parametrised by @xmath45 , which quantifies the level of degeneracy : the fwhm of the momentum distribution within this theory goes from the maxwell - boltzmann prediction for @xmath46 to @xmath47 for @xmath48 , and this crossover spans more than one order of magnitude in @xmath49 , or equivalently in @xmath11 at fixed @xmath12 .
the classical field theory correctly describes the quasi - condensation crossover only if it shares a common plateau with the ideal bose gas theory at @xmath50 in the degenerate ideal bose gas regime , where @xmath51 . as seen in fig .
[ fig.compqmc](a ) , this is not the case at all for @xmath31 : the highly degenerate ideal bose gas regime is not very well identified for such a small value of @xmath12 as far as the momentum distribution is concerned , so that the classical field theory does not correctly describe the quasi - condensation crossover . for @xmath52 , one needs therefore a more accurate treatment than the above - mentioned classical field approximation .
we make use of qmc simulations to provide numerically exact results for the lieb - liniger model at finite temperature .
bosons in continuous space can be efficiently simulated making use of the density matrix renormalization group in one dimension @xcite and monte carlo techniques in arbitrary dimensions .
variational and diffusion monte carlo results for the first - order correlation function at @xmath53 are reported in ref .
@xcite . at finite temperature
, one can in principle make use of path - integral monte carlo simulations @xcite , although we are not aware of explicit calculations for the lieb - liniger model .
this latter technique requires the discretization of the imaginary - time dimension ( trotter approximation ) , and its influence on the simulation results has to be carefully removed . here
, we adopt a rather complementary approach , discretizing the spatial dimension instead @xcite . indeed , bosonic lattice models lend themselves to efficient qmc simulations free of any trotter approximation . by discretizing the field operator @xmath54 , where @xmath55 is the lattice spacing , @xmath56 , @xmath57 are bosonic operators at site @xmath58 , and @xmath59
, we obtain the bose - hubbard hamiltonian @xmath60,\end{aligned}\ ] ] with @xmath61 , and parameters @xmath62 , @xmath63 , @xmath64 .
we also include the presence of an external potential @xmath65 , discretized to give @xmath66 .
the lieb - liniger parameters @xmath11 and @xmath12 read @xmath67 where @xmath68 is the lattice filling .
we study the bose - hubbard approximation to the lieb - liniger hamiltonian by making use of stochastic series expansion ( sse ) qmc @xcite , extensively used to investigate lattice bosons .
the lieb - liniger model is correctly recovered when the lattice spacing is much smaller than the correlation length in the system . in the classical ideal bose gas regime
, this condition amounts to require @xmath69 . on the other hand , in the degenerate ideal bose gas regime and quasi - condensate regime ,
we need @xmath70 , implying @xmath71 and @xmath72 respectively .
the first - order correlation function and the momentum distribution are efficiently sampled with sse - qmc during the directed - loop update @xcite .
the lattice simulation reproduces faithfully the momentum distribution of the lieb - liniger model for momentum @xmath73 .
[ fig.compqmc ] shows qmc calculations for a homogeneous gas at @xmath31 , for values of @xmath11 that span the quasi - condensation crossover . in fig .
[ fig.compqmc](a ) the fwhm of the momentum distribution calculated with qmc is compared to that of the classical field and of the ideal bose gas models .
the qmc results follow the ideal bose gas prediction almost until the fwhm of the ideal bose gas prediction crosses that of the classical field prediction at @xmath74 , and then converges towards the classical field prediction .
the disagreement between qmc and the above theories never exceeds 20% .
these results mitigate the `` failure '' of the classical field approximation : as long as the fwhm is concerned , an approximate model where one uses the ideal bose gas prediction for @xmath75 and the classical field prediction for @xmath76 , would give the correct predictions within 20% error .
investigation of the full momentum distribution is shown in fig .
[ fig.compqmc](b ) for three different values of @xmath11 . on the ideal bose gas side of the quasi - condensation crossover , for @xmath77
, the qmc momentum distribution follows closely the ideal bose gas behaviour .
the classical field result is recovered for @xmath78 , while the momentum width is still about 10% above the bogoliubov prediction . for @xmath79 ,
we find that the tails of the momentum distribution agree well with the ideal bose gas prediction , while the width of the distribution is narrowed by about 20% .
finally , note that , although the highly degenerate ideal bose gas prediction @xmath80 has not been reached before the gas undergoes the quasi - condensation crossover , the momentum distribution at @xmath81 is much narrower than the maxwell - boltzmann prediction ( dash - dotted line in fig .
[ fig.compqmc](a ) ) , so that the effect of degeneracy within the ideal bose gas regime is already substantial .
while the previous results hold for homogeneous infinite systems , in the experiment described below , the gas is trapped longitudinally in a harmonic confinement of frequency @xmath82 . as long as the correlation length @xmath83 is much smaller than the cloud size @xmath84
, however , a local density approximation ( lda ) is valid : at a given position @xmath2 , the local value of an observable @xmath85 is that of a homogeneous gas ( @xmath86 ) at the chemical potential @xmath87 , and the global value for a trapped system ( @xmath88 ) can be obtained by adding the contributions of each slice of the gas , _
i.e. _ , @xmath89 therefore , within this approximation and for a given value of @xmath6 ( fixing the peak density ) , the normalized momentum distribution is independent of @xmath90 , while the total atom number scales as @xmath91 . for a gas whose peak linear density lies in the crossover from the degenerate ideal bose gas regime to the quasi - condensate regime
, the lda condition @xmath92 writes @xmath93 @xcite . at much higher densities ,
when almost the whole cloud lies within the quasi - condensate regime , the lda condition gives @xmath94 @xcite .
both conditions are fulfilled in the data shown below , so that the lda is expected to be valid .
using qmc , we compute the momentum distribution ( normalised to the total atom number ) of a harmonically trapped gas for different chemical potentials ( _ i.e. _ different peak linear densities ) and temperatures . to shorten the computational time ,
calculations are performed for trapping frequencies five times larger than the experimental values , while still satisfying the lda condition , as verified numerically .
the computed momentum distribution are multiplied by five to be compared to experimental data .
linear interpolation between the calculated distributions permits the calculation of the momentum distribution for any chemical potential @xmath6 and temperature @xmath8 , and enables us to perform fits to the measured distribution .
we check _ a posteriori _ that the interpolation is correct up to a few percent .
, @xmath95 , @xmath96 and @xmath97 realise the longitudinal trapping potential of frequency @xmath98 .
the transverse confinement of frequency @xmath99 is ensured by the three wires carrying a current modulated at @xmath100 khz .
( b , c ) focusing sequences . the focusing potential , harmonic to order @xmath101 in @xmath2 , is applied during @xmath102 before the longitudinal potential is turned off . after a free evolution during the focusing time @xmath103
, an absorption image is taken . without guiding ,
see ( b ) , the transverse confinement is removed during the whole focusing time , though a 1 ms ramp down to @xmath104 hz is carried out to decrease the transverse expansion . in the presence of guiding ,
see ( c ) , @xmath105 is kept at @xmath106 hz for a time @xmath107 , during which the cloud does not fall .
numerical values in ( b ) and ( c ) correspond to data of fig .
[ fig.results ] and [ fig.levitation ] respectively . ]
one - dimensional atomic clouds are realised in our experiment using an atom - chip setup @xcite .
more specifically , we load @xmath108 rb@xmath109 atoms in the modulated guide described in @xcite .
the longitudinal harmonic potential has an oscillation frequency of about @xmath110 hz and is obtained by dc currents running through wires @xmath96 and @xmath97 ( see fig .
[ fig.config_fils ] ) .
the transverse potential is provided by three wires carrying a @xmath100 khz sinusoidal current .
the transverse oscillation frequency can be varied from a few hundred hz to tens of khz by changing the current amplitude
. an important feature of this design is the independence between the longitudinal and transverse confinements .
after @xmath111 ms of radio - frequency evaporation , we let the cloud reach thermal equilibrium for @xmath112 ms , after which no remaining breathing can be observed . to characterise the atomic cloud , we perform _ in situ _ density fluctuations measurements , analysing the noise in an ensemble of absorption images as described in @xcite . to measure the momentum distribution , one can monitor the free expansion of the gas after all confining potentials have been turned off .
the rapid transverse expansion amounts to an effective instantaneous suppression of the interactions , so that the longitudinal expansion reflects faithfully the initial momentum distribution in the trap .
however , reaching the far field regime where the longitudinal density profile is homothetic to the momentum distribution requires unrealistically large field of view and long expansion time for our experimental parameters .
we thus use the so called focusing technique , already applied to 1d @xcite and 2d systems @xcite .
its principle is described below .
a strong longitudinal harmonic potential is applied for a very short time during which atoms do not have time to move but acquire a longitudinal momentum shift @xmath113 , proportional to their distance @xmath2 to the trap center .
then , the atomic interactions are suppressed by switching off the transverse confinement , and the longitudinal confinement is removed so that the cloud starts a free evolution . after a focusing time @xmath114 , the longitudinal density distribution @xmath115 is homothetic to the initial momentum distribution @xmath116 @xcite . more precisely , the time sequence we use is drawn in fig .
[ fig.config_fils ] .
the longitudinal focusing potential is realized by a four - wire configuration that cancels out the non harmonic terms up to the 6th order , thus minimizing aberrations .
it has an oscillation frequency @xmath117 hz and it is applied for about @xmath118 ms . after the focusing pulse
, we ramp the transverse frequency down to @xmath106 hz in @xmath119 ms before completely turning off the transverse trapping potentials and letting the cloud evolve freely .
the @xmath119 ms transverse ramp is used to limit the transverse velocity spread , so that the transverse final expansion of the cloud is reduced and the signal to noise ratio ( snr ) is increased .
we verified that this ramp is however quick enough to leave the longitudinal velocity distribution unchanged .
after free evolution during the focusing time @xmath120 ms , we take an absorption image of the cloud and extract the longitudinal density distribution , from which we deduce the initial momentum distribution .
the focusing time @xmath103 is adjusted by minimising the width of the density distribution of cold samples .
moreover , the width of very cold clouds saturates at a value that gives us the momentum resolution @xmath121 .
this resolution is significantly smaller than the width of the data in fig .
[ fig.results ] .
on the other hand , a better resolution may be necessary for a momentum distribution that is much narrower .
this can be achieved by a longer focusing time @xmath103
. however , the depth of field of our optical system limits the maximum free fall under gravity to about @xmath122 mm , corresponding to a free fall time of about @xmath123 ms .
nevertheless , larger focusing times can be obtained if a transverse potential is maintained during the focusing time , strong enough to overcome gravity ( which acts in the transverse plane in our experiment ) , but weak enough so that atomic interactions are negligible .
more precisely the guiding scheme we use is the following .
after having performed the longitudinal focusing pulse and switched off the longitudinal potential , we ramp the transverse frequency down in @xmath119 ms , similar to the procedure described before .
the final transverse frequency is @xmath124 hz , which is the lowest value that allows the atoms to remain transversally trapped .
we then hold the cloud for a guiding time @xmath107 , at the end of which we switch off the transverse confinement and let the cloud fall freely for @xmath125
. the guiding does not affect the longitudinal free evolution as long as the effect of interactions stays negligible .
we have checked that this is indeed the case for the explored parameter range by measuring the longitudinal density profile at focus @xmath115 of identical clouds , either without guiding and a focusing time @xmath126 ms , or with the above guiding scheme and @xmath127 ms , where @xmath128 ms and @xmath129 ms .
the functions @xmath130 , shown on fig .
[ fig.levitation ] for a cloud of @xmath131 atoms at @xmath132 nk , give indeed the same momentum distribution .
the validation of this technique paves the way towards high resolution measurements of momentum distribution of 1d gases .
fig . [ fig.results ] shows experimental results referring to two different parameter sets ( particle number , temperature and trapping potential ) . in both cases ,
the momentum distribution is obtained without the guiding scheme , with a focusing time of @xmath133 ms . data in fig .
[ fig.results](a ) correspond to a cloud initially trapped with a transverse oscillation frequency @xmath134 khz and a longitudinal frequency of @xmath135 hz , measured using parametric heating and dipole oscillations respectively .
since the cloud temperature and chemical potential are much smaller than the transverse level spacing , the system is well described by a 1d gas with a coupling constant @xmath136 , where @xmath137 nm is the 3d scattering length @xcite . in fig .
[ fig.results ] ( a,1 - 2 ) , we show the measured momentum distribution ( points ) with a qmc fit ( solid lines ) that yields @xmath138 nk , corresponding to @xmath139 .
note that the shape is reproduced up to a few per cent .
we report in fig .
[ fig.validity_c_field ] the segment @xmath140 $ ] corresponding to the domain explored by the data : @xmath141 corresponds to the peak density , and the segment length is such that 80% of the atoms lie in a region of density larger than that associated to @xmath142 .
we find that @xmath143 , where @xmath14 is the interaction parameter at the quasi - condensation crossover .
thus the cloud lies quite deeply in the quasi - condensate regime .
however , the measured momentum distribution is substantially broader than that predicted using the quasi - condensate lorentzian @xmath144 together with the lda , shown as the dashed line in fig .
[ fig.results](a,1 ) . on the other hand , for such a value of @xmath141 , the momentum distribution shows slowly decreasing tails , compatible with a lorentzian behavior within a few percent in a wide range of momenta .
this is seen in fig .
[ fig.results](a,2 ) where the dashed - dotted line shows a @xmath35 decrease . for our snr of about 80 , data are compatible with a lorentzian behavior at large @xmath145 .
as shown in @xcite , an independent and reliable thermometry can be carried out using _ in situ _ density fluctuations , provided the pixel size @xmath146 is both much larger than the correlation length of density fluctuations , and much smaller than the typical length scale of the variation of the mean density profile : the atom number fluctuations are then given by the thermodynamic relation @xmath147 , where @xmath148 is the equation of state of homogeneous 1d gases , known exactly using the yang - yang calculation @xcite , and the derivative is computed for @xmath6 such that @xmath149 is the local density .
the above criterion is well fulfilled in our experiment and the measured density fluctuations are shown in fig .
[ fig.results](a,3 ) . for linear densities
larger than 25 atoms per pixel , fluctuations are almost independent of the density , which confirms that the center part of the gas lies within the quasi - condensate regime . indeed , in this regime @xmath150 , so that the thermodynamic relation reduces to @xmath151 , which does not depend on @xmath152 .
the fluctuations computed using the thermodynamic relation and the yang - yang equation of state , for the temperature @xmath138 nk ( obtained by fitting the momentum distribution with the qmc results ) are shown as solid line .
the data are in agreement with this prediction . with our snr , the density fluctuations based thermometry has a precision of about 20% ( see gray curves in fig .
[ fig.results](a,3 ) ) , less precise than the thermometry obtained by fitting qmc calculations to the momentum distribution .
finally , we compare in fig .
[ fig.results](a,4 ) the measured _ in situ _ density profile to the qmc density profile ( within lda ) for @xmath138 nk ( dashed line ) .
we find a good agreement for most of the profile , although the measured data show higher wings .
since the central part of the cloud lies deep in the quasi - condensate regime , a large part of the cloud follows the thomas - fermi profile , and all the information on the temperature lies in the _ small _ wings .
this renders yang - yang thermometry based on density profile less precise and extremely sensitive to the shape of the wings . here
, a yang - yang fit to the density profile gives @xmath153 nk , a value 40% higher than that extracted from the qmc fit , incompatible with the measured momentum distribution or the density fluctuation measurements .
this discrepancy , and thus the presence of the inflated wings , may come from the anharmonicity of the potential due to its residual roughness @xcite .
alternatively , it may also indicate a lack of perfect thermal equilibrium .
while the previous results probe mainly the quasi - condensate regime , we also probe the ideal bose gas side of the quasi - condensation crossover , _
i.e. _ data with @xmath154 .
the ideal bose gas regime shows a very different behaviour from the trivial maxwell - boltzmann prediction only for a large @xmath12 parameter , for which the quasi - condensation crossover occurs for an already highly degenerate gas .
if one wishes to preserve the 1d condition @xmath155 , large @xmath12 parameters can be accessed only by decreasing the transverse confinement @xcite .
we thus reduced the transverse confinement to 2.1 khz .
data are shown in fig .
[ fig.results](b ) .
no saturation of the density fluctuations is seen on the _ in situ _ fluctuation measurements , which indicates that the gas does not lie in the quasi - condensate regime .
fluctuations however rise well above the poissonian level ( shown as dashed - dotted line on fig . [ fig.results](b,3 ) ) so that the gas is highly degenerate .
contrary to the data of fig .
[ fig.results](a ) , we now have @xmath156 which is of the order of unity , so that the transverse excited states contribute to the measured fluctuations and momentum distribution .
we take into account the population of transverse excited states , assuming that they behave as independent ideal 1d bose gases , while the transverse ground state is treated as a lieb - liniger gas .
this modified lieb - liniger model has been used with success to describe density profiles @xcite and density fluctuations @xcite , and has been applied to predict the rms width of momentum distribution @xcite .
here we use our qmc calculations to describe the transverse ground state .
a fit using the above strategy reproduces well the measured momentum distribution ( see fig .
[ fig.results](b,1 ) ) , and yields the temperature @xmath157 nk , corresponding to @xmath158 .
the segment in the phase space @xmath159 explored by the data is shown in fig .
[ fig.validity_c_field ] : the peak linear density is close to the quasi - condensation crossover density and most of the cloud lies in the degenerate ideal bose gas regime . the contribution of the excited states to the momentum distribution is seen in fig .
[ fig.results](b,1 ) : it is only 10% in the center but it rises to almost 50% in the wings around @xmath160 . fig .
[ fig.results](b,2 ) shows that the theoretical momentum distribution decreases faster than @xmath35 at large momenta .
this is mainly due to the contribution of the excited states , which have approximately gaussian momentum tails .
however , with our snr of about 50 , no deviation from a @xmath35 behavior can be identified .
we report in fig .
[ fig.results](b,2 ) the fluctuations expected for the temperature @xmath161 nk ( obtained from the fit of the momentum distribution with qmc ) .
they are in agreement with the measured fluctuations .
note however that , for these parameters , the uncertainty of thermometry based on fluctuations is about 30% ( see gray lines in fig .
[ fig.results](b,3 ) ) . the density profile expected for @xmath161
nk is not far from the measured one ( see fig .
[ fig.results](b,4 ) ) .
in contrast to case ( a ) , yang - yang thermometry based on the profile is less sensitive to tiny modifications of the wings , since the profile of the high - density regions is also affected by the temperature .
thus , a thermometry based on the profile is expected to be quite precise and the potential roughness is expected to have a smaller effect .
a yang - yang fit to the experimental profile yields a temperature of @xmath162 nk , different only by 10% from the temperature deduced from the momentum distribution .
we have shown that for @xmath52 , a simple 1d classical field approximation where the momentum distribution is computed without energy cutoff , fails to describe the momentum distribution of weakly interacting 1d bose gases in the crossover from ideal bose gas to quasi - condensate .
thus , we performed qmc calculations to investigate the crossover .
experimentally , we measure the momentum distribution with the focusing technique , which is improved by a guiding scheme for enhanced resolution . we show that the temperature deduced from the momentum distribution is in agreement with an independent fluctuation - based thermometry .
we find that the result of bogoliubov theory is not appropriate for parameters that are close to the ones relevant to this article , even when the cloud lies quite deeply in the quasi - condensate regime . in @xcite
, it has been proposed to use the mean kinetic energy deduced from measured momentum distribution in order to extract the temperature using the modified yang - yang calculation . for the data presented in this article , with a snr of about 50 ,
the measured momentum distribution shows tails compatible with a @xmath35 behavior both in the quasi - condensate regime and in the degenerate ideal bose gas regime .
therefore , extraction of the mean kinetic energy from the data is impossible .
we argue that this technique was applicable in @xcite because of the more 3d nature of the gas : for quasi 1d gases , as discussed in section [ sec.resultsb ] , the contribution of the transverse excited states leads to fast decaying tails , which enable the extraction of the mean kinetic energy per particle .
we also remark that neither our measurements nor the qmc calculations display the asymptotic behaviour of @xmath0 , which only appears at even larger momenta for our parameters . on the other hand ,
we believe that this behaviour could be more readily discernable in the strongly interacting regime , where the @xmath0 tail should contain a larger proportion of the atoms .
this work opens many perspectives for the study of 1d gases .
first of all , we show that momentum distribution measurements provide a precise thermometry for 1d bose gases .
while in this article we use qmc calculations to fit the temperature of experimental data , a rough estimate of the cloud temperature could be performed with a lower numerical cost .
indeed , the comparison with the exact qmc momentum distribution shows that a model which combines the ideal bose gas theory and the classical field approximation gives the correct width within 20% precision .
second , momentum distribution measurements are essential to characterize more complex systems .
for instance , in the presence of a lattice , it enables the investigation of the correlation properties at the mott and/or pinning transitions .
contrary to bragg spectroscopy @xcite , where a single momentum component is probed at each shot , the focusing method as well as time of flight method gives access to the whole distribution at the same time , which allows for noise correlation measurements in momentum space @xcite .
finally , recording the time evolution of the momentum distribution is essential to monitor the non - equilibrium dynamics and address the issue of thermalization in 1d closed quantum systems .
for example , the dynamics resulting from a quench of the 1d coupling constant can be investigated .
the measure of the momentum distribution of impurities would also permit the study of impurity dynamics , which is currently under intense investigation @xcite .
the authors thank karen kheruntsyan for providing the data from the yang - yang calculations .
this work was supported financially by the triangle de la physique , by the anr grant anr-09-nano-039 - 04 , the austro - french fwr - anr project i607 , the arc discovery project grant no .
dp110101047 , and the coqus graduate school of the fwf .
quantum fluctuations lead to an algebraic decay of @xmath20 as @xmath163 for distances @xmath164 , where @xmath165 is the healing length . for
the decay to be noticeable over this range , one requires @xmath166 . in terms of the analog quantum problem , @xmath167 , where the sum is carried out over the eigenstates of energy @xmath168 , and the factors @xmath169
are obtained by wavefunction overlaps .
the fourier transform is a sum of lorentzian , which shows @xmath35 tails ( uniform convergence is ensured by the fast decay of the @xmath169 s at large energy ) .
s. tung , g. lamporesi , d. lobser , l. xia , and e. a. cornell , phys .
lett . * 105 * , 230408 ( 2010 ) , in this experiment , a somehow different technique is employed , where a harmonic potential is kept during the focusing .
deviations from the harmonic potential by about 30 nk is sufficient to explain the presence of those wings .
this value is consistent with the residual roughness we estimated from the measured density variations of very elongated quasi - condensates confined in a longitudinal quartic potential . | we investigate the momentum distribution of weakly interacting 1d bose gases at thermal equilibrium both experimentally and theoretically .
momentum distribution of single 1d bose gases is measured using a focusing technique , whose resolution we improve via a guiding scheme .
the momentum distribution compares very well with quantum monte carlo calculations for the lieb - liniger model at finite temperature , allowing for an accurate thermometry of the gas that agrees with ( and improves upon ) the thermometry based on _ in situ _ density fluctuation measurements .
the quasi - condensation crossover is investigated via two different experimental parameter sets , corresponding to the two different sides of the crossover .
classical field theory is expected to correctly describe the quasi - condensation crossover of weakly interacting gases .
we derive the condition of validity of the classical field theory , and find that , in typical experiments , interactions are too strong for this theory to be accurate .
this is confirmed by a comparison between the classical field predictions and the numerically exact quantum monte carlo calculations . |
and grating and ccd observations of the cores of clusters previously believed to be harboring cooling flows find very little gas cooling below @xmath8 kev ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
this major discovery has led to the widely discussed picture where feedback on the hot intracluster medium ( icm ) from an agn in the central cluster galaxy is responsible for stifling the cooling flow , though the details of the heating process remain elusive ( e.g. , for reviews see * ? ? ?
* ; * ? ? ?
about 70% of clusters previously classified as cooling flows possess radio emission from the central galaxy ( e.g. , * ? ? ?
observations indicate that such clusters also have x - ray cavities and filamentary structures in the x - ray images that are related to the radio emission ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
what about clusters indicated to be cooling flows from previous low - resolution x - ray observations that do not display radio properties typical of cooling flows ? we have previously analyzed observations of two such systems : a2029 @xcite and a2589 @xcite .
a2589 has no observed radio emission associated with its cd galaxy ; and , although the cd in a2029 does possess a wide - angle - tail ( wat ) radio source , such sources are generally not found in cooling flows @xcite , though their morphology provides evidence for relative motions between the radio source and icm @xcite .
both of these clusters have high density cores with temperatures significantly lower than at larger radii .
( although the relatively low - quality data of a2589 are consistent with an isothermal gas , a new observation clearly indicates that the temperature profile increases with radius zappacosta et al .
2005 , in preparation . )
these clusters also display no spectroscopic evidence for cooling flows and have unusually symmetric x - ray images without the pronounced surface brightness irregularities observed in the cores of `` radio loud '' cool - core clusters ( e.g. , * ? ? ?
finally , cool - core clusters also appear to preferentially display central metallicity enhancements , while flat metallicity profiles are characteristic of non - cooling flow , generally disturbed , systems ( e.g. , * ? ? ?
* ; * ? ? ?
a2589 and a2029 both have pronounced central metallicity enhancements further attesting to their comparatively relaxed states .
observations suggest that cooling flows are destroyed by large turbulent motions in the icm since major merging activity in clusters anti - correlates with cooling mass - flow rate @xcite , a conclusion supported by theoretical studies ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
likewise , theoretical studies ( e.g. , * ? ? ?
* ; * ? ? ?
* ) suggest that wat radio morphology may also be a signature of hot shocks in on - going cluster mergers . since the x - ray images of a2029 and a2589 are very regular all the way into their cores ( though see [ conc ] ) , and yet there is no evidence for icm cooling or radio emission typical of cool - core clusters , in the context of the agn feedback paradigm we are viewing these clusters at a special time after they have settled down from past mergers but very soon before the onset of agn heating .
here we examine a observation of a644 , another `` radio quiet '' cluster that was previously classified as a cooling flow based on analysis of both x - ray imaging and spectral data @xcite .
a644 is the only cooling flow cluster with large inferred @xmath9 ( @xmath10 100 @xmath11 yr@xmath12 ) , which was not detected in the radio study of ( * ? ? ?
* @xmath13 3 x10@xmath14 w hz@xmath12 ) .
we wish to examine how its x - ray properties compare to the other cooling flow clusters with atypical radio properties and , in particular , whether it demonstrates a clearer connection between past merger activity and the suppression of a cooling flow .
the redshift of a644 ( @xmath15 ) corresponds to an angular diameter distance of 277 mpc and @xmath16 kpc assuming @xmath17 , @xmath18 , and @xmath19 .
the paper is organized as follows . in [ obs ]
we present the observation and discuss the data reduction .
the analysis of the image and spectral data of the acis - i are presented in
[ image ] and [ spec ] respectively .
we calculate the gravitating matter distribution in [ mass ] . finally , in [ conc ] we present our conclusions .
a644 was observed with the acis - i ccd array for approximately 30 ks during ao-2 as part of the guest observer program .
we reduced the data using the ciao v3.1 ( with caldb v2.28 ) and heasoft v5.3.1 software packages .
we followed the standard ciao threads and reprocessed the level-1 events data to make use of the latest calibration information , including correcting for charge - transfer inefficiency and a time - dependent gain shift . from regions of least source contamination of the ccds
we extracted a light - curve ( 5.0 - 10.0 kev ) to identify periods of high background .
the observation was remarkably quiescent , and after removing a small period of modestly increased background rate , the total exposure used for subsequent analysis was 28.9 ks .
since the cluster emission fills the entire acis - i field , we adopted for our default analysis the standard background events files provided in the caldb collected from suitable blank fields .
for comparison to results obtained using these standard background templates , we also modeled the background directly . using regions as far away from the cluster center as possible
, we extracted a spectrum ( with point sources masked out ) and fitted it with a combined source and background model ( e.g. , * ? ? ? * ) .
for the source , we adopted a thin thermal plasma component ( ) with @xmath20 kev and half - solar metallicity .
the background was modeled with components for the cosmic x - ray background ( cxb ) and instrumental background : two soft thermal plasmas and a hard power law for the cxb ; a broken power law and two gaussians for the instrumental background .
comparison of results using the alternative background is presented in [ spec ] .
in figure [ fig.image ] we display the acis - i image of a644 in the 0.3 - 7.0 kev band .
the image was corrected for exposure variations using an exposure map created for a monochromatic energy 1 kev , approximately the counts - weighted average energy of the spectrum . for presentation purposes
only , the image was also adaptively smoothed using the ciao task csmooth with default parameter settings .
the diffuse emission is seen to fill the entire acis - i field .
the morphology of most of the cluster is quite regular and approximately elliptical in shape
. visual inspection of the central @xmath21 region of the ( smoothed ) x - ray surface brightness reveals some irregular , non - azimuthally symmetric features ( figure [ fig.image ] ) .
these deviations from a symmetric configuration , though statistically significant , do not represent large fluctuations ( @xmath22 ) in the mean surface brightness level . in particular , the data do not provide clear evidence for distinct subclumps in the core region
. we also searched for azimuthal fluctuations in the core by constructing a hardness ratio map ( and also through spectral fitting in sectors , see [ spec ] ) .
we did not find any corresponding azimuthal fluctuations either in the core or on larger scales .
instead , we found the spectrally softer core gives way at larger radius to a spectrally harder surrounding medium .
this radial variation is consistent with the radially increasing temperature profile ( see [ spec ] ) .
we measure the centroid and the ellipticity of the x - ray surface brightness using the moment method described by @xcite and implemented in our previous x - ray studies of galaxies and clusters ( e.g. , * ? ? ?
this iterative method is equivalent to computing the ( two - dimensional ) principal moments of inertia within an elliptical region .
the ellipticity is defined by the square root of the ratio of the principal moments , and the position angle is defined by the orientation of the larger principal moment .
following our previous study of the ellipticity of the data of ngc 720 @xcite we removed point sources and replaced them with smoothly distributed diffuse emission using the ciao task dmfilth .
the centroid of the x - ray surface brightness shifts significantly ( e.g. , @xmath23 kpc between 15 - 500 kpc ) , mostly directed toward the south . at the smallest radii
the centroid is located within the region of image irregularities in the core mentioned above . outside of the core region
the x - ray isophotes are very regular and moderately elliptical .
the ellipticity gently falls from @xmath24 at @xmath25 kpc to @xmath26 at @xmath27 kpc , where @xmath28 is the semi - major axis of the elliptical aperture within which the moments are calculated . over
this radius range the position angle ( measured n through e ) of the elliptical apertures is quite steady , varying between a maximum of @xmath29 degrees ( at @xmath25 kpc ) and a minimum of @xmath30 degrees ( at @xmath31 kpc ) .
overall , the image properties suggest a relaxed cluster outside about @xmath32 kpc but a disturbed cluster within @xmath33 kpc .
we examined the radial surface brightness profile ( with point sources masked out ) located about the x - ray peak and also about the centroid computed within @xmath27 kpc . in both cases the surface brightness profile is fitted fairly well out to @xmath34 kpc by a @xmath35 model : @xmath36 kpc , @xmath37 , @xmath38 , 129 dof ( peak ) and @xmath39 kpc , @xmath40 , @xmath41 , 128 dof ( centroid ) .
the largest fit residuals ( @xmath42 ) lie within the innermost region ( @xmath43 ) and are modest ( @xmath44 ) elsewhere .
we note that the parameters @xmath45 and @xmath35 increase systematically when we include data from increasingly larger radii or exclude data from the very central region .
both of these issues likely account for the different values obtained from previous studies which included data at larger radii ; e.g. , using the pspc @xcite obtain @xmath46 kpc and @xmath47 .
c|ll|ll|c|c|c|c + & & & @xmath48 & @xmath49 & @xmath50 + annulus & ( arcsec ) & ( kpc ) & ( arcsec ) & ( kpc ) & ( kev ) & ( solar ) & ( 10@xmath51 @xmath52 ) & @xmath53/dof ) + + 1 & 0.0 & 0.0 & 29.3 & 39.3 & @xmath54 & @xmath55 & @xmath56 & @xmath57 + 2 & 29.3 & 39.3 & 43.5 & 58.5 & @xmath58 & @xmath59 & @xmath60 & @xmath61 + 3 & 43.5 & 58.5 & 56.6 & 76.0 & @xmath62 & @xmath63 & @xmath64 & @xmath65 + 4 & 56.6 & 76.0 & 69.1 & 92.9 & @xmath66 & @xmath67 & @xmath68 & @xmath69 + 5 & 69.1 & 92.9 & 95.9 & 128.9 & @xmath70 & @xmath71 & @xmath72 & @xmath73 + 6 & 95.9 & 128.9 & 127.2 & 170.9 & @xmath74 & @xmath75 & @xmath76 & @xmath77 + 7 & 127.2 & 170.9 & 180.8 & 242.9 & @xmath78 & @xmath79 & @xmath80 & @xmath81 + 8 & 180.8 & 242.9 & 248.2 & 333.5 & @xmath82 & @xmath83 & @xmath84 & @xmath85 + 9 & 248.2 & 333.5 & 364.1 & 489.2 & @xmath86 & @xmath87 & @xmath88 & @xmath89 + + we focus our attention on the spectral properties of the hot gas obtained within a series of concentric circular annuli centered either on the x - ray centroid or peak .
below we also briefly discuss attempts to constrain azimuthal variations in the spectral properties .
the widths of the circular annuli were chosen so that the temperatures were determined to similar precision in each radial bin . for the centroid case we list the annuli in table [ tab.spec ] .
annuli 1 - 4 contain @xmath90 background - subtracted counts while bins 5 - 9 contain twice that amount in the 1.5 - 7 kev band .
we restrict our analysis to energies above 1.5 kev because of uncertainties arising from excess absorption above the galactic value discussed below , and because the gas temperature lies well above 1.5 kev within all annuli . for each annulus
counts - weighted response matrices were generated using the ciao tasks mkwrmf and mkarf .
we fitted the background - subtracted spectrum with an thermal plasma modified by galactic absorption ( @xmath91 , @xcite ) to each annulus .
the free parameters are temperature , iron abundance , and normalization ( emission measure ) .
( the fitted abundance is technically a metallicity since we fit the iron abundance as a free parameter and the abundances of all the other elements tied to iron in their solar ratios .
however , iron dominates the metallicity . )
the spectral fitting was performed with ( 11.3.1i , * ? ? ?
* ) using the @xmath92 minimization method .
hence , we rebinned all spectral channels to have a minimum of 20 counts per energy bin ( necessary for the validity of the @xmath92 method ) and a signal - to - noise ratio of at least 3 .
the solar abundances in are taken to be those given by @xcite which use the correct `` new '' photospheric value for iron which agrees also with the value obtained from solar - system meteorites ( e.g. , * ? ? ?
* ) . to estimate the statistical errors on the fitted parameters we simulated spectra for each annulus using the best - fitting models and fit the simulated spectra in exactly the same manner as done for the actual data . from 100 monte carlo simulations
we compute the standard deviation for each free parameter which we quote as the `` @xmath93 '' error .
( we note that these @xmath93 error estimates generally agree very well with those obtained using the standard @xmath94 approach in . )
all quoted errors are @xmath93 unless stated otherwise .
the radial profiles of temperature and iron abundance are shown in figure [ fig.tfe ] ; table [ tab.spec ] gives the parameters obtained from the spectral fits for the centroid case ; figure [ fig.spec ] shows the acis - i spectra of two representative annuli and the associated best - fitting one - temperature model .
the simple one - temperature model provides a good description of the spectral data in all annuli . for @xmath95
kpc the spectral fits for both the centroid and peak case agree very well . over this region
the temperature is consistent with a constant value @xmath96 kev as found also by previous studies of a644 with ( e.g. , * ? ? ?
* ; * ? ? ?
in contrast to these previous studies , the data clearly reveal that the iron abundance declines with radius from a value of @xmath97 at @xmath98 kpc to @xmath99 at @xmath100 kpc .
previous studies of a644 were consistent with a constant iron abundance profile ( e.g. , * ? ? ?
* ; * ? ? ?
* ) within the large errors .
we note also that the average iron abundance from ( @xmath101 , scaled to our solar reference ) obtained by @xcite agrees well with our results . for @xmath95
kpc the declining radial profile of iron abundance , approximately isothermal gas profile , and regular image morphology are very characteristic of relaxed `` cool core '' clusters ( e.g. , * ? ? ?
* ; * ? ? ?
inspection of figure [ fig.tfe ] reveals that the region @xmath102 kpc is more complex .
the iron abundance profile is approximately constant for the centroid case , with marginal evidence for a central dip . in the peak case ,
the profile declines and then appears to increase in the central bin : the value in the central bin ( @xmath103 solar ) is inconsistent with that in the adjacent bin at the @xmath104 level .
the temperature profile for @xmath102 kpc does not correspond to a typical cool - core cluster . in the centroid case , the temperature is consistent with isothermal . in the peak case ,
the temperature declines inward to @xmath105 kpc like a cool - core cluster but then rises significantly in the central bin ( @xmath106 kev ) .
the 99% confidence lower limit on the temperature estimated using @xmath107 is @xmath108 kev , still larger than the minimum value of @xmath109 kev reached near @xmath110 kpc .
a power law fit to the temperature profile of the inner three radial bins is inconsistent with a constant value at the @xmath6 level and is highly inconsistent with the declining temperature central profiles characteristic of cool core clusters . despite the good quality of the fits
, we examined whether they could be improved further . allowing abundances other than iron to vary
separately did not yield noticeable improvement .
adding another temperature component also had no effect on the fits , even when energies down to 0.5 kev were included .
consequently , we found no need for an additional multiphase cooling flow component in the spectra .
previous studies were divided between indicating a massive cooling flow ( @xmath111 , * ? ? ? * ) and no cooling flow ( @xmath112 , * ? ? ?
* ) . using a standard model of a gas cooling at constant pressure as implemented in our previous studies ( e.g. , * ? ?
* ) , our results indicate a best - fitting cooling rate of zero with 90% upper limits of @xmath113 in all annuli except annulus 1 ( referring to the centroid case ) , where @xmath114 ( 90% confidence ) .
( we note that sometimes evidence for a cooling flow and multiphase gas in clusters has been claimed based on the analysis of ccd spectra extracted from a large aperture covering , e.g. , the entire putative cooling flow region ( e.g. , * ? ? ?
however , the distribution of temperatures produced by the radial variation of a single - phase medium within the putative cooling flow aperture , as well as that contributed by the projection of higher temperature gas outside the aperture if the data are not deprojected , generally can not be distinguished from a two - temperature medium or a cooling flow ( e.g. , * ? ?
* ; * ? ? ?
hence , we have explored the need for multiphase gas using the narrowest annuli allowed by the data quality . )
although the hardness ratio map displayed no significant azimuthal variations ( [ image ] ) , we searched for azimuthal spectral variation by dividing up the annuli into four equal sectors we rebinned the radial bins by a factor of two to achieve enough signal for interesting constraints .
when fitting a single component to each sector we found no evidence for azimuthal spectral variations , consistent with the findings from the hardness ratio map .
the temperatures and iron abundances in each sector within an annulus were found to be consistent within their @xmath115 confidence limits .
in particular , we do not find the @xmath116 kev temperature spike in a region a few arcminutes to the east of the center reported by @xcite from .
( the orientations of our sectors were oriented approximately to match those of sectors 2 - 5 of @xcite . ) finally , we have studied the sensitivity of the results to various sources of systematic error which we summarize below : _ deprojection : _ we performed a non - parametric deprojection using the `` onion - peeling '' method as implemented in our previous studies ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
that is , one begins by determining the spectral model in the bounding annulus and then works inward by subtracting off the spectral contributions from the outer annuli .
this technique tends to introduce fluctuations between parameters in adjacent radial bins .
we found that to obtain results of comparable quality to the projected analysis it was necessary either to fix the temperature at a constant value near 8 kev or restrict the radial variation in the the iron abundance so that its logarithmic slope was @xmath117 . after imposing either of these restrictions
we obtain results consistent with those reported above for the projected analysis both in terms of fit quality and parameter values .
( we note that these results for the temperatures and iron abundances obtained using our implementation of the `` onion - peeling '' method are very consistent with those obtained using the deprojection scheme provided in using the projct routine and require the same restrictions on the temperatures and iron abundances noted above . )
_ galactic column density and bandwidth : _ when energies below @xmath118 kev are included in the fits then a model assuming the nominal foreground galactic column density ( @xmath119 ) overestimates the flux at low energies , especially near 0.5 - 0.6 kev .
if is allowed to vary the fits are improved substantially within each annulus ; e.g. , in annulus 4 @xmath120 for 228 dof .
the best - fitting column densities are @xmath121 about twice the galactic value .
since the excess column densities apparently do not vary with radius it is likely that the excess is associated with an underestimate of the galactic column .
( using the background models described below we verified that the inferred excess column density is not the result of over - subtracting the soft background using the standard templates of nominally blank fields . )
our result is similar to that reported by an study of a478 @xcite .
compared to the results presented above for energies above 1.5 kev , the temperature and iron abundance are mostly unaffected when fitting down to 0.5 kev with a column density about twice the galactic value .
the most significant change is a systematic decrease of @xmath122 kev in the temperature values in all annuli , but the shape of the radial temperature profile is preserved .
although we expect the excess column density is of galactic origin we preferred to confine our analysis to energies above 1.5 kev to avoid sensitivity to the precise value .
_ background : _ as described in [ obs ] , for comparison to results obtained using the standard blank background fields we also constructed a model background by fitting the total ( source plus background ) spectrum of regions of the acis - i as far away from the center of a644 as possible .
we found that results obtained using this modeled background agreed extremely well with those of the standard blank fields .
as expected , by far the largest differences were observed in annulus 9 ; e.g. , @xmath123 kev using the modeled background compared to @xmath124 kev for the standard blank fields .
_ plasma code : _ we investigated the sensitivity of our results to the plasma code using the model .
the quality of the fits and the temperature values were found to be very consistent within the @xmath93 errors .
the iron abundance values also were consistent with those obtained using the code , though the values obtained with the code were generally smaller by 10%-20% .
these results are consistent with previous studies @xcite . _
calibration : _ we explored the sensitivity of our results to the version of the calibration using previous versions of ciao ( v3.0.2 ) and the caldb ( v2.26 ) .
we found the results on the fitted spectral parameters to be consistent with those quoted above within the estimated @xmath125 statistical errors .
lcccccccc + + + + + & & @xmath126 & & @xmath127 & @xmath128 & & @xmath129 + model & @xmath53/dof ) & ( kpc ) & @xmath35 & ( @xmath130 g cm@xmath51 ) & ( kpc ) & @xmath132 & ( @xmath130 g cm@xmath51 ) + + + @xmath133 & 12.1/3 & @xmath134 & @xmath135 & @xmath136 + @xmath137 & 4.3/0 & @xmath138 & @xmath139 & @xmath140 & @xmath141 & @xmath142 & @xmath143 + + + + + + + & & @xmath144 & + model & @xmath53/dof ) & ( kev ) & @xmath145 + power law & 2.7/4 & @xmath146 & @xmath147 + + cccccccc + & best & @xmath148statistical & @xmath148background & @xmath148plasma & @xmath148bandwidth & @xmath148calibration & @xmath149 + + @xmath150 & 6.1 & @xmath151 & @xmath152 & @xmath153 & @xmath152 & @xmath154 & @xmath155 + ( mpc ) & 2.9 & @xmath156 & @xmath157 & @xmath158 & @xmath159 & @xmath160 & @xmath160 + + + to determine the gravitating mass distribution from x - ray observations requires the hot gas to be in hydrostatic equilibrium . both n - body simulations and gravitational lensing studies confirm the reliability of x - ray mass measurements , particularly for clusters with regular x - ray image morphologies ( e.g. , for a review see * ? ? ?
the assumption of hydrostatic equilibrium should be valid outside the central @xmath161 kpc region of a644 .
outside the core region the derived radial variations in the density and temperature of the hot gas agree for both the cases where the cluster center is defined as the surface brightness peak or centroid .
consequently , we exclude the central three annuli ( defined in table [ tab.spec ] ) from our default analysis , though for comparison we summarize results obtained using all annuli .
the approach we use to calculate the gravitating mass distribution follows closely our previous studies @xcite . as is standard we assume spherical symmetry to provide a spherically averaged mass profile appropriate for comparison to other observations and to the spherically averaged mass profiles obtained from cosmological simulations . in the equation of hydrostatic equilibrium
we evaluate the derivatives of the three - dimensional gas density ( ) and temperature ( @xmath48 ) using simple parameterized models as discussed below . for our default analysis we projected parameterized models of the three - dimensional quantities , and @xmath48 , and fitted these projected models to the results obtained from our analysis of the data projected on the sky ( table [ tab.spec ] ) .
in this manner we obtained good constraints on the three - dimensional radial profiles of and @xmath48 .
although our analysis closely follows our previous studies , we note two minor improvements .
first , when fitting models to the radial profiles of the gas density and temperature to the data binned in annuli on the sky ( as given in table [ tab.spec ] ) we integrated the models over each radial bin ( rather than only evaluating at a single point within the bin ) to provide a consistent comparison .
second , when projecting models of the gas density and temperature along the line of sight we also included the radial variation in the plasma emissivity ( @xmath162 ) . in figure [ fig.profile ]
we show the radial profile of the emission - weighted projection of @xmath163 ; i.e. , the `` norm '' parameter ( see caption to table [ tab.spec ] ) of the model divided by the area of the annulus . a single @xmath35 model provides a good description of the data with fit residuals @xmath164 ( table [ tab.gastemp ] ) .
because the inner three annuli were excluded , a marginally larger value of @xmath35 was obtained compared to the surface brightness fits ( [ image ] ) . adding a second @xmath35-model component
provides small improvement in the fits with smaller residuals ( @xmath165 ) than the @xmath133 model ( table [ tab.gastemp ] ) ; the @xmath137 model is displayed in figure [ fig.profile ] .
although the parameters for the @xmath137 model are not as well constrained as those for the @xmath133 model we adopt the @xmath137 model for our analysis of the mass distribution because it describes the data a little better . note that if the inner three annuli are included then the parameters obtained are consistent with those in table [ tab.gastemp ] but are somewhat better constrained . for the case where the annuli are oriented about the emission peak the second component of a @xmath137 model ( fitted to all annuli ) matches very well the @xmath133-model results obtained for the centroided case listed in table [ tab.gastemp ] , demonstrating the similarity of the gas density profiles obtained outside the core region for the peak and centroid case .
the radial profile of the emission - weighted projection of @xmath48 is shown in figure [ fig.profile ] ( middle panel ) along with the best - fitting power - law fit .
the simple power law is an excellent fit ( table [ tab.gastemp ] ) and demonstrates that the temperature profile is consistent with isothermal within the estimated @xmath93 errors .
notice that despite the inner three annuli being excluded from the fits the model describes those data points well .
we constructed the mass profile using the models just described for the gas density and temperature .
following previous studies ( e.g. , * ? ? ?
* ) , for each annulus listed in table [ tab.gastemp ] we assign a single ( three - dimensional ) radius value , @xmath166^{2/3}$ ] , where @xmath167 and @xmath168 are respectively the inner and outer radii of the ( two - dimensional ) annulus .
the radial mass profile is plotted in figure [ fig.profile ] using the @xmath137-model for and the power law model for @xmath48 .
( we emphasize that although the mass values for the inner three data points are shown , being derived from the gas density and temperature models , they are not included in the fits . )
the nfw model is a good smooth fit when excluding the inner three radial bins ( figure [ fig.profile ] ) .
we obtain @xmath169 for 4 dof with fit residuals @xmath170 .
because the mass data points are correlated it is necessary to recalibrate the @xmath92 statistic to examine goodness of fit using our monte carlo simulations ( [ spec ] ) ; i.e. , for each set of parameters for the gas density and temperature profiles obtained for each simulation , a mass profile is constructed and then fitted with an nfw model .
these simulations indicate the the @xmath92 value quoted above is within one standard deviation of the expected value . in table
[ tab.mass ] we give the parameters of the nfw fit and the estimated statistical and systematic errors .
the estimates for the systematic errors are mostly based on our discussion in [ spec ] , and we have also included the differences obtained when using a @xmath133 model to parameterize the profile .
( note we do not provide an estimated error arising from the deprojection method because the statistical errors are considerably larger when using the onion - peeling method . ) the derived concentration parameter ( @xmath171 ) and virial radius ( @xmath172 mpc ) imply a virial mass of @xmath173 appropriate for a massive galaxy cluster .
the value of @xmath150 is consistent with that expected from cdm ( e.g. , * ? ? ?
* ; * ? ? ?
we mention that within the central region ( i.e. , the central three annuli ) , where the assumption of hydrostatic equilibrium is suspect , the mass density profile is flatter than nfw ; i.e. , the cumulative mass profile at small radii lies below nfw as seen in figure [ fig.profile ]
. however , for the case where the annuli are centered about the emission peak we find the mass profile within the central @xmath174 kpc is highly uncertain but consistent with nfw .
( to obtain a reasonable smooth fit to the centrally peaked temperature profile in this case , we modeled the profile with two power laws with exponential cut - offs at both ends . ) the smaller mass values in the core for the centroid case are attributed to the flatter density and temperature profiles obtained with respect to the peak case ; i.e. , when the annuli are not positioned about the emission peak radial differences in spectral quantities are smeared out .
outside the core ( @xmath0 kpc @xmath1 ) the hot icm has properties consistent with a ( relaxed ) cool - core cluster out to the largest radii investigated ( @xmath2 kpc @xmath175 ) .
the x - ray surface brightness is very regular , moderately elliptical in shape , with no evidence of substructure ; previous studies generally indicated a relaxed cluster on such scales ( e.g. , * ? ? ?
the radial profile of the density of the icm is fitted well with a single @xmath35 model , with a double @xmath35 model providing slight improvement .
the temperature profile is consistent with isothermal or gently rising ( e.g. , * ? ? ?
* ; * ? ? ?
the entropy factor ( @xmath176 , e.g. , @xmath177 kev @xmath178 at @xmath179 ) is consistent with those observed in cool - core clusters ( e.g. , * ? ? ?
the iron abundance declines with increasing radius ( e.g. , * ? ? ?
the gravitating mass profile is well fitted by an nfw profile with reasonable concentration parameter for a halo of its virial mass .
finally , the gas fraction ( @xmath180 at @xmath179 ) is quite typical of relaxed clusters observed by and produced in cdm simulations ( e.g. , * ? ? ?
* ) . inside the core
, however , the data reveal an irregular , disturbed system .
the x - ray image displays asymmetrical isophotes about the emission peak which are offset from the global emission centroid by @xmath181 kpc . when oriented about the emission peak , the temperature profile of the icm declines towards the center ( down to @xmath105 kpc ) , very much resembling a cool - core cluster .
but at small radii the temperature profile turns around and increases toward the center ; i.e. , the inner temperature profile is inconsistent with a constant at the @xmath6 level .
both the iron abundances and entropy profiles mirror the shape of the temperature profile with an excess at the center .
since a644 does not possess significant radio emission @xcite , and the image does not show any x - ray cavities , the suppression of a cooling flow can not be attributed to the present day feedback from an agn . nor
is it easily ascribed to heating by a past radio outburst from the black hole in the cd , because the synchrotron lifetime of 1 ghz - emitting cosmic rays is @xmath182 yr for a magnetic field strength of a few @xmath183gauss in the cluster icm .
since this energy loss timescale is comparable to the sound - crossing time for the cluster core ( and thus to the time needed to heat the core ) , the core should remain hot for about the same time as the 1 ghz radio emission remains visible after heating is terminated .
it is even more difficult to explain heating of the `` hot spot '' ( i.e. , the central `` hot '' radial bin oriented about the emission peak see figure 2 ) by an agn in the recent past , because an even shorter time since the cessation of cosmic ray injection is required due to its smaller physical size .
nevertheless , a search for a steep - spectrum radio source would be worthwhile .
the absence of low - frequency radio emission ( e.g. , vla low - frequency sky survey ; * ? ? ?
* ) would be an even more powerful test against the agn heating model in a644 , because 100 mhz - emitting cosmic rays have three times longer lifetimes in comparable magnetic fields .
moreover , although the higher temperature in the `` hot spot '' appears to be accompanied by a sharp increase in iron abundance , supernovae are a most unlikely heating source for even this small portion of the core because the required number of supernovae would need to be accompanied by a recent , galaxy - sized ( @xmath184 ) starburst .
neither a blue galaxy nor a luminous emission - line nebula are visible in the cluster core of a644 @xcite . however , the morphological irregularities in the core of the x - ray image are very likely the result of merging . the cd galaxy ( 2masx j08172559 -0730455 )
is located @xmath185 kpc south of the x - ray emission peak nearly along the direction of the centroid shift noted in [ image ] .
it is also @xmath186 kpc displaced from the centroid of the x - ray emission computed from larger radii .
evidently the cd galaxy and core icm are `` sloshing '' in the potential well of the main cluster , and this energy has prevented the establishment of a cooling flow .
the metallicity peak measured within the innermost radial bin probably indicates that previous merging did not destroy the initial peak , created when it was spatially coincident with the cd galaxy .
this is because , as argued above , the peak is unlikely the result of very recent supernova enrichment . also , the visual appearance of the a644 galaxies is quite diffuse , stretching @xmath187 2 mpc linearly across the sky , suggesting a system far from relaxation .
this morphology strengthens the argument for an on - going merger .
the properties of a644 ( this paper ) , a2029 @xcite , and a2589 @xcite have important similarities .
their x - ray images indicate a peak which is offset from the centroid of the global x - ray halo , with a644 displaying the most extreme shift ( @xmath188 kpc offset ) compared to the others ( a2589 : @xmath189 kpc , a2029 : @xmath190 kpc ) .
other than these offsets , the x - ray emission is very regular in each case with no indications of cavities as are seen in cool - core clusters with agn . )
surface brightness deviations from a smooth elliptical model in the central @xmath191 kpc of a2029 using the ao-1 image , and claim that the wat morphology anitcorrelates with these small fluctuations . using a more recent , deeper ( 80 ks ) chandra exposure we confirm @xmath192 deviations in the surface brightness ( zappacosta et al .
2005 , in prep ) , but we do not confirm a strong anticorrelation with the wat morphology as seen in other cool - core clusters . compared to other cool - core clusters with obvious central disturbances these @xmath192 deviations in the surface brightness , which translate to even smaller ( @xmath193 ) deviations in the gas density , indicate that the core of a2029 is in a much more relaxed state . for comparison
, we find that perseus , a cool - core cluster with an obvious distrurbance in the core , displays up to 50% fluctuations in the surface brightness .
it should be added that , unlike a2029 , perseus also has strong azimuthal fluctuations in the temperature and metallicity which , however , apparently consipire to produce approximate azimuthal balance in the gas pressure and therefore approximate hydrostatic equilibrium @xcite . ]
two of these clusters have no detected agn radio emission , while a2029 has a wat morphology agn , very unusual ( unique to our knowledge ) for a cool - core cluster . finally
, the lack of nebular emission lines in the cd galaxies are characteristic of non - cool - core clusters @xcite . of these three clusters
only a644 possesses a central temperature peak .
although a644 appears to be unique in that its temperature profile declines toward the center and then rises , there are a few reports of clusters in the literature with temperature profiles measured by either or that peak at their centers . the clusters a2218 @xcite and a3921 @xcite are violent , advanced mergers with with pronounced asymmetrical surface brightness and temperature distributions uncharacteristic of a644 , a2029 , and a2589 .
another cluster with a reported central temperature peak is a401 @xcite , which is in the early stages of a major merger ( with a399 ) .
nevertheless , @xcite argue that the properties of a401 are determined primarily from its own hierarchical formation before interaction with a399 . in this context
it is interesting that a401 has a weak radio halo since we have suggested that radio halos form only in massive clusters where a violent merger has proceeded fully into the cluster core @xcite .
if the other clusters had a similar history , their lack of radio halos indicate they are presently at much more evolved states than a401 . in the paradigm wherein the energy output from an agn disrupts cooling flows
, it is expected that just prior to the onset of agn feedback a cluster will have the characteristics of a cool - core cluster but without significant radio emission or x - ray cavities .
indeed , the clusters a644 , a2029 , and a2589 are excellent candidates for this pre - feedback state , with a644 being at a relatively early state and a2029 at the very latest evolutionary state , where an agn outburst has already occurred but little or no heating from agn - generated cosmic rays has commenced .
the extremely advanced evolutionary states of a2029 and a2589 make them ideal for studies of their core mass profiles using the assumption of hydrostatic equilibrium .
we thank a. lewis for his participation in this project at early stages .
thanks a. venketasen for useful discussions of cluster heating mechanisms .
d.a.b . acknowledges partial support from nasa under grant nag5 - 13059 issued through the office of space science astrophysics data program .
partial support for this work was also provided by the national aeronautics and space administration through chandra award number go1 - 2130x issued by the chandra x - ray observatory center , which is operated by the smithsonian astrophysical observatory for and on behalf of the national aeronautics and space administration under contract nas8 - 03060 . | we present an analysis of a acis - i observation of the massive galaxy cluster a644 .
this cluster was previously classified as a cooling flow , but no radio emission has been detected from its cd galaxy . outside the core ( @xmath0 kpc @xmath1 ) the hot icm has properties consistent with a ( relaxed ) cool - core cluster out to the largest radii investigated ( @xmath2 kpc @xmath3 ) . over this region
the gravitating mass profile is described well by a navarro - frenk - white profile with concentration parameter , @xmath4 , and virial radius , @xmath5 mpc . however ,
inside the core the temperature and entropy profiles reverse their inward radial decline and rise at the center ; the inner temperature profile is inconsistent with a constant at the @xmath6 level . although the core region does not display x - ray cavities or filamentary structures characteristic of radio - loud , cool - core clusters , the peak of the x - ray emission is offset from that of the centroid of the global x - ray halo by @xmath7 kpc .
the position of the cd galaxy lies approximately between the x - ray peak and centroid , further testifying to a merger origin for the properties of the x - ray emission in the core .
we discuss the implications of a644 and the small number of radio - quiet , cool - core clusters for the agn feedback paradigm to suppress cooling flows in clusters . |
low mass x - ray binaries ( lmxbs ) are usually characterized by weak magnetic fields . in these systems
the accretion disk can reach regions close to the compact object , where the matter is subject to a strong gravitational field .
this results in complex power spectra with characteristic frequencies , determined by the dynamical time scales close to the compact object , up to more than a thousand hz .
these timing structures can be used to probe the physics near the compact object , and timing is a key tool to perform such studies . because several features are usually simultaneously present in the power spectra of lmxbs , we need to understand the entire timing structure in order to construct reliable theories for the different timing signals .
power spectra of lmxbs show significant changes as a function of the source intensity and spectral hardness . based on their spectral and temporal behavior
, lmxbs can be divided into two classes , the so - called z and atoll sources ( hasinger & van der klis 1989 ) , whose names are derived from the tracks they trace out in an x - ray color - color diagram .
the z sources are thought to have higher mass accretion rates and perhaps higher magnetic fields than the atoll sources .
these two classes of sources show a temporal behavior that strongly depends on their position in the color - color diagram ( hasinger & van der klis 1989 ) . in particular , atoll sources can be found in the so - called island state , where the power spectrum is characterized by a strong band - limited noise component , called the high frequency noise ( hfn ) , which extends up to tens of hz in frequency , or in the banana state , where the power spectrum is dominated by a power - law noise below 1 hz , usually called very low frequency noise ( vlfn ) .
it is thought that these states reflect different mass accretion rate , which increases from the island to the banana ( see van der klis 1995 ) . in the island state
the power spectrum usually shows other features , such as a low frequency quasi - periodic oscillation ( hereafter lfqpo ) between 20 and 40 hz .
this lfqpo , which may be similar to the horizontal branch oscillation ( hbo ) in z sources , has been variously interpreted as due to the lense - thirring precession of material in the innermost disk region ( stella & vietri 1998 ) , to radial oscillations in the boundary layer between the neutron star surface and the accretion disk ( titarchuk & osherovich 1999 ) , or in terms of the magnetospheric beat frequency model ( alpar & shaham 1985 ; lamb et al .
1985 ; see also psaltis et al .
1999a ) .
observations with the rossi x - ray timing explorer ( rxte ) have shown kilohertz quasi - periodic oscillations ( khz qpo ) in the persistent emission of 20 lmxbs , with frequencies ranging from a few hundred hz up to @xmath2 hz ( see van der klis 2000 for a review ) . usually two khz peaks are simultaneously observed .
the difference between the frequencies of the two simultaneous khz qpos is , in many cases , consistent with being constant . during type - i x - ray bursts in some atoll sources ,
nearly coherent oscillations were also detected at frequencies between 330 and 590 hz , interpreted as the spin frequency of the neutron star ( strohmayer et al .
1996 ) . in these cases
the frequency separation between the two simultaneous khz qpos is similar to the frequency of the burst oscillations , or half that value ( see strohmayer , swank , & zhang 1998 for a review ) , suggesting a beat frequency mechanism as the origin of the khz qpos ( strohmayer et al .
1996 ; miller , lamb , & psaltis 1998 ) . however , the peak separation is sometimes significantly smaller than the frequency of the burst oscillations ( mndez , van der klis , & van paradijs 1998 ) , and decreases as the frequency of the khz qpos increases ( van der klis et al .
1997 ; mndez et al .
1998 ; mndez & van der klis 1999 ) .
other models for the khz qpos are based on general relativistic periastron precession at the innermost disk boundary ( stella & vietri 1999 ) , or on relaxation oscillations that occur at the centrifugal barrier around the neutron star ( titarchuk , lapidus , & muslimov 1998 ) .
the broad - band noise component , the lfqpo and the khz qpos seem to be correlated with each other and with the position of the source in the color - color diagram .
the frequency of the lfqpo in z - sources ( i.e. the hbo ) is correlated with the khz qpo frequency ( e.g. van der klis et al .
1997 ; wijnands et al . 1998 ; see also psaltis , belloni , & van der klis 1999b , and references therein ) , as is the lfqpo in atoll sources ( e.g. stella & vietri 1998 ; ford & van der klis 1998 ; van straaten et al .
the lfqpo in z sources may differ from that in atoll sources , because they might be the second harmonic instead of the fundamental ( cf .
jonker et al .
1998 ; wijnands & van der klis 1999 ) . the break frequency of the broad - band noise component in atoll sources is also correlated with the khz qpo frequency ( ford & van der klis 1998 ; van straaten et al .
2000 ; reig et al .
2000 ) and such a component may also be present in z - sources ( wijnands & van der klis 1999 ) .
the correlation between the lfqpo and the khz qpos seems to be general , including also the @xmath3 hz qpos observed in black hole candidates ( bhcs ; psaltis et al .
another general correlation seems to exist between the frequency of the break of the broad - band noise component and the frequency of the lfqpo ( wijnands & van der klis 1999 ) .
a detailed study of the khz qpos in several atoll sources revealed that there is a correlation between frequency and position in the color - color diagram ( see e.g. mndez & van der klis 1999 ; kaaret et al .
1999 ; mndez 1999 ) or spectral shape ( ford et al .
1997a ; kaaret et al . 1998 ) .
these correlations suggest that all these properties depend on one parameter that could be the mass accretion rate ( cf .
hasinger & van der klis 1989 ) . on the other hand ,
the relation between the frequency of the khz qpos and count rate is complex ( ford et al .
1997b ; yu et al .
1997 ; zhang et al . 1998 ; mndez 1999 , and reference therein ) , suggesting that the x - ray flux is not a good indicator of the mass accretion rate .
one of the most interesting sources in this context is the bursting source 4u 172834 .
it belongs to the class of the atoll sources , and shows temporal variability on all time scales resulting in a complex power spectrum .
it was observed with the exosat satellite in the island state ( hasinger & van der klis 1989 ) , but a great deal of new information about this source has recently come from rxte observations .
4u 172834 shows khz qpos ( strohmayer , zhang , & swank 1996 ) and burst oscillations at 363 hz ( strohmayer et al .
the peak separation between the khz qpos in this source is always smaller than 363 hz , and decreases significantly at higher inferred accretion rates ( mndez & van der klis 1999 ) .
besides these high - frequency features , the power spectrum of this source also shows a broad - band noise component ( with a break frequency around 10 hz ) , a lfqpo ( between 20 and 40 hz ; strohmayer et al .
1996 ) and a bump around 100 hz ( ford & van der klis 1998 ) .
both the centroid frequency of the lfqpo and the frequency of the break of the broad - band noise seem to be correlated with the frequency of the khz qpos ( ford & van der klis 1998 ) . in this paper we present new results about the correlations between noise and qpos in 4u 172834 , obtained from a study of the temporal behavior of this source as a function of the position in the color - color diagram . for this analysis we used all the rxte observations performed in 1996 and 1997 , that are now publicly available
( results from these data were already published by strohmayer et al .
1996 ; strohmayer , zhang , & swank 1997 ; ford & van der klis 1998 ; mndez & van der klis 1999 ) .
we find that , while the lfqpo frequency is correlated to both the khz qpo and break frequencies at low inferred mass accretion rates , at higher accretion rate there is a jump in this correlation , probably because a new lfqpo appears at the position of the break , and another noise component appears in the power spectrum .
we analyzed data of 4u 172834 from the public rxte archive , using observations performed in 1996 between february 15 and march 1 , 1996 on may 3 , and 1997 between september 23 and october 1 .
the log of these observations is presented in table 1 . from the instruments on board rxte , we only used data from the proportional counter array ( pca ) , selecting intervals for which the elevation angle of the source above the earth limb was greater than 10 degrees .
we discarded those data ( @xmath4% of the total ) where 1 or 2 of the 5 pca detectors were switched off .
several bursts occurred during these observations , but we excluded those intervals ( @xmath5 s around each burst ) from our analysis .
we used the standard 2 mode data ( 16 s time resolution ) to produce light curves and the color - color diagram , and event data ( with a time resolution of 16 @xmath6s or 122 @xmath6s ) to produce fourier power spectra as described below .
we used the standard pca background model 2.1b to extract background subtracted light curves binned at 16 s in the energy bands @xmath7 , @xmath8 , @xmath9 , @xmath10 kev .
we defined the soft color as the count rate ratio in the bands @xmath8 kev and @xmath7 kev , and the hard color as the ratio of count rates in the bands @xmath10 kev and @xmath9 kev , and produced a color - color diagram . because the observations span two years , we took into account the gain changes of the pca between epoch 1 ( until 1996 march 21 ) and epoch 3 ( from 1996 april 15 to 1999 march 22 ) , as well as the additional continuous gain change that occurs within each epoch . as it can be seen in table 1 , 18 observations ( @xmath11 of the observation time ) took place in february - march 1996 during epoch 1 ,
one observation occurred in may 1996 during epoch 3 , and the remaining 9 observations occurred in september - october 1997 , again during epoch 3 . to correct the data for changes of the gain of the pca , we proceeded as described in kuulkers et al .
( 1994 ) : we selected rxte observations of crab obtained close to the dates of our observations , and calculated x - ray colors for crab . between february 1996 and october 1997 the soft and hard colors of crab increased by @xmath12% and @xmath13% respectively , and between may 1996 and october 1997 the same colors increased by @xmath13% and @xmath4% , respectively , so we applied these ( multiplicative ) corrections to the colors of 4u172834 .
the resulting color - color diagram is shown in figure 1 .
the reason to use crab colors to estimate changes in the instrumental response is that the spectrum of crab is supposed to be constant , and therefore changes in the colors of crab are caused by changes in the instrumental response .
however , the corrections calculated in this way do not take into account the difference in spectral shape between crab and 4u172834 .
we used spectral fits of 4u172834 and the response matrices to test the magnitude of this effect .
the correction factors obtained in this way are very similar to those calculated using crab .
they are @xmath14% larger in soft color and @xmath4% larger in hard color than the crab - based correction factors , less than the statistical errors of the colors from each 16 s interval ( @xmath15% in both soft color and hard color ) .
we preferred to use the crab - based correction as it is independent of any model of either instrumental response or source spectrum . to study the temporal variability of the source as a function of the position in the color - color diagram we divided this diagram into 19 intervals , as shown in figure 1 .
the intervals were chosen in order to have a good number of intervals , and good statistics in each interval . for each interval
we computed a power spectrum as follows : we divided the pca light curve into segments 256 s long for which we calculated fourier power spectra with no selection on energy , using a nyquist frequency of 2048 hz .
because every color point spans 16 s , each power spectrum corresponds to 16 color points . due to the statistical uncertainties and to the motion of the source in the color - color diagram
, a power spectrum could be spread over more than one interval ( usually 2 to 3 intervals ) . for each interval
, we averaged together all power spectra that fall in that interval for more than 30% of the time , weighing each of them by the fraction of the time it actually spent in the interval .
we used 50% of the time in the upper banana ( intervals 18 and 19 ) , where the motion of the source in the color - color diagram is faster .
this selection criterion allows some overlap between the nearest intervals .
we checked that this overlap does not affect our results .
in fact , the results we discuss in this paper do not significantly change if we use more restrictive selection criteria ( without overlap between intervals ) , although qpo features are usually 1020% narrower in those cases .
the power spectra , one for each interval selected in the color - color diagram , are indicated with numbers from 1 to 19 starting from the island state ( upper right corner in the color - color diagram ) .
each interval contains from 10 to 134 power spectra .
the high frequency part of each power spectrum ( @xmath16 hz , where neither noise nor qpos are known to be present ) was used to estimate the poisson noise , which was subtracted before performing the spectral fitting . in figures 2a and 2b
we show some representative power spectra ( normalized as in leahy 1983 ) , corresponding to the intervals 4 , 8 , 10 , 11 , 12 , 13 , 14 , and 19 of the color - color diagram of figure 1 .
power spectra 111 are typical of the island state of atoll sources , while 14 and 19 correspond to the lower and the upper banana respectively ( see hasinger & van der klis 1989 for definitions ) . in spectra
111 we can distinguish several components : the hfn , with a break between 2 and 20 hz , the lfqpo , between 10 and 40 hz , a bump around 100 hz , and one or two khz qpos . we used a broken power law , @xmath17 where @xmath18 is the power - law index below the break and @xmath19 is the power - law index above the break , to describe the broad - band noise , and lorentzians to describe the lfqpo , the 100 hz peaked noise , and the khz qpos ( when present ) .
the characteristic frequencies of all these features are seen to increase from intervals 1 to 9 . in intervals
1013 a complex change is seen to occur in the power spectra .
one way to describe this change is as follows : the break of the power law becomes sharper , and a bump is seen to occur on top of this break , which gradually turns into a qpo , pushing back the break to lower frequency . at the same time the original lfqpo , while still visible , weakens .
we will discuss below how to describe this transition in terms of fits of the power spectra . in intervals 14 and higher ,
besides the previous components , we also see the vlfn , which we fitted using a power law , @xmath20 , with index @xmath21 .
the power spectrum of interval 19 appears much simpler than the others , consisting only of two noise components , namely the vlfn and the hfn . in this case
the hfn appears to have a different shape from the one we observe in the island state , because it is peaked , i.e. the power law has a positive index before the break .
following hasinger & van der klis ( 1989 ) , in this case we modeled the hfn using a power law with @xmath22 plus an exponential cutoff ( @xmath23 hz ) .
for those intervals where the frequencies of the khz qpos were high enough that the khz qpos do not affect the parameters of the low frequency features and vice versa , we fitted the khz qpos in a limited frequency range ( 5002048 hz ) where no other component was present .
we then fitted the other components using the full frequency range and fixing the parameters of the khz qpos at the values obtained in the high frequency range .
in particular , for interval 7 and for intervals 10 to 12 we did this with upper khz qpo , and for intervals 13 to 17 we did it for both khz qpos . for the other intervals we fitted all the components
( with all the parameters free ) in the full frequency range .
the broad - band noise , the lfqpo , the 100 hz peaked noise , and the khz qpos are detected in 17 of the 19 intervals in the color - color diagram ( 1 to 17 ) , with the exception of the lower khz peak that is absent in the first 9 intervals .
the vlfn is detected in intervals 14 to 19 .
the broad - band noise is also detected ( with a different shape ) in intervals 18 and 19 . in table 2
we report the results of these fits and in section 3 we discuss all these results in more detail . in figures 2a and 2b
we also plot the best fit model and the corresponding residuals ( in units of @xmath24 ) for each power spectrum . for intervals 4 and 5
we obtain high reduced @xmath25 : 4.5 and 4.1 ( for 131 degrees of freedom ) , respectively .
looking at the residuals in figure 2a , it is apparent that for these two intervals the main problem is the low frequency part of the power spectrum ( below @xmath15 hz , i.e. the frequency of the break ) , where the shape of the noise is curved , and is badly approximated by a power law ( see also section 2.2 , where we try an alternative model for the noise component ) .
also , the abrupt change of the slope of the broken power law around the break frequency does not represent well the smooth change of the slope in the data there .
some residuals are also visible near the lfqpo ( between 10 and 30 hz ) due to the shape of the qpo , that is sharper than a lorentzian . for intervals 15 and 16
we obtain reduced @xmath25 of 2.3 and 2.6 ( for 132 degrees of freedom ) , respectively .
for these two intervals the largest residuals appear in the region of the khz qpos .
in particular , the leading wing of the lower khz qpo is not well fitted by the model .
for all the other power spectra the reduced @xmath25 ranges from 1.3 to 2.0 ( for @xmath26 degrees of freedom ) .
these values are also rather large , but this is not surprising given the high statistics of the power spectra and their complexity
. nevertheless , for these other intervals the residuals in figure 2a , b are usually within 3 @xmath24 .
most of the time the largest residuals occur in the low frequency part of the power spectrum ( due to the curved shape of the noise ) , or around the break ( due to the sharp shape of the model at the break frequency , compared to the smoother shape of the data there ) . from all the above we
conclude that the high values of @xmath25 we obtain in some cases are mainly due to the detailed shape of the noise component , which is not so well approximated by a broken power law .
however , this does not affect strongly the parameters of the qpos , where the residuals are usually flat , or the behavior of the break frequency of the broad - band noise component , even though a broken power law is sometimes an oversimplification of the complex shape of these power spectra . under the applied model , our estimates of the statistical errors on the fit parameters , which use @xmath27 , for a one - sigma single parameter error ,
therefore provide a good estimation of the total error for the qpo parameters .
for the noise parameters systematic effects probably dominate .
these will be addressed below .
the model described above gave the best fit for most of the intervals .
it is also the most commonly used model in the literature , which allows us to compare our results to those obtained for other sources .
however , we note that this model is not the only possible choice . for some intervals other models can fit the data equally well , and more complex models can fit better .
this non - uniqueness of the model contributes systematic uncertainties to the parameters . in the next sections we try different models for the power spectra , and we discuss how the results can be affected by using different models .
our conclusion will be that , independently of the model chosen to describe the spectra , in intervals 1012 a major change gradually occurs in the power spectra . in the model just described this change shows up as a jump in the frequency and rms amplitude of the hfn and the lfqpo ( cf .
5 ) between intervals 12 and 13 . other models can place this jump earlier ( between 9 and 10 ) .
the results presented in the previous section were obtained by fitting the power spectra using always the same function .
while this facilitates the study of the power spectral parameters as a function of the position in the color - color diagram , we must note that the power spectra of intervals 10 to 12 are more complex than the others , and extra components may be required to fit them properly ( cf .
2 ) . for those intervals
, there is an excess of power , that can be fitted by an extra qpo , at about the break frequency of the broken power law that we used to fit the hfn .
for instance , if we add this extra qpo to our fit function of interval 12 , we get lfqpos at @xmath28 hz and @xmath29 hz , while the break frequency of the hfn decreases to @xmath30 hz , compared to @xmath31 hz when we only fit one lfqpo to the power spectrum .
this extra lfqpo significantly improves the fit , as @xmath25 decreases from 192 for 131 degrees of freedom to 149 for 128 degrees of freedom ( the f - test probability is @xmath32 ) .
the addition of this extra qpo does not change the parameters of the other lfqpo , the 100 hz peaked noise , and the khz qpos ; the differences between the values of these parameters using these two different models are well within the errors . however , the inclusion of this extra lfqpo changes the parameters of the hfn . both the break frequency we obtain for interval 12 using this model ( @xmath33 hz ) and the frequency of the extra lfqpo ( @xmath31 hz )
are in line with the results we obtain for interval 13 .
the change in the shape of the power spectra , that shows up in the model with only one lfqpo as a jump in the lfqpo frequency at interval 1213 ( to be discussed below ) , can be detected by the more complex model already in lower intervals .
the results reported in figures 2a and 2b and in table 2 refer to the model with only one lfqpo . for the first intervals , especially 4 and 5 , the power spectra are not flat at low frequencies , and therefore not well fitted by a power law ( cf .
2 ) ; we then get better fits using a lorentzian instead of a broken power law . for interval 4 , @xmath34 using the broken power law , and @xmath35 using the lorentzian , giving a statistically significant improvement (
although the @xmath25 is still large ) .
the lorentzian fit becomes complicated , however , for intervals 1012 .
the sharp break there ( see fig .
2 ) can not be fitted with the smoother lorentzian .
if we fit a lorentzian anyway , this sharp edge appears as a qpo ; two low frequency qpos are now needed to fit these spectra . for interval 11 , @xmath36 for the broken power law and @xmath37 using the lorentzian plus qpo for the broad - band noise , again a statistically significant improvement .
note that an extra lfqpo also improves the fit of intervals 1012 when the broken power law is used to model the hfn ( see section 2.1 ) . for the other intervals the fits are equally good whether we use a broken power law or a lorentzian to represent the broad - band noise component .
the centroid frequency of the lorentzian used to fit the broad - band noise goes from @xmath38 hz to @xmath39 hz for intervals 1 to 6 , and it is around 3.2 ( @xmath40 ) hz for intervals 9 to 13 .
its rms amplitude ranges from 15 to 20% up to interval 10 ; then it jumps down to 10% and continues decreasing to @xmath15% .
this is the same jump that we see in the rms amplitude of the broken power law at interval 1213 ( fig .
5 ) : it occurs at interval 910 now because in intervals 1012 the necessary extra qpo ( centered near the break ) absorbs some of the power . comparing the half width at half maximum ( hwhm ) of the lorentzian ( plus the centroid frequency of the lorentzian ) and @xmath41 in each interval
, we see that there is a good correspondence between these two parameters up to interval 9 , with the hwhm of the lorentzian being a little bit larger than @xmath42 . in intervals 10 to 12 @xmath42 increases ( see table 2b ) while the hwhm of the lorentzian is almost constant around 10 hz . again , this is due to the extra qpo at @xmath43 hz . for intervals
13 to 17 both the hwhm of the lorentzian and @xmath42 are almost constant : @xmath42 is @xmath1 hz and the hwhm of the lorentzian is @xmath44 hz .
we conclude that the choice of either of these two models to fit the broad - band noise component does not affect the behavior of the parameters that we obtain , except for the width of the noise after interval 12 .
in fact , the parameters of all the other components that we obtain using a lorentzian instead of the broken power law for the broad - band noise are consistent ( usually within @xmath45 ; in a few cases the parameters of the lfqpo are within @xmath46 ) with the values already discussed , and show the same behavior when plotted against the frequency of the upper khz qpo .
the addition of an extra lfqpo near 20 hz in intervals 1012 , whether we use a lorentzian for the hfn or a broken power law , modifies the width of the noise ( break frequency or lorentzian hwhm ) and its rms amplitude , which jump to lower values ( @xmath1 hz and @xmath47 , respectively ) .
these results are in line with the parameters of the hfn and lfqpo in interval 13 and higher , indicating that a gradual change in the shape of the power spectra begins in intervals 1012 .
finally , we also tried to fit these spectra using zero - centered lorentzians for the noise and the bump at 100 hz , as in olive et al .
( 1998 ) for the source 1e 17243045 .
we did not obtain good fits , the principal reason being that the 100 hz peaked noise can not be well fitted by a zero - centered lorentzian for most of the intervals .
we now describe the results from the fit with the model described in section 2 .
we will note differences with the other fit function where appropriate in section 4 .
in figure 3a we plot the frequencies of both khz qpos ( upper panel ) and their rms amplitudes ( lower panel ) as a function of the interval number in the color - color diagram .
we also show the khz qpo rms amplitudes as a function of the corresponding khz qpo frequency in figure 3b .
while in the power spectra of intervals 10 to 17 we see both khz qpos , in the power spectra of intervals 1 to 9 ( at low inferred mass accretion rate ) we only see one of them , which in principle could be either the upper or the lower peak . however , when we plot the rms amplitude of the khz qpos vs. the position in the color - color diagram for all the intervals in which we detect khz qpos ( fig .
3a , lower panel ) , the rms amplitude of the single khz qpo connects smoothly with the rms amplitude of the upper khz qpo , whereas at the same time the rms amplitude of the lower khz qpo is significantly lower .
the same behavior is also visible in figure 3b , where we plot the rms amplitude of each khz qpo as a function of the corresponding frequency .
we therefore conclude that when we only see one of the khz qpo , it is always the upper peak ( see also ford & van der klis 1998 ; mndez & van der klis 1999 ) . the frequencies of the khz qpo range from @xmath48 to @xmath49 hz for the upper peak , and from @xmath5 to @xmath50 hz for the lower peak .
the frequencies of both qpos are well correlated to the position of the source in the color - color diagram , although the frequency of the upper khz qpo seems to jump from @xmath5 to @xmath51 hz between intervals 6 and 7 .
this jump does not seem to be present in the plot of khz qpo frequency versus position of the source in the color - color diagram ( indicated with @xmath52 ) shown in mndez & van der klis ( 1999 ; fig .
4 ) . however , this is because the jump is hidden by the intrinsic fluctuations of the khz qpo frequency in the plot of mndez & van der klis ( 1999 ) .
if we average together points with similar values of @xmath52 in their plot , we find that the average frequency of the khz qpo remains more or less constant at @xmath5 hz for @xmath53 , similar to what we observe for intervals 36 , and then jumps to @xmath51 hz for @xmath54
. for both khz qpos the rms amplitude ranges from 3% to @xmath55 , but while the rms amplitude of the upper peak generally decreases with frequency , the rms amplitude of the lower peak increases with frequency ( except for the last point , which is significantly lower than the previous one ) .
the width of the upper peak generally decreases with frequency .
this implies higher values of the quality factor @xmath56 ( which ranges from @xmath57 to @xmath58 ) , and therefore higher coherence of the signal , at higher frequency . for the lower peak no clear trend is evident .
in this case q ranges between 3.5 and 7.4 and is @xmath59 in the last interval in which the lower peak is detected .
we note that in each interval we averaged many power spectra , and therefore the measured width of the khz qpos could be affected by changes in their centroid frequencies within the interval . in particular , we can estimate the effect of the frequency variations on the width of these features , considering the relation found for the frequency of the khz qpos versus interval number , and knowing that each power spectrum can be spread over at most three intervals . for interval numbers
higher than 7 , we find that the frequency of the khz qpos varies by @xmath60 hz across 3 intervals .
this is comparable with the measured widths of the khz qpos in interval 7 and higher where , therefore , the widths of the khz qpos are likely considerably affected by this broadening effect . for interval numbers lower than 7 , the frequency of the khz qpos varies by @xmath61 hz across 3 intervals .
this is much smaller than the measured width of the khz qpo ( more than 200 hz fwhm ) in these intervals , indicating that the khz qpos may be intrinsically broader at low frequencies .
the 100 hz peaked noise is a broad feature , with a fwhm that ranges from 100 to more than 200 hz .
its frequency ( fig .
4 , upper panel ) seems to fluctuate around 130 hz , without any clear correlation with the khz qpo frequencies , although at higher values of the upper khz qpo frequency ( @xmath62 hz ) it seems to increase with @xmath63 .
figure 4 ( lower panel ) shows the rms amplitude of the 100 hz bump as a function of the frequency of the upper khz qpo . for @xmath64 hz
the rms amplitude of the 100 hz bump stays constant at @xmath65% , and for @xmath62 hz it decreases to @xmath66% .
it seems to be anti - correlated with the corresponding centroid frequency , at least at higher khz qpo frequencies .
the broken power - law index below the break ( @xmath18 ) ranges from @xmath67 to @xmath68 , and the index above the break ( @xmath19 ) is almost constant around 1.5 ( the typical error is 0.1 ) , except for intervals 10 to 12 where it ranges between 2 and 2.5 .
this is because the shape of the power spectrum at the frequency of the break becomes sharp in these intervals , and probably another component begins to appear here ( see section 2.1 ) . in figure 5 ( upper panel )
we plot the break frequency of the broad - band noise component ( @xmath42 ) vs. the frequency of the upper khz qpo .
the break frequency increases from @xmath4 to @xmath69 hz when @xmath63 increases from @xmath48 hz to 900 hz . beyond this frequency , @xmath42 jumps down to 5 hz and then remains constant around 7 hz .
the rms amplitude of the broad - band noise component ( measured in the whole frequency range ; fig .
5 , lower panel ) generally decreases as the source moves in the color - color diagram from the island to the banana , as expected . in particular
, it remains at @xmath70 up to @xmath71 hz where it starts decreasing ; at @xmath72 hz , it jumps down to 5% , and continues decreasing down to @xmath73% as @xmath63 increases .
similar jumps at the same value of @xmath74 hz are also visible in the frequency and rms amplitude of the lfqpo ( fig . 5 , upper and lower panel respectively ) . for @xmath75
hz the lfqpo frequency increases from 8 to 46 hz as @xmath63 increases . at @xmath76
hz it jumps down to @xmath31 hz , it then starts increasing again with @xmath63 , and finally seems to saturate .
the rms amplitude of the lfqpo initially decreases from 15% to 3% as @xmath63 increases up to 900 hz ; at that point the rms amplitude of the lfqpo jumps up to 10% and then decreases again from 10% to 5% as @xmath63 increases .
table 2b shows the parameters of the vlfn as a function of the interval number in the color - color diagram .
the rms amplitude of the vlfn increases ( from @xmath0% to @xmath77 ) with increasing inferred mass accretion rate , while the index of the power law is almost constant at 1.7 . for intervals 18 and 19 we fitted the broad - band noise using a power law ( with a positive index around 1.5 ) with an exponential cutoff at a frequency of @xmath65 hz . the rms amplitude of this component is around 2% .
the results illustrated in the previous section show that both the frequency and the rms amplitude of the khz qpos and x - ray colors are all well correlated with each other .
it is thought that the position in the color - color diagram reflects different mass accretion rate ( see e.g. priedhorsky et al .
1986 ; hasinger & van der klis 1989 ; hasinger et al . 1990
; van der klis 1994 ; kaaret et al .
1998 ) . moreover , various models of the khz qpos associate their frequencies with the keplerian frequency at the inner rim of the accretion disk ( miller et al . 1998 ; titarchuk et al . 1998 ; stella & vietri 1999 ) . the inner radius of the disk is thought to decrease with increasing mass accretion rate ( miller et al .
1998 ) , suggesting that higher khz qpo frequencies correspond to higher accretion rate .
these considerations suggest that the mass accretion rate determines both the position of the source in the color - color diagram and the related timing behavior ( cf .
hasinger & van der klis 1989 ) . in the first 9 power spectra
we only see one khz qpo . in principle
we can not say , in this case , if it is the lower or upper peak .
ford & van der klis ( 1998 ) tentatively identified this qpo as the upper peak .
this identification is in line with the result of mndez & van der klis ( 1999 ) that , in general , the single khz qpo shows the same correlation with the inferred mass accretion rate as the upper peak .
this identification is strengthened by our results presented here that there is a smooth trend of rms amplitude with qpo frequency if the single qpos are identified with the upper of the two khz qpos ( see fig .
this confirmation is important because the correlation of the lfqpo frequency with the frequency of the khz qpos lines up with that of other sources only if we make the identification as we do here ( see e.g. psaltis et al .
1999b ) .
we find that the parameters of the lfqpo and the broad - band noise are correlated with the frequency of the khz qpos in a complex way .
as we can see in figure 5 , the frequencies of the lfqpo and of the break are correlated with the frequency of the upper khz qpo up to @xmath78 hz , which corresponds to interval 12 , just before the appearance of the vlfn ( see table 2b ) . beyond this frequency
there is a jump in both the break and the lfqpo frequencies , confirming the trend already pointed out by ford & van der klis ( 1998 ) .
after the jump the frequency of the lfqpo seems to increase and then saturate at @xmath79 hz while , except for the first point , @xmath41 seems to be constant around 7 hz .
if for intervals 1012 we use a model with an extra lfqpo ( see sections 2.1 and 2.2 ) , the break of the hfn jumps to lower values , @xmath33 hz . these values are similar to the break frequency measured in interval 13 and higher , indicating that the change in power spectral shape already begins in intervals 1012 ( corresponding to @xmath80 hz ) .
we also note that the frequency of the extra lfqpo ( @xmath31 hz ) is similar to that of the lfqpo in intervals 13 , suggesting that the lfqpo in intervals 13 and higher is the extra lfqpo present in 1012 , while the lfqpo at higher frequencies ( present up to interval 12 ) disappears . as mentioned in section 2 , looking carefully at the power spectra ( fig .
2 ) , something is seen to happen starting at interval 10 , namely an excess appears on top of the break at a frequency that is around half that of the lfqpo , while the lfqpo itself is fading as can be seen from its rms amplitude ( fig .
5 , lower panel ) . in interval 12 two qpos , one at 20 hz ( where we put the break frequency of the broken power law ) and one at 46 hz ,
could be present ( see sec .
so , the jump in the parameters of the lfqpo at @xmath81 hz can be due to the fact that beyond this frequency we measure the parameters of what seems to be another lfqpo , which begins to appear at the frequency of the break already in intervals 10 to 12 . at the same point
, there is a jump in the parameters of the hfn ( see fig . 5 at @xmath76 hz ) .
this change can be identified in the power spectra : in interval 13 the hfn fits what appears to be another noise component characterized by a different behavior and a drastically lower break frequency ( @xmath1 hz ) .
some evidence of the emergence of this noise is already visible in the earlier power spectra ( see sec .
jumps are also visible in the rms amplitudes of the noise component and of the lfqpo at @xmath82 hz ( fig . 5 , lower panel ) .
we note that at the same frequency similar variations ( more gradual in this case ) are also present in the rms amplitudes of the upper khz qpo and the 100 hz peaked noise , which both decrease from about 10% to @xmath66% ( fig .
3b and 4 ) .
a possible interpretation of the behavior of the lfqpo and broken power - law break frequencies is that the noise component at low inferred accretion rates turns into a qpo at higher accretion rates , while what we call the lfqpo at low accretion rates gradually disappears at higher accretion rates . in figure 5 ( top )
we can see that the centroid frequency of the lfqpo after the jump is in line with that of the break , as expected in this interpretation .
a continuity is also visible between the rms amplitudes of the broken power law before the jump and the lfqpo after the jump ( fig . 5 , bottom ) .
we observe that the alternative models for intervals 10 to 12 , containing an extra lfqpo , are also consistent with this interpretation , although the continuity between the rms amplitudes of the broken power law and the lfqpo is not observed in this case , because some of the power is absorbed by the extra lfqpo . after the jump the value of @xmath42 seems to be constant around 7 hz , which is in the same range of frequencies of the normal branch oscillations ( nbos ) observed in the z sources .
no qpo like the nbo are usually observed in atoll sources .
for this reason these qpos were associated with high accretion rates .
recently wijnands et al .
( 1999 ) found that 4u 182030 shows a 7 hz qpo at its highest mass accretion rate levels ( see also dotani et al .
going from the lower to the upper banana , in this source the broad - band noise seems to evolve into a broad peaked noise and then into a qpo around 7 hz ( wijnands et al .
this noise near 7 hz seems quite similar to the noise after the jump shown by 4u 172834 .
so , in 4u 172834 when mass accretion rate increases we see a noise component near 20 hz turn into a qpo , while a new noise component appears at 7 hz .
this component at 7 hz is similar to the one in 4u 182030 , which in that source turns into a qpo when mass accretion rate increases further .
we note that the noise we find after the jump in 4u 172834 seems to become peaked , i.e. before interval 13 the index @xmath18 of the broken power law ( below the break ) is @xmath83 , but after the jump at interval 13 , @xmath18 becomes negative ( see table 2b ) .
a similar behavior was also observed in 4u 0614 + 09 ( van straaten et al . 2000 ) . in order to compare our results on 4u 172834 with similar results from other sources , we plotted in figure 6 the frequencies of the break of the broad - band noise component ( filled triangles ) and of the lfqpo ( open diamonds ) on a logarithmic scale as a function of the lower khz qpo frequency @xmath84 .
for the first 9 intervals we calculated the frequency of the lower khz qpo by subtracting an assumed constant peak separation of 363 hz from the frequency of the upper khz qpo . the frequency separation between the twin khz qpos in 4u 172834 is known to be constant for @xmath84 between 600 and 800 hz ( mndez & van der klis 1999 ) , and we are here extrapolating this result down to lower frequencies .
we caution that it is unknown whether this peak separation remains constant down to the lowest khz qpo frequencies .
for instance , the relativistic precession model for the khz qpos ( stella & vietri 1999 ) predicts that the peak separation should decrease at low frequencies ( probably below @xmath85 hz , depending on the mass of the compact object ) .
this might explain the deviation of the points at the lowest frequencies from the observed correlations in figure 6 .
a smaller peak separation would move these points towards higher values of @xmath84 , closer to the observed correlations . using the interpretation described above
, we can distinguish three low frequency features , that are plotted in the diagram shown in figure 6 .
the first one is the lfqpo at low accretion rate , that is correlated with the lower khz qpo frequency ( and therefore with the inferred mass accretion rate ) .
as already known ( ford & van der klis 1998 ) , it follows a power law of index @xmath86 ( the solid line drawn in figure 6 ) , the same as observed for the hbos of the z sources ( psaltis et al .
1999a , b .
note that line in fig . 2 of psaltis et al .
1999b has an index of 1.05 ) .
the second feature is the break of the broad - band noise which turns into a lfqpo at high accretion rate .
this feature also seems to be correlated with the frequency of the lower khz qpo .
it follows a power law with index @xmath57 ( the dashed line in fig . 6 ) .
the third feature is the break frequency at high accretion rate , which we identify with the nbo of the z sources and the 7 hz qpo of 4u 182030 .
variabilities such as qpos and broad noise components , have been observed in different kinds of accreting x - ray sources , like z sources , atoll sources and bhcs .
recently some authors ( wijnands & van der klis 1999 ; psaltis et al .
1999b ) have studied correlations between these components in order to understand if they could be explained by a single mechanism .
psaltis et al .
( 1999b ) propose that two kinds of variability can be identified , one similar to the hbo of the z sources and one similar to the khz qpo , both in neutron star and in bhc systems , with different coherences and range of frequencies . the data used by psaltis et al .
( 1999b ) are also shown in figure 6 ( gray points ) . in the case of the bhc systems
a @xmath3 hz qpo is often accompanied by a @xmath87 hz broad noise .
psaltis et al .
( 1999b ) suggest to identify the first qpo with the hbo and the noise with the `` khz '' qpo . when these two kinds of variabilities are plotted one against the other ( pbk diagram ; psaltis et al .
1999b ) a strong correlation appears that seems to extend the correlation observed for the z sources over three orders of magnitude .
in addition , the pbk diagram shows two more branches below this main branch , a second branch described by the bhcs xte 1550564 and gro 165540 , and the atoll source 4u 160852 , and a third branch occupied by the nbo of the z sources . if we compare our results on 4u 172834 with the plot shown in psaltis et al .
( 1999b ) , we can see that our dashed line describing the break frequency at low mass accretion rate and the lfqpo at high mass accretion rate coincides with the second branch in pbk , and that @xmath42 after the jump coincides with the third branch ( see fig . 6 ) .
another general correlation seems to exist between the frequencies of the lfqpo and the break for z sources , atoll sources and bhcs , that is discussed in wijnands & van der klis ( 1999 ) .
these data are plotted in figure 7 ( gray points ) . in order to compare our results with this correlation ( hereafter wk correlation )
we plotted in figure 7 the lfqpo frequency as function of @xmath42 for 4u 172834 ( filled triangles ) .
we see that all the points before the jump follow the wk correlation , whereas the four points after the jump lie at @xmath88 hz and @xmath89 hz in figure 7 , slightly above the main relation , i.e. in the part of the diagram occupied by the z sources .
these correspond to higher accretion rates and follow a similar , but slightly shifted correlation , perhaps because the noise component is different in these sources ( wijnands & van der klis 1999 ) .
the position of these points in the diagram also seems to suggest that , in 4u 172834 after the jump , another kind of variability is present , characterized by a different behavior .
note that with our proposed identification of @xmath42 at low mass accretion rate with the second branch in the pbk diagram , we can now understand the wk correlation as the correlation between the main branch and the second branch in the pbk diagram . by plotting in the pbk diagram the break frequency vs. the lower khz qpo frequency for
other atoll sources this second branch will probably become further populated . to summarize , we find that at low inferred mass accretion rate the frequency of the lfqpo and @xmath42 are well correlated with the frequency of both khz qpos ( and therefore with the inferred mass accretion rate ) .
these frequencies describe two branches in the pbk diagram that are related through the wk relation . at high inferred mass accretion rate the previous lfqpo disappears and the broad - band noise changes into a qpo with a centroid frequency that corresponds to the break frequency of the previous noise . in the pbk diagram
this new lfqpo falls in the same branch as the break frequency before the change .
a new broad - band noise appears at these high accretion rates , characterized by a break frequency around 7 hz .
this could correspond to the nbo observed in z sources , and also in the atoll source 4u 182030 , at high accretion rate .
all these results seem to indicate that there is a tight relation between noise and qpos .
in fact there is evidence that noise could change into a qpo and vice versa .
moreover , their characteristic frequencies show a precise dependence on inferred accretion rate , which shows up as the three branches in the pbk diagram .
the jump in the frequency of the lfqpo can be understood in the framework of the magnetospheric beat frequency model assuming that the neutron star magnetic field is approximately but not exactly dipolar ( e.g. , a dipole offset from the center of the star ) . in this case ,
at lower accretion rates , when the magnetospheric radius is far from the neutron star , the gas in the disk feels primarily the dipolar field .
this is symmetric , and the lfqpo frequency will be double the difference between the orbital frequency at the magnetospheric radius and the spin frequency . at higher accretion rates , when the magnetospheric radius moves closer to the stellar surface , the gas in the disk feels primarily the offset , asymmetric field and the beat frequency will be the difference between the orbital and the spin frequency .
however , a phenomenology where lfqpo can mutate into broad - band noise and vice versa is probably more easily accommodated by models in which the lfqpo arises due to disk fluctuations ( e.g. , titarchuk & osherovich 1999 ; psaltis & norman 1999 ) than by those in which they are produced by orbital motion at the magnetospheric radius ( psaltis et al .
1999a ) or lense - thirring precession of free - orbiting self - luminous blobs at the inner disk whose orbital radius is the upper khz qpo frequency ( stella & vietri 1998 ) , particularly if the khz qpo maintain a high q while at the same time the lfqpo turns into broad noise .
we point out , however , that disk fluctuation models do appear to be able to provide variability at or near the frequencies predicted by such orbital models ( alpar & yilmaz 1997 ; psaltis & norman 2000 ) .
the 100 hz broad bump seems _ not _ to follow these general correlations , but instead to remain relatively constant in frequency at low inferred mass accretion rate . at higher accretion rate
the frequency increases , while the rms amplitude decreases .
we note that this component might be similar to the qpos between 67 and 300 hz that have been found in several bhcs ( morgan , remillard , & greiner 1997 ; remillard et al .
1999a ; remillard et al .
1999b ; homan et al .
a similar qpo near 100 hz is also seen in the atoll source 4u 0614 + 09 ( van straaten et al .
2000 ) .
the authors thank f. lamb and l. burderi for useful and stimulating discussions .
this work was supported by the italian space agency ( asi ) , by the ministero della ricerca scientifica e tecnologica ( murst ) , by the netherlands research school for astronomy ( nova ) , the netherlands organization for scientific research ( nwo ) under contract number 614 - 51 - 002 and the nwo spinoza grant 08 - 0 to e. p. j. van den heuvel .
mm is a fellow of the consejo nacional de investigaciones cientficas y tcnicas de la repblica argentina .
this research has made use of data obtained through the high energy astrophysics science archive research center online service , provided by the nasa / goddard space flight center .
alpar , m. a. , & shaham , j. 1985 , nature , 316 , 239 alpar , m. a. , & yilmaz , a. , 1997 , newa , 2 , 225 dotani , t. , mitsuda , k. , makishima , k. , jones , m. h. , 1989 , pasj , 41 , 577 ford , e. , et al . , 1997a , apj , 486 , l47 ford , e. , et al .
, 1997b , apj , 475 , l123 ford , e. , & van der klis , m. 1998 , apj , 506 , l39 hasinger , g. , & van der klis , m. 1989 , a&a , 225 , 79 hasinger , g. , van der klis , m. , ebisawa , k. , dotani , t. , mitsuda , k. , 1990 , a&a , 235 , 131 homan , j. , wijnands , r. , van der klis , m. , belloni , t. , van paradijs , j. , klein - wolt , m. , fender , r. , mndez , m. , 2000 , apj , submitted ( astro - ph/0001163 ) jonker , p g. , wijnands , r. , van der klis , m. , psaltis , d. , kuulkers , e. , lamb , f. k. , 1998 , apj , 499 , l191 kaaret , p. , yu , w. , ford , e. c. , zhang , s. n. , 1998 , apj , 497 , l93 kaaret , p. , piraino , s. , bloser , p. f. , ford , e. c. , grindlay , j. e. , santangelo , a. , smale , a. p. , zhang , w. , 1999 , apj , 520 , l37 kuulkers , e. , van der klis , m. , oosterbroek , t. , asai , k. , dotani , t. , van paradijs , j. , lewin , w. h. g. , 1994 , a&a , 289 , 795 lamb , f. k. , shibazaki , n. , alpar , m. a. , shaham , j. 1985 , nature , 317 , 681 leahy , d. , et al .
1983 , apj , 266 , 160 mndez , m. , van der klis , m. , wijnands , r. , ford , e. c. , van paradijs , j. , vaughan , b. a. 1998 , apj , 505 , l23 mndez , m. , van der klis , m. , van paradijs .
j. 1998 , apj , 506 , l117 mndez , m. , & van der klis , m. 1999 , apj , 517 , l51 mndez , m. 2000 , proc .
19th texas symposium on relativistic astrophysics and cosmology , ed .
j. paul , t. montmerle , & e. aubourg ( amsterdam : elsevier ) , 15/16 ( astro ph/9903469 ) miller , m. c. , lamb , f. k. , psaltis , d. 1998 , apj , 508 , 791 morgan , e. h. , remillard , r. a. , greiner , j. , 1997 , apj , 482 , 993 olive , j. f. , barret , d. , boirin , l. , grindlay , j. e. , swank , j. h. and smale , a. p. 1998 , a&a 333 , 942 priedhorsky , w. , hasinger , g. , lewin , w. h. g. , middleditch , j. , parmar , a. , stella , l. , white , n. , 1986 , apj , 306 , l91 psaltis , d. , et al . 1999a , apj , 520 , 763 psaltis , d. , belloni , t. , & van der klis , m. 1999b , apj , 520 , 262 psaltis , d. , norman , c. , 2000 , apj , in press reig , p. , mndez , m. , van der klis , ford , e. c. 2000 , apj , 530 , 916 remillard , r. a. , morgan , e. h. , mcclintock , j. e. , bailyn , c. d. , orosz , j. a. , 1999a , apj , 522 , 397 remillard , r. a. , et al .
, 1999b , apj , 517 , 127 stella , l. , & vietri , m. 1998 , apj , 492 , l59 stella , l. , & vietri , m. 1999 , phys .
, 82 , l17 strohmayer , t. e. , swank , j. h. , zhang , w. 1998 , apj , 503 , l147 strohmayer , t. , zhang , w. , swank , j. 1996 , iau circ .
6319 strohmayer , t. e. , zhang , w. , swank , j. h. , smale , i. , titarchuk , l. , day , c. , lee , u. 1996 , apj , 469 , l9 strohmayer , t. e. , zhang , w. , swank , j. h. 1997 , apj , 487 , l77 titarchuk , l. , & osherovich , v. , 1999 , apj , 518 , l95 titarchuk , l. , lapidus , i. , muslimov , a. 1998 , apj , 499 , 315 van der klis , m. 2000 , ann.rev.a&a , in press ( astro - ph/0001167 ) van der klis , m. , wijnands , r. a. d. , horne , k. , chen , w. 1997 , apj , 481 , l97 van der klis , m. 1995 , in _ x - ray binaries _ , lewin w. h. g. , van paradijs j. , van den heuvel e. p. j. eds . , cambridge astrophysics series , cambridge , p. 252
van der klis , m. , 1994 , a&a , 283 , 469 van straaten , s. , ford , e. c. , van der klis , m. , mndez , m. , kaaret , p. 2000 , apj , in press wijnands , r. , van der klis , m. , rijkhorst , e. 1999 , apj , 512 , l39 wijnands , r. , & van der klis , m. 1999 , apj , 514 , 939 wijnands , r. , mndez , m. , van der klis , m. , psaltis , d. , kuulkers , e. , lamb , f. k. 1998 , apj , 504 , l35 yu , w. , et al . , 1997 ,
apj , 490 , l153 zhang , w. , jahoda , k. , kelley , r. l. , strohmayer , t. e. , swank , j. h. , zhang , s. n. , 1998 , apj , 495 , l9 c|c|c start time & total time & averaged rate + ( utc ) & ( ksec ) & ( c / s ) + 15/02/96 11:51 & 25 & 1863 + 15/02/96 18:59 & 18 & 1948 + 16/02/96
00:03 & 20 & 2060 + 16/02/96 06:26 & 15 & 2172 + 16/02/96 15:49 & 21 & 2160 + 16/02/96 22:15 & 10 & 2199 + 18/02/96 11:10 & 25 & 1609 + 18/02/96 19:15 & 15 & 1516 + 22/02/96 11:33 & 25 & 996 + 22/02/96 19:12 & 15 & 997 + 23/02/96 21:16 & 30 & 978 + 24/02/96 05:17 & 5 & 1090 + 24/02/96 17:51 & 25 & 1022 + 25/02/96 01:15 & 12 & 1034 + 25/02/96 20:51 & 25 & 1080 + 26/02/96 04:02 & 15 & 1093 + 29/02/96 23:12 & 0.8 & 1200 + 01/03/96 00:13 & 20 & 1193 + 03/05/96 14:03 & 2.1 & 1126 + 23/09/97 23:50 & 1.2 & 2190 + 24/09/97 09:14 & 25 & 2582 + 24/09/97 17:15 & 4 & 2422 + 26/09/97 12:30 & 25 & 1775 + 27/09/97 09:19 & 25 & 1506 + 30/09/97 04:34 & 15 & 1799 + 30/09/97 10:57 & 8 & 1793 + 01/10/97 06:09 & 25 & 1659 + 01/10/97 14:10 & 3.5 & 1683 + & & & & + interval & rate & frequency & fwhm & rms & frequency & fwhm & rms & frequency & fwhm & rms + number & ( c / s ) & ( hz ) & ( hz ) & ( % ) & ( hz ) & ( hz ) & ( % ) & ( hz ) & ( hz ) & ( % ) + 1 & 1315 & 387@xmath9018 & 230@xmath9073 & [email protected] & - & - & - & 64@xmath9037 & 350@xmath90110 & [email protected] + 2 & 1285 & 397@xmath9014 & 270@xmath9037 & [email protected] & - & - & - & 110@xmath9020 & 269@xmath9061 & [email protected] + 3 & 1209 & 466@xmath9021 & 295@xmath9052 & [email protected] & - & - & - & 163@xmath9015 & 205@xmath9061 & [email protected] + 4 & 1156 & [email protected] & 282@xmath9017 & [email protected] & - & - & - & [email protected] & 164@xmath9014 & [email protected] + 5 & 1155 & [email protected] & 287@xmath9018 & [email protected] & - & - & - & [email protected] & 183@xmath9016 & [email protected] + 6 & 1151 & [email protected] & 219@xmath9026 & [email protected] & - & - & - & [email protected] & 220@xmath9033 & [email protected] + 7 & 1375 & [email protected] & 206@xmath9033 & [email protected] & - & - & - & 153@xmath9011 & 118@xmath9047 & [email protected] + 8 & 1510 & [email protected] & [email protected] & [email protected] & - & - & - & 106@xmath9021 & 211@xmath9038 & [email protected] + 9 & 1605 & [email protected] & [email protected] & [email protected] & - & - & - & 72@xmath9024 & 239@xmath9027 & [email protected] + 10 & 1722 & [email protected] & [email protected] & [email protected] & 511@xmath9017 & 98@xmath9048 & [email protected] & [email protected] & 190@xmath9015 & [email protected] + 11 & 1739 & [email protected] & [email protected] & [email protected] & 559@xmath9011 & 88@xmath9032 & [email protected] & [email protected] & [email protected] & [email protected] + 12 & 1732 & [email protected] & [email protected] & [email protected] & 599@xmath9014 & 139@xmath9030 & [email protected] & [email protected] & 149@xmath9012 & [email protected] + 13 & 1865 & [email protected] & 158@xmath9012 & [email protected] & 673@xmath9010 & 191@xmath9033 & [email protected] & [email protected] & 98@xmath9010 & [email protected] + 14 & 1871 & [email protected] & 166@xmath9027 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & 87@xmath9014 & [email protected] + 15 & 1907 & [email protected] & 148@xmath9019 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & 94@xmath9016 & [email protected] + 16 & 1861 & 1129@xmath9013 & 117@xmath9022 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & 102@xmath9016 & [email protected] + 17 & 1939 & 1158@xmath9018 & 129@xmath9056 & [email protected] & [email protected] & [email protected] & [email protected] & 210@xmath9032 & 180@xmath9074 & [email protected] + 18 & 2500 & - & - & - & - & - & - & - & - & - + 19 & 2600 & - & - & - & - & - & - & - & - & - + ccccccccccc & & & & + interval & @xmath18 & break or cutoff & @xmath19 & rms & frequency & fwhm & rms & @xmath91 & rms & @xmath92 + number & @xmath93 & ( hz ) & & ( % ) & ( hz ) & ( hz ) & ( % ) & & ( % ) & + 1 & @[email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & - & - & 241 ( 131 ) + 2 & @[email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & - & - & 273 ( 131 ) + 3 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & - & - & 270 ( 131 ) + 4 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & - & - & 587 ( 131 ) + 5 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & - & - & 538 ( 131 ) + 6 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & - & - & 267 ( 131 ) + 7 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & - & - & 219 ( 134 ) + 8 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & - & - & 282 ( 131 ) + 9 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & - & - & 254 ( 131 ) + 10 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & - & - & 275 ( 131 ) + 11 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & - & - & 241 ( 131 ) + 12 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & - & - & 192 ( 131 ) + 13 & @[email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & - & - & 181 ( 134 ) + 14 & @[email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & 203 ( 132 ) + 15 & @[email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & 305 ( 132 ) + 16 & @[email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & 348 ( 132 ) + 17 & @xmath99@xmath9019 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & 191 ( 132 ) + 18 & - & [email protected] & @[email protected] & [email protected] & - & - & - & [email protected] & [email protected] & 179 ( 139 ) + 19 & - & [email protected] & @[email protected] & [email protected] & - & - & - & [email protected] & [email protected] & 184 ( 139 ) + fig .
representative leahy normalized power spectra of 4u 172834 corresponding to different intervals in the color - color diagram of figure 1 .
interval numbers are indicated . for each interval the power spectrum and the best fit model ( including always at most one lfqpo , see text ) are shown in the upper panel , and the corresponding residuals ( in units of @xmath24 ) are shown in the lower panel .
[ fig2a ] fig . 3a .
frequencies ( upper panel ) and rms amplitudes ( lower panel ) of the khz qpos as a function of the interval number in the color - color diagram .
different symbols are used for the lower peak ( filled triangles ) and the upper peak ( open diamonds ) .
open squares represent intervals where we only detected one of the khz qpos .
[ fig3a ] fig .
parameters of the hfn and lfqpo as a function of the frequency of the upper khz qpo .
different symbols are used for the hfn ( filled triangles ) and lfqpo ( open diamonds ) .
upper panel : break frequency of the hfn and centroid frequency of the lfqpo .
lower panel : rms amplitudes of the hfn ( calculated from 0 hz to infinity ) and of the lfqpo .
[ fig5 ] fig . 6 .
frequencies of the lfqpo ( open diamonds ) and of the break ( filled triangles ) as a function of the frequency of the lower khz qpo , @xmath84 . in the first 9 intervals @xmath84 is calculated by subtracting the assumed constant peak separation of 363 hz from the frequency of the upper khz qpo that has been detected . for comparison ,
data for several lmxbs ( including z sources , atoll sources and black hole candidates , from psaltis et al .
1999b ) are also shown ( gray points ) .
correlations between these frequencies are visible : the solid line is a power law of index 0.95 ( psaltis et al .
1999b ) and the dashed line is a power law of index 1.5 .
[ fig6 ] fig . 7 .
centroid frequency of the lfqpo as a function of the break frequency of the hfn ( filled triangles ) . for comparison , data for several lmxbs ( including z sources , atoll sources and black hole candidates , from wijnands & van der klis 1999 ) are also shown ( gray points ) .
these frequencies follow the wk correlation , except for four points at @xmath88 hz and @xmath102 hz , corresponding to frequencies of the upper khz qpo higher than 900 hz . | we study the timing properties of the bursting atoll source 4u 172834 as a function of its position in the x - ray color - color diagram . in the island part of the color - color diagram ( corresponding to the hardest energy spectra ) the power spectrum of 4u
172834 shows several features such as a band - limited noise component present up to a few tens of hz , a low frequency quasi - periodic oscillation ( lfqpo ) at frequencies between 20 and 40 hz , a peaked noise component around 100 hz , and one or two qpos at khz frequencies .
in addition to these , in the lower banana ( corresponding to softer energy spectra ) we also find a very low frequency noise ( vlfn ) component below @xmath0 hz . in the upper banana ( corresponding to the softest energy spectra ) the power spectra
are dominated by the vlfn , with a peaked noise component around 20 hz .
we find that the frequencies of the khz qpos are well correlated with the position in the x - ray color - color diagram .
for the frequency of the lfqpo and the break frequency of the broad - band noise component the relation appears more complex .
these frequencies both increase when the frequency of the upper khz qpo increases from 400 to 900 hz , but at this frequency a jump in the values of the parameters occurs .
we interpret this jump in terms of the gradual appearance of a qpo at the position of the break at high inferred mass accretion rate , while the previous lfqpo disappears .
simultaneously , another kind of noise appears with a break frequency of @xmath1 hz , similar to the nbo of z sources .
the 100 hz peaked noise does not seem to correlate with the position of the source in the color - color diagram , but remains relatively constant in frequency .
this component may be similar to several 100 hz qpos observed in black hole binaries . |
nearly all low - mass stars are believed to have magnetic fields that influence their surroundings and evolution .
young stars with ages less than 1020 myr exhibit high - energy activity in the form of hot coronal loops , flares , accretion shocks , and open - field regions with winds or jet - like outflows ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . because the magnetic fields of the star and disk are threaded together with one another , it is often difficult to disentangle the contributions from various proposed sources of activity . on the other hand ,
the situation may be greatly simplified for older stars that have evolved past the classical t tauri phase .
these stars have lost their dense gas disks , and thus the major remaining contributor to the star s ultraviolet and x - ray emission is the presence of magnetic coronal loops . in that case , the subsequent evolution of a star s _ coronal heating _ as it starts main sequence hydrogen burning may no longer involve drastic changes in the source regions , but instead just be the result of a gradual evolution in parameters related to its magnetohydrodynamic ( mhd ) dynamo @xcite .
pre - main - sequence stars with debris disks are ideal targets for studying the earliest stages of non - accretion - driven magnetic activity . in this paper
we focus on au microscopii ( hd 197481 , gj 803 ) , a nearby m1ve flare star with a well - resolved debris disk .
having an effective temperature of @xmath03500 k , au mic probably is not fully convective like the later - type m dwarfs .
thus , its mhd dynamo may be qualitatively similar to those of more massive stars like the sun .
au mic is bright in x - rays @xcite , rich in ultraviolet flaring phenomena @xcite , and surrounded by an edge - on dust disk with a mass of roughly 1 @xmath1 and a spatial extent similar to the kuiper belt in our solar system ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
recently , @xcite observed au mic at millimeter wavelengths with the atacama large millimeter / submillimeter array ( alma ) .
in addition to the kuiper - like dust belt , they were able to distinguish a compact central emission peak with a flux of @xmath0320 @xmath2jy .
the origin of this component is not yet known .
the m dwarf s stellar photosphere would generate a blackbody flux of only about 60 @xmath2jy at @xmath3 mm , a factor of 5 smaller than the observed emission peak . in order for a star - sized blackbody to be responsible for the observed emission , it would need to have a temperature of roughly 17,000 k. this kind of dominant chromospheric emission was predicted by @xcite for cool evolved giants , but those stars have much larger emitting areas and no significant coronal emission .
an m dwarf like au mic may have a thin chromosphere underneath its hot corona , but its optical depth is not likely to be high enough to generate a photosphere - like blackbody spectrum .
@xcite suggested that the central emission peak of au mic may be produced by an inner `` asteroid belt '' of cool dust grains or planetesimals within @xmath03 au of the star .
only about 0.01 @xmath1 worth of dust material needs to be present to generate the observed emission peak ; this is comparable to the mass of the asteroid belt in our solar system . @xcite
also computed upper limits on the required temperatures of silicate grains in this proposed belt and found values of 3575 k. this range is the same order of magnitude as the expected dust temperatures at a distance of @xmath4 au from an m dwarf . in this paper
we propose an alternate explanation for the alma central emission component of au mic .
since this star has such strong x - ray emission , we explored the possibility that a collection of _ hot coronal loops _ on its surface could be responsible for similarly strong thermal emission at 1.3 mm .
figure [ fig01 ] illustrates the suggested mhd circumstellar environment , where magnetic loops with a continuous distribution of sizes are assumed to fill the corona with @xmath5 k plasma .
we aim to create a single model of these loops that reproduces the available observations at x - ray and millimeter wavelengths .
@xcite also explored the use of radio continuum measurements as constraints to models of m dwarf coronal heating . in a way
, this paper follows on from the general ideas described by @xcite , but our models are based on self - consistent physical sources of plasma heating inspired by recent advances in understanding the sun s well - resolved corona . in section [ sec : model ] of this paper , we describe the model of coronal heating that we apply to au mic .
section [ sec : radio ] summarizes how we synthesize observable quantities in the x - ray , radio , and millimeter wavelength bands for later comparison with observations . in section [ sec : results ] we present the computed coronal loop parameters for au mic as well as synthetic x - ray luminosities and radio / mm spectra . section [ sec : conc ] concludes the paper with a brief summary of the main results and suggestions for future improvements .
in this section we describe a model of turbulence - driven coronal activity motivated by the present - day sun .
the model adopts a time - steady description of coronal heating and does not attempt to simulate the full range of intermittent and chaotic mhd dynamics that is evident in high - resolution solar images ( e.g. , * ? ? ?
nonetheless , such a phenomenological description of turbulence has been shown to accurately reproduce many properties of coronal heating and wind acceleration for the sun @xcite .
also , @xcite showed how this model performs well in predicting the x - ray activity of young stars undergoing active accretion , and @xcite used the model to compute the mass - loss rates of stellar winds for stars with a wide range of ages and masses .
the basic idea is that magnetic field lines at the stellar photosphere are continually jostled by a stochastic source of mechanical energy , and this gives rise to mhd waves that propagate up to larger heights and eventually dissipate . in tandem with wave generation ,
the random footpoint motions produce an increase in the overall magnetic energy via the twisting , shear , and braiding of field lines ( e.g. , * ? ? ?
the transport and dissipation of magnetic energy from both waves and braiding can be described using the unifying language of _ turbulent cascade _ @xcite .
the presence of radiative cooling in such a system produces a thermal instability that often leads to the coexistence of a cool ( @xmath6 k ) chromosphere and a hot ( @xmath7 k ) corona . in the ionized corona
, heat conduction also makes for a `` thermostat '' effect in which the time - steady spatial distribution of @xmath8 is much smoother than the spatial distribution of the heating rate .
there are several possible sources of stochastic mechanical energy at the photospheric base .
convective granulation surely plays a major role for most low - mass stars @xcite . for stars undergoing active accretion
, an additional source of energy can exist in the form of ripples from the impact of gas clumps falling down from above @xcite .
it has also been suggested that the presence of planets in a magnetized stellar wind may give rise to mhd fluctuations that propagate back down to the star ( e.g. , * ? ? ?
for the purposes of this paper , we will treat the total available mechanical energy flux at the photosphere as a free parameter .
we describe the star itself by its fundamental parameters ( mass , radius , luminosity , and metallicity ) and ignore rotation .
the dynamics of the jostled magnetic flux tubes are described by three additional parameters defined at the photosphere : mass density @xmath9 , magnetic field strength @xmath10 , and the mean energy flux @xmath11 of alfvn waves that propagate upwards as a result of the stochastic motion . to compute the density as a function of effective temperature @xmath12 and gravity @xmath13 , we used the standard criterion that the rosseland mean optical depth @xmath14 should have a value of @xmath15 in the photosphere , which demands @xmath16 the rosseland mean opacity @xmath17 was interpolated from tables given by @xcite , and we defined the density scale height in the photosphere as @xmath18 where @xmath19 is boltzmann s constant and @xmath20 is the mass of a hydrogen atom .
@xcite found that , for cool - star photospheres , the mean atomic weight @xmath2 is a relatively slowly varying function of @xmath12 ; we used their parameterization . to specify the photospheric magnetic field strength
, we assumed that @xmath10 remains linearly proportional to the so - called equipartition field strength ( i.e. , the value at which magnetic pressure balances gas pressure ) .
we used the proportionality constant found from observations by @xcite to specify @xmath21 for the footpoints of coronal loops .
the alfvn wave energy flux can be written in terms of the surface velocity amplitude @xmath22 of transverse mhd waves , with @xmath23 , and the alfvn speed is given by @xmath24 .
we can make an initial estimate of the expected velocity amplitude using turbulent convection models .
@xcite used the models of @xcite for main sequence and evolved stars .
we find below , however , that for au mic the @xcite models predict an amplitude that is far too small to explain its observed coronal heating and x - ray activity .
( our understanding of the magnetic dynamos of young , rapidly rotating stars is still incomplete ; see section [ sec : conc ] . ) in general , we assume that magnetic flux tubes fill only a fraction @xmath25 of the stellar photosphere , so that the mean magnetic flux density of the star would be given by @xmath26 . for the sun , @xmath25 varies between about @xmath27 and @xmath28 over location and activity cycle @xcite . the local magnetic field strength @xmath29 inside a flux tube drops rapidly from @xmath10 to @xmath30 as a function of increasing height .
however , for young stars in the `` saturated '' part of the age activity
rotation diagram , @xcite and @xcite found that @xmath31 is not a bad approximation to make .
we will assume this for au mic as well ( see additional discussion in section [ sec : results ] ) , and this also allows us to assume that @xmath10 remains constant over each coronal loop .
we note that @xmath25 describes the filling factor of all magnetic flux tubes , of both polarities , that pass through the stellar surface . by itself
, our assumption of @xmath31 does not directly imply a prediction for the fraction of the magnetic field lines that are open ( presumably to the stellar wind ) versus those that are are closed .
the open / closed fraction depends on both the overall level of gas pressure in the stellar wind and the spatial patterns of imbalance between flux tubes with positive and negative polarities @xcite .
below , we do assume the entire surface is covered by loops of varying lengths , but the coronal heating and thus the x - ray and radio / mm emission from the longest loops is likely to be not too different from the emission from open - field regions ( see , e.g. , * ? ? ?
to specify the rate of plasma heating , we use a general expression for the rate of energy flux in the cascade from large to small eddies that was derived from analytic and numerical studies of mhd turbulence .
following @xcite , we define the heating rate ( in units of power generated per unit volume ) as @xmath32 and @xmath33 .
the quantity @xmath34 is a transverse correlation length for the largest driving eddies , and is typically found to be of the same order of magnitude as the radius of the flux tube ( e.g. , * ? ? ?
* ) . by convention
, @xmath35 is the half - length of the coronal loop .
nonzero values of the exponent @xmath36 describe departures from an ideal @xcite hydrodynamic cascade ; such departures occur because of the specific mhd nature of wave - packet collisions that comprise the eddies in the presence of a strong background field . in closed loops
, @xmath36 can vary between 1.5 and 2 .
our expression for the dependence of @xmath36 on the wave amplitude @xmath22 and alfvn speed @xmath37 was derived from the models of @xcite . even though equation ( [ eq : qturb ] )
utilizes constant values for @xmath22 , @xmath34 , @xmath35 , and @xmath10 ( in the definition of @xmath37 ) , the density @xmath38 is known to decrease rapidly as a function of height in a stellar atmosphere .
thus , because @xmath39 , the heating rate itself varies with density as @xmath40 .
for @xmath41 the heating rate is constant ( as was assumed by * ? ? ? * ) , and for @xmath42 the heating is stronger at the high - density footpoints .
the values of @xmath38 and @xmath37 at the `` coronal base '' ( i.e. , just above the sharp transition region between chromosphere and corona ) are not the same as the corresponding values in the photosphere ; these quantities are computed self - consistently by the model and are not specified as inputs . in order to determine the time - steady coronal response to a given heating rate @xmath43
, we must consider how the injected heat is transported along the field by conduction and lost by radiation .
we use the energy balance model of @xcite to solve for the temperature @xmath8 and electron number density @xmath44 as a function of distance along a given loop of half - length @xmath35 ( see also * ? ? ?
* ; * ? ? ?
these quantities are specified by first computing the peak temperature @xmath45 at the top of the loop and the base pressure @xmath46 .
@xcite parameterized the heating rate as @xmath47 , where @xmath48 and @xmath46 are assumed to be constants as a function of distance . for the heating model of equation ( [ eq : qturb ] ) ,
the exponents are given by @xmath49 .
@xcite derived two scaling laws that allow one to solve for any two of the four loop quantities ( @xmath35 , @xmath45 , @xmath46 , @xmath48 ) if the other two are provided . in our case , these scaling laws are given by @xmath50 and the unspecified constants of proportionality are described in detail by @xcite .
these constants depend on the exponent @xmath36 , the heat conductivity coefficient @xmath51 erg @xmath52 s@xmath53 k@xmath54 , and the assumed properties of radiative cooling . for the latter
, we assumed @xmath55 , with a value of @xmath56 @xmath57 s@xmath53 k@xmath58 , which we enhanced slightly over the solar value because of the higher metallicity of au mic ( see equation 21 of * ? ? ? * ) .
in order to model the thermal state of a given loop with half - length @xmath35 , we use equation ( [ eq : qturb ] ) to determine @xmath48 in terms of the other input parameters , and we use the scaling laws given in equation ( [ eq : scale ] ) to solve for @xmath45 and @xmath46 .
the density at the coronal base is determined mainly by @xmath46 , but it needs to be specified initially in order to compute @xmath37 and @xmath36 in equation ( [ eq : qturb ] ) .
thus , we solve for these quantities iteratively by starting with the initial guess that the density is given by @xmath9 .
cycling back through the equations gives rise to a converged and self - consistent value of @xmath38 ( and thus also @xmath37 and @xmath36 ) typically within 1015 iterations , and we run it for a total of 50 iterations to ensure accuracy .
this converged value of the coronal base density is often 36 orders of magnitude smaller than @xmath9 , which produces typical basal @xmath44 values of @xmath59@xmath60 @xmath61 .
once @xmath45 and @xmath46 are known for a given loop , we solved for the spatial dependence of temperature and density along the loop length @xmath62 .
@xcite found that the temperature dependence @xmath63 can be written as the inverse of an incomplete beta function ; we used equation ( 25 ) of @xcite to obtain @xmath63 , and then we combined it with the ideal gas equation of state to solve for @xmath64 as a function of @xmath46 and @xmath63 .
the minimum value of @xmath8 at the coronal base was fixed at @xmath65 k.
lastly , we must take account of the fact that there should be a continuous distribution of loop lengths @xmath35 across the surface of the star .
we specify a normalized probability distribution @xmath66 that describes the chances of finding a loop at any given value of @xmath35 at a random location on the star .
the shape of @xmath66 depends on many details about the strength and topology of the magnetic field . the large - scale magnetic geometry of m dwarfs is beginning to be understood observationally ( e.g. , * ? ? ?
* ) , but we know from the sun that the x - ray and uv emission is dominated by activity on spatial scales much smaller than have been resolved so far on other stars . thus , we follow @xcite and assume a power - law probability distribution , @xmath67 , with sharp cutoffs below minimum and maximum loop lengths @xmath68 and @xmath69 .
the exponent @xmath70 in the loop - length distribution is another key free parameter of this model .
the shortest loops are defined geometrically as those for which the central hole in the torus shrinks to zero ; i.e. , @xmath71 , where @xmath72 is the modeled poloidal radius of the loop ( see section [ sec : results ] ) .
the longest loops are assumed to have lengths of order the stellar radius .
we note that some young stars may have loops that extend out to even larger heights @xcite . nonetheless , it is not known what fraction of the very longest loops remain stable and closed in the presence of a stellar wind .
thus , we set @xmath73 and assume that the coronal emission is relatively insensitive to the details of what happens for the small fraction of the star covered by the longest loops .
to simulate a full distribution of loops covering the surface of au mic , we constructed 100 coronal models on a logarithmic grid in @xmath35 that spans the @xmath03.5 orders of magnitude between @xmath68 and @xmath69 .
we synthesized the observable quantities described below for the 100 models individually assuming for each that the star is filled with loops of a given length
then we convolved them together using @xmath66 to produce a result that is weighted properly over the statistical distribution of loop sizes .
we computed the total x - ray luminosity @xmath74 under the optically thin assumption that all radiation escapes from the emitting regions . to maintain continuity with past observations of cool - star x - rays
, we chose to use the response function of the _ rosat _ position sensitive proportional counter ( pspc ) given by @xcite .
this function has nonzero sensitivity between about 0.1 and 2.4 kev , with a minimum around 0.3 kev that separates the hard and soft bands .
@xcite presented the temperature - dependent radiative loss rate @xmath75 that is consistent with this response function , which is then used to estimate the x - ray luminosity , @xmath76 \ , \lambda_{\rm x}[t(z ) ] \,\ , .\ ] ] we integrated down through the length of each loop , from @xmath77 to @xmath78 , under the simplifying geometric assumption that the loop is oriented vertically .
in other words , we assumed that each loop has been `` snipped in two '' and the two ends are pointing radially upward .
we also ignored foreshortening effects that would alter the path lengths through loops close to the stellar limb . calculating
the emission at radio and millimeter wavelengths was slightly more complicated than the x - ray luminosity because we can no longer assume an optically thin emitting region .
thus , we solved the diffuse term in the formal solution to the equation of radiative transfer by integrating both the optical depth @xmath79 and the specific intensity @xmath80 simultaneously , with @xmath81 @xmath82 and we made the standard assumption that the source function is given by the rayleigh - jeans tail of the local planck function . because the plasma conditions vary rapidly as a function of position @xmath83 along the loop , we did _ not _ make the other standard assumption that the source function is constant over the emitting area . as above , we integrated down from @xmath77 to @xmath78 , so that the resulting optical depth increases as @xmath83 decreases . for wavelengths @xmath84 between 0.01 and 100 cm , we assumed the radiation is dominated by thermal free - free emission ( bremsstrahlung ) and that the opacity is given by @xmath85 in cgs units ( see * ? ? ?
* ; * ? ? ?
for the gaunt factor ( i.e. , the natural logarithm term above ) , we used a fully ionized approximation that should apply for @xmath86 k. equation ( [ eq : freefree ] ) also assumes @xmath87 , where @xmath88 is plasma frequency in hz ; this condition is satisfied easily for the alma submillimeter wavelengths of interest and is satisfied marginally for longer radio wavelengths . for a star at distance @xmath89
, we converted the surface - averaged specific intensity @xmath80 to flux @xmath90 by assuming that short loops ( with @xmath91 ) generally dominate the emission and that there is no limb brightening or limb darkening .
thus , we used @xmath92 .
table 1 gives the basic stellar parameters that we used either as inputs ( upper part ) or as after - the - fact constraints ( lower part ) on the models .
the photospheric density @xmath9 and magnetic field strength @xmath10 were computed using the equations discussed above .
the photospheric turbulence length scale @xmath34 and flux tube radius @xmath72 were scaled down from canonical solar values of 300 km and 200 km , respectively , by the ratio of photospheric scale heights .
the solar value of @xmath34 comes from models of alfvn wave damping in the fast solar wind @xcite and @xmath72 comes from measurements of magnetic bright points @xcite ; both are related closely to the horizontal sizes of magnetic features sitting between the granulation cells .
recent convection models ( e.g. , * ? ? ?
* ; * ? ? ?
* ) show that the diameters of granules remain linearly proportional to the vertical scale height @xmath93 over a wide range of stellar parameters . for additional discussion of how rapid rotation may invalidate some traditional ideas about convection and magnetic activity .
] we assume that intergranular features like magnetic flux tubes scale with @xmath93 similarly as the cells . with the values given above
, we found that the coronal loops of au mic span about 3.5 orders of magnitude in length between @xmath94 km and @xmath95 km .
lll [ table01 ] distance & 9.91 pc & @xcite + @xmath96 & 0.5 @xmath97 & @xcite + @xmath98 & 0.838 @xmath99 & @xcite + @xmath12 & 3493 k & @xcite + @xmath100}$ ] & 0.154 & @xcite + @xmath9 & @xmath101 g @xmath61 & equation ( [ eq : taur ] ) + @xmath10 & 2200 g & equation ( [ eq : beq ] ) + @xmath34 & 182 km & see text + @xmath72 & 121 km & see text + @xmath74 & @xmath102 erg s@xmath53 & @xcite + @xmath90 ( @xmath3 mm ) & 320 @xmath103 60 @xmath2jy & @xcite + @xmath90 ( @xmath104 cm ) & @xmath105 210 @xmath2jy & @xcite + @xmath90 ( @xmath106 cm ) & @xmath105 120 @xmath2jy & @xcite + for au mic , we assumed the photospheric magnetic filling factor @xmath25 is equal to 1 .
au mic appears to be safely inside the `` saturated '' regime of stellar activity that is consistent with this assumption .
its rotation period @xmath107 days implies a dimensionless rossby number ( i.e. , @xmath108 , where @xmath109 is the convective turnover time ) of 0.050.1 @xcite .
figure 7(b ) of @xcite shows @xmath110 for a large sample of cool stars in this range of rossby number .
au mic also has an x - ray luminosity ratio @xmath111 of 0.0015 , which is in the saturated part of the empirical age activity
rotation diagram ( e.g. , * ? ? ?
@xcite estimated the filling factor of `` solar - like active region '' plasma in the corona of au mic to be roughly 0.91 .
although this measurement does not constrain the _
photospheric _ value of @xmath25 , it helps to show that the coronal flux tubes eventually fill the entire circumstellar volume .
the one remaining major parameter that is not yet determined for au mic is the surface flux of alfvn waves .
@xcite used the models of @xcite to produce a fitting formula that gives this surface flux as a function of a star s @xmath12 and @xmath112 .
for the parameters of au mic this formula gives @xmath113 erg @xmath114 s@xmath53 , which is equivalent to @xmath115 km s@xmath53 . on the other hand , both observations of microturbulent broadening @xcite and three - dimensional convection simulations @xcite
show vigorous overturning motions with velocities of order 12 km s@xmath53 for m dwarfs . for the sun ,
the granulation velocities and surface wave amplitudes are roughly of the same order of magnitude as one another , but we do not know if this holds true for m dwarfs .
thus , we treat @xmath22 as a free parameter , but we also note that values of order 12 km s@xmath53 may be the most realistic .
we computed models of coronal heating for a grid of loops having lengths @xmath35 between @xmath116 and @xmath117 km , and footpoint wave amplitudes @xmath22 between 0.02 and 200 km s@xmath53 .
figure [ fig02 ] shows the resulting dependence of @xmath45 and @xmath46 on these two parameters .
each of these models had its own iterated value of the @xmath36 exponent , and the largest values ( @xmath118 ) tended to occur for the lowest velocity amplitudes and the longest loops ; the smallest values ( @xmath119 ) occurred for the largest amplitudes and shortest loops .
the overall dependence of @xmath45 and @xmath46 on the two varied parameters is close to what is expected from the scaling laws given in section [ sec : model ] .
for example , if we use the average value of @xmath120 for this set of models , the scaling laws give @xmath121 which agrees well with the plotted results .
the curves shown in figure [ fig02 ] are not exact power laws because @xmath36 itself is not constant . in order to begin the process of testing our coronal models against real observations of au mic
, we assembled together the temperature and density distributions for individual loop lengths into surface - averaged models using the power - law length distribution @xmath67 discussed above .
various measurements of solar features have constrained the value of @xmath70 to be of order 22.5 @xcite , and the @xcite study of t tauri star x - ray emission also found that @xmath122 produced the most reasonable results . for completeness
, we varied @xmath70 between 0 and 2.5 , noting that values of the exponent greater than 2.5 produce distributions that are only marginally different from those with @xmath122 ( i.e. , once it is peaked sharply at the shortest length scales , making it even steeper does not change the resulting weighting significantly ) .
figure [ fig03](a ) shows contours of @xmath74 computed for the _ rosat _ pspc band , and for a large grid of 250 @xmath22 values by 250 @xmath70 values .
amplitudes in the grid because there were no combinations of parameters in this region that produced agreement with both the x - ray and millimeter observations .
the @xmath123 grid encompasses the ranges @xmath124 km s@xmath53 and @xmath125 . ] for any single value of @xmath70 , the x - ray luminosity first increases with increasing @xmath22 , because the temperatures and densities in the loops are also increasing , but then @xmath74 begins to decrease because the loops become too hot to produce significant emission in the relatively soft pspc passband .
the dotted red contour in figure [ fig03](a ) denotes the parameters that produce agreement with the observed value of @xmath126 @xcite .
the contours in figure [ fig03](b ) show constant values of the free - free emission flux @xmath90 computed at @xmath3 mm .
the thickest contour highlights the alma flux measured by @xcite for the central emission component of au mic .
the corresponding observational contour for @xmath74 is reproduced from figure [ fig03](a ) , and it is noteworthy that the two empirical constraints intersect one another ( roughly perpendicularly ) at a single parameter value .
this best - fitting model has a basal velocity amplitude @xmath127 km s@xmath53 and a loop distribution power - law exponent @xmath128 .
we note also that this model has a value of @xmath22 that falls within the region of likely microturbulence values ( 12 km s@xmath53 ) for early m - type dwarfs @xcite .
the best - fitting model contains loops with peak temperatures between 3.7 mk ( for @xmath68 ) and 7.2 mk ( for @xmath69 ) .
@xcite produced a fit of the full _ chandra _ spectrum of au mic with three plasma components having temperatures of 3.37 , 7.78 , and 17.3 mk .
our range of modeled @xmath45 values is reasonably consistent with the first two _ chandra _ components .
the hottest observed x - ray component may be related to intermittent nonthermal stellar flaring , which we did not include in our model .
the best - fitting model also exhibits loop - top electron densities between @xmath129 @xmath61 ( for @xmath68 ) and @xmath130 @xmath61 ( for @xmath69 ) , with the densities increasing as one goes down from the loop tops .
@xcite found upper limits on electron densities from _ chandra _ spectroscopy of au mic that are close to the above values . for lines of and , formed at roughly 2 mk and 7 mk respectively , @xcite found upper limits of @xmath131 and @xmath132 @xmath61 .
we can also combine our model densities with the @xmath45 values given above to determine the range of gas pressures in the loops and compare them with observations .
we assumed that @xmath46 remained constant over the length of each loop , and the values spanned from 760 dyne @xmath114 ( for @xmath68 ) to 1.8 dyne @xmath114 ( for @xmath69 ) .
@xcite found a best fit for the uv - derived differential emission measure of au mic to occur for a constant pressure of @xmath03 dyne @xmath114 , which falls inside the range of our modeled loops .
we also computed the free - free emission at longer radio wavelengths , for which there are no firm detections of au mic in quiescence .
both @xcite and @xcite gave non - detection upper limits that we can use as further tests of our model .
figure [ fig04](a ) carries over the two empirical contours from figure [ fig03 ] and also divides the parameter space into two regions : one in which the modeled emission is consistent with the @xcite upper limit for @xmath133 cm , and one in which the modeled emission exceeds that limit .
our best - fitting solution sits comfortably in the region that is consistent with all of the observed radio upper limits .
figure [ fig04](b ) shows representative radio and millimeter spectra for a set of seven models that fall along the observational @xmath74 contour shown in figure [ fig04](a ) .
each of these models reproduces the observed x - ray luminosity of au mic , but only one of them ( coincidentally , the one with the lowest value of @xmath22 ) agrees with the alma flux at 1.3 mm .
four of these seven models have radio fluxes that fall below the @xcite and @xcite upper limits .
it is noteworthy that a factor - of - two better measurement of the radio continuum at 110 cm would provide a much more stringent test of this model .
the shapes of the spectra in figure [ fig04](b ) convey interesting information about the spatial variations in temperature and density .
at the lowest wavelengths , the optical depth @xmath79 is small ( even when integrating all the way down to the coronal base ) , so the @xmath134 dependence of the source function cancels out with the @xmath135 in the opacity .
thus , the spectrum has a canonically flat shape that reflects only the slow wavelength dependence of the gaunt factor .
as the wavelength increases , the optical depth at the coronal base gets larger , and more of the coronal loops become optically thick . in that limit ,
the intensity is capped at a maximum value of @xmath136 if the temperature in the emitting region was a constant , this would give a power - law spectrum going as @xmath137 .
a steepening to this shape is indeed seen around @xmath1385 cm , but then it flattens out again .
this is because the optical depth keeps increasing as @xmath84 increases , and the level of the @xmath139 `` radio photosphere '' moves up from the coronal base to the tops of the loops . because @xmath63 increases from the base to the loop - tops ,
the emitting temperature to be used in equation ( [ eq : thick ] ) is an increasing function of @xmath84 , too .
this gives rise to an absolute value of @xmath80 that grows progressively larger than it would have been had the emitting temperature remained constant .
presumably as @xmath84 is increased even further , all of the loop tops will become optically thick , the emergent free - free emission will be dominated only by @xmath45 , and the spectrum will start declining again as @xmath137 .
lastly , we mention that the resulting values of x - ray and radio / mm emissivity do depend on the assumed value of the photospheric magnetic filling factor @xmath25 .
we ran a series of models for the standard parameter choices of @xmath127 km s@xmath53 and @xmath128 , where @xmath25 was varied over two orders of magnitude from 0.01 to 1 .
the x - ray luminosity scales roughly as @xmath140 , and the free - free emission flux at @xmath3 mm scales as @xmath141 .
these order - unity power law exponents imply that the uncertainties in the coronal loop parameters determined above may be roughly of the same magnitude as the uncertainty in our knowledge of @xmath25 for this star .
in this paper we began the process of exploring whether a physically motivated model of coronal loop heating could reproduce the observed properties of au mic .
we adapted an existing model of mhd turbulent dissipation to the environment of an m1 dwarf star and treated the surface velocity amplitude @xmath22 of stochastic alfvnic motions as a free parameter . for a value of @xmath142 km s@xmath53 ( which is consistent with the expected granular microturbulence velocity ) and a power - law probability distribution of loop lengths @xmath35 that goes roughly as @xmath143
, the model successfully predicts the observed x - ray luminosity and millimeter - wave emission peak .
the model is also consistent with radio upper limits measured at quiescent ( non - flaring ) time periods for au mic .
it may be possible to use independent estimates of the stellar wind mass loss rate of au mic to say more about the verisimilitude of the coronal heating model .
@xcite determined that a mass loss rate of order @xmath144 @xmath97 yr@xmath53 for au mic would generate sufficient pressure to transport dust to the outermost parts of its observed disk .
we used the stellar wind model presented by @xcite to compute how the mass loss rate of au mic depends on the assumed value of @xmath22 in the treatment of coronal heating .
the `` default '' value of @xmath145 km s@xmath53 obtained from the convective turbulence model of @xcite gives rise to a mass loss rate of only @xmath146 yr@xmath53 .
however , if we increase @xmath22 to the empirically determined value of 1.5545 km s@xmath53 ( see white circles in figures [ fig03][fig04 ] ) , the @xcite mass loss rate increases to @xmath147 yr@xmath53 .
this is close to the value required by @xcite to explain the outer disk s dust diffusion .
it is at least circumstantial evidence that our loop model , `` calibrated '' with @xmath142 km s@xmath53 , reflects what is actually going on in the outer atmosphere of au mic .
although our model accurately reproduces the magnitude of the alma 1.3 mm central emission peak , we do not yet have a definitive way to distinguish between this model and the idea of an inner asteroid belt proposed by @xcite .
for example , neither model predicts significant mid - infrared excess centered on the star , and indeed none is seen for au mic @xcite .
it is possible , of course , that au mic has _ both _ an active stellar corona and a bright inner source of dusty debris .
additional information about the spectral energy distribution of the central peak resolved by alma would be extremely helpful to putting limits on the relative contributions of these two suggested explanations for the 1.3 mm emission .
also , better observational limits on the spatial extent of the central peak would be helpful , since the dust belt is expected to be several orders of magnitude larger in size than the source of coronal emission .
in addition to better observational constraints , there are also several ways that the models can be improved .
we limited ourselves to just the time - steady component of coronal heating and not the more intermittent flaring , but the latter is known to be a non - negligible component of the total x - ray @xcite and ultraviolet @xcite emission . because we did not model flares , we also ignored the potential contributions of nonthermal electrons and gyromagnetic emission to the radio and millimeter spectra ( see * ? ? ?
* ; * ? ? ?
we did not simulate the full three - dimensional magnetic geometry of the corona of au mic ; instead we treated its topological complexity using a simple statistical distribution of loop sizes .
future refinements to our predictions of radio , millimeter , and x - ray emission from stellar coronae should include improvements to each of the above approximations .
finally , the success of any ab initio coronal model depends on our understanding of magnetic dynamos in low - mass stars .
young , rapidly rotating stars are suspected of having more vigorous convective motions and more frequent magnetic flux emergence than older stars such as the sun ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the turbulent convection models we have used so far @xcite do not yet include these effects , but our semi - empirical determination of @xmath142 km s@xmath53 may help put limits on how active the surfaces of young stars actually are .
still , we do not know what the granulation pattern and the distribution of magnetic flux tubes really looks like on the surface of a saturated - activity star .
it is also possible that rapid rotation strongly affects the distribution of loop lengths @xcite .
m dwarfs such as au mic are also close to the dividing line between having a radiative core and being fully convective .
the qualitative properties of rotation , dynamos , and large - scale magnetic fields are believed to change substantially across this dividing line @xcite .
these `` hidden '' aspects of stellar interiors may have significant impacts on the high - energy activity of a star like au mic and its surrounding dust and debris .
the authors gratefully acknowledge adriaan van ballegooijen , nancy brickhouse , and hans moritz gnther for many valuable discussions .
this paper made use of the following alma data : ads / jao.alma#2011.0.00142.s .
alma is a partnership of eso ( representing its member states ) , nsf ( usa ) and nins ( japan ) , together with nrc ( canada ) and nsc and asiaa ( taiwan ) , in cooperation with the republic of chile .
the joint alma observatory is operated by eso , aui / nrao , and naoj .
the national radio astronomy observatory is a facility of the national science foundation operated under cooperative agreement by associated universities , inc .
this research made extensive use of nasa s astrophysics data system and the simbad database operated at cds , strasbourg , france .
we also thank the anonymous referee for constructive suggestions that improved this paper .
beeck , b. , schssler , m. , & reiners , a. 2011 , in asp conf .
448 , 16th cambridge workshop on cool stars , stellar systems , and the sun , ed .
c. johns - krull , m. browning , & a. west ( san francisco , ca : asp ) , 1071 | many low - mass pre - main - sequence stars exhibit strong magnetic activity and coronal x - ray emission .
even after the primordial accretion disk has been cleared out , the star s high - energy radiation continues to affect the formation and evolution of dust , planetesimals , and large planets .
young stars with debris disks are thus ideal environments for studying the earliest stages of non - accretion - driven coronae . in this paper
we simulate the corona of au mic , a nearby active m dwarf with an edge - on debris disk .
we apply a self - consistent model of coronal loop heating that was derived from numerical simulations of solar field - line tangling and magnetohydrodynamic turbulence .
we also synthesize the modeled star s x - ray luminosity and thermal radio / millimeter continuum emission . a realistic set of parameter choices for au mic produces simulated observations that agree with all existing measurements and upper limits .
this coronal model thus represents an alternative explanation for a recently discovered alma central emission peak that was suggested to be the result of an inner `` asteroid belt '' within 3 au of the star .
however , it is also possible that the central 1.3 mm peak is caused by a combination of active coronal emission and a bright inner source of dusty debris . additional observations of this source s spatial extent and spectral energy distribution at millimeter and radio wavelengths will better constrain the relative contributions of the proposed mechanisms . |
one of the most promising projects in todays high energy physics is an electron - positron linear collider , for example tesla @xcite . at such a linear collider one should be able to polarize the electrons with the same technology as at the slc to up to @xmath7 . at tesla
it should also be possible to run with positrons polarized up to @xmath8 @xcite . with a luminosity of @xmath9 at energies close to the peak of the z - resonance tesla
could produce @xmath10 z - bosons in about 70 days of running @xcite
. in this scenario , often referred to as gigaz , the measurements already performed at the electron - positron collider experiments at lep and slc could be redone with increased precision .
an interesting topic is the test of the cp symmetry in z decays .
there is already a number of theoretical ( @xcite-@xcite and references therein ) and experimental @xcite-@xcite studies of this subject . in the present paper
we will study a flavor - diagonal z decay where cp - violating effects within the standard model ( sm ) are estimated to be very small @xcite .
thus , looking for cp violation in such z decays means looking for new physics beyond the sm . for a model - independent systematic analysis of cp violation in z
decays we use the effective lagrangian approach as described in @xcite .
of particular interest are z decays involving heavy leptons or quarks .
thus , the process z @xmath11 , which is sensitive to effective cp - violating couplings in the @xmath1 vertex , has been analysed theoretically in @xcite and experimentally in @xcite .
no significant deviation from the sm has been found .
if cp - violating couplings are introduced in the @xmath12 vertex , they will , because of gauge invariance of qcd , appear in the @xmath13 vertex as well . but the @xmath2 vertex could in principle contain new coupling parameters .
the analysis of the 4 jet decays of the z boson involving @xmath3 quarks looks into both , 4- and 5-point vertices .
this has been investigated theoretically in @xcite and experimentally in @xcite .
also in this case no significant deviation from the sm has been found . in this paper
we review the results of the calculations of the processes and including cp - violating couplings , with at least two of the jets originating from a @xmath3 or @xmath14 quark , for the gigaz scenario assuming longitudinal beam polarization for electrons and positrons .
all details of the calculation for unpolarized @xmath15 , @xmath16 beams can be found in @xcite .
finally we make an estimate for models with excited quarks and show that one can obtain stringent bounds on their mass .
this , however , requires the introduction of a new type of strong interactions for quarks . in chapter 2 we briefly review the theoretical framework of our computations .
models with excited quarks are discussed in chapter 3 .
next , in chapter 4 , we define cp - odd tensor and vector observables and calculate their sensitivities to anomalous couplings .
achievable bounds on the mass of excited quarks are presented .
our conclusions can be found in chapter 5 .
for a model independent study of cp violation in 3 jet and 4 jet decays of the z boson we use the effective lagrangian approach as explained in @xcite .
we could add to the sm lagrangian @xmath17 a cp - violating term @xmath18 containing all cp - odd local operators with a mass dimension @xmath19 ( electroweak symmetry breaking ) that can be constructed with sm fields .
however , it turns out that quite a number of such coupling terms can contribute to the reactions analysed here . to keep the analysis manageable we restrict ourselves to coupling terms involving the z and the @xmath3 quarks and in addition any number of gluons .
then the effective cp - violating lagrangian with @xmath20 relevant to our analysis is : @xmath21 \\ & + & [ \ ; h_{vb}\ : \bar{b}(x)\ : t^a\ : \gamma^{\nu}\ : b(x ) + h_{ab}\ : \bar{b}(x)\ : t^a\ : \gamma^{\nu}\ : \gamma_5\ : b(x)\ ; ] \ ; z^{\mu}(x)\ : g^a_{\mu\nu}(x)\;\;,\end{aligned}\ ] ] where @xmath22 denotes the @xmath3 quark field , @xmath23 and @xmath24 represent the field of the z boson and the field strength tensor of the gluon , respectively , and @xmath25 are the generators of @xmath26 @xcite . in ( [ lcp ] )
@xmath27 is the weak dipole moment and @xmath28 , @xmath29 are cp - violating vector and axial vector chirality conserving coupling constants . as effective coupling constants in @xmath18 the parameters @xmath27 , @xmath28 , @xmath29 are real .
they are related to form factors of vertices but should not be confused with the latter ( see e. g. @xcite ) .
the lagrangian @xmath18 is required to be invariant under the electromagnetic and strong gauge group @xmath30 .
we do not require explicit gauge invariance under the complete @xmath31 group of the electroweak interaction , since we consider the theory after electroweak symmetry breaking .
the couplings of ( [ lcp ] ) can , however , always be generated from higher dimensional @xmath32 invariant couplings involving suitable higgs fields , see @xcite .
the coupling constants in ( [ lcp ] ) are then proportional to powers of the higgs vacuum expectation value . also , in theories
which we do not want to exclude a priori from our discussion where @xmath33 and @xmath34 are composite objects the @xmath32 group has not necessarily a fundamental meaning ( see the review @xcite ) .
even if compositeness is not favoured by the lep data @xcite this option for going beyond the sm should certainly be investigated again at the gigaz factory .
information on the spin of the final state partons is hardly available experimentally .
thus , we consider as observables only the parton s energies and momenta .
then , effects linear in the dipole form factor @xmath27 are suppressed by powers of @xmath35 .
so angular correlations of the jets in and z @xmath0 4 jets are only sensitive to the couplings @xmath28 and @xmath29 .
the corresponding vertices following from @xmath18 are shown in figure [ fig : vert ] . because the non - abelian field strength tensor has a term quadratic in the gluon fields the @xmath1- and @xmath36-vertices
are related .
we define dimensionless coupling constants @xmath37 using the z mass as the scale parameter by @xmath38 where @xmath39 , @xmath40 and @xmath41 is the weak mixing angle . for numerical calculations we set @xmath42gev , @xmath43 and the fine structure constant and @xmath44 at the z mass to @xmath45 and @xmath46 @xcite .
our calculations are carried out in leading order of the cp - violating couplings of @xmath18 and the sm couplings .
a non - vanishing @xmath3 quark mass of @xmath47gev is included quark . in our leading order calculation
we could as well use the running @xmath3 mass at @xmath48 : @xmath49 gev @xcite .
this would result only in minimal changes in our correlations . ] ; masses of @xmath50 , @xmath51 , @xmath52 , @xmath53 quarks are neglected .
in this chapter we discuss the possible generation of chirality conserving cp - violating interactions as introduced in the previous chapter in models with excited quarks .
excitations of quarks would be natural in a scenario where quarks have substructure and participate in a new type of strong interaction .
this type of models and effects from excited quarks at hadron colliders have for instance been discussed in @xcite . in particular , we assume here that @xmath3 quarks have excited partners @xmath54 , which could have spin @xmath55 or @xmath56 . for simplicity
we consider a @xmath54 of spin @xmath55 and mass @xmath57 .
due to higher order dimensional operators in composite models chirality - conserving @xmath58 couplings at the scale of gigaz energies are a priori possible ( see e.g. @xcite ) .
because of colour gauge invariance we expect the @xmath59 couplings to be chirality - flipping dipole couplings .
then , couplings @xmath37 as introduced in ( [ lcp ] ) can be generated by the following effective interactions of @xmath54 and @xmath3 quarks , @xmath33 bosons and gluons : @xmath60 here @xmath61 , @xmath62 and @xmath63 are complex parameters , which can be expected to be of order one if the underlying dynamics is strongly interacting .
in addition to @xmath63 , the chromoelectric dipole transition form factor @xmath64 , there will be in general also a chromomagnetic transition form factor @xmath65 which is omitted here for brevity .
the couplings @xmath37 have been calculated @xcite in this model from the diagrams of the type shown in fig .
[ fig : excquark ] for @xmath66 : @xmath67
in our study we assume that one is able to flavor - tag the @xmath3 quarks and to measure their momenta .
this is justified due to the extremely good @xmath3-tagging capabilities foreseen at tesla @xcite .
for instance , the impact parameter resolution at tesla is expected to be about a factor 10 better than at lep @xcite .
the definition of a 3 and 4 jet sample requires the introduction of resolution cuts .
we use jade cuts @xcite requiring @xmath68 with @xmath69 the angle between the momentum directions of any two partons ( @xmath70 ) and @xmath71 , @xmath72 their energies in the z rest system .
we study our cp - violating couplings using cp - odd observables constructed from the momentum directions of the @xmath3 and @xmath14 quarks , @xmath73 and @xmath74 ( see @xcite ) : @xmath75 @xmath76 with @xmath77 , @xmath78 the cartesian vector indices in the z rest system . the observables @xmath79 transform as tensor components , @xmath80 as vector components . for polarized @xmath81 beams and our rotationally invariant cuts ( [ jade ] ) their expectation values are then proportional to the z tensor polarization @xmath82 and vector polarization @xmath83 , respectively . for all definitions concerning the @xmath33
density matrix see section 2.1 of @xcite . defining the positive @xmath84-axis in the @xmath15 beam direction
, we have @xmath85 @xmath86 where @xmath87 and @xmath88 with @xmath89 and @xmath90 the weak vector and axial vector @xmath91 couplings . @xmath92 and @xmath93 are the longitudinal polarizations for positron and electron , respectively , measured in the direction of the particle s velocity .
we have @xmath94 . from ( [ s ] ) ( [ s30 ] ) we see that the components @xmath95 and @xmath96 are the most sensitive ones .
note that the tensor observables do _ not _ change their sign upon charge misidentification ( @xmath97 ) whereas the vector observables do .
thus , it is only for the measurement of the latter that charge identification is indispensable .
we have computed the expectation values of the observables ( [ ten ] ) , ( [ vec ] ) for different jade cuts ( [ jade ] ) , as function of @xmath98 and @xmath99 where @xmath100 the expectation value of a cp - odd observable @xmath101 has the following general form : @xmath102 here @xmath103 are constants , @xmath104 and @xmath105 denote the corresponding and decay widths in the sm and in the theory with sm plus cp - violating couplings , respectively .
note that terms quadratic in the anomalous couplings are cp - even .
thus on the r.h.s . of ( [ cpoddobs ] )
they only emerge in @xmath105 . in an experimental analysis one has two options .
the first one is to study directly the expectation values @xmath106 which have a non - linear dependence on , .
this dependence becomes linear for small anomalous couplings . in the following
we will neglect non - linear terms , that is assume @xmath107 .
the other option is to take @xmath104 from the theoretical calculation , @xmath105 and @xmath106 from the experimental measurement .
the quantity @xmath108 is then an observable strictly linear in the anomalous couplings , which has obvious advantages . for unpolarised @xmath109
beams a non - zero value @xmath110 for one of our cp - odd observables above is an unambiguous indicator of cp violation . for longitudinally polarised beams this holds if possible chirality flipping interactions at the @xmath111 vertex which do not exist in the sm
are neglected .
see @xcite for an extensive discussion of this point . from the measurement of a single observable ( [ cpoddobs ] ) we can get a simple estimate of its sensitivity to by assuming @xmath112 .
the error on a measurement of is then to leading order in the anomalous couplings : @xmath113 where @xmath114 is the number of events within cuts .
similarly , assuming @xmath115 we get the error on as @xmath116 a measure for the sensitivity of to ( ) is then @xmath117 ( @xmath118 ) . in very good approximation
, it was found for and that the tensor observables are only sensitive to @xmath119 and the vector observables only to @xmath120 . a detailed discussion about that can be found in @xcite .
a measurement of , has to produce a mean value larger than @xmath121 ( [ dhb ] ) , @xmath122 ( [ dhbn ] ) to be able to claim a non - zero effect at the 1 s. d. level .
we have calculated the sensitivities to @xmath119 and @xmath120 for the tensor ( [ ten ] ) and vector ( [ vec ] ) observables varying the jet resolution parameter .
comparing with optimal observables it was found for unpolarized beams @xcite that these simple observables ( [ ten],[vec ] ) reach nearly optimal sensitivities .
therefore optimal observables are not considered in the following .
we assume a total number of @xmath123 z decays for unpolarized beams , following the gigaz scenario .
the number @xmath114 of events within cuts which is available for the analysis is then given by @xmath124 with @xmath125 being the total z decay width . solely due to higher statistics in gigaz of about a factor of one hundred compared to the sum of the four lep experiments , the sensitivity to the cp violating couplings increases by a factor 10 , as can be seen from ( [ dhb ] , [ dhbn ] ) .
the inverse sensitivities to these cp - odd couplings as calculated from ( [ dhb ] ) and ( [ dhbn ] ) , respectively , are shown in fig .
[ fig : del_asym_3jets ] for and in fig .
[ fig : del_asym_4jets ] for for different longitudinal beam polarizations .
the sensitivity decreases with increasing @xmath126 for all observables due to the decrease in number of events available . because the expectation value of the tensor observable does not depend on longitudinal polarization ( [ sij ] ) ,
the differences in for different polarization choices reflect only the change in statistics . for @xmath127 and @xmath128 the enhancement of the z production rate is largest .
the differences in reflect both the change in statistics and the modification of the expectation value due to polarization ( [ s3 ] ) . for @xmath127 and @xmath128
the sensitivity increases by more than a factor of six compared to unpolarized beams .
a convenient choice of the polarizations can even lead to a better sensitivity of the vector observable to than of the tensor observable to .
in contrast , an unsuitable choice of the polarizations could kill any sensitivity of the vector observable .
this is illustrated in figs .
[ fig : del_sym_3jets ] and [ fig : del_sym_4jets ] for and , respectively : the inverse sensitivities are shown as a function of the positron polarization assuming @xmath129 . for @xmath130 the expectation value for the vector observable and therefore the sensitivity to vanishes . for the tensor
observable this can not happen because the sensitivity to depends on the polarization only due to the change in the total number of z decays .
it should be stressed that in this article we present a tree - level calculation .
thus next - to - leading order qcd corrections are not taken into account .
for the sm part they can be found in @xcite for including non - vanishing @xmath3 quark masses and in @xcite for for massless quarks .
qcd corrections to the anomalous couplings ( [ lcp ] ) could be calculated using ( 0,0 ) ( -230,65 ) ( 0,0 ) ( -230,65 ) ( -140,-60 ) ( -140,-333 ) ( 0,0 ) ( -230,65 ) ( 0,0 ) ( -230,65 ) ( 0,0 ) ( -230,65 ) ( 0,0 ) ( -230,65 ) the methods of effective field theories ( see for instance @xcite ) . however , because here we always consider ratios of expectation values , see ( [ cpoddobs][dhbn ] ) , these corrections can be expected to cancel to some extent and to lead only to moderate changes to the numbers given .
this should hold at least for not too low values of @xmath131 . from @xcite one
finds higher order qcd corrections to become important for @xmath132 .
thus , to be on the safe side one should restrict the analysis to @xmath133 . if a measurement of , produces a mean value lower than @xmath121 ( [ dhb ] ) , @xmath122 ( [ dhbn ] ) a non - zero effect at the 1 s. d. level can not be claimed and therefore an upper limit on these couplings can be derived . using ( [ hexcf ] )
this can be translated into lower bounds on the excited quark mass @xmath57 .
assuming @xmath134 these bounds are shown in fig .
[ fig : mb_asym_3jets ] for and in fig .
[ fig : mb_asym_4jets ] for for different longitudinal beam polarizations . in @xcite at the @xmath135 confidence level excited quarks with mass between 80 and 570 gev and between 580 and 760
gev were excluded . in @xcite
the lower limit @xmath136 gev on the masses of excited quarks was given .
( [ dhb ] ) , @xmath137 ( [ dhbn ] ) to be able to claim a non - zero effect . from ( [ hexcf ] )
one derives that the mass limits at the 1 s. d. level given in fig .
[ fig : mb_asym_3jets ] and [ fig : mb_asym_4jets ] have to be divided by a factor @xmath138 to get the limits at the 2 s. d. level .
] however , these results apply to excited @xmath50 and @xmath51 quarks only and do not exclude a lighter @xmath54 quark .
in this paper , we have reviewed calculations concerning the search for cp violation in the 3 jet and 4 jet decays of the z boson with at least two of the jets originating from @xmath3 and @xmath14 quarks .
we have studied a cp - violating contact interaction with a vector and axial vector coupling , ( [ lcp ] ) , ( [ hhat ] ) .
we have discussed how such couplings can be generated in models with an excited @xmath3 quark , @xmath54 .
such couplings can also arise at one loop level in multi - higgs extensions of the standard model @xcite .
longitudinal beam polarization is included .
we studied a tensor and vector observable which can be used for the measurement of the anomalous couplings .
while the sensitivity of the tensor observable to cp - violating effects is only affected by the variation of statistics due to beam polarization given a certain integrated luminosity , the expectation value of the vector observable itself changes by the factor @xmath139 ( [ s3 ] ) .
if flavor tagging of @xmath3 and jets is available then , with a total number of @xmath10 z decays and choosing a cut parameter @xmath140 , the anomalous coupling constant ( [ hat ] ) can be determined with an accuracy of order 0.004 ( ) and 0.008 ( ) at 1 s. d. level using the tensor observable @xmath95 ( [ ten ] ) for the measurement . here
, @xmath141 distinction is not necessary
. these accuracies are close to the ones which already can be obtained with unpolarized beams .
if in a measurement a non - zero effect at the 1 s. d. level is not observed excited quark masses @xmath57 lower than 1.4 tev ( ) and 0.94 tev ( ) can be excluded if appropriate couplings are of a size characteristic of a strong interaction . if @xmath141 distinction is experimentally realizable , which should be the case at a future linear collider , the coupling constant ( [ tilde ] ) can be measured with an accuracy of order 0.0015 ( ) and 0.003 ( ) using the vector observable @xmath142 ( [ vec ] ) and choosing @xmath127 and @xmath128 as longitudinal polarizations of positron and electron , respectively . in case of a non - observation of an effect at the 1 s. d. level excited quark masses @xmath57 lower than 2.2 tev ( ) and 1.5 tev ( ) can be excluded if the relevant couplings are of a size characteristic of a strong interaction . comparing 3 and 4 jet analyses we found that the sensitivity to the anomalous coupling was roughly constant as function of the cut parameter for @xmath143 in the 3 jet case .
for the 4 jet case the sensitivity was found to increase as decreases .
for @xmath144 the 4 jet sensitivity was found to become equal to that from 3 jets .
of course in an experimental analysis one should try to make both 3 and 4 jet analyses in order to extract the maximal possible information from the data . in our theoretical investigations we assumed always @xmath145 efficiencies and considered the statistical errors only .
assuming systematic errors to be of the same size as the statistical ones , the accuracies in the determinations of , discussed above should indeed be better by more than one order of magnitude than those derived from lep .
as shown in @xcite this will , for instance , give valuable information on the scalar sector in multi - higgs extensions of the standard model .
that interesting information on models with excited quarks can be derived as well has been discussed in detail here .
we would like to thank a. brandenburg , f. nagel , t. ohl , s. schmitt and m. spira for valuable discussions .
we also thank a. brandenburg , a. denner , g. pasztor and m. spira for carefully reading the manuscript .
f. richard
, j. r. schneider , d. trines and a. wagner , `` tesla technical design report part i : executive summary '' , [ hep - ph/0106314 ] ; + j. a. aguilar - saavedra _ et al . _
( ecfa / desy lc physics working group collaboration ) , `` tesla technical design report part iii : physics at an @xmath81 linear collider '' , .
o. nachtmann and c. schwanenberger , eur .
j. * c 9 * ( 1999 ) 565 .
o. nachtmann and c. schwanenberger , eur .
j. * c 13 * ( 2000 ) 315 .
d. buskulic et al . , ( aleph collaboration ) , phys .
* b 297 * ( 1992 ) 459 .
m. steiert , `` suche nach cp - verletzenden effekten in hadronischen 3-jet ereignissen mit bottom flavour @xmath146 '' , diploma thesis , university of heidelberg ( unpublished ) ; + rainer liebisch , `` suche nach cp - verletzenden effekten auerhalb des standardmodells i m zerfall @xmath147 '' , diploma thesis , university of heidelberg ( unpublished ) .
h. harari , phys .
* 104 * ( 1984 ) 159 .
k. hagiwara _ et al .
_ ( particle data group collaboration ) , phys . rev . *
d 66 * ( 2002 ) 010001 .
g. kopp , d. schaile , m. spira and p. m. zerwas , z. phys .
c * 65 * ( 1995 ) 545 .
a. brandenburg , p. n. burrows , d. muller , n. oishi and p. uwer , phys .
* b 468 * ( 1999 ) 168 ; + g. rodrigo , m. s. bilenky and a. santamaria , nucl .
* b 554 * ( 1999 ) 257 ; + m. jamin and a. pich , nucl . phys .
* 74 * ( 1999 ) 300 ; + p. abreu _ et al .
_ , ( delphi collaboration ) , phys .
* b 418 * ( 1998 ) 430 ; + r. barate _
( aleph collaboration ) , eur .
j. * c 18 * ( 2000 ) 1 ; + g. abbiendi _ et al . _
( opal collaboration ) , eur .
j. * c 21 * ( 2001 ) 411 .
u. baur , i. hinchcliffe and d. zeppenfeld , int .
j. mod . phys . * a 2 * ( 1987 ) 1285 ; + u. baur , m. spira and p. zerwas , phys . rev .
* d 42 * ( 1990 ) 815 .
t. han and j. l. hewett , phys .
d 60 * ( 1999 ) 074015 .
k. mnig , `` electroweak physics at a linear collider z - factory '' , lc - phsm-1999 - 002-tesla . s. bethke _
( jade collaboration ) , phys .
* b 213 * ( 1988 ) 235 .
m. diehl , o. nachtmann and f. nagel , eur .
j. * c 27 * ( 2003 ) 375 ; m. diehl , o. nachtmann and f. nagel , `` probing triple gauge couplings with transverse beam polarisation in @xmath149 '' , hep - ph/0306247 .
w. bernreuther , a. brandenburg and p. uwer , phys .
* 79 * ( 1997 ) 189 ; + a. brandenburg and p. uwer , nucl . phys .
* b 515 * ( 1998 ) 279 .
s. weinzierl , j. phys .
* g 26 * ( 2000 ) 654 ; + l. j. dixon and a. signer , phys . rev . * d 56 * ( 1997 ) 4031 ; + z. nagy and z. trocsanyi , phys . rev .
* d 59 * ( 1999 ) 014020 [ erratum - ibid . * d 62 * ( 2000 ) 099902 ] ; + j. m. campbell , m. a. cullen and e. w. glover , eur .
j. * c 9 * ( 1999 ) 245 ; + s. weinzierl and d. a. kosower , phys .
d 60 * ( 1999 ) 054028 .
s. weinberg , `` the quantum theory of fields '' , volume ii , cambridge university press 1996 . | * abstract * we review cp - violating effects in z @xmath0 3 jet and z @xmath0 4 jet decays , assuming the presence of cp - violating effective @xmath1 and @xmath2 couplings .
longitudinal beam polarization is included in the studies .
we propose a direct search for such cp - violating couplings by using various cp - odd observables .
the data of a future linear collider running at the z - resonance in the so - called gigaz option should give significant information on the couplings .
finally we show that stringent bounds on the mass of excited @xmath3 quarks can be derived if appropriate couplings are of a size characteristic of a strong interaction .
desy 03101 + lc th2003 - 097 + hep - ph/0308198 + * cp violation in the 3 jet and 4 jet decays of the + z boson at gigaz * + o. nachtmann@xmath4 and c. schwanenberger@xmath5 + _ @xmath4 institut fr theoretische physik , universitt heidelberg , philosophenweg 16 , + d69120 heidelberg , germany + @xmath6 deutsches elektronen - synchrotron desy , notkestr .
85 , d22603 hamburg , germany _ |
quasars , powered by accretion of material onto the central supermassive black holes ( smbhs ) of active galaxies , are among the most luminous objects in the universe . in the current paradigm of galaxy formation and evolution
, quasars represent a key phase in the evolutionary history of massive galaxies .
essentially all spheroidal systems are believed to harbour massive black holes .
the masses of the central smbhs are shown to correlate with many properties of their host galaxies , e.g. , the stellar velocity dispersion of the bulge , the bulge mass and the bulge luminosity ( e.g. , kormendy & richstone 1995 ; magorrian et al . 1998 ; ferrarese & merritt 2000 ; gebhardt et al .
2000 ; tremaine et al .
2002 ; kormendy et al .
present - day smbhs are thought to have gained most of their mass via accretion during an active nuclear phase , e.g. the quasar phase .
the implication is that black hole growth and star - formation activity are inextricably linked , at least for galaxies which experience ` wet ' major mergers and a quasar - like period of rapid black hole growth ( see kormendy & ho 2013 for a review ) .
a probable scenario is that the galaxy initially forms in a gas - rich rotationally supported system . as the host dark matter halo grows , some mechanisms , such as major merger or disk / bar instability trigger a period of rapid , dust obscured star - formation activity ( starburst ) which produces a stellar bulge .
at the same time , the infalling gas fuelling the starburst is also feeding the central black hole . after some time the quasar becomes optically visible ( unobscured ) , and soon after that the on - going star formation is quenched on a short time - scale , perhaps via radiative or mechanical feedback from the central engine ( see fabian 2012 for a review ) .
the cosmic star - formation history is shown to be similar to the evolving luminosity density of quasars ( which is related to the bh accretion history ) at most redshifts ( e.g. , boyle & terlevich 1998 ; franceschini et al .
1999 ; merloni et al .
2004 ; shankar et al . 2007 ; silverman et al . 2008 ; wall , pope & scott 2008 ) , which adds further support to this picture of the galaxy - black hole connection . in this paper
, we probe the co - evolution of black hole growth and star - formation activity using a cross - correlation analysis between optically selected type 1 quasars over the redshift range @xmath17 and the cosmic infrared background ( cib ) in the far - infrared ( fir)/sub - millimetre ( sub - mm ) . the former represents some of the most luminous and massive black holes found to date .
the latter is dominated by emission from dusty star - forming galaxies ( dsfgs ) over a similar redshift range to the quasars ( valiante et al .
2009 ; bethermin et al .
2011 ; bethermin et al .
2012 ; viero et al .
the large - scale cross - correlation signal of quasars and star - forming galaxies is determined by the effective linear bias ( or equivalently , the effective host halo mass ) of both sets of tracers of the underlying large - scale structure .
the small - scale cross - correlation signal constrains the halo occupation distributions ( hods ) of star - forming galaxies in quasar - hosting dark matter halos and the hods of quasars in halos hosting star - forming galaxies .
in contrast , the auto - correlation function constrains the hods of each population independently , not the hods of one population in the presence of the other .
therefore , the cross - correlation function is arguably more suited to studying the co - evolution of the quasars and the star - forming galaxies .
quasars are very luminous active galactic nuclei ( agn ) and hence can be seen out to great distances .
they are great tracers of the large - scale structure . to measure their clustering properties reliably , large - volume surveys are required , due to the low spatial density of quasars .
the sloan digital sky survey ( sdss ; york et al .
2000 ) has produced some of the largest quasar samples to date and that is what we use in this paper .
fir and sub - mm observations are critical for studying star - forming galaxies , as dust absorbs the optical / ultraviolet light and reradiates in the fir / sub - mm .
dust emission makes up the cib , which contains about half of the extragalactic background radiation ( excluding the cosmic microwave background ) , i.e. the total power originated from stars and agn throughout the cosmic history ( puget et al .
1996 ; fixsen et al . 1998 ; lagache et al .
the _ herschel space observatory _
( pilbratt et al . 2010 ) has carried out imaging surveys with unprecedented size and depth at fir / sub - mm wavelengths .
thanks to its sensitivity and mapping speed , it allows us for the first time to probe the bolometric power of dsfgs at high redshifts in large volumes .
this paper is organised as follows . in section 2
, we give an overview of the _ herschel_-spire maps and sdss dr7 and dr9 qso catalogues used in our analysis . in section 3 ,
we first describe the estimator used to calculate the cross - correlation signal between the qsos and the cib .
we then investigate the potential contamination of galactic dust emission by explicitly calculating the cross - correlation signal between our quasar samples and the iras 100 @xmath5 data in our fields . finally , we present the cross - correlation results between various quasar samples and the spire maps at 250 , 350 and 500 @xmath5 . in section 4 ,
a simple halo model is employed to interpret the measured qso
cib cross - correlation signal and constrain the hods of quasars in halos that also host star - forming galaxies . further discussion and conclusions of our results
are given in section 5 .
we assume an @xmath18 km s@xmath19 mpc@xmath19 , @xmath20 , and @xmath21 cosmology and use the ab magnitude scale throughout the paper .
we use cib maps at 250 , 350 , and 500 @xmath5 ( 1200 , 857 , and 600ghz , respectively ) from the two _ herschel _ surveys in the sdss stripe 82 region , i.e. the _ herschel _ stripe 82 survey ( hers ; viero et al .
2013b ) and the hermes large - mode survey ( helms ; oliver et al .
maps were observed with the spectral and photometric imaging receiver ( spire ) instrument ( griffin et al .
2010 ) aboard _
hers covers 79 deg@xmath22 along the sdss stripe 82 to a 5-@xmath23 depth of 65.0 , 64.5 and 74.0 mjy beam@xmath19 at 250 , 350 and 500 @xmath5 , respectively .
helms covers 270 deg@xmath22 to a a 5-@xmath23 depth of 64.0 , 53.0 and 76.5 mjy beam@xmath19 at 250 , 350 and 500 @xmath5 .
the joint hers and helms areal coverage between -10@xmath24 and 37@xmath24 ( ra ) covers the full @xmath25 subset of stripe 82 that has the lowest level of galactic dust emission ( or cirrus ) foregrounds ( i.e. , with @xmath26 ) .
the spire data , obtained from the _ herschel _ science archive , were reduced using the standard esa software and the custom - made software package , smap ( levenson et al .
2010 , viero et al .
maps were made using an updated version of smap / shim , which is an iterative map - maker designed to optimally separate large - scale noise from signal . for more details
, please refer to viero et al .
( 2013b ) .
the maps have a tangent plane ( tan ) projection with pixel sizes of 6@xmath27 , 8.33@xmath27 and 12@xmath27 at 250 , 350 , and 500 @xmath5 , respectively . in comparison ,
the full width at half maximum ( fwhm ) of the spire beams are 18.1@xmath27 , 25.2@xmath27 and 36.6@xmath27 at 250 , 350 , and 500 @xmath5 , respectively ( swinyard et al . 2010 ) .
the spire maps are converted from their native unit of @xmath28 to @xmath29 by dividing them by the effective beam areas which are 0.9942 , 1.765 , and 3.730 @xmath30steradians at 250 , 350 , and 500 @xmath5 , respectively ( spire observers manual ) .
we note that no colour corrections from a flat - spectrum point - source calibration are applied , as they were shown in viero et al .
( 2013a ) to be negligible .
[ fig : hershelms_sky ] shows the three - colour spire images of the hers and helms regions .
we can clearly see white foreground clouds of galactic cirrus emission , which have been shown to correlate with hi emission from gass 21-cm data in viero et al .
( 2013b ) . in this paper
, we also include spire maps of a subset of the hermes level 5 and level 6 fields , which have significant overlaps with our sdss quasar samples .
in particular , the hermes botes field ( 11.3 deg@xmath22 ) and the lockman - swire field ( 15.2 deg@xmath22 ) overlap with the sdss dr7 qso sample .
the botes field and the _ xmm_-lss field ( 21.6 deg@xmath22 ) overlap with the sdss dr9 qso sample .
[ fig : lss14_sky ] shows the three - colour spire images of the botes , lockman - swire and _
xmm_-lss field . out of these three fields ,
the _ xmm_-lss field is the most contaminated by cirrus emission , but all three fields have significantly less cirrus contamination than the hers or helms field .
in addition , the level 5 and 6 fields are significantly deeper than hers or helms .
the 5-@xmath23 noise level in the _
xmm_-lss and botes field is 25.8 , 21.2 and 30.8 mjy at 250 , 350 and 500 @xmath5 , respectively .
the lockman - swire field is even deeper and the 5-@xmath23 noise level is 13.6 , 11.2 and 16.2 mjy at 250 , 350 and 500 @xmath5 , respectively .
botes filed ( top ) , the 15.2 deg@xmath22 lockman - swire field ( middle ) and the 21.6 deg@xmath22 _
xmm_-lss field ( bottom).,width=321,height=854 ] the cib at sub - mm wavelengths is dominated by emission from dsfgs over a wide range of redshift , while other potential sources of signal , e.g. , the cosmic microwave background ( cmb ) , sunyaev - zeldovich ( sz ) effect , and radio galaxies , are subdominant .
one exception is the emission from galactic dust ( cirrus ) which can not be ignored at fir / sub - mm wavelengths .
therefore , the spire map signal can be modelled , to first order , as the sum of the following three terms : the emission from dsfgs ; galactic cirrus emission ; and instrument noise .
thus we can write @xmath31 we expect that only the emission from dsfgs have intrinsic correlation with the quasars .
this is because some quasars could be strong ir / sub - mm emitters from dust heated by the accreting black hole , as well as star formation in the host galaxy ( e.g. lutz et al .
2008 ; shi et al . 2009 ; clements et al .
2009 ; serjeant et al . 2010 ; bonfield et al . 2011 ; dai et al .
but additionally dsfgs are expected to occur as satellite galaxies in the dark matter halos where quasars are found .
a last effect is that dark matter halos correlated with the quasar - hosting halo are also expected to host dsfgs .
the spire instrumental noise should not correlate with quasars and therefore only contributes to the noise in the measured cross - correlation signal between quasars and the cib . similarly , galactic dust emission should not have any intrinsic correlation with the quasars .
however , due to the selection of the quasars , there might be ( anti-)correlation between the observed quasar number density field and the intensity of the cirrus emission .
there are two effects of galactic dust on the selection of qsos .
the first effect is reddening which means qsos with extreme reddening will not be selected as targets for spectroscopy .
the second effect is extinction , which can cause the qsos to drop out of the target list even before they reach the colour selection stage .
qso populations with steeper number counts will be more sensitive to extinction .
to explicitly check the potential correlation between the quasar samples and galactic cirrus emission , we make use of the 100 @xmath5 maps from the improved reprocessing of the _ iras _ ( neugebauer et al . 1984 )
survey ( iris ; miville - deschenes & lagache 2005 ) , a data set which corrects the original plates for calibration , zero level and striping problems .
the iris 100 @xmath5 maps in the hermes level 5 and level 6 fields are discussed in viero et al . (
2013 ) and the iris 100 @xmath5 maps in hers and helms are discussed in hajian et al .
( in prep . ) .
the 100 @xmath5 maps in all of our fields are shown in appendix [ appendix1 ] .
here we have assumed that the dominant signal in the 100 @xmath5 maps comes from galactic cirrus emission , not the dsfgs comprising the cib . in other words ,
these 100 @xmath5 maps are used as a galactic cirrus tracer in our fields data may not be used as a reliable galactic cirrus tracer as a result of significant contamination from the cib ( pnin et al .
2012 ) . ] .
visual inspection of the maps in appendix [ appendix1 ] confirms that extended diffuse galactic cirrus emission are the dominant structures ( long filamentary chains ) in the 100 @xmath5 maps , especially in the hers and helms regions .
we have also computed the mean value of the 100 @xmath5 maps in the helms , hers , _
xmm_-lss , botes and lockman - swire field to be 3.03 , 2.84 , 2.30 , 1.42 and 1.09 mjy sr@xmath19 , respectively .
if we adopt the dirbe measurement of the cosmic ir background at 100 @xmath5 which is @xmath32 nw m@xmath33 sr@xmath19 or equivalently @xmath34 mjy sr@xmath19 ( dole et al .
2006 ) , then we can estimate that the fractional contribution of the cib to the 100 @xmath5 maps in the helms , hers , _
xmm_-lss , botes and lockman - swire field is 16% , 17% , 21% , 34% and 44% , respectively . for our cross - correlation analysis ,
we use quasars from both the sdss dr7 ( abazajian et al . 2009 ) and the sdss - iii dr9 ( ahn et al . 2012 ) spectroscopic surveys . the spectroscopic sdss dr7 quasar catalogue ( dr7q ; schneider et al .
2010 ; shen et al .
2010 ) contains 105,783 spectroscopically confirmed quasars with @xmath35-band absolute magnitude @xmath36 and @xmath37 over an area of 9,380 deg@xmath22 .
it is the culmination of the sdss - i and sdss - ii quasar surveys .
the quasar target selection is described in detail by richards et al .
quasar candidates are selected based on their fluxes and colors in sdss bands . at low - redshift , @xmath38 ,
quasar targets are selected based on their location in @xmath39-color space and at high - redshift , @xmath40 , targets are selected from @xmath41-color space .
quasar candidates passing the @xmath39-color selection are selected to a flux limit of @xmath42 , but since high - redshift quasars are rare , objects lying in regions of color - space corresponding to quasars at @xmath40 are targeted to @xmath43 ( psf magnitudes corrected for galactic extinction based on the dust maps of schlegel , finkbeiner & davis 1998 ) . the data release 9 quasar catalogue ( dr9q ; pris et al .
2012 ) is the first quasar catalogue from the baryon oscillation spectroscopic survey ( boss ; dawson et al .
2012 ) of the sloan digital sky survey iii ( eisenstein et al . 2011 ) .
this catalogue contains 87,822 quasars over 3,275 deg@xmath22 .
the primary science goal for the high-@xmath44 boss quasar survey was to measure the baryon acoustic oscillations in the lyman-@xmath45 forest ( busca et al .
2013 ; slosar et al .
2013 ; delabuc et al .
2014 ) . as such
a priority was set on obtaining a high - surface density of @xmath46 quasars in order to maximise the number of lyman-@xmath45 forest lines of sight observable in the blue .
to achieve this high - surface density , quasar candidates were selected in a heterogenous manner , down to a magnitude limit of @xmath47 or @xmath48 ( ross et al .
quasars were selected based on their optical fluxes and colours , as well as their near - infrared , ultraviolet fluxes and radio properties ; ross et al .
( 2012 ; and references therein ) has full details .
however , a uniform selection was performed on a subsample of the dr9 quasar targets .
this uniform sample was selected by the
extreme deconvolution " ( xd ; bovy , hogg & roweis 2009 ) algorithm which is applied to model the distribution in flux and flux uncertainty of a training set of quasars and stars ( xdqso ; bovy et al . 2011 ) .
the dr9q has a flag which indicates if a quasar was selected in this uniform manner . in this paper
we select all quasars in the dr9q catalog with uniform @xmath49 " which represent a statistical sample for clustering studies ( pris et al .
2012 ; white et al . 2012 ) . additionally , we limit the sdss dr9 quasar sample to be in the redshift range @xmath50 .
the median redshift of the final selected sdss dr9 quasar sample is @xmath51 .
stripe 82 is an equatorial region that the sloan digital sky survey ( sdss ) has repeatedly imaged up to 80 times ( abazajian et al .
2009 ) giving coadded optical data @xmath522 magnitudes deeper than the single epoch sdss observations .
the stripe 82 field covers approximately a 270 deg.@xmath22 area ( @xmath53ra@xmath54 , @xmath55 dec @xmath56 ) . for the sdss dr7q ,
since the stripe 82 imaging is deeper we are able to select quasars to @xmath57 for the cross - correlation study .
we further restrict our analysis to sdss dr7 quasars in the redshift range @xmath58 , with a median redshift of 1.4 .
the redshift cut is used to avoid the redshift deserts of quasars at @xmath59 and @xmath60 due to contamination from f and gk stars ( fan 1999 ; richards et al .
2002 , 2006 ) .
one complication for sdss - iii boss dr9 is that a set of quasar selection methods , and not the uniform xdqso method , was used in the stripe 82 region .
consequently , the white et al .
( 2012 ) completeness corrections are not suitable in stripe 82 . however ,
during boss observations , stripe 82 was observed at a high target density ( palanque - delabrouille et al .
2011 ) and the actual completeness of @xmath50 quasars is close to @xmath61 , down to the @xmath47 or @xmath48 limiting magnitude ( palanque - delabrouille et al .
2013 ; ross et al .
2013 ) .
.number of selected sdss dr7 and dr9 qsos found in hers , helms , botes , lockman - swire and _ xmm_-lss regions .
the lockman - swire field does not overlap with the dr9 quasar sample and the _ xmm_-lss field does not overlap with the dr7 quasar sample .
note that the number of dr9 qsos in the hers and helms stripe 82 regions goes through a final change in section 3.2 .
the numbers inside the brackets represent the final number of dr9 quasars in these two regions .
the median redshift for the dr7 and dr9 qsos is 1.4 and 2.5 , respectively . [
cols="<,<,<",options="header " , ]
we present the first cross - correlation measurement between optically selected type 1 sdss dr7 and dr9 quasars and the cosmic infrared background ( cib ) measured by the _ herschel_-spire instrument at 250 , 350 and 500 @xmath5 .
the distribution of the quasars at @xmath0 ( sdss dr7 quasars : @xmath58 ; sdss - iii dr9 quasars : @xmath50 ) covers the redshift range where we expect most of the cib to originate .
we detect the sub - millimetre ( sub - mm ) emission of the optical quasars which dominates on small angular scales as well as the correlated sub - mm emission from dusty star - formation galaxies ( dsfgs ) dominant on larger angular scales .
a simple halo model is used to interpret the measured cross - correlation signal between the quasars and the cib . at a median redshift of 1.4 ,
the mean sub - mm flux densities of the dr7 quasars is @xmath62 , @xmath63 and @xmath64 mjy at 250 , 350 and 500 @xmath5 respectively . at a median redshift of 2.5 ,
the mean sub - mm flux densities of the dr9 quasars is @xmath65 , @xmath66 and @xmath67 mjy at 250 , 350 and 500 @xmath5 respectively . by fitting a modified blackbody
sed combined with a mid - infrared power - law template to the mean sub - mm fluxes , we find that the higher - redshift dr9 quasars have a higher mean infrared luminosity than the dr7 quasars , possibly due to elevated star - formation activity and/or black hole accretion or a larger contribution to the infrared luminosity from the accreting bh .
further investigations into the power source of the infrared luminosity of the quasars using data at other wavelengths is beyond the scope of this paper .
we find that the correlated sub - mm emission includes both the emission from satellite dsfgs in the same dark matter halo as the central quasar ( the one - halo term ) and the emission from dsfgs in separate halos which are correlated with the quasar - hosting halo ( the two - halo term ) .
the amplitude of the one - halo term of the correlated emission is about 10 times smaller than the sub - mm emission of the quasars themselves , implying that the satellite dusty star - forming galaxies are much fainter and have a lower sfr than the central quasars .
we infer from the halo model that the satellite fraction of the sdss dr7 quasars ( @xmath68 $ ] ) at a median redshift @xmath69 of 1.4 is @xmath10 , and the host halo mass scale for the central and satellite quasars is @xmath70 m@xmath12 and @xmath71 m@xmath12 respectively .
the satellite fraction of the sdss - iii dr9 quasars ( @xmath72 $ ] ) at a median redshift @xmath69 of 2.5 is @xmath14 , and the host halo mass scale for the central and satellite quasars is @xmath73 m@xmath12 and @xmath74 m@xmath12 respectively .
therefore , the typical halo environment of the quasars is similar to that of the dsfgs over the redshift range probed .
this is expected if dusty starburst and quasar activity are evolutionarily linked phenomena .
several aspects of our study could be improved and/or extended . in this paper , we have used the cross - correlation signal between the quasar samples and the cib maps to constrain the halo occupation statistics of the quasars .
a more complete approach would be to combine the auto - correlation statistics ( of the quasars and the cib maps ) and the cross - correlation statistics ( between the two tracers ) to constrain the connection between different types of galaxies and the underlying dark matter halos simultaneously .
we leave this project to a future paper when we have a better understanding of the qso selection effects in the stripe 82 region .
donoso et al .
( 2013 ) recently reported a difference in the halo environment between obscured and unobscured agn .
we can provide independent constraints by looking at the cross - correlation between obscured and unobscured agn and the cib ( wang et al .
2014 , in prep . )
lw and pn acknowledge support from an erc stg grant ( degas-259586 ) .
spire has been developed by a consortium of institutes led by cardiff univ .
( uk ) and including univ .
lethbridge ( canada ) ; naoc ( china ) ; cea , lam ( france ) ; ifsi , univ .
padua ( italy ) ; iac ( spain ) ; stockholm observatory ( sweden ) ; imperial college london , ral , ucl - mssl , ukatc , univ .
sussex ( uk ) ; caltech , jpl , nhsc , univ .
colorado ( usa ) .
this development has been supported by national funding agencies : csa ( canada ) ; naoc ( china ) ; cea , cnes , cnrs ( france ) ; asi ( italy ) ; mcinn ( spain ) ; snsb ( sweden ) ; stfc ( uk ) ; and nasa ( usa ) .
funding for the sdss and sdss - ii has been provided by the alfred p. sloan foundation , the participating institutions , the national science foundation , the u.s .
department of energy , the national aeronautics and space administration , the japanese monbukagakusho , the max planck society , and the higher education funding council for england .
the sdss web site is http://www.sdss.org/. the sdss is managed by the astrophysical research consortium for the participating institutions .
the participating institutions are the american museum of natural history , astrophysical institute potsdam , university of basel , university of cambridge , case western reserve university , university of chicago , drexel university , fermilab , the institute for advanced study , the japan participation group , johns hopkins university , the joint institute for nuclear astrophysics , the kavli institute for particle astrophysics and cosmology , the korean scientist group , the chinese academy of sciences ( lamost ) , los alamos national laboratory , the max - planck - institute for astronomy ( mpia ) , the max - planck - institute for astrophysics ( mpa ) , new mexico state university , ohio state university , university of pittsburgh , university of portsmouth , princeton university , the united states naval observatory , and the university of washington .
funding for sdss - iii has been provided by the alfred p. sloan foundation , the participating institutions , the national science foundation , and the u.s .
department of energy office of science .
the sdss - iii web site is http://www.sdss3.org/. sdss - iii is managed by the astrophysical research consortium for the participating institutions of the sdss - iii collaboration including the university of arizona , the brazilian participation group , brookhaven national laboratory , carnegie mellon university , university of florida , the french participation group , the german participation group , harvard university , the instituto de astrofisica de canarias , the michigan state / notre dame / jina participation group , johns hopkins university , lawrence berkeley national laboratory , max planck institute for astrophysics , max planck institute for extraterrestrial physics , new mexico state university , new york university , ohio state university , pennsylvania state university , university of portsmouth , princeton university , the spanish participation group , university of tokyo , university of utah , vanderbilt university , university of virginia , university of washington , and yale university .
the iris 100 @xmath5 maps in the hers and helms field are plotted in fig .
[ fig : iris100_hershelms ] . the iris
100 @xmath5 maps in the botes , lockman - swire and _
xmm_-lss field are plotted in fig .
[ fig : iris100_lss14 ] .
visual inspection confirms that extended diffuse galactic cirrus emission are the dominant structures ( long filamentary chains ) in these 100 @xmath5 maps , especially in the hers and helms regions .
the mean value of the 100 @xmath5 maps in the helms , hers , _
xmm_-lss , botes and lockman - swire field is 3.03 , 2.84 , 2.30 , 1.42 and 1.09 mjy sr@xmath19 , respectively . adopting the dirbe measurement of the cosmic ir background at 100 @xmath5 which is @xmath32 nw m@xmath33 sr@xmath19 or equivalently @xmath34 mjy sr@xmath19 ( dole et al .
2006 ) , we can estimate that the fractional contribution of the cib to the 100 @xmath5 maps in the helms , hers , _
xmm_-lss , botes and lockman - swire field is 16% , 17% , 21% , 34% and 44% , respectively .
maps in the botes ( top ) , lockman - swire ( middle ) and _ xmm_-lss ( bottom ) field.,title="fig:",width=340,height=292 ] maps in the botes ( top ) , lockman - swire ( middle ) and _ xmm_-lss ( bottom ) field.,title="fig:",width=340,height=292 ] maps in the botes ( top ) , lockman - swire ( middle ) and _ xmm_-lss ( bottom ) field.,title="fig:",width=340,height=292 ]
we have a quasar catalogue from which build the quasar density map @xmath75 , which gives the number of quasars in each pixel of the map .
the mean value of this map is @xmath76 , where @xmath77 is the number of objects in the catalogue and @xmath78 is the number of pixels in the map .
the spire map @xmath79 has a mean value @xmath80 which is set to zero , as spire has no sensitivity to the absolute zero - level on the sky .
we retain the mean value in what follows to keep the equations general .
the covariance of @xmath75 and @xmath79 is defined as : @xmath81 where the sums are over the two map dimensions and @xmath82 . as shown in marsden et al .
( 2009 ) , this is equivalent to what is often called a `` stack '' , when normalised by @xmath83 , the average number of quasars per pixel : @xmath84 we can extend eq . [ eqn : cov ] to describe a stacked thumbnail , made by making a pixel - by - pixel average of map cutouts centred on each source . this is equivalent to calculating the covariance of two maps with a relative shift by @xmath85 pixels : @xmath86 in order to account for the edges of the maps , the sum over @xmath87(@xmath88 ) runs from the larger of 1 and @xmath35 ( @xmath89 ) to the smaller of @xmath90 ( @xmath91 ) and @xmath92 ( @xmath93 ) .
@xmath94 is thus a map with indices running from @xmath95 to @xmath96 for an @xmath97 thumbnail map .
we can then take an azimuthal average of this map to create the radial profile of the stack : @xmath98 where @xmath99 is the set of pixels for which : @xmath100 and @xmath101 is the number of pixels in @xmath99 . using eq .
[ eqn : stack ] , the radial stack @xmath102 is then : @xmath103 in section 3.1 , we use a modification of the landy & szalay ( 1993 ) two - point correlation function estimator , @xmath104 here , @xmath105 and @xmath106 represent the quasar density map and spire map , called @xmath75 and @xmath79 above . @xmath107 and
@xmath108 represent random realisations of these two datasets .
the quantities @xmath109 are the same as @xmath110 in eq .
[ eqn : rk ] above , although without the subtractions of the map means in eq .
[ eqn : covij ] ; for convenience , we refer to the non mean - subtracted version of eq .
[ eqn : rk ] as @xmath111 .
the first term in eq .
[ eqn : wtheta ] can then be precisely defined as @xmath112 .
the remaining terms all involve one or more random distributions , and thus in each term , the two quantities are uncorrelated . for uncorrelated distributions ,
the expectation value of the product of the two variables is equal to the product of the expectation values of each variable , @xmath113 , and so the remaining terms are simply to the product of the mean values , @xmath114 .
we can thus simplify eq .
[ eqn : wtheta ] : @xmath115 by expanding eqn [ eqn : covij ] , eq . [ eqn : radstack ] can similarly be rewritten in terms of @xmath116 : @xmath117.\ ] ] the denominator of eq .
[ eqn : wtheta ] is zero when the spire map mean is zero , and so we modify the equation by multiplying both sides by the mean @xmath80 .
we see that the algorithm used in section 3.1 is thus exactly equivalent to eq .
[ eqn : radstack ] , the azimuthally - averaged stacked thumbnail image . | we present the first cross - correlation measurement between sloan digital sky survey ( sdss ) type 1 quasars and the cosmic infrared background ( cib ) measured by _ herschel_-spire .
the distribution of the quasars at @xmath0 covers the redshift range where we expect most of the cib to originate .
we detect the sub - millimetre ( sub - mm ) emission of the quasars , which dominates on small scales , as well as correlated emission from dusty star - forming galaxies ( dsfgs ) dominant on larger scales .
a simple halo model is used to interpret the measured cross - correlation signal .
the mean sub - mm flux densities of the dr7 quasars ( median redshift @xmath1 ) is @xmath2 , @xmath3 and @xmath4 mjy at 250 , 350 and 500 @xmath5 , respectively , while the mean sub - mm flux densities of the dr9 quasars ( @xmath6 ) is @xmath7 , @xmath8 and @xmath9 mjy .
we find that the correlated sub - mm emission includes both the emission from satellite dsfgs in the same halo as the central quasar ( the one - halo term ) and the emission from dsfgs in separate halos that are correlated with the quasar - hosting halo ( the two - halo term ) .
the amplitude of the one - halo term is about 10 times smaller than the sub - mm emission of the quasars , implying the the satellite dsfgs have a lower star - formation rate than the central quasars .
we infer that the satellite fraction ( i.e. the fraction of quasars hosted by satellite galaxies in the halo ) for the dr7 quasars ( @xmath1 ) is @xmath10 and the host halo mass scale for the central and satellite quasars is @xmath11 m@xmath12 and @xmath13 m@xmath12 , respectively . at a median redshift of 2.5 , the satellite fraction of the dr9 quasars is @xmath14 and the host halo mass scale for the central and satellite quasars is @xmath15 m@xmath12 and @xmath16 m@xmath12 , respectively .
thus , the typical halo environment of the sdss type 1 quasars is found to be similar to that of dsfgs over the redshift range probed , which supports the generally accepted view that dusty starburst and quasar activity are evolutionarily linked phenomena .
submillimetre : galaxies quasars : general galaxies : evolution galaxies : haloes galaxies : high - redshift quasars : supermassive black holes |
one of the most interesting recent developments in the study of supernovae ( sne ) is the discovery of some very energetic supernovae , whose kinetic energy ( ke ) exceeds @xmath3erg , about 10 times the ke of normal core - collapse sne ( hereafter @xmath4erg ) .
the most luminous and powerful of these objects , the type ic supernova ( sn ic ) 1998bw , was probably linked to the gamma - ray burst grb 980425 ( galama et al .
1998 ) , thus establishing for the first time a connection between gamma - ray bursts ( grbs ) and the well - studied phenomenon of core - collapse sne .
however , sn 1998bw was exceptional for a sn ic : it was as luminous at peak as a sn ia , indicating that it synthesized @xmath5 @xmath2 of @xmath0ni , and its ke was estimated at @xmath6 erg ( iwamoto et al .
1998 ; woosley , eastman , & schmidt 1999 ) .
because of its large ke , sn 1998bw was called a hypernova ( hn ) " . subsequently , other
hypernovae " of type ic have been discovered or recognised , such as sn 1997ef ( iwamoto et al .
2000 ; mazzali , iwamoto & nomoto 2000 ) , sn 1997dq ( matheson et al .
2001 ) , sn 1999as ( knop et al . 1999 ) , and sn 2002ap ( mazzali et al .
although these sne ic did not appear to be associated with grbs , most recent `` hypernova '' sn 2003dh is clearly associated with grb 030329 ( stanek et al .
2003 ; hjorth et al . 2003 ; kawabata et al .
figures [ eps1 ] and [ eps2 ] show the near - maximum spectra and the absolute v - light curves of these hypernovae .
these objects span a wide range of properties , although they all appear to be highly energetic compared to normal core - collapse sne .
sn 1999as is the most luminous supernova ever discovered , reaching a peak magnitude @xmath7 , while the brightness of sn 2002ap appears to be similar to that of normal core collapse sne .
the analysis of these various objects suggests that the ke may be related to the progenitor s main - sequence mass , which was probably @xmath8 @xmath2 for sn 1999as , @xmath9 @xmath2 for sn 1998bw , @xmath10 @xmath2for sn 1997ef , and @xmath11 @xmath2 for sn 2002ap .
another possible hypernovae , although of type iin , were sn 1997cy and 1999e , which was also estimated to have a large mass ( @xmath12 @xmath2 ; germany et al .
2000 ; turatto et al .
2000 ; rigon et al .
these mass estimates place hypernovae at the high - mass end of sn progenitors , as they are always larger than the mass of the progenitors of normal core - collapse sne , which is @xmath13 @xmath2 . in the following sections , we examine the properties of these hypernovae as derived from optical light curves and spectra and discuss what may be the discriminating factor for the birth of a hypernova or the connection with a grb .
we then focus on nucleosynthesis in hypernovae , which is quite distinct from the case of ordinary supernovae , thus making a unique contribution to galactic chemical evolution .
-band light curves of sne 1998bw ( _ open circles _ ) , 1997ef ( _ open triangles _ ) , 2002ap ( _ stars _ ) , and 1994i ( _ filled circles _ ) ( mazzali et al . 2002 ) . ]
in the early spectra of sn 1998bw , only few features were present .
absorption features were very broad , if present at all , while broad peaks were present at 4000 , 5,000 , 6,000 and 8,000 (galama et al .
the absence of any hydrogen lines , of any clear he i features , and of a strong si ii 6355 line indicated that sn1998bw should be classified as a type ic sn , despite some deviation from previously known objects of this class , mostly due to the extreme line width .
sne ic are thought to be the result of core - collapse - induced explosions of c+o stars , which are massive stars that have lost their h - rich envelope and almost all of their he shell , through either a wind or binary interaction ( nomoto et al .
1994 ) .
sn 1998bw is located in a spiral arm of the barred spiral galaxy eso 184-g82 , for which a distance is measured as @xmath14 mpc from its redshift velocity 2550 km s@xmath15 ( galama et al . 1998 ) and a hubble constant @xmath16 km s@xmath15 mpc@xmath15 .
then , the peak absolute luminosity is estimated to be @xmath17 ergs sec@xmath15 , which is about ten times brighter than typical core - collapse sne ( sne ib / ic , or ii ) . assuming that the grb was at the same distance as sn 1998bw , the gamma - ray fluence of grb980425 reported by the batse ( cgro ) group , @xmath18 erg @xmath19 ( kippen et al .
1998 ) , corresponds to a burst energy @xmath20 ergs ( in gamma - rays ) .
this is four orders of magnitude smaller than for average grbs , implying that either grb980425 might not be a typical burst or that it may not be related to sn 1998bw .
however , the recent case of grb030329/sn2003dh has provided a solid evidence of connection between the ordinary grbs and hypernovae ( stanek et al .
2003 ; hjorth et al . 2003 ; kawabata et al .
wang & wheeler ( 1998 ) studied the correlation between sne and grbs in the literature systematically .
extensive follow - up observations of sn 1998bw have brought us invaluable information on this peculiar sn and put rather strong constraints on corresponding theoretical models .
the early light curve of sn 1998bw has been modeled based on spherically symmetric explosions .
various investigations reached the similar conclusion that the early light curve can be successfully reproduced by the explosion of a massive c+o star with a kinetic energy more than ten times larger than a canonical sn explosion ( iwamoto et al .
1998 ; woolsey et al .
figure [ eps5 ] compares the light curve of sn 1998bw with those of other sne ic and their model light curves ( table 1 ) .
sn 1998bw showed a very early rise .
it had a luminosity of @xmath21 erg s@xmath15 already at day 1 .
this rapid rise requires the presence of @xmath0ni near the surface .
since spherically symmetric explosion models produce @xmath0ni in deep inner layers of the ejecta , they can only reproduce this behaviour if extensive mixing is postulated to have occurred , dredging @xmath0ni up to outer layers ( iwamoto et al . 1998 ; nakamura et al . 2001a ) .
the light curve of sn 1998bw reaches the peak on about day 16 , with the peak absolute magnitude being comparable to that of normal sne ia .
we obtain a qualitative guess of model parameters such as ejecta mass @xmath22 , explosion kinetic energy @xmath23 erg , and the @xmath0ni mass @xmath24 by employing an analytic solution by arnett ( 1982 ) . at early times , the bolometric luminosity @xmath25 of a `` compact '' sn that is powered by radioactive decay of @xmath0ni is written as @xmath26 where @xmath27 days and @xmath28 erg g@xmath15 are the decay time of @xmath0ni and the energy deposited per gram of @xmath0ni , respectively , and the function @xmath29 is given by @xmath30 it is assumed that @xmath0ni is distributed homogeneously and that @xmath31-rays are all trapped in the ejecta .
this latter assumption is correct at the early phases we are considering .
the dimensionless variable @xmath32 , where @xmath33 is the elapsed time and @xmath34 is the characteristic time of the light curve @xmath35 , where @xmath36 and @xmath37 are the hydrodynamical time scale and the diffusion time scale of optical photons through the ejecta , respectively .
the variable @xmath38 is defined as @xmath39 . for the principal mode of diffusion (
arnett 1982 ; pinto & eastman 2000 ) , @xmath34 turns out to be @xmath40 where @xmath41 is the effective opacity .
hereafter , the symbol @xmath42 denotes @xmath43 .
the ejecta are assumed to be a sphere of constant density expanding homologously , and thus @xmath22 and @xmath44 erg are given by @xmath45 , @xmath46 , where @xmath47 , @xmath48 , and @xmath49 are the radius , density , and surface velocity of the ejecta , respectively .
-2 cm .models and their parameters for sne ic .
co138e1 is an ordinary sn ic model , in which a c+o star of @xmath50 ( which is the core of a @xmath51 main - sequence star ) explodes with @xmath52 ergs and @xmath53 .
@xmath54 (= 2 @xmath2 in this case ) denotes the mass cut , which corresponds to the mass of the compact star remnant , either a neutron star or a black hole .
co138e50 , co138e30 , and co138e7 are hypernova models , in which the progenitor c+o star is the same as co138e1 but explodes with different energies .
the mass cut is chosen so that the ejected mass of @xmath0ni is the value required to explain the observed peak brightness of sn 1998bw . [ cols= " < ,
< , < , < , < , < , < " , ] [ tab : pop3 ] we have shown that the ejecta of core - collapse supernova explosions of massive stars can well account for the abundances of emp stars
. we can put further constraint on the typical mass of pop iii .
the abundances of all observed emp stars including the most metal - poor one are inconsistent with the abundance pattern of pair - instability supernovae ( pisne ) . for typical emp stars and cs22948 - 037 , enrichment by pisne can not be consistent with the observed abundant zn / fe and co / fe ratios ( umeda & nomoto 2002 ; heger & woosley 2002 ) .
for he0107 - 5240 and other c - rich emp stars , pisne enrichment is difficult to account for the large c / mg ratios .
therefore the supernova progenitors that are responsible for the formation of emp stars are in the range of @xmath55 @xmath2 , but not more massive than 130 @xmath2 .
this upper limit depends on the stability of massive stars as will be discussed below . to determine the upper limit mass of the zero age main sequence ( zams ) , we analyze a linear non - adiabatic stability of massive ( @xmath56 - @xmath57 ) pop iii stars using a radial pulsation code ( nomoto et al .
because cno elements are absent during the early stage of their evolution , the cno cycle does not operate and the star contracts until temperature rises sufficiently high for the @xmath58 reaction to produce @xmath59c .
we calculate that these stars have @xmath60 , and the central temperature @xmath61 on their zams .
we also examine the models of pop i stars for comparison .
table [ tab : pop3 ] shows the results for our analysis .
the critical mass of zams pop iii star is @xmath62 while that of pop i star is @xmath63 .
this difference comes from very compact structures ( with high @xmath64 ) of pop iii stars .
stars more massive than the critical mass will undergo pulsation and mass loss .
we note that the @xmath65-folding time of instability is much longer for pop iii stars than pop i stars with the same mass , and thus the mass loss rate is much lower .
these results are consistent with ibrahim , boury , & noels ( 1981 ) and baraffe , heger , & woosley ( 2001 ) .
however , the absence of the indication of pisne may imply that these massive stars above 130 @xmath2 undergo significant mass loss , thus evolving into fe core - collapse rather than pisne .
in this paper , we first describe how the basic parameters of hypernova sn 1998bw are derived from observations and modeling , and discuss the properties of other hypernovae individually .
these hypernovae seem to come from rather massive stars , being more massive than @xmath1 20 - 25 @xmath2 on the main - sequence , thus forming black holes . on the other hand ,
there are some examples of massive sne with only a small kinetic energy .
we suggest that stars with non - rotating black holes are likely to collapse `` quietly '' ejecting a small amount of heavy elements ( faint supernovae ) .
in contrast , stars with rotating black holes are likely to give rise to very energetic supernovae ( hypernovae ) .
we present distinct nucleosynthesis features of these two types of `` black - hole - forming '' supernovae .
nucleosynthesis in hypernovae are characterized by larger abundance ratios ( zn , co , v , ti)/fe and smaller ( mn , cr)/fe .
nucleosynthesis in faint supernovae is characterized by a large amount of fall - back , yielding large [ @xmath66/fe ] .
we show that the abundance pattern of the recently discovered most fe deficient star , he0107 - 5240 , and other extremely metal - poor carbon - rich stars are in good accord with those of black - hole - forming supernovae , but not pair - instability supernovae .
this suggests that black - hole - forming supernovae made important contributions to the early galactic ( and cosmic ) chemical evolution .
finally we discuss the nature of first ( pop iii ) stars .
deng , j. , hatano , k. , nakamura , t. , maeda , k. , nomoto , k. , nugent , p. , aldering , g. , & branch , d. 2001 , in _ new century of x - ray astronomy , asp conference series , 251 _ , eds .
h. inoue , & h. kunieda ( asp ; san francisco ) , 238 nomoto , k. , maeda , k. , umeda , h. , ohkubo , t. , deng , j. , mazzali , p. 2003
, in _ iau symp 212 , a massive star odyssey , from main sequence to supernova _ , eds . van der hucht , a. herrero , & c. esteban ( asp , san francisco 2003 ) , 395 ( astro - ph/0209064 ) | during the last few years , a number of exceptional core - collapse supernovae ( sne ) have been discovered .
although their properties are rather diverse , they have the common feature that at least some of their basic parameters ( kinetic energy of the explosion , mass of the ejecta , mass of the synthesized @xmath0ni ) , and sometimes all of them , are larger , sometimes by more than an order of magnitude , than the values typically found for this type of sne .
therefore , these sne have been given the collective classification of ` hypernovae ' .
the best known object in this class is sn 1998bw , which owes its fame to its likely association with the gamma - ray burst ( grb ) 980425 . in this paper , we first describe how the basic parameters of sn 1998bw can be derived from observations and modeling , and discuss the properties of other hypernovae individually .
these hypernovae seem to come from rather massive stars , being more massive than @xmath1 20 - 25 @xmath2 on the main - sequence , thus forming black holes . on the other hand , there are some examples of massive sne with only a small kinetic energy .
we suggest that stars with non - rotating black holes are likely to collapse `` quietly '' ejecting a small amount of heavy elements ( faint supernovae ) .
in contrast , stars with rotating black holes are likely to give rise to very energetic supernovae ( hypernovae ) .
we present distinct nucleosynthesis features of these two types of `` black - hole - forming '' supernovae .
nucleosynthesis in hypernovae are characterized by larger abundance ratios ( zn , co , v , ti)/fe and smaller ( mn , cr)/fe .
nucleosynthesis in faint supernovae is characterized by a large amount of fall - back .
we show that the abundance pattern of the recently discovered most fe deficient star , he0107 - 5240 , and other extremely metal - poor carbon - rich stars are in good accord with those of black - hole - forming supernovae , but not pair - instability supernovae .
this suggests that black - hole - forming supernovae made important contributions to the early galactic ( and cosmic ) chemical evolution .
finally we discuss the nature of first ( pop iii ) stars .
49 pages , to be published in `` stellar collapse '' ( astrophysics and space science ; kluwer ) ed . c. l. fryer ( 2003 )
supernovae , hypernovae , nucleosynthesis , chemical evolution , gamma - ray bursts |
the strongly interacting matter physics serves currently as a pivot topic in the fundamental many - body theory or statistical mechanics context .
an interesting observable that characterizes this kind of physics is the unitary limit thermodynamics .
taking the feshbach resonance techniques , it is now possible to magnetically tune the two - body inter - particle interaction strength .
increasing the scattering length of fermions from @xmath2 to @xmath3 resulting in bound boson systems can exhibit the crossover from bardeen - cooper - schrieffer(bcs ) to bose - einstein condensation(bec ) .
this will lead to the universal thermodynamics of the bethe - peierls contact interaction fermions system with a zero - range , infinite scattering length interaction . to solve the universal resonant superfluidity thermodynamic problem , at least in principle ,
more theoretical efforts are urgent to understand the detailed dynamics by going beyond the mean field theory . at the crossover point
, the dimensionless coefficient @xmath4 relates the total energy @xmath5 with the fermi kinetic energy @xmath6 . here , @xmath7 is the bare fermion mass while @xmath8 is the fermi momentum .
the determination of the fundamental constant @xmath4 as a challenging many - body topic attracts much attention theoretically / experimentally .
it is presently accepted to be in the region @xmath9 . compared with the old values ,
the updating experiments converge more towards @xmath10 . the ultra - cold fermion atoms gas physics offers a plausible perspective in looking for the general nonperturbative statistical field theory methods . in the strongly interacting regime @xmath11 ,
especially right at the unitary limit point @xmath12 , the infrared singular nature of the two - body scattering amplitude @xmath13 ( with @xmath14 being the relative wave - vector magnitude of the colliding particles ) rules out any conventional perturbative expansion techniques .
it is generally believed that there is not a simple analytical method to compute the universal coefficient @xmath4 , although this topic stimulates many growing arduous theoretical efforts . by noting that an exact many - body theory concerns the behavior at unitarity is presently not available
, it is not surprising that the existed analytical results about the universal coefficient @xmath4 differ from each other considerably(@xmath15 ) .
it is worthy noting that there are still uncertain discrepancies among the powerful quantum monte - carlo works themselves as well as the experimental ones existed in the literature . in refs .
, we make an almost first principle detour to attack the universal constant @xmath4 with the general dirac quantum many - body formalism .
the analytical circumvention is motivated by the strongly instantaneous coulomb correlation thermodynamics discussions in a compact nuclear confinement environment based on the relativistic continuum nuclear many - body framework .
the infrared correlation characteristic in the unitary limit promotes us to use the _ universal _ strongly instantaneous long range interaction to attack the infinite contact interaction thermodynamics .
the obtained analytical result @xmath4 is comparable with other theoretical quantum monte - carlo attempts .
it is derived through a medium dependent interaction by accounting for the interior correlation with a counteracting or anti - correlation external source static potential .
however , the auxiliary quantities pressure @xmath16 and chemical potential @xmath17 do not obey the profound universal hypothesis in the unitary limit with @xmath1 . according to the universal hypothesis at unitarity with @xmath1 ,
the ratio of pressure to energy density should be still that of non - interacting fermion gas .
therefore , the physical pressure should be @xmath18 instead of @xmath19 obtained in refs . and consequently affects the important sound speed .
the chemical potential should be @xmath20 instead of @xmath21 .
intriguingly , the physical helmholtz free energy density is not affected . at @xmath1
, it can be @xmath22 .
therefore the underlying energy(-momentum ) conservation law is still exactly ensured even with the respective discrepancy of the auxiliary pressure and chemical potential .
this observation instructs us to reanalyze the physical pressure and system chemical potential carefully .
the hugenholtz - van hove(hvh ) theorem characterizes the fundamental thermodynamic relation between energy density and pressure .
it is well known that any effective medium dependent interaction with brueckner - bethe - goldstone(bbg ) techniques can easily violate this relation . the standard many - body technique to remove this tortuous difficulty is considering the quantum rearrangement effects of the single particle energy spectrum on the physical chemical potential and pressure .
the role of the single particle energy spectrum shift will be explored in the strongly interacting unitary limit . to elucidate the physical chemical potential and pressure(equation of state ) at both finite temperature and density transparently with a more concise non - relativistic recipe for the ultra - cold atomic physics is the present main motivation .
up to now , quantum statistical field theory of the strongly interacting matter at finite temperature and density is still a challenge .
more concretely , the exact determination of equation of state plays an important role in looking for the detailed knowledge about the collective excitation modes measured with the ultra - cold atomic physics experiments .
the present work is organized as follows . in sec .
[ sec2 ] , the adopted low energy effective theory formalism and corresponding nonperturbative approach for the strongly interacting unitary gas thermodynamics are described . through taking into account the single particle energy rearrangement effect ,
the unitary fermions thermodynamics with the dynamically self - consistent dyson - schwinger equations is presented in sec .
moreover , the universal thermodynamics at unitarity with @xmath1 as well as the concrete comparisons with existed results is discussed in sec .
section [ sec5 ] includes the discussion and prospective remarks .
the summary conclusion is made in the final sec .
[ sec6 ] . in this work ,
the calculations are performed in the non - relativistic contact interaction formalism . throughout the paper ,
the natural units @xmath23 are employed .
the usual bare bethe - peierls hamiltonian with a global @xmath24 symmetry addressing the many - body ground state energy can be found in the literature .
the in - medium behavior associated with the many - body correlation characteristic is the key , while a nonperturbative approach is crucial for the strongly interacting unitary regime physics .
motivated by previous works , we work with the medium - scaling universal functional version [ hamiltonian ] h=&&-d^3x _
^*(x ) ( -_r[n])_(x ) + & & + 2 d^3x_^*(x)^*_(x ) _ ( x)_(x ) . in , m is the bare fermion mass with @xmath25 representing the ( hyperfine-)spin projection .
the physical many - body renormalized interaction potential @xmath26 $ ] is a functional of density and temperature , from which one can define the in - medium @xmath27-wave scattering length @xmath28 u_eff^*[n , t ] .
the bare interaction matrix @xmath29\rightarrow u_0 $ ] is directly related with the free - space scattering length @xmath30 with the low energy born approximation for the zero range contact interaction . considering the medium influence of the surrounding environment on the single particle energy spectrum properties , we have included the rearrangement term @xmath31 $ ] in .
it is introduced through the auxiliary quantity @xmath32 indicated by the tilde symbol @xmath33=\mu_\alpha -\mu_{r\alpha}[n]$ ] .
it means that the collective correlation effects are further accounted for as a single particle potential formalism in the spirit of density functional theory , while the true lagrange multiplier is @xmath34 . as a functional of density and temperature ,
its concrete medium dependence can be derived analytically from the thermodynamic relation for the fixed particle number density . theoretically ,
the derivation of the many - body renormalized effective interaction @xmath35 is beyond the bare bethe - peierls contact interaction hamiltonian itself , i.e. , it should be derived from the more underlying physical law .
firstly , from the point of view of the contact interaction fermions physics , it is extremely instructive to recall the primary intermediate vector boson ( ivb ) hypothesis in weak interaction theory .
it is well known that the local intermediate vector boson theory is related with the current - current(cc ) contact interaction version through the corresponding connection : @xmath36 . in the low energy limit @xmath37 , the two ivb and cc theories are ingrainedly identical . to model the low energy unitary limit physics ,
@xmath38 by taking @xmath39 to be an arbitrary _ large _ charge or with an arbitrary _ small _ mass gap @xmath40
. consequently , the counteracting environment associated with an additional @xmath24 conserved electric charge " can be introduced to attack the challenging infinity .
constrained by the opposite charge neutralizing background , the effective interaction @xmath35 is found to appear as an instantaneously anti - screening formalism .
this particular analogy and/or discursion makes the coupling constant @xmath39 enter simultaneously in the denominator and numerator of fractions in the relevant thermodynamic expressions .
the assumed electric charge @xmath39 is reduced by the physical constraint or charge neutralizing stability condition and its concrete magnitude does not affect the final unitary physical results . therefore , there are not any regularization or expansion parameters violating the intriguing unitarity in gauging the infrared singular divergence .
this spirit will be further explored in this more concise non - relativistic recipe . moreover , corresponding to counteracting the deviation tendency of charge neutralizing , there exists a similar general le chatelier s principle in thermodynamics .
this profound stability principle accounts for frustrating an instantaneous departure from equilibrium with an alternating _ minus _ function of the surrounding environment . first of all , as we will see , the counteracting - antiscreening formalism makes it possible for us to use the algebra equations to characterize the classical thermodynamics with the significant quantum correlation effects , but of crucial importance is that the many - body renormalization effects are introduced through a mirror background with an alternating minus function to avoid the theoretical double counting trouble .
the following calculations based on can be characterized by the comprehensive feynman diagrams presentation with the generalized coupled dyson - schwinger equations(dses ) formalism as indicated by fig.[fig1 ] .
the polarization tensor @xmath41 is calculated through the non - relativistic random phase approximation(rpa ) with the full fermion propagator @xmath42 as described by fig.[fig1]b ^ * ( _
n , * p*)&&=t _ _ 1 _ k g ( _ 1 , * k * ) g(_1+_n , * k*+*p * ) + & & = -_k , with @xmath43 . here , @xmath44 is the fermi - dirac distribution function given explicitly in .
after employing the analytical continuation for the matsubara frequency @xmath45 according to @xmath46 , the @xmath47 determines the dispersion relation of the excitation mode a ( the composite boson / dimer ) in - medium propagating .
the polarization function obtained in this way is the standard in - medium lindhard density - density or spin - spin response function for the one component fermions gas . in this work ,
we care about the low energy long wavelength thermodynamics of the fully pairing correlation occasion , which is related to the static infrared limit [ debye ] ^ * ( 0 , |*p*| 0)=-=1t()^3/2 _ 1/2[y ] .
the @xmath48 appearing in is the physical non - relativistic debye screening mass squared with unit electric charge and reflects the quantum fluctuation contributions consequently .
its magnitude is the spin - spin or density - density correlation response susceptibility in the static infrared limit . at @xmath49 ,
@xmath48 can be reduced to the familiar formalism m_d^2|_t=0=. different from the multiplier - system physical chemical potential @xmath50 , the symbol @xmath51 employed in above equations is the effective(total ) chemical potential .
it characterizes the additional many - body collective correlation effects and will be defined explicitly in the following discussion for the full fermion propagator . here and afterwards , the function identifier @xmath52 $ ] stands for the standard fermi integral mathematica polylogarithm function @xmath53 $ ] .
meanwhile , we will use the shorting notation @xmath54 for the sake of conciseness throughout the paper . with the polarization function @xmath55
, the rearrangement interaction matrix is derived to be [ eff - p ] u^*_eff[n , t]&&=. at @xmath1 , it can be reduced to u^*_eff[n]=. the @xmath56 is the un - perturbated density of states on the fermi surface for one component fermions , which is one half of the physical debye screening mass squared . without a loss of generality and conformity , the screening mass squared notion @xmath48 instead of the density of states will be used to characterize the general finite temperature thermodynamics . compared with the conventional screened interaction customarily applied in the weak coupling perturbative loop diagram ring / ladder resummations [ ind - p ] u_ind(_n , * p*)&&=,there is a key dynamically mirrored _ minus _ sign difference in the denominators of and , which leads to the quite different physical motivations and calculational details .
the full fermion propagator @xmath42 given with the instantaneously anti - screening interaction matrix tells us all the information of the ground state energy or quasi - particle distribution function .
it is calculated with the proper self - energy @xmath57 through the instantaneous approximation as indicated by fig.[fig1]a [ propagator ] g_&&=g_0+g_0^ * g _ , + ^*(0,0)&&=u_eff^*(0,|| 0)n_=u_eff^*[n]2 . here , @xmath58 is the total particle number density @xmath59 for the unpolarized / fully pairing symmetric trapped dilute atomic fermi gas .
the self - energy @xmath60 is four - momentum independent in the instantaneous approximation .
it means that the self - energy will shift the chemical potential of the particle distribution function , i.e. , the correlation contributions to thermodynamics are characterized through the phase space energy - momentum distribution _
deformation_. from the full fermion propagator with ,
the particle number and energy densities are given by nv&&=t_k_n e^i_n ( _ n,*k * ) , + ev & & = t_k_n e^i_n 12 ( i_n + + ) ( _ n,*k * ) .
although we once again consider the linear contact formalism , the lowest order diagrams characterize the thermodynamics with the highly nonlinear / turbulent contributions of the density and temperature fluctuations due to taking the medium dependent interaction potential .
after summing the matsubara discrete frequency @xmath45 with the residual theorem , the particle density @xmath61 and energy density @xmath62 are reduced to the compact formalisms @xmath63,&\\ \label{energy}\epsilon&=2\int_k \0{{\bf k}^2}{2m}f_{\bf k}+\0{\pi a _ { eff}}{m}n ( t,\lambda)^2&\no\\ & = -3t(\0{m t}{2 \pi } ) ^{3/2 } \mbox{li}_{5/2}[y]+\0{\pi a _ { eff}}{m}n^2 , & \end{aligned}\ ] ] with the entropy density @xmath64 [ entropy ] s&&=-2_k @xmath65 + & & = 1t()^3/2@xmath66 - 5t\mbox{li}_{5/2}[y]$ ] .
the in - medium effective scattering length is a_eff = a1-a . in deriving above equations
, we have defined the quasi - particle fermi - dirac distribution function with the total / effective chemical potential @xmath67 through fig.[fig1]a with [ fermi - dirac ] f_*k*&&= , + & & = -^*(0,0 ) , with @xmath68 being the inverse temperature .
the expressions for the entropy density and the first energy density term in are similar to those of the non - interacting fermion gas .
the total chemical potential @xmath67 is determined for the given particle number density @xmath58 and temperature @xmath69 .
the above algebra equations are completely _
closed_. from the helmholtz free energy density @xmath70 , the pressure is given by [ pressure - d ] p=-@xmath71_t , n=-@xmath72_t = n^2@xmath73_t . at @xmath1 with entropy density @xmath74 , @xmath75 is reduced to the important formalism[p - e ] p = n^2 @xmath76 .
the equation determines the fundamental relation between the energy density and pressure according to the hvh theorem , which plays a vital role for characterizing the symmetric nuclear matter saturation property .
as widely discussed in the nuclear many - body literature , the hvh theorem constraint will lead to the rearrangement terms in the pressure and chemical potential dealt with any kind of in - medium interaction potential .
presently , their exact analytical expressions can be derived in a thermodynamic way instead of doing the numerical interpolation .
let us further address the pressure @xmath75 . from given with
the medium dependent interaction potential , one can have [ pressure ] p=&&n^2+c(t , ) @xmath77 ^ 2 n^3+p_ideal(t , ) , + with p_ideal(t,)=&&-2t@xmath78 ^ 3/2_5/2[y ] .
in , the first term represents the expected direct contribution of the induced interaction , while the last one results from the fermi kinetic energy .
the essential second term is due to the rearrangement contributions , which can be easily neglected .
it is worthy noting that this rearrangement term is vanishing within the standard mean field theory approach .
the discussion for the physical system chemical potential / multiplier is in line with that of pressure = @xmath79_t , v=@xmath80_t=@xmath81_t .
its analytical expression can be further reduced to [ chemical ] = n+c ( t,)@xmath77 ^ 2 n^2 + .
there is an additional(second ) term in the system chemical potential resulting from the intriguing rearrangement effect of the single - particle degrees of freedom analogous to that of pressure .
explicitly , the rearrangement terms in the pressure and chemical potential are similar . for the general finite temperature occasion ,
the rearrangement factor @xmath82 appearing in and is given by ( t,)m_d@xmath83_t=. from the pressure and chemical potential analytical expressions , one can see that the rearrangement effects play the positive repulsive " role .
these effects prevent the system from collapsing in the strongly coupling attractive unitary limit .
furthermore , the in - medium spontaneously generated rearrangement effects are exactly counteracted(easily neglected ) in the helmholtz free energy density @xmath84 .
this reflects an important physical fact that the in - medium off - shell(spontaneous lorentz violation ) formalism can approach the quantum correlation effects on the universal thermodynamics , while the thermodynamics first law constraint ensures the energy conservation .
exactly , the classical thermodynamic consistencies are satisfied with the generalized self - consistent dses .
from the general analytical expressions , one can give the @xmath1 results with the in - medium interaction u_eff^*= a_eff=. with @xmath85 and the rearrangement factor ( _ f)=m_d= , the physical internal energy density , pressure and the system chemical potential are @xmath86 in the strongly interacting universal unitary limit @xmath87 , the ratios of energy density , pressure and chemical potential to those of non - interacting fermion gas are @xmath0 .
this @xmath0 is exactly consistent with the existed theoretical results and in agreement with the updating experimental ones . in the weak coupling limit , the analytical thermodynamic expressions readily return to the lowest order mean field theory results which neglect the quantum fluctuation / correlaion contributions . at unitarity and with @xmath1 ,
the pressure is p=445m(3 ^2)^2/3n^1 + 2/3=p_ideal , with which the sound speed is calculated to be v==23 v_fg = v_fg .
the @xmath88 is that for the ideal non - interacting fermion gas with @xmath89 being the fermi speed .
a detailed information of the sound speed of the homogeneous system is very crucial for determining the frequencies of collective modes in the realistic trapped systems . with @xmath1 ,
the ratio of pressure to binding energy is still @xmath90 of the non - interacting fermions gas in the unitary limit as pointed out in the literature .
the hvh theorem ensures the thermodynamic consistency dealt with the in - medium interaction potential in addressing the strongly interacting physics .
the same ratio is counterintuitive to some extent , which indicates that the universal thermodynamic properties with the infrared long range quantum fluctuation / correlation in the strongly interacting limit are extremely profound .
the numerical results for @xmath91 can be readily given in the unitary limit regime with the explicit analytical expressions , from which one can see that the collective strongly correlating and rearrangement effects influence the chemical potential and pressure simultaneously .
however , the internal energy and helmholtz free energy densities do not directly rely on the intriguing rearrangement effect even with a careful numerical check .
in addressing the unitary fermions physics , the essential task is how to incorporate the significant collective density and temperature fluctuation / correlation contributions into the thermodynamics in a reasonably controllable way .
the physical connotations are crystal clear in this analytical attempt : on one hand , the many - body surrounding environment modulates the bare two - body interaction , which is characterized by the in - medium renormalized interaction matrix @xmath29 $ ] . on the other hand ,
the collective correlation effects further shift the single particle energy spectrum , which is described by the corresponding single particle correlation potential @xmath92 $ ] .
the crucial energy - momentum conservation law is guaranteed by the _ simultaneous _ consideration of these two collective correlation effects .
the fundamental thermodynamic consistency is the basic validity requirement of any scientific theoretical approaches , which can be easily neglected .
the generic infrared singularity of the bare two - body scattering amplitude at unitarity provides the scaling basis , which leads to the thermodynamic universal hypothesis .
this scaling property implies that the thermodynamic expressions can be very simple in principle , i.e , the analytical formulae should be similar to those for the ideal non - interacting fermions gas . in turn
, this infrared singularity makes any potential theoretical approaches nontrivial
. essentially , the unitary characteristic rules out any naive expansions or conventional perturbative loop diagram ring / ladder resummation calculations . in this work , we have obtained the @xmath1 dimensionless universal coefficient @xmath0 for the @xmath93 unitary fermion gas .
it is in the regime of the reasonable simulation / experimental values @xmath94 .
however , as discussed recently in the literature , it is yet far from the ending of looking for its realistic exact value .
the exact determination of this fundamental constant will influence the detailed knowledge about the collective excitation modes measured in the realistic ultra - cold unitary atomic fermions .
consequently , more and more theoretical / experimental efforts are in fact most expected to make it certain precisely . to understand the comprehensive thermodynamic properties exhibited by these novel trapped strongly interacting quantum systems
, there is still a faraway route before us to constitute a consistent microscopic theory and corresponding analytical calculational framework .
in summary , based on the dyson - scwhinger equations formalism , the ground state of the @xmath93 ultra - cold unitary fermion gas has been calibrated in detail in terms of the quantum ising - like bethe - peierls contact interaction model .
the universal factor is calculated to be @xmath0 . in addressing the universal resonant thermodynamics
, the medium dependent interaction through a dynamically twisted minus sign caused by an assumed mirror background makes it possible for us to incorporate the landau pole contributions with the fermi surface instability . with the spontaneously generated rearrangements due to the single particle energy spectrum shift , the additional instantaneous correlation potential ensures the classical thermodynamic consistency .
finally , we want to emphasize that there are not any infrared and/or ultraviolet cutoffs or regularization and/or expansion parameters obviously violating the significant unitarity , i.e. , the bare @xmath95 drops out in the unitary limit . | based on previous works , analytical calculational procedures for dealing with the strongly interacting fermions ground state are further developed through a medium dependent potential in terms of the bethe - peierls contact interaction model .
the methods are exact in the unitary limit regime and they lead to the self - consistent equations analogous to the hartree ones . the single particle energy spectrum rearrangement effects on the thermodynamics due to the hugenholtz - van hove theorem constraint are addressed . these effects lead to an additional instantaneous correlation potential contribution to the system physical chemical potential and pressure , i.e. , equation of state , which enforces the classical thermodynamic consistency .
the dyson - schwinger equations represent implicitly the various bethe - goldstone expansion ones . in a thermodynamically self - consistent way
, the universal dimensionless factor is analytically calculated to be @xmath0 , which defines the ratio of the unitary fermions energy density to that of the ideal non - interacting ones at @xmath1 . |
in the nonsupersymmetric @xmath9 extension of the standard @xmath10 model ( sm ) of particle interactions , the higgs sector must be enlarged from the one @xmath3 scalar doublet of the sm .
there are several ways to do this , as discussed comprehensively in ref .
@xcite . in the canonical approach ,
a higgs triplet is used to break @xmath2 at a large scale , and @xmath11 gets a large majorana mass .
a higgs bidoublet is then added to break @xmath3 , and all fermions obtain dirac masses , with @xmath12 getting a small seesaw mass . in this scenario ,
the @xmath2 breaking scale is presumably beyond the reach of present accelerators , such as the large hadron collider ( lhc ) .
even if we try to lower this scale , the canonical model has severe difficulties with flavor - changing neutral currents , in contradiction with what is experimentally observed .
the purpose of this paper is to elaborate on a simple alternative @xcite , where the @xmath2 breaking scale may be lowered to 1 tev , using the inverse seesaw mechanism for neutrino mass @xcite .
we choose a higgs sector which contains only one @xmath0 bidoublet and one @xmath2 doublet [ but no @xmath3 doublet ] .
of course , flavor - changing neutral higgs couplings are still unavoidable . however , as we show in this paper , a scenario exists where they are sufficiently suppressed . since the @xmath2 charged - current mixing matrix is unknown , we consider two scenarios , where it is given either by @xmath4 ( scenario i ) or @xmath5 ( scenario ii ) . we consider the bounds on the corresponding scalar particle masses from @xmath6 and @xmath7 mixing , as well as @xmath8
we find that , whereas in scenario i , they are of order 10 tev , as in other left - right models , they may be well below 1 tev in scenario ii , thus allowing them to be within reach of detection at the forthcoming large hadron collider ( lhc ) .
the fermion content of the minimal @xmath13 gauge model is well - known , i.e. @xmath14 where the @xmath15 charge is normalized to @xmath16 so that the electric charge is given by @xmath17 . here
a neutral fermion singlet @xmath18 is also added per family , which will have important implications for the neutrino masses , as shown below . to obtain masses for the quarks and leptons ,
a higgs bidoublet @xmath19 is needed . in a nonsupersymmetric model , which is being considered here , the dual of @xmath20 , i.e. @xmath21
must also be used . to break @xmath22 to @xmath23 , an @xmath2 higgs doublet @xmath24
is added , which also links @xmath25 with @xmath26 to form a dirac mass @xmath27 .
since @xmath26 is a gauge singlet , it is also allowed to have a majorana mass @xmath28 ; hence the @xmath29 neutrino mass matrix spanning @xmath30 is of the form @xmath31 where @xmath32 is the usual dirac mass linking @xmath33 to @xmath11 through @xmath34 and @xmath35 .
a quick look at the above shows clearly that if @xmath36 , then lepton number is conserved with a linear combination of @xmath12 and @xmath26 forming a dirac fermion with @xmath11 , and the orthogonal combination is exactly massless .
this means that it is natural for @xmath28 to be small , thereby triggering the inverse seesaw mechanism , resulting in @xmath37 note that there is no entry in eq .
( 7 ) linking @xmath12 and @xmath26 because the @xmath3 higgs doublet is absent .
this is important for the validity of eq .
( 8) . instead of the canonical seesaw formula @xmath38 , which is small if @xmath27 is large , eq .
( 8) lets @xmath39 be small if @xmath28 is small , even if @xmath27 is not too large .
thus the inverse seesaw mechanism is suitable for bringing down the scale of @xmath2 breaking to 1 tev , with verifiable phenomenology at the lhc .
note also that the mixing of @xmath12 with @xmath26 is of order @xmath40 which may now be nonnegligible and results in deviations from unitarity @xcite of the neutrino mixing matrix . the most general higgs potential consisting of @xmath41 , @xmath20 , and @xmath42 is given by @xmath43 ^ 2 + { 1 \over 2 } \lambda_2 tr(\phi^\dagger \phi \phi^\dagger \phi ) \nonumber \\ & + & { 1 \over 8 } \lambda_3 \ { [ tr(\phi^\dagger \tilde \phi)]^2 + [ tr(\tilde{\phi}^\dagger \phi)]^2 \ } + { 1 \over 2 } \lambda_4 [ tr(\phi^\dagger \phi)][tr(\phi^\dagger \tilde \phi + \tilde{\phi}^\dagger
\phi ) ] \nonumber \\ & + & f_1 \phi_r^\dagger ( \tilde{\phi}^\dagger \tilde{\phi } ) \phi_r + f_2 \phi_r^\dagger ( \phi^\dagger \phi ) \phi_r + f_3 \phi_r^\dagger ( \phi^\dagger \tilde{\phi } + \tilde{\phi}^\dagger \phi ) \phi_r,\end{aligned}\ ] ] where all parameters have been chosen real for simplicity .
let @xmath44 and @xmath45 , then the minimum of @xmath46 is given by @xmath47 where @xmath48 and @xmath49 satisfy @xmath50 @xmath51 + v_2 ( \mu^2 + f_3 v_r^2 + \lambda_4 v_2 ^ 2 ) = 0,\ ] ] @xmath52 + v_1 ( \mu^2 + f_3 v_r^2 + \lambda_4 v_1 ^ 2 ) = 0.\ ] ] a solution exists where @xmath53 , i.e. @xmath54 with @xmath55 fine tuning is of course unavoidable . in the limit @xmath56 , the physical higgs bosons are @xmath57 and @xmath58 with masses squared @xmath59 and three linear combinations of @xmath60 , @xmath61 , @xmath62 , with mass - squared matrix @xmath63 the structure of the scalar sector leads in general to both @xmath64 and @xmath65 mixing .
the former vanishes in the limit @xmath66 and so will be suppressed for the above choice of vacuum expecation values .
in contrast , the @xmath67 mixing term is proportional to @xmath68 , which is unacceptably large . to cancel this contribution ,
a simple possibility is to add a higgs bidoublet @xmath69 with vacuum expectation value @xmath70 . in that case , the choice @xmath71 ( for @xmath72 ) will lead to zero mixing at tree level ; details are given in the appendix . note that @xmath73 will not affect the @xmath74 parameter ( at tree level ) in precision electroweak measurements , nor will it contribute to quark or lepton masses .
in particular , it does not link @xmath12 with @xmath26 in eq .
( 7 ) , otherwise the inverse seesaw mechanism would be invalidated .
the present experimental limits on @xmath75 and @xmath76 are respectively 715 and 860 gev .
since both @xmath20 and @xmath77 couple to the quarks and leptons , flavor - changing interactions through the exchange of neutral higgs scalars are unavoidable . the question is whether they can be suppressed @xcite . consider the yukawa terms @xmath78 in the limit @xmath56 , both @xmath79 and @xmath80 quark masses come from only @xmath81 .
hence @xmath82 where @xmath83 and @xmath84 are unitary matrices , with @xmath85 being the quark mixing matrix for the known left - handed charged currents and that for their unknown right - handed counterparts .
this means that in the basis of quark mass eigenstates , the structure of flavor - changing neutral currents through scalar exchange is determined , i.e. @xmath86 for the @xmath79 quarks , and @xmath87 for the @xmath80 quarks .
hence @xmath60 behaves as the sm higgs boson , and at tree level , all flavor - changing effects come from @xmath88 , whereas in one loop , there are also contributions from @xmath89 .
note that for @xmath90 , this electroweak doublet has the common mass of @xmath91 . in the lepton sector , the analog of @xmath92 is unknown because the neutrino mass matrix depends on @xmath32 , @xmath27 , and @xmath28 .
in fact , we could choose @xmath32 to be diagonal in the @xmath93 basis and still have the freedom to obtain the observed neutrino mixing matrix from @xmath27 and @xmath28 . in that case
, @xmath88 would have no flavor - changing leptonic interactions .
we now apply eq .
( 22 ) to @xmath6 and @xmath7 mixing . in the two scenarios
i and ii considered for the @xmath94 matrix mentioned in the introduction , the @xmath88 couplings are of the form : @xmath95 we use the formulae presented in ref .
the mass difference of a neutral meson and its antiparticle is written in terms of its sm and other contributions : @xmath96 where @xmath97 , and @xmath98 denotes the sm ( one - loop ) contribution , and @xmath99 is everything else . in our case ,
the latter comes from the flavor - changing @xmath100 couplings .
the resulting expression for the mass difference is then given by @xmath101\ ] ] where the constant @xmath102 includes strong - interaction effects , and the coefficients @xmath103 include next - to - leading qcd corrections , while the functions @xmath104 denote the wilson coefficients of the ope expansion for the relevant hadronic matrix elements .
let us consider first case ( i ) of our model , i.e. @xmath4 . here the wilson coefficients @xmath105 are equal : @xmath106,\ ] ] and suppressed because the mass difference between @xmath62 and @xmath58 is small compared to their sum , whereas @xmath107 is of the form : @xmath108,\ ] ] which has no such suppression . in case ( i ) , the various @xmath109 s in each system are also the same : @xmath110 , where @xmath111 we have also assumed for simplicity that all the @xmath92 entries are real .
obviously there are large contributions coming from those terms proportional to @xmath112 or @xmath113 . however , there is also a natural suppression for the @xmath114 and @xmath115 wilson coefficients , because their contributions are proportional to the effective @xmath116 propagator , i.e. @xmath117 . whereas @xmath118 is a mass eigenstate , @xmath119 is not , but if @xmath120 and @xmath121 are small in eq .
( [ m_scalar ] ) , then it is approximately so , and their combined contribution for @xmath90 is naturally suppressed , i.e. @xmath122 this suppression persists even if @xmath120 and @xmath121 are not neglected .
we simply replace @xmath123 by @xmath124 this feature of our model would allow @xmath48 to be at the tev scale , without running into conflict with present data on @xmath6 and @xmath7 mixing as far as @xmath114 and @xmath115 are concerned . unfortunately , this suppression does not work for @xmath125 , which is proportional to @xmath126 .
however , as we show below in case ( ii ) , the @xmath125 coefficients are further suppressed by light quark masses in the @xmath109 s , which allows @xmath88 to be lighter than 1 tev . in case ( ii ) , i.e. @xmath127 , the @xmath109 values are related by @xmath128 , with @xmath129 from the above , it is clear that whereas @xmath125 is not suppressed by eq .
( [ propagator ] ) , it is much smaller than what it is in case ( i ) , because of the smallness of @xmath130 . as mentioned in ref .
@xcite , there are large theoretical uncertainties associated with these expressions . to make an estimate
, we simply require the absolute value of the contribution of new physics to be less than the corresponding experimental value . in
what follows we shall obtain bounds for the combination of parameters : @xmath131 and @xmath132 .
let us define : @xmath133 and @xmath134 , where @xmath135 is the approximate mass of @xmath136 or @xmath137 , and @xmath138 is a meassure of the splitting between their squared masses .
using eqs .
( [ wilsona ] ) and ( [ wilsonb ] ) , we obtain the following general expression : @xmath139 \frac{p^{sll}_1 \delta^2 } { m^4_2 } = \frac { 3 } { 8 s_x } \delta m_{x}^{exp}.\ ] ] for the @xmath6 system , @xmath140 , with @xmath141 mev , @xmath142 mev , @xmath143 and @xmath144 . at the scale @xmath145 gev , @xmath146 , @xmath147 , @xmath148 mev .
notice that in case ( i ) , both @xmath149 and @xmath150 dominate over the @xmath151 and @xmath152 contributions .
threfore the resulting bound is not sensitive to the parameter @xmath153 , and the bound on @xmath135 is given by @xmath154 for the @xmath7 systems , we take the corresponding parameters from the particle data group @xcite , so that for @xmath155 : @xmath156 these results are in agreement with @xcite . in case ( ii ) , if we take @xmath157 ( i.e. only the @xmath158 contribution ) , we obtain a much smaller bound for the @xmath159 system , i.e. @xmath160 tev . however , for the @xmath161 , and @xmath162 systems , the same procedure yields the bounds @xmath163 gev , respectively .
thus for the @xmath161 system , it seems more appropriate to consider @xmath164 , in which case the bound becomes : @xmath165 tev . to evaluate the contribution of @xmath89 to @xmath8 , we consider the relevant terms in eq .
( 22 ) . for case ( i ) , i.e. @xmath4 , the important ones are @xmath166 for case
( ii ) , i.e. @xmath5 , they are @xmath167 the sm contribution ( from @xmath168 exchange ) is of the form @xmath169 which is classified @xcite as @xmath170 . using the above interactions ,
there is only one such contribution coming from @xmath88 exchange , i.e. @xmath171 and @xmath172 , which is proportional to @xmath173 in both cases ( i ) and ( ii ) , assuming that @xmath174 , which is of course a very good approximation .
in contrast to the usual two - higgs - doublet model , the @xmath175 contribution is suppressed here because it is proportional to @xmath176 . as for @xmath177 ,
i.e. @xmath178 , both @xmath88 and @xmath175 have contributions proportional to @xmath173 , but since the @xmath8 rate is proportional to @xmath179 the latter can be safely ignored
. using ref .
@xcite , we find @xmath180 \\ a_{\phi^0_2 } & \sim & \frac{m^2_t}{m^2_{\phi^0_2 } } [ -\frac{1}{3 } f_1(x_b)]\end{aligned}\ ] ] where @xmath181 , @xmath182 , and the functions @xmath183 are given by @xmath184 we now require the amplitude ratio @xmath185 to be less than 10% , so that it is well within the experimental accuracy .
this translates to an estimated lower bound for @xmath186 of about 200 gev , as shown in fig .
[ fig ] . vs @xmath186.,width=491 ]
we have studied in this paper a simple nonsupersymmetric left - right extension of the standard model .
the asymmetric higgs sector of this model consists of one @xmath0 bidoublet and one @xmath2 doublet [ but no @xmath3 doublet ] . with the addition of neutral fermion singlets , the inverse seesaw mechanism for neutrino mass
is naturally implemented , suggesting that the @xmath2 breaking scale may be lowered to 1 tev .
we then analyzed the unavoidable problem of flavor - changing couplings of the neutral higgs bosons of this model and showed that in the limit of @xmath187 , these effects are naturally suppressed in case ( ii ) ( @xmath5 ) [ but not in case ( i ) ( @xmath4 ) , which has the same constraint as other left - right models that the @xmath2 breaking scale is above 10 tev .
] from @xmath6 and @xmath7 mixing , we find @xmath188 to be consistent with less than about 1 tev in case ( ii ) . from @xmath189 , we find @xmath186 to be above 200 gev .
the new particles of this model , i.e. @xmath190 , @xmath76 , the heavy pseudo - dirac neutral fermion of mass @xmath27 from the pairing @xmath26 with @xmath11 , and the heavy higgs particles @xmath61 and @xmath89 , are all consistent with having masses below 1 tev in case ( ii ) and are potentially observable at the lhc .
this work was supported in part by the u. s. department of energy under grant no .
de - fg03 - 94er40837 , by uc - mexus , and by conacyt .
we also thank m. holthauseen for comments and pointing out a discrepancy in the first version of this paper .
with only the bidoublet @xmath20 , our model exhibits an unavoidable @xmath191 mixing term proportional to @xmath192 , implying thus a very large value of @xmath48 .
this can be remedied by enlarging the scalar sector through the addition of another bidoublet @xmath193 of @xmath194 , = ~(1,2,2,-1 ) , and its corresponding dual @xmath195 .
we list in this appendix the modifications resulting from its addition .
let the neutral components of @xmath20 , @xmath41 , and @xmath196 acquire vacuum expectation values = v_1,2 , = v_r , = v_3 , and denote the neutral gauge bosons associated with @xmath197 by @xmath198 and the @xmath199 gauge boson by @xmath200 , with @xmath201 and @xmath202 the gauge couplings for @xmath203 and @xmath204 respectively .
the resulting mass - squared matrix in the @xmath205 basis is then given by ^2 = 2 .
the photon @xmath206 , the neutral gauge boson @xmath207 of the sm , and the new @xmath76 are then linear combinations , determined according to = ; = e , where @xmath208 and the weak - mixing angle @xmath209 and the proton charge @xmath210 are defined by = , = + + . in terms of these fields , the above @xmath29 mass - squared matrix is reduced to a @xmath211 one , spanning only @xmath212 with entries ^2 & = & 2 ( v_1 ^ 2 + v_2 ^ 2 + v_3 ^ 2 ) ; ^2 & = & 2 ; & = & - 2 , where @xmath213 is the @xmath67 mixing term . for the charged vector bosons , the analogous mass terms are ^2 & = & 2 ^2 ( v_1 ^ 2 + v_2 ^ 2 + v_3 ^ 2 ) ; ^2 & = & 2 ^2 ( v_1 ^ 2 + v_2 ^ 2 + v_3 ^ 2 + v_r^2 ) ; & = & - v_1 v_2 . note that the @xmath74 parameter is one at tree - level , i.e. @xmath214 , in the absence of mixing , i.e. @xmath215 .
this can be achieved by taking @xmath216 as already discussed in the text , and requiring v_3 ^ 2 = ^2 v_1 ^ 2 ( 1 - 2 ^ 2 _
w)v_1 ^ 2 .[eq : v3 ] without this cancellation from @xmath73 , @xmath217 would have been unacceptably large . with the addition of @xmath196 , more terms occur in the higgs potential : v_x & = & m_^2 + _ 1 | |^2 + _ 2 | |^2 + _ 3 | ^|^2 & + & _ 4 + _ 5 ( ) ^2 + _ 6 ( ) ( x^x ) & + & _ 7 + _ 8 | |^2 ( x^x ) + _ 9 ( ) & + & _ 10 + _ 11 + _ 12 ( ) & + & _ 13 ( ) ^2 + _ 14 , where @xmath218 .
the full potential is then @xmath219 .
the minimum value of @xmath46 , which we denote by @xmath220 , occurs when the various neutral fields are set equal to their corresponding s : v_0 & = & m_r^2 v_r^2 + m^2 ( v_1 ^ 2 + v_2 ^ 2 ) + 2 ^2 v_1 v_2 + 1 2 _
r v_r^4 + 1 2 _ 1 ( v_1 ^ 2+v_2 ^
2)^2 + 1 2 _ 2 ( v_1 ^ 4 + v_2 ^ 4 ) & + & _ 3 v_1 ^ 2 v_2 ^ 2 + 2 _ 4 ( v_1 ^ 2 + v_2 ^ 2 ) v_1 v_2 + f_1 v_1 ^ 2 v_r^2 + f_2 v_2 ^ 2 v_r^2 + 2 f_3 v_1 v_2 v_r^2 & + & m_^2 v_3 ^ 2 + _ 9 v_1 ^ 2 v_3 ^ 2 + 2 _ 7 v_1 v_2 v_3 ^ 2 + _ 12 v_2 ^ 2 v_3 ^ 2 + _ 6 ( v_1 ^ 2 + v_2 ^ 2 ) v_3 ^ 2 + _ 13 v_3 ^ 4 & + & f_5 v_3 ^ 4 + 2 _ 11 v_1 v_3 v_r^2 + _ 14 v_3 ^ 2 v_r^2 + _ 8 v_3 ^ 2 v_r^2 , where @xmath221 satisfy 0 & = & v_1[m^2 + f_1 v_r^2 + ( _ 1 + _ 2 ) v_1 ^ 2 + ( _ 1 + _ 3 ) v_2 ^ 2 + 3 _ 4 v_1 v_2 ] + v_2 ( ^2 + f_3 v_r^2 + _ 4 v_2 ^ 2 ) & & + v_3 [ ( _ 6 + _ 9 ) v_1 v_3 + _ 7 v_2 v_3 + _ 11 v_r^2 ] ; 0 & = & v_2[m^2 + f_2 v_r^2 + ( _ 1 + _ 2 ) v_2 ^ 2 + ( _ 1 + _ 3 ) v_1 ^ 2 + 3 _ 4 v_1 v_2 ] + v_1 ( ^2 + f_3 v_r^2 + _ 4 v_1 ^ 2 ) & & + v_3 [ ( _ 6 + _ 12 ) v_2 v_3 + _ 7 v_1 v_3 ] ; 0 & = & v_r ( m_r^2 + _ r v_r^2 + f_1 v_1 ^ 2 + f_2 v_2 ^ 2 + 2 f_3 v_1
v_2 ) + v_3 [ ( _ 8 + _ 14 ) v_r v_3 + 2 _ 11 v_1 v_r ] ; 0 & = & v_3[m_^2 + ( _ 6 + _ 9 ) v_1 ^ 2 + 2 _ 7 v_1 v_2 + ( _ 12 + _ 6 ) v_2 ^ 2 + 2 ( _ 13 + _ 5 ) v_3 ^ 2 & & + ( _ 14 + _ 8) v_r^2 ] + _ 11 v_1 v_r^2 . [ eq : min ] using this notation , the vacuum expectation values have the following solution with @xmath53 : v_2 & & , & & v_1 ^ 2 & = & , & & v_r^2 & = & , where @xmath222 was introduced in ( [ eq : v3 ] ) , and @xmath223 is determined by that same equation . in order for ( [ eq : min ] ) to be consistent with ( [ eq : v3 ] )
, the parameters in the potential must also satisfy m_^2 [ _ 1 ( _ 1 + _ 2 ) - _ 4 _ r ] = m^2 [ _ 2 ( _ 1 + _ 2 ) - _ 3 _ r ] + m_r^2 ( _ 1 _ 3 - _ 2 _ 4 ) .
there are many ways to obtain the desired hierarchy , @xmath224 for example , let @xmath225 and @xmath226 ; then @xmath227 , @xmath228 , and @xmath229 . in the limit @xmath56 , the ( unnormalized ) would - be goldstone fields associated with the @xmath230 and @xmath231 vector bosons are , respectively , g & = & i m ( _ 1 ^ 0 + _ 2 ^ 0 ) ; g & = & i m ; g^+_r & = & _ r^+ + ( u _ 1^+ - _ 2^+ ) ; g^+_l & = & _ 1^+ + _ 2^+ ; where =
; = , = , = 2 ; = u .
the physical scalars and their corresponding masses can be obtained from the potential in a straightforward manner : there is a single doubly - charged field , 3 singly charged fields and 6 ( real ) neutral fields . in obtaining the various expressions we have assumed ( [ eq : hi ] ) .
the results are presented in table [ tab : mass ] : they indicate that the field @xmath232 has a mass @xmath233 and plays the role of the sm higgs boson ; the other physical scalars have masses of order @xmath234 . d. wyler and l. wolfenstein , nucl .
* b218 * , 205 ( 1983 ) .
r. n. mohapatra and j. w. f. valle , phys . rev . *
d34 * , 1642 ( 1986 ) .
e. ma , phys .
* b191 * , 287 ( 1987 ) .
e. ma , phys . rev . *
d80 * , 013013 ( 2009 ) ; and references therein .
e. ma , mod .
a24 * , 2161 ( 2009 ) ; and references therein . m. pospelov , phys . rev . * d56 * , 259 ( 1997 ) ; y. zhang , h. an , x. ji , and r. n. mohapatra , nucl . phys . *
b802 * , 247 ( 2008 ) .
a. dedes and a. pilaftsis , phys
* d67 * , 015012 ( 2003 ) .
particle data group , c. amsler _ et al .
_ , phys . lett . *
b667 * , 1 ( 2008 ) .
s. bertolini , f. borzumati , a. masiero , and g. ridolfi , nucl . phys . *
b353 * , 591 ( 1991 ) . | the asymmetric higgs sector of one @xmath0 bidoublet @xmath1 and one @xmath2 doublet [ but no @xmath3 doublet ] is considered in a nonsupersymmetric left - right extension of the standard model ( sm ) of particle interactions . the inverse seesaw mechanism for neutrino mass
is naturally implemented with the addition of fermion singlets , allowing thereby the possibility of breaking @xmath2 at the tev scale .
flavor - changing neutral higgs couplings to quarks are studied in two scenarios , where the @xmath2 charged - current mixing matrix is given either by @xmath4 ( scenario i ) or @xmath5 ( scenario ii ) .
we consider the bounds on these scalar particle masses from @xmath6 and @xmath7 mixing , as well as @xmath8 .
we find that , whereas in scenario i , they are of order 10 tev , as in other left - right models , they may be well below 1 tev in scenario ii , thus allowing them to be within reach of detection at the forthcoming large hadron collider ( lhc ) .
ucrhep - t473 + dcp-09 - 03 + october 2009 * asymmetric higgs sector and + neutrino mass in an @xmath2 model + * |
the potential of electromagnetic probes ( real and virtual photons ) to characterize the strongly interacting quark gluon plasma is well known since it was first pointed out by shuryak @xcite .
they are unique in providing information about the plasma temperature , the in - medium properties of hadrons and the mechanism of chiral symmetry restoration @xcite .
the phenix detector was designed with special emphasis on the measurement of such probes @xcite and has produced the first rhic results on dileptons @xcite and direct photons @xcite .
capitalizing on the flexibility of the rhic facility , phenix is engaged in a systematic study of electromagnetic probes . in these proceedings ,
i present the most recent phenix results on these probes , including the nuclear modification factor r@xmath2 and the elliptic flow v@xmath3 of direct photons in au+au collisions , the measurement of direct photons in d+au collisions and the first results obtained with the hbd on the measurement of @xmath0 pairs in au+au collisions at @xmath1 = 200 gev .
the r@xmath2 of direct photons up to p@xmath4 = 20 gev / c was first reported by the phenix experiment @xcite .
the preliminary results showed an intriguing suppression trend at high p@xmath4 , above @xmath5 12 gev / c , albeit with very large statistical and systematic uncertainties , that triggered discussions about possible initial state and isospin effects @xcite . to qualitatively improve this result ,
the r@xmath2 was recalculated using the statistically improved 2006 p+p data @xcite together with a reanalysis of the au+au data that combines the information of both the pbgl and pbsc calorimeters .
the final results presented in fig .
[ fig : raa - photons ] show that r@xmath2 is consistent with 1 up to p@xmath4 = 20 gev / c and for all centralities expressed by the number of participant nucleons n@xmath6 @xcite thus resolving the long - standing issue of the behavior of the direct photons r@xmath2 at high p@xmath4 .
phenix has measured direct photons in p+p and au+au collisions at @xmath1 = 200 gev@xcite . the p+p yield is well reproduced by nlo pqcd calculations down to p@xmath4 = 1 gev / c , whereas the au+au data exhibit a strong excess of photons ( in the p@xmath4 range of 1 - 3 gev / c ) beyond the p+p yield scaled by the number of binary collisions n@xmath7 .
this excess has an exponential shape with a slope parameter t = 221 @xmath8 mev in central ( 0 - 20% ) collisions and has been interpreted as thermal radiation from the medium thus providing the first information about the temperature of the system averaged over the space - time evolution of the collision .
using hydrodynamical models one can infer an initial temperature of t@xmath9 = 300 to 600 mev depending on the assumed formation time ( @xmath10 = 0.6 - 0.15 fm / c ) of the system .
this is a seminal result . in order to further strengthen the interpretation in terms of thermal radiation , phenix measured the direct photons in d+au collisions to check whether the observed excess in au+au collisions could originate from cold nuclear matter effects in the initial state .
the direct photons in d+au were measured via three independent methods : virtual photons , @xmath11 tagging and the statistical subtraction @xcite .
all three methods yielded consistent results as shown in the left panel of fig .
[ fig : photons - dau ] .
the figure also demonstrates that the minimum bias ( mb ) d+au data are well reproduced by the fit to the p+p data scaled by n@xmath7 and that there is no excess of direct photons in the p@xmath4 range of 1 - 3 gev / c . the difference with the au+au system is better illustrated in the right panels of fig .
[ fig : photons - dau ] .
the right upper panel demonstrates that the nuclear modification factor in d+au collisions is consistent with unity whereas the right lower panel shows a comparison of the r@xmath2 in au+au and d+au collisions .
the large excess of direct photons observed in au+au at low p@xmath4 is not observed in d+au and one can then conclude that it is not due to initial state effects .
this result reinforces the interpretation of the au+au excess as thermal radiation from the sqgp .
phenix has measured the elliptic flow v@xmath3 of direct photons in au+au collisions at @xmath1 = 200 gev @xcite .
the magnitude and shape of v@xmath3 vs. p@xmath4 are very similar to those measured for pions .
a large v@xmath3 is observed at p@xmath4 @xmath12 4 gev / c where the yield is dominated by thermal photons , whereas at higher p@xmath4 , where the yield is dominated by prompt photons , v@xmath3 is consistent with 0 as illustrated in fig .
[ fig : v2-photons ] ( open circles ) .
this is a surprising result .
thermal radiation implies early emission whereas a large v@xmath3 implies late emission .
no wonder therefore that models have difficulties in simultaneously reproducing the yield and the large v@xmath3 values of direct photons .
the importance of this result prompted phenix to seek confirmation by performing a completely independent analysis with different systematic uncertainties based on external conversion photons .
the new analysis allowed to extend the p@xmath4 range of the measurements down to 0.5 gev / c and yielded consistent results with the previous one in the p@xmath4 range of overlap as illustrated in fig . [ fig : v2-photons ] .
this provides an important confirmation of the unexpected large v@xmath3 values of low p@xmath4 direct photons in au+au collisions .
phenix has performed the first measurements of the @xmath0 pair continuum in p+p @xcite and au+au collisions @xcite at @xmath1 = 200 gev .
the p+p results are well reproduced by the cocktail of known sources whereas the au+au data show a strong enhancement at low masses m = 0.15 - 0.75 gev / c@xmath13 with respect to a similar cocktail adjusted to au+au collisions .
the enhancement is concentrated at the most central 0 - 20% collisions and reaches a factor of @xmath14 for the 0 - 10% centrality bin and @xmath15 for mb collisions . in spite of numerous attempts ,
this result remains a challenge for theory .
in particular , all models that successfully reproduced the low mass pair enhancement observed by the ceres and na60 experiments at the sps energies fail to reproduce the phenix result @xcite raising the speculation of a new , yet unidentified , source of dileptons .
the phenix result is limited by large uncertainties , both statistical and systematic , due to the large combinatorial background and the resulting low signal to background ratio s / b and low cocktail to background ratio c / b .
the latter represents an objective figure of merit for the sensitivity of the measurement to the known hadronic sources in the absence of any enhancement . for the mb data , the s / b ratio is @xmath51/200 and the c / b is @xmath51/1000 . in order to qualitatively improve the measurement of low - mass dileptons ,
phenix embarked on an upgrade program .
a hadron blind detector ( hbd ) was developed with the purpose to reduce the combinatorial background .
the hbd is a windowless cherenkov detector operated with pure cf@xmath16 , in a proximity focus configuration .
the detector consists of a 50 cm long radiator directly coupled to a triple gem detector which has a csi photocathode evaporated on the top face of the upper - most gem foil and pad readout at the bottom of the gem stack .
the detector is located around the beam pipe in the field free region generated by operating the inner and outer coils of the phenix central arms in the @xmath17 configuration . in this mode
, the inner coil counteracts the main field of the outer coil creating an almost field - free region close to the vertex , extending out to @xmath550 - 60 cm in the radial direction @xcite .
the hadron blindness property of the hbd is achieved by operating the detector in the so - called reverse bias mode . in this mode
, the entrance mesh is set at a lower negative voltage with respect to the top gem and consequently the ionization electrons produced by a charged particle in the drift region between the mesh and the gem are mostly repelled towards the mesh @xcite . the choice of cf@xmath16 as both the radiator and the detector gas in a windowless configuration results in an unprecedented large bandwidth extending from @xmath5110 nm given by the cf@xmath16 cut - off , up to @xmath5210 nm given by the threshold of csi .
this large bandwidth translates into a large figure of merit n@xmath18 and a large number of photo - electrons per electron .
the main task of the hbd is to recognize and reject tracks originating from @xmath19 conversions or @xmath11 dalitz decays , where only the @xmath20 or the @xmath21 is detected in the central arms .
the strategy is to exploit the fact that the opening angle of electron pairs from these sources is very small compared to the pairs from light vector mesons . in the field - free region , this angle is preserved and the pairs produce two close or overlapping hits . therefore by applying an opening angle cut or a double amplitude cut one can reject a large fraction of the conversions and @xmath11 dalitz decays , while keeping most of the signal .
the detector was successfully operated in 2009 and 2010 and collected data in p+p and au+au collisions , respectively .
the data analysis led to a measured value of n@xmath22330 @xmath23 by far higher than in any other existing gas cherenkov counter .
this translates into 20 photo - electrons per single electron incident in the detector , allowing a good separation between single and double hits , a key requirement of the hbd to identify pairs from conversions and @xmath11 dalitz decays @xcite .
the preliminary results shown below exhibit an improvement of the s / b of @xmath55 with respect to the previous measurements without the hbd . for the p+p data analysis a procedure very similar to the one described in @xcite was used .
significant changes were made for the au+au data analysis .
in particular , two parallel and independent analysis streams were followed .
the two streams yielded results that are in agreement within their uncertainties , thus providing a crucial consistency check .
a full account of the two streams is far beyond the scope of these proceedings . only a few aspects and details of the analysis chain that yielded the best results in terms of signal efficiency are described here . in this first pass through the data we chose to apply strong run qa and strong fiducial cuts .
this has the benefit of homogenizing the response of the central arm detectors over time thereby reducing the number of corrections necessary to account for time variations of the various subsystems .
on the other hand this approach involves a large price in statistics and pair reconstruction efficiency .
cf@xmath16 has a strong scintillation line at 160 nm , right in the middle of the hbd sensitive bandwidth .
this results in a large occupancy of the detector for central collisions that requires an underlying event subtraction before attempting to identify the electrons hits in the hbd .
two different methods were used for this subtraction yielding consistent results .
the hbd single electron detection efficiency achieved was @xmath2490% for the 40 - 60% centrality bin gradually decreasing to 65% for the most central bin of 0 - 10% while rejecting 90% of the fake tracks in both cases .
the hbd provides strong electron identification capability in addition to the rich and emcal detectors of the central arms . instead of applying a series of independent one - dimensional cuts that ultimately result in a poor efficiency , we used a neural network for electron identification @xcite .
the neural network was trained on samples of hijing events covering all centralities . a siginificant improvement of the global single electron track efficiency ( of @xmath5 50% for the peripheral 60 - 92% events and
@xmath5 20% for the most central 0 - 10% events ) was obtained with the neural network as compared to the standard electron identification cuts .
a neural network was also used to inspect the vicinity of electron hits and decide whether the hbd signal corresponds to a single or a double electron hit .
the combinatorial background was subtracted in two separate steps .
first the background resulting from the random combination of @xmath20 and @xmath21 in the same event was subtracted using a mixed event technique .
the normalization of the unlike mixed event backround is obtained by normalizing the like - sign mixed event spectrum to the like - sign foreground spectrum in a phase space region which is free from correlations ( mass @xmath24 0.55 gev / c@xmath13 to avoid cross pairs @xcite and opening angle @xmath12 120@xmath25 to avoid jet correlations ) .
after this first step , the foreground unlike - sign spectrum contains both correlated background pairs and signal pairs and the foreground like - sign spectra contain only correlated background pairs . in the second step , the residual like - sign spectra , corrected for the acceptance difference between @xmath26 and @xmath17 pairs ,
are subtracted from the residual unlike - sign spectrum to obtain the raw signal yield @xcite .
this approach is relatively simple but results in a factor of @xmath27 larger statistical error as opposed to the case in which the various background sources are determined a priori and subtracted one after the other , without relying on the measured like - sign spectra .
the raw signal is then corrected for acceptance and reconstruction efficiency to yield the invariant dilepton mass spectrum within the ideal acceptance of the phenix detector .
conservative systematic errors were assigned to the various steps of the analysis and in particular to the combinatorial background subtraction .
the au+au data were analyzed in the same five centrality bins as in the previoulsy published phenix dilepton results @xcite . in these proceedings
we are presenting the results for peripheral ( 60 - 92% ) , semi - peripheral ( 40 - 60% ) and semi - central ( 20 - 40% ) events . results from the two most central bins ( 0 - 10% and 10 - 20% ) are not yet available . the results are compared to a cocktail containing all the known hadronic sources . the exodus event generator @xcite is used to simulate the photonic sources ( dalitz decays of light neutral mesons : @xmath28 , @xmath29 , @xmath29@xmath30 @xmath31 and @xmath32 ) and the non - photonic sources ( di - electron decays of the light vector mesons : @xmath33 , @xmath34 , @xmath35 ) .
the correlated pairs from semi - leptonic decays of heavy flavor ( charm and bottom ) mesons are generated using the mc@nlo package @xcite which calculates the initial hard scattering at the next to leading order .
after generating the various sources , the cocktail is filtered through the ideal acceptance of the phenix detector and smeared with the detector resolution .
@xmath28 is the dominant electron source and also the fundamental input for exodus . for each centrality bin
, we use @xmath11 p@xmath4 spectra based on phenix measurements of neutral and charged pions . for other mesons , the shape of the p@xmath4 distribution
is obtained assuming m@xmath4 scaling .
the absolute normalization is provided by the meson to @xmath11 ratio at high p@xmath4 @xcite .
the j/@xmath36 line shape for the p+p cocktail is obtained from a full monte carlo simulation where j/@xmath36 decays into @xmath0 are propagated through the full chain of reconstruction and analysis cuts using a geant based simulation of the phenix detector . for the j/@xmath36 in au+au ,
the shape and yield are taken from the p+p measurement , the latter being scaled by n@xmath7 and the measured nuclear modification factor r@xmath2 @xcite . therefore the comparison cocktail vs data for the j/@xmath36 in au+au is reduced to a consistency check between the previously published and the present j/@xmath36 yields , without and with hbd , respectively . figure [ fig : mass - spectrum - pp ] shows the invariant mass spectrum of @xmath0 pairs within the phenix detector acceptance measured in the 2009 p+p run with the hbd .
the results are compared to the expected yield from the cocktail of known sources .
the bottom panel shows the ratio of data to cocktail and demonstrates a good understanding of the measured spectrum .
the present result is fully consistent with the previously published one without the hbd @xcite and provides a crucial proof of principle for usage of the hbd .
this analysis is now being repeated using all the tools of the new analysis chain developed for the au+au data analysis in order to fully assess the benefit of these tools .
18 cm figure [ fig : mass - spectra - auau ] shows the invariant mass spectra of @xmath0 pairs within the phenix detector acceptance measured in au+au collisions in the 2010 run with the hbd , in three centrality bins : 60 - 92% ( left panel ) , 40 - 60% ( middle panel ) and 20 - 40% ( right panel ) .
the results are compared to the expected yield from the cocktail of known sources , briefly described in the previous section .
the bottom panels show the ratio of data to cocktail . in the most peripheral bin
the measured spectrum is in quite good agreement with the expected yield . for semi - peripheral and semi - central events
there seems to be a small enhancement of the measured yield with respect to the cocktail that increases with centrality , both in the low - mass region ( lmr , m = 0.15 - 0.75 gev / c@xmath13 ) and the intermediate mass region ( imr , m = 1.2 - 2.8 gev / c@xmath13 ) .
however , no clear statement can be made at this stage with the present level of very conservative errors .
pairs in au+au collisions at @xmath1 = 200 gevmeasured in the 2010 run with the hbd in three centrality bins .
the data points show statistical ( vertical bars ) and systematic ( shades ) uncertainties separately .
the shaded band represents the systematic uncertainty of the cocktail.,title="fig:",width=162 ] pairs in au+au collisions at @xmath1 = 200 gevmeasured in the 2010 run with the hbd in three centrality bins .
the data points show statistical ( vertical bars ) and systematic ( shades ) uncertainties separately .
the shaded band represents the systematic uncertainty of the cocktail.,title="fig:",width=162 ] pairs in au+au collisions at @xmath1 = 200 gevmeasured in the 2010 run with the hbd in three centrality bins .
the data points show statistical ( vertical bars ) and systematic ( shades ) uncertainties separately .
the shaded band represents the systematic uncertainty of the cocktail.,title="fig:",width=162 ] there are significant differences between the 2010 run with the hbd and the 2004 run without it , that prevent a direct comparison of the results . in the 2010 run
the central coils were operated in the @xmath17 configuration , as opposed to @xmath37 in 2004 , resulting in larger acceptance of low p@xmath4 tracks .
there is more radiative tail in the j/@xmath36 , @xmath38 and @xmath34 line shapes due to the hbd material in front of the central arms .
there is also a difference in the cocktails . in the 2010 run
, the open heavy flavor contribution is calculated using the mc@nlo package whereas in the 2004 data analysis we used pythia .
this results in a calculated yield in the imr which is @xmath516% larger with the mc@nlo than with pythia . to account for these differences , we compare in fig .
[ fig : run4-run10 ] the data / cocktail ratio for the integrated yield in the lmr and the imr as obtained in the 2004 and 2010 runs for the three centrality bins presented here . the 2004 pythia yield in the imr
is scaled up by 1.16 to account for the difference with the mc@nlo calculations .
the figure shows consistent results between the two runs within their uncertainties .
phenix is pursuing its systematic study of electromagnetic probes exploiting the flexibility of the rhic facility .
recent achievements presented here , include final results on r@xmath2 of direct photons @xcite , independent confirmation of the large v@xmath3 values of direct photons and first measurement of direct photons in d+au collisions . the preliminary dilepton analysis using the hbd , with strong qa cuts and conservative error estimates , yielded consistent results with those previously published @xcite .
the final analysis is expected to show a significant improvement in the statistical and systematic uncertainties and to include the two most central bins .
shuryak , , 150 ( 1978 ) . for recent theoretical and experimental reviews see the papers of c. gale ( p. 6
- 74 ) , r. rapp et al .
( p. 4 - 1 ) and
i. tserruya ( p. 4
- 43 ) in `` relativistic heavy ion physics '' , published in landolt - boernstein , vol . i-23 ( 2010 ) .
k. adcox et al .
( phenix collaboration ) , , 469 ( 2001 ) .
a. adare et al .
( phenix collaboration ) , , 034911 ( 2010 ) .
a. adare et al .
( phenix collaboration ) , , 132301 ( 2010 ) .
t. awes for the phenix collaboration , j. phys .
g*35 * , 104007 ( 2008 ) .
f. arleo , jhep * 0609 * , 015 ( 2006 ) .
a. adare et al .
( phenix collaboration ) , , 072008 ( 2012 ) .
s. afanasiev et al .
( phenix collaboration ) , , 152302 ( 2012 ) .
a. adare et al .
( phenix collaboration ) , arxiv:1208.1234 .
a. adare et al .
( phenix collaboration ) , , 122302 ( 2012 ) .
a. adare et al .
( phenix collaboration ) , , 313 ( 2009 ) .
a. kozlov et al , , 345 ( 2004 ) .
z. fraenkel et al . , ,
466 ( 2005 ) .
w. anderson et al . , , 35 ( 2011 ) .
the neural network was implemented using the class tmultilayerperceptron from the root package : http://root.cern.ch/root/html530/tmultilayerperceptron.html s. frixione and b.r .
webber , jhep 0206:029 , 2002 a. adare et al .
( phenix collaboration ) , , 232301 ( 2007 ) . | most recent phenix results on electromagnetic probes are presented including first preliminary results obtained with the hadron blind detector ( hbd ) on @xmath0 invariant mass spectra from au+au collisions at @xmath1 = 200 gev . |
our inspiration is to understand the canonical model of the moduli space @xmath9 of stable curves of genus @xmath10 .
this is known to be of general type for @xmath11 and @xmath12 @xcite . in these cases
, we can consider the canonical ring @xmath13 which is finitely generated by a fundamental conjecture of birational geometry , recently proven in @xcite .
then the corresponding projective variety @xmath14 is birational to @xmath9 and is called its _ canonical model_.
there has been significant recent progress in understanding canonical models of moduli spaces . for moduli spaces @xmath15 of principally polarized abelian varieties of dimension @xmath16 , the canonical model exists and is equal to the first voronoi compactification @xcite .
unfortunately , no analogous results are known for @xmath9 , even for @xmath17 . our approach is to approximate the canonical models with _ log canonical models_. consider @xmath18\cap { \mathbb q}$ ] so that @xmath19 is an effective @xmath20-divisor .
we have the graded ring @xmath21 and the resulting projective variety @xmath22 our previous paper @xcite describes @xmath0 explicitly for large values of @xmath23 . for simplicity
we assume that @xmath24 : small genera cases have been considered in @xcite .
for @xmath25 , @xmath0 is equal to @xmath9 .
the first critical value is @xmath26 : @xmath27 is the coarse moduli space of the moduli stack @xmath28 of pseudostable curves @xcite .
a pseudostable curve may have cusps but they are not allowed to have _ elliptic tails _ , i.e. , genus one subcurves meeting the rest of the curve in one point .
furthermore , there is a divisorial contraction @xmath29 induced by the morphism @xmath30 of moduli stacks which replaces an elliptic tail with a cusp .
furthermore , @xmath31 provided @xmath32 .
this paper addresses what happens when @xmath33 .
given a sufficiently small positive @xmath34 , we construct a small contraction and its flip : @xmath35 ^ -{\psi } & & \mg(\frac7{10}-\e ) \ar[dl]_-{\psi^+}\\ & \mg(\frac7{10 } ) & \\ } \ ] ] the resulting spaces arise naturally as geometric invariant theory ( git ) quotients and admit partial modular descriptions .
we construct @xmath4 as a git quotient of the chow variety of bicanonical curves ; it parametrizes equivalence classes of _ c - semistable curves_. we defer the formal definition , but these have nodes , cusps , and tacnodes as singularities .
the flip @xmath5 is a git quotient of the hilbert scheme of bicanonical curves ; it parametrizes equivalence classes of _ h - semistable curves _ , which are c - semistable curves not admitting certain subcurves composed of elliptic curves ( see definition [ d : h - stable ] ) .
we may express the flip in geometric terms ( figure [ fig : gf ] ) : let @xmath36 denote an _
elliptic bridge _ , where @xmath37 is smooth of genus @xmath38 , @xmath39 is smooth of genus one , and @xmath37 meets @xmath39 at two nodes @xmath40 and @xmath41 .
let @xmath42 be a tacnodal curve of genus @xmath10 , with normalization @xmath37 and conductor @xmath43 . in passing from @xmath44 to @xmath5 , we replace @xmath45 with @xmath42 .
note that the descent data for @xmath42 includes the choice of an isomorphism of tangent spaces @xmath46 the collection of such identifications is a principal homogeneous space for @xmath47 .
when @xmath45 is a generic elliptic bridge , the fiber @xmath48 ; see proposition [ p : const - tac ] for an explicit interpretation of the endpoints . here
we offer a brief summary of the contents of this paper ; a more detailed roadmap can be found in section [ ss : detailed ] .
section [ s : gitch ] is devoted to a general discussion of the git of chow points and hilbert points .
the main applications are the analysis of tautological classes and polarizations on the hilbert scheme , the resulting formulas for hilbert - mumford indices ( proposition [ prop : curveresult ] ) , and cycle maps ( corollary [ coro : hilbtochow ] ) .
we also recall various formulations of the hilbert - mumford one - parameter - subgroup criterion .
section [ s : basin ] is a brief review of the basin - of - attraction techniques used in this paper .
these are important for analyzing when distinct curves are identified in the git quotients . section [ s : cms ] discusses , in general terms , how to obtain contractions of the moduli space of stable curves from git quotients of hilbert schemes .
the resulting models of the moduli space depend on the choice of linearization ; we express the polarizations in terms of tautological classes . section [ s : propchstable ] summarizes basic properties of c - semistable curves : embedding theorems and descent results for tacnodal curves . section [ s : ubc ] offers a preliminary analysis of the git of the hilbert scheme and the chow variety of bicanonically embedded curves of genus @xmath49 . then in section [ s : automorphisms ] we enumerate the curves with positive - dimensional automorphism groups .
section [ s : fgitq ] applies this to give a git construction of the flip @xmath7 .
section [ s : hardwork ] offers a detailed orbit closure analysis , using basins of attractions and a careful analysis of the action of the automorphism group on tangent spaces .
the main application is a precise description of the semistable and stable bicanonical curves , proven in section [ s : psa ] . throughout
, we work over an algebraically closed field @xmath50 , generally of characteristic zero .
however , sections [ s : gitch ] and [ s : propchstable ] are valid in positive characteristic .
* acknowledgments : * the first author was partially supported by national science foundation grants 0196187 , 0134259 , and 0554491 , the sloan foundation , and the institute of mathematical sciences of the chinese university of hong kong .
the second author was partially supported by the korea institute for advanced study and the korea science and engineering foundation ( kosef ) grant funded by the korea government ( most ) ( no .
r01 - 2007 - 000 - 10948 - 0 ) .
we owe a great deal to s. keel , who helped shape our understanding of the birational geometry of @xmath9 through detailed correspondence .
we are also grateful to d. abramovich , y. kawamata , i. morrison , b.p .
purnaprajna , m. simpson , d. smyth , and d. swinarski for useful conversations .
in this paper , we will use four stability conditions : deligne - mumford stability @xcite , schubert pseudostability @xcite , c-(semi)stability , and h-(semi)stability .
we recall the definition of _ pseudostability _ , which is obtained from deligne - mumford stability by allowing ordinary cusps and prohibiting elliptic tails : [ d : p - stable]@xcite a complete curve is _ pseudostable _ if 1 .
it is connected , reduced , and has only nodes and ordinary cusps as singularities ; 2 .
admits no _ elliptic tails _ , i.e. , connected subcurves of arithmetic genus one meeting the rest of the curve in one node ; 3 . the canonical sheaf of the curve is ample .
the last condition means that each subcurve of genus zero meets the rest of the curve in at least three points . before formulating the notions of c- and h-(semi)stability ,
we need the following definition : [ d : eb ] an _ elliptic bridge _ is a connected subcurve of arithmetic genus one meeting the rest of the curve in two nodes . in our stability analysis
, we will require additional technical definitions : [ d : ech ] an _ open elliptic chain _ of length @xmath51 is a two - pointed projective curve @xmath52 such that * @xmath53 where each @xmath54 is connected of genus one , with nodes , cusps or tacnodes as singularities ; * @xmath54 intersects @xmath55 at a single tacnode @xmath56 , for @xmath57 ; * @xmath58 if @xmath59 ; * @xmath60 are smooth points with @xmath61 and @xmath62 ; * @xmath63 is ample .
an open elliptic chain of length @xmath51 has arithmetic genus @xmath64 .
[ d : echain ] let @xmath45 be a projective connected curve of arithmetic genus @xmath65 , with nodes , cusps , and tacnodes as singularities .
we say @xmath45 _ admits an open elliptic chain _ if there is an open elliptic chain @xmath66 and a morphism @xmath67 such that * @xmath68 is an isomorphism over @xmath69 onto its image ; * @xmath70 are nodes of @xmath45 ; we allow the case @xmath71 , in which case @xmath45 is said to be a _ closed elliptic chain_. @xmath45 admits a _ weak elliptic chain _ if there exists @xmath67 as above with the second condition replaced by * @xmath72 is a tacnode of @xmath45 and @xmath73 is a node of @xmath45 ; or * @xmath71 is a tacnode of @xmath45 , in which case @xmath45 is said to be a _
closed weak elliptic chain_. now we are in a position to formulate our main stability notions : [ d : c - stable ] a complete curve @xmath45 is said to be _ c - semistable _ if 1 .
@xmath45 has nodes , cusps and tacnodes as singularities ; 2 .
@xmath74 is ample ; 3 .
a connected genus one subcurve meets the rest of the curve in at least two points ( not counting multiplicity ) .
it is said to be _ c - stable _ if it is c - semistable and has no tacnodes or elliptic bridges .
[ d : h - stable ] a complete curve @xmath45 of genus @xmath10 is said to be _ h - semistable _ if it is c - semistable and admits no elliptic chains .
it is said to be _ h - stable _ if it is h - semistable and admits no weak elliptic chains .
[ r : whatiscstable ] a curve is c - stable if and only if it is pseudostable and has no elliptic bridges .
table [ tab : stability ] summarizes the defining characteristics of the stability notions .
.stability notions [ cols="^,^,^,^,^,^",options="header " , ] [ default ] we start with some preliminary results .
recall from @xcite that @xmath75 denotes the functorial contraction and @xmath76 the induced morphism on coarse moduli spaces , which contracts the divisor @xmath77 .
[ l : logterm ] for @xmath78 , @xmath79 and @xmath80 are log terminal and @xmath81 since @xmath82 , the locus in @xmath83 parametrizing curves with nontrivial automorphisms has codimension @xmath84 @xcite .
( of course , we have already collapsed @xmath85 . )
thus the coarse moduli map @xmath86 is unramified in codimension one and @xmath87 for each @xmath88 .
we have the log discrepancy equation @xcite @xmath89 and the pull back @xmath90 where @xmath91 designates proportionality .
since @xmath92 is smooth and @xmath93 is normal crossings , the pair @xmath94 is log terminal .
the discrepancy equation implies that @xmath79 is log terminal for @xmath95 .
applying the ramification formula @xcite to ( [ e : ramify ] ) ( or simply applying ( * ? ? ?
* a.13 ) ) , we find that @xmath80 is also log terminal .
since @xmath77 is @xmath96-exceptional , for each cartier divisor @xmath97 on @xmath83 and @xmath98 we have @xmath99 .
this implies that @xmath100 we shall construct the contractions by using the powerful results of @xcite : [ p : makepsi ] for @xmath101\cap \bq$ ] , there there exists a birational contraction @xmath102 it contracts the codimension - two strata @xmath103 where 1 .
@xmath104 ; 2 .
@xmath105 , @xmath106 , by collapsing the loci @xmath107 corresponding to varying @xmath108 .
we shall see in corollary [ c : excep - psi ] that @xmath8 is an isomorphism away from @xmath109 .
recall that @xmath110 is ample provided @xmath111 ; this is part of the assertion that @xmath112 for @xmath111 ( * ? ? ?
* theorem 1.2 ) .
however , @xmath113 is nef but not ample @xcite . indeed
, the pull - back to @xmath9 @xmath114 can be analyzed using the classification of one - dimensional boundary strata by faber @xcite and gibney - keel - morrison @xcite .
it is ` f - nef ' , in the sense that it intersects all these strata nonnegatively , and is therefore nef by @xcite .
later on , we will list the strata meeting it with degree zero .
we apply kawamata basepoint freeness @xcite : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ let @xmath115 be a proper kawamata log terminal pair with @xmath37 effective .
let @xmath116 be a nef cartier divisor such that @xmath117 is nef and big for some @xmath118 .
then @xmath119 has no basepoint for all @xmath120 .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ for our application , @xmath116 is a cartier multiple of @xmath121 and @xmath122 for small positive @xmath123 .
the resulting morphism is denoted @xmath8 .
we claim that @xmath8 is birational . to establish the birationality
, we show that each curve @xmath124 meeting the interior satisfies @xmath125 the moriwaki divisor @xmath126 meets each such curve nonnegatively ( * ? ? ?
* theorem b ) .
we can write @xmath127 each of these coefficients is positive : clearly @xmath128 and since @xmath129 , @xmath130 thus we have @xmath131 where the last inequality reflects the fact that the torelli morphism is nonconstant along @xmath132 .
we verify the image of @xmath8 equals @xmath4 .
the log discrepancy formula ( [ e : logdisc ] ) implies @xmath133 however , since @xmath77 is @xmath134-exceptional adding it does not change the space of global sections , whence @xmath135 finally , we offer a preliminary analysis of the locus contracted by @xmath8 .
the main ingredient is the enumeration of one - dimensional boundary strata in @xcite ( see also @xcite ) .
we list the ones orthogonal to @xmath136 ; any stratum swept out by these classes is necessarily contracted by @xmath8 . in the second and third cases @xmath137 denotes a varying @xmath138-pointed curve of genus zero parametrizing the stratum . 1 . families of elliptic tails , which sweep out @xmath85 and correspond to the extremal ray contracted by @xmath139 .
2 . attach a @xmath140-pointed curve of genus @xmath141 and a @xmath140-pointed curve @xmath142 of genus @xmath38 to @xmath137 and stabilize .
contracting this and the elliptic tail stratum collapses @xmath143 along the @xmath144 s corresponding to fixing @xmath142 and varying the other components .
attach a @xmath145-pointed curve @xmath146 of genus @xmath147 , a @xmath145-pointed curve @xmath148 of genus @xmath149 , and a @xmath140-pointed curve of genus @xmath141 to @xmath137 and stabilize . contracting this and
the elliptic tail stratum collapses @xmath150 along the @xmath144 s corresponding to fixing @xmath151 and varying the other components .
thus the codimension - two strata @xmath152 are all contracted by @xmath8 .
consider the chow variety of degree @xmath154 curves of genus @xmath10 in @xmath155 .
let @xmath156 denote the closure of the bicanonically embedded smooth curves of genus @xmath10 .
similarly , let @xmath157 denote the closure of these curves in the hilbert scheme .
[ p : cyclebicanonical ] the cycle class map @xmath158 induces a morphism of git quotients @xmath159 where the hilbert scheme has the asymptotic linearization introduced in [ subsect : cphs ] .
this is a special case of corollary [ coro : hilbtochow ] , which applies quite generally to cycle - class maps from hilbert schemes for chow varieties .
( see [ subsect : cphs ] for background information on the cycle class map . )
let @xmath160 and @xmath161 denote the resulting git quotients @xmath162 and @xmath163 , and @xmath164 the morphism of proposition [ p : cyclebicanonical ] .
[ t : main12 ] let @xmath165 be a small positive number .
there exist isomorphisms @xmath166 and @xmath167 such that the induced morphism @xmath168 is the flip of @xmath8 .
we thus obtain a modular / git interpretation of the flip : @xmath169 ^ -{\psi } & & \mg(\frac7{10}-\e ) \simeq \mg^{hs } \ar[dl]_-{\psi^+}\\ & \mg(\frac7{10})\simeq \mg^{cs } & \\ } \ ] ] for c - semistable curves , @xmath170 is very ample and has no higher cohomology ( proposition [ p : embedcstable ] ) .
the image in @xmath155 is said to be _ bicanonically embedded_. [ t : chow ] the semistable locus @xmath171 ( resp .
stable locus @xmath172 ) corresponds to bicanonically embedded c - semistable ( resp .
c - stable ) curves . unlike in @xmath9 and @xmath173 , nonisomorphic curves may be identified in the quotient @xmath174 .
for example , if a c - semistable curve @xmath175 consists of a genus @xmath38 curve @xmath37 meeting in two nodes @xmath176 with an elliptic curve @xmath39 , then it is identified with any tacnodal curve obtained by replacing @xmath39 with a tacnode . in section [ s : hardwork ] , we shall give a complete classification of strictly semistable curves and the curves in their orbit closure .
[ t : hilbert ] the semistable locus @xmath177 ( resp .
stable locus @xmath178 ) with respect to the asymptotic linearization corresponds to bicanonically embedded h - semistable ( resp .
h - stable ) curves .
one difference from the case of chow points is that tacnodal curves may well be hilbert stable .
for instance , when @xmath49 _ irreducible _ bicanonical h - semistable curves are necessarily h - stable .
when @xmath179 , a bicanonical h - semistable curve is hilbert strictly semistable if and only if it has a tacnode @xcite .
when @xmath180 , every h - semistable curve is h - stable and the moduli functor is thus separated .
since c - stable curves are h - stable ( see proposition [ prop : hilbtochow ] ) and pseudostable ( see remark [ r : whatiscstable ] ) , we have [ c : excep - psi ] @xmath8 and @xmath153 are isomorphisms over the locus of c - stable curves .
thus @xmath8 is a small contraction with exceptional locus @xmath181 and @xmath153 is a small contraction with exceptional locus @xmath182 , the h - semistable curves with tacnodes .
thus the geometry of the flip is as indicated in figure [ fig : gf ] : @xmath183 precisely when @xmath45 is the ` pseudostable reduction ' of @xmath42 .
the proof of theorems [ t : chow ] and [ t : hilbert ] is rather intricate , so we give a bird s eye view for the reader s convenience .
\(1 ) the following implications are straightforward : * from the definitions , it is clear that : @xmath184 * general results on linearizations of chow and hilbert schemes ( proposition [ prop : hilbtochow ] ) imply @xmath185 and @xmath186 \(2 ) we next prove that non c - semistable ( resp .
non h - semistable ) curves are chow unstable ( resp .
hilbert unstable ) .
the main tool is the stability algorithm ( proposition [ prop : stabcrit2 ] ) .
* non c - semistable curves can be easily destabilized by one - parameter subgroups ( [ s : ubc ] ) .
we obtain @xmath187 * we show that if a curve @xmath45 admits an _
open rosary _ of even length ( see definition [ d : rosary ] ) , then there is a 1-ps @xmath188 coming from the automorphism group of the rosary such that the _ @xmath189th hilbert point _
@xmath190_m$ ] ( definition [ d : mth - hilb - pt ] ) is unstable with respect to @xmath188 for all @xmath191 ( proposition [ p : hs - o - ros ] and proposition [ p : hs - cr-1br ] ) . *
if @xmath45 admits an elliptic chain , then it is contained in the _ basin of attraction _ @xmath192_m)$ ] ( see definition [ d : basin ] ) of a curve @xmath193 admitting an open rosary of even length such that @xmath194_m , \rho ) < 0 $ ] . hence such curves are hilbert unstable ( propositions [ p : hs - echain ] and [ p : basin - cr-1br ] ) and we obtain @xmath195 \(3 ) we prove c - semistable @xmath196 chow semistability " , and use it to establish
h - semistability @xmath196 hilbert semistability " . *
the only possible chow semistable replacement of a c - stable curve is itself ( see theorem [ t : partial ] ) .
thus c - stable curves are chow stable and hence hilbert stable .
* we show that any strictly c - semistable curve @xmath45 is contained in a basin of attraction of a distinguished c - semistable curve @xmath197 with one - parameter isomorphism such that @xmath198 ( see proposition [ p : degenerate ] ) .
indeed , we choose @xmath197 so that it has closed orbit in the locus of c - semistable points ( cf .
proposition [ p : c - minorbit ] ) . *
if @xmath45 is strictly c - semistable , its pseudo - stabilization @xmath37 has elliptic bridges . for any such @xmath37
, there is a distinguished strictly c - semistable curve @xmath197 such that its basins of attraction contain every c - semistable replacement for @xmath37 .
futhermore , _ every _ possible chow - semistable replacement for @xmath37 is contained in some basin of attraction @xmath199 with @xmath200 . since one of these
must be chow semistable , every one is chow semistable ( see lemma [ l : flat limit ] ) . *
the hilbert semistable curves form a subset of the set of chow semistable curves .
we first identify the chow semistable curves admitting one - parameter subgroups that are hilbert - destabilizing .
then we show that any curve that is hilbert unstable but chow semistable arises in the basin of attraction of such a curve .
these basins of attraction consist of the curves that are c - semistable but not h - semistable .
thus the h - semistable curves are hilbert semistable ( [ ss : hs ] ) .
let @xmath201 for some @xmath202-dimensional vector space @xmath203 . throughout this section ,
let @xmath204 be a one - parameter subgroup and @xmath205 be homogeneous coordinates that diagonalize the @xmath188-action so that @xmath206 we have the associated one - parameter subgroup @xmath207 @xmath208 given @xmath209 , the _ hilbert - mumford index _ is given by ( cf .
@xcite ) : @xmath210 we say that @xmath211 is ( semi)stable with respect to @xmath212 if @xmath213 .
a fundamental theorem of git is that @xmath211 is git ( semi)stable if and only if it is ( semi)stable with respect to every 1-ps of @xmath214 .
we will sometimes abuse terminology and say that @xmath211 is ( semi)stable with respect to @xmath188 when it is ( semi)stable with respect to @xmath212 .
we briefly recall mumford s interpretation of hilbert - mumford criterion for chow stability of projective varieties @xcite .
let @xmath215 be a projective variety and @xmath188 be a one - parameter subgroup of @xmath216 with weights @xmath217 .
let @xmath218 be the ideal sheaf of @xmath219 $ ] such that @xmath220 = \left(\begin{array}{c}\mbox{the $ \ss_x$-submodule of $ \ss_x(1)[t]$ generated by } \\
\mbox{$t^{r_i}x_i$ , \ , $ i = 0 , \dots , n$. } \end{array}\right)\ ] ] the _ hilbert - samuel multiplicity _
@xmath221 is the normalized leading coefficient of @xmath222 where @xmath223 is the invertible @xmath219$]-module @xmath224 $ ] .
then the hilbert - mumford criterion can be translated in terms of @xmath221 as follows : [ t : mumford ] ( * ? ? ? * theorem 2.9 ) the chow point of @xmath225 is stable ( resp .
semistable ) if and only if @xmath226 for any one - parameter subgroup @xmath204 with weights @xmath227 .
we shall make frequent use of the following lemma which describes how the hilbert - samuel multiplicity is affected by the singular points : [ l : balance ] ( * ? ? ?
* lemma 1.4 ) let @xmath225 be a reduced curve in @xmath228 and @xmath229 be its normalization . 1 .
@xmath230 , where @xmath231 denotes the normalized leading coefficient of @xmath232 .
2 . suppose that @xmath233 for all @xmath234 where @xmath235 is the natural valuation of @xmath236 .
then @xmath237 we shall also use the following lemma that addresses the case in which @xmath225 is degenerate and @xmath188 acts on @xmath225 trivially .
[ l : degree ] ( * ? ? ? * lemma 1.2 ) let @xmath225 be an @xmath51-dimensional variety in @xmath238 .
let @xmath188 be a 1-ps of @xmath239 such that @xmath240 , @xmath241 .
suppose that @xmath242 vanish on @xmath225 and @xmath243 .
then @xmath244 let @xmath245 be a projective variety with hilbert polynomial @xmath246 .
choose an integer @xmath189 sufficiently large so that * @xmath247 has no higher cohomology ; * the natural map @xmath248 is surjective .
[ d : mth - hilb - pt ] the _ @xmath189th hilbert point @xmath249_m$ ] of @xmath225 _ is defined @xmath250_m : = \left [ \sym^mv^ * \to \gm(\co_x(m ) ) \right ] \in \gr(p(m),\sym^mv ) \inj \bp(\bigwedge^{p(m ) } \sym^mv).\ ] ] note that @xmath225 is determined by @xmath249_m$ ] provided @xmath225 is cut out by forms of degree @xmath189 .
@xmath225 is said to be _
@xmath189-hilbert stable ( resp .
semistable ) _
if @xmath249_m$ ] is git stable ( semistable ) with respect to the natural @xmath214 action on @xmath251 .
we refer the reader to @xcite for detailed discussion of an algorithm ( and a macaulay 2 implementation ) using grbner basis to determine whether a variety is @xmath189-hilbert ( semi)stable with respect to a given one - parameter subgroup .
we sketch the main results here . for any given @xmath252
, @xmath253 denotes the monomial order defined by declaring @xmath254 if 1 .
@xmath255 ; 2 .
@xmath256 and @xmath257 ; 3 .
@xmath256 , @xmath258 and @xmath259 in the lexicographic order . in particular , given a one - parameter subgroup @xmath188 with the weight vector @xmath260
, the monomial order @xmath261 means the graded lexicographic order associated to the weight @xmath262 .
given a monomial @xmath263 , the @xmath188-weight is defined @xmath264 for each polynomial @xmath265 , let @xmath266 denote the largest term of @xmath265 with respect to @xmath261 .
for an ideal @xmath267 , we let @xmath268 .
let @xmath269 be a homogeneous ideal with graded pieces @xmath270 .
the monomials @xmath271 of degree @xmath189 not contained in @xmath272 form a basis for @xmath273 .
we reformulate gieseker s stability criterion for hilbert points @xcite in these terms : [ prop : stabcrit2 ] the hilbert - mumford index of @xmath249_m$ ] with respect to a one - parameter subgroup @xmath204 with weights @xmath274 is given by @xmath275_m , \rho ) = \frac{m p(m)}{n+1 } \sum r_i -\sum_{j=1}^{p(m ) } wt_{\rho}(x^{a(j)})\ ] ] where @xmath276 index the monomials of degree @xmath189 not contained in @xmath272 .
in particular , @xmath249_m\in \bp(\bigwedge^{p(m ) } \sym^mv)$ ] is stable ( resp .
semistable ) under the natural @xmath214-action if and only if for any one - parameter subgroup @xmath188 we have @xmath277 let @xmath278 be the connected component of the hilbert scheme containing @xmath225 , @xmath279 the universal family , @xmath280 the natural projection , and @xmath281 the polarization . a coherent sheaf @xmath282 on @xmath228
is said to be _
@xmath116-regular _ in the sense of castelnuovo and mumford ( * ? ?
14 ) if @xmath283 for each @xmath284 .
suppose that the ideal sheaf @xmath285 is @xmath116-regular .
it follows that for each @xmath286 * @xmath287 is surjective ; * @xmath288 for each @xmath284 ; and also * @xmath289 is surjective ; * @xmath290 for @xmath284 .
in particular , @xmath291 is @xmath292-regular .
conversely , if @xmath291 is @xmath292-regular and @xmath293 then @xmath285 is @xmath116-regular .
@xcite there exists an @xmath294 such that _ every _
@xmath249 \in hilb$ ] has @xmath116-regular ideal sheaf ( * ? ? ?
* ch.14 ) .
then for each @xmath295 we get a closed embedding ( * ? ? ?
15 ) @xmath296 & \mapsto & [ x]_m \end{array}\ ] ] the universal quotient bundle @xmath297 satisfies @xmath298 and on taking determinants we find @xmath299 recall the _
tautological divisor classes _ developed in @xcite and ( * ? ? ?
* theorem 4 ) : there exist cartier divisors @xmath300 on @xmath278 such that @xmath301 where @xmath51 is the dimension of subschemes parametrized by @xmath278 .
that is , the determinant of cohomology of @xmath302 can be expressed as a polynomial in the tautological class .
this is a relative version of the hilbert polynomial of @xmath303 over @xmath278 .
it follows that the polarizations introduced above satisfy : @xmath304 using these formulas , we extend our definition : for each @xmath305 , write @xmath306 in many situations the tautological divisors satisfy a dependence relation : [ prop : depend ] let @xmath278 denote a connected component of the hilbert scheme parametrizing subschemes in @xmath228 of dimension @xmath51 and @xmath300 the tautological divisors on @xmath278 . let @xmath307 denote an open subset corresponding to subschemes @xmath225 where the following hold : * @xmath308 has no higher cohomology ; * the restriction map @xmath309 is an isomorphism . over @xmath310
we have the relation @xmath311 .
in particular , if @xmath312 then equation [ eqn : expansion ] takes the form @xmath313 let @xmath314 be the universal family embedded in @xmath315 .
our first assumption implies @xmath316 is locally free and @xmath317 the second assumption implies we have a trivialization ( cf .
@xcite ) @xmath318 in particular , it follows that @xmath311 .
we have seen that @xmath278 admits an embedding into @xmath319 for @xmath320 . in practice , we are usually interested in subsets of @xmath278 , that exclude degenerate subschemes with very high castelnuovo - mumford regularity : [ prop : hphs ] let @xmath321 denote the open subset parametrizing @xmath249 \in hilb$ ] satisfying : * @xmath247 has no higher cohomology ; * @xmath322 is surjective .
let @xmath323 denote the universal family restricted to this subset .
then we have @xmath324 and there exists a morphism @xmath325 & \mapsto & [ x]_m \end{array}\ ] ] such that @xmath326 .
@xmath327 contains the open subset parametrizing subschemes @xmath225 with @xmath189-regular ideal sheaf @xmath328 and @xmath329-regular structure sheaf @xmath291 . in particular , @xmath330 for @xmath320 .
the first assertion is just an application of equation [ eqn : taut ] in the special situation when @xmath331 for @xmath332 .
the morphism @xmath333 is just the classifying map for the surjection of locally free sheaves @xmath334 again , if @xmath335 is the universal quotient bundle then @xmath336 , and taking determinants gives the equality of line bundles .
applying the functorial properties of the hilbert - mumford index @xcite we obtain : [ coro : compareindex ] retain the notation of proposition [ prop : hphs ] .
suppose that @xmath249 \in hilb^{\circ , m}$ ] and @xmath337 is a one - parameter subgroup as before such that @xmath338 \in hilb^{\circ , m}.\ ] ] then we have @xmath339_m,\rho)=\mu^{\lambda_m}([x],\rho ) .
\label{eqn : equalindices}\ ] ] we compare the geometric invariant theory of the hilbert points @xmath249_m , m\gg 0 $ ] with that of the chow point @xmath340 .
let @xmath341 denote the corresponding _ chow variety _
i.e. , the image of the hilbert scheme under the morphism @xcite @xmath342 & \mapsto & ch(x ) \end{array}.\ ] ] this is equivariant under the natural actions of @xmath214 . by ( * ? ? ?
* theorem 4 ) , we obtain the proportionality @xmath343 and thus @xmath344 in other words , the sequence @xmath345\}$ ] converges to the pull back of the chow polarization in the projectivized nron - severi group of the hilbert scheme .
let @xmath346 denote the locus of points stable and semistable under the @xmath214-action .
for each @xmath320 , let @xmath347 the locus of points stable and semistable under the @xmath214-action linearized by @xmath348 .
the ample cone of @xmath278 admits a finite decomposition into locally - closed cells , such that the semistable locus is constant for linearizations taken from a given cell ( * ? ? ? * theorem 0.2.3(i ) ) .
in particular , @xmath349 and @xmath350 are constant for @xmath320 ; these are loci of the points _ stable and semistable with respect to the asymptotic linearization_. while the linearization is not well - defined , its locus of semistable points is ! moreover ,
applying functoriality of stability ( * ? ? ?
* theorem 2.1 ) , we find [ prop : hilbtochow ] let @xmath351 be a variety .
if @xmath225 is chow stable then @xmath225 is @xmath189-hilbert stable for @xmath320 .
if @xmath225 is @xmath189-hilbert semistable for @xmath320 then @xmath225 is chow semistable .
[ coro : hilbtochow ] assume the cycle map induces an @xmath214-equivariant map @xmath352 which is the case for @xmath320 .
then we obtain a natural morphism of git quotients @xmath353 given a graded ideal @xmath354 , the _ filtered hilbert function _ is defined @xmath355 where the sum is taken over the monomials of degree @xmath189 not contained in @xmath356 . for a closed subscheme @xmath357 ,
we define @xmath358 [ prop : fhf ] the filtered hilbert function @xmath359 is a polynomial @xmath360 for @xmath286 , the castelnuovo - mumford regularity of @xmath291 .
this polynomial is called the _ filtered hilbert polynomial_. since @xmath286 we have an embedding @xmath361 and proposition [ prop : hphs ] implies @xmath339_m,\rho)=\mu^{\lambda_m}([x],\rho).\ ] ] equation [ eqn : expansion ] gives @xmath362 fixing the point and the one - parameter subgroup , @xmath363 is a homomorphism in the line bundle variable @xcite and we have @xmath364,\rho)=\sum_{i=0}^r \binom{m}{i } \mu^{l_i}([x],\rho).\ ] ] the result follows from equation [ eqn : stabcrit ] . in this section ,
we assume @xmath278 parametrizes schemes of pure dimension one . here
equation [ eqn : taut ] takes the form @xmath365 [ prop : curveresult ] let @xmath366 denote the open subset parametrizing @xmath249 \in hilb$ ] satisfying : * @xmath225 is connected of pure dimension one ; * @xmath309 is an isomorphism ; * @xmath291 is @xmath140-regular .
then for each @xmath367 we have @xmath339_m,\rho)=\mu^{\lambda_m}([x],\rho)= ( m-1)\left[(3-m)\mu^{\lambda_2}([x],\rho)+(m/2 - 1)\mu^{\lambda_3}([x],\rho)\right ] .
\label{eqn : curveindices}\ ] ] proposition [ prop : depend ] gives the relation @xmath311 over @xmath368 , and equation [ eqn : expansion2 ] gives @xmath369 under our regularity hypothesis , proposition [ prop : hphs ] applies for each @xmath191 and @xmath370,\rho ) = ( m-1 ) \left [ \mu^{l_1}([x],\rho ) + \frac m2 \mu^{l_2}([x],\rho)\right]\ ] ] is a polynomial for @xmath367 ( see proposition [ prop : fhf ] ) .
one can obtain ( [ eqn : curveindices ] ) by expressing @xmath371,\rho)$ ] and @xmath372,\rho)$ ] in terms of @xmath373,\rho)$ ] and @xmath374,\rho)$ ] and plugging them in ( [ e : hismu ] ) .
[ d : basin ] let @xmath225 be a variety with @xmath375 acting via @xmath376 with fixed points @xmath377 . for each @xmath378 , the _ basin of attraction _
is defined @xmath379 when @xmath225 is smooth and projective this can be interpreted via the biaynicki - birula decomposition ( * ? ? ?
* theorem 4.3 ) : consider the decomposition @xmath380 of the fixed points into connected components . then there is a unique locally closed @xmath188-invariant decomposition @xmath381 and morphisms @xmath382 such that * @xmath383 for each @xmath384 ; * @xmath385 is an affine bundle ; * for each @xmath386 , the tangent space @xmath387 is the subspace over which @xmath188 acts with nonnegative weights . for @xmath388
we have @xmath389 .
the importance of this decomposition for the analysis of semistable points is clear from the following proposition which is well known to experts . given a point on a projective variety @xmath390 , let @xmath391 denote an affine lift , i.e. , a point in the affine cone over @xmath225 lying over @xmath211 .
[ p : whybasin ] suppose that @xmath392 is a reductive linear algebraic group acting on a projective variety @xmath225 and @xmath97 is a @xmath392-linearized ample line bundle .
suppose @xmath393 be semistable points mapping to the same point in the git quotient @xmath394 .
then there exists a semistable point @xmath395 with the following properties : * the orbit @xmath396 is closed , or equivalently , the stabilizer @xmath397 is reductive ; * there exists @xmath398 , one - parameter subgroups @xmath399 of @xmath400 , and lifts @xmath401 and @xmath402 of @xmath403 and @xmath404 such that @xmath405 since @xmath403 and @xmath404 are identified in the git quotient , any homogeneous invariant vanishing on @xmath403 automatically vanishes on @xmath404 , and vice versa .
consider the orbit closures @xmath406 and @xmath407 in @xmath225 .
their orbit closures meet ( * ? ? ?
* proposition 7 , pp .
254 ) : @xmath408 and moreover there exist @xmath401 and @xmath402 lying over @xmath403 and @xmath404 in the affine cone over @xmath225 such that @xmath409 ( this is essentially the fact that invariants separate orbit closures in affine space , e.g. , ( * ? ? ?
* corollary 1.2 , pp .
pick @xmath410 generating a _ closed _ orbit of the intersection .
recall matsushima s criterion @xcite : suppose @xmath392 is a reductive algebraic group and @xmath411 a closed subgroup ; the homogeneous space @xmath412 is affine if and only if @xmath413 is reductive .
this gives the equivalence of the two conditions on @xmath414 .
we apply ( * ? ? ?
* theorem 1.4 ) to the closed @xmath392-invariant set @xmath415 : there exist one - parameter subgroups @xmath416 and @xmath417 such that @xmath418 clearly there exists @xmath398 such that @xmath419 setting @xmath420 , we obtain the desired result . also , as far as stability is concerned , the points in a basin of attraction are all equivalent if the attracting point is strictly semistable with respect to the 1-ps : [ l : flat limit ] let @xmath392 , @xmath225 , @xmath97 be as in proposition [ p : whybasin ] .
let @xmath421 and suppose there exists an @xmath395 and a one - parameter subgroup @xmath188 of @xmath392 such that @xmath422 . if @xmath423 then @xmath424 is semistable with respect to @xmath97 if and only if @xmath211 is semistable with respect to @xmath97 .
assume that @xmath225 is embedded in @xmath238 by sections of @xmath97 and @xmath425 be an affine lift of @xmath211 .
since @xmath423 , @xmath426 has a specialization , say @xmath427 , which corresponds to @xmath428 .
let @xmath429 be a @xmath392-invariant section of @xmath97 .
then @xmath430 and it follows that @xmath431 if and only if @xmath432 .
let @xmath424 be a point in @xmath225 and @xmath433 be a one - parameter subgroup fixing @xmath424 .
it follows directly from the definition of the hilbert - mumford index that if @xmath434 then every point in the basin of attraction @xmath435 is unstable .
this observation can be used to classify unstable points in certain situations ; our approach is similar to @xcite : [ p : destabilize ] let @xmath392 , @xmath225 , @xmath97 be as in proposition [ p : whybasin ] and @xmath116 a second @xmath392-linearized ample line bundle .
let @xmath436 be semistable with respect to @xmath97 but unstable with respect to @xmath437 for each rational @xmath438 .
then there exists a point @xmath395 having the following properties : 1 .
@xmath424 is strictly semistable with respect to @xmath97 ; 2 .
there exists a one - parameter subgroup @xmath439 such that @xmath440 3 .
@xmath441 .
that is , every strictly semistable point that becomes unstable after perturbing @xmath97 can be destabilized by a one - parameter subgroup acting via automorphisms of a point strictly semistable with respect to @xmath97 .
let @xmath188 be a one - parameter subgroup with @xmath442 , which exists by the hilbert - mumford criterion .
let @xmath443 denote the corresponding limit point in @xmath225 .
clearly , @xmath444 and @xmath441 .
for the convenience of the reader , we recall the standard semistable replacement theorem : [ t : ss - replacement ] retain the assumptions of lemma [ l : flat limit ] .
assume that @xmath392 is reductive so the git quotient scheme @xmath445 exists .
let @xmath132 be a smooth curve , @xmath446 a closed point , and @xmath447 be a regular morphism .
then there exists a covering @xmath448 branched only over @xmath141 and @xmath449 , @xmath450 , such that * there is a regular morphism @xmath451 ; * @xmath452 for all @xmath453 .
two c - semistable curves @xmath454 and @xmath455 are said to be _ c - equivalent _ , denoted @xmath456 , if there exists a curve @xmath457 ( which we may assume has reductive automorphism group ) and one - parameter subgroups @xmath399 of @xmath458 with @xmath459 such that the basins of attraction @xmath460 and @xmath461 contain chow - points of curves isomorphic to @xmath454 and @xmath455 respectively .
we define _ h - equivalence _ , denoted @xmath462 , in an analogous way .
lemma [ l : flat limit ] shows that these equivalence relations respect the semistable and unstable loci .
proposition [ p : whybasin ] shows that for git - semistable curves @xmath456 if and only if @xmath463 and @xmath464 yield the same point of @xmath161 ; the analogous statement holds for h - equivalence .
let @xmath465 denote the universal curve over the moduli stack of curves of genus @xmath10 . for each @xmath466
we have the vector bundle @xmath467 of rank @xmath468 write @xmath469 and use @xmath470 to designate @xmath471 .
consider the multiplication maps @xmath472 for each @xmath473 .
we have the chern - class identities @xmath474 where @xmath91 designates proportionality .
these divisor classes were introduced by viehweg @xcite and cornalba - harris @xcite and their significance is explained by the following fact : [ prop : descendhilb ] assume that @xmath475 when @xmath476 .
consider the hilbert scheme @xmath278 of degree @xmath477 curves of genus @xmath10 in @xmath478 .
let @xmath479 denote the closure of the @xmath480-pluricanonically embedded smooth curves of genus @xmath10 .
suppose that @xmath348 ( introduced in [ subsect : githp ] ) is well - defined and ample on @xmath481 .
consider the open subsets @xmath482 corresponding to @xmath480-canonically embedded deligne - mumford stable curves that are git stable and git semistable with respect to @xmath348 .
let @xmath483 denote their images in moduli .
then @xmath348 descends to a multiple of @xmath484 along @xmath485 .
this restricts to an ample divisor on the coarse moduli space @xmath486 .
@xcite ) we illustrate how @xmath348 descends to @xmath485 .
let @xmath487 denote the universal family .
the multiplication map ( [ eqn : multiplication ] ) on moduli is obtained by descent from the multiplication map over @xmath481 @xmath488 as in proposition [ prop : depend ] , we have a trivialization @xmath489 thus the divisor class @xmath490 is proportional to @xmath491 as for the ampleness , the coarse moduli space @xmath486 of @xmath492 can be identified with an open subset of the git quotient @xmath493 @xmath348 descends to a polarization of this quotient .
mumford ( * ? ? ? * theorem 5.10 ) showed that the grothendieck - riemann - roch formula gives @xmath494 we therefore find @xmath495asymptotically as @xmath496 , we obtain the proportionality @xmath497combining proposition [ prop : descendhilb ] and equation [ eqn : hilbtochow ] , we obtain : [ prop : descendchow ] assume that @xmath475 when @xmath476 .
consider the chow variety @xmath498 of degree @xmath477 curves of genus @xmath10 in @xmath478 .
let @xmath499 denote the closure of the @xmath480-pluricanonically embedded curves of genus @xmath10 .
consider the open subsets @xmath500 corresponding to @xmath480-canonically embedded deligne - mumford stable curves that are chow stable and chow semistable respectively .
let @xmath501 denote their images in moduli .
then the polarization descends to a multiple of @xmath502 or @xmath503 the restriction to the coarse moduli space @xmath504 is ample .
mumford has proven that @xmath505 for each @xmath506 ( * ? ? ?
* theorem 5.1 ) .
proposition [ prop : hilbtochow ] then guarantees that @xmath507 for all @xmath320 .
proposition [ prop : descendchow ] then implies that @xmath508 is ample for @xmath509 ( * ? ? ?
* corollary 5.18 ) .
cornalba and harris @xcite established the sharp result : @xmath508 is ample if and only if @xmath510 .
we are primarily interested in situations where not all deligne - mumford stable curves have stable hilbert / chow points . here
git yields alternate birational models of the moduli space .
[ thm : getcontraction ] retain the notation of propositions [ prop : descendhilb ] and [ prop : descendchow ] with the convention that @xmath511 in the chow case .
suppose that * the complement to the deligne - mumford stable curves in the git - semistable locus @xmath512 ( resp .
@xmath513 ) has codimension @xmath84 ; * there exist deligne - mumford stable curves in the git - stable locus @xmath514 ( resp @xmath515 )
. then there exists a birational contraction @xmath516 regular along the deligne - mumford stable curves with git - semistable hilbert ( resp .
chow ) points . if @xmath517 is the polarization on the git quotient induced by @xmath348 then the moving divisor @xmath518 where @xmath519 is the subgroup generated by @xmath520-exceptional divisors .
a rational map of proper normal varieties is said to be a _
birational contraction _ if it is birational and its inverse has no exceptional divisors .
note that propositions [ prop : descendhilb ] and [ prop : descendchow ] cover the case where @xmath520 is an isomorphism .
our assumptions can be written in the notation of propositions [ prop : descendhilb ] and [ prop : descendchow ] : * @xmath521 ( resp .
@xmath522 ) has codimension @xmath84 ; * @xmath523 .
the git quotient morphism @xmath524 identifies the stack - theoretic quotient @xmath525 $ ] with @xmath526 .
this gives a birational map @xmath527 we define @xmath520 as its inverse .
we establish that @xmath520 is regular along @xmath528 : we have an @xmath529-equivariant morphism @xmath530 which descends to @xmath531 recall the universal property of the coarse moduli space : any morphism from a stack to a scheme factors through its coarse moduli space . in our context , this gives @xmath532 furthermore , the total transform of @xmath533 is contained in the complement @xmath534 ( @xmath535 ) , which has codimension @xmath84 .
thus any divisorial components of @xmath533 are @xmath520-exceptional divisors .
similary , @xmath536 has no exceptional divisors : these would give rise to divisors in the complement to @xmath537 in the semistable locus .
we now analyze @xmath538 in the rational picard group of @xmath9 .
( since @xmath9 has quotient singularities , its weil divisors are all @xmath539-cartier . )
if @xmath540 is very ample on the git quotient then @xmath541 induces @xmath520 , i.e. , @xmath541 has no fixed components and is generated by global sections over @xmath542 .
now @xmath538 is proportional to @xmath484 over @xmath542 and formula ( [ eqn : piccongruence ] ) follows .
[ p : embedcstable ] if @xmath65 and @xmath45 is a c - semistable curve of genus @xmath10 over @xmath50 , then @xmath543 and @xmath544 is very ample for @xmath545 . for the rest of this paper ,
when we refer to the chow or hilbert point of a c - semistable curve @xmath45 it is with respect to its bicanonical embedding in @xmath546 .
our argument follows ( * ? ? ?
* theorem 1.2 ) . by serre duality ,
@xmath547 vanishes if @xmath548 vanishes .
the restriction of @xmath549 to each irreducible component @xmath550 has negative degree because @xmath551 is ample .
it follows that @xmath552 , hence @xmath553 . to show that @xmath544 is very ample for @xmath475
, it suffices to prove for all @xmath554 that @xmath555 let @xmath556 denote the partial normalization of any singularities at @xmath211 and @xmath557 .
when @xmath211 is singular , a local computation gives @xmath558 if @xmath211 is a cusp and @xmath559 its preimage then @xmath560 if @xmath211 is a node or tacnode and @xmath561 the preimage points then @xmath562 thus in each case we can express @xmath563 for a suitable invertible sheaf @xmath564 on @xmath42 .
moreover , we have an inclusion @xmath565 with cokernel @xmath566 supported in @xmath567 of length @xmath568 . for instance , if both @xmath211 and @xmath557 are smooth then @xmath569 if both @xmath211 and @xmath557 are singular and @xmath570 then @xmath571 suppose that for each irreducible component @xmath572 , the degree @xmath573
. then @xmath574 and the desired vanishing follows .
we therefore classify situations where @xmath575 which divide into the following cases : 1 .
@xmath576 , @xmath577 , @xmath578 ; 2 .
@xmath576 , @xmath579 , @xmath580 ; 3 .
@xmath581 , @xmath577 , @xmath580 .
we write @xmath582 .
we enumerate the various possibilities .
we use the assumption that @xmath45 is c - semistable and thus has no elliptic tails . in cases ( a ) and ( b )
, @xmath37 is necessarily isomorphic to @xmath583 and meets the rest of @xmath45 in either three nodes or in one node and one tacnode .
after reordering @xmath211 and @xmath557 , we have the following subcases : 1 .
@xmath584 a node or tacnode of @xmath45 ; 2 .
@xmath585 a node or tacnode of @xmath225 and @xmath586 a smooth point of @xmath45 ; 3 .
@xmath587 a smooth point of @xmath45 and @xmath588 .
@xmath589 smooth points of @xmath45 . in case ( c ) ,
@xmath37 may have arithmetic genus zero or one : 1 . @xmath590 with @xmath589 smooth points of @xmath45 ; 2 .
@xmath37 of arithmetic genus one with @xmath589 smooth points of @xmath45 ; 3 .
@xmath37 of arithmetic genus one , @xmath591 a node or cusp of @xmath37 , and @xmath592 . in subcase ( c1 )
, @xmath37 meets the rest of @xmath45 in either four nodes , or in two nodes and one tacnode , or in two tacnodes . in subcases ( c2 ) and ( c3 ) , @xmath37 meets the rest of @xmath45 in two nodes . except in case ( c3 )
, @xmath593 is an isomorphism . for subcases ( b1 ) , ( c1 ) , and ( c2 )
, @xmath594 is an isomorphism .
moreover , @xmath566 is supported along @xmath37 so @xmath564 has negative degree along any other irreducible components of @xmath45 .
there are other components because the genus of @xmath45 is at least three .
thus elements of @xmath595 restrict to elements of @xmath596 that vanish at the points where @xmath37 meets the other components , i.e. , in at least two points . since @xmath597 or @xmath145
, we conclude @xmath598 . for subcase ( c3 ) , @xmath594 is not an isomorphism but @xmath566 is still supported along @xmath599 . as before , @xmath564 has negative degree along other irreducible components of @xmath42 , and elements of @xmath600 restrict to elements of @xmath601 vanishing where @xmath599 meets the other components .
there are at least two such points but @xmath602 , so we conclude that @xmath603 . in case ( a ) , we have @xmath604 . subcases ( a1 ) and
( a2 ) are similar to ( b1 ) and ( c1 ) : @xmath566 is supported along @xmath599 so elements in @xmath600 restrict to elements of @xmath601 vanishing at the points where @xmath599 meets the other components .
there is at least one such point , e.g. , the singularity not lying over @xmath211 , hence @xmath603 .
subcase ( a3 ) is more delicate . if @xmath599 is the unique component such that @xmath605 then the arguments of the previous cases still apply
however , the support of @xmath566 might not be confined to a single component .
we suppose there are two components @xmath606 and @xmath607 as described in ( a3 ) , such that @xmath608 .
since the genus of @xmath45 is @xmath609 , @xmath45 can not just be the union of @xmath606 and @xmath607 ; there is at least one additional component meeting each @xmath610 at some point @xmath611 , and the restriction of @xmath564 to this component has negative degree .
thus elements of @xmath600 restrict to elements of @xmath612 vanishing at @xmath611 , which are necessarily zero .
[ c : versal ] let @xmath613 be a c - semistable bicanonical curve .
* @xmath614 is 2-regular . *
the hilbert scheme is smooth at @xmath190 $ ] .
* let @xmath615 denote the singularities of @xmath45 and @xmath616 their versal deformation spaces .
then there exists a neighborhood @xmath617 of @xmath190 $ ] in the hilbert scheme such that @xmath618 is smooth .
proposition [ p : embedcstable ] yields @xmath619 which gives the regularity assertion
. this vanishing also implies ( * ? ? ?
* i.6.10.1 ) @xmath620 since the singularities of @xmath45 are local complete intersections we have @xmath621 thus the hilbert scheme is unobstructed at @xmath190 $ ] ( see ( * ? ? ? *
i.2.14.2 ) ) .
the assertion about the map onto the versal deformation spaces is ( * ? ? ? * i.6.10.4 ) .
[ c : all - m ] let @xmath622 be a bicanonical c - semistable curve and @xmath623 denote the curve to which @xmath624 specializes .
if @xmath623 is a bicanonical c - semistable curve then @xmath625_m,\rho)= ( m-1)\left[(3-m)\mu([c]_2,\rho)+(m/2 - 1)\mu([c]_3,\rho)\right].\ ] ] thus @xmath190_m$ ] is stable ( resp .
strictly semistable , resp .
unstable ) with respect to @xmath188 for each @xmath191 if and only if @xmath626_3,\rho ) \ge 2 \mu([c]_2,\rho ) > 0 $ ] ( resp .
@xmath626_3,\rho ) = \mu([c]_2,\rho ) = 0 $ ] , resp .
@xmath626_3,\rho ) \le 2 \mu([c]_2,\rho ) < 0 $ ] . )
the chow point @xmath627 is stable ( resp .
strictly semistable , resp .
unstable ) with respect to @xmath188 if and only if @xmath626_3,\rho)-2\mu([c]_2,\rho ) > 0 $ ] ( resp .
@xmath628 , resp . @xmath629 . ) by proposition [ p : embedcstable ] , a bicanonical c - semistable is 2-regular and the assertion on the hilbert points follows immediately from equation ( [ eqn : curveindices ] ) .
equations ( [ eqn : chowpolarization ] ) and ( [ eqn : expand ] ) allow us to interpret the hilbert - mumford index of the chow point in terms of the leading coefficient of @xmath626_m,\rho)$ ] as a polynomial in @xmath189 .
let @xmath45 be a curve with a tacnode @xmath51 , i.e. , a singularity with two smooth branches intersecting with simple tangency .
let @xmath630 be the partial normalization of @xmath45 at @xmath51 and @xmath631 the conductor .
the descent data from @xmath142 to @xmath632 consists of a choice of isomorphism @xmath633 identifying the tangent spaces to the branches .
functions on @xmath45 pull back to functions @xmath265 on @xmath37 satisfying @xmath634 and @xmath635 .
varying the descent data gives a one - parameter family of tacnodal curves : [ p : const - tac ] let @xmath37 be a reduced curve and @xmath636 distinct smooth points with local parameters @xmath637 and @xmath638 . each invertible linear transformation @xmath639 can be expressed @xmath640 for some @xmath641 ; let @xmath642 denote the corresponding identification .
then there exists a family @xmath643 , a section @xmath644 , and a morphism @xmath645 ^ -{\nu}\ar[dr ] & & \cc\ar[dl ] \ , \\ & \bg_m & & } \ ] ] such that 1 .
@xmath646 restricts to an isomorphism @xmath647 2 . for each @xmath648
, @xmath649 is a tacnode and @xmath650 its partial normalization ; 3 .
the descent data from @xmath142 to @xmath651 is given by @xmath652 . every tacnodal curve
normalized by @xmath142 occurs as a fiber of @xmath653 if @xmath37 is projective of genus @xmath38 then each @xmath654 has genus @xmath10 .
we sketch the construction of @xmath655 : @xmath652 tautologically yields an identification over @xmath375 @xmath656 which is the descent data from @xmath657 to @xmath655 .
fiber - by - fiber , we get the universal family of tacnodal curves normalized by @xmath142
. we will extend @xmath643 to a family of tacnodal curves @xmath658 .
first , observe that the graph construction gives an open embedding @xmath659 where @xmath660 corresponds to @xmath661 $ ] and @xmath662 corresponds to @xmath663 $ ] .
however , the identification ( [ eqn : iotamap ] ) fails to extend over all of @xmath583 ; indeed , it is not even defined at @xmath664 $ ] and its inverse is not defined at @xmath665 $ ] .
we therefore blow up @xmath666,q\times [ 1,0]}d \times \bp^1\ ] ] and consider the sections @xmath667 extending @xmath668 and @xmath669 . now ( [ eqn : iotamap ] ) extends to an identification @xmath670 [ p : limit - tac ] retain the notation of proposition [ p : const - tac ] .
there exists an extension @xmath671 where @xmath672 denotes the family of curves obtained from @xmath673 and @xmath674 by descent , @xmath675 the tacnodal section , and @xmath676 the resulting morphism . the new fiber @xmath677 ( resp .
@xmath678 ) is normalized by @xmath679 ( resp .
@xmath680 . )
we say that the tacnodes in the family @xmath681 are _ compatible _ , and that two curves are _ compatible _ if one can be obtained from the other by replacing some tacnodes by compatible tacnodes .
in this section , we show that if a curve is not c - semistable then it has unstable chow point : [ p : chow - imply - c ] if @xmath682 is chow semistable then @xmath613 is c - semistable .
we prove this by finding one - parameter subgroups destabilizing curves that are not c - semistable .
many statements in this section are fairly direct generalizations of results in @xcite and @xcite . a chow semistable bicanonical curve @xmath45 can not have a triple point , since @xmath683 and this implies that @xmath45 is chow unstable by proposition 3.1 of @xcite .
we need to show that among the double points , only nodes , ordinary cusps and tacnodes are allowed .
if @xmath45 has a non - ordinary cusp , then it is chow unstable .
suppose that @xmath45 has a non - ordinary cusp at @xmath40 .
let @xmath684 be the normalization , @xmath685 and assume @xmath686 $ ] .
recall that the singularity at @xmath40 is determined by the vanishing sequence @xmath687 which is the strictly increasing sequence determined by the condition @xmath688 @xmath45 has a cusp at @xmath40 if and only if the vanishing sequence @xmath689 is of the form @xmath690 , and it has an ordinary cusp if it is of the form @xmath691 .
hence if @xmath45 has a non - ordinary cusp at @xmath40 , then we can choose coordinates @xmath692
such that @xmath693 , @xmath694 , @xmath695 , and @xmath696 let @xmath697 be the one - parameter subgroup such that @xmath698 , where the weights are : @xmath699 then @xmath700 for all @xmath234 , and it follows from lemma [ l : balance ] that @xmath701 while @xmath702 the assertion now follows from theorem [ t : mumford ] .
suppose @xmath45 has a singularity at @xmath40 such that @xmath703/(y^2 - x^{2s } ) , \ , s \geq 3.\ ] ] then @xmath45 is unstable .
let @xmath704 be the normalization , @xmath705 since the two branches of @xmath45 agree to order @xmath429 at @xmath40 , we may choose coordinates @xmath692 such that @xmath706 let @xmath188 be the one - parameter subgroup of @xmath239 with weights @xmath707 .
then we have @xmath708 and by lemma [ l : balance ] , @xmath709 which is strictly greater than @xmath710 if @xmath45 has a multiple component , @xmath45 is chow unstable .
let @xmath454 be a component of @xmath45 with multiplicity @xmath711 .
choose a smooth non - flex point @xmath712 such that @xmath40 does not lie in any other component . since @xmath40 is smooth on @xmath713 , we may choose coordinates @xmath692 such that @xmath714 let @xmath188 be the one - parameter subgroup of @xmath239 with weights @xmath715 .
then we have @xmath716 this yields the inequality @xmath717 whereas @xmath718 we first prove that the semistable limit of a one parameter family of smooth bicanonical curves is bicanonical : [ p : pol - ss - limit ] let @xmath719 $ ] be a family of chow semistable curves of genus @xmath10 such that the generic fibre @xmath720 is smooth . if @xmath721}$ ] is an embedding such that @xmath722}^{\ten 2}$ ] then @xmath723}^{\ten 2}$ ] .
by @xcite , nonsingular bicanonical curves are chow stable .
hence any chow semistable curve is a limit of nonsingular bicanonical curves and proposition [ p : pol - ss - limit ] implies that if @xmath45 is not bicanonical , then @xmath724 . in particular
, a chow semistable curve does not have a smooth rational component meeting the rest of the curve in @xmath725 points .
mumford proved the statement for the @xmath480-canonical curves for @xmath726 , and his argument can be easily modified to suit our purpose .
it is an easy consequence of ( ii ) of the following proposition , which , in mumford s words , says that the degrees of the components of @xmath45 are roughly in proportion to their _ natural _ degrees .
let @xmath727 be a connected curve of genus @xmath10 and degree @xmath154 .
then * @xmath45 is embedded by a non - special complete linear system .
* let @xmath728 be a decomposition of @xmath45 into two sets of components such that @xmath729 and @xmath730 ( counted with multiplicity ) .
then @xmath731 mumford s argument goes through in the bicanonical case except for the proof of @xmath732 .
if @xmath733 , then by clifford s theorem we have @xmath734 and the chow semistability of @xmath45 forces @xmath735 combining the two , we obtain @xmath736 if @xmath737 , then @xmath738 , hence @xmath454 is rational and @xmath732 . if @xmath739 , then @xmath740 which is absurd since @xmath741 and @xmath742 .
we need to justify our use of clifford s theorem here , as chow semistable bicanonical curves have cusps and tacnodes .
we shall sketch the proof of gieseker and morrison @xcite and highlight the places where modifications are required to accommodate the worse singularities ( @xcite assumes that @xmath45 has only nodes ) .
let @xmath743 be a reduced curve with nodes , cusps and tacnodes .
let @xmath97 be a line bundle generated by sections .
if @xmath744 , then there is a subcurve @xmath745 such that @xmath746 suppose that @xmath747 and @xmath748 .
let @xmath454 be the union of components where @xmath749 does not vanish entirely and @xmath750 be the intersection points of @xmath454 and @xmath751 .
assume that @xmath752 s are ordered so that @xmath753 are tacnodes .
then we have @xmath754 we claim that @xmath749 restricts to give a homomorphism from @xmath755 to @xmath756 : let @xmath752 be a tacnode and let @xmath757 be the irreducible component containing @xmath752 . since @xmath749 vanishes entirely on @xmath37 , @xmath749 must vanish to order @xmath84 at @xmath752 on @xmath454 .
likewise , @xmath749 must vanish at each node .
it follows that @xmath758 factors through @xmath759 .
let @xmath760 be a basis of @xmath761 such that @xmath762 , and let @xmath763 be a basis for @xmath764 such that @xmath765 does not vanish at the support of @xmath766 and at any singular points .
it is shown in @xcite that @xmath767 , & [ s_1 , t_2 ] , & [ s_1 , t_3 ] , & \dots , & [ s_1 , t_p ] , \\ & [ s_2 , t_1 ] , & [ s_3 , t_1 ] , & \dots , & [ s_r , t_1]\\ \end{array}\ ] ] are linearly independent sections of @xmath768 , which implies that @xmath769 .
combining it with the riemann - roch gives the desired inequality .
let @xmath45 be a deligne - mumford stable curve with an elliptic tail @xmath770 .
then @xmath771 fails to be very ample along @xmath39 and thus @xmath45 does not admit a bicanonical embedding . in particular , @xmath45 does not arise in git quotients of the chow variety / hilbert scheme of bicanonical curves . here , we focus on curves with an elliptic subcurve meeting the rest of the curve in a _
tacnode_. [ p : g1 - 1tac ] let @xmath772 be a bicanonical curve consisting of a rational curve @xmath39 with one cusp , a rational curve @xmath773 and a genus @xmath38 curve @xmath37 such that @xmath40 is a tacnode and @xmath41 is a node .
then @xmath45 is chow unstable with respect to a one - parameter subgroup coming from its automorphism group . restricting @xmath774 we get @xmath775 since @xmath776 , we can choose coordinates so that @xmath777 and @xmath778 .
@xmath39 and @xmath773 can be parametrized by @xmath779 \mapsto [ s^4 , s^2t^2 , st^3 , t^4 , 0 , \dots , 0]\ ] ] and @xmath780 \mapsto [ 0 , 0 , uv , u^2 , v^2 , 0 , \dots , 0].\ ] ] the cusp is at @xmath781 $ ] , @xmath782 $ ] , and @xmath783 $ ] .
let @xmath188 be the 1-ps with weight @xmath784 .
we have @xmath785 on @xmath39 ( and @xmath773 ) , we have @xmath786 for all @xmath234 where @xmath787 is the valuation of @xmath788 ( and @xmath789 respectively ) and @xmath790 are weights of @xmath188 . by lemma [ l : balance ] , @xmath791 and @xmath792 . on @xmath773 , @xmath793 and @xmath794 .
since @xmath188 acts trivially on @xmath37 with weight @xmath140 , we use lemma [ l : degree ] and obtain @xmath795 combining them all , we obtain @xmath796 let @xmath797 be a bicanonical curve consisting of a genus one curve @xmath798 and a genus @xmath38 curve @xmath37 meeting in one tacnode @xmath40 . then @xmath42 is chow unstable . in view of proposition [ p : g1 - 1tac ]
, it suffices to show that @xmath42 is in the basin of attraction of @xmath799 with respect to @xmath188 .
consider the induced action on the local versal deformation space of the cusp @xmath800 $ ] which is given by @xmath801 where @xmath802 and @xmath803 .
the @xmath375 action is given by @xmath804 and the basin of attraction contains arbitrary smoothing of the cusp . on the other hand ,
the local versal deformation space of the tacnode @xmath805 $ ] is given by @xmath806 where @xmath807 so that @xmath375 acts on @xmath808 with weight @xmath809 and the basin of attraction does not contain any smoothings of the tacnode .
at the node @xmath783 $ ] , the local versal deformation space is @xmath810 where @xmath211 may be taken to be @xmath811 and @xmath375 acts with weight @xmath812 on the branch of @xmath773 and trivially on @xmath37 .
thus the induced action on the deformation space has weight @xmath812 , and the basin of attraction contains arbitrary smoothing of the node .
let @xmath45 be a bicanonical curve . by proposition [ prop : hilbtochow ] , @xmath45 is chow semistable if it is hilbert semistable .
note that by definition , if @xmath45 does not admit an elliptic chain , then @xmath45 is c - semistable if and only if it is h - semistable . combining this with proposition [ p : chow - imply - c ] ,
we obtain : [ p : hilb - imply - h ] if a bicanonical curve is hilbert semistable and does not admit an elliptic chain , then it is h - semistable .
we shall have completed the implication @xmath813 once we prove that a hilbert semistable curve does not admit an elliptic chain .
we accomplish this in proposition [ p : echain - unstable ] and corollary [ c : basin - cr-1br ] .
in this section , we classify c - semistable curves with infinite automorphisms . [ d : rosary ] an _ open rosary _ @xmath814 of length @xmath51 is a two - pointed connected curve @xmath815 such that * @xmath816 where @xmath817 is a smooth rational curve , @xmath818 ; * @xmath817 and @xmath819 meet each other in a single tacnode @xmath56 , for @xmath57 ; * @xmath820 if @xmath59 ; * @xmath821 and @xmath822 are smooth points .
an open rosary of length @xmath51 has arithmetic genus @xmath823 .
note that an open rosary of length @xmath824 is naturally an open elliptic chain of length @xmath825 .
we say that a curve @xmath45 _ admits an open rosary _ or length @xmath51 if there is a @xmath140-pointed open rosary @xmath815 and a morphism @xmath826 such that * @xmath68 is an isomorphism onto its image over @xmath827 ; * @xmath70 are nodes of @xmath45 ; we allow the case @xmath71 . a _ closed rosary _
@xmath45 is a curve admitting @xmath67 as above with the second condition replaced by * @xmath828 at a tacnode of @xmath45 .
a _ closed rosary with broken beads _ is a curve expressible as a union of open rosaries .
[ rem : rosary ] if @xmath45 admits an open rosary of length @xmath829 then @xmath45 admits a weak elliptic chain .
if @xmath51 is even then @xmath45 admits an elliptic chain .
thus a closed rosary of even length is also a closed weak elliptic chain .
[ p : rosary - aut ] consider the closed rosaries of genus @xmath830 .
if the genus is even then there is a unique closed rosary @xmath45 ( of the given genus ) and the automorphism group @xmath831 is finite .
if the genus is odd then the closed rosaries depend on one modulus and the connected component of the identity @xmath832 is isomorphic to @xmath375 .
there is a unique open rosary @xmath833 of length @xmath51 .
if @xmath834 denotes the automorphisms fixing @xmath40 and @xmath41 then @xmath835 it acts on tangent spaces of the endpoints with weights satisfying @xmath836 let @xmath45 be a closed @xmath51-rosary obtained by gluing @xmath51 smooth rational curves @xmath837\}$ ] so that @xmath838 let @xmath42 be another such rosary with the gluing data @xmath839 consider the morphism @xmath840 between the normalizations of @xmath45 and @xmath42 given by @xmath841 \mapsto [ \b_i s_i ' , t_i'].$ ] for @xmath265 to descend to an isomorphism from @xmath45 to @xmath42 , the following is necessary and sufficient : @xmath842 this gives rise to @xmath843 and @xmath844 . solving for @xmath845 , we get @xmath846 when @xmath51 is odd , there is no constraint and all @xmath51-rosaries are isomorphic .
when @xmath847 , @xmath848 forces the condition @xmath849 this means that the @xmath850-rosaries are parametrized by @xmath851 to describe the automorphisms we take @xmath852 .
when @xmath51 is odd we get @xmath853 for each @xmath234 which implies that @xmath832 is trivial . when @xmath854 we get a unique solution @xmath855 and thus @xmath856 .
the open rosary case entails exactly the same analysis , except that we omit the gluing datum @xmath857 associated with the end points .
thus we get a @xmath375-action regardless of the parity of @xmath51 . our assertion on the weights at the distinguished points @xmath40 and @xmath41 follows from the computation above of the action on tangent spaces . by
_ breaking the @xmath234th bead _ of a rosary ( open or closed ) , we mean replacing @xmath817 with a union @xmath858 of smooth rational curves meeting in a node such that @xmath859 meets @xmath860 in a tacnode @xmath861 and @xmath862 meets @xmath819 in a tacnode @xmath863 ( figure [ f : breaking - bead ] ) .
[ p : hsinf ] a c - semistable curve @xmath45 of genus @xmath864 has infinite automorphisms if and only if 1 .
@xmath45 admits an open rosary of length @xmath84 , or 2 .
@xmath45 is a closed rosary of odd genus ( possibly with broken beads ) .
we have already seen in proposition [ p : rosary - aut ] that closed rosaries of odd genus have infinite automorphisms .
let @xmath45 be a c - semistable curve of genus @xmath865 that is not a closed rosary .
for @xmath45 to have infinitely many automorphisms , it must have a smooth rational component , say @xmath454 . to satisfy the stability condition and
still give rise to infinite automorphisms , @xmath454 has to meet the rest of the curve in one node and a tacnode , or in two tacnodes .
we examine each case below : \(1 ) @xmath454 meets the rest in one node @xmath866 and in a tacnode @xmath867 : for the automorphisms of @xmath454 to extend to automorphisms of @xmath45 , the irreducible component @xmath868 containing @xmath867 must be a smooth rational component - this follows easily from that an automorphism of @xmath45 lifts to an automorphism of its normalization . also , @xmath455 has to meet the rest of the curve in one point @xmath869 other than @xmath867 since otherwise @xmath870 would be an elliptic tail ( or @xmath871 and @xmath45 is of genus two ) .
\(2 ) @xmath454 meets the rest in two tacnodes @xmath866 and @xmath867 : for the automorphisms to extend to @xmath45 , the components @xmath872 containing @xmath866 and @xmath873 containing @xmath867 must be smooth rational curves .
hence @xmath45 contains @xmath874 which is a rosary of length three .
moreover , @xmath193 and @xmath455 do not intersect : if they do meet , say at @xmath869 , then either @xmath875 and the genus of @xmath45 is of genus three ( if @xmath869 is a node ) or @xmath45 is a closed rosary if @xmath869 is a tacnode . iterating , we eventually produce an open rosary @xmath876 of length @xmath829 containing @xmath454 as a bead .
[ c : hsinf1 ] an h - semistable curve @xmath45 of genus @xmath877 has infinite automorphisms if and only if 1 .
@xmath45 admits an open rosary of odd length @xmath878 , or 2 .
@xmath45 is a closed rosary of odd genus ( possibly with broken beads ) .
let @xmath45 be a c - semistable curve and suppose @xmath37 is a deligne - mumford stabilization of @xmath45 .
in other words , there exists a smoothing of @xmath45 @xmath879 such that @xmath880 in the moduli space of stable curves . here , a smoothing is a flat proper morphism to a smooth curve with distinguished point @xmath881 such that @xmath882 and the generic fiber is smooth .
our classification result ( proposition [ p : hsinf ] ) has the following immediate consequence : [ c : hsinf ] suppose @xmath45 is a c - semistable curve with infinite automorphism group
. then the deligne - mumford stabilization and the pseudo - stabilization of @xmath45 admit an elliptic bridge .
indeed , @xmath45 necessarily admits a tacnode which means that its stabilization contains a connected subcurve of genus one meeting the rest of the curve in two points .
we will eventually give a complete description of the semistable and stable points of @xmath156 and @xmath157 .
for our immediate purpose , the following partial result will suffice : [ t : partial ] if @xmath45 is c - stable , i.e. , a pseudostable curve admitting no elliptic bridges , then @xmath883 . thus the hilbert point @xmath190_m
\in \hg2^{s , m}$ ] for @xmath320 .
the git - stable loci @xmath172 and @xmath884 , contain the nonsingular curves by @xcite .
( we discussed the relation between chow and asymptotic hilbert stability in section [ subsect : cphs ] . )
recall that proposition [ p : embedcstable ] guarantees that c - semistable curves admits bicanonical embeddings .
in particular , this applies to pseudostable curves without elliptic bridges .
suppose that @xmath45 is a singular pseudostable curve without elliptic bridges .
assume that @xmath627 is _ not _ in @xmath172 .
if @xmath627 is strictly semistable then it is c - equivalent to a semistable curve @xmath42 with infinite automorphism group .
it follows that @xmath45 is a pseudo - stabilization of @xmath42 , and we get a contradiction to corollary [ c : hsinf ] .
suppose @xmath627 is unstable and let @xmath42 denote a semistable replacement ( see theorem [ t : ss - replacement ] ) . by uniqueness of the pseudo - stabilization
, @xmath42 is not pseudostable but has @xmath45 as its pseudo - stabilization .
it follows that @xmath42 has a tacnode .
however , the pseudo - stabilization of such a curve necessarily contains an elliptic bridge . with our current partial understanding of the git of bicanonical curves
, we are ready to prove theorem [ t : main12 ] .
our main task is to establish isomorphisms ( [ eqn : tmain1 ] ) and ( [ eqn : tmain2 ] ) .
proposition [ p : makepsi ] established the existence of a birational contraction morphism @xmath885 .
the first step here is to show that @xmath886 and @xmath887 are birational contractions of @xmath9 and small contractions of @xmath173 . in particular
, we may identify the divisor class groups of these git quotients with the divisor class group of @xmath173 ( which in turn is a subgroup of the divisor class group of @xmath9 ) .
furthermore , we obtain @xmath888 and lemma [ l : logterm ] gives @xmath889 the second step is to compute the induced polarizations of @xmath886 and @xmath887 in the divisor class group of @xmath173 .
this will show that @xmath890 ( resp .
@xmath891 ) is ample on @xmath886 ( resp .
@xmath887 ) . isomorphisms ( [ eqn : tmain1 ] ) and ( [ eqn : tmain2 ] ) then follow from ( [ e : smallproj ] ) . to realize our git quotients as contractions of @xmath9
, we apply theorem [ thm : getcontraction ] in the bicanonical case . consider the complement of the deligne - mumford stable curves @xmath892 in the git - semistable locus @xmath171 ; we must show this has codimension @xmath84 . since @xmath893 and @xmath894 is an isomorphism where @xmath895 denotes the cycle class map from @xmath278 to @xmath896 , the analogous statement for the hilbert scheme follows immediately .
proposition [ p : chow - imply - c ] implies @xmath897 parametrizes * pseudostable curves that are not deligne - mumford stable , i.e. , those with cusps ; and * c - semistable curves with tacnodes .
the cuspidal pseudostable curves have codimension two in moduli ; the tacnodal curves have codimension three .
indeed , a generic tacnodal curve of genus @xmath10 is determined by a two - pointed curve @xmath52 of genus @xmath38 and an isomorphism @xmath898 .
we conclude there exist rational contractions @xmath899 and @xmath900 .
it remains to show that we have small contractions @xmath901 and @xmath902 . to achieve this
, we must establish that @xmath77 is the unique exceptional divisor of @xmath903 ( resp .
@xmath904 ) .
the exceptional locus of @xmath903 ( resp .
@xmath904 ) lies in the complement to the git - stable curves in the moduli space @xmath905 chow stable points are asymptotically hilbert stable ( cf .
proposition [ prop : hilbtochow ] ) , i.e. , @xmath906 when @xmath907 .
it suffices then to observe that @xmath85 is the unique divisorial component of @xmath908 , which is guaranteed by theorem [ t : partial ] .
theorem [ thm : getcontraction ] gives moving divisors on @xmath9 inducing the contractions @xmath903 and @xmath904 .
we apply equation ( [ e : lm ] ) , which in our situation takes the form @xmath909 this approaches @xmath910 as @xmath911 .
thus we have @xmath912 and @xmath913 using the identity @xmath914 we obtain ( for @xmath320 ) @xmath915 and @xmath916 it follows then that @xmath917 and @xmath918 the proof of theorem [ t : main12 ] will be complete if we can show that @xmath153 is the flip of @xmath8 . more precisely , for small positive @xmath165 , @xmath153 is a small modification of @xmath83 with @xmath919 ample . since @xmath161 and @xmath160 ar both small contractions of @xmath83 , @xmath153 is small as well .
and the polarization we exhibited on @xmath160 gives the desired positivity , which completes the proof of theorem [ t : main12 ] .
in this section , we analyze whether c - semistable curves are git - semistable with respect to the one - parameter subgroups of their automorphism group .
we shall also use deformation theory to classify the curves that belong to basins of attraction of such curves .
our analysis will focus primarily on the hilbert points .
indeed , corollary [ c : all - m ] shows that we can recover the sign of the hilbert - mumford index of the chow point from the indices of the hilbert points . and
in view of the cycle map @xmath920 , if @xmath190_m \in a_{\rho}([c^{\star}]_m)$ ] for @xmath907 then @xmath921 .
[ p : hs - o - ros ] let @xmath922 be a c - semistable curve of genus @xmath10 consisting of a genus @xmath923 curve @xmath37 meeting the genus @xmath51 curve @xmath773 in two nodes @xmath866 and @xmath924 where @xmath925 is a rosary of length @xmath830 , and @xmath926 and @xmath927 .
there is a one - parameter subgroup @xmath188 coming from the automorphisms of @xmath45 of the rosary @xmath773 such that for all @xmath367 , 1 .
@xmath626_m,\rho ) = 0 $ ] if @xmath51 is even ; 2 .
@xmath626_m,\rho ) = -m + 1 $ ] if @xmath51 is odd . in particular ,
@xmath45 is hilbert unstable if @xmath773 is of even length and strictly semistable otherwise .
an application of corollary [ c : all - m ] then yields : [ c : cs - o - ros ] let @xmath45 and @xmath188 be as in proposition [ p : hs - o - ros ]
. then @xmath45 is chow strictly semistable with respect to @xmath188 and @xmath928 . upon restricting @xmath74 to @xmath37 and each component of @xmath97
, we get * @xmath929 ; * @xmath930 ; * @xmath931 ; * @xmath932 , @xmath933 .
hence we may choose coordinates @xmath692 , @xmath934 , such that 1 .
@xmath935 is parametrized by @xmath936 \mapsto [ s_1 ^ 2 , s_1t_1 , t_1 ^ 2 , 0 , \dots , 0];\ ] ] 2 .
@xmath937 is parametrized by @xmath938\mapsto [ \underbrace{0 , \dots , 0}_{3r-2 } , s_{r+1}t_{r+1 } , s_{r+1}^2 , t_{r+1}^2 , 0 , \dots , 0].\ ] ] 3 . for @xmath939
, @xmath940 is parametrized by @xmath941 \mapsto [ \underbrace{0 , \dots , 0}_{3j-5 } , s_j^3t_j , s_j^4 , s_j^2t_j^2 , s_jt_j^3 , t_j^4 , 0 , \dots , 0].\ ] ] 4 .
@xmath37 is contained in the linear subspace @xmath942 and @xmath943 $ ] and @xmath944 $ ] . from the parametrization
, we obtain a set of generators for the ideal of @xmath97 : @xmath945 and for @xmath946 , @xmath947 in proposition [ p : rosary - aut ] we showed that @xmath375 acts on the open rosary via automorphisms . with respect to our coordinates , this is the one - parameter subgroup @xmath188 with weights : @xmath948 by considering the parametrization , it is easy to see that @xmath45 is stable under the action of @xmath188
. now we shall enumerate the degree two monomials in the initial ideal of @xmath45 . from ( [ e : gen1 ] ) and ( [ e : gen2 ] ) , we get the following monomials in @xmath949 : @xmath950 the weights of these @xmath951 monomials sum up to give @xmath952 the total weight @xmath953 of all degree two monomials in @xmath954 is @xmath955 hence the degree two monomials in @xmath949 that are _ not _ in the initial ideal contributes , to the total weight , @xmath956 the rest of the contribution comes from the monomials supported on the component @xmath37 : these are the degree two monomials in @xmath957 that vanish at @xmath866 and @xmath958 .
the number of such monomials is , by riemann - roch , @xmath959 since @xmath960 for all @xmath961 , these monomials contribute @xmath962 to the sum . combining ( [ e : l - deg2 ] ) and
( [ e : d - deg2 ] ) , we find the sum of the weights of the degree two monomials not in @xmath963 to be @xmath964 on the other hand , the average weight is @xmath965 hence by ( [ eqn : stabcrit ] ) , we find that @xmath626_2 , \rho ) = 0 $ ] if @xmath51 is even and @xmath626_2,\rho ) = -1 $ ] if @xmath51 is odd .
we enumerate the degree three monomials in the same way : the degree three monomials in @xmath949 that are in the initial ideal are the multiples of ( [ e : inl - deg2 ] ) together with @xmath966 that come from the linear relation @xmath967 from this , we find that the degree three monomials in @xmath954 that are not in the initial ideal contribute @xmath968 the contribution from @xmath37 is @xmath969 hence the grand total is @xmath970 on the other hand , the average weight is @xmath971 using ( [ eqn : stabcrit ] ) , we compute @xmath625_3 , \rho ) = \begin{cases } 0 , \quad \mbox { if $ r$ is even;}\\ -2 , \quad \mbox{if $ r$ is odd . }
\end{cases}\ ] ] corollary [ c : all - m ] implies @xmath625_m,\rho ) = \begin{cases}0 ,
\quad \mbox { if $ r$ is even;}\\ -m+1 , \quad \mbox{if $ r$ is odd . } \end{cases}\ ] ] for each @xmath191 .
let @xmath45 and @xmath773 be as in the previous section .
let @xmath972 be homogeneous coordinates on @xmath817 .
we may assume that @xmath973 ; \
, a_{r+1 } = [ 1 , 0 ] ; \ , a_i = \begin{cases } [ 1,0 ] = \infty \ , \ , { \mbox on } \ , \ , l_i\\ [ 0,1 ] = 0 \ , \ , { \mbox on } \ , \ , l_{i+1 } \\ \end{cases } \\\ ] ] consider the @xmath375 action associated to @xmath188 .
the action on @xmath773 is given by @xmath974 ) \mapsto [ x_i , t^{(-1)^{i-1 } } y_i]$ ] on each @xmath817 .
hence it induces an action on the tangent space @xmath975 given by @xmath976 there is an induced @xmath375 action on the hilbert scheme and @xmath157 .
corollary [ c : versal ] asserts that a neighborhood of @xmath190 $ ] in the hilbert scheme dominates the product of the versal deformation spaces .
these inherit a @xmath375 action as well , which we shall compute explicitly .
\(a ) _ @xmath375 action on the versal deformation spaces of nodes @xmath866 and @xmath977 _ : let @xmath978 be a local parameter at @xmath866 on @xmath37 .
we have @xmath979 as a local parameter at @xmath866 on @xmath980 and the local equation at @xmath866 on @xmath45 is @xmath981 .
hence the action on the node @xmath866 is given by @xmath982 and the action on the versal deformation space is @xmath983 likewise , at @xmath924 , the action on the node is @xmath984 where @xmath985 is a local parameter at @xmath924 on @xmath37 , and the action on the versal deformation space is @xmath986 \(b ) _ @xmath375 action on the versal deformation space of a tacnode @xmath56 _ : at @xmath56 , the local analytic equation is of the form @xmath987 where @xmath988 and @xmath989 in @xmath990\oplus k[[x_{i+1}/y_{i+1}]]$ ] and the @xmath375 action at the tacnode is given by @xmath991 therefore the action on the versal deformation space is @xmath992 from these observations , we conclude that the basin of attraction of @xmath45 with respect to @xmath188 contains arbitrary smoothings of @xmath993 but no smoothing of @xmath994 for all @xmath995 .
we have established : [ p : hs - echain ] retain the notation of proposition [ p : hs - o - ros ] and assume that @xmath320 . 1 .
if @xmath51 is even ( i.e. , the length of the rosary is odd ) then @xmath996_m)$ ] ( resp .
@xmath997_m)$ ] ) parametrizes the curves consisting of @xmath37 and a weak elliptic chain @xmath42 of length @xmath998 meeting @xmath37 in a node at @xmath866 and in a tacnode at @xmath924 ( resp . in a tacnode at @xmath866 and in a node at @xmath924 ) ( figure [ f : basin - o - ros - even1 ] ) ; 2 .
if @xmath51 is odd ( i.e. , the length of the rosary is even ) then @xmath996_m)$ ] ( resp .
@xmath997_m)$ ] ) parametrizes the curves consisting of @xmath37 and an elliptic chain @xmath42 of length @xmath999 ( resp .
length @xmath1000 ) meeting @xmath37 in a node ( resp .
tacnode ) at @xmath866 and @xmath924 ( figure [ f : basin - o - ros - odd ] ) .
when @xmath312 , @xmath1001_m)$ ] consists of tacnodal curves normalized by @xmath37 .
it follows from proposition [ p : hs - echain ] and proposition [ p : hs - o - ros ] that [ p : echain - unstable ] if a bicanonical curve admits an open elliptic chain then it is hilbert unstable .
in particular , a bicanonical curve with an elliptic bridge is hilbert unstable .
the closed case can be found in proposition [ p : basin - cr-1br ] and corollary [ c : basin - cr-1br ] .
[ p : hss - c - ros ] let @xmath45 be a bicanonical closed rosary of even length @xmath51 .
then @xmath45 is hilbert strictly semistable with respect to the one - parameter subgroup @xmath1002 arising from @xmath831 .
the relevant one - parameter subgroup was introduced in proposition [ p : rosary - aut ] .
restricting @xmath1003 to each component @xmath817 , we find that each @xmath817 is a smooth conic in @xmath155 .
we can choose coordinates @xmath692 such that @xmath817 is parametrized by * @xmath1004 \mapsto [ \underbrace{0 , \dots , 0}_{3(i-1 ) } , s_i^3t_i , s_i^4 , s_i^2t_i^2 , s_it_i^3 , t_i^4 , 0 , \dots , 0]$ ] , @xmath1005 ; * @xmath1006 \mapsto [ s_rt_r^3 , t_r^4 , 0 , \dots , 0 , s_r^3t_r , s_r^4 , s_r^2t_r^2]$ ] the normalization of @xmath45 admits the automorphisms given by @xmath1007 \mapsto [ \a^{sgn(i)}s_i , \ , \a^{1-sgn(i)}t_i ] , \quad sgn(i ) : = i - 2\lfloor i/2 \rfloor,\ ] ] for @xmath57 and @xmath1008 \mapsto [ s_r , \a t_r ] $ ] . the one - parameter subgroup @xmath188 associated to this automorphism has weights @xmath1009 the sum of the weights
@xmath1010 is @xmath1011 if @xmath51 is even and @xmath1012 if @xmath51 is odd . from the parametrization
, we obtain a set of generators for the ideal of @xmath45 : @xmath1013 we have the following @xmath1014 degree two monomials that are in the initial ideal : @xmath1015 the sum of the weights of these monomials is @xmath1016 .
it follows that the sum of the weights of the monomials _ not _ in the initial ideal is @xmath1017 which is precisely @xmath1018 hence @xmath626_2 , \rho ) = 0 $ ] .
we shall now enumerate the degree three monomials in the initial ideal of @xmath45 . together with the monomials divisible by the monomials from ( [ e : incr ] ) , we have the initial terms @xmath1019 that come from the grbner basis members @xmath1020 the degree three monomials in ( [ e : inl - syz - deg3 ] ) and the degree three monomials divisible by monomials in ( [ e : incr ] ) have total weight @xmath1021 .
this agrees with the average weight @xmath1022 therefore , @xmath626_2 , \rho ) = 0 = \mu([c]_3 , \rho ) $ ] and @xmath45 is @xmath189-hilbert strictly semistable by corollary [ c : all - m ] .
since @xmath1023 , a one - parameter subgroup coming from @xmath831 is of the form @xmath1024 for some @xmath1025 , and we have @xmath625_m , \rho^a ) = a \ , \mu([c]_m , \rho ) = 0.\ ] ] [ p : gen - ss - c ] retain the notation of proposition [ p : hss - c - ros ] .
then the basin of attraction @xmath996_m)$ ] parametrizes the closed weak elliptic chains of length @xmath998 ( figure [ f : basin - c - ros ] ) .
we use the parametrization from the proof of proposition [ p : hss - c - ros ] .
@xmath45 has tacnodes @xmath1026 $ ] , @xmath1027 . from the parametrization , we find that the local parameters @xmath1028 at @xmath56 to the two branches is acted upon by @xmath188 with weight @xmath1029 .
it follows that @xmath188 acts on the versal deformation space @xmath1030 of the tacnode @xmath56 with weights @xmath1031 .
hence the basin of attraction @xmath996)$ ] has arbitrary smoothings of @xmath56 for odd @xmath234 but no nontrivial deformations of @xmath56 for even @xmath234 .
closed rosaries with broken beads of _ even _ genus are unstable : [ p : hs - cr-1br ] let @xmath1032 be an odd number and @xmath1033 the curve obtained from a closed rosary of length @xmath51 by breaking a bead
. then there exists a one - parameter subgroup @xmath188 of @xmath1034 with @xmath1035_m,\rho ) = 1-m $ ] for each @xmath191 , so @xmath1036_m$ ] is unstable . furthermore , @xmath1037 is strictly semistable with respect to @xmath188 .
note that @xmath1033 is unique up to isomorphism and it can parametrized by @xmath1038we give the set of monomials the graded @xmath188-weighted lexicographic order , where @xmath188 is the one - parameter subgroup with the weight vector @xmath1039 a grbner basis for @xmath1033 is : @xmath1040 and for @xmath1041 , @xmath1042 together with the following degree three polynomials @xmath1043 the degree two initial monomials are : @xmath1044and for @xmath1041 , @xmath1045the sum of the weights of the monomials in ( [ e:1bead - deg2 - 1 ] ) is @xmath1046 , whereas the monomials in ( [ e:1bead - deg2 - 2 ] ) contribute @xmath1047 to the total weight of the monomials in the initial ideal .
the total weight of all degree two monomials is @xmath1048 .
hence the weights of all degree two monomials not in the initial ideal sum up to @xmath1049 on the other hand , the average weight @xmath1050 is @xmath1051 .
it follows from proposition [ prop : stabcrit2 ] that @xmath1035_2 , \rho ) = ( -(28r-13 ) + 28r-14 ) = -1 $ ] .
the degree three monomials divisible by the ones in the lists ( [ e:1bead - deg2 - 1 ] ) , ( [ e:1bead - deg2 - 2 ] ) contribute @xmath1052 to the total weight of the monomials in the initial ideal . on the other hand ,
the monomials @xmath1053 coming from the degree three grbner basis members ( [ e : deg3-gb ] ) contribute @xmath1054 .
the sum of the weights of all degree three monomials is @xmath1055 .
hence the total weight of the degree three monomials not in the initial ideal is @xmath1056 on the other hand , the average weight is @xmath1057 by proposition [ prop : stabcrit2 ] , the hilbert - mumford index is @xmath1035_3 , \rho ) = - ( 66r - 31 ) + 66r - 33 ) = - 2.$ ] since @xmath1035_2 , \rho ) = 2 \mu([c_r]_3 , \rho ) < 0 $ ] , it follows from corollary [ c : all - m ] that @xmath1033 is @xmath189-hilbert unstable for all @xmath367 .
indeed , we find that @xmath1058_m,\rho)=1-m\ ] ] for each @xmath191 and @xmath1059 .
[ p : basin - cr-1br ] let @xmath1033 and @xmath188 be as in proposition [ p : hs - cr-1br ] .
then the basin of attraction @xmath996_m)$ ] parametrizes closed elliptic chains @xmath52 of length @xmath999 such that @xmath71 ( figure [ f : basin - c - ros-1broken ] ) .
[ c : basin - cr-1br ] a closed elliptic chain is hilbert unstable . at the node ,
the local analytic equation is given by @xmath1060 and @xmath375 acts on the local parameters @xmath1061 and @xmath1062 with weight @xmath1063 .
hence @xmath375 acts on the local versal deformation space ( defined by @xmath1064 ) with weight @xmath1065 . at the adjacent tacnode , @xmath375 acts on the tangent space
to the two branches with positive weights : the tangent lines are traced by @xmath1066 and @xmath1067 .
in fact , @xmath375 acts on the local versal deformation of the tacnode ( defined by @xmath1068 ) with a positive weight vector @xmath1069 .
similar analysis reveals that @xmath375 acts on the subsequent tacnode with a negative weight vector @xmath809 . using the symmetry of the rosary
, we can conclude that @xmath375 acts on the local versal deformation space of the tacnodes with weight vector alternating between @xmath1070 and @xmath1071 .
the assertion now follows .
our main goal is a complete description of all c - equivalence and h - equivalence classes . throughout , each c - semistable curve @xmath45
is embedded bicanonically ( cf . proposition [ p : embedcstable ] ) @xmath1072 and we consider the corresponding chow points @xmath682 and hilbert points @xmath190_m \in \hg2 , m \gg 0 $ ] .
to summarize : * if @xmath45 is c - stable ( resp .
h - stable ) , then the equivalence class of @xmath45 is trivial : it coincides with the @xmath1073 orbit of @xmath627 ( resp .
@xmath190_m$ ] for @xmath320 ) . *
if @xmath45 is strictly c- or h - semistable its equivalence class is nontrivial .
we shall identify the unique closed orbit curve and describe all equivalent curves .
* since closed orbit curves are separated in a good quotient @xcite , we have a complete classification of curves identified in the quotient spaces @xmath1074 and @xmath1075 .
* @xmath1076 consists of connected genus - one curves @xmath1077 such that @xmath54 meets @xmath55 in a node @xmath56 , @xmath1078 ; * @xmath1079 if @xmath1080 ; * @xmath61 and @xmath1081 are smooth points . 1 .
fix a subset @xmath1083 of the genus - one subcurves arising in elliptic bridges .
2 . choose a subset of the nodes of @xmath45 lying on @xmath1084 consisting of points of the following types : * if @xmath1085 for some distinct @xmath1086 then the node where they intersect must be included .
* nodes where the @xmath1087 meet other components may be included .
3 . replace each of these nodes by a smooth @xmath583 ( for any point of our subset ) or by a chain of two smooth @xmath583 s ( only for points of the first type ) .
precisely , let @xmath1088 denote the curve obtained by normalizing our set of nodes and then joining each pairs of glued points with a @xmath583 or a chain of two @xmath583 s with one component meeting each glued point .
4 . let @xmath1089 denote the proper transform of @xmath1087 , which are pairwise disjoint in @xmath1088 .
replace each @xmath1089 with a tacnode .
precisely , write @xmath1090 and consider a morphism @xmath1091 such that * @xmath1092 is an isomorphism and @xmath1093 is the normalization ; * for @xmath384 , @xmath646 contracts @xmath1089 to a tacnode of @xmath42
. the generic curve @xmath42 produced by this procedure does not admit components isomorphic to @xmath583 containing a node of @xmath42 .
we introduce @xmath583 s in step ( 3 ) only to separate two adjacent contracted elliptic components .
when enumerating the c - semistable curves , it is convenient to use a graph that is similar to the dual graph : we use a single line to denote a node and a double line to denote a tacnode ( e.g. figure [ f : graph ] ) .
( of lemma [ l : possibilities ] ) our hypotheses give a flat family @xmath1100 \tag{$\dag$}\ ] ] whose generic fibre @xmath1101 is smooth and the special fibre @xmath1102 is @xmath42 .
furthermore , after a base change @xmath1103 \\ t & \mapsto & t_1^s \end{array}\ ] ] there exists a birational modification over @xmath1104 @xmath1105 such that @xmath1106 is @xmath45 .
in other words , @xmath1107 is the _ pseudostable reduction _ of @xmath1108 ; we replace each tacnode by an elliptic bridge and contract any rational component that meets the rest of the curve in fewer than three points .
let @xmath1109 be the normalization of the graph of @xmath1110 , with @xmath1111 and @xmath1112 the projections to @xmath1113 and @xmath1114 respectively : @xmath1115_-{\pi_1}\ar[rd]^-{\pi_2 } & \\ \mcl c & & \mcl c ' \times_b b_1 \\ } \ ] ] by @xcite , @xmath1109 is flat over @xmath132 and @xmath1116 is reduced . an argument similar to @xcite yields * the exceptional locus of @xmath1112 is a disjoint union of connected genus - one subcurves @xmath1117 that arise as proper transforms of components of elliptic bridges in @xmath45 .
each component is mapped to a tacnode of @xmath42 . *
the exceptional locus of @xmath1111 is a union of chains of rational curves of length one or two @xmath1118 that arise as proper transforms of rational components of @xmath42 meeting the rest of the curve in two points ( either two tacnodes or one node and one tacnode ) .
each component is mapped to a node of @xmath45 contained in an elliptic bridge .
we analyze the generic curves arising from our procedure .
suppose there is a component isomorphic to @xmath583 meeting the rest of the curve in a node and a tacnode .
corollary [ c : versal ] implies we can smooth the node to get a c - semistable curve .
the smoothed curve also arises from our procedure .
[ r : subsets ] lemma [ l : possibilities ] yields a bijection between subsets @xmath1119 and generic configurations of the locus of curves arising from our procedure .
indeed , there is a unique generic configuration contracting the curves @xmath1120 .
[ p : degenerate ] let @xmath45 be strictly c - semistable without tacnodes and @xmath1082 the genus - one subcurves of @xmath45 arising as components of elliptic bridges .
let @xmath197 be the curve obtained from @xmath45 by replacing each @xmath54 with an open rosary @xmath1121 of length two .
then there exists a one - parameter subgroup @xmath1122 such that @xmath1123 and @xmath1124 .
the analysis of proposition [ p : rosary - aut ] makes clear that our description of @xmath197 determines it uniquely up to isomorphism .
furthermore , we have @xmath1128 with basis @xmath1129 ; here @xmath1130 denotes the one - parameter subgroup acting trivially on @xmath1131 and with weight @xmath145 on the tangent spaces @xmath1132 and @xmath1133 . as explained in
[ s : ba - o - ros ] , it acts with negative weights on the versal deformation space of the tacnode of @xmath1134 . consider the one - parameter subgroup @xmath1135 which acts with positive weights on each of the tacnodes .
the basin of attraction analysis of proposition [ p : hs - echain ] shows that @xmath1136 parametrizes those curves obtained by replacing each open rosary of @xmath197 with an elliptic bridge / chain of length one .
this includes our original curve @xmath45 .
section [ s : ba - o - ros ] gives the action of @xmath1140 on the versal deformations of the singularities of @xmath197 .
it acts with weights @xmath1141 on the versal deformation space of the tacnode on @xmath1134 . at a node ( @xmath752 or @xmath1142 )
lying on a single open rosary @xmath1134 of length two , it acts with weight @xmath1143 . for nodes on two open rosaries @xmath1134 and @xmath1144 , it acts with weight @xmath1145 .
restrict attention to one - parameter subgroups with weights @xmath1146 for each @xmath1147 .
these naturally divide up into @xmath1148 equivalence classes , depending on the signs of the @xmath1149 .
let @xmath1150 denote those indices with @xmath1151 .
just as in the proof of proposition [ p : hs - echain ] , the basin of attraction @xmath1152 does not contain smoothings of the tacnodes in @xmath1153 but does contain all smoothings of the remaining tacnodes . choosing the negative @xmath1149 suitably large in absolute value , we can assume each @xmath1154 , so the nodes where two rosaries meet are smoothed provided at least one of the adjacent tacnodes is _ not _ smoothed .
thus @xmath1152 consists of the c - semistable curves obtained by smoothing all the tacnodes _ not _ indexed by @xmath1155 , as well as the nodes on the rosaries containing one of the remaining tacnodes ( indexed by @xmath1155 ) .
the generic member of the basin equals the generic configuration indexed by @xmath1155 , as described in remark [ r : subsets ] .
it follows that _ each _ curve @xmath42 enumerated in lemma [ l : possibilities ] appears in the the basin of attraction of @xmath1156 for a suitable one - parameter subgroup @xmath1157 .
let @xmath42 denote a strictly c - semistable curve , with tacnodes and/or elliptic bridges .
assume that @xmath42 is chow unstable and let @xmath1100\ ] ] be a smoothing .
let @xmath1158 be a chow semistable reduction of this family ( see theorem [ t : ss - replacement ] ) and @xmath45 the pseudostable reduction .
reversing the steps outlined in lemma [ l : possibilities ] , we see that @xmath45 is obtained by replacing each tacnode of @xmath42 ( or @xmath1158 ) with an elliptic bridge and then pseudo - stabilizing .
let @xmath197 denote the curve obtained from @xmath45 in proposition [ p : degenerate ] , which guarantees that @xmath1159 and @xmath1160 are contained in basins of attraction @xmath199 and @xmath1161 respectively .
moreover , since @xmath1162 lemma [ l : flat limit ] implies that @xmath42 ( resp .
@xmath1158 ) is chow semistable iff @xmath45 is chow semistable .
this contradicts our assumption that @xmath42 is chow unstable .
if @xmath45 is a strictly c - semistable curve with closed orbit then @xmath1163 where @xmath1164 is the number of tacnodes and the superscript @xmath1165 denotes the connected component of the identity .
assume that @xmath42 is a strictly semistable curve with closed orbit .
let @xmath45 be a pseudostable reduction and @xmath197 the curve specified in proposition [ p : degenerate ] , so the chow point of @xmath42 is in the basin of attraction of the chow point of @xmath197 . since @xmath197 is chow semistable ,
we conclude that @xmath1166 .
conversely , suppose @xmath42 is a curve satisfying the three conditions of proposition [ p : c - minorbit ] .
again , let @xmath45 be a pseudostable reduction of @xmath42 and @xmath197 the curve obtained in proposition [ p : degenerate ] , so that @xmath42 is in the basin of attraction of @xmath197 for some one - parameter subgroup @xmath1140 .
note that @xmath197 also satisfies the conditions of proposition [ p : c - minorbit ] .
the basin of attraction analysis in section [ s : ba - o - ros ] implies that any nontrivial deformation of @xmath197 in @xmath199 induces a nontrivial deformation of at least one of the singularities of @xmath197 sitting in an open rosary .
there are three cases to consider : first , we could deform the tacnode on one of the rosaries @xmath1134 .
however , then the rosary @xmath1134 deforms to an elliptic bridge in @xmath42 that is not a length two rosary , which yields a contradiction .
therefore , we may assume that none of the tacnodes in @xmath197 is deformed in @xmath42 .
second , we could smooth a node where length two rosaries meet .
however , this would yield a rosary in @xmath42 of length @xmath609 .
finally , we could smooth a node where a length two rosary @xmath1134 meets a component not contained in an rosary .
however , the tacnode of @xmath1134 then deforms to a tacnode of @xmath42 not on any length two rosary .
suppose that @xmath45 is a h - semistable bicanoncal curve . by definition
it is also c - semistable and thus chow - semistable by the analysis of section [ ss : csbc ] . of course
, strictly chow - semistable points can be hilbert unstable , and we classify these in two steps .
first , we enumerate the curves @xmath193 with strictly semistable chow point such that there exists a one - parameter subgroup @xmath1167 destabilizing the hilbert point of @xmath193 , i.e. , with @xmath194_m,\rho)<0 $ ] for @xmath907 .
second , we list the curves that are in the basins of attraction @xmath1168_m)$ ] , which are also guaranteed to be hilbert unstable by the hilbert - mumford one - parameter subgroup criterion .
proposition [ p : destabilize ] shows these are _ all _ the hilbert unstable curves : if @xmath45 is a c - semistable bicanonical curve that is hilbert unstable then its chow point is contained in the basin of attraction @xmath1169 of a chow semistable curve @xmath193 with closed orbit such that @xmath194_m , \rho ) < 0.$ ] suppose that @xmath193 has open rosaries @xmath1171 .
each contributes @xmath375 to the automorphism group of @xmath193 and @xmath1172 .
let @xmath1173 denote the nodes in the intersection @xmath1174 .
the automorphism coming from @xmath1175 gives rise to a one - parameter subgroup @xmath1176 where the second @xmath375 means the @xmath234th copy in the product @xmath1177 and @xmath1178 is the stabilizer group .
we assume that @xmath1179 are open rosaries of even length and @xmath1180 are of odd length . for @xmath1181 ,
the weights of @xmath1130 on the versal deformation spaces of @xmath752 and @xmath1142 have the same sign ( see [ s : ba - o - ros ] ) .
we normalize @xmath1130 so that this weight is negative .
given our one - parameter subgroups @xmath1182 with @xmath194_m,\rho)<0 $ ] , we can expand @xmath1183 so that @xmath1184_m , \rho ) = \sum_{i=1}^k a_i \ , \mu([c_0]_m , \rho_i ) + \sum_{i = k+1}^\ell b_i \ , \mu([c_0]_m , \rho_i ) < 0.\ ] ] we have already computed these terms : proposition [ p : hs - o - ros ] implies that @xmath194_m,\rho_i)=0 $ ] for @xmath1185 and @xmath194_m,\rho_i)=1-m$ ] for @xmath1186 .
thus in order for the sum to be negative , we must have @xmath1187 for some @xmath1186 .
in particular , there is at least one rosary of even length .
proposition [ p : hs - echain ] implies that the basin of attraction @xmath1168_m)$ ] contains curves with elliptic chains , which are not h - semistable .
we are left with the case of a closed rosary @xmath1033 of even genus with one broken bead .
there is a unique one - parameter subgroup @xmath188 of the automorphism group , and we choose the sign so that it destabilizes @xmath1033 ( cf .
proposition [ p : hs - cr-1br ] ) .
the basin of attraction analysis in proposition [ p : basin - cr-1br ] again shows that the curves with unstable hilbert points admit elliptic chains .
we shall now prove that if @xmath45 is h - stable then it is hilbert stable .
if @xmath45 is hilbert strictly semistable , then it belongs to a basin of attraction @xmath192_m)$ ] where @xmath193 is a hilbert semistable curve with infinite automorphisms and @xmath188 is a 1-ps coming from @xmath1170 . by corollary
[ c : hsinf ] , @xmath193 admits an open rosary of odd length @xmath878 or is a closed rosary of even length @xmath877 .
but we showed in propositions [ p : hs - echain ] and [ p : gen - ss - c ] that any curve in the basin of such @xmath193 has a weak elliptic chain and hence is not h - stable .
suppose @xmath42 is strictly h - semistable .
we shall show that there exists a curve @xmath1189 satisfying the conditions of proposition [ p : h - minorbit ] and a one - parameter subgroup @xmath1140 of @xmath1190 such that @xmath1191_m \in a_{\rho'}([c^{\ast}]_m)$ ] for @xmath907 and @xmath1192_m,\rho')=0 $ ] .
assume first that @xmath42 is a closed weak elliptic chain with @xmath51 components , with arithmetic genus @xmath1193 .
let @xmath1189 denote a closed rosary with beads @xmath1194 and tacnodes @xmath1195 .
proposition [ p : rosary - aut ] implies @xmath1196 , generated by a one - parameter subgroup @xmath188 acting on the versal deformation spaces of the @xmath1197 with positive weights and the @xmath1198 with negative weights .
proposition [ p : gen - ss - c ] implies @xmath1199_m)$ ] contains the closed weak elliptic chains of length @xmath51 .
now assume that @xmath42 is not a closed weak elliptic chain but contains maximal closed weak elliptic chains @xmath1200 of lengths @xmath1201 .
let @xmath1202 ( resp .
@xmath1203 ) denote the node ( resp .
tacnode ) where @xmath1204 meets the rest of the curve .
let @xmath1189 be the curve obtained from @xmath42 by replacing each @xmath1205 with a chain of @xmath1206 open rosaries of length three .
precisely , write @xmath1207 and let @xmath1208 denote a chain of @xmath1206 open rosaries of length three joined end - to - end
. then @xmath1189 is obtained by gluing @xmath1209 to @xmath37 via nodes at @xmath1202 and @xmath1203 .
one special case requires further explanation : if @xmath42 admits an irreducible component @xmath1210 meeting the rest of @xmath42 at two points @xmath1142 and @xmath1203 then we contract this component in @xmath1189 .
there are examples where the construction of @xmath1189 involves components being contracted .
let @xmath454 and @xmath455 be smooth and connected of genus @xmath84 and let @xmath1211 and @xmath1212 be elliptic . consider the curve @xmath42 @xmath1213 replacing the weak elliptic chains with rosaries of length three yields @xmath1214 which is not h - semistable .
contracting the middle @xmath583 , we obtain @xmath1189 : @xmath1215 we return to our proof : the curve @xmath1189 has @xmath1219 essentially repeating the argument of proposition [ p : degenerate ] , using the one - parameter subgroup analysis of proposition [ p : hs - o - ros ] and the basin - of - attraction analysis of proposition [ p : hs - echain ] ( or proposition [ p : hs - cr-1br ] and [ p : basin - cr-1br ] in the generate case ) , we obtain a one - parameter subgroup @xmath1140 in the automorphism group such that @xmath1191_m \in a_{\rho'}([c^{\ast}]_m)$ ] for @xmath907 and @xmath1192_m,\rho')=0 $ ] .
we now show that the curves enumerated in proposition [ p : h - minorbit ] all have closed orbits . due to ( * ? ? ?
* theorem 1.4 ) , it suffices to show none of these are contained in the basin of attraction of any other .
suppose that @xmath1220 and @xmath1221 are such that @xmath1222_m \in a_{\rho}([c^{\ast}_1]_m)\ ] ] for some one - parameter subgroup @xmath188 of @xmath1223 .
a nontrivial deformation of @xmath1220 necessarily deforms one of the singularities of @xmath1220 .
if the singularity is a tacnode on a length - three open rosary , the resulting deformation admits a weak elliptic chain that is not contained in a chain of length - three rosaries .
if the singularity is a node where two length - three open rosaries meet then the deformation admits a weak elliptic chain _ not _ contained in a chain of length - three open rosaries .
however , there is one case that requires special care : suppose that @xmath1220 is a closed chain of @xmath51 rosaries of length three @xmath1224 where @xmath1225 and @xmath814 meet at a node @xmath1226 ; this has arithmetic genus @xmath1193 .
let @xmath1221 denote a closed rosary of genus @xmath1193 , which is a deformation of @xmath1220 .
we need to insure that @xmath1227_m \not \in a_{\rho}([c^{\ast}_1]_m)\ ] ] for any one - parameter subgroup @xmath188 of @xmath1228_m)^{\circ}$ ] .
we can express @xmath1229 where @xmath1230 acts trivially except on @xmath1144 and has weights @xmath812 and @xmath1063 on @xmath1231 and @xmath1232 .
( here @xmath1233 acts with weights @xmath812 and @xmath1063 on @xmath1234 and @xmath1235 . )
however , assuming @xmath188 is nontrivial , one of the following differences @xmath1236 is necessarily negative ; for simplicity , assume @xmath1237 .
it follows that @xmath188 acts with negative weight on the versal deformation of the node @xmath1238 , thus deformations in @xmath1239_m)$ ] can not smooth @xmath1238 .
we conclude that deformations in the basin of attraction of @xmath1220 can not smooth each node , which yields ( [ eqn : tricky ] ) brendan hassett .
classical and minimal models of the moduli space of curves of genus two . in _ geometric methods in algebra and number theory _ , volume 235 of _ progress in mathematics _ , pages 160192 .
birkhuser , boston , 2005 .
jnos kollr and shigefumi mori .
, volume 134 of _ cambridge tracts in mathematics_. cambridge university press , cambridge , 1998 . with the collaboration of c. h. clemens and a. corti , translated from the 1998 japanese original .
jnos kollr . , volume 32 of _ ergebnisse der mathematik und ihrer grenzgebiete .
3 . folge . a series of modern surveys in mathematics [ results in mathematics and related areas .
3rd series .
a series of modern surveys in mathematics]_. springer - verlag , berlin , 1996 .
david mumford , john fogarty , and frances kirwan .
, volume 34 of _ ergebnisse der mathematik und ihrer grenzgebiete ( 2 ) [ results in mathematics and related areas ( 2)]_. springer - verlag , berlin , third edition , 1994 . | we give a geometric invariant theory ( git ) construction of the log canonical model @xmath0 of the pairs @xmath1 for @xmath2 $ ] for small @xmath3 .
we show that @xmath4 is isomorphic to the git quotient of the chow variety bicanonical curves ; @xmath5 is isomorphic to the git quotient of the asymptotically - linearized hilbert scheme of bicanonical curves . in each case
, we completely classify the ( semi)stable curves and their orbit closures .
chow semistable curves have ordinary cusps and tacnodes as singularities but do not admit elliptic tails .
hilbert semistable curves satisfy further conditions , e.g. , they do not contain elliptic bridges . we show that there is a small contraction @xmath6 that contracts the locus of elliptic bridges .
moreover , by using the git interpretation of the log canonical models , we construct a small contraction @xmath7 that is the mori flip of @xmath8 . |
[ sintro ] genome - wide association studies ( gwas ) constitute a popular approach for investigating the association of common single nucleotide polymorphisms ( snps ) with complex diseases .
usually , a large number of snp are tested across the genome . great interest lies in improving power of testing snp effects by borrowing additional biological information .
indeed , a major criticism of genetic association studies lies in its agnostic style [ @xcite ] : none of the biological knowledge was encoded in the standard genetic association analyses . to overcome such limitations ,
multi - marker analysis has been advocated to integrate biological information into statistical analyses and to decrease the number of tests [ @xcite ; @xcite ] .
analysis using snp - sets grouped by physical locations has better performance than the standard single snp analysis in reanalyzing the breast cancer gwas data [ @xcite ] .
snps can also be grouped into a snp - set according to biological pathways , in which a gene harmonizes with other genes to exert biological functions .
the two factors we try to bridge in genetic association studies are snps and disease risk . despite the success of snp - set analyses in assembling multiple snps based on biological information ,
mechanistic pathways between snps ( snp - sets ) and disease are still neglected .
given the availability of multiple sources in genomic data ( e.g. , gene expression and snps ) [ @xcite ; @xcite ] , it is desirable to perform joint analysis by integrating multiple sources of genomic data .
here we combine the information of snps and gene expression by introducing gene expression as a mediator in the causal pathway from snps to disease .
_ biologically _ , gene expression can be determined by the dna genotype [ @xcite ; @xcite ; @xcite ] and that gene expression can also affect disease risk [ @xcite ] .
moreover , results from the snp - set analysis augmented by a biological model can be more scientifically meaningful .
_ statistically _ , gene expression can help explain variability of the effect of snps on disease when there exists an effect of snps on disease via gene expression and thus increases the power of detecting the overall effect of snps on disease risk .
snps that regulate mrna expression of a gene are so - called expression quantitative trait loci ( eqtl ) [ @xcite ] . statistically , eqtl snps can be viewed as the snps that are correlated with mrna expression of a gene .
_ cis_-eqtl snps are the snps that are within or around the corresponding gene , and _
trans_-eqtl snps are those that are far away or even on different chromosomes .
numerous genome - wide eqtl analyses have been reported to comprehensively capture such a dna rna ( i.e. , snps - gene expression ) association in the genome in different tissues and organisms [ @xcite ; @xcite ; @xcite ] .
eqtl results can be external information to prioritize the discovery of susceptibility loci in genome - wide association studies [ @xcite ; @xcite ; @xcite ] .
methods are available to integrate multiple genomic data to draw causal inference on a biological network [ @xcite ; @xcite ; @xcite ; @xcite ] .
we focus in this paper on _ joint analysis _ of multiple eqtl snps of a gene and their corresponding mrna expression for their effects on disease phenotypes .
compared with multi - snp analyses , this approach further incorporates eqtls into genetic association studies and accounts for a biological process ( from dna to rna ) within a gene to improve power .
this paper is motivated by an asthma genome - wide association study of subjects of british descent ( mrc - a ) , in which the association between snps at the gene and the risk of childhood asthma was investigated [ @xcite ; @xcite ] .
the mrc - a data set consists of 108 cases and 50 controls with both snp genotype ( illumina 300k ) and gene expression ( affymetrix hu133a 2.0 ) data available .
the original genome - wide study reported that the 10 typed snps on chromosome 17q21 where _
ormdl3 _ is located were strongly associated with childhood asthma in mrc - a data , and the results were validated in several other independent studies .
the authors also found that each of these 10 snps was highly correlated with gene expression of _
ormdl3 _ , which is also associated with asthma .
the 10 snps , _
ormdl3 _ expression and asthma status can be illustrated as the @xmath1 , @xmath2 and @xmath3 , respectively , in figure [ fig1 ] .
instead of analyzing snp - expression , expression - asthma and snp - asthma associations separately and univariately , here we are interested in assessing the overall genetic effect of _ ormdl3 _ on the occurrence of childhood asthma , by jointly analyzing snp and gene expression data and accounting for the possibility that the _ ormdl3 _ gene expression might be a causal mediator for the association of the snps in the _ ormdl3 _ gene and asthma risk .
our ultimate goal is to integrate multiple sources of genomic data for genetic association analyses .
is a set of correlated exposures , for example , snp set ; @xmath2 is a mediator , for example , gene expression ; @xmath3 is an outcome , for example , disease ( yes / no ) ; and @xmath4 are covariates , including the true and potential confounders . ] in this paper we propose to jointly model a set of snps within a gene , a gene expression and disease status , where a logistic model is used to model the dependence of disease status on the snp - set and the gene expression , and a linear model is used for the dependence of the gene expression on the snp - set , both adjusting for covariates .
we are primarily interested in testing whether a gene , whose effects are captured by snps and/or gene expression , is associated with a disease phenotype .
we formulate this hypothesis in the causal mediation analysis framework [ @xcite ; @xcite ( @xcite ) ; @xcite ] .
note that the previous work on causal mediation analysis is mostly focused on estimation .
we use the joint model to derive the direct and indirect effects of a snp - set mediated through gene expression on disease risk . for eqtl snps ,
we show that the total effect of a gene on a disease captured by a set of snps and a gene expression corresponds to the total effects of the snp - set , which are the combined direct effects and indirect effects of the snps mediated through the gene expression , on a disease .
this framework allows us to study how the use of gene expression data can enhance power to test for the total effect of a snp - set on disease risk .
we study the impact of model misspecification using the conventional snp - only models when the true model is that both the snps and the gene expression affect the disease outcome . for non - eqtl snps
, the null hypothesis simply corresponds to the joint effects of the snps and the gene expression .
due to a potentially large number of snps within a gene and some of them possibly being highly correlated , that is , in high linkage disequilibrium ( ld ) , conventional tests , such as the likelihood ratio test , for the total effects of multiple snps and a gene expression , do not perform well .
we propose in this paper a variance component test to assess the overall effects of a snp set and a gene expression on disease risk , under the null that the test statistic follows a mixture of @xmath0 distributions , which can be approximated analytically or empirically using a resampling - based perturbation procedure [ @xcite ; @xcite ] .
as the true disease model is often unknown , we construct an omnibus test to improve the power by accommodating different underlying disease models .
the rest of the paper is organized as follows . in section [ sec2 ]
we introduce the joint model for snps , a gene expression and disease as well as the null hypothesis of no joint effect of the snps and the gene expression on a disease phenotype . in section [ sec3 ]
we propose a variance component score test for the total effect of snps and gene expression , and construct an omnibus test to maximize the test power across different underlying disease models . in section [ sec4 ]
we interpret the null hypothesis and study the assumptions within the framework of causal mediation modeling for eqtl snps and non - eqtl snps . in section [ sec5 ]
we evaluate the finite sample performance of the proposed test using simulation studies and show that the omnibus test is robust and performs well in different situations .
in section [ sec6 ] we apply the proposed method to study the overall effect of the _ ormdl3 _ gene contributed by both the snps and the gene expression on the risk of childhood asthma , followed by discussions in section [ sec7 ] .
[ smodel ] the statistical problem is to jointly model the effect of a set of snps and a gene expression on the occurrence of a disease .
assume for subject @xmath5 ( @xmath6 ) an outcome of interest @xmath7 is dichotomous ( e.g. , case / control ) , whose mean is associated with @xmath8 covariates ( @xmath9 , with the first covariate to be 1 , i.e. , the intercept ) , @xmath10 snps in a snp - set ( @xmath11 ) , mrna expression of a gene ( @xmath12 ) and possibly the interactions between the snps and the gene expression as @xmath13 where @xmath14 , and @xmath15 are the regression coefficients for the covariates , the snps , the gene expression , and the interactions of the snps and the gene expression , respectively .
a snp - set and gene expression pair can be defined in multiple ways .
for example , @xmath16 can be the snps in a gene and @xmath2 is the mrna expression of the gene . in the asthma data example , @xmath16 are the 10 typed snps around _ ormdl3 _ and @xmath2 is the expression of _
ormdl3_. alternatively , one can choose the snp - set / expression pair based on the eqtl study : eqtl snp - set / corresponding gene expression . as snps can affect gene expression
[ @xcite ; @xcite ; @xcite ] , for each subject @xmath5 , we next consider a linear model for the continuous gene expression @xmath12 ( i.e. , the mediator ) , which depends on the @xmath8 covariates ( @xmath9 ) and the @xmath10 snps ( @xmath11 ) : @xmath17 where @xmath18 and @xmath19 are the regression coefficients for the covariates and the snps , respectively , and @xmath20 follows a normal distribution with mean @xmath21 and variance @xmath22 . again
, the @xmath10 snps can be the snps within a gene or the eqtl snps corresponding to a gene for which the expression level is measured .
our goal is to test for the total effect of a gene captured by the snps in a set @xmath16 and a gene expression @xmath2 on @xmath3 , which can be written using the regression coefficients in model ( [ ymodel ] ) as @xmath23 note that the null hypothesis ( [ null ] ) only involves the parameters in the @xmath24 $ ] model ( [ ymodel ] ) .
we use the @xmath25 $ ] model in section [ sinterpret ] to facilitate interpretation of the null hypothesis ( [ null ] ) and to study the assumptions within the causal mediation analysis framework . throughout the paper
, we term this null hypothesis as a test for the _ total effect of a gene_. later , in section [ sinterpret ] we will show that it corresponds to the _ total effect of snps _ for eqtl snps and simply to the joint effect of snps and expression for non - eqtl snps .
[ stest ] [ ssvct ] we propose in this section to test for the null hypothesis of no total effect of a gene ( [ null ] ) under model ( [ ymodel ] ) . as
the number of snps ( @xmath10 ) in a gene might be large and some might be highly correlated ( due to linkage disequilibrium ) , the likelihood ratio test ( lrt ) or multivariate wald test for the null hypothesis ( [ null ] ) uses a large degree of freedom ( df ) and has limited power . to overcome this problem , we assume under model ( [ ymodel ] ) the regression coefficients of the individual main snp effects @xmath26 are independent and follow an arbitrary distribution with mean 0 and variance @xmath27 , and the snp - by - expression interaction coefficients @xmath28 ( @xmath29 ) are independent and follow another arbitrary distribution with mean 0 and variance @xmath30 .
the resulting outcome model ( [ ymodel ] ) hence becomes a logistic mixed model .
the problem for testing the null hypothesis ( [ null ] ) becomes a joint test of variance components ( @xmath31 ) and the scalar regression coefficient for the fixed gene expression effect ( @xmath32 ) in the induced logistic mixed models as @xmath33 and @xmath32 .
one can easily show that the scores for @xmath27 , @xmath34 and @xmath30 under the induced logistic mixed models are @xmath35 where @xmath36 , @xmath37 , @xmath38 , @xmath39 and @xmath40 ; @xmath41 and @xmath42 is the mean of @xmath7 under @xmath43 , and @xmath44 is the maximum likelihood estimator of @xmath45 under the null model@xmath46 to combine the three scores to test for the null hypothesis @xmath47 and @xmath32 , one may consider the conventional score statistic @xmath48 , where @xmath49 and @xmath50 is the efficient information matrix of @xmath51 .
however , this approach has several major limitations .
first , notice that the score of the regression coefficient of gene expression @xmath52 is a linear function of @xmath3 , while the scores of the variance components of main effects of snps and snp - by - expression interactions @xmath27 and @xmath30 are quadratic functions of @xmath3 .
hence , @xmath52 has a different scale from @xmath53 .
it follows that @xmath53 are likely to dominate @xmath52 .
a combination of the three scores using @xmath54 is hence not desirable .
second , @xmath54 involves quartic functions of @xmath55 and the information matrix @xmath50 involves the 8th moment of @xmath55 .
hence , calculations of @xmath54 are not stable , and it is difficult to analytically approximate the null distribution of @xmath54 .
we hence propose the following weighted sum of three scores as the test statistic for the null hypothesis ( [ null ] ) : @xmath56\\[-8pt ] & = & n^{-1}({\mathbf{y}}-\bmuhat_0)^t \bigl(a_1\mathbb{ss}^t+a_2 { \mathbf{gg}}^t+a_3\mathbb{cc}^t\bigr ) ( { \mathbf{y}}- \bmuhat_0),\nonumber\end{aligned}\ ] ] where @xmath57 are some weights .
@xmath58 is a nice quadratic function of @xmath59 .
hence , its null distribution can be easily approximated by a mixture of @xmath0 distributions .
different weights can be chosen . with an equal weight @xmath60
, @xmath58 is equivalent to the variance component test for @xmath61 by assuming @xmath62 , @xmath34 and @xmath63 follow a common distribution with mean zero and variance @xmath61 .
however , this common distribution assumption is strong , as @xmath16 , @xmath2 and @xmath64 generally have different scales and so do their effects @xmath65 , @xmath34 and @xmath63 .
notice that @xmath66 are all quadratic functions of @xmath59 in similar forms ; we propose to weight each term @xmath66 using the inverse of the square root of their corresponding variances .
this allows each weighted term to have variance 1 and be comparable .
specifically , the variances for @xmath66 are @xmath67 , @xmath68 , @xmath69 , respectively , where @xmath70 denotes the component - wise multiplication of conformable matrices @xmath71 and @xmath72 , @xmath73 denotes a vector of ones , the diagonal and off - diagonal elements of @xmath74 are @xmath75 and @xmath76[\hat\mu_{0i'}(1-\hat\mu _ { 0i'})]$ ] , respectively [ @xcite ] .
we derive the asymptotic distribution of @xmath58 under the null hypothesis ( [ null ] ) by accounting @xmath77 is a function of @xmath78 , which is the maximum likelihood estimate of @xmath79 under the null model ( [ nullmodel ] ) .
define @xmath80 = n^{-1 } { \mathbf{u}}^t{\mathbf{w}}{\mathbf{u}},\ ] ] where @xmath81 , @xmath82 , @xmath83 , @xmath84 .
we show in section 3 of the supplementary material [ @xcite ] that under the null hypothesis ( [ null ] ) , the score test statistic @xmath58 converges in distribution to @xmath85 , where @xmath86 is a random vector following @xmath87 , @xmath88 ) and @xmath89 is the @xmath90th row of @xmath91 $ ] .
this means under the null hypothesis @xmath58 follows a mixture of @xmath0 distributions , which can be approximated using a scaled @xmath0 distribution by matching the first two moments [ @xcite ] as @xmath92 , where @xmath93 $ ] and @xmath94 ^ 2/\operatorname{var}(q)$ ] , and the expressions of @xmath95 and @xmath96 are given in section 3 of the supplementary material [ @xcite ] .
alternatively , one can approximate the mixture of @xmath0 distribution using the characteristic function inversion method [ @xcite ] .
[ somnibus ] so far we derive the test statistic @xmath58 under the outcome model specified in ( [ ymodel ] ) , which assumes the disease risk of @xmath3 depends on snps , gene expression and their interactions .
denote this @xmath58 by @xmath97 .
suppose that the disease risk of @xmath3 depends on snps and gene expression but not their interactions , or even depends only on snps , then it is more powerful to test for the total snp effect using the test statistics @xmath98 and @xmath99 , respectively . under these simpler disease models ,
the test statistic @xmath97 loses power as it tests for extra unnecessary parameters . on the other hand ,
if the disease risk indeed depends on snps , expression and their interactions , performing tests only using snps @xmath100 or main effects @xmath101 will lose power , compared to @xmath97 .
since in reality we do not know the underlying true disease model , it is difficult to choose a correct model .
it is hence desirable to develop a test that can accommodate different disease models to maximize the power .
moreover , in a genome - wide association study , it is almost impossible that one disease model is true for tens of thousands of genes .
thus , we further propose an omnibus test where we identify the strongest evidence among the three different models with : ( 1 ) only snps , ( 2 ) snps and gene expression , and ( 3 ) snps , gene expression and their interactions .
specifically , we calculate the @xmath10-value under each of the three models , then compute the minimum of the three @xmath10-values and compare the observed minimum @xmath10-value to its null distribution . because of the complicated correlation among @xmath97 , @xmath101 and @xmath100 , it is difficult to analytically derive the null distribution of the minimum @xmath10-value . to this end , we resort to a resampling perturbation procedure . as shown in section [ ssvct ] , @xmath58 converges in distribution to @xmath85 .
the empirical distribution of @xmath102 can be estimated using the resampling method via perturbation [ @xcite ; @xcite ] .
the perturbation method approximates the target distribution of @xmath58 by generating random variables @xmath103 from the estimated asymptotic distribution of @xmath86 .
this perturbation procedure is also called the score - based wild bootstrap [ @xcite ] .
specifically , set @xmath104 , where @xmath105 s are independent @xmath106 . by generating independent @xmath107 repeatedly
, the distribution of @xmath103 conditional on the observed data is asymptotically the same as that of @xmath86 .
denoted by @xmath108 , where @xmath109 is the number of perturbations , it follows that the empirical distribution of the @xmath110 is the same as that of @xmath102 asymptotically .
the @xmath10-value can be approximated using the tail probability by comparing @xmath111 with the observed @xmath58 .
hence , one can calculate the @xmath10-values of @xmath97 , @xmath101 and @xmath112 by setting @xmath113 , @xmath114 and @xmath115 and generate their perturbed realizations of the null counterpart as @xmath116 , @xmath117 and @xmath118 , and compare them with corresponding observed values , respectively .
note that for each perturbation @xmath119 , the random normal perturbation variable @xmath120 is the same across the three tests such that the correlation among @xmath97 , @xmath101 and @xmath100 can be preserved .
the @xmath10-value of the omnibus test can be easily calculated using the perturbation method .
let @xmath121 , @xmath122 and @xmath123 be the three @xmath10-values using the three statistics , where @xmath124 , @xmath125 and @xmath126 .
the null distribution of the minimum @xmath10-value , @xmath127 , can be approximated by @xmath128 , @xmath129 ( @xmath130 . the @xmath10-value of the omnibus test hence can be calculated by comparing the observed minimum @xmath10-value @xmath131 with its empirical null distribution @xmath132 .
note that different from permutation where the observed data are shuffled and resampled to calculate the test statistic @xmath58 , the perturbation procedure resamples from the asymptotic null distribution of @xmath58 without recalculating @xmath58 using the shuffled data .
thus , it is much more efficient than the permutation method .
using a cpu ( 2.53 ghz ) to run 100 genes ( each with 10 snps ) and 100 cases/100 controls , the computation time is 134.10 and 809.76 seconds for the perturbation and permutation methods ( both with 200 resampling ) , respectively .
furthermore , covariates can be easily adjusted using the perturbation method , but adjustment is more difficult using permutation .
specifically , the permutation - based @xmath10-values calculated by simply permuting snps and gene expression fail if snps / gene expression are correlated with covariates .
[ sinterpret ] [ scmm ] to understand the null hypothesis of no total effect of a gene captured by snps in a gene and a gene expression and the underlying assumptions , we discuss in this section how to interpret the null in ( [ null ] ) within the causal mediation analysis framework .
causal interpretation can be helpful for understanding genetic etiology of diseases as well as for applications in pharmaceutical research [ @xcite ] .
although genotype is essentially fixed at , it is at that time effectively randomized , conditional on parental genotypes , and could be viewed as subject to have hypothetical intervention .
the statistical problem of jointly modeling a set of snps , a gene expression and a disease can be presented as a causal diagram [ @xcite ; @xcite ] in [ fig1 ] and be framed using a causal mediation model [ @xcite ( @xcite ) ; @xcite ] based on counterfactuals [ @xcite ( @xcite ) ] .
@xcite and @xcite have used the causal mediation analysis for epidemiological and social science studies , respectively , where the exposure of interest is univariate .
one can decompose the _ total effect _ ( te ) of snps into the _ direct effect _ ( de ) and the _ indirect effect _ ( ie ) .
the _ direct effect _ of snps is the effect of the snps on the disease outcome that is not through gene expression , whereas the _ indirect effect _ of the snps is the effect of the snps on the disease outcome that is through the gene expression . within the causal mediation analysis framework
, we derive in the supplementary material the te , de and ie of the snps on the disease outcome [ @xcite ] .
specifically , we define the total effect ( te ) of snps as @xmath133 that is , the equation ( [ ymodel ] ) marginalizes over gene expression @xmath2 . in section 1 of the supplementary material [ @xcite ] , we show that for rare diseases , the total effect of the snps on the log odds ratio ( or ) of disease risk can be expressed in terms of the regression coefficients in models ( [ ymodel ] ) and ( [ gmodel ] ) and is approximately equal to @xmath134\\[-8pt ] & & { } + \tfrac{1}{2}\sigma_g^2({\mathbf{s}}_1 + { \mathbf{s}}_0)^t{\bolds{\gamma}}({\mathbf{s}}_1-{\mathbf{s}}_0)^t{\bolds{\gamma}}.\nonumber\end{aligned}\ ] ] we can express the de and ie of the snps on the log odds ratio of disease risk in terms of the regression coefficients in models ( [ ymodel ] ) and ( [ gmodel ] ) . for rare diseases , they are , respectively , approximately equal to @xmath135 \nonumber\\[-8pt]\label{de}\\[-8pt ] & & { } + \tfrac{1}{2}\sigma_g^2({\mathbf{s}}_1 + { \mathbf{s}}_0)^t{\bolds{\gamma}}({\mathbf{s}}_1- { \mathbf{s}}_0)^t{\bolds{\gamma}},\nonumber \\ \mathrm{ie}&=&({\mathbf{s}}_1-{\mathbf{s}}_0)^t{\bolds { \delta}}\bigl(\beta_g+{\mathbf{s}}_1^t{\bolds{\gamma } } \bigr ) .
\label{ie}\end{aligned}\ ] ] these are derived in section 1 of the supplementary materials using counterfactuals under the assumptions of no unmeasured confounding [ @xcite ] .
the sum of the direct and indirect effects is equal to the total effect of the snps , that is , @xmath136 . as shown in the supplementary material [ @xcite ] and discussed in section 3.2 , identification of the total effect
requires a much weaker assumption than those required for the direct and indirect effects .
[ ssnull ] under the assumption that the gene expression @xmath2 is associated with the snps @xmath16 ( i.e. , eqtl snps ) , that is , @xmath137 , using equations ( [ de ] ) and ( [ ie ] ) , the test for the joint effects of snps in a snp set @xmath16 and a gene expression @xmath2 on @xmath3 , that is , the total effect of a gene , is equivalent to a test for the total snp effect on the outcome ( @xmath3 ) .
in fact , for eqtl snps , which have nonzero effects on expression @xmath2 ( i.e. , @xmath137 ) , using the expressions of de and ie in ( [ de ] ) and ( [ ie ] ) , one can show that the null hypothesis of no direct and indirect genetic ( snp ) effects is equivalent to the null hypothesis ( [ null ] ) that all the regression coefficients ( @xmath138 , @xmath34 and @xmath139 ) equal zero : @xmath140 the null hypothesis ( [ null ] ) that all the regression coefficients ( @xmath138 , @xmath34 and @xmath139 ) are equal to zero is also equivalent to the null hypothesis of no total effect of the snps provided @xmath141 if @xmath138 or @xmath34 is not 0 for eqtl snps ( @xmath142 , that is , @xmath143 we show in section 1.4 of the supplementary material that the null hypothesis ( [ null1 ] ) requires only the assumption of no unmeasured confounding for the effect of eqtl snps ( @xmath16 ) on the outcome ( @xmath3 ) after adjusting for the covariates ( @xmath144 ) [ @xcite ] .
most genetic association studies make this assumption . in other words
, we make no stronger assumption than standard snp only analyses for testing the null hypothesis of no total effect of the snp set in a gene
. note that in models ( [ ymodel ] ) and ( [ gmodel ] ) we allow other covariates ( @xmath4 ) to affect both the gene expression and the disease .
if the covariates @xmath4 affect both expression and disease , ignoring @xmath4 may cause confounding in estimating de and ie . as shown in figure [ fig1
] , if arrows from @xmath4 to @xmath2 and @xmath3 exist and @xmath4 is not controlled for , assumption ( 2 ) in section 1.2 of the supplementary material is violated [ @xcite ] .
but if the covariates @xmath4 , the common causes of expression and disease , do not affect the snps @xmath1 ( no arrow from @xmath4 to @xmath1 ) , the estimation and hypothesis testing for te is still valid . however , if there does exist an effect of @xmath4 on @xmath1 , then it violates the above assumption of no unmeasured confounding for the @xmath16@xmath3 association and , thus , the test or estimation for te will be biased .
if snps have no effect on gene expression ( @xmath145 ) , that is , they are not eqtl snps , then there is no indirect effect of the snps on @xmath3 , so that the null hypothesis of no total effect of a gene ( @xmath146 ) is not equivalent to testing for no total snp effect on @xmath3 ( @xmath147 ) . in this case
, what the null hypothesis , @xmath146 , tries to evaluate is simply whether there exists a joint effect of the given set of snps @xmath16 and the given gene expression @xmath2 , and possibly their interactive effect , on disease risk .
to test for such a joint effect , we need the first two assumptions regarding no unmeasured confounding in section 1.2 of the supplementary material : no unmeasured confounding of the snps on the outcome and no unmeasured confounding of the gene expression on the outcome [ @xcite ] .
[ ssgwas ] in standard genetic association analysis , we usually fit the following snp only model : @xmath148 which does not take gene expression into account , but simply considers the association between the outcome and snps adjusting for covariates .
note for the special case where snp , @xmath149 , is univariate , the model ( [ gwas ] ) corresponds to single snp analysis , the most common approach in gwas .
@xcite and @xcite have developed tests for a snp - set for @xmath150 under ( [ gwas ] ) , which can be more powerful than individual snp tests for the association between the joint effects of the snps in a gene and the outcome by borrowing information across snps within a gene , especially when the snps are in good linkage disequilibrium ( ld ) . assuming the true models that depend on both snps and a gene expression are specified in ( [ ymodel ] ) and ( [ gmodel ] )
, we study in this section how @xmath151 in the misspecified standard snp only model ( [ gwas ] ) is related to the regression parameters @xmath152 and @xmath139 in the true model ( [ ymodel ] ) and what the null hypothesis @xmath153 under ( [ gwas ] ) tests for . to focus on the fundamental issues and for simplicity
, we first discuss the case of no interaction effect between snps and gene expression on disease risk , that is , @xmath154 in model ( [ ymodel ] ) . under
the true @xmath155 $ ] and @xmath25 $ ] models in ( [ ymodel ] ) and ( [ gmodel ] ) assuming no @xmath156 interaction ( @xmath154 ) , by plugging ( [ gmodel ] ) into ( [ ymodel ] ) , the true @xmath155 $ ] model can be rewritten as @xmath157 . integrating out @xmath158
, we have the true @xmath159 $ ] model as @xmath160\approx c \bigl\{{\mathbf{x}}_i^t ( { \bolds{\alpha}}+ \beta_g{\bolds{\phi}})+{\mathbf{s}}_i^t ( { \bolds{\beta}}_s+\beta_g{\bolds{\delta } } ) \bigr\ } , \label{trueysmodel}\ ] ] where @xmath161 [ @xcite ] .
a comparison of ( [ gwas ] ) with ( [ trueysmodel ] ) shows that @xmath162 and that the effect of @xmath163 versus @xmath164 on the outcome @xmath3 under the snp only model ( [ gwas ] ) corresponds to @xmath165 , which is proportional to the total effect of snps in ( [ te ] ) when @xmath154 .
it follows that testing for @xmath166 in the snp only model ( [ gwas ] ) is approximately equivalent to testing for no total effect of the snps .
however , if there exists a snp - by - expression interaction on @xmath3 and the snps are eqtl snps , the naive snp only analysis using ( [ gwas ] ) does not provide obvious correspondence to the total snp effect . as shown in section 2 of the supplementary material [ @xcite ] , the induced true @xmath159 $ ] model in this setting follows @xmath167\\[-8pt ] & & \qquad \approx c_{i}^ * \bigl\{{\mathbf{x}}_i^t ( { \bolds{\alpha}}+{\bolds{\phi } } \beta_g)+{\mathbf{s}}_i^t ( { \bolds{\beta}}_s+{\bolds{\delta}}\beta_g ) + { \mathbf{x}}_i^t { \bolds{\phi}}{\mathbf{s}}_i^t{\bolds{\gamma}}+{\mathbf{s}}_i^t { \bolds{\delta}}{\mathbf{s}}_i^t{\bolds{\gamma } } \bigr\ } , \nonumber\end{aligned}\ ] ] where @xmath168 .
this implies that if the @xmath169 $ ] follows the interaction model ( [ ymodel ] ) , the induced true @xmath170 $ ] model depends not only on the linear terms of @xmath4 and @xmath1 but also on the cross - product terms of @xmath4 and @xmath1 and the second order term of @xmath1 .
a comparison of ( [ gwas ] ) with ( [ trueysint ] ) shows that the standard snp only model ( [ gwas ] ) misspecifies the functional form of the true @xmath170 $ ] . the test for @xmath171 under the misspecified snp only model ( [ gwas ] )
will still be valid for testing the total effects of snps , because under the null the two models are the same .
however , the misspecified model is subject to power loss , compared to the test based on the correctly specified model . with only an interaction effect ( @xmath172 , @xmath173 , @xmath32 ) , ( [ trueysint ] ) can be written as @xmath174 \approx c_{i}^ * \{{\mathbf{x}}_i^t{\bolds{\alpha } } + { \mathbf{x}}_i^t{\bolds{\phi}}{\mathbf{s}}_i^t{\bolds{\gamma}}+{\mathbf{s}}_i^t{\bolds{\delta}}{\mathbf{s}}_i^t{\bolds{\gamma } } \ } , $ ] where @xmath175 . if we assume this is the true model and fit the conventional gwas model ( [ gwas ] ) to test for the snp effect , the test is again still valid under the null , but loses power under the alternative .
the approach here differs in several ways from that based on mendelian randomization [ smith and ebrahim ( @xcite and @xcite ) ] in which genetic markers ( snps ) are instrumental variables to assess the effect of an exposure ( in our case , a gene expression value ) on an outcome .
here we are interested in using a gene expression to increase power for testing for the total effect of snps on a disease outcome .
furthermore , mendelian randomization makes the assumption that snps do not have an effect on an outcome except through an exposure ( e.g. , gene expression in our case ) , in other words , no direct effect .
no such assumption is being made here .
this is because we are interested in testing for a different effect , that is , the effect of snps , rather than the effect of an exposure ( gene expression ) on disease risk .
[ ssimulation ] to make the simulation mimic the motivating asthma data [ @xcite ] , we simulated data using the _ ormdl3 _ gene on chromosome 17q21 .
we generated the snp data in the _ ormdl3 _ gene by accounting for its linkage disequilibrium structure using hapgen based on the ceu sample [ @xcite ] .
the genomic location used to generate the snp data is between 35.22 and 35.39 mb on chromosome 17 , which contains 99 hapmap snps .
ten of the 99 hapmap snps are genotyped on the illumina humanhap300 array , that is , 10 typed snps .
to generate gene expression and the disease outcome , we assumed there is one causal snp @xmath176 and varied the causal snp among the 99 hapmap snps in each simulation . in section [ sec5.4 ]
we further perform a simulation study assuming three causal snps . for subject @xmath5 , gene expression @xmath12
was generated by the linear regression model @xmath177 , @xmath178 .
the outcome @xmath7 was generated by the logistic model @xmath179 the parameters and the range of @xmath180 , @xmath34 and @xmath181 were based on the empirical estimates from analysis of the asthma data . for each simulation
, we first generated a cohort with 1000 subjects , and 100 cases and 100 controls were randomly selected from the 1000 subjects to form a case control sample .
two sets of simulations were performed . in the first set , we selected the snp rs8067378 as the causal snp , as this snp
is highly associated with asthma in the original gwas [ @xcite ] . for each configuration of @xmath180 , @xmath34 , @xmath181 and @xmath182
, we generated 2000 data sets to calculate the empirical size and power .
in the second set of simulation , the causal snp was chosen one at a time out of the 99 hapmap snps .
for each selected causal snp , we generated 1000 data sets and evaluated statistical power for two different disease models : @xmath183 or @xmath184 . in both simulation settings
, we used the 10 typed snps of the gene on the illumina chip to form the snp - set for the model ( [ ymodel ] ) , that is , @xmath185 , in calculating the test statistics @xmath100 , @xmath101 , @xmath97 and the omnibus test . for @xmath101 and @xmath97 , both weighted and unweighted methods were investigated , where @xmath186 , @xmath187 and @xmath188 for the weighted statistic , and @xmath189 for the unweighted statistic .
the @xmath10-values were calculated using the scaled @xmath0 approximation , the davies method by inverting the characteristic function [ @xcite ] and the perturbation procedure with 500 perturbations .
the results of these approximations were very similar at the significance level of 0.05 .
we performed the omnibus test by combining the evidence from @xmath100 , weighted @xmath101 and weighted @xmath97 .
[ ssinglesnp ] we first evaluated the sizes of the proposed score tests , where the null distribution was approximated by either the scaled @xmath0 approximation or the perturbation procedure ( table [ tab1 ] ) .
type i errors are well protected using both approximation methods under the three models with statistics @xmath190 . the empirical size is close to 0.05 for the omnibus test and the three models . as the results using different approximation methods are similar at the level of 0.05 , we only present in the following the empirical power using the perturbation method .
we also evaluate the performance of the proposed tests using the characteristic function inversion method [ @xcite ] and the perturbation method at smaller sizes ( @xmath191 and @xmath192 ) ( table [ tab1 ] of the supplementary material [ @xcite ] ) , and find the methods perform well .
@@lcccc@ & & + & & + & * unweighted * & * weighted * & * unweighted * & * weighted * + + snps & & + snps and expression & 4.80 & 4.95 & 4.80 & 4.80 + snps , expression and interaction & 4.75 & 4.60 & 4.60 & 4.35 + omnibus test & & 5.15 + + snps & & + snps and expression & 4.87 & 4.90 & 4.83 & 4.87 + snps , expression and interaction & 4.60 & 4.80 & 4.77 & 5.07 + omnibus test & & 5.13 + [ cols="^,^ " , ] we performed additional simulations to assess how model misspecification influences our proposed test .
gene expression @xmath12 was generated without the normality assumption @xmath193 , @xmath194 .
two outcome models are explored .
the first model generates the outcome @xmath7 by the logistic model assuming nonlinear effects of snps and @xmath2 as @xmath195 , and the second model generates @xmath7 by a probit model @xmath196 .
although the model is not correctly specified in our proposed test under these settings , the joint analyses of snps and expression still outperform the snp - only analyses when the gene expression contributes to the risk of developing disease .
similarly , the performance of the omnibus test is very close to the optimal test obtained under the true model for different scenarios ( figure [ fig4 ] ) .
we conducted two additional simulation studies by varying the number of causal variants and ld structures .
the pattern of the results from these additional studies is very similar to what is presented above ( figures [ fig1 ] and [ fig2 ] in the supplementary material [ @xcite ] ) .
the first additional study is similar to the study in section [ ssinglesnp ] , except that there are three causal snps in the _ ormdl3 _ gene instead of a single causal snp . using the same ten typed snps for the analyses
, we again found that the test performs the best when the model is correctly specified and the omnibus test approaches the optimal test obtained under the true model with limited power loss ( figure [ fig1 ] of the supplementary material [ @xcite ] ) .
similar to analyses in section [ smultiplesnp ] , the second additional study investigates the performance of the proposed test at 15q2415q25.1 , where snps have a different ld pattern from the _
ormdl3 _ gene .
assuming one causal snp at a time , we used the same ten typed snps to perform our proposed test .
again , the test performs the best when the model is correctly specified , and the omnibus test is robust and approaches the optimal test obtained by assuming the true model , and the power depends on the correlation of the causal snp and the typed snps ( figure [ fig2 ] of the supplementary material [ @xcite ] ) .
[ sdataanalysis ] we applied the proposed testing procedures to reinvestigate the genetic effects of the _ ormdl3 _ gene on the risk of childhood asthma in the mrc - a data [ @xcite ; @xcite ] .
this subset of the data contained 108 asthma cases and 50 controls where we have complete data of the 10 typed snps and gene expression of _ ormdl3_. the snp data were genotyped using the illumina 300k chip and the gene expression was collected using the affymetrix hu133a 2.0 .
we analyzed the data using both additive and dominant modes : in the additive mode , the genotype was coded as the number of the minor allele ( i.e. , 0 , 1 , 2 ) , whereas in the dominant mode , the genotype was coded as whether or not the minor allele was present ( i.e. , 0 , 1 ) .
@@lccccc@ & * multivariate * & * bonferroni- * & * permutation- * & * vct * & * vct * + & * wald * & * adjusted @xmath197 * & * adjusted @xmath197 * & * unweighted * & * weighted * + + snps & 0.122 & 0.102 & 0.039 & + snps , gene & 0.342 & 0.194 & 0.057 & 0.039 & 0.033 + snps , gene and interaction & 0.013 & 0.303 & 0.093 & 0.0025 & 0.0028 + omnibus test & & + + snps & 0.018 & 0.015 & 0.018 & + snps , gene & 0.094 & 0.031 & 0.018 & 0.0040 & 0.0040 + snps , gene and interaction & 0.131 & 0.098 & 0.048 & 0.0035 & 0.0023 + omnibus test & & + we applied the proposed tests for the total snp effect of _
ormdl3 _ using the snp and gene expression data .
there are strong associations between the snps and the gene expression ( 8 out of 10 with @[email protected] and the other two with @xmath10-values of 0.076 and 0.21 ) , that is , the snps are eqtl snps .
we considered six test statistics : @xmath100 , unweighted @xmath101 , weighted @xmath101 , unweighted @xmath97 , weighted @xmath97 , and omnibus test ( table [ tab2 ] ) .
we compared our methods with the standard multivariate or univariate methods : the multivariate wald test , which has 10 , 11 and 21 degrees of freedom under the three models ( snps only , snps and gene expression , and snps , gene expression and interactions ) .
we also included in the comparison the test using the smallest @xmath10-value from the 10 single snp analyses with the bonferroni adjustment or the adjustment using the permutation procedure to account for the correlation among the snps .
the results in table [ tab2 ] show that our proposed methods give smaller @xmath10-values compared to the standard testing procedures .
the test @xmath97 , which accounts for the effects of snps , gene expression and their interactions , gives the smallest value compared to the tests only using snps , in both additive and dominant modes .
for example , the @xmath10-values using weighted @xmath97 and @xmath100 are 0.0028 and 0.044 , respectively , using the additive snp model .
the omnibus test calculated using the perturbation procedure by computing the minimum @xmath10-value from @xmath100 , weighted @xmath101 and weighted @xmath97 also provides a more significant signal than those only considering snps , with the @xmath10-value being 0.0055 .
these results are consistent with the findings in simulation studies .
we also performed genome - wide analyses for both snp - sets and single snp .
we first paired eqtl snps with their corresponding gene expression [ @xcite ] and performed snp - only analyses and our proposed method .
for single snp analyses , after adjustment of multiple comparison using false discovery rate [ fdr ; @xcite ] , 56 snps with [email protected] were identified in snp - only analyses and 97 snps were identified from the proposed omnibus test . for snp - set analyses
, we grouped eqtl snps that correspond to the same gene as a snp - set , and we identified 5 and 15 snp - sets ( [email protected] ) from snp - only analyses and omnibus tests , respectively .
[ sdiscussion ] we proposed in this paper to integrate snp and gene expression data to improve power for genetic association studies .
the major contributions of this paper are as follows : ( 1 ) to formulate the data integration problem of different types of genomic data as a mediation problem ; ( 2 ) to propose a powerful and robust testing procedure for the total effect of a gene contributed by snps and a gene expression ; and ( 3 ) to relax the assumptions required for mediation analyses in the test for the total effect .
specifically , as shown in figure [ fig1 ] , we are able to integrate the information of snps and gene expression as a biological process through the mediation model .
our proposed variance component score test for the total effect of snps and a gene expression circumvents the instability of estimation of the joint effects of multiple snps and gene expression , because only the null model needs to be fit .
mediation analysis to estimate direct and indirect effects generally requires additional unmeasured confounding assumptions , and previous work mainly focused on estimation .
here we focus on testing for the total effect of a gene using snps and a gene expression . for eqtl snps , we show that the total effect of a gene contributed by snps and a gene expression is equivalent to the total effect of snps , which is the sum of direct and indirect effects of snps mediated through gene expression . testing for the total snp effect only requires one assumption : no unmeasured confounding for the effect of snps on the outcome , which is the same assumption as the standard gwas and , thus , no stronger assumption is required .
we characterize the relation among snps , gene expression and disease risk in the framework of causal mediation modeling .
this framework allows us to understand the null hypothesis of no total effect of a gene contributed by snps and gene expression , and the underlying assumptions of the test for both eqtl snps and non - eqtl snps .
we propose a variance component score test for the total effects of a gene on disease .
this test allows to jointly test for the effects of snps , gene expression and their interactions .
we showed that the proposed test statistic follows a mixture of @xmath0 distributions asymptotically , and proposed to approximate its finite sample distribution using a scaled @xmath0 distribution , a characteristic function inversion method or a perturbation method .
we considered three tests : using only snps ( @xmath100 ) , snps and gene expression main effects ( @xmath101 ) , and snps , gene expression and their interactions ( @xmath97 ) .
our simulation study shows that all three tests have the correct type i error for testing for the overall snp effect .
the relative power of these tests depends on the underlying true relation between the predictors ( snps , gene expression and their interactions ) and disease .
as the underlying biology is often unknown , we further constructed the omnibus test that identifies the most powerful test among the three disease models , and proposed to use the perturbation method to calculate the value for the omnibus test .
further , the test using only the main effects of snps and gene expression loses limited power compared to the omnibus test and can be used as a simple alternative .
our results also show that to test for the total effects of a gene , the tests that incorporate both snp and gene expression information , such as @xmath101 and @xmath97 , are more powerful if snps are associated with gene expression than if they are not .
in other words , it is even more beneficial to incorporate gene expression data with snp data to detect genetic effects on disease if gene expression is a good causal mediator for the snps . to achieve this ,
a natural way is to select snps located within or at the neighborhood of a gene , since it has been well destablished that the snp within a gene can alter its expression value via transcription regulation [ @xcite ] .
alternatively , one can restrict the joint snp - expression analysis to known eqtl snps .
if selection of eqtl snps is based on statistical significance , one also needs to be aware of the possibility that cis - action may be a confounding effect of snps on array hybridization [ @xcite ] .
we mainly focus on testing for the total effect of a gene in this paper .
the proposed method can be easily extended to test for direct and indirect effects separately for eqtl snps .
using equation ( [ de ] ) , to test for the direct effect of the snps , one can test @xmath199 .
using the notation in equation ( [ q - stat ] ) , one can test this null hypothesis using the statistic @xmath200 , where the null model is a logistic model with @xmath201 and @xmath202 . to test for the indirect effects of the snps , using equation ( [ ie ] )
, one can test @xmath203 . using the notation in equation ( [ q - stat ] )
, one can test this null hypothesis using the statistic @xmath204 , where the null model is a logistic model with @xmath201 and @xmath205 .
as the number of snps @xmath205 might be large and some snps might be highly correlated ( with high ld ) , standard regression to fit the null model might not work well .
one can fit the null model using ridge regression . to perform these tests
, one will need to make the four unmeasured confounding assumptions required for estimating direct and indirect effects of snps stated in section 1.2 of the supplementary material [ @xcite ] .
gene expression may not be the only mediator for the relation between snps and disease .
other biomarkers , such as dna methylation , proteins , metabolites of the gene product in the blood , immunological or biochemical markers in the serum , and environmental factors can also serve as potential mediators , depending on the context or the disease to be studied .
for instance , epigenetic variations have been reported to exert heritable phenotypic effects [ @xcite ] .
furthermore , our proposed test can be applied to address many other scientific questions as long as there exists a causal relationship as illustrated in figure [ fig1 ] .
for example , the snp - gene - disease relations can be replaced by the dna copy number - protein - cancer stage ( early vs. late ) in tumor genomics studies to assess if copy number can have any effect on the clinical stage of cancer .
it is advantageous to set up a biologically meaningful model before applying our proposed test procedure , which makes the best use of the prior knowledge .
the authors would like to thank the editor , the associate editor and the referees for their helpful comments that have improved the paper . | genetic association studies have been a popular approach for assessing the association between common single nucleotide polymorphisms ( snps ) and complex diseases . however ,
other genomic data involved in the mechanism from snps to disease , for example , gene expressions , are usually neglected in these association studies . in this paper
, we propose to exploit gene expression information to more powerfully test the association between snps and diseases by jointly modeling the relations among snps , gene expressions and diseases .
we propose a variance component test for the total effect of snps and a gene expression on disease risk .
we cast the test within the causal mediation analysis framework with the gene expression as a potential mediator .
for eqtl snps , the use of gene expression information can enhance power to test for the total effect of a snp - set , which is the combined direct and indirect effects of the snps mediated through the gene expression , on disease risk .
we show that the test statistic under the null hypothesis follows a mixture of @xmath0 distributions , which can be evaluated analytically or empirically using the resampling - based perturbation method .
we construct tests for each of three disease models that are determined by snps only , snps and gene expression , or include also their interactions . as the true disease model is unknown in practice , we further propose an omnibus test to accommodate different underlying disease models .
we evaluate the finite sample performance of the proposed methods using simulation studies , and show that our proposed test performs well and the omnibus test can almost reach the optimal power where the disease model is known and correctly specified . we apply our method to reanalyze the overall effect of the snp - set and expression of the _ ormdl3 _
gene on the risk of asthma . , |
semiconductor lasers are enabling components in multiple platform applications spanning optical communication networks to laser surgery and sensing .
recent works include impressive advances in high - speed lasers with low power consumption , high - power vertical external cavity surface emitting lasers and high - speed beam steering with phased vertical cavity laser arrays .
significant advances have been made in nitride based lasers , record - high temperature operation quantum dot lasers , and the field of nanolasers with ultralow volume and threshold is coming to technological maturity @xcite .
of special importance for next generation applications such as laser radars , is the design of photonically integrated semiconductor laser arrays that consist of a very large number of properly coupled photonic emitters @xcite .
it is well known that phase locking of an array of diode lasers is a highly effective method in beam shaping because it increases the output power and reduces the overall needed lasing threshold .
recent work on phase - locked laser arrays through global antenna mutual coupling has employed custom made nano - lasers @xcite .
moreover , reconfigurable semiconductor laser networks based on diffractive coupling using talbot geometry have been studied on commercially available vertical cavity diode lasers @xcite . in the present work ,
we are interested in the collective behavior of a large array of semiconductor lasers with nearest - neighbor coupling .
the crucial parameters for the observed dynamics are the coupling strength and the relative optical frequency detuning between the lasers , which introduces realistic inhomogeneities into the system .
our focus , in particular , is to identify the parameter regions where chimera states emerge and subsequently characterize these states using suitable classification measures @xcite .
chimera states were first reported for identical and symmetrically coupled phase oscillators in @xcite . for over a decade now , a number of works has been dedicated to this phenomenon of coexisting synchronous and asynchronous oscillatory behavior ( see @xcite and references within ) .
the latest developments in this field involve their study in physical , higher - dimensional systems beyond phase oscillators , their experimental verification @xcite , their robustness against system inhomogeneities @xcite , their existence in stochastic systems @xcite , and their manipulation through control techniques @xcite .
coupled lasers have been extensively studied in terms of nonlinear dynamics @xcite and @xcite ( with references therein ) and synchronization phenomena @xcite , but works on chimera states in laser networks have appeared only recently . in @xcite chimera states
were reported both theoretically and experimentally in a virtual space - time representation of a single laser system subject to long delayed feedback .
furthermore , so - called `` small chimeras '' were numerically observed in a network of four globally delay - coupled lasers in @xcite , for both small and large delays .
such chimeras exist for very small network sizes and do not require nonlocal coupling in order to emerge .
in our study we use neither nonlocal , nor global coupling but simple nearest - neighbor interactions which is physically plausible for lasers , e. g. , grown on a single chip .
this coupling realization is less expensive computationally .
moreover , it revises the general belief that nonlocal coupling is essential for the existence of chimeras @xcite .
we will show that the crucial parameter for the collective behavior in our system is the frequency detuning between the coupled lasers .
the effect of detuning has been examined before in @xcite but with respect to in- and anti - phase synchronization .
moreover , transitions from complete to partial synchronization ( optical turbulence ) were explained , for a small array of three lasers . here
, we address the emergence of the hybrid phenomenon of chimera states in a _
large _ laser array and provide a quantification of these patterns using newly developed classification measures @xcite .
the work is organized as follows : in sec .
[ model ] we present the model equations and the bifurcation diagrams for two coupled lasers with the coupling strength and the detuning as control parameters . in sec .
[ sec_3 ] the collective behavior for a laser array without detuning is discussed , whereas in sec . [ yes_detuning ] various patterns including chimera states
are presented and classified in the presence of detuning . moreover , the effect of the system size and the initial conditions on the observed dynamics is addressed . in the concluding section
we summarize our results and discuss open problems .
our system consists of an array of @xmath0 locally coupled semiconductor lasers .
the evolution of the slowly varying complex amplitudes @xmath1 of the electric fields and the corresponding population inversions @xmath2 are given by @xcite : @xmath3 the amplitude - phase coupling is modeled by the linewidth enhancement factor @xmath4 , @xmath5 is the ratio of the lifetime of the electrons in the upper level and that of the phonons in the laser cavity .
the normalized angular frequency @xmath6 measures the optical frequency detuning of laser @xmath7 from a common reference .
the lasers are pumped electrically with the excess pump rate @xmath8 which is @xmath9 above laser threshold .
the coupling strength @xmath10 is a control parameter used to tune the dynamics of the system . by using polar coordinates @xmath11 and separating real from imaginary part , we get : @xmath12 \nonumber \\
\frac{d\phi_{i}}{dt}&=&-ian_{i}-\eta[\frac{e_{i+1}}{e_{i}}\cos(\delta\phi_{i+1}+\delta\omega_{i+1}t ) + \frac{e_{i-1}}{e_{i}}\cos(\delta\phi_{i-1}+\delta\omega_{i-1}t ) ] \nonumber \\ t\frac{dn_{i}}{dt}&=&p - n_{i}-(1 + 2n_{i}){e_{i}}^2 \label{eq2}\end{aligned}\ ] ] where @xmath13 , @xmath14 , @xmath15 , @xmath16 . for the special case of _ two lasers
_ and in the absence of detuning , eqs .
( [ eq2 ] ) have the following fixed points : @xmath17 @xmath18 to investigate the stability of these steady states we introduce small perturbations and linearize eqs .
( [ eq2 ] ) about their steady - state values @xcite .
the routh - hurwitz criterion is used to determine the parameter value regions in which the steady - state solutions are stable .
after some calculations we find that the fixed point of eq .
( [ eq4 ] ) is stable under the condition : @xmath19 and the fixed point of eq .
( [ eq3 ] ) is stable for : @xmath20 has been rescaled to the relaxation oscillation frequency @xmath21 .
blue color refers to the first laser and red to the second one .
other parameters are : @xmath22.[fig_1 ] ] figure [ fig_1 ] depicts a numerically obtained bifurcation diagram of the maxima and minima of the amplitude of the oscillating electric field .
the control parameter is the coupling strength normalized to the relaxation oscillation frequency @xmath23 @xcite .
a hopf bifurcation occurs at @xmath24 . as the coupling
is increased the limit cycle exists until @xmath25 .
after that , a period - doubling cascade takes place , leading to chaos .
the system remains chaotic until the approximate vale of @xmath26 and then enters a new limit cycle which is stable up to @xmath27 , which is followed by a new period doubling cascade into a second chaotic region . .
the difference in the detuning @xmath28 has been rescaled by the relaxation oscillation frequency @xmath21 .
other parameters : @xmath29 and @xmath4 .
[ fig_2 ] ] apart from the coupling strength , another crucial parameter is the optical frequency detuning and its correlation with the amplitude instability and mutual coherence of the light emitted by the laser . for both solid state @xcite and semiconductor lasers @xcite , the complexity of the system increases immensely by introducing detuning .
we have found that the most relevant parameter is actually the _ difference _ between the laser detunings rather than their individual values .
the bifurcation diagram of fig .
[ fig_2 ] shows the maxima and minima of the electric field amplitude in dependence of @xmath30 , rescaled by the free relaxation frequency @xmath21 .
this has been repeated for various values of the coupling strength ( figs .
[ fig_2](a - d ) ) . we observe that in a certain range of @xmath31 values the amplitude of the laser oscillations increases significantly .
moreover , for large coupling strengths ( figs .
[ fig_2](c , d ) ) the behavior of the system is rich and complex in dynamical responses .
it is also noticeable that although some @xmath10 values render the system chaotic in the case without detuning ( see fig . [ fig_1 ] ) , for the same coupling strengths the dynamics is regular in the presence of detuning ( figs
. [ fig_2](d ) ) .
the situation is much more complicated when we consider larger arrays . in the case of @xmath0 coupled lasers
, it can be found that the critical coupling strength , for the special case of the anti - phase region ( see eq .
( [ eq6 ] ) for two coupled lasers ) changes to @xcite : @xmath32 as @xmath0 increases , the critical coupling decreases roughly as @xmath33 and reaches a limiting value at large @xmath34 which is half of that corresponding to @xmath35 . throughout this work
, we will consider an array of @xmath36 lasers .
the numerical integration has been done by using the fourth order runge - kutta algorithm . for faster numerical calculations we can rescale eq .
( [ eq2 ] ) with the laser relaxation oscillations frequency .
in this section we focus on the influence of the coupling strength on the collective behavior , in the absence of detuning . in fig .
[ fig_3 ] snapshots of the amplitude of the electric field are shown at @xmath37 , where @xmath38 is the period of the relaxation oscillation of the free running diode laser .
lasers for different coupling strengths without detuning : @xmath39 .
other parameters : @xmath29 and @xmath4 .
[ fig_3 ] ] according to eq .
( [ eq7 ] ) , the hopf bifurcation for our laser array occurs at the value @xmath40 . slightly above this value ,
the system demonstrates a self - organized pattern ( see fig.[fig_3 ] , ( a ) ) : the laser array splits into two sub - systems with each laser having a phase difference equal to @xmath41 with its nearest neighbors ( anti - phase synchronization @xcite ) .
this pattern gradually vanishes with increasing coupling strength and the system becomes fully incoherent ( figs .
[ fig_3](c - d ) .
the influence of the detuning for two coupled lasers was illustrated in the previous section in fig .
[ fig_3 ] . for the case of an array of lasers
, we incorporate detuning in the following way : @xmath42 where @xmath43 is a constant . ,
( b ) @xmath44 , ( c ) @xmath45 . the red circle denotes the steady state solution where the amplitude of the oscillations .
other parameters : @xmath46 and @xmath4 ( movies s1-s3 ) .
[ fig_4 ] ] with this distribution , the differences of the detuning have a simple form : @xmath47 @xcite .
additionally , we redefine @xmath48 as @xmath49 .
it is possible to realize different forms of synchronization depending on the coupling strength .
one case is full synchronization , where @xmath50 holds for all lasers @xmath51 ( see fig.[fig_4](a ) , bottom ) .
the behavior is therefore similar to that of the uncoupled system since the whole array ends up in the steady state ( each laser is lasing with constant intensity equal to @xmath52 ) . in a partially synchronized state
the amplitudes are different in one or more lasers ( see fig .
[ fig_4](b ) , bottom ) and in the unsynchronized state there is no fixed amplitude relation between the oscillators ( see fig .
[ fig_4](c ) , bottom ) . in fig .
[ fig_4](a - c ) ( top ) we can see all of these states depicted in the complex unit circle . the red circle denotes the steady state solution where the amplitude of the oscillations is constant . in fig .
[ fig_4](a ) the amplitudes are locked to this value , while the phases of the individual lasers are randomly distributed over the steady state solution circle .
this case corresponds to amplitude ( intensity ) synchronization .
the opposite situation is full asynchrony , displayed by fig .
[ fig_4](c ) where both amplitude and phase exhibit incoherent behavior .
the intermediate case is shown in fig .
[ fig_4](b ) where an amplitude - chimera @xcite is illustrated through the coexistence of partial amplitude locking and incoherence . ( for more information see movies s1-s3 corresponding to fig .
[ fig_4](a - c ) ) .
our next step is to characterize the observed states by using suitable measures .
recently , kemeth _
_ presented a classification scheme for chimera states @xcite .
for measuring spatial coherence , in particular , they introduced a quantity called _ local curvature _ which may be calculated at each time instance .
this is done by applying the discrete laplacian @xmath53 on the spatial data of the amplitude of the electric field : @xmath54 in the synchronization regime the local curvature is close to zero while in the asynchronous regime it is finite and fluctuating .
therefore , if @xmath55 is the normalized probability density function of @xmath56 , @xmath57 measures the relative size of spatially coherent regions in each temporal realization . for a fully synchronized system @xmath58 , while for a totally incoherent system it holds that @xmath59 .
a value between @xmath60 and @xmath61 of @xmath57 indicates coexistence of synchronous and asynchronous lasers . over time .
( a ) @xmath62 , ( b ) @xmath44 , ( c ) @xmath45 . other parameters as in fig .
[ fig_4 ] .
[ fig_6 ] ] the quantity @xmath55 is time - dependent .
complementary to the local curvature we also calculate the spatial extent occupied by the coherent lasers which is given by the following integral : @xmath63 where @xmath64 is a threshold value distinguishing between coherence and incoherence which is related to the maximum curvature and is system - dependent
. we will apply these measures in order to classify the observed patterns and we will discuss their dependence on the coupling strength @xmath49 and the detuning parameter @xmath43 . figure [ fig_5 ] shows the spatio - temporal evolution of the local curvature corresponding to the states of fig .
[ fig_4 ] . in the fully synchronized case
the local curvature is equal to zero ( fig.[fig_5](a ) ) . in fig .
[ fig_5](b ) we have the case of an amplitude - chimera state .
we see that this is not a stationary pattern since the local curvature oscillates in time .
the fully incoherent states is shown in fig.[fig_5](c ) , where the local curvature attains higher values . in fig .
[ fig_6 ] , the time evolution @xmath65 for all three cases of fig.[fig_5 ] is plotted .
we see that for the case of fig.[fig_5](b ) @xmath66 oscillates in an irregular manner , and therefore the corresponding amplitude chimera states are _ turbulent _ according to the classification of @xcite .
the other two curves ( a ) and ( c ) refer to full synchronization and full incoherence , respectively .
apart from the detuning , the coupling strength has also an effect on the synchronization patterns observed in our system . in fig .
[ fig_7 ] the temporal mean of @xmath65 ( averaged over @xmath67 ) is plotted in the ( @xmath68 ) parameter space .
the initial conditions of the phases are randomly distributed between @xmath69 and @xmath41 , while for the electric field amplitudes and the population inversions they are chosen identical for all lasers : @xmath70 , @xmath71 .
the labels ( a ) , ( b ) and ( c ) mark the coordinates corresponding to figs
. [ fig_5](a ) , [ fig_5](b ) and [ fig_5](c ) , respectively .
it is clear , that the parameter space is separated in two main domains , one of @xmath72 values close to unity which corresponds to full coherence and contains point ( a ) , and one of @xmath72 values tending to zero which corresponds to full incoherence and contains point ( c ) . on the boundary between these two areas , lies a small region where the amplitude chimeras arise .
note that , due to multistability , the mapping of the dynamical patterns may slightly change with different choice of initial conditions .
the qualitative result , however , will be the same .
for example , in fig.[fig_8 ] ( a ) , we plot @xmath72 for a system with all initial phases randomly distributed but fixed , except those of laser 50 and 150 , which we vary from @xmath60 to @xmath73 .
clearly , the exact values of @xmath72 change but remain within the range allowing for chimera states . finally , the question of system size is addressed . in our simulations
we observe that the behavior of the system does not change significantly when increasing @xmath0 from @xmath36 to @xmath74 .
this is illustrated in fig.[fig_8 ] ( b ) .
after @xmath75 the temporal mean @xmath76 remains constant in time .
from this fact we can conclude that , for an appropriately large system , the formation of chimera states is size - independent . as a function of the system size normalized to @xmath77 ) .
( b ) ( @xmath78)-projection .
initial phases for all lasers are random and fixed while @xmath79 and @xmath80 are varied .
the color bar shows the value of the temporal mean @xmath76 .
parameters : @xmath81 and @xmath4 .
other parameters as in fig .
[ fig_4 ] .
[ fig_8 ] ]
in conclusion , we have found amplitude chimera states in a large one - dimensional network of semiconductor lasers by properly modifying the optical frequency detuning .
local coupling is sufficient to generate these states . by using suitable classification measures
we have quantified the observed dynamics . due to the system
s multistability , even a slight change in the initial conditions may produce different values for these measures .
however , the range of the obtained values ensures the existence of chimeras , the nature of which is turbulent .
the system size also has an effect on the calculated values , which saturate for arrays with more that 200 emitters .
a systematic study in the optical frequency detuning and coupling strength parameter space , shows that the region of chimera states lies between full synchronization and desynchronization . for future studies it would be worthwhile to explore the effects introduced by noise as well as the laser pump power which is the most conveniently accessible control parameter in chip scale diode systems .
such investigations may have multiple technological applications regarding next generation photonic emitters that provide on demand diverse states like turbulent chimeras .
this work was partially supported by the european union seventh framework program ( fp7-regpot-2012 - 2013 - 1 ) under grant agreement no 316165 , the ministry of education and science of the russian federation in the framework of the increase competitiveness program of nust `` misis '' ( no .
k2 - 2015 - 007 ) , and the ministry of education and science of the republic of kazakhstan via contract number 339/76 - 2015 .
00 m. j. heck , j. f. bauters , m. l. davenport , j. k. doylend , j. siddharth , g. kurczveil , s. srinivasan , y. tang , and j. e. bowers , ieee journal of selected topics in quantum electronics * 19 * , 6100117 ( 2013 ) . | semiconductor laser arrays have been investigated experimentally and theoretically from the viewpoint of temporal and spatial coherence for the past forty years . in this work , we are focusing on a rather novel complex collective behavior , namely chimera states , where synchronized clusters of emitters coexist with unsynchronized ones .
for the first time , we find such states exist in large diode arrays based on quantum well gain media with nearest - neighbor interactions .
the crucial parameters are the evanescent coupling strength and the relative optical frequency detuning between the emitters of the array . by employing a recently proposed figure of merit for classifying chimera states , we provide quantitative and qualitative evidence for the observed dynamics .
the corresponding chimeras are identified as _ turbulent _ according to the irregular temporal behavior of the classification measure .
such studies may be the springboard for designing next generation photonic emitters providing on demand diverse waveforms .
semiconductor laser , arrays , chimera states , coupling , nearest - neighbor , detuning , turbulent |
the study of minimal representation - infinite @xmath0-algebras with @xmath0 an algebraically closed field was one of the central themes of the representation theory around 1984 with contributions by bautista , gabriel , roiter , salmeron , bongartz , fischbacher and many others .
recent investigations of bongartz @xcite provide a new impetus for analyzing the module category of such an algebra and seem to yield a basis for a classification of these algebras .
here is a short summery of this development .
first of all , there are algebras with a non - distributive ideal lattice , such algebras have been studied already 1957 by jans @xcite .
second , there are algebras with a good universal cover @xmath1 and such that @xmath1 has a convex subcategory which is a tame concealed algebra of type @xmath2 or @xmath3 these were the algebras which have been discussed by bautista , gabriel , roiter and salmeron in @xcite ( we say that the universal cover is _ good _ provided it is a galois cover with free galois group and is interval - finite ) .
as bongartz now has shown , the remaining minimal representation - infinite algebras also have a good cover @xmath1 , but all finite convex subcategories of @xmath1 are representation - finite .
these are the algebras which will be discussed here .
we will show that such an algebra is special biserial and we will provide a full classification of the special biserial algebras which are minimal representation - infinite .
let us recall the definition : a finite dimensional @xmath0-algebra is said to be _ special biserial _
( see @xcite ) provided it is morita equivalent to the path algebra of a quiver @xmath4 with relations with the following properties : \(1 ) any vertex of @xmath4 is endpoint of at most two arrows , and also starting point of at most two arrows .
\(2 ) if two different arrows @xmath5 and @xmath6 start in the endpoint of the arrow @xmath7 , then at least one of the paths @xmath8 is a relation .
( 2@xmath9 ) if two different arrows @xmath7 and @xmath10 end in the starting point of the arrow @xmath5 , then at least one of the paths @xmath11 is a relation . note that the composition of an arrow @xmath7 with endpoint @xmath12 and an arrow @xmath5 with starting point @xmath12 is here denoted by @xmath13 , one should visualize the situation as follows : @xmath14 from 0.8 0.8 to 0.2 0.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.8 1.8 to 1.2 1.2 \put{$\alpha$ } at 1.7 1.3 \put{$\gamma$ } at 0.7 0.3 \endpicture}\ ] ] the definition of a special biserial algebra looks quite technical , but actually there are a lot of natural examples of algebras which turn out to be of this kind .
note that a special biserial algebra is hereditary if and only if it is morita equivalent to the path algebra of a quiver of type @xmath15 or @xmath16 , where the cyclic orientation of @xmath17 has to be excluded in order to get a finite dimensional algebra .
special biserial algebras were first studied by gelfand and ponomarev @xcite : they have provided the methods in order to classify all the indecomposable representations of such an algebra .
this classification shows that special biserial algebras are always tame ( see also @xcite ) and usually they are of non - polynomial growth .
for the structure of the auslander - reiten quiver of a special biserial algebra we refer to @xcite .
the aim of the present paper is to describe the special biserial algebras which are minimal representation - infinite and to exhibit the corresponding module categories .
we can assume that the defining relations of the special biserial algebras to be considered are monomials ( since otherwise we will obtain an indecomposable module which is both projective and injective , but minimal representation - infinite algebras do not have indecomposable modules which are both projective and injective ) .
if @xmath18 is a finite dimensional algebra , a simple module @xmath19 is said to be a _
provided @xmath19 is neither projective nor injective , and such that @xmath19 does not occur as a composition factor of a module of the form @xmath20 , where @xmath21 is indecomposable and not simple .
if @xmath18 is given by a quiver with relations , then the simple module @xmath22 corresponding to a vertex @xmath12 is a node if and only if @xmath12 is neither a sink nor a source and given an arrow @xmath7 which ends in @xmath12 and an arrow @xmath5 which starts in @xmath12 , then @xmath13 is a relation .
there is a well - known procedure @xcite to resolve nodes : for any algebra @xmath18 , there is an algebra @xmath23 without nodes such that @xmath18 and @xmath23 are stably equivalent : in case @xmath18 is given by a quiver with relations , one just replaces any vertex @xmath12 with @xmath22 a node by two vertices @xmath24 such that @xmath25 is a sink and @xmath26 a source .
a vertex @xmath12 of a quiver will be called an _ @xmath27-vertex _ provided
@xmath12 has @xmath27 neighbors ( this means that there are @xmath28 arrows ending in @xmath12 and @xmath29 arrows starting in @xmath12 and @xmath30 ; observe that in this way , the loops at @xmath12 are counted twice ) .
assume the @xmath0-algebra @xmath18 is special biserial and minimal representation - infinite .
then any @xmath31-vertex of the quiver of @xmath18 is a node . if we want to classify algebras which are minimal representation - infinite , it is sufficient to deal with algebras without a node , since an algebra @xmath18 is minimal representation - infinite if and only if the node - free algebra @xmath23 is minimal representation - infinite , see section 7 .
a finite dimensional hereditary algebra of type @xmath32 will be said to be a _
cycle algebra_. the main task of the paper will be to define two classes of finite dimensional algebras , the so - called barbell algebras and wind wheel algebras , see sections 5 and 6 , respectively .
these algebras are obtained from cycle algebras by a construction which we call barification ( see section 4 ) and adding , if necessary , suitable zero relations . the special biserial algebras which are minimal representation - infinite and have no nodes are the cycle algebras , the barbell algebras with non - serial bars and the wind wheel algebras .
a minimal representation - infinite algebra is special biserial if and only if its universal cover @xmath33 is good and any finite convex subcategory of @xmath33 is representation - finite .
the first part of these notes is devoted to a proof of theorems 1.1 , 1.2 and 1.3 .
the second part provides information on the module categories of the minimal representation - infinite special biserial algebras .
as we have mentioned already , all special biserial algebras are tame .
dealing with the minimal representation - infinite ones , we encounter both algebras of non - polynomial growth ( namely the barbell algebras ) as well as domestic ones ( the hereditary algebras of type @xmath32 as well as all the wind wheel algebras ) , note that the domestic ones all are even @xmath34-domestic . here
, an algebra is said to be _
@xmath27-domestic _ in case there are precisely @xmath27 primitive 1-parameter family of indecomposable modules ( and additional `` isolated '' indecomposables ) .
not much is known about domestic algebras @xmath18 in general , not even about 1-domestic algebras ! the wind wheel algebras @xmath35 provide new examples of @xmath34-domestic algebras such that the auslander - reiten quiver has an arbitrary finite number of non - regular components ( 12.5 ) as well as having non - regular components with arbitrary ramification ( 12.6 ) .
let us stress that the examples which we present all have loewy length @xmath36 . also , we will describe in detail the corresponding auslander - reiten quilt @xmath37 of a wind wheel
, it is obtained from the set of auslander - reiten components which contain string modules by inserting suitable infinite dimensional algebraically compact indecomposable modules .
we will see that @xmath37 is a connected orientable surface with boundary , its euler characteristic is @xmath38 where @xmath39 is the number of bars . a further property of the wind wheels @xmath35 with @xmath39 bars seems to be of interest : let @xmath21 be a primitive homogeneous and absolutely indecomposable @xmath35-module , and @xmath40 the factor ring of @xmath41 modulo the ideal of endomorphisms with semisimple image , then @xmath40 is of dimension @xmath42 , thus arbitrarily large .
in general , we will show : let @xmath18 be a @xmath0-algebra which is minimal representation - infinite and special biserial .
then any complete sectional path is a mono ray , an epi coray or the concatenation of an epi coray with a mono ray .
this implies in particular the following : if @xmath43 are indecomposable @xmath18-modules with an irreducible monomorphism @xmath44 and an irreducible epimorphism @xmath45 , then @xmath46 whereas we present in part i full proofs for the main results , the discussion in part ii is less complete , several of the ( sometimes tedious ) combinatorial verifications are left to the reader
. * acknowledgment . *
the classification of the minimal representation - infinite special biserial algebras was first announced at the trondheim conference 2007 and then presented in lectures at several places .
the author is indebted to various mathematicians for helpful comments . at the icra workshop tokyo 2010 ,
the author gave a sequence of lectures dealing with minimal representation - infinite algebras in general . the following text written for the workshop proceedings
restricts the attention again to the special biserial algebras .
* part i. the algebras *
given a quiver @xmath4 with vertex set @xmath47 and arrow set @xmath48 and a set @xmath49 of monomial relations ( monomial relations are paths of length at least 2 ) , we consider ( usually finite ) words using as letters the arrows of the quivers and formal inverses of these arrows , the set of such words will be denoted by @xmath50 ( and just by @xmath51 if no relations are given ) . in case the algebra @xmath52 is given by the quiver @xmath4 with relations @xmath49 , we also dare to write @xmath53 instead of @xmath50 ( but this is an abuse of notation ) . here is the proper definition : let @xmath54 be the quiver obtained from @xmath4 by adding formal inverses of the arrows ( given an arrow @xmath7 with starting point @xmath55 and terminal point @xmath56 we denote by @xmath57 a formal inverse of @xmath58 with starting point @xmath59 and terminal point @xmath60 given such a formal inverse @xmath61 one writes @xmath62 ) .
we consider paths in the quiver @xmath54 , those of length @xmath63 are of the form @xmath64 ( one may consider @xmath65 just as the sequence @xmath66 , but it will be convenient , to delete the brackets and the colons ) .
in addition , there are the paths of length zero corresponding to the vertices . by definition ,
the _ inverse _ of @xmath67 is @xmath68 ; a subword of @xmath65 is of the form @xmath69 ( with @xmath70 or else a vertex which is starting or terminal point for some @xmath71 .
the elements of @xmath50 are the paths @xmath67 in @xmath54 which satisfy the following conditions : ( w1 ) we have @xmath72 for all @xmath73 ( w2 ) no subword of @xmath65 or its inverse belongs to @xmath49 .
the elements of @xmath50 will be called _ words _ for the quiver @xmath4 with the relations @xmath49 , the arrows and their formal inverses will be called the _ letters_.
a word @xmath67 is said to be _ direct _ provided all the letters @xmath71 are arrows , and _ inverse _ provided @xmath74 is direct . a word which is either direct of inverse is said to be _ serial .
_ two letters @xmath75 will be said to have the same direction , if both are direct or if both are inverse letters .
we say that a word @xmath65 is _ without repetition _ provided no letter appears twice in @xmath65 . given a word without repetition and a letter @xmath76 , then both @xmath76 and
@xmath77 may appear in @xmath65 ; in this case we will say that the edge @xmath78 occurs twice in @xmath65 . given a word @xmath79 of length @xmath80
, we will write @xmath81 and @xmath82 . given words @xmath83 such that the starting point of @xmath84 is the endpoint of @xmath85 and @xmath86 are different , but have the same direction , then we will say that the pair @xmath87 is _ attracting . _
note that for an attracting pair @xmath87 , the composition @xmath88 is a word again .
finally , recall that a word @xmath65 is called _ cyclic _ provided it contains both direct and inverse letters and such that also @xmath89 is a word . a cyclic word @xmath65
is said to be _ primitive _ provided it is not of the form @xmath90 with @xmath91 an infinite sequence @xmath92 using our letters will be called a @xmath93-word provided all the finite subsequences @xmath94 are words .
similarly , a double infinite sequence @xmath95 is said to be a @xmath96-word provided all the finite subsequences @xmath97 are words .
this report deals mainly with quivers @xmath4 with a set @xmath49 of monomial relations which yield a special biserial algebra @xmath52 . in this case , the finite dimensional @xmath52-modules are easy to construct and to characterize , this classification goes back to gelfand and ponomarev @xcite .
there are two kinds of indecomposable modules , the string modules and the band modules . starting with any word @xmath98 of length @xmath27
, there is an indecomposable module @xmath99 of length @xmath100 , called a string module .
in addition , there are one - parameter families of indecomposable @xmath52-modules which are constructed starting with a primitive cyclic word @xmath65 as well as a finite dimensional vector space @xmath101 with an automorphism @xmath102 such that the pair @xmath103 is indecomposable ; the modules @xmath104 are called the band modules .
if @xmath101 is one - dimensional and @xmath102 is the multiplication by @xmath105 , then we write @xmath106 instead of @xmath106 . for an outline of these constructions
we refer to @xcite .
we first describe the hereditary algebras of type @xmath32 .
they also will be used in order to construct the barbell as well as the wind wheel algebras .
we start with a function @xmath107 ; if necessary , we call such a function an _ orientation sequence _ of length @xmath27 . in order to specify @xmath108
, we usually will write just the sequence @xmath109 , or the corresponding sequence of signs @xmath110 and @xmath111 . note that this means that we consider @xmath108 as a word of length @xmath27 in the letters @xmath110 and @xmath111
. this interpretation explains also the following conventions : assume there is given an orientation sequence @xmath108 of length @xmath27 .
we say that @xmath108 starts with @xmath112 and ends with @xmath113 we write @xmath114 for the function with @xmath115 for @xmath116 .
given a further orientation sequence @xmath117 say of length @xmath118 , let @xmath119 be the orientation sequence of length @xmath120 with @xmath121 for @xmath122 and @xmath123 for @xmath124 .
the orientation sequences which we are interested in will be obtained by starting with a word @xmath125 , where @xmath4 is a quiver with monomial relations @xmath49 and looking at @xmath126 defined by @xmath127 if @xmath71 is a direct letter and @xmath128 otherwise .
let @xmath108 be an orientation sequence .
we attach to @xmath108 the hereditary algebra @xmath129 with the following quiver : its vertices are @xmath130 , and there is an arrow @xmath131 in case @xmath132 and @xmath133 in case @xmath134 the algebra @xmath129 is finite dimensional if and only if @xmath108 is not constant .
the algebras @xmath129 with @xmath135 not constant , will be called the _ cycle algebras_. for example , if @xmath136 then @xmath129 is the path algebra @xmath137 from 0.8 0.2 to 0.2 0.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.8 0 to 1.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 0.2 to 2.8 0.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 1.8 to 2.8 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.8 2 to 1.2 2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.2 1.2 to 0.8
1.8 \put{$\alpha_1 $ } at 0.3 0.3 \put{$\alpha_2 $ } at 1.5 0.2 \put{$\alpha_3 $ } at 2.7 0.3 \put{$\alpha_4 $ } at 2.7 1.7 \put{$\alpha_5 $ } at 1.5 2.2 \put{$\alpha_6 $ } at 0.3 1.7 \endpicture}\ ] ] always @xmath138 is a primitive cyclic word .
starting from a hereditary algebra of type @xmath32 , the further algebras will be obtained by identifying some subquivers and adding zero relations .
the essential part of the construction will be described now .
let @xmath4 be a quiver with relations .
let @xmath139 be pairwise different @xmath140-vertices such that @xmath141 as well es @xmath142 are neighbors , for all @xmath143 thus there are letters @xmath144 for @xmath145 such that @xmath146 end in @xmath147 , or @xmath148 respectively , and @xmath149 start in @xmath147 , or @xmath148 respectively , for @xmath150 we assume that the letters @xmath151 have the same direction , for any @xmath152 , where s @xmath153 have different direction , and also @xmath154 have different direction .
we assume in addition that the letters @xmath144 for @xmath152 are not involved in any relation .
let @xmath155 and @xmath156 the _ barification _ of @xmath157 and @xmath158 is defined as follows : we identify the vertex @xmath159 with @xmath160 for @xmath161 , and label the new vertex again @xmath159 ; also , we identify the arrow @xmath162 between @xmath163 with the arrow between @xmath164 and label it again @xmath162 .
we add as new relation the compositions @xmath165 as well as @xmath166 . if necessary
, we will denote the new quiver with relations by @xmath167 the subquiver of @xmath168 given by the identified vertices and arrows is called a _ bar _ ( at least if @xmath169 ) .
if @xmath170 , then we just identify two @xmath140-vertices of @xmath4 in order to form a @xmath31-vertex . if @xmath169 , then we identify sequences of @xmath140-vertices and obtain from the identification of @xmath171 with @xmath172 a @xmath36-vertex , then several @xmath140-vertices , and finally as the identification of @xmath173 with @xmath173 again a @xmath36-vertex . note that in case we start with a quiver @xmath4 which is special biserial , the new quiver @xmath168 with relations again will be special biserial . here is a schematic example with @xmath174 .
we indicate the relevant parts @xmath175 and @xmath176 , but we do not specify what happens further ( we just draw a box ) @xmath177 from 2.3 4 to 2.7 4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.7 4 to 3.3 4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.7 4 to 4.3 4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.3 4 to 5.7 4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.7 4 to 6.3 4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.7 4 to 7.3 4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.7 0 to 2.3 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.7 0 to 3.3 0 \arrow < 1.5 mm
> [ 0.25,0.75 ] from 4.7 0 to 4.3 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.3 0 to 5.7 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.7 0 to 6.3 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.3 0 to 7.7 0 \put { } at 0 -3 \setlinear \plot 2.5 4.5 1.5 4.5 1.5 -3 8.5 -3 8.5 4.5 7.5 4.5 7.5 -1 2.5 -1 2.5 4.5 / \setshadegrid span < .4 mm > \vshade 1.5 -3 4.5 < , z , , > 2.5 -3 4.5 < z , z , , > 2.55 -3 -1 < z , z , , > 7.45 -3 -1 < z , z , , > 7.5 -3 4.5 < z , , , > 8.5 -3 4.5 / \endpicture}\ ] ] the barification yields a quiver of the following form : @xmath178 from 2.3 0.5 to 2.8 0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.8 -.1 to 2.3 -.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.7 0 to 3.3 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.7 0 to 4.3 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.3 0 to 5.7 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.7 0 to 6.3 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.7 0.5 to 7.2 0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.2 -.1 to 7.7 -.5 \setsolid \setdots < .7 mm >
\setquadratic \plot 7.45 -0.2 7.35 0 7.45 0.2 / \plot 2.55 -0.2 2.65 0 2.55 0.2 / \setlinear \setsolid \plot 2.5 0.7 1.5 0.7 1.5 -1.3 8.5 -1.3 8.5 0.7 7.5 0.7 7.5 -.7 2.5 -.7 2.5 0.7 / \setshadegrid span < .4 mm > \vshade 1.5 -1.3 0.7 < , z , , > 2.5 -1.3 0.7 < z , z , , >
2.55 -1.3 -.7 < z , z , , > 7.45 -1.3 -.7 < z , z , , > 7.5 -1.3 0.7 < z , , , > 8.5 -1.3 0.7 / \endpicture}\ ] ] ( the box is not changed ) .
definition : consider orientation sequences @xmath179 and assume that both @xmath108 and @xmath117 start and end with @xmath110 .
we start with the hereditary algebra @xmath180 , and construct the barification using the two copies of @xmath181 , this will be the _ barbell algebra _ @xmath182 .
the subquiver given by ( the identified copies of ) @xmath181 will be called its _
bar_. * example 1 : * start with @xmath183 then @xmath184 has the following shape : @xmath185 from 0 0.2 to 0 0.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.8 0 to 0.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1 0.8 to 1 0.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.8 1 to 0.2 1 \put{$\alpha=
\alpha_1 $ } at -0.5 0.5 \put{$\alpha_2 $ } at 0.5 -0.3 \put{$\alpha_3 = \gamma$ } at 1.5 0.5 \put{$\alpha_4 $ } at 0.5 1.3 \endpicture}\ ] ] after the identification of @xmath186 and @xmath187 , we write @xmath10 for the identified arrow : @xmath188 from 1.7 0 to 0.3 0 \circulararc 320 degrees from -0.05 0.2 center at -0.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.09 0.3 to -0.05 0.2 \circulararc -330 degrees from 2.05 0.2 center at 2.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.09 -0.3 to 2.05 -0.2 \put{$\alpha$ } at -2 0 \put{$\beta$ } at 1 0.4 \put{$\gamma$ } at 4 0 \setdots < .7 mm >
\setquadratic \setdots < .7 mm >
\setquadratic \plot -0.4
-0.4 -0.3 0 -0.4 0.4 / \plot 2.4 -0.4 2.3 0 2.4 0.4 / \endpicture}\ ] ] here , the bar is just one arrow ( namely @xmath10 ) , thus serial .
a barbell algebra is minimal representation - infinite if and only if the bar is not serial . in case
the bar @xmath181 is direct , there are arrows @xmath189 such that @xmath190 is a word for the barbell algebra .
if we add this word as a relation , we obtain an algebra which still is representation - infinite ( it is a wind wheel algebras as discussed in the next section , thus @xmath34-domestic ) .
in example 1 , the bar was serial , thus this barbell algebra was not minimal representation - infinite . here is an example of a barbell with non - serial bar : * example 2 .
* we start with @xmath191 from 0.2 0.8 to 0.8 0.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.8 0.8 to 1.2 0.2 \circulararc 310 degrees from 0.2 1.1 center at 0 1.5 \circulararc 310 degrees from 2.2 1.1 center at 2 1.5 \arrow < 1.5
mm > [ 0.25,0.75 ] from -0.25 1.14 to -0.2 1.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.75 1.14 to 1.8 1.1 \setdots < .5 mm >
\setquadratic \plot -0.25 1.3 0 1.25 .25 1.3 / \plot 1.75 1.3 2 1.25 2.25 1.3 / \endpicture}\ ] ] in order to construct this algebra , we can start with @xmath192 @xmath193 then @xmath184 has the following shape : @xmath194 from -0.175 0.8 to -0.15 0.85 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.175 0.2 to 2.15 0.15 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.2 0 to 0.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.8 0 to 1.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.2 1 to 0.8
1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.8 1 to 1.2 1 \circulararc -80 degrees from -0.15 0.15 center at .2 0.5 \circulararc 80 degrees from 2.15 0.15 center at 1.8 0.5 \put{$\alpha_1 $ } at -0.5 0.5 \put{$\alpha_2 $ } at 0.5 -0.2 \put{$\alpha_3 $ } at 1.5 -0.2 \put{$\alpha_4 $ } at 2.5 0.5 \put{$\alpha_5 $ } at 0.5 1.2 \put{$\alpha_6 $ } at 1.5 1.2 \endpicture}\ ] ] here the bar ( given by the arrows @xmath195 ) is not serial , thus the algebra is minimal representation - infinite .
a _ wind wheel algebra _ @xmath35 is given by a cyclic word @xmath65 without repetition which is of the form @xmath196 with words @xmath197 of length at least 1 and such that there is a ( necessarily fixed point free ) involution @xmath198 on the set @xmath199 with the following properties : ( ww1 ) the words @xmath200 are serial and @xmath201 .
( ww2 ) the edges appearing in some @xmath202 occur only once in @xmath65 , those occurring in some @xmath200 occur twice in @xmath65 ( namely in @xmath200 and in @xmath203 ) .
( ww3 ) the pairs @xmath204 are attracting , the pairs @xmath205 are not attracting ( here , @xmath206 .
note that the factorization into the subwords @xmath207 and the permutation @xmath198 are uniquely determined by @xmath65 , thus we can write @xmath208 .
the algebra @xmath209 is obtained from the @xmath32-algebra @xmath210 by identifying the path @xmath200 with @xmath211 ( barification ) , for all @xmath212 and using additional zero relations as follows : let @xmath200 be direct , and @xmath213 ( thus @xmath214 )
. then the barification relations are @xmath215 ( recall that for any path @xmath202 , we denote by @xmath216 its first letter , by @xmath217 the last one ) . and
we take in addition also the paths @xmath218 as relations .
thus , there are @xmath219 monomial relations of length @xmath140 as well as @xmath39 long relations ( of length at least @xmath36 ) . as in the case of a barbell algebra , a subquiver given by ( the identified copies of ) some @xmath200
will be called a _
bar_. our example 1 yields the wind wheel algebra for the following word @xmath220 its quiver with relations is as follows : @xmath188 from 1.7 0 to 0.3 0 \circulararc 320 degrees from -0.05 0.2 center at -0.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.09 0.3 to -0.05 0.2 \circulararc -330 degrees from 2.05 0.2 center at 2.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.09 -0.3 to 2.05 -0.2 \put{$\alpha$ } at -2 0 \put{$\beta$ } at 1 0.4 \put{$\gamma$ } at 4 0 \setdots < .7 mm >
\setquadratic \setdots < .7 mm >
\setquadratic \plot -0.4
-0.4 -0.3 0 -0.4 0.4 / \plot 2.4 -0.4 2.3 0 2.4 0.4 / \plot 0 -.7 1 -.3 2 -.7 / \put{with $ \alpha^2 = \gamma^2 = \alpha\beta\gamma = 0 $ } at 7 0 \endpicture}\ ] ] further examples are presented at the end of part i. there is a a canonical map @xmath221 defined as follows : write @xmath222 with letters @xmath71 for the quiver of @xmath209 , then @xmath71 may be considered as an arrow of the quiver of @xmath210 .
we set @xmath223 and extend this multiplicatively .
given a bar @xmath157 of @xmath209 , there are uniquely defined letters @xmath224 such that both @xmath225 and @xmath226 are attracting pairs .
we call @xmath227 the _ closure _ of the bar @xmath157
. a word in @xmath228 does not belong to the image of @xmath181 if and only if it contains the closure of a bar as a subword .
the wind wheel algebra @xmath208 is domestic with only one primitive cyclic word , namely @xmath65 .
if @xmath39 is the number of bars , then there are precisely @xmath39 non - periodic ( but biperiodic ) @xmath96-words : write @xmath229 , where @xmath157 is a bar
. then @xmath230 is such a @xmath96-word .
the @xmath96-word @xmath231 determines uniquely the central part @xmath227 where @xmath232 and @xmath233 , thus it determines the bar @xmath157 ; indeed , this word is the only @xmath96-word which contains a subword of the form @xmath234 such that both @xmath225 and @xmath226 are attracting pairs .
let @xmath208 be a wind wheel algebra with @xmath39 bars .
let @xmath105 .
then the endomorphism ring of @xmath235 is a radical square zero algebra with radical dimension @xmath39 , and the only endomorphism of @xmath21 with semisimple image is the zero endomorphism . thus we see that the wind wheels provide examples of @xmath34-domestic algebras @xmath18 with a primitive homogeneous and absolutely indecomposable @xmath18-module @xmath21 such that the factor ring of @xmath41 modulo the ideal of endomorphisms with semisimple image is of arbitrarily large dimension .
any bar @xmath236 provides an endomorphism of @xmath21 with image @xmath237 , these endomorphisms form a basis of the radical of @xmath238
we want to present the proof of theorem 1.1 .
the process of resolving a node @xmath12 can be visualized as follows : @xmath239 from 1.5 3.5 to 2.8 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.5 3.5 to 2.8 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3 3.5 to 3 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.5 3.5 to 3.2 2.2 \put{$\circ$ } at 3 2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.9 1.8 to 2.5 0.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 1.8 to 4.2 0.5 \put{$a$ } at 3.5 2 \put{$\lambda$ } at -0.7 3.5 \setquadratic \setdots < .5 mm > \plot 2.3 2.5 2.2 2 2.7 1.3 / \plot 2.9 2.7 2.4 2.1 2.75 1.5 / \plot 3.5 2.7 2.6 2.2 2.8 1.7 / \plot 2.4 2.7 3.7 2.5 3.6 1.5 / \plot 3.1 2.8 3.9 2.4 3.6 1.35 / \plot 3.9 2.8 4.1 2.2 3.7 1.2 / \setlinear \setshadegrid span < 0.5 mm > \hshade 0 0 5 < , , z , z > 0.99 0 5 < , , z , z > 1 0 1 < , , z , z > 3 0 1 < , , z , z > 3.01 0 5 < , , z , z > 4 0 5 / \endpicture } at 0 0 \put{\beginpicture \multiput { } at 0 0 5 4 / \plot 0 0 0 4 5 4 5 3 1 3 1 1 5 1 5 0 0 0 / \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.5 3.5 to 2.3 2.2 \arrow
< 1.5 mm > [ 0.25,0.75 ] from 3 3.5 to 2.5 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.5 3.5 to 2.7 2.2 \put{$\circ$ } at 2.5 2 \put{$\circ$ } at 3.5 2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.4 1.8 to 2.7 0.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.7 1.8 to 4.2 0.5 \put{$a_+$ } at 2.1 1.8 \put{$a_-$ } at 4.1 2.2 \put{$\lambda'$ } at -0.7 3.5 \setshadegrid span < 0.5 mm > \hshade 0 0 5 < , , z , z > 0.99 0 5 < , , z , z > 1 0 1 < , , z , z > 3 0 1 < , , z , z > 3.01 0 5 < , , z , z > 4 0 5 / \endpicture } at 9 0 \endpicture}\ ] ] we replace the vertex @xmath12 by two vertices labeled @xmath25 and @xmath26 such that @xmath25 becomes a sink , @xmath26 a source : all arrows of @xmath18 will be kept , however , if an arrow of @xmath18 ends in @xmath12 , then in @xmath240 it ends in @xmath25 , whereas if an arrow of @xmath18 starts in @xmath12 , then in @xmath240 it starts in @xmath26 . since all the paths @xmath13 with @xmath7 ending in @xmath12 ( and therefore @xmath5 starting in @xmath12 ) are relations for @xmath18 , there is a minimal set of relations consisting of these paths as well as of a set @xmath241 of relations which do not pass through @xmath12 ( a relation is a linear combination of paths and we say that the relation passes through @xmath12 provided at least one of the paths contains a subpath @xmath13 with @xmath7 ending in @xmath12 ) .
it is the set @xmath241 which is used as set of relations for @xmath240 .
the important feature of this construction is the following : there is a canonical functor @xmath242 which yields a bijection between the indecomposable @xmath18-modules and the indecomposable @xmath240-modules different from the simple @xmath240-module @xmath243 ) there is the following quite obvious assertion : assume @xmath18 is a finite dimensional algebra with a node @xmath12 .
let @xmath240 be obtained from @xmath18 by resolving the node .
then @xmath18 is minimal representation - infinite if and only if @xmath240 is minimal representation - infinite .
thus , if we want to classify algebras which are minimal representation - infinite , it is sufficient to deal with algebras without a node .
recall that a word @xmath65 which is neither direct nor inverse is called cyclic , provided @xmath244 is a word .
let @xmath245 be a cyclic word with @xmath7 an arrow .
then at least one of the words @xmath246 , @xmath247 is a cyclic word .
first , assume that neither @xmath248 nor @xmath157 is direct , write @xmath249 and @xmath250 with @xmath251 all being direct and of maximal possible length . since @xmath65 is a cyclic word , noth @xmath252 and @xmath253 are words .
assume that @xmath246 is not a cyclic word , then there is a zero relation which is a subword of @xmath254 .
since no subword of @xmath252 is a zero relation , we conclude that @xmath255 for some word @xmath256 with @xmath257 also @xmath258 is a word .
this implies that @xmath247 is a cyclic word .
now assume @xmath248 is direct . since @xmath65 is not direct , we know that @xmath157 can not be direct .
as above , write @xmath250 with @xmath259 direct and of maximal possible length . since @xmath65 is a cyclic word
, there is no zero relation which is a subword of @xmath260 .
but the direct word @xmath258 is a subword of @xmath260 , thus we see that there is no subword of @xmath258 is a zero relation , thus @xmath247 is a cyclic word .
this completes the proof . as a consequence ,
we see : if @xmath65 is a cyclic word of minimal length , then any arrow can occur in @xmath65 at most once as a direct letter , and at most once as an inverse letter .
in particular , the length of @xmath65 is bounded by @xmath261 , where @xmath12 is the number of arrows .
( a typical example of a cyclic word of minimal length which contains both an arrow as well as its inverse is given by example 1 . )
let @xmath12 be a @xmath31-vertex , with arrows @xmath262 ending in @xmath12 and arrows @xmath263 starting in @xmath12 such that the words @xmath264 and @xmath265 are relations .
let @xmath65 be a cyclic word of smallest possible length .
assume @xmath65 contains @xmath13 as a subword .
up to rotation , we can assume that @xmath65 starts with @xmath13 .
assume @xmath65 also contains @xmath266 say @xmath267 , for some words @xmath268 .
then @xmath269 is a cyclic word of shorter length , a contradiction .
similarly , if @xmath270 , for some words @xmath271 , then @xmath272 is a cyclic word of shorter length .
now assume that @xmath65 contains @xmath273 as a subword .
then @xmath274 where @xmath268 are words such that @xmath248 starts with @xmath7 and ends with @xmath6 ( it may be that @xmath275 , or else @xmath276 )
. then we consider @xmath277 .
this is again a cyclic word , and it contains neither @xmath273 or its inverse as a subword .
namely , according to lemma 1 , @xmath273 was contained just once as a subword of @xmath65 , and this composition has been destroyed when we built @xmath278 .
also no new composition has been created .
finally , we have to consider the case that @xmath65 does not contain @xmath273 .
since @xmath65 has to contain @xmath10 and @xmath6 , it must contain @xmath279 and @xmath280 by lemma 1 , @xmath65 contains @xmath13 only once , and it can not contain @xmath281 , since otherwise we apply the previous considerations to @xmath74 .
let @xmath282 , then the subword @xmath283 does not contain @xmath13 or its inverse .
similarly , write @xmath284 , then @xmath285 is a subword of @xmath244 and does not contain @xmath13 or its inverse .
we form the word @xmath286 .
this is a cyclic word which does not contain @xmath13 or its inverse .
this shows that the factor algebra with the added relations @xmath13 and @xmath273 is still representation - infinite .
this completes the proof .
we can assume that we deal with a special biserial algebra @xmath18 with no 4-vertex .
let @xmath65 be a cyclic word of minimal length , thus no letter occurs twice .
given an arrow @xmath7 , we will say that the edge @xmath287 occurs once in @xmath65 if precisely one of the letters @xmath288 occurs in @xmath65 , and that it occurs twice if both occur ; these are the only possibilities , since we can assume that any arrow or its inverse occurs in @xmath65 .
we can assume that all the vertices of @xmath4 are 2-vertices or 3-vertices ( note that the support of a cyclic word can not contain a @xmath34-vertex ) . in case
all the vertices are 2-vertices , then we deal with a hereditary algebra of type @xmath32 .
thus we can assume that there is at least one @xmath36-vertex .
if @xmath12 is a 3-vertex , there can be only one zero - relation of length two passing through @xmath12 , since otherwise the word @xmath65 could not pass through @xmath12 .
thus , we deal with the following local situation @xmath289 from 1.2 0.8 to 1.8 0.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.8 1.8 to 1.2 1.2 \put{$\alpha$ } at 1.7 1.4 \put{$\delta$ } at 1.7 0.6 \put{$\eta$ } at 0.4 1.2 \setquadratic \setdots < .5 mm > \plot 1.4 0.7 1.3 1 1.4 1.3 / \endpicture}\ ] ] _ then @xmath290 occurs twice in @xmath65 whereas @xmath291 and @xmath292 occur just once .
_ proof : since @xmath287 as well as @xmath292 both have to occur at least once , this yields two different subwords of @xmath65 involving @xmath290 .
but if @xmath287 or @xmath292 would occur twice , we would obtain at least three letters of the form @xmath181 and @xmath293 , impossible . in this way
, we see that there are edges which occur once , as well as edges which occur twice .
this shows clearly the structure of the word @xmath65 .
up to rotation , we can assume that @xmath65 starts in a 3-vertex , and that the inverse of the first letter does not occur in @xmath65 .
of course , then the last letter yields an edge which occurs twice .
we cut @xmath65 into pieces @xmath294 such that any @xmath202 uses edges which occur only once , whereas the @xmath200 use edges which occur twice .
we obtain an involution @xmath198 on the set @xmath295 with @xmath296 ( note that in case @xmath76 is a letter which occurs in some @xmath200 , and @xmath77 occurs in @xmath297 , then necessarily @xmath213 , thus we define @xmath298 _ the involution @xmath198 has no fixed point . _
namely , @xmath299 for any word @xmath157 : in case @xmath157 has odd length , just consider the middle letter , it can not be both direct and inverse ; in case @xmath157 has even length , say @xmath300 with letters @xmath71 , then @xmath301 , which is excluded .
this shows that @xmath302 is even .
let us look at the behavior of the compositions @xmath303 and @xmath304 and compare this with the compositions @xmath305 and @xmath306 , taking into account that @xmath307 .
the composition @xmath303 occurs at the same 3-vertex as @xmath306 , thus precisely one of the pairs @xmath205 and @xmath308 is attracting . similarly , precisely one of the pairs @xmath204 and @xmath309 is attracting . claim : _ if one of the words @xmath200 is not serial , then @xmath310 _ let us assume that @xmath311 and that one of the words @xmath200 is not serial , say @xmath312 such that the pair ( @xmath313 is attracting .
two different cases have to be considered .
the first case is the following : @xmath314 where all the @xmath315 start and end with edges which occur only once .
we can assume that the pair @xmath316 is not attracting .
otherwise , @xmath317 is not attracting , and we replace the given word @xmath65 first by @xmath74 and then by a rotated one in order to obtain a similar situation .
we claim that we can replace @xmath85 at the beginning by the word by @xmath318 , thus dealing with @xmath319 this is a word .
namely , since the pair @xmath316 is not attracting , the pair @xmath317 is attracting , thus the same is true for the inverse . also , it is a cyclic word , since the pair @xmath320 , thus also @xmath321 is attracting .
it remains to observe that this new cyclic word uses less arrows : all the edges of @xmath85 which have multiplicity 1 in @xmath65 have disappeared .
this contradicts that @xmath4 is minimal representation - infinite .
let us now discuss the second case , where @xmath322 lies in between @xmath157 and @xmath323 .
@xmath324 where again all the @xmath315 start and end with edges which occur only once .
again , we can assume that the pair @xmath316 is not attracting ( otherwise , the pair @xmath325 is not attracting and we invert and rotate @xmath65 ) .
this time , we claim that @xmath326 is a cyclic word .
the first part is a subword of @xmath65 , the inverse of the second part is also a subword of @xmath65 .
thus , both words @xmath327 and @xmath328 exist and they can be composed , since @xmath329 is an attracting pair . also , it is a cyclic word , since @xmath330 is an attracting pair : by assumption , @xmath316 is not attracting , thus @xmath325 is an attracting pair , and therefore the same is true for its inverse @xmath331 the case @xmath332 with @xmath333 non - serial yields a barbell .
thus , we now assume that all the words @xmath200 are serial .
claim : ( @xmath334 ) _ if @xmath335 is not attracting , then also @xmath336 is not attracting , but @xmath337 is attracting . _
let @xmath338 and assume that @xmath335 is not attracting , but @xmath339 is attracting .
let @xmath340 and @xmath341 , thus @xmath342 and @xmath343 .
we claim that @xmath344 is a cyclic word .
of course , @xmath345 is a subword of @xmath65 .
also , @xmath346 is the inverse of a subword of @xmath65 . since we assume that @xmath347 is attracting , also the inverse @xmath348 is attracting , and thus also @xmath349 is attracting .
since @xmath335 is not attracting , the pair @xmath350 is attracting .
this shows that the concatenation of @xmath351 with @xmath278 or even with @xmath352 provides no problem .
therefore we see that @xmath353 is a cyclic word .
note that @xmath65 has the serial word @xmath354 as a subword , whereas this is no longer a subword of @xmath353 .
this means that we can use @xmath355 as an additional relation and still have the cyclic word @xmath353 .
this contradicts the assumption that we deal with a minimal representation - infinite algebra .
this contraction shows that @xmath336 is not attracting , but then @xmath337 has to be attracting .
claim : _ if @xmath335 is not attracting , then also @xmath356 is not attracting . _
thus , let us assume that @xmath356 is attracting , whereas @xmath335 is not .
we already have seen in ( @xmath334 ) that @xmath336 is not attracting , thus @xmath357 we know from ( @xmath334 ) that @xmath337 is attracting .
by assumption , also @xmath356 is attracting .
consider the case that @xmath358 @xmath359 thus we may replace @xmath360 by @xmath361 and obtain the cyclic word @xmath362 but here we use less arrows : all the arrows in @xmath363 have disappeared . finally , we have to deal with the case that @xmath364 @xmath365 and take the cyclic word @xmath366 this exists , since the pairs @xmath367 and @xmath368 both are attracting . in this case , we lose for example all the arrows which occur in @xmath369 .
altogether we see that @xmath65 defines a wind wheel algebra .
let @xmath18 be special biserial , let @xmath4 be its quiver . if @xmath18 is minimal representation - infinite , then the ideal of relations is generated by a set @xmath49 of zero relations , thus the universal cover @xmath18 of @xmath18 is given by the universal cover @xmath370 of the quiver @xmath4 and the set @xmath371 of lifted relations .
it follows that @xmath1 is special biserial with a quiver which is a tree , and the indecomposable @xmath1-modules are string modules with support a quiver of type @xmath15 .
it follows that any finite convex subcategory is representation - finite .
conversely , assume now that @xmath18 is a finite dimensional basic @xmath0-algebra which is minimal representation - infinite , that the universal cover @xmath1 of @xmath18 is good and that any finite convex subcategory of @xmath372 is representation - finite .
we want to show that @xmath18 is special biserial .
let @xmath4 be the quiver of @xmath18 and @xmath370 the quiver of @xmath1 .
we denote by @xmath373 the covering functor ( often called push - down functor , or forgetful functor ) . for any finite dimensional @xmath0-algebra @xmath52 ,
let @xmath374 be the number of ( isomorphism classes of ) simple @xmath52-modules .
note that @xmath372 is not of bounded representation type , since otherwise also @xmath18 would be of bounded , thus finite , representation type .
thus we see that there are indecomposable representations @xmath21 of @xmath372 of arbitrarily large length .
given such a representation @xmath21 say of length @xmath375 , its support algebra @xmath33 is a representation - directed algebra , thus @xmath376 .
this shows that there are indecomposable representations @xmath21 of @xmath18 whose support has with arbitrarily large cardinality .
the sincere representation - directed algebras with large support have been classified by bongartz ( see @xcite or @xcite ) . such an algebra @xmath33 has a convex subalgebra @xmath377 which is given by a quiver of type @xmath15 ( without any relation ) , such that @xmath378 thus we see : @xmath372 has convex subcategories @xmath377 which are given by a quiver of type @xmath15 without relations , where @xmath27 is arbitrarily large .
let @xmath379 the loewy length of @xmath18 .
choose a convex subcategory @xmath377 of @xmath372 which is of the form @xmath15 and without relations , where @xmath380 and let @xmath21 the the ( unique ) since representation of @xmath377 .
note that the indecomposable @xmath377-modules are string modules , they are given by words using as letters the arrows of the quiver of @xmath377 .
since @xmath381 , it follows easily that there is a simple @xmath18-module @xmath19 such that @xmath382 \ge 3 $ ] . but
this implies that there is an indecomposable @xmath377-module @xmath383 which is given by a word of the form @xmath384 , with arrows @xmath385 and @xmath386 such that @xmath387 and @xmath388 are different arrows of @xmath4 and end in the same vertex of @xmath4 ( namely the support vertex of @xmath19 ) .
we denote by @xmath52 the support algebra of @xmath383 , it is given by a quiver of type @xmath389 without any relation , and @xmath390 has a one - parameter family of simple submodules @xmath391 such that the @xmath18-modules @xmath392 are indecomposable and pairwise non - isomorphic .
since @xmath18 is minimal representation - infinite , it follows that @xmath390 is faithful , thus we obtain all the vertices and all the arrows of @xmath4 by applying @xmath393 to the vertices and arrows of @xmath52 , respectively .
since we now know that the vertices and the arrows of @xmath4 are images under @xmath393 of vertices and arrows in the support of @xmath52 , it follows that the support of @xmath52 contains a fundamental domain for the action of the galois group on @xmath370 .
in particular , we see that we can compose the word @xmath65 with a word @xmath278 such that @xmath394 and so on
. now assume there is a vertex @xmath395 of @xmath4 with 3 arrows ending in @xmath395 , say @xmath396 .
looking at the universal cover , we obtain arrows @xmath397 ending in the same vertex @xmath398 but the arrow @xmath399 is the first letter of a word @xmath400 of length @xmath36 which yields a string module for @xmath372 .
similarly , the arrow @xmath401 is the first letter of a word @xmath402 of length @xmath36 which also yields a string module for @xmath372 .
but the union of the support of the words @xmath403 and @xmath5 is a subquiver ( without any relation ) of @xmath370 of type @xmath404 .
this contradicts the assumption that any finite subcategory of @xmath1 is representation - finite .
the dual argument shows that any vertex of @xmath4 is starting point of at most two arrows .
now assume that the vertex @xmath395 of @xmath4 is endpoint of the arrows @xmath405 and starting point of the arrow @xmath5 and that neither @xmath13 nor @xmath264 is a zero relation .
we looking again at the universal cover , and obtain arrows @xmath406 ending in a vertex @xmath407 , as well as @xmath408 starting in the vertex @xmath407 .
since @xmath13 is not a zero relation , we see that @xmath409 must be a subword of @xmath65 or @xmath74 , in particular we can prolong it to a word @xmath410 of length @xmath31 so that we have a corresponding string module @xmath411 similarly , we prolong @xmath412 to a word @xmath413 of length @xmath31 with string module @xmath414 .
we consider the union of the support of the words @xmath403 ; it is a subquiver ( without any relation ) of @xmath370 , again of type @xmath404 , thus again we obtain a contradiction to the assumption that any finite subcategory of @xmath1 is representation - finite .
first , we present a second example of a barbell algebras with non - serial bar . * example 3 . * start with @xmath192 @xmath415 then @xmath180 has the following shape : @xmath416 from -0.175 0.8 to -0.15 0.85 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.8 0 to 0.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 0 to 1.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.8 0.13 to 2.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 1 to 2.8 0.87 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.8 1 to 0.2 1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 1 to 1.8 1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 0.75 to 3.15 0.75 \circulararc -80 degrees from -0.15 0.15 center at .2 0.5 \put{$\alpha_1 $ } at -0.5 0.5 \put{$\alpha_2 $ } at
0.5 -0.2 \put{$\alpha_3 $ } at 1.5 -0.2 \put{$\alpha_4 $ } at 2.6 -0.1 \put{$\alpha_5 $ } at 3.8 0.5 \put{$\alpha_6 $ } at 2.6 1.1 \put{$\alpha_7 $ } at 0.5 1.2 \put{$\alpha_8 $ } at 1.5 1.2 \ellipticalarc axes ratio 2.5:1 170 degrees from 3.15 0.25 center at 3.1 0.5 \endpicture}\ ] ] and @xmath182 will be @xmath417 from 0.8 0 to 0.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 0 to 1.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.7 -0.5 to 2.2 -0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 0.1 to 2.7 0.5 \ellipticalarc axes ratio 1:1 180 degrees from 3.1 -0.55 center at 3.1 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.15 0.545 to 3.1 0.55 \circulararc 320 degrees from -0.05 0.2 center at -0.4 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.095 0.3 to -0.05 0.2 \multiput { } at -1 0.5 4 -0.5 / \setdots < .7 mm >
\setquadratic \plot -0.2 -0.3 -0.15 0 -0.2 0.3 / \plot 2.5 -0.2 2.4 0 2.5 0.2 / \endpicture}\ ] ] again , the bar ( given by the arrows @xmath418 is not serial .
the next examples are wind wheel algebras .
* example 4 .
* the wind wheel algebra for the word @xmath65 @xmath419 the permutation is @xmath420 the short zero relations are @xmath421 the long zero relations are @xmath422 the corresponding wind wheel algebra @xmath208 is : @xmath423 from 0 2.8 to 0 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4 1.2 to 4 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.2 0.9 to 1.8 0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.8 0.9 to 2.2 0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.8 3.9 to 0.2 3.1 \arrow < 1.5
mm > [ 0.25,0.75 ] from 2.2 3.9 to 3.8 3.1 \arrow < 1.5
mm > [ 0.25,0.75 ] from 3.8 2.8 to 0.2 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.2 2.8 to 3.8 1.2 \setdots<.7 mm > \plot 3.4 2.8 3.4 3.2 / \plot 3.4 0.8 3.4 1.2 / \plot 0.6 2.8 0.6 3.2 / \plot 0.6 0.8 0.6 1.2 / \endpicture}\ ] ] with the further relations @xmath424 and @xmath425 there are two bars , they are given by the arrows @xmath7 and @xmath108 . in order to construct @xmath35 ,
one starts with the following orientation sequence : @xmath426 and constructs the hereditary algebra @xmath427 : @xmath428 from 0.2 0 to 0.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 0 to 1.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.8 0.8 to 2.2 0.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3 1.8 to 3 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3 2.2 to 3 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 3 to 2.8 3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 3 to 1.8 3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.8 2.8 to 0.2 2.2 \arrow < 1.5
mm > [ 0.25,0.75 ] from 0 1.8 to 0 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0 0.2 to 0 0.8 \endpicture}\ ] ] thus , we start with a quiver which can be drawn either as a zigzag ( with arrows pointing downwards ) , where the left end and the right end have to be identified , or else as a proper cycle : @xmath429 from 0.2 0 to 0.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 0 to 1.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.8 0.8 to 2.2 0.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3 1.8 to 3 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3 2.2 to 3 2.8 \arrow < 1.5
mm > [ 0.25,0.75 ] from 2.2 3 to 2.8 3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 3 to 1.8 3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.8 2.8 to 0.2 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0 1.8 to 0 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0 0.2 to 0 0.8 \setdashes < 1 mm > \plot 0.5 -0.5 0.5 0.5 2.5 0.5 2.5 -0.5 0.5 -0.5 / \plot 0.5 2.5 0.5 3.5 2.5 3.5 2.5 2.5 0.5 2.5 / \setshadegrid span < 0.5 mm > \vshade -.5 0.5 2.5 .5 0.5 2.5 / \vshade 2.5 0.5 2.5 3.5 0.5 2.5 / \endpicture } at 9 0 \put{\beginpicture \setcoordinatesystem units < .6cm,.4 cm > \multiput{$\bullet$ } at -0 0 -1 1 -2 2 -3 1 -4 0 -6 2 -7 1 -8 0 -9 1 -10 2 -12 0 / \plot -0.3 0.3 -0.7 0.7 / \plot -1.3 1.3 -1.7 1.7 / \plot -2.3 1.7 -2.7 1.3 / \plot -3.3 0.7
-3.7 0.3 / \plot -4.3 0.3 -5.7 1.7 / \plot -6.3 1.7 -6.7 1.3 / \plot -7.3 0.7 -7.7 0.3 / \plot -8.3 0.3 -8.7 0.7 / \plot -9.3 1.3 -9.7 1.7 / \plot -10.3 1.7 -11.7 0.3 / \setdashes < 1
mm > \plot -10 1 -9 2 -7 0 -8 -1 -10 1 / \plot -1 2 -2 3 -4 1 -3 0 -1 2 / \setshadegrid span < 0.5 mm > \vshade -8 1 1 < z , z , , > -7 0 2 < z , z , , > -6 1 3 < z , z , , > -5 2 2 / \vshade -2 1 1 < z , z , , > -1 0 2 < z , z , , > 0 -1 1 < z , z , , > 1 0 0 / \plot -12 -1 -12 1.5 / \plot 0 -1 0 1.5 / \endpicture } at 0 0 \endpicture}\ ] ] and we barify on the one hand the two subquivers which are enclosed in rectangular boxes , on the other hand also the two subquivers with shaded background . in both cases , the barification yields an identification of a projective serial module of length 2 with an injective serial module of length 2 .
* example 5 .
* we start with the following orientation sequence : @xmath430 and construct the hereditary algebra @xmath427 : @xmath431 from 0.2 0.8 to 0.8 0.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 0 to 1.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.8 .8 to 2.2 0.2 \arrow < 1.5 mm >
[ 0.25,0.75 ] from 3 1.8 to 3 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.8 2.2 to 2.2 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.8 3 to 1.2 3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.2 2 . to 0.8 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0 1.2 to 0 1.8 \endpicture}\ ] ] the quiver which we obtain is @xmath432 from 0 0.8 to 0 0.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2 0.8 to 2 0.2 \circulararc -310 degrees from 0.2 -.1 center at 0 -.5 \circulararc -310 degrees from 2.2 -.1 center at 2 -.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.25 -.14 to -0.2 -.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.75 -.14 to 1.8 -.1 \setquadratic \plot 0.2 1.2 1 1.4 1.8 1.2 / \plot 0.2 0.8 1 0.6 1.8 0.8 / \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.75 1.22 to 1.8 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.25 0.78 to 0.2 0.8 \setdots < .5 mm >
\setquadratic \plot -0.25 -.3 0 -.25 .25 -.3 / \plot 1.75 -.3 2 -.25 2.25 -.3 / \plot 0.5 0.8 0.4 1 0.5 1.2 / \plot 1.5 0.8 1.6 1 1.5 1.2 / \put { } at 0 -1 \endpicture}\ ] ] with the additional relations : @xmath433 and @xmath434 the bars are given by the arrows @xmath435 and @xmath436 * example 6 .
* as in example 4 , consider again @xmath426 but now we take @xmath437 @xmath438 from 0.2 0 to 0.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 0 to 1.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.8 0.8 to 2.2 0.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3 1.8 to 3 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3 2.2 to 3 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 3 to 2.8 3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 3 to 1.8 3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.8 2.8 to 0.2 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0 1.8 to 0 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0 0.2 to 0 0.8 \endpicture}\ ] ] here is the quiver : @xmath439 from 0 0.2 to 0 0.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2 0.8 to 2 0.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.8 0.4 to 0.2 0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.2 -.1 to 0.8 -.4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 0.4 to 1.8 0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.8 -.1 to 1.2 -.4 \circulararc 310 degrees from 0.2 1.1 center at 0 1.5 \circulararc 310 degrees from 2.2 1.1 center at 2 1.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.25 1.14 to -0.2 1.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.75 1.14 to 1.8 1.1 \setdots < .5 mm >
\setquadratic \plot -0.25 1.3 0 1.25 .25 1.3 / \plot 0.5 -0.2 0.4 0 0.5 0.2 / \plot 1.75 1.3 2 1.25 2.25 1.3 / \plot 1.5 -0.2 1.6 0 1.5 0.2 / \endpicture}\ ] ] with the additional relations @xmath440 and @xmath441 .
the bars are again the arrows @xmath435 and @xmath442 . *
the module categories *
we consider the algebras @xmath443 where @xmath108 is not constant , so that @xmath444 is finite dimensional .
the structure of the auslander - reiten quiver of @xmath444 is well - known : there is the preprojective and the preinjective component , the remaining components are regular tubes .
the string modules form four components of the auslander - reiten quiver , namely the preprojective component , the preinjective component , and two tubes ( we call them _ string tubes _ ) ; the remaining components ( the _ band tubes _ ) are homogeneous .
note that the string tubes may be homogeneous or exceptional !
the auslander - reiten quiver of @xmath444 looks as follows : @xmath445 one should stress that in this case all the auslander - reiten components are ( considered as simplicial complexes , thus as topological spaces ) surfaces with boundary ; all are homeomorphic to @xmath446 .
we consider @xmath447 and depict here , for later reference , the four components which contain string modules , always we draw two fundamental domains inside the universal cover of the component , they are separated by dashed lines ( horizontal ones for the preprojective and the preinjective component , vertical ones for the regular components ) . note that instead of arrows , we show edges , the orientation is from left to right .
@xmath448 at -15 28 \put{component $ \mathcal p$ } [ l ] at -15 26 \put{the preinjective } [ r ] at 45 28 \put{component $ \mathcal q$ } [ r ] at 45 26 \put{\beginpicture \multiput { } at 0 0 10 16 / \plot 0 12 4 16 / \plot 1 11 6 16 / \plot 2 10 8 16 / \plot 3 9 10 16 / \plot 0 4 10 14 / \plot 1 3 10 12 / \plot 2 2 10 10 / \plot 3 1 10 8 / \plot 4 0 10 6 / \plot 6 0 10 4 / \plot 8 0 10 2 / \plot 0 4 4 0 / \plot 1 5 6 0 / \plot 2 6 8 0 / \plot 3 7 10 0 / \plot 0 12 10 2 / \plot 1 13 10 4 / \plot 2 14 10 6 / \plot 3 15 10 8 / \plot 4 16 10 10 / \plot 6 16 10 12 / \plot 8 16 10 14 / \setdashes < 1 mm > \plot 4 0 11 0 / \plot 4 8 11 8 / \plot 4 16 11 16 / \endpicture } at 3 20 \put{\beginpicture \multiput { } at 0 0 10 16 / \plot 6 0 7 1 / \plot 4 0 6 2 / \plot 2 0 5 3 / \plot 0 0 8 8 / \plot 0 2 7 9 / \plot 0 4 6 10 / \plot 0 6 5 11 / \plot 0 8 8 16 / \plot 0 10 6 16 / \plot 0 12 4 16 / \plot 0 14 2 16 / \plot 6 16 7 15 / \plot 4 16 6 14 / \plot 2 16 5 13 / \plot 0 16 8 8 / \plot 0 14 7 7 / \plot 0 12 6 6 / \plot 0 10 5 5 / \plot 0 8 8 0 / \plot 0 6 6 0 / \plot 0 4 4 0 / \plot 0 2 2 0 / \setdashes < 1
mm > \plot -1 0 8 0 / \plot -1 8 8 8 / \plot -1 16 8 16 / \endpicture } at 27 20 \endpicture}\ ] ] the two exceptional tubes : @xmath449 recall that a module is called _ serial _ provided it has a unique composition series .
we consider the serial @xmath444-modules , where @xmath444 is a cycle algebra .
the following assertions are easy to verify : any serial @xmath444-module @xmath21 is projective or regular or injective .
if @xmath21 is a serial @xmath444-module of length at least two , and @xmath21 is projective , then @xmath450 belongs to @xmath451 or @xmath452 if @xmath21 is a serial @xmath444-module of length at least two , and @xmath21 is injective , then @xmath453 belongs to @xmath451 or @xmath452 in the barification process , we barify a projective serial module @xmath237 of length at least @xmath140 with @xmath454 say in @xmath451 and an injective serial module @xmath455 of the same length with @xmath456 in @xmath457
. our convention for distinguishing @xmath451 and @xmath457 will be the following : given an indecomposable projective @xmath444-module @xmath458 with radical @xmath459 , where @xmath460 and @xmath461 are serial modules , we fix the order @xmath462 and assume that the module @xmath463 as well as the composition factors of @xmath464 are simple regular objects of @xmath451 , whereas the module @xmath465 as well as the composition factors of @xmath466 are simple regular objects of @xmath457 : @xmath467
let @xmath444 be hereditary of type @xmath32 , and @xmath468 a corresponding wind wheel algebra .
let us look at the restriction functor @xmath469 .
first , let us recall the shape of the category @xmath470 .
@xmath471 we have shaded the homogeneous tubes : they remain untouched ; whereas the other four components are cut ( between rays or corays ) into pieces and these pieces are embedded ( with some overlap ) into a component which contain in addition so - called quarters . this cut - and - paste process will now be explained .
we consider the wind wheel algebra @xmath209 for the word @xmath472 thus we start with the quiver @xmath473 @xmath474 from 0.8 0.2 to 0.2 0.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.8 1.8 to 0.2 1.2 \arrow < 1.5
mm > [ 0.25,0.75 ] from 1.8 0 to 1.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.8 2 to 1.2 2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.8 0 to 2.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.8 2 to 2.2 2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.8 1.2 to 3.2 1.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.8 0.8 to 3.2 0.2 \put{${\scriptstyle } 0 $ } at 0 0.5 \put{${\scriptstyle } 1'$ } at 1 2.5 \put{${\scriptstyle } 2'$ } at 2 2.5 \put{${\scriptstyle } 3'$ } at 3 2.5 \put{${\scriptstyle } 1 $ } at 1 -0.5 \put{${\scriptstyle } 2 $ } at 2 -0.5 \put{${\scriptstyle } 3 $ } at 3 -0.5 \put{${\scriptstyle } 4 $ } at 4 0.5 \put{${\scriptstyle } \alpha$ } at 0.5 0 \put{${\scriptstyle } \beta_1'$ } at 0.4 2.1 \put{${\scriptstyle } \beta_2'$ } at 1.5 2.7 \put{${\scriptstyle } \beta_3'$ } at 2.5 2.7 \put{${\scriptstyle } \beta_1 $ } at 1.5 -.7 \put{${\scriptstyle } \beta_2 $ } at 2.5 -.7 \put{${\scriptstyle } \beta_3 $ } at 3.6 -.1 \put{${\scriptstyle } \gamma$ } at 3.5 2 \endpicture}\ ] ] and barify the subquivers @xmath475 and @xmath476 .
we obtain in this way the wind wheel algebra @xmath208 @xmath477 from 1.7 0 to 0.3 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.7 0 to 2.3 0 \arrow < 1.5 mm >
[ 0.25,0.75 ] from 5.7 0 to 4.3 0 \circulararc 320 degrees from -0.05 0.2 center at -0.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.09 0.3 to -0.05 0.2 \circulararc
-330 degrees from 6.05 0.2 center at 6.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.09 -0.3 to 6.05 -0.2 \put{$\alpha$ } at -2 0 \put{$\beta_1 $ } at 1 0.4 \put{$\beta_2 $ } at 3 0.4 \put{$\beta_3 $ } at 5 0.4 \put{$\gamma$ } at 8 0 \setdots < .7 mm >
\setquadratic \setdots < .7 mm > \setquadratic \plot -0.4 -0.4 -0.3 0 -0.4 0.4 / \plot 6.4 -0.4 6.3 0 6.4 0.4 / \plot -0.3 -.8 3 -.3 6.3 -.8 / \endpicture}\ ] ] with @xmath478 .
recall that we know : _ there is precisely one non - periodic ( but biperiodic ) @xmath96-word , namely @xmath479 where @xmath480 and @xmath481 _
@xmath482 with @xmath483 and @xmath484 ( the word @xmath278 is obtained from @xmath65 by rotation , the word @xmath353 by rotation and inversion ) . here
we see the reason why we call these algebras the wind wheels : we consider the word @xmath485 as a pair of opposite rotor blades .
_ the restriction functor @xmath486 has the following properties : _
\(1 ) _ indecomposable modules are sent to indecomposable modules . _
\(2 ) _ corresponding modules on the two @xmath487-quivers which yield the bar become isomorphic , otherwise non - isomorphy is preserved .
_ \(3 ) _ the indecomposable @xmath35-modules which are not in the image of the functor are the string modules for words which contain @xmath488 as a subword .
_ note that @xmath489 is the closure of the bar @xmath483 , as defined in section 6 . in order to outline the cut - and - paste process , we start with the auslander - reiten components containing string modules , as shown above .
we assume that @xmath451 is the tube which contains at the boundary the simple @xmath444-modules @xmath490 as well as the serial module with composition factors @xmath491 , whereas @xmath457 is the tube which contains at the boundary the simple @xmath444-modules @xmath492 and the serial module with composition factors @xmath493 .
let us look at the full subcategories of @xmath470 with modules with support in @xmath494 on the one hand ( see the left picture ) as well as those with support in @xmath495 on the other hand ( the right picture ) and describe the role of the various modules inside @xmath470 : @xmath496 at 4.5 2.3 \endpicture } at 0 0 \put{\beginpicture \multiput{$\bullet$ } at 0 0 1 1 2 0 2 2 3 1 3 3 4 0 4 2 5 1 6 0 / \plot 0 0 3 3 6 0 / \plot 1 1 2 0 4 2 / \plot 2 2 4 0 5 1 / \setshadegrid span < 0.7 mm > \hshade 0 0 4 2 2 2 / \plot 0.3 -0.3 2.3 1.7 / \plot -.3 0.3 1.7 2.3 / \circulararc 180 degrees from 2.3 1.7 center at 2 2 \circulararc 180 degrees from -.3 0.3 center at 0 0 \plot 3.7 1.7 5.7 -.3 / \plot 4.3 2.3 6.3 .3 / \circulararc 180 degrees from 4.3 2.3 center at 4 2 \circulararc 180 degrees from 5.7 -.3 center at 6 0 \circulararc 360 degrees from 3.3 3.3 center at 3 3 \plot 3.3 1.3 4.9 -.3 1.1 -.3 2.7 1.3 / \circulararc 90 degrees from 3.3 1.3 center at 3 1 \put{$1'$ } at 0 -.7 \put{$2'$ } at 2 -.7 \put{$3'$ } at 4 -.7 \put{$4 $ } at 6 -.7 \put{$m(b')$ } at 3 3.8 \put{${\operatorname{rad}}m(b')$ } [ r ] at 1.5 2.3 \endpicture } at 9 0 \endpicture}\ ] ] in the left picture , the shaded area marks those modules which belong to @xmath451 , whereas the remaining modules ( those which form the left boundary ) are projective , thus in @xmath497 in the right picture , the shaded area marks those modules which belong to @xmath457 , the remaining ones are injective , thus in @xmath498 according to property ( 2 ) , any module of the left triangle is identified under @xmath181 with the corresponding module of the right triangle .
let us look at the various components of @xmath470 which contain string modules .
we will add bullets @xmath499 in order to mark the position of the indecomposables with support contained either in @xmath494 or else in @xmath500 we are going to cut these components into suitable pieces : these are the dashed areas seen in the pictures .
first , we exhibit the preprojective component ( left ) and the preinjective component ( right ) : @xmath501 next , the two regular components ( the boundary of any of the two components contains three simple modules ; they are labeled ) : @xmath502 as we have mentioned , under the restriction functor @xmath469 , some serial @xmath444-modules become isomorphic , namely , the ten @xmath444-modules with support in in the subquiver with vertices @xmath494 are identified with the corresponding ten @xmath444-modules with support in in the subquiver with vertices @xmath495 . in a first step ,
we make the identification of the nine pairs consisting of modules of length at most 3 .
we obtain the following partial translation quiver : @xmath503 here we denote by @xmath504 the rays coming from @xmath505 , and so on . note that we did not yet identify the points labeled @xmath237 and @xmath455 , these are @xmath444-modules of length 4 which are identified under the restriction functor .
now let us make this last identification , and insert the @xmath35-modules which are not in the image of @xmath506 @xmath507 what are the additional modules ?
these are the string modules for words which contain a completed bar as a subword .
these modules form quarters ( as introduced in @xcite ) , namely the four shaded areas in the picture above .
the four quarters can be rearranged in order to be parts of a tile , similar to those exhibited in @xcite , p 54f , this will be explained in the next section .
of special interest seem to be the four encircled module , the corner modules for the quarters
. it may be worthwhile to identify explicitly the four corner modules ( in the presentation of this component given above , we have encircled these modules ) : @xmath508 in our example both the socle and the top of all corner modules are of length 2 . in general , the corner modules for the quarters * i * and * iv * may have a socle of length 3 , and dually , the corner modules for the quarter * iii * and * iv * may have a top of length 3 , as the following description shows : for the quarter * i * , the corner module @xmath509 is obtained from @xmath510 by adding hooks on the left and on the right .
dually , for the quarter * iii * , the corner module @xmath511 is obtained from @xmath454 by adding cohooks on the left and on the right . for the quarter * iv * we obtain the corner module @xmath512 by adding to @xmath237 a hook on the left , a cohook on the right . in our case
we have @xmath513 , where @xmath514 is the completion of @xmath236 . the corner module for the quarter * ii * has been denoted here by @xmath515 , in our example , we start with @xmath516 and add a cohook on the left and a hook on the right , in order to obtain @xmath383 however , this rule makes sense only in case @xmath516 is non - zero , thus in case the bar module @xmath237 is of length at least @xmath36 . in general ,
let @xmath517 be the boundary module in @xmath451 which has the same socle as @xmath237 and @xmath518 the boundary module in @xmath457 which has the same top as @xmath237
. then @xmath383 has a filtration @xmath519 with @xmath520 in case @xmath237 is of length @xmath140 , say @xmath521 where @xmath10 is an arrow , then we deal with an exact sequence @xmath522 this is one of the auslander - reiten sequences involving string modules and having an indecomposable middle term , namely that corresponding to the arrow@xmath10 , see @xcite .
this description of the corner modules shows that all of them are related to the following auslander - reiten sequence @xmath523 for @xmath524 , where @xmath525 is the annihilator of @xmath237 .
recall that we have used the auslander - reiten quiver of @xmath524 as our gluing device , let us mark the auslander - reiten sequence in question : @xmath526 at 1.5 2.2 \put{$m(b)/{\operatorname{soc}}$ } [ l ] at 4.5 2.2 \put{${\operatorname{rad}}m(b)/{\operatorname{soc}}$ } at 3 .6 \put { } at 0 -.3 \endpicture}\ ] ] in case @xmath237 is of length at least @xmath36 we deal with a square , if it is of length 2 , then with a triangle . in section 13.2
we will see in which way this square or triangle is enlarged in @xmath527 .
now we consider the general case of @xmath39 bars .
any of the components @xmath528 will be cut into @xmath39 pieces , and always the pieces will be indexed the the set @xmath529 of the direct bars .
write @xmath530 where all the words @xmath315 start with a direct letter and end with an inverse letter , and such that any @xmath315 ends with an inverse bar , say @xmath531 .
we denote by @xmath532 the cyclic permutation with @xmath533 similar to the case @xmath534 , we remove arrows from the preprojective component , but now we want to retain @xmath39 connected pieces @xmath535 with @xmath536 .
the piece @xmath535 is supposed to contain the projective modules starting with @xmath510 and ending with @xmath537 .
here is a picture of @xmath535 : @xmath538 next , consider the regular component @xmath451 ; according to our convention , this is the component which contains the simple modules @xmath539 .
again , we remove arrows in order to obtain @xmath39 pieces consisting of full corays ; the piece @xmath540 with index @xmath236 shall contain the modules @xmath541 up to @xmath542 .
@xmath543 similarly , we consider a word @xmath278 obtained from @xmath65 by cyclic rotation , such that @xmath544 where any @xmath545 starts with an inverse letter and ends with a direct letter , and ends in a ( direct ) bar , say the bar @xmath546 and we denote by @xmath547 the permutation which sends @xmath546 to @xmath548 now we cut the preinjective component in order to get pieces made up of corays .
the piece @xmath549 has to contain @xmath454 up to @xmath550 .
here is a picture of @xmath549 .
@xmath551 finally , we consider the regular component which contains the simple modules @xmath552 .
again , we remove arrows in order to obtain @xmath39 pieces consisting now of full rays .
the piece @xmath553 indexed by @xmath236 contains the rays starting at @xmath554 up to @xmath555 @xmath556 altogether we have cut the four string components of @xmath470 into flat pieces : any such component yields @xmath39 pieces .
the gluing of these pieces is done by identifying @xmath444-modules which become isomorphic under the restriction functor @xmath557 ( and finally we will have to add various quarters ) .
as in the case @xmath534 , we first will look at the proper subfactors of the bars . identifying the corresponding @xmath444-modules
, we obtain partial translation quivers which are planar : any of the components @xmath558 has been cut into @xmath39 pieces , and the identification process will use one piece of each kind , in order to obtain @xmath39 planar partial translation quivers of the following form ( @xmath559 ) : @xmath560 it is important to observe that in contrast to the case @xmath534 , the modules labeled @xmath455 and @xmath561 ( corresponding to bars @xmath562 and @xmath563 ) now may be different ! we obtain in this way a permutation @xmath393 of the bars such that @xmath564 .
but we know that @xmath565 and @xmath566 now @xmath567 means that @xmath568 , thus @xmath569(b').\ ] ] this shows : @xmath570.\ ] ] of course , as we have mentioned already , we still have to add the indecomposable @xmath35-modules which do not belong to the image of @xmath181 . we know that there are precisely @xmath39 non - periodic ( but biperiodic ) @xmath96-words ; they give rise to @xmath571 quarters which have to be inserted as in the case @xmath534 . before making the final identifcations ,
let us attach the quarters of type * iv * to the pieces of the form @xmath549 .
in this way , we obtain @xmath39 partial translation quivers of the form @xmath572 ( the visualization on the right hand side takes into account the embedding of these auslander - reiten components into the corresponding `` auslander - reiten quilt '' which we will discuss in the next section . ) the partial translation quivers are sewn together in the same way as one constructs the riemann surfaces of the @xmath27-the root functions in complex analysis ( taking into account the permutation @xmath393 ) . for example , we may obtain a @xmath36-ramified component which roughly will have the following shape : @xmath573 we will call such a component with @xmath574 leaves an _ @xmath574-ramified component of type @xmath575 . _
let @xmath35 be a wind wheel with @xmath39 bars .
then : any non - regular auslander - reiten component is an @xmath574-ramified component of type @xmath575 with @xmath576 if @xmath577 are the non - regular auslander - reiten components of @xmath35 and @xmath578 is @xmath579-ramified , for @xmath580 , then @xmath581 we may assume that @xmath582 , thus we deal with a partition and we call this partition @xmath583 the _ ramification sequence _ of @xmath35 .
we are going to present a wind wheel with @xmath39 bars which has @xmath39 non - regular auslander - reiten components ( all being necessarily @xmath34-ramified : the ramification sequence is @xmath584 ) .
here is the quiver : @xmath585 from 0 0.8 to 0 0.2 \circulararc 155 degrees from 0 2 center at 0 1.5 \circulararc -155 degrees from 0 2 center at 0 1.5 \arrow < 1.5
mm > [ 0.25,0.75 ] from -0.22 1.05 to -0.17 1.02 \setdots < .7 mm >
\setquadratic \plot -.4 1.4 0 1.3 .4 1.4 / \endpicture } at 0 0 \put{\beginpicture \put { } at 0 2 \put{$2 $ } at 0 0 \put{$3 $ } at 0 1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0 0.8 to 0 0.2 \circulararc 155 degrees from 0 2 center at 0 1.5 \circulararc -155 degrees from 0 2 center at 0 1.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.22 1.05 to -0.17 1.02 \setdots < .7 mm >
\setquadratic \plot -.4 1.4 0 1.3 .4 1.4 / \endpicture } at 2 0 \put{\beginpicture \put { } at 0 2 \put{$2t\!-\!2 $ } at 0 0 \put{$2t\!-\!1 $ } at 0 1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0 0.8 to 0 0.2 \circulararc 123 degrees from 0 2 center at 0 1.5 \circulararc -125 degrees from 0 2 center at 0 1.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.47 1.33 to -0.42 1.25 \setdots < .7 mm >
\setquadratic \plot -.4 1.4 0 1.3 .4 1.4 / \endpicture } at 6 0 \plot -0.3 -.95 -.6 -.95 / \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.6 -.95 to 0.4 -.95 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3 -.95 to 2.4 -.95 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.5 -.95 to 5 -.95 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7 -.95 to 6.5 -.95 \circulararc 180 degrees from -.6 -.95 center at -.6 -1.3 \circulararc -180 degrees from 7 -.95 center at 7 -1.3 \plot -.6 -1.65 7 -1.65 / \setdots < .7 mm >
\setquadratic \plot -.4 -1.1 0 -1.3 .4 -1.1 / \plot 1.6 -1.1 2 -1.3 2.4 -1.1 / \plot 5.4 -1.1 6 -1.3 6.6 -1.1 / \plot -0.4 0 -0.2 -.45 -.4 -.8 / \plot 1.6 0 1.8 -.45 1.6
-.8 / \plot 5.5 0 5.7 -.45 5.5 -.8 / \endpicture}\ ] ] for example , for @xmath586 , the primitive cyclic word is @xmath587 all the non - regular auslander - reiten components of this algebra look similar , here is one of these components : @xmath588 from 0.3 4.3 to 0.7 4.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 4.3 to 2.7 4.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 5.3 to 1.7 5.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.3 5.3 to 3.7 5.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.3 6.3 to 4.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.3 6.3 to 6.6 6.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from 8.3 6.3 to 8.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.3 6.7 to 5.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.3 6.7 to 7.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 9.3 6.7 to 9.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 10.3 5.7 to 10.7 5.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 11.3 4.7 to 11.7 4.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 5.7 to 12.7 5.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 4.7 to 13.7 4.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.3 0.3 to 4.7 0.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.3 0.3 to 6.7 0.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 8.3 0.3 to 8.7 0.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.3 0.7 to 5.7 0.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.3 0.7 to 7.7 0.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 9.3 0.7 to 9.7 0.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.3 1.3 to 5.7 1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.3 1.3 to 7.7 1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.3 1.7 to 6.7 1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 8.3 1.7 to 8.7 1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.3 2.3 to 6.7 2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.3 2.7 to 7.7 2.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 6.3 to 0.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 6.3 to 2.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 10.3 6.3 to 10.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 6.3 to 12.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 4.3 to 12.7 4.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 2.3 to 0.7 2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 11.3 5.3 to 11.7 5.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 5.3 to 13.7 5.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 3.3 to 13.7 3.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 3.3 to 1.7 3.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 3.7 to 0.7 3.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 5.7 to 0.7 5.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 4.7 to 1.7 4.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 6.7 to 1.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 5.7 to 2.7 5.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.3 6.7 to 3.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 11.3 6.7 to 11.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 6.7 to 13.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 3.7 to 12.7 3.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 2.7 to 13.7 2.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.4 3.9 to 6.7 3.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.4 2.9 to 5.7 2.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.4 1.9 to 4.7 1.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.2 0.8 to 3.7 0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.3 3.1 to 11.4 3.9 \arrow < 1.5 mm > [ 0.25,0.75 ] from 8.3 2.1 to 12.4
2.9 \arrow < 1.5 mm > [ 0.25,0.75 ] from 9.3 1.1 to 13.4 1.9 \plot 10.3 0.1 13.8 0.8 / \setdots<1 mm > \plot 4.4 6 5.6 6 / \plot 6.4 6 7.6 6 / \plot 8.4 6 9.6 6 / \setshadegrid span < 0.7 mm > \endpicture}\ ] ] the inserted quarters are labeled * i , ii , iii , iv , * with a bar @xmath236 as an index : all the modules in such a quarter are of the form @xmath589 , where @xmath157 is a word which contains @xmath514 as a subword ( such a quarter will later be seen as part of the tile @xmath590 ) .
first , let us present an example with a @xmath36-ramified component .
here is the quiver with the zero relations of length 2 ( in addition all paths of length 3 are zero relations ) : @xmath591 from 0 0.8 to 0 0.2 \circulararc 155 degrees from 0 2 center at 0 1.5 \circulararc -155 degrees from 0 2 center at 0 1.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.22 1.05 to -0.17 1.02 \setdots
< .7 mm > \setquadratic \plot -.4 1.4 0 1.3 .4 1.4 / \endpicture } at 0 0 \put{\beginpicture \put { } at 0 2 \put{$2 $ } at 0 0 \put{$3 $ } at 0 1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0 0.8 to 0 0.2 \circulararc 155 degrees from 0 2 center at 0 1.5 \circulararc -155 degrees from 0 2 center at 0 1.5 \arrow < 1.5
mm > [ 0.25,0.75 ] from -0.22 1.05 to -0.17 1.02 \setdots < .7 mm >
\setquadratic \plot -.4 1.4 0 1.3 .4 1.4 / \endpicture } at 2 0 \put{\beginpicture \put { } at 0 2 \put{$4 $ } at 0 0 \put{$5 $ } at 0 1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0 0.8 to 0 0.2 \endpicture } at 4 0 \put{\beginpicture \put { } at 0 2 \put{$6 $ } at 0 0 \put{$7 $ } at 0 1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0 0.8 to 0 0.2 \circulararc 155 degrees from 0 -1 center at 0 -.5 \circulararc -155 degrees from 0 -1 center at 0 -.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.22 -.05 to -0.17 .02 \setdots < .7 mm > \setquadratic \plot -.4 -.4 0 -.3 .4 -.4 / \endpicture } at 6 0 \setquadratic \plot 0 -1.3 2 -2 4 -1.3 / \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.6 -.95 to 0.4 -.95 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.6 -.95 to 2.4 -.95 \setquadratic \plot 5.6 0.4 5 0.6 4.4 0.4 / \plot 5.6 0.0 5 -.2 4.4 0.0 / \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.45 0.43 to 4.4 .4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.55 .0 to 5.6 .03 \setdots < .7 mm >
\setquadratic \plot .6 -1.5 0.2 -1.2 .6 -1.1 / \plot 1.6 -1.1 2 -1.3 2.4 -1.1 / \plot 3.4 -1.5 3.8 -1.2 3.4 -1.1 / \plot 4.7 0.4 4.4 0.2 4.7 0.0 / \plot 5.3 0.4 5.6 0.2 5.3 0.0 / \endpicture}\ ] ] the primitive cyclic word is @xmath592 let us exhibit one part of the @xmath36-ramified auslander - reiten component ( as before , the inserted quarters are labeled * i , ii , iii , iv , * with a bar @xmath236 as an index ) : @xmath593 from 11.3 3.3 to 11.6 3.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 2.3 to 12.6 2.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 1.3 to 13.7 1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 14.3 .3 to 14.7 .7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 11.3 2.7 to 11.7 2.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 1.7 to 12.7 1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 .7 to 13.7 .3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 4.3 to 0.7 4.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 4.3 to 2.7 4.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 5.3 to 1.7 5.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.3 5.3 to 3.7 5.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.3 6.3 to 4.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.3 6.3 to 6.6 6.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from 8.3 6.3 to 8.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.3 6.7 to 5.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.3 6.7 to 7.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 9.3 6.7 to 9.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 10.3 5.7 to 10.7 5.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 11.3 4.7 to 11.7 4.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 5.7 to 12.7 5.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 4.7 to 13.7 4.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 0.3 to 0.7 0.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 0.3 to 2.7 0.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.3 0.3 to 4.7 0.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 0.7 to 1.7 0.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.3 0.7 to 3.7 0.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.3 0.7 to 5.7 0.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 1.3 to 1.7 1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.3 1.3 to 3.7 1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 1.7 to 2.7 1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.3 1.7 to 4.7 1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 2.3 to 2.7 2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.3 2.7 to 3.7 2.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 6.3 to 0.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 6.3 to 2.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 10.3 6.3 to 10.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 6.3 to 12.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 4.3 to 12.7 4.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 2.3 to 0.7 2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 11.3 5.3 to 11.7 5.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 5.3 to 13.7 5.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 3.3 to 13.7 3.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 3.3 to 1.7 3.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 3.7 to 0.7 3.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 5.7 to 0.7 5.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 4.7 to 1.7 4.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 6.7 to 1.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 5.7 to 2.7 5.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.3 6.7 to 3.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 11.3 6.7 to 11.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 6.7 to 13.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 3.7 to 12.7 3.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 2.7 to 13.7 2.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.4 3.6 to 2.7 3.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.4 2.6 to 1.7
2.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.4 1.6 to 0.7 1.3 \setdots<1 mm > \plot 4.4 6 5.6 6 / \plot 6.4 6 7.6 6 / \plot 8.4 6 9.6 6 / \setshadegrid span < 0.7 mm > \endpicture}\ ] ] this concerns the part of the component containing the module @xmath594 the leaves containing the modules @xmath595 and @xmath596 look similar .
these three leaves together form a component , namely a 3-ramified component of type @xmath597 .
@xmath598 from -.7 -0.7 to -.3 -.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -1.6 to -1.3 -1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -2.4 to -1.3 -2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.4 -0.4 to 6.7 -.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -1.4 to 7.7 -1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -2.7 to 7.7 -2.4 \setshadegrid span < .7 mm >
\vshade -3 -4 1 < z , z , , > -2 -4 1 / \vshade -2 -4 -2 < z , z , , > -1 -4 -3 / \vshade -2 -2 1 < z , z , , > -0 0 1 < z , z , , > 6 0 1 < z , z , , > 7.99 -2 1 < z , z , , > 8 -4 1 < z , z , , > 9 -4 1 / \vshade 7 -4 -3 < z , z , , > 8 -4 -2 / \multiput { } at 0 -4 0 1 / \endpicture } at 0 0 \put{\beginpicture \put{$5 $ } at 0 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 7 $ } at 0 2 \put{${\scriptstyle } 6 $ } at 1 1 \put{${\scriptstyle } 6 $ } at 2 0 \endpicture } at 2 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 7 $ } at 0 2 \put{${\scriptstyle } 5 $ } at 1 1 \put{${\scriptstyle } 4 $ } at 2 0 \endpicture } at 4 0 \put{$0 $ } at 6 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 2 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 2 0 \endpicture } at 7 -1 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 2 $ } at 0 0 \put{${\scriptstyle } 3 $ } at 1 1 \put{${\scriptstyle } 3 $ } at 2 2 \put{${\scriptstyle } 2 $ } at 3 1 \put{${\scriptstyle } 0 $ } at 4 0 \endpicture } at 8 -2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 2 $ } at 1 1 \put{${\scriptstyle } 3 $ } at 2 2 \endpicture } at 7 -3 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 5 $ } at 1 1 \put{${\scriptstyle } 7 $ } at 2 0 \endpicture } at -1 -1 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 5 $ } at 0 2 \put{${\scriptstyle } 7 $ } at 1 1 \put{${\scriptstyle } 6 $ } at 2 0 \put{${\scriptstyle } 6 $ } at 3 1 \put{${\scriptstyle } 7 $ } at 4 2 \endpicture } at -2 -2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 6 $ } at 1 1 \put{${\scriptstyle } 7 $ } at 2 2 \endpicture } at -1 -3 \setdots<1 mm > \plot 0.4 0 1.6 0 / \plot 2.4 0 3.6 0 / \plot 4.4 0 5.6 0 / \setsolid \arrow < 1.5 mm > [ 0.25,0.75 ] from -.7 -0.7 to -.3 -.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -1.6 to -1.3 -1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -2.4 to -1.3 -2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.4 -0.4 to 6.7 -.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -1.4 to 7.7 -1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -2.7 to 7.7 -2.4 \setshadegrid span < .7 mm > \vshade -3 -4 1 < z , z , , > -2 -4 1 / \vshade -2 -4 -2 < z , z , , > -1 -4 -3 / \vshade -2 -2 1 < z , z , , > -0 0 1 < z , z , , > 6 0 1 < z , z , , > 7.99 -2 1 < z , z , , > 8 -4 1 < z , z , , > 9 -4 1 / \vshade 7 -4 -3 < z , z , , > 8 -4 -2 / \multiput { } at 0 -4 0 1 / \endpicture } at 0 -6 \put{\beginpicture \put{$3 $ } at 0 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at 0 2 \put{${\scriptstyle } 2 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 2 0 \endpicture } at 2 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at 0 2 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 2 0 \endpicture } at 4 0 \put{$2 $ } at 6 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 2 $ } at 2 0 \endpicture } at 7 -1 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 6 $ } at 0 0 \put{${\scriptstyle } 7 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 4 $ } at 3 1 \put{${\scriptstyle } 2 $ } at 4 0 \endpicture } at 8 -2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 6 $ } at 1 1 \put{${\scriptstyle } 7 $ } at 2 2 \endpicture } at 7 -3 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at 1 1 \put{${\scriptstyle } 3 $ } at 2 0 \endpicture } at -1 -1 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at 0 2 \put{${\scriptstyle } 3 $ } at 1 1 \put{${\scriptstyle } 2 $ } at 2 0 \put{${\scriptstyle } 4 $ } at 3 1 \put{${\scriptstyle } 5 $ } at 4 2 \endpicture } at -2 -2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \endpicture } at -1 -3 \setdots<1 mm > \plot 0.4 0 1.6 0 / \plot 2.4 0 3.6 0 / \plot 4.4 0 5.6 0 / \setsolid \arrow < 1.5 mm > [ 0.25,0.75 ] from -.7 -0.7 to -.3 -.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -1.6 to -1.3 -1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -2.4 to -1.3 -2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.4 -0.4 to 6.7 -.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -1.4 to 7.7 -1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -2.7 to 7.7 -2.4 \setshadegrid span < .7 mm > \vshade -3 -4 1 < z , z , , > -2 -4 1 / \vshade -2 -4 -2 < z , z , , > -1 -4 -3 / \vshade -2 -2 1 < z , z , , > -0 0 1 < z , z , , > 6 0 1 < z , z , , > 7.99 -2 1 < z , z , , > 8 -4 1 < z , z , , > 9 -4 1 / \vshade 7 -4 -3 < z , z , , > 8 -4 -2 / \multiput { } at 0 -4 0 1 / \endpicture } at 0 -12 \endpicture}\ ] ] in addition there is a second non - regular auslander - reiten component , namely the component containing the module @xmath599 ; it is @xmath34-ramified .
the boundary looks as follows : @xmath600 from -.7 -0.7 to -.3 -.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -1.6 to -1.3 -1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.4 -2.1 to 2.7 -3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.4 -0.4 to 6.7 -.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -1.4 to 7.7 -1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.4 -3 to 7.4 -2.1 \setshadegrid span < .7 mm > \vshade -3 -4 1 < z , z , , > -2 -4 1 / \vshade -2 -4 -2 < z , z , , > 3 -4 -3 / \vshade -2 -2 1 < z , z , , > -0 0 1 < z , z , , > 6 0 1 < z , z , , > 7.99 -2 1 < z , z , , > 8 -4 1 < z , z , , > 9 -4 1 / \vshade 3 -4 -3 < z , z , , > 8 -4 -2 / \multiput { } at 0 -4 0 1 / \endpicture}\ ] ] we use galois coverings of this wind wheel in order to exhibit wind wheels with arbitrary ramification .
let us consider the galois covering obtained by an @xmath601-covering of the cycle of length @xmath140 , thus we deal with a wind wheel of the following shape ( for @xmath602 , we have to identify the upper left hand arrow with the upper right hand arrow in order to have an arrow @xmath603 ) : @xmath604 from 0 -.2 to 0 -.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2 -.2 to 2 -.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4 -.2 to 4 -.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6 -.2 to 6 -.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 8 -.2 to 8 -.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 10 -.2 to 10 -.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1 -2.8 to 1 -2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3 -2.8 to 3 -2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5 -2.8 to 5 -2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7 -2.8 to 7 -2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 9 -2.8 to 9 -2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 11 -2.8 to 11 -2.2 \circulararc -155 degrees from 0 -2 center at 0 -1.5 \circulararc 155 degrees from 0 -2 center at 0 -1.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.22 -1.04 to 0.17 -1.02 \circulararc -155 degrees from 4 -2 center at 4 -1.5 \circulararc 155 degrees from 4 -2 center at 4 -1.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.22 -1.04 to 4.17 -1.02 \circulararc -155 degrees from 8 -2 center at 8 -1.5 \circulararc 155 degrees from 8 -2 center at 8 -1.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from 8.22 -1.04 to 8.17 -1.02 \setdots < .7 mm > \setquadratic \plot -.4 -1.4 0 -1.3 .4 -1.4 / \plot 3.6 -1.4 4 -1.3 4.4 -1.4 / \plot 7.6 -1.4 8 -1.3 8.4 -1.4 / \setsolid \circulararc -155 degrees from 1 -4 center at 1 -3.5 \circulararc 155 degrees from 1 -4 center at 1 -3.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.22 -3.04 to 1.17 -3.02 \circulararc -155 degrees from 3 -4 center at 3 -3.5 \circulararc 155 degrees from 3 -4 center at 3 -3.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.22 -3.04 to 3.17 -3.02 \circulararc -155 degrees from 5 -4 center at 5 -3.5 \circulararc 155 degrees from 5 -4 center at 5 -3.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.22 -3.04 to 5.17 -3.02 \circulararc -155 degrees from 7 -4 center at 7 -3.5 \circulararc 155 degrees from 7 -4 center at 7 -3.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.22 -3.04 to 7.17 -3.02 \circulararc -155 degrees from 9 -4 center at 9 -3.5 \circulararc 155 degrees from 9 -4 center at 9 -3.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from 9.22 -3.04 to 9.17 -3.02 \circulararc -155 degrees from 11 -4 center at 11 -3.5 \circulararc 155 degrees from 11 -4 center at 11 -3.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from 11.22 -3.04 to 11.17 -3.02 \setdots < .7 mm > \setquadratic \plot .6 -3.4 1 -3.3 1.4 -3.4 / \plot 2.6 -3.4 3 -3.3 3.4 -3.4 / \plot
4.6 -3.4 5 -3.3 5.4 -3.4 / \plot 6.6 -3.4 7 -3.3 7.4 -3.4 / \plot 8.6 -3.4 9 -3.3 9.4 -3.4 / \plot 10.6 -3.4 11 -3.3 11.4 -3.4 / \setsolid \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.7 -1.3 to 1.3 -1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.7 -1.7 to 2.3 -1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.4 -2 to 2.6 -2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.7 -1.3 to 5.3 -1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.7 -1.7 to 6.3 -1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.4 -2 to 6.6 -2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 9.7 -1.3 to 9.3 -1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 10.7 -1.7 to 10.3 -1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 9.4 -2 to 10.6 -2 \arrow < 1.5 mm > [ 0.25,0.75 ] from -.4 0 to -1 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.6 0 to .4 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.6 0 to 2.4 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.6 0 to 4.4 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.6 0 to 6.4 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 9.6 0 to 8.4 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 11 0 to 10.4 0 \setdots < .7 mm >
\plot 2.45 -1.5 2 -1.4 1.55 -1.5 / \plot 6.45 -1.5 6 -1.4 5.55 -1.5 / \plot 10.45 -1.5 10 -1.4 9.55 -1.5 / \plot 1.5 -1.6 1.4 -1.8 1.7 -1.9 / \plot 2.5 -1.6 2.6 -1.8 2.3 -1.9 / \plot 5.5 -1.6 5.4 -1.8 5.7 -1.9 / \plot
6.5 -1.6 6.6 -1.8 6.3 -1.9 / \plot 9.5 -1.6 9.4 -1.8 9.7 -1.9 / \plot 10.5 -1.6 10.6 -1.8 10.3 -1.9 / \plot -.4 0.2 0 0.3 .4 .2 / \plot 1.6 0.2 2 0.3 2.4 .2 / \plot 3.6 0.2 4 0.3 4.4 .2 / \plot 5.6 0.2 6 0.3 6.4 .2 / \plot 7.6 0.2 8 0.3 8.4 .2 / \plot 9.6 0.2 10 0.3 10.4 .2 / \endpicture}\ ] ] the corresponding primitive cyclic word is @xmath605 let us show the leaves which contain the modules @xmath606 , they are quite similar to those seen above the only difference occurs on the right hand side of the upper leaf and in the upper row of the middle leaf : @xmath607 from -.7 -0.7 to -.3 -.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -1.6 to -1.3 -1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -2.4 to -1.3 -2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.4 -0.4 to 6.7 -.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -1.4 to 7.7 -1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -2.7 to 7.7 -2.4 \setshadegrid span < .7 mm > \vshade -3 -4 1 < z , z , , > -2 -4 1 / \vshade -2 -4 -2 < z , z , , > -1 -4 -3 / \vshade -2 -2 1 < z , z , , > -0 0 1 < z , z , , > 6 0 1 < z , z , , > 7.99 -2 1 < z , z , , > 8 -4 1 < z , z , , > 9 -4 1 / \vshade 7 -4 -3 < z , z , , > 8 -4 -2 / \multiput { } at 0 -4 0 1 / \endpicture } at 0 0 \put{\beginpicture \put{$5 $ } at 0 0 \put { \beginpicture \setcoordinatesystem units < .2cm,.18 cm > \put{${\scriptstyle } { } ' 7 $ } at 0 2 \put{${\scriptstyle } { } ' 6 $ } at 1 1 \put{${\scriptstyle } { } ' 6 $ } at 2.1 0 \endpicture } at 2 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } { } ' 7 $ } at 0 2 \put{${\scriptstyle } 5 $ } at 1 1 \put{${\scriptstyle } 4 $ } at 2 0 \endpicture } at 4 0 \put{$0 $ } at 6 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 2 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 2 0 \endpicture } at 7 -1 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 2 $ } at 0 0 \put{${\scriptstyle } 3 $ } at 1 1 \put{${\scriptstyle } 3 $ } at 2 2 \put{${\scriptstyle } 2 $ } at 3 1 \put{${\scriptstyle } 0 $ } at 4 0 \endpicture } at 8 -2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 2 $ } at 1 1 \put{${\scriptstyle } 3 $ } at 2 2 \endpicture } at 7 -3 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 5 $ } at 1 1 \put{${\scriptstyle } 7 $ } at 2 0 \endpicture } at -1 -1 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 5 $ } at 0 2 \put{${\scriptstyle } 7 $ } at 1 1 \put{${\scriptstyle } 6 $ } at 2 0 \put{${\scriptstyle } 6 $ } at 3 1 \put{${\scriptstyle } 7 $ } at 4 2 \endpicture } at -2 -2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 6 $ } at 1 1 \put{${\scriptstyle } 7 $ } at 2 2 \endpicture } at -1 -3 \setdots<1 mm > \plot 0.4 0 1.6 0 / \plot 2.4 0 3.6 0 / \plot 4.4 0 5.6 0 / \setsolid \arrow < 1.5 mm > [ 0.25,0.75 ] from -.7 -0.7 to -.3 -.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -1.6 to -1.3 -1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -2.4 to -1.3 -2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.4 -0.4 to 6.7 -.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -1.4 to 7.7 -1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -2.7 to 7.7 -2.4 \setshadegrid span < .7 mm > \vshade -3 -4 1 < z , z , , > -2 -4 1 / \vshade -2 -4 -2 < z , z , , > -1 -4 -3 / \vshade -2 -2 1 < z , z , , > -0 0 1 < z , z , , > 6 0 1 < z , z , , > 7.99 -2 1 < z , z , , > 8 -4 1 < z , z , , > 9 -4 1 / \vshade 7 -4 -3 < z , z , , > 8 -4 -2 / \multiput { } at 0 -4 0 1 / \endpicture } at 0 -6 \put{\beginpicture \put{$3 $ } at 0 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at 0 2 \put{${\scriptstyle } 2 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 2 0 \endpicture } at 2 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at 0 2 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 2 0 \endpicture } at 4 0 \put{$2 $ } at 6 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 2 $ } at 2 0 \endpicture } at 7 -1 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 6 $ } at 0 0 \put{${\scriptstyle } 7 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 4 $ } at 3 1 \put{${\scriptstyle } 2 $ } at 4 0 \endpicture } at 8 -2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 6 $ } at 1 1 \put{${\scriptstyle } 7 $ } at 2 2 \endpicture } at 7 -3 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at 1 1 \put{${\scriptstyle } 3 $ } at 2 0 \endpicture } at -1 -1 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at 0 2 \put{${\scriptstyle } 3 $ } at 1 1 \put{${\scriptstyle } 2 $ } at 2 0 \put{${\scriptstyle } 4 $ } at 3 1 \put{${\scriptstyle } 5 $ } at 4 2 \endpicture } at -2 -2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \endpicture } at -1 -3 \setdots<1 mm > \plot 0.4 0 1.6 0 / \plot 2.4 0 3.6 0 / \plot 4.4 0 5.6 0 / \setsolid \arrow < 1.5 mm > [ 0.25,0.75 ] from -.7 -0.7 to -.3 -.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -1.6 to -1.3 -1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -2.4 to -1.3 -2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.4 -0.4 to 6.7 -.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -1.4 to 7.7 -1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -2.7 to 7.7 -2.4 \setshadegrid span < .7 mm > \vshade -3 -4 1 < z , z , , > -2 -4 1 / \vshade -2 -4 -2 < z , z , , > -1 -4
-3 / \vshade -2 -2 1 < z , z , , > -0 0 1 < z , z , , > 6 0 1 < z , z , , > 7.99 -2 1 < z , z , , > 8 -4 1 < z , z , , > 9 -4 1 / \vshade 7 -4 -3 < z , z , , > 8 -4 -2 / \multiput { } at 0 -4 0 1 / \endpicture } at 0 -12 \endpicture}\ ] ] as before , the upper leaf and the middle leaf are sewn together ( both contain the module @xmath608 ) , similarly , the middle leaf and the lower leaf are sewn together ( both contain the module @xmath609 ) . but the change of the right hand side of the upper leaf is important , since it means that the upper leave and the lower leave no longer are sewn together ( the lower leaf contains the module @xmath610 , the upper one the shifted module @xmath611 ) .
it follows that for the @xmath601-fold covering @xmath612 leaves are sewn together and form a @xmath612-ramified component ( this is the auslander - reiten component which contains the modules @xmath613 and their shifts under the galois group ) . what happens with the remaining non - regular component ( the @xmath34-ramified one ) ?
thus , let us start to calculate the auslander - reiten component which contains the module @xmath614 .
here is the relevant part of the boundary : @xmath615 from -.7 -0.7 to -.3 -.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -1.6 to -1.3 -1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.4 -0.4 to 6.7 -.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -1.4 to 7.7 -1.7 \setshadegrid span < .7 mm > \vshade -3 -4 1 < z , z , , > -2 -4 1 / \vshade -2 -4 -2 < z , z , , > -1 -4 -3 / \vshade -2 -2 1 < z , z , , > -0 0 1 < z , z , , > 6 0 1 < z , z , , > 7.99 -2 1 < z , z , , > 8 -4 1 < z , z , , > 9 -4 1 / \vshade 7 -4 -3 < z , z , , > 8 -4 -2 / \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at 1 1 \put{${\scriptstyle } 1 $ } at 2 2 \endpicture } at 7 -3 \put { \beginpicture \setcoordinatesystem units < .2cm,.18 cm > \put{${\scriptstyle } 0'$ } at 1 1 \put{${\scriptstyle } 1'$ } at 2 2 \endpicture } at -1 -3 \arrow <
1.5 mm > [ 0.25,0.75 ] from -1.6 -2.4 to -1.3 -2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.4 -2.7 to 7.7 -2.4 \multiput { } at 0 -4 0 1 / \endpicture}\ ] ] it follows that the galois shifts of the module @xmath614 all lie in one component , and this is a component of type @xmath575 which is @xmath601-ramified .
this shows that any ramification does occur .
we have seen above , that the sewing of the leaves is accomplished via the permutation @xmath616.$ ] thus , we see : _ all the non - regular components are @xmath34-ramified if and only if the permutations @xmath617 and @xmath49 commute _ ( in particular , this will be the case if these permutations coincide , as in example 12.5 ) . in general , we see that we do not get all the possible permutations for @xmath393 .
the mathematics behind it , is as follows : in the symmetric group @xmath618 , we fix one @xmath39-cycle as @xmath49 and form for any @xmath39-cycle @xmath617 the commutator @xmath619 $ ] : _ these are the permutations which arise for the sewing procedure . _
a partition of @xmath39 is the ramifcation sequence of a wind wheel if and only if it is the cycle partition for the commutator of two @xmath39-cycles .
( by definition , the cycle partition of a permutation has as parts the lengths of the cycles when written as a product of disjoint cycles . ) for @xmath620 , the group @xmath618 is commutative , thus we get as @xmath393 only the identity .
this means : for @xmath621 , we always get two non - regular components , both being @xmath34-ramified . for @xmath622 ,
the group @xmath618 is no longer commutative , however the @xmath36-cycles commute , thus again the only commutator @xmath623 $ ] is the identity , thus again we see that we only get @xmath34-ramified components
. the first case where one can obtain an @xmath574-ramified component with @xmath624 is @xmath625 an explicit example has been discussed in 12.6 . for @xmath626
one checks easily that the possible ramification sequences are @xmath627 and @xmath628 . for @xmath586 , they are @xmath629 , @xmath630 and @xmath631 ( for example , the commutator of the permutations @xmath632 and @xmath633 has the cycle partition @xmath630 , that of @xmath632 and @xmath634 has the cycle partition @xmath629 ) . in particular , we see that for @xmath635 , there are no @xmath140-ramified components . for @xmath636 ,
the commutator of @xmath637 and @xmath638 has the cycle partition @xmath639 .
it seems that a description of the structure of the commutators @xmath640,$ ] where @xmath617 and @xmath49 are @xmath39-cycles in @xmath618 is not known ( but see the related investigations @xcite ) .
auslander - reiten quilts have been considered until now only for suitable special biserial algebras @xmath18 .
a general definition can be given in case @xmath18 is a 1-domestic special biserial algebra , see @xcite : the vertices are ( finite or infinite ) words , and there are arrows , meshes , but also a convergence relation .
the auslander - reiten quilt considers not only the indecomposable @xmath18-modules of finite length , but also related indecomposable @xmath18-modules of infinite length which are algebraically compact .
the main objective is to sew together auslander - reiten components which contain string modules , using @xmath93-words and @xmath96-words , the @xmath96-words yield tiles .
we recall from @xcite some considerations concerning the auslander - reiten quilt of a special biserial algebra .
the poset @xmath641 is the ordered sum of @xmath93 and @xmath642 ; inserting a limit point @xmath643 in the middle , we obtain the completion @xmath644
@xmath645 we may consider @xmath644 as a topological space , namely as a closed interval ; correspondingly , we will consider @xmath646 as a square or better as a lozenge .
the center @xmath647 is the vertex @xmath648 , the vertices on the diagonals ( some are marked by @xmath499 ) are the pairs @xmath649 and @xmath650 with @xmath651 . in a rather obvious way , we can consider @xmath652 also as a translation quiver and as in @xcite we will call it a _ tile _ , see the following picture on the left : @xmath653 the middle picture presents the translation subquiver @xmath654 obtained from the left picture by deleting the diagonals through the center .
it may be considered as the disjoint union of four parts , the _ quarters _ * i , ii , iii , iv , * see the right picture . here , we also have marked the four _ corners _ using circles .
given any bar @xmath236 , we obtain such a tile by considering all the ( finite or infinite ) words containing the completed bar @xmath514 , thus by considering all the ( finite or infinite ) subwords of @xmath485 .
the word @xmath485 itself will be the center of the tile , the infinite words form the diagonals through the center . if we look only at the finite words , we obtain in this way the four quarters . as we have seen in @xcite
, tiles can occur as hammocks ( the bridges in @xcite yield such a hammock ) . here
we encounter a different situation where tiles appear .
the tiles which occur when dealing with minimal representation - finite special biserial algebras have the special property that all the irreducible maps in the quarter * i * are monomorphisms , whereas those in the quarter * iii * are epimorphisms .
there is one bar @xmath236 , thus one tile @xmath590 .
the finite - dimensional modules in the tile are those which do not lie in the image of the restriction functor @xmath655 thus the string modules @xmath589 where @xmath157 is a finite word which contains the completed bar @xmath514 . any tile yields four quarters , and we have seen above that the corresponfing corner modules are related to the following auslander - reiten sequence of @xmath524 , where @xmath525 is the annihilator of @xmath237 ( in case @xmath237 is of length @xmath140 we deal with a triangle , otherwise with a square ) : @xmath526 at 1.5 2.2 \put{$m(b)/{\operatorname{soc}}$ } [ l ] at 4.5 2.2 \put{${\operatorname{rad}}m(b)/{\operatorname{soc}}$ } at 3 .6 \put { } at 0 -.3 \endpicture}\ ] ] we are going to outline in which way this square or triangle is enlarged in @xmath527 .
it is the tile @xmath652 which may be considered as being inserted into this square alternatively , we may say that we add a border to the tile : @xmath656 at -.6 2 \put{$m(b)/{\operatorname{soc}}$ } [ l ] at 4.6 2 \put{$m(b)$ } at 2 4.7 \put{${\operatorname{rad}}m(b)/{\operatorname{soc}}$ } at 2 -.7 \plot 1.8 -.2 2 0 2.2 -.2 / \plot 1.8 4.2 2 4 2.2 4.2 / \plot -.2 1.8 0 2 -.2 2.2 / \plot 4.2 2.2 4 2 4.2 1.8
/ \plot 0.9 1.1 0 2 0.9 2.9 / \plot 1.1 3.1 2 4 2.9 3.1 / \plot 3.1 2.9 4 2 3.1 1.1 / \plot 2.9 0.9 2 0 1.1 0.9 / \put{\bf i } at 1 2 \put{\bf ii } at 2 1 \put{\bf iii } at 3 2 \put{\bf iv } at 2 3 \multiput{$\bigcirc$ } at 2 0 0 2 2 4 4 2 / \multiput{$\bullet$ } at 1 1 1 3 3 1 3 3 0.8 0.8 0.8 3.2 3.2 0.8 3.2 3.2 / \multiput{$\blacksquare$ } at 2 2 / \plot 0.8 0.8 3.2 3.2 / \plot 0.8 3.2 3.2 0.8 / \put { } at 0 0 \put { } at 4 4 \endpicture } at 0 0 \put{\beginpicture \plot -.4 2 .7 0.9 / \plot .9 0.7 1.8 -.2 / \plot 2.2 -.2 3.1 0.7 / \plot 3.3 0.9 4.2 1.8 / \plot 1.8 -.2 2 0 2.2 -.2 / \plot 0.9 1.1 0 2 / \plot 4 2 3.1 1.1 / \plot 2.9 0.9 2 0 1.1 0.9 / \put{\bf i } at 1 2 \put{\bf ii } at 2 1 \put{\bf iii } at 3 2 \put{$\bigcirc$ } at 2 0 \multiput{$\bullet$ } at 1 1 3 1 0.8 0.8 3.2 0.8 / \multiput{$\blacksquare$ } at 2 2 / \plot 0.8 0.8 2 2 / \plot 2 2 3.2 0.8 / \setdots
< .7 mm > \plot 1.8 -.2 2.2 -.2 / \put { } at 0 0 \put { } at 4 4 \put{$n_0 $ } [ r ] at 1.7 -.4 \put{$n_\infty$ } [ l ] at 2.3 -.4 \multiput{${\scriptstyle } \bullet$ } at 1.8 -.2 2.2 -.2 / \endpicture } at 7 -.25 \endpicture}\ ] ] as we have noted above , dealing with the corner of the quarter * ii * we have to distinguish whether @xmath516 is non - zero ( as in the left picture ) or zero ( the right picture shows the changes ) ; in the zero case we have marked the modules @xmath517 and @xmath518 , where @xmath517 is the boundary module in @xmath451 which has the same socle as @xmath657 and @xmath518 is the boundary module in @xmath457 which has the same top as @xmath237 .
we have exhibited in 12.5 a wind wheel with 5 bars , thus there are 5 tiles .
let us present at least one of the tiles , say @xmath658 in more detail : @xmath659 from 0.4 0.4 to 0.6 0.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.4 1.4 to 1.6 1.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.4 0.6 to 1.6 0.4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.4 -.4 to 0.6 -.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.4 -1.4 to 1.6 -1.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.4 -.6 to 1.6 -.4 \plot 2.3 2.3 2.5 2.5 / \plot 2.3 -2.3 2.5 -2.5 / \plot 2.3 1.7 2.5 1.5 / \plot 2.3 -1.7 2.5 -1.5 / \plot 2.3 0.3 2.5 0.5 / \plot 2.3 -.3 2.5 -.5 / \setdots < 1 mm > \plot 2.7 2.7 3 3 / \plot 2.7 -2.7 3 -3 / \endpicture } at -6 0 \put{\beginpicture \multiput { } at 0 2 -2 -2 / \put { \beginpicture \setcoordinatesystem units < .15cm,.15
cm > \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at 0 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at -2 2 \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at -1 1 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at -5 3 \put{${\scriptstyle } 0 $ } at -4 2 \put{${\scriptstyle } 2 $ } at -3 1 \put{${\scriptstyle } 3 $ } at -2 2 \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at -2 2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at -2 2 \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \put{${\scriptstyle } 4 $ } at 4 0 \put{${\scriptstyle } 2 $ } at 5 1 \put{${\scriptstyle } 3 $ } at 6 2 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at -2 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \put{${\scriptstyle } 4 $ } at 4 0 \put{${\scriptstyle } 2 $ } at 5 1 \put{${\scriptstyle } 3 $ } at 6 2 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at -1 -1 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \put{${\scriptstyle } 4 $ } at 4 0 \put{${\scriptstyle } 2 $ } at 5 1 \put{${\scriptstyle } 3 $ } at 6 2 \put{${\scriptstyle } 3 $ } at 7 1 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at -2 -2 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.6 0.6 to -0.4 0.4 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 1.6 to -1.4 1.4 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 0.4 to -1.4 0.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.6 -.6 to -0.4 -.4 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -1.6 to -1.4 -1.4 \arrow < 1.5
mm > [ 0.25,0.75 ] from -1.6 -.4 to -1.4 -.6 \plot -2.3 2.3 -2.5 2.5 / \plot -2.35 -2.35 -2.5 -2.5 / \plot -2.3 1.7 -2.5 1.5 / \plot -2.3 -1.7 -2.5 -1.5 / \plot -2.3 0.3 -2.5 0.5 / \plot -2.3 -.3 -2.5 -.5 / \setdots < 1 mm > \plot -2.7 2.7 -3 3 / \plot -2.7 -2.7 -3 -3 / \endpicture } at 6 0 \put{\beginpicture \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at 0 2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \put{${\scriptstyle } 4 $ } at 4 0 \put{${\scriptstyle } 2 $ } at 5 1 \put{${\scriptstyle } 3 $ } at 6 2 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at -1 1 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \put{${\scriptstyle } 4 $ } at 4 0 \put{${\scriptstyle } 2 $ } at 5 1 \put{${\scriptstyle } 3 $ } at 6 2 \put{${\scriptstyle } 3 $ } at 7 1 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at -2 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 2 $ } at -3 1 \put{${\scriptstyle } 3 $ } at -2 2 \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at 1 1 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 2 $ } at -3 1 \put{${\scriptstyle } 3 $ } at -2 2 \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \put{${\scriptstyle } 4 $ } at 4 0 \put{${\scriptstyle } 2 $ } at 5 1 \put{${\scriptstyle } 3 $ } at 6 2 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at 0 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at -4 2 \put{${\scriptstyle } 2 $ } at -3 1 \put{${\scriptstyle } 3 $ } at -2 2 \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at 2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 0.4 to -1.4 0.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from -.6 1.4 to -.4 1.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from -.6 0.6 to -.4 .4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.4 1.6 to 0.6 1.4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.4 0.6 to 1.6 0.4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.4 0.4 to 0.6 0.6 \plot -2.5 -0.5 -2.3 -0.3 / \plot -1.7 -0.3 -1.5 -.5 / \plot -0.5 -0.5 -.3 -.3 / \plot 2.5 -0.5 2.35 -0.35 / \plot 1.7 -0.3 1.5 -.5 / \plot 0.5 -0.5 .3 -.3 / \setdots < 1 mm > \plot -3 -1 -2.7 -.7 / \plot 3 -1 2.7 -.7 / \endpicture } at 0 6 \put{\beginpicture \multiput { } at -2 0 2 -2 / \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \put{${\scriptstyle } 4 $ } at 4 0 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at 0 -2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at -2 2 \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \put{${\scriptstyle } 4 $ } at 4 0 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at -1 -1 \put { \beginpicture \setcoordinatesystem units
< .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at -5 3 \put{${\scriptstyle } 0 $ } at -4 2 \put{${\scriptstyle } 2 $ } at -3 1 \put{${\scriptstyle } 3 $ } at -2 2 \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \put{${\scriptstyle } 4 $ } at 4 0 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at -2 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at -2 2 \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \put{${\scriptstyle } 4 $ } at 4 0 \put{${\scriptstyle } 2 $ } at 5 1 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at 0 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \put{${\scriptstyle } 4 $ } at 4 0 \put{${\scriptstyle } 2 $ } at 5 1 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at 1 -1 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put { } at -3 3 \put{${\scriptstyle } 3 $ } at -1 3 \put{${\scriptstyle } 2 $ } at 0 2 \put{${\scriptstyle } 4 $ } at 1 1 \put{${\scriptstyle } 5 $ } at 2 2 \put{${\scriptstyle } 5 $ } at 3 1 \put{${\scriptstyle } 4 $ } at 4 0 \put{${\scriptstyle } 2 $ } at 5 1 \put{${\scriptstyle } 3 $ } at 6 2 \put{${\scriptstyle } 3 $ } at 7 1 \put{${\scriptstyle } 2 $ } at 8 0 \plot 0 1 1 0 3 2 2 3 0 1 / \endpicture } at 2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.6 -0.4 to -1.4 -0.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from -.6 -1.4 to -.4 -1.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from -.6 -0.6 to -.4 -.4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.4 -1.6 to 0.6 -1.4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.4 -0.6 to 1.6 -0.4 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.4 -0.4 to 0.6 -0.6 \plot -2.5 0.5 -2.3 0.3 / \plot -1.7 0.3 -1.5 .5 / \plot -0.5 0.5 -.3 .3 / \plot 2.5 0.5 2.3 0.3 / \plot 1.7 0.3 1.5 .5 / \plot 0.5 0.5 .3 .3 / \setdots
< 1 mm > \plot -3 1 -2.7 .7 / \plot 3 1 2.7 .7 / \endpicture } at 0 -6 \put{\bf i$_{45}$ } at -2.5 0 \put{\bf ii$_{45}$ } at 0 -2.5 \put{\bf iii$_{45}$ } at 2.5 0 \put{\bf iv\!$_{45}$ } at 0 2.5 \setdots < 1 mm > \plot -3.5 3.5 -2 2 / \plot -1 1 1 -1 / \plot 2 -2 3.5 -3.5 / \plot -3.5 -3.5 -2 -2 / \plot -1 -1 1 1 / \plot 2 2 3.5 3.5 / \put{$\blacksquare$ } at 0 0 \multiput{$\bullet$ } at -3.5 -3.5 -2.5 -2.5 3.5 3.5 2.5 2.5 / \multiput{$\circ$ } at 3.5 -3.5 2.5 -2.5
-3.5 3.5 -2.5 2.5 / \put{adding hooks } [ r ] at -6 -2.6 \put{on the left } [ r ] at -6.3 -3 \put{deleting cohooks } [ r ] at -2 -6.6 \put{on the left } [ r ] at -2.3 -7 \put{adding hooks } [ l ] at 2 -6.6 \put{on the right } [ l ] at 2.1 -7 \put{deleting cohooks } [ l ] at 6 -2.6 \put{on the right } [ l ] at 6.3 -3 \setsolid \plot -8 -1.4 -7.2 -2.2 / \arrow < 1.5 mm > [ 0.25,0.75 ] from -6.2 -3.2 to -5.4 -4 \plot -4 -5.4 -3.2 -6.2 / \arrow < 1.5 mm > [ 0.25,0.75 ] from -2.2 -7.2 to -1.4 -8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.2 -2.2 to 8 -1.4 \plot 1.4 -8 2.2 -7.2 / \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 -6.2 to 4 -5.4 \endpicture}\ ] ] the vertices of such a tile @xmath590 , with @xmath236 a bar , are all the ( finite or infinite ) words which contain the completed bar @xmath514 as a subword .
thus , in our case we deal with the words which contain the word units < .15cm,.15 cm > at 0 2 at 1 1 at 2 2 at 3 1 0 1 1 0 3 2 2 3 0 1 / as a subword .
there is precisely one @xmath660-word of this form , namely @xmath485 , it lies in the center and is marked by the black square @xmath647 .
there are many @xmath661-words , they lie on the two diagonals through the center .
note that the @xmath660-word as well as all the @xmath661-words are not periodic , since any word in @xmath590 contains @xmath514 only once as a subword .
there are two kinds of @xmath661-words : on the northwest - southeast diagonal they are marked by a circle @xmath662 , these are the words of the form @xmath663 with @xmath157 a finite word .
the maximal periodic subword of @xmath395 starts with the bar @xmath236 ; one easily checks that @xmath395 is contracting in the sense of @xcite .
those on the northeast - southwest diagonal are marked by a bullet @xmath499 , these are the words of the form @xmath664 again with @xmath157 a finite word . here the maximal periodic subword starts with @xmath665 and @xmath666 turns out to be expanding .
as we know , the corner modules for the quarters are related to the exact sequence @xmath667 we obtain a border for @xmath590 : @xmath668 at 1.7 -.4 \put{$455 = n_\infty$ } [ l ] at 2.3 -.4 \multiput{${\scriptstyle } \bullet$ } at 1.8 -.2 2.2 -.2 / \put{${\operatorname{rad}}45= s(4)$ } [ r ] at -.6 2 \put{$s(5 ) = 45/{\operatorname{soc}}$ } [ l ] at 4.6 2 \put{$45 $ } at 2 4.7 \setsolid \plot 1.8 -.2 2 0 2.2 -.2 / \plot 1.8 4.2 2 4 2.2 4.2 / \plot -.2 1.8 0 2 -.2 2.2 / \plot 4.2 2.2 4 2 4.2 1.8 / \plot 0.9 1.1 0 2 0.9 2.9 / \plot 1.1 3.1 2 4 2.9 3.1 / \plot 3.1 2.9 4 2 3.1 1.1 / \plot 2.9 0.9 2 0 1.1 0.9 / \put{\bf i$_{45}$ } at 1 2 \put{\bf ii$_{45}$ } at 2 1 \put{\bf iii$_{45}$ } at 3 2 \put{\bf iv$_{45}$ } at 2 3 \multiput{$\bigcirc$ } at 2 0 0 2 2 4 4 2 / \multiput{$\bullet$ } at 1 1 3 3 0.8 0.8 3.2 3.2 1.2 1.2 2.8 2.8 / \multiput{$\circ$ } at 1 3 3 1 0.8 3.2 3.2 0.8 1.2 2.8 2.8 1.2 / \multiput{$\blacksquare$ } at 2 2 / \plot 0.8 0.8 3.2 3.2 / \plot 0.8 3.2 3.2 0.8 / \put { } at 0 0 \put { } at 4 4 \endpicture}\ ] ] here , @xmath517 is the boundary module in @xmath451 which has the same socle as @xmath610 and @xmath518 is the boundary module in @xmath457 which has the same top as @xmath610 .
we can further enlarge the picture by adding rays from @xmath505 and @xmath457 , as well as corays from @xmath451 and @xmath669 . in the case
@xmath534 , we obtain in this way the complete module category ( of course , opposite edges have to be identified ) : @xmath670 at 0 4 \put{$b '' = \lambda^{-1 } b$ } [ l ] at 4 4 \plot -.6 1.4 1.4 -.6 / \plot 2.6 4.6 4.6 2.6 / \plot -.6 2.6 1.4 4.6 / \plot 2.6 -0.6 4.6 1.4 / \setsolid \plot 1.8 -.2 2 0 2.2 -.2 / \plot 1.8 4.2 2 4 2.2 4.2 / \plot -.2 1.8 0 2 -.2 2.2 / \plot 4.2 2.2 4 2 4.2 1.8 / \plot 0.9 1.1 0 2 0.9 2.9 / \plot 1.1 3.1 2 4 2.9 3.1 / \plot 3.1 2.9 4 2 3.1 1.1 / \plot 2.9 0.9 2 0 1.1 0.9 / \put{\bf i } at 1 2 \put{\bf ii } at 2 1 \put{\bf iii } at 3 2 \put{\bf iv } at 2 3 \multiput{$\bigcirc$ } at 2 0 0 2 2 4 4 2 / \multiput{$\bullet$ } at 1 1 3 3 0.8 0.8 0.6 0.6 3.2 3.2 3.4 3.4 1.2 1.2 2.8 2.8 3.6 3.6 0.4 0.4 / \multiput{$\circ$ } at 1 3 3 1 0.8 3.2 3.2 0.8 0.6 3.4 3.4 0.6 1.2 2.8 2.8 1.2 3.6 0.4 0.4 3.6 / \multiput{$\blacksquare$ } at 2 2 / \plot 0.4 0.4 3.6 3.6 / \plot 0.4 3.6 3.6 0.4 / \put { } at 0 0 \put { } at 4 4 \endpicture}\ ] ] in particular , we get the following neighboring relation for tiles : @xmath671 according to @xcite , for any almost periodic @xmath661-word or biperiodic @xmath660-word @xmath395 there exists an indecomposable algebraically compact module @xmath672 . in case @xmath395 is a contracting @xmath661-word , the module @xmath672 is the usual string module for @xmath395 , whereas for @xmath395 an expanding @xmath661-word , one needs to complete the string module in order to obtain the module @xmath672 .
the @xmath660-words @xmath485 are mixed words ( the right hand side is contracting , the left hand side expanding ) , thus the module @xmath673 is obtained from the corresponding string module by a partial completion : the left hand side has to be completed , the right hand side not .
let us observe that for our enlargement by adding rays from @xmath674 and corays from @xmath675 , we also have to invoke the adic modules for the component @xmath451 and the prfer modules for the component @xmath457 .
in our example 12.5 with @xmath586 , five non - regular auslander - reiten components have been obtained . using infinite words ,
thus infinite dimensional representations , we see that these auslander - reiten components have to be arranged as follows : @xmath676 here , the left boundary has to be identified with the right boundary , and the lower dashed line with the ( slightly rotated ) upper dashed line .
the quilt which we obtain in this way is a torus with 5 holes .
let us summarize : the picture above presents the quilt of our wind wheel , it exhibits on a surface ( finite and infinite ) words which give rise to relevant indecomposable algebraically compact modules .
the black squares @xmath647 mark the non - periodic @xmath96-words , the circles @xmath662 and the bullets @xmath499 mark the @xmath677-words .
if we delete the infinite words , we obtain the 5 auslander - reiten components which contain string modules , all being shown here as squares with a hole in the middle . on the other hand , for any bar @xmath236 , we also spot easily the tile @xmath590 , it is a square with center @xmath485 .
let us stress that almost all , but * not all * indecomposable algebraically compact modules are used in the construction of the auslander - reiten quilt of a wind wheel .
here is the list of the additional modules : \(1 ) the generic module , \(2 ) the prfer modules for the tube @xmath678 and \(3 ) the adic modules for the tube @xmath452 there is the following general observation : let @xmath18 be a wind wheel with @xmath39 bars .
the auslander - reiten quilt @xmath37 of @xmath18 is a connected surface with boundary , its euler characteristic is @xmath679 this result can be interpreted as follows : let @xmath680 be the number of non - regular auslander - reiten components of @xmath18 , thus @xmath680 is the number of components of the boundary of @xmath37 and we have @xmath681 .
there is a ( connected ) compact riemann surface @xmath682 without boundary and with euler characteristic @xmath683 such that @xmath37 is obtained from @xmath682 by inserting @xmath680 holes . for @xmath684 ( as in the examples 12.8 ) ,
the surface @xmath682 has euler characteristic @xmath685 , thus it is a torus .
our cut - and - paste procedure presents @xmath37 as being obtained from @xmath686 pieces of the following form @xmath687 where the edge @xmath236 will be part of the boundary , whereas the remaining edges have to be identified in pairs .
looking at the vertices , we have to distinguish the endpoints @xmath52 and @xmath33 of the boundary edge @xmath236 and the remaining ones : the endpoints of the boundary edges are identified pairwise , whereas always four of the remaining ones yield a vertex of @xmath37 .
let @xmath688 be the number of vertices , edges and faces respectively .
there are @xmath689 faces , there are @xmath690 edges and @xmath691 vertices , thus @xmath692 the auslander - reiten quilt @xmath37 of any wind wheel is connected and orientable .
compact surfaces are often presented by a polygon with an even number of edges and an identification rule for pairs of the edges , this rule is shown by a word using the edges ( and their inverses ) as letters , this takes into account the orientation in which the edges are identified .
let @xmath693 be obtained from @xmath37 by filling the holes . clearly , we obtain @xmath694 by taking @xmath39 squares ( corresponding to the tiles ) and identifying pairs of edges .
the neighboring relation for the tiles shows in which way we have to identify the edges .
first , we use the permutation @xmath617 in order to obtain the following rectangle : @xmath695 at 13.6 7 \put{$x_2 $ } at 6.8 8 \put{$\rho(x_2)$ } [ l ] at 10.6 4 \put{$x_1 $ } at 9.8 11 \put{$\rho(x_1)$ } [ l ] at 4.6 -2 \setsolid \arrow < 2 mm > [ 0.25,0.75 ] from 1 1 to 2 2 \arrow < 2 mm > [ 0.25,0.75 ] from 7 7 to 8 8 \arrow < 2 mm > [ 0.25,0.75 ] from 10 10 to 11 11 \arrow < 2 mm > [ 0.25,0.75 ] from 13 11 to 14 10 \arrow < 2 mm > [ 0.25,0.75 ] from 1 -1 to 2 -2 \arrow < 2 mm > [ 0.25,0.75 ] from 4 -2 to 5 -1 \arrow < 2 mm > [ 0.25,0.75 ] from 10 4 to 11 5 \arrow < 2 mm > [ 0.25,0.75 ] from 13 7 to 14 8 \endpicture}\ ] ] this already shows the connectedness .
but we also see that the further identifications are achieved by the word @xmath696 it is well - known ( and easy to see ) that we can change the word to a product of commutators , but this means that @xmath697 is orientable . as we have seen , the auslander - reiten quilt of any wind wheel is orientable .
for example , in section 13.4 , we have exhibited a the quilt of a wind wheel @xmath35 which is a torus with holes .
but the category @xmath527 contains as a full subcategory the module category @xmath698 of the following algebra @xmath699 , and the auslander - reiten quiver of @xmath699 is obviously ( homeomorphic to ) a mbius strip : @xmath700 from 0 0.8 to 0 0.2 \circulararc 155 degrees from 0 2 center at 0 1.5 \circulararc -155 degrees from 0 2 center at 0 1.5 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.22 1.05 to -0.17 1.02 \setdots < .7 mm >
\setquadratic \plot -.4 1.4 0 1.3 .4 1.4 / \endpicture } at -1 0 \put{\beginpicture \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at 0 2 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 2 0 \endpicture } at -1 5 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 0 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 1 $ } at 2 0 \endpicture } at -1 3 \put{$0 $ } at 0 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 0 0 \endpicture } at 0 2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at 1 0 \put{${\scriptstyle } 1 $ } at 0 1 \endpicture } at 0 4 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 1 \put{${\scriptstyle } 1 $ } at 1 2 \put{${\scriptstyle } 1 $ } at 2 1 \put{${\scriptstyle } 0 $ } at 3 0 \endpicture } at 1 1 \put{$1 $ } at 1 3 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at 0 2 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 2 0 \endpicture } at 2 0 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 0 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 1 $ } at 2 0 \endpicture } at 2 2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at 0 1 \put{${\scriptstyle } 1 $ } at 1 0 \endpicture } at 3 1 \put { \beginpicture \setcoordinatesystem units
< .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 0 \put{${\scriptstyle } 1 $ } at 1 1 \endpicture } at 3 3 \put{$0 $ } at 3 5 \put{$1 $ } at 4 2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 1 \put{${\scriptstyle } 1 $ } at 1 2 \put{${\scriptstyle } 1 $ } at 2 1 \put{${\scriptstyle } 0 $ } at 3 0 \endpicture } at 4 4 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at 0 2 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 2 0 \endpicture } at 5 5 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 0 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 1 $ } at 2 0 \endpicture } at 5 3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -.8 4.8 to -.2 4.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from -.8 3.2 to -.2 3.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from -.8 2.8 to -.2 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from .2 3.8 to .8 3.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from .2 2.2 to .8 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from .2 1.8 to .8 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from .2 .2 to .8 .8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 2.8 to 1.8 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 1.2 to 1.8 1.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 .8 to 1.8 .2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 2.2 to 2.8 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 1.8 to 2.8 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 .2 to 2.8 .8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 4.8 to 3.8 4.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 3.2 to 3.8 3.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 2.8 to 3.8 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 1.2 to 3.8 1.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.2 4.2 to 4.8 4.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.2 3.8 to 4.8 3.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.2 2.2 to 4.8 2.8 \setdashes <
2 mm > \plot 0 5 0 4.3 / \plot 0 2.3 0 3.7 / \plot 0 1 0 1.7 / \plot 3 4 3 3.3 / \plot 3 1.3 3 2.7 / \plot 3 0 3 .7 / \setshadegrid span < .7 mm > \vshade -1 2 5 < z , z , , > 0 2 4 < z , z , , > .99 1 3 < z , z , , > 1 0 3 < z , z , , > 2 0 3 < z , z , , > 3 1 3 < z , z , , > 3.99 2 4 < z , z , , > 4 2 5 < z , z , , > 5 2 5 / \vshade 0 0 0 < z , z , , > 1 0 1 / \vshade 3 5 5
< z , z , , > 4 4 5 / \endpicture } at 5 0 \endpicture}\ ] ] here , the vertical dashes lines mark a fundamental domain . instead of looking at the wind wheel considered in section 12.5 ,
let @xmath35 now be the smallest possible wind wheel , with two vertices @xmath701 , a loop @xmath7 at the vertex @xmath685 , a loop @xmath5 at the vertex @xmath34 and an arrow @xmath702 ( and the relations @xmath703 ) .
we want to analyze the embedding of @xmath698 into @xmath704 some parts of the auslander - reiten quiver of @xmath699 can be identified in the quilt of @xmath35 without problems : @xmath705 from .2 3.8 to .8 3.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from .2 2.2 to .8 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from .2 1.8 to .8 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from .2 .2 to .8 .8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 2.8 to 1.8 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 1.2 to 1.8 1.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 .8 to 1.8 .2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 2.2 to 2.8 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 1.8 to 2.8 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 .2 to 2.8 .8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 4.8 to 3.8 4.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 3.2 to 3.8 3.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 2.8 to 3.8 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 1.2 to 3.8 1.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.2 4.2 to 4.8 4.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.2 3.8 to 4.8 3.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.2 2.2 to 4.8 2.8 \setdashes <
2 mm > \plot 0 5 0 4.3 / \plot 0 2.3 0 3.7 / \plot 0 1 0 1.7 / \plot 3 4 3 3.3 / \plot 3 1.3 3 2.7 / \plot 3 0 3 .7 / \setdots < 1 mm > \plot 1.4 3 2.6 3 / \plot 4.4 2 5 2 / \plot 3.4 5 4.6 5 / \setshadegrid span < .7 mm > \vshade 0 0 0 < z , z , , > 2 0 2 < z , z , , > 3 1 3 < z , z , , > 4 2 2 / \setsolid \circulararc 360 degrees from 2.4 0 center at 2 0 \circulararc 360 degrees from 0.3 0 center at 0 0 \circulararc 360 degrees from 4.3 2 center at 4 2 \circulararc 360 degrees from 3.4 3 center at 3 3 \endpicture } at 0 0 \put{\beginpicture \setcoordinatesystem units <
1cm,1 cm > \put{$0 $ } at -1 2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 1 \put{${\scriptstyle } 1 $ } at 1 2 \put{${\scriptstyle } 1 $ } at 2 1 \put{${\scriptstyle } 0 $ } at 3 0 \endpicture } at -.5 2.5 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 1 \put{${\scriptstyle } 0 $ } at 1 0 \endpicture } at -.5 1.5 \put{$1 $ } at 5 2 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at 0 2 \put{${\scriptstyle } 0 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 2 0 \endpicture } at 1.5 -0.5 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at 0 2 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 2 0 \endpicture } at 2.5 -0.5 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 0 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 1 $ } at 2 0 \endpicture } at 1.5 4.5 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 1 \put{${\scriptstyle } 0 $ } at 1 0 \put{${\scriptstyle } 1 $ } at 2 1 \endpicture } at 2.5 4.5 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at 0 1 \put{${\scriptstyle } 1 $ } at 1 0 \endpicture } at 4.5 1.5 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 0 \put{${\scriptstyle } 1 $ } at 1 1 \endpicture } at 2 5 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at 0 2 \put{${\scriptstyle } 0 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 2 0 \put{${\scriptstyle } 1 $ } at 3 1 \endpicture } at 4.5 2.5 \setsolid \plot 1.8 -.2 2 0 2.2 -.2 / \plot 1.8 4.2 2 4 2.2 4.2 / \plot -.2 1.8 0 2 -.2 2.2 / \plot 4.2 2.2 4 2 4.2 1.8 / \plot 0.9 1.1 0 2 0.9 2.9 / \plot 1.1 3.1 2 4 2.9 3.1 / \plot 3.1 2.9 4 2 3.1 1.1 / \plot 2.9 0.9 2 0 1.1 0.9 / \multiput{$\bullet$ } at 1 1 1 3 3 1 3 3 0.5 0.5 0.5 3.5 3.5 0.5 3.5 3.5 / \multiput{$\blacksquare$ } at 2 2 / \plot 0.5 0.5 3.5 3.5 / \plot 0.5 3.5 3.5 0.5 / \plot -.2 1.2 0.4 0.6 / \plot 0.6 0.4 1.2 -.2 / \plot -.8 1.8 -.7 1.7 / \plot -.2 2.8 0.4 3.4 / \plot 0.6 3.6 1.2 4.2 / \plot -.8 2.2 -.7 2.3 / \plot 4.2 1.2 3.6 0.6 / \plot 3.4 0.4 2.8 -.2 / \plot 4.2 2.8 3.6 3.4 / \plot 3.4 3.6 2.8 4.2 / \plot 4.8 1.8 4.7 1.7 / \plot 4.85 2.15 4.75 2.25 / \plot 1.7 4.7 1.85 4.85 / \plot 2.3 4.7 2.15 4.85 / \setdots < 1 mm > \plot 1.7 -.5 2.3 -.5 / \setshadegrid span < .7 mm > \hshade -0.5 1.5 2.5 < , , z , z > 2 -1 5 < , , z , z > 5 2 2 / \setsolid \circulararc 360 degrees from 2.9 -.5 center at 2.5 -.5 \circulararc 360 degrees from -1.3 2 center at -1 2 \circulararc 360 degrees from 2.4 5 center at 2 5 \circulararc 360 degrees from 5.3 2 center at 5 2 \endpicture } at 7 0 \endpicture}\ ] ] all the maps in the shaded part of @xmath698 ( see the left picture ) are nicely factorized in @xmath527 , see the right picture , note that on the right we see the bordered tile for @xmath35 . of course , this concerns also the following shaded part on the left ( actually , we deal with the same part of @xmath698 ) : @xmath705 from .2 3.8 to .8 3.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from .2 2.2 to .8 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from .2 1.8 to .8 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from .2 .2 to .8 .8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 2.8 to 1.8 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 1.2 to 1.8 1.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 .8 to 1.8 .2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 2.2 to 2.8 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 1.8 to 2.8 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 .2 to 2.8 .8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 4.8 to 3.8 4.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 3.2 to 3.8 3.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 2.8 to 3.8 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 1.2 to 3.8 1.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.2 4.2 to 4.8 4.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.2 3.8 to 4.8 3.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.2 2.2 to 4.8 2.8 \setdashes <
2 mm > \plot 0 5 0 4.3 / \plot 0 2.3 0 3.7 / \plot 0 1 0 1.7 / \plot 3 4 3 3.3 / \plot 3 1.3 3 2.7 / \plot 3 0 3 .7 / \setdots < 1 mm > \plot 1.4 3 2.6 3 / \plot 4.4 2 5 2 / \plot 3.4 5 4.6 5 / \plot 0.4 0 1.6 0 / \setshadegrid span < .7 mm > \hshade 2 0 0 < , , z , z > 3 -1 1 < , , z , z > 5 -3 -1 / \setsolid \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 1 $ } at 0 2 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 0 $ } at 2 0 \endpicture } at -1 5 \put { \beginpicture \setcoordinatesystem units < .15cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 0 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 1 $ } at 2 0 \endpicture } at -1 3 \arrow < 1.5 mm > [ 0.25,0.75 ] from -.8 4.8 to -.2 4.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from -.8 3.2 to -.2 3.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from -.8 2.8 to -.2 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.8 4.2 to -1.2 4.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from -1.8 3.8 to -1.2 3.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from -2.8 4.8 to -2.2 4.2 \endpicture } at 0 0 \endpicture}\ ] ] it is the following square in @xmath698 which is difficult to recover in the auslander - reiten quilt of @xmath35 : @xmath705 from .2 3.8 to .8 3.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from .2 2.2 to .8 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from .2 1.8 to .8 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from .2 .2 to .8 .8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 2.8 to 1.8 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 1.2 to 1.8 1.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 .8 to 1.8 .2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 2.2 to 2.8 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 1.8 to 2.8 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 .2 to 2.8 .8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 4.8 to 3.8 4.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 3.2 to 3.8 3.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 2.8 to 3.8 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 1.2 to 3.8 1.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.2 4.2 to 4.8 4.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.2 3.8 to 4.8 3.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.2 2.2 to 4.8 2.8 \setdashes <
2 mm > \plot 0 5 0 4.3 / \plot 0 2.3 0 3.7 / \plot 0 1 0 1.7 / \plot 3 4 3 3.3 / \plot 3 1.3 3 2.7 / \plot 3 0 3 .7 / \setdots < 1 mm > \plot 1.4 3 2.6 3 / \plot 4.4 2 5 2 / \plot 3.4 5 4.6 5 / \plot 0.4 0 1.6 0 / \setshadegrid span < .7 mm > \vshade 0 2 2 < z , z , , > 1 1 3 < z , z , , > 2 2 2 / \setsolid \put{$f$ } at 0.35 1.25 \put{$g$ } at 1.6 2.75 \endpicture } at 0 0 \endpicture}\ ] ] to be more precise : we should distinguish between the maps pointing upwards and those pointing downwards , the maps pointing upwards have already been discussed since they are part of the bordered tile , thus let us concentrate on the maps pointing downwards , they are labeled @xmath706 and @xmath707 in the picture .
we want to see in which way these maps can be factorized in the category @xmath527 . first , consider the map @xmath708 we can factor is as follows ( always additions or deletions on the right ) : @xmath709 from 0.5 0 to 1.5 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.5 0 to 3.5 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.5 0 to 5.5 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.5 0 to 7.5 0 \put{$f_1 $ } at 1 0.5 \put{$f_2 $ } at 3 0.5 \put{$f_3 $ } at 5 0.5 \put{$f_4 $ } at 7 0.5 \put{$\cdots$ } at 8 0 \put{$\cdots$ } at 10 0 \put{\beginpicture \setcoordinatesystem units < .1cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 0 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 1 $ } at 2 2 \put{${\scriptstyle } 0 $ } at 3 1 \endpicture } at 15.8 0 \put{\beginpicture \setcoordinatesystem units < .1cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 0 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 1 $ } at 2 2 \put{${\scriptstyle } 0 $ } at 3 1 \put{${\scriptstyle } 0 $ } at 4 0 \put{${\scriptstyle } 1 $ } at 5 1 \put{${\scriptstyle } 1 $ } at 6 2 \endpicture } at 14 0 \put{\beginpicture \setcoordinatesystem units < .1cm,.15 cm > \put{${\scriptstyle } 0 $ } at 0 0 \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 1 $ } at 2 2 \put{${\scriptstyle } 0 $ } at 3 1 \put{${\scriptstyle } 0 $ } at 4 0 \put{${\scriptstyle } 1 $ } at 5 1 \put{${\scriptstyle } 1 $ } at 6 2 \put{${\scriptstyle } 0 $ } at 7 1 \endpicture } at 12 0 \arrow
< 1.5 mm > [ 0.25,0.75 ] from 10.5 0 to 11.5 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.5 0 to 13.5 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 14.5 0 to 15.5 0 \put{${}_3f$ } at 11 0.5 \put{${}_2f$ } at 13 0.5 \put{${}_1f$ } at 15 0.5 \endpicture}\ ] ] similarly , we look at the map @xmath710 and factor it ( again always additions or deletions on the right ) : @xmath711 from 0.5 0 to 1.5 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.5 0 to 3.5 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.5 0 to 5.5 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.5 0 to 7.5 0 \put{$g_1 $ } at 1 0.5 \put{$g_2 $ } at 3 0.5 \put{$g_3 $ } at 5 0.5 \put{$g_4 $ } at 7 0.5 \put{$\cdots$ } at 8 0 \put{$\cdots$ } at 10 0 \put{\beginpicture \setcoordinatesystem units < .1cm,.15 cm > \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 1 $ } at 2 2 \put{${\scriptstyle } 0 $ } at 3 1 \endpicture } at 15.8 0 \put{\beginpicture \setcoordinatesystem units < .1cm,.15 cm > \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 1 $ } at 2 2 \put{${\scriptstyle } 0 $ } at 3 1 \put{${\scriptstyle } 0 $ } at 4 0 \put{${\scriptstyle } 1 $ } at 5 1 \put{${\scriptstyle } 1 $ } at 6 2 \endpicture } at 14 0 \put{\beginpicture \setcoordinatesystem units < .1cm,.15 cm > \put{${\scriptstyle } 1 $ } at 1 1 \put{${\scriptstyle } 1 $ } at 2 2 \put{${\scriptstyle } 0 $ } at 3 1 \put{${\scriptstyle } 0 $ } at 4 0 \put{${\scriptstyle } 1 $ } at 5 1 \put{${\scriptstyle } 1 $ } at 6 2 \put{${\scriptstyle } 0 $ } at 7 1 \endpicture } at 12 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 10.5 0 to 11.5 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.5 0 to 13.5 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 14.5 0 to 15.5 0 \put{${}_3g$ } at 11 0.5 \put{${}_2g$ } at 13 0.5 \put{${}_1g$ } at 15 0.5 \endpicture}\ ] ] now if @xmath712 is odd , then the maps @xmath713 and @xmath714 both are obtained by the addition of a hook , thus they are irreducible , and both @xmath715 and @xmath716 are obtained by the deletion of a cohook , thus they also are irreducible . but for @xmath712 even , all the maps @xmath717 belong to the infinite radical @xmath718 ( and actually not to @xmath719 ) . by definition
( see for example @xcite ) , the infinite radical @xmath718 is the intersection of the powers @xmath720 with @xmath721 it follows that the maps @xmath722 belong to all powers of the infinite radical , thus to @xmath723 . the sequences of maps displayed here show that @xmath706 factors through the adic module given by the ( expanding ) word @xmath724 , whereas @xmath707 factors through the prfer module given by the word ( contracting ) word @xmath725 .
note that both these modules are indecomposable algebraically compact modules which are not used in the quilt .
let us try to following the factorization in the quilt .
first , we consider again @xmath706 .
looking at the maps @xmath713 we have noted already that those with odd index are irreducible , they belong to @xmath504 , whereas those with even index factor through an upwards path in the tile @xmath652 .
similarly , the maps @xmath715 with odd index are irreducible maps inside @xmath726 , whereas those with even index factor throgh a downwards path in the tile @xmath652 : @xmath727 from 0.1 1.7 to 2.4 4 \arrow < 2.5 mm > [ 0.25,0.75 ] from 0.3 1.5 to 2.6 3.8 \arrow < 2.5 mm > [ 0.25,0.75 ] from 0.6 1.2 to 2.9 3.5 \arrow < 2.5 mm > [ 0.25,0.75 ] from 0.7 1.1 to 3.0 3.4 \put{$f_2 $ } at 1.0 2.43 \put{$f_4 $ } at 1.25 2.2 \setdots < .5 mm > \plot 2 3 2.15 2.85 / \plot 2.35 2.65 2.5 2.5 / \plot 1.0 2.0 1.15 1.85 / \plot 1.35 1.65 1.5 1.5 / \endpicture } at 0 0 \put{\beginpicture \multiput { } at 0 4.4 0 -.4 / \multiput{$\mathcal r'_\infty$ } at 2.8 0.35 .35 2.7 / \plot -.4 1.6 1.6 -.4 / \plot 2.4 4.4 4.4 2.4 / \plot -.4 2.4 1.6 4.4 / \plot 2.4 -0.4 4.4 1.6 / \setsolid \plot 1.8 -.2 2 0 2.2 -.2 / \plot 1.8 4.2 2 4 2.2 4.2 / \plot -.2 1.8 0 2 -.2 2.2 / \plot 4.2 2.2 4 2 4.2 1.8 / \plot 0.9 1.1 0 2 0.9 2.9 / \plot 1.1 3.1 2 4 2.9 3.1 / \plot 3.1 2.9 4 2 3.1 1.1 / \plot 2.9 0.9 2 0 1.1 0.9 / \multiput{$\bigcirc$ } at 2 0 0 2 2 4 4 2 / \multiput{$\bullet$ } at 1 1 3 3 0.8 0.8 0.6 0.6 3.2 3.2 3.4 3.4 1.2 1.2 2.8 2.8 / \multiput{$\circ$ } at 1 3 3 1 0.8 3.2 3.2 0.8 0.6 3.4 3.4 0.6 1.2 2.8 2.8 1.2 / \multiput{$\blacksquare$ } at 2 2 / \plot 0.6 0.6 3.4 3.4 / \plot 0.6 3.4 3.4 0.6 / \put { } at 0 0 \put { } at 4 4 \arrow < 2.5 mm > [ 0.25,0.75 ] from 0.1 2.3 to 2.4 0 \arrow < 2.5 mm > [ 0.25,0.75 ] from 0.3 2.5 to 2.6 0.2 \arrow < 2.5 mm > [ 0.25,0.75 ] from 0.6 2.8 to 2.9 .5 \arrow < 2.5 mm > [ 0.25,0.75 ] from 0.7 2.9 to 3.0 .6 \put{${}_2f$ } at .6 1.8 \put{${}_4f$ } at 1 1.8 \setdots < .5 mm >
\plot 2 1 2.15 1.15 / \plot 2.35 1.35 2.5 1.5 / \plot 1.0 2.0 1.15 2.15 / \plot 1.35 2.35 1.5 2.5 / \endpicture } at 5.2 0 \endpicture}\ ] ] what is of importance is the change of the direction which we encounter : as long as we deal with the maps @xmath713 we work with maps pointing upwards , but after we have passed the adic module ( which is hidden ) we deal with the maps @xmath715 and they point downwards .
there is the similar feature for the map @xmath707 : @xmath728 from 1.7 .1 to 4 2.4 \arrow < 2.5 mm > [ 0.25,0.75 ] from 1.5 .3 to 3.8 2.6 \arrow < 2.5 mm > [ 0.25,0.75 ] from 1.2 .6 to 3.5 2.9 \arrow < 2.5 mm > [ 0.25,0.75 ] from 1.1 .7 to 3.4 3.0 \put{$g_2 $ } at 2.43 1.0 \put{$g_4 $ } at 2.2 1.2 \setdots < .5 mm > \plot 3 2 2.85 2.15 / \plot 2.65 2.35 2.5 2.5 / \plot 2.0 1.0 1.85 1.15 / \plot 1.65 1.35 1.5 1.5 / \endpicture } at 0 0 \put{\beginpicture \multiput { } at 0 4.4 0 -.4 / \multiput{$\mathcal q'$ } at 3.65 1.25 1.3 3.6 / \plot -.4 1.6 1.6 -.4 / \plot 2.4 4.4 4.4 2.4 / \plot -.4 2.4 1.6 4.4 / \plot 2.4 -0.4 4.4 1.6 / \setsolid \plot 1.8 -.2 2 0 2.2 -.2 / \plot 1.8 4.2 2 4 2.2 4.2 / \plot -.2 1.8 0 2 -.2 2.2 / \plot 4.2 2.2 4 2 4.2 1.8 / \plot 0.9 1.1 0 2 0.9 2.9 / \plot 1.1 3.1 2 4 2.9 3.1 / \plot 3.1 2.9 4 2 3.1 1.1 / \plot 2.9 0.9 2 0 1.1 0.9 / \multiput{$\bigcirc$ } at 2 0 0 2 2 4 4 2 / \multiput{$\bullet$ } at 1 1 3 3 0.8 0.8 0.6 0.6 3.2 3.2 3.4 3.4 1.2 1.2 2.8 2.8 / \multiput{$\circ$ } at 1 3 3 1 0.8 3.2 3.2 0.8 0.6 3.4 3.4 0.6 1.2 2.8 2.8 1.2 / \multiput{$\blacksquare$ } at 2 2 / \plot 0.6 0.6 3.4 3.4 / \plot 0.6 3.4 3.4 0.6 / \put { } at 0 0 \put { } at 4 4 \arrow < 2.5 mm > [ 0.25,0.75 ] from 1.7 3.9 to 4
. 1.6 \arrow < 2.5 mm > [ 0.25,0.75 ] from 1.5 3.7 to 3.8 1.4 \arrow < 2.5 mm > [ 0.25,0.75 ] from 1.2 3.4 to 3.5 1.1 \arrow < 2.5 mm > [ 0.25,0.75 ] from 1.1 3.3 to 3.4 1 . \put{${}_2g$ } at 2.15 3.3 \put{${}_4g$ } at 1.8 3.2 \setdots < .5 mm > \plot 2.85 1.85 3 2 / \plot 2.55 1.55 2.7 1.7 / \plot 1.9 2.9 2.05 3.05 / \plot 1.55 2.55 1.7 2.7 / \endpicture } at 5.2 0 \endpicture}\ ] ] here we deal first with the maps @xmath714 pointing wards , and after we have passed the hidden prfer module we deal with the maps @xmath716 and they point again downwards .
altogether we may say that the maps @xmath706 and @xmath707 are embedded into the quilt of @xmath35 with a kind of crossing , so that the shaded parts are connected by a square which is folded over : @xmath729 from .2 .2 to .8 .8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 1.2 to 1.8 1.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 2.2 to 2.8 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 .8 to 1.8 .2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 1.8 to 2.8 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 .2 to 2.8 .8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 2.8 to 3.8 2.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.2 1.2 to 3.8 1.8 \setdashes < 2 mm > \plot 3 3.5 3 3.3 / \plot 3 1.3 3 2.7 / \plot 3 0 3 .7 / \setshadegrid span < .7 mm > \vshade 0 0 0 < z , z , , > 2 0 2 < z , z , , > 3 1 3 < z , z , , > 4 2 2 / \put { } at 0 3.5 \endpicture } at 0 0 5 0 / \setquadratic \plot 2.2 .6 3.3 1 4.6 0.6 / \plot 1.4 1.4 3.1 1 4 .
0 / \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.55 0.63 to 4.61 0.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4 0 to 4.05 -.2 \put{$f$ } at 2 1.6 \put{$g$ } at 2.5 .55 \endpicture}\ ] ]
barbell algebras are of non - polynomial growth . given an arrow @xmath7 , we denote by @xmath730 the set of cyclic words starting with @xmath7 and ending in an inverse letter ( for all the words in @xmath730 , the last letter is a fixed one , namely the inverse of the only arrow different from @xmath7 which has the same end point as @xmath7 ) .
clearly , @xmath730 is a semigroup .
note that the given algebra is non - domestic ( and then even of non - polynomial growth ) if and only if there exists an arrow @xmath7 such that @xmath730 is non - empty and not cyclic ( [ r1 ] , proposition 2 and its proof ) . here
we take @xmath731 we assume that the length of @xmath732 is @xmath733 , respectively .
let @xmath734 , @xmath735 and @xmath736
. then both @xmath737 and @xmath738 are elements of @xmath730 .
this shows that @xmath182 is of non - polynomial growth .
we consider the algebra given in example 2 .
the non - regular component looks as follows : @xmath739 from 0.3 0.3 to 0.7 0.7 \arrow < 2 mm > [ 0.25,0.75 ] from 2.3 0.3 to 2.7 0.7 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 0.3 to 6.7 0.7 \arrow < 2 mm > [ 0.25,0.75 ] from 0.3 2.3 to 0.7 2.7 \arrow < 2 mm > [ 0.25,0.75 ] from 2.3 2.3 to 2.7 2.7 \arrow < 2 mm > [ 0.25,0.75 ] from 0.3 4.3 to 0.7 4.7 \arrow < 2 mm > [ 0.25,0.75 ] from 2.3 4.3 to 2.7 4.7 \arrow < 2 mm > [ 0.25,0.75 ] from 4.3 4.3 to 4.7 4.7 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 4.3 to 6.7 4.7 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 2.3 to 6.7 2.7 \arrow < 2 mm > [ 0.25,0.75 ] from 3.3 3.3 to 3.7 3.7 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 1.3 to 1.7 1.7 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 3.3 to 1.7 3.7 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 1.3 to 7.7 1.7 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 3.3 to 7.7 3.7 \arrow < 2 mm > [ 0.25,0.75 ] from 1.4 -0.6 to 1.7 -0.3 \arrow < 2 mm > [ 0.25,0.75 ] from 3.4 -0.6 to 3.7 -0.3 \arrow < 2 mm > [ 0.25,0.75 ] from 5.4 -0.6 to 5.7 -0.3 \arrow < 2 mm > [ 0.25,0.75 ] from 7.4 -0.6 to 7.7 -0.3 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 0.7 to 1.7 0.3 \arrow < 2 mm > [ 0.25,0.75 ] from 3.3 0.7 to 3.7 0.3 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 0.7 to 7.7 0.3 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 2.7 to 1.7 2.3 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 2.7 to 7.7 2.3 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 4.7 to 7.7 4.3 \arrow < 2 mm > [ 0.25,0.75 ] from 5.3 4.7 to 5.7 4.3 \arrow < 2 mm > [ 0.25,0.75 ] from 3.3 4.7 to 3.7 4.3 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 4.7 to 1.7 4.3 \arrow < 2 mm > [ 0.25,0.75 ] from 8.4 5.6 to 8.7 5.3 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 3.7 to 6.7 3.3 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 1.7 to 6.7 1.3 \arrow < 2 mm > [ 0.25,0.75 ] from 0.3 1.7 to 0.7 1.3 \arrow < 2 mm > [ 0.25,0.75 ] from 2.3 1.7 to 2.7 1.3 \arrow < 2 mm > [ 0.25,0.75 ] from 0.3 3.7 to 0.7 3.3 \arrow < 2 mm > [ 0.25,0.75 ] from 2.3 3.7 to 2.7 3.3 \arrow < 2 mm > [ 0.25,0.75 ] from 0.4 5.6 to 0.7 5.3 \arrow < 2 mm > [ 0.25,0.75 ] from 2.4 5.6 to 2.7 5.3 \arrow < 2 mm > [ 0.25,0.75 ] from 4.4 5.6 to 4.7 5.3 \arrow < 2 mm > [ 0.25,0.75 ] from 6.4 5.6 to 6.7 5.3 \plot 1.3 5.3 1.6 5.6 / \plot 3.3 5.3 3.6 5.6 / \plot 5.3 5.3 5.6 5.6 / \plot 7.3 5.3 7.6 5.6 / \plot 8.3 4.3 8.6 4.6 / \plot 8.3 2.3 8.6 2.6 / \plot 8.3 0.3 8.6 0.6 / \plot 8.3 3.7 8.6 3.4 / \plot 8.3 1.7 8.6 1.4 / \plot 8.3 -0.3 8.6 -0.6 / \plot 6.3 -0.3 6.6 -0.6 / \plot
4.3 -0.3 4.6 -0.6 / \plot 2.3 -0.3 2.6 -0.6 / \plot 0.3 -0.3 0.6 -0.6 / \setdots<1 mm > \plot 4.5 0 5.7 0 / \plot 4.45 4 5.8 4 / \setshadegrid span < 0.7 mm > \vshade 0 -0.8 5.6 < , z , , > 2 -0.8 5.6 / \vshade 2 2 5.6 < z , z , , > 4 4 5.6 < z , z , , > 6 4 5.6 < z , z , , > 7 3 5.6 / \vshade 2 -0.8 2 < z , z ,
, > 4 -0.8 0 < z , z , , > 6 -0.8 0 < z , z , , > 7 -0.8 1 / \vshade 7 -0.8 5.6 < z , z , , > 8.6 -0.8 5.6 / \vshade 6 2 2 < z , z , , > 7 1 3 / \setdashes < 2 mm > \plot 3.3 2.7 5.7 0.3 / \plot 4.3 3.7 5.7 2.3 / \plot 3.3 1.3 5.7 3.7 / \plot 4.3 0.3 5.7 1.7 / \endpicture}\ ] ] the component contains 10 of the 12 string modules which are boundary modules ; the remaining two string modules which are boundary modules are the serial string modules of length 3 ( with composition factors going up @xmath740 and @xmath741 , respectively ) .
they form the boundary of a stable tube of rank 2 ; the boundary meshes are those provided by the arrows @xmath742 and @xmath743 respectively : @xmath744 from 0.3 0.3 to 0.7 0.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 0.3 to 2.7 0.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 0.7 to 1.7 0.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.3 0.7 to 3.7 0.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 2.7 to 1.7 2.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.3 2.7 to 3.7 2.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 1.3 to 1.7 1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.3 1.3 to 3.7 1.7 \arrow < 1.5
mm > [ 0.25,0.75 ] from 0.3 2.3 to 0.7 2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 2.3 to 2.7 2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 1.7 to 0.7 1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 1.7 to 2.7 1.3 \setdashes
< 1 mm > \plot 0 0.3 0 1.7 / \plot 4 0.3 4 1.7 / \plot 0 2.3 0 3 / \plot 4 2.3 4 3 / \setshadegrid span < 0.7 mm > \vshade 0 0 3 4 0 3 / \endpicture}\ ] ] let us have another look at the non - regular component .
the picture shows nicely a phenomenon which has attracted a lot of attention lately , in some other context , namely when dealing with cluster tilted algebras .
let us recall the relevant facts : given a cluster tilted algebra @xmath18 , the category @xmath745 is obtained from the corresponding cluster category by factoring out a cluster tilting object @xcite . looking at a vertex @xmath12 of the quiver of @xmath18 , the corresponding indecomposable projective @xmath18-module @xmath746 and the corresponding indecomposable injective @xmath18-module @xmath747 satisfy @xmath748 where @xmath749 is the auslander - reiten translation in the cluster category ( if we denote by @xmath750 the auslander - reiten translation in the category @xmath745 , then @xmath751 for any indecomposable non - projective @xmath18-module , whereas , of course , @xmath752 for @xmath21 indecomposable projective ) . as we see in the picture , the non - regular component is a translation quiver which can be considered as part of a regular translation quiver @xmath753 obtained by adding a new vertex @xmath754 for every projective vertex @xmath755 , such that the translate of @xmath755 is @xmath754 and the translate of @xmath754 is an injective vertex .
let us define a function @xmath706 on the set of vertices of @xmath753 as follows : if @xmath395 is an old vertex , let @xmath756 be the length of the corresponding module , if @xmath757 is a new vertex , let @xmath758 then @xmath706 satisfies the following property : @xmath759 for all vertices @xmath355 of @xmath753 ( one may say that such a function with values in @xmath96 is cluster - additive
, see @xcite ) .
other similarities with cluster tilted algebras ( see @xcite ) should be mentioned : the barbell algebras are gorenstein algebras of gorenstein dimension @xmath34 und the stable category of cohen - macaulay modules is calabi - yau of cy - dimension @xmath36 . for our example 2 , here are the minimal injective resolutions of the indecomposable projective modules : @xmath760 let @xmath761 be the full subcategory of all torsionless modules ( by definition , a module is torsionless if it can be embedded into a projective module ) and @xmath762 the full subcategory of all projective modules .
we have to calculate the factor category @xmath763 since we deal with a @xmath34-gorenstein algebra , @xmath764 is a triangulated category with auslander - reiten translation .
it is not difficult to check that the only indecomposable modules which are torsionless and not projective are the two serial modules of length 2 with socle @xmath765 , we denote them by @xmath766 and @xmath767 thus , @xmath761 has the following auslander - reiten quiver : @xmath768 from 0.2 0.2 to 0.8 0.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 0.8 to 1.8 0.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 1.7 to 0.8 1.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 2.3 to 0.8 2.8 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.2 3.8 to 0.8 3.2 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.2 3.2 to 1.8 3.8 \setdots < 1 mm > \plot 0.3 0 1.7 0 / \plot 0.3 4 1.7 4 / \setshadegrid span < 0.5 mm > \hshade 0 0 2 1 1 1 / \hshade 3 1 1 4 0 2 / \setdashes
< 1 mm > \plot 0 0.4 0 0.7 / \plot 0 -.4 0 -.7 / \plot 2 0.4 2 0.7 / \plot 2 -.4 2 -.7 / \plot 0 4.4 0 4.7 / \plot 0 3.6 0 3.3 / \plot 2 4.4 2 4.7 / \plot 2 3.6 2 3.3 / \endpicture}\ ] ] the dashed line indicate that we have to identify vertices of the triangles exhibited : note that both serial modules @xmath769 and @xmath770 are shown twice , these are the vertices which have to be identified .
it follows that the ( triangulated ) category @xmath764 has just two indecomposable objects , both being fixed under the suspension functor as well as under the auslander - reiten translation functor ( so that @xmath764 is the product of two copies of the stable module category of the algebra @xmath771 = k[t]/\langle t^2\rangle$ ] of dual numbers ) , the auslander - reiten quiver of @xmath764 looks as follows : @xmath772 thus , we deal with a triangulated category for which both the suspension functor as well as the auslander - reiten translation functor are the identity functor .
this means that @xmath764 is @xmath36-calabi - yau , and indeed @xmath27-calabi - yau for any @xmath27 .
since the module category of a barbell algebra shares so many properties with the module category of a cluster tilted algebra , one may wonder whether also for a barbell algebra @xmath18 the module category @xmath745 is obtained from a triangulated category @xmath773 by forming @xmath774 for some object @xmath775 in @xmath776 . as idun reiten has pointed out , this is indeed the case if we deal with a barbell algebra @xmath18 with two loops ( as in our running example 2 ) provided we assume that the characteristic of @xmath0 is different from @xmath36 : such an algebra is 2-cy - tilted ( this means : the endomorphism ring of some cluster tilting object of a 2-calabi - yau category , @xcite ) .
namely , if @xmath18 is a barbell algebra with two loops @xmath777 in its quiver @xmath4 and if the characteristic of @xmath0 is different from @xmath36 , then @xmath18 is the jacobian algebra @xmath778 , where @xmath35 is the potential @xmath779 , see @xcite , thus one can apply theorem 3.6 of amiot @xcite .
recall that a ( finite or infinite ) path @xmath780 in the auslander - reiten quiver of a finite dimensional algebra is called _ sectional _ provided @xmath781 is not isomorphic to @xmath782 for all possible @xmath712 .
such a path will be called _
maximal _ provided it is not a proper subpath of some sectional path .
an infinite sectional path involving only monomorphisms will be called a _ mono ray , _ an infinite path involving only epimorphisms will be called an _ epi coray ; _ of course , mono rays start with some module , epi corays end in a module . note that for auslander - reiten components of the form @xmath783 as well as for stable tubes , all maximal sectional paths are mono rays and epi corays .
let @xmath18 be a @xmath0-algebra which is minimal representation - infinite and special biserial .
then any maximal sectional path is a mono ray , an epi coray or the concatenation of an epi coray with a mono ray .
assume that @xmath18 is minimal representation - infinite and special biserial .
let @xmath43 be indecomposable @xmath18-modules with an irreducible monomorphism @xmath784 and an irreducible epimorphism @xmath45 .
then @xmath785 we may assume that there are no nodes : namely , if all the sectional paths of @xmath23 are as mentioned , the same has to be true for @xmath18 : the only maximal sectional paths for @xmath18 to be looked at are those passing through the node .
resolving the node we will obtain sectional paths which are not double infinite paths , thus by the assumption on @xmath23 , we will deal with an epi coray ending in the node and a mono ray starting in the node , thus with a concatenation as listed .
if @xmath18 is hereditary of type @xmath16 , thus a cycle algebra , then any maximal sectional path is a mono ray , an epi coray .
we only have to look at the preprojective component and the preinjective component .
but if @xmath786 is a non - zero map between indecomposable preprojective modules , then @xmath706 has to be always a monomorphism : otherwise , the kernel of @xmath706 would have negative defect , and since the defect of @xmath460 is @xmath787 , it would follow that the image of @xmath706 is a non - zero submodule of @xmath788 with non - negative defect , a contradiction .
the dual argument shows that the maximal sectional paths in the preinjective component are epi corays .
next , let us look at the wind wheels : again , only the non - regular components have to be considered .
but we know how to construct these components : we use mono rays from the preprojective component and the tube @xmath457 as well as epi corays from the tube @xmath451 and from the preinjective component , and in addition rays and corays in the tiles .
but all the maximal sectional paths in the four quarters of a tile are mono rays and epi corays ( in the quarter * i * we have only mono rays , in * iii * only epi corays , whereas * ii * and * iv * have both mono rays and epi corays .
finally , let us look at the barbells .
there is only one non - regular component which has to be treated separately .
the band modules lie in homogeneous tubes and there will be an additional regular tube containing string modules .
what really is of interest are the remaining components @xmath773 , they are of the form @xmath789 .
let us look at the example 2 ( the general case is similar ) .
let @xmath790 be the gei module ( @xcite ) for @xmath773 ( it is the unique module in @xmath773 of minimal length ) and one easily observes that @xmath157 is a word of the form @xmath791 .
it is easy to see that all the modules @xmath792 for @xmath793 are obtained from @xmath794 by adding hooks both on the left and on the right ; similarly , all the modules @xmath795 for @xmath793 are obtained from @xmath796 by adding cohooks both on the left and on the right . but
this implies that all the maximal sectional paths in @xmath773 are concatenation of an epi coray with a mono ray .
it may be helpful to call an indecomposable @xmath18-module a _ valley _
module if it is the concatenation vertex for a sectional path which is the concatenation of an epi coray with a mono ray , and to exhibit corresponding pictures : always we encircle the valleys .
first , we present a non - regular component of a wind wheel : @xmath797 from 0.3 4.3 to 0.7 4.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 4.3 to 2.7 4.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 5.3 to 1.7 5.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.3 5.3 to 3.7 5.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.3 6.3 to 4.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.3 6.3 to 6.6 6.6 \arrow < 1.5 mm > [ 0.25,0.75 ] from 8.3 6.3 to 8.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.3 6.7 to 5.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.3 6.7 to 7.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 9.3 6.7 to 9.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 10.3 5.7 to 10.7 5.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 11.3 4.7 to 11.7 4.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 5.7 to 12.7 5.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 4.7 to 13.7 4.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.3 0.3 to 4.7 0.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.3 0.3 to 6.7 0.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 8.3 0.3 to 8.7 0.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.3 0.7 to 5.7 0.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.3 0.7 to 7.7 0.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 9.3 0.7 to 9.7 0.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.3 1.3 to 5.7 1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.3 1.3 to 7.7 1.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.3 1.7 to 6.7 1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 8.3 1.7 to 8.7 1.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.3 2.3 to 6.7 2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.3 2.7 to 7.7 2.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 6.3 to 0.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.3 6.3 to 2.7 6.7 \arrow < 1.5
mm > [ 0.25,0.75 ] from 10.3 6.3 to 10.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 6.3 to 12.7 6.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 4.3 to 12.7 4.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 2.3 to 0.7 2.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 11.3 5.3 to 11.7 5.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 5.3 to 13.7 5.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 3.3 to 13.7 3.7 \arrow < 1.5
mm > [ 0.25,0.75 ] from 1.3 3.3 to 1.7 3.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 3.7 to 0.7 3.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.3 5.7 to 0.7 5.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 4.7 to 1.7 4.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.3 6.7 to 1.7 6.3 \arrow < 1.5
mm > [ 0.25,0.75 ] from 2.3 5.7 to 2.7 5.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.3 6.7 to 3.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 11.3 6.7 to 11.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 6.7 to 13.7 6.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 12.3 3.7 to 12.7 3.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 13.3 2.7 to 13.7 2.3 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.4 3.9 to 6.7 3.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.4 2.9 to 5.7 2.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.4 1.9 to 4.7 1.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 0.2 0.8 to 3.7 0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 7.3 3.1 to 11.4 3.9 \arrow < 1.5 mm > [ 0.25,0.75 ] from 8.3 2.1 to 12.4
2.9 \arrow < 1.5 mm > [ 0.25,0.75 ] from 9.3 1.1 to 13.4 1.9 \plot 10.3 0.1 13.8 0.8 / \setdots<.1 cm > \plot 4.4 6 5.6 6 / \plot 6.4 6 7.6 6 / \plot 8.4 6 9.6 6 / \setshadegrid span < 0.7 mm > \endpicture}\ ] ] the valleys may be considered as the natural places where to cut such a component into pieces .
of course , in our cut - and - paste process , we followed this rule .
the second example is the non - regular component of the barbell given as example 2 : @xmath798 from 0.3 0.3 to 0.7 0.7 \arrow < 2 mm > [ 0.25,0.75 ] from 2.3 0.3 to 2.7 0.7 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 0.3 to 6.7 0.7 \arrow < 2 mm > [ 0.25,0.75 ] from 0.3 2.3 to 0.7 2.7 \arrow < 2 mm > [ 0.25,0.75 ] from 2.3 2.3 to 2.7 2.7 \arrow < 2 mm > [ 0.25,0.75 ] from 0.3 4.3 to 0.7 4.7 \arrow < 2 mm > [ 0.25,0.75 ] from 2.3 4.3 to 2.7 4.7 \arrow < 2 mm > [ 0.25,0.75 ] from 4.3 4.3 to 4.7 4.7 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 4.3 to 6.7 4.7 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 2.3 to 6.7 2.7 \arrow < 2 mm > [ 0.25,0.75 ] from 3.3 3.3 to 3.7 3.7 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 1.3 to 1.7 1.7 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 3.3 to 1.7 3.7 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 1.3 to 7.7 1.7 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 3.3 to 7.7 3.7 \arrow < 2 mm > [ 0.25,0.75 ] from 1.4 -0.6 to 1.7 -0.3 \arrow < 2 mm > [ 0.25,0.75 ] from 3.4 -0.6 to 3.7 -0.3 \arrow
< 2 mm > [ 0.25,0.75 ] from 5.4 -0.6 to 5.7 -0.3 \arrow < 2 mm > [ 0.25,0.75 ] from 7.4 -0.6 to 7.7 -0.3 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 0.7 to 1.7 0.3 \arrow < 2 mm > [ 0.25,0.75 ] from 3.3 0.7 to 3.7 0.3 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 0.7 to 7.7 0.3 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 2.7 to 1.7 2.3 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 2.7 to 7.7 2.3 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 4.7 to 7.7 4.3 \arrow < 2 mm > [ 0.25,0.75 ] from 5.3 4.7 to 5.7 4.3 \arrow < 2 mm > [ 0.25,0.75 ] from 3.3 4.7 to 3.7 4.3 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 4.7 to 1.7 4.3 \arrow < 2 mm > [ 0.25,0.75 ] from 8.4 5.6 to 8.7 5.3 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 3.7 to 6.7 3.3 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 1.7 to 6.7 1.3 \arrow < 2 mm > [ 0.25,0.75 ] from 0.3 1.7 to 0.7 1.3 \arrow < 2 mm > [ 0.25,0.75 ] from 2.3 1.7 to 2.7 1.3 \arrow < 2 mm > [ 0.25,0.75 ] from 0.3 3.7 to 0.7 3.3 \arrow < 2 mm > [ 0.25,0.75 ] from 2.3 3.7 to 2.7 3.3 \arrow < 2 mm > [ 0.25,0.75 ] from 0.4 5.6 to 0.7 5.3 \arrow < 2 mm > [ 0.25,0.75 ] from 2.4 5.6 to 2.7 5.3 \arrow < 2 mm > [ 0.25,0.75 ] from 4.4 5.6 to 4.7 5.3 \arrow < 2 mm > [ 0.25,0.75 ] from 6.4 5.6 to 6.7 5.3 \plot 1.3 5.3 1.6 5.6 / \plot 3.3 5.3 3.6 5.6 / \plot 5.3 5.3 5.6 5.6 / \plot 7.3 5.3 7.6 5.6 / \plot 8.3 4.3 8.6 4.6 / \plot 8.3 2.3 8.6 2.6 / \plot 8.3 0.3 8.6 0.6 / \plot 8.3 3.7 8.6 3.4 / \plot 8.3 1.7 8.6 1.4 / \plot 8.3 -0.3 8.6 -0.6 / \plot 6.3 -0.3 6.6 -0.6 / \plot 4.3 -0.3 4.6 -0.6 / \plot 2.3 -0.3 2.6 -0.6 / \plot 0.3 -0.3 0.6 -0.6 / \setdots<.1
cm > \plot 4.5 0 5.7 0 / \plot 4.45 4 5.8 4 / \endpicture}\ ] ] of course , when dealing with a barbell and look at a regular component @xmath773 of string modules , say with gei - module @xmath21 , then the valley modules are precisely those which lie on the sectional paths which contain @xmath21 . in all these components , the valleys
provide a clear division into regions with common growth pattern .
for example , in the regions on the left , all irreducible maps are epimorphisms , whereas in the regions on the right , all are monomorphisms .
* part iii .
* the appendix collects some remarks related to the investigations presented above .
first , we show an example of an algebra which may be considered as a twisted version of a barbell .
consider the following algebra : @xmath799 from 1.7 0 to 0.3 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.7 0.9 to 2.3 0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.7 -.9 to 2.3 -.1 \circulararc 320 degrees from -0.05 0.2 center at -0.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from -0.09 0.3 to -0.05 0.2 \put{$\alpha$ } at -2 0 \put{$\beta$ } at 1 0.4 \put{$\gamma$ } at 3 0.9 \put{$\gamma'$ } at 3 -.9 \setdots
< .7 mm > \setquadratic \setquadratic \plot -0.4
-0.4 -0.3 0 -0.4 0.4 / \endpicture}\ ] ] ( or , more generally , the corresponding algebras where @xmath7 and @xmath10 are replaced by longer paths ) .
note that the universal covering are the `` dancing girls '' of brenner - butler .
this is a gorenstein algebra of gorenstein dimension 1 , the minimal injective resolutions of the indecomposable projective modules are as follows : @xmath800 and here is the central part of the non - regular component : @xmath801 from 0.3 4.3 to 0.7 4.7 \arrow < 2 mm > [ 0.25,0.75 ] from 0.3 2.3 to 0.7 2.7 \arrow < 2 mm > [ 0.25,0.75 ] from 0.3 4.3 to 0.7 4.7 \arrow < 2 mm > [ 0.25,0.75 ] from 2.3 4.3 to 2.7 4.7 \arrow < 2 mm > [ 0.25,0.75 ] from 4.3 4.3 to 4.7 4.7 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 4.3 to 6.7 4.7 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 2.3 to 6.7 2.7 \arrow < 2 mm > [ 0.25,0.75 ] from 5.3 3.3 to 5.7 3.7 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 1.3 to 1.7 1.7 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 3.3 to 1.7 3.7 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 1.3 to 7.7 1.7 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 3.3 to 7.7 3.7 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 2.7 to 1.7 2.3 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 2.7 to 7.7 2.3 \arrow < 2 mm > [ 0.25,0.75 ] from 7.3 4.7 to 7.7 4.3 \arrow < 2 mm > [ 0.25,0.75 ] from 5.3 4.7 to 5.7 4.3 \arrow < 2 mm > [ 0.25,0.75 ] from 3.3 4.7 to 3.7 4.3 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 4.7 to 1.7 4.3 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 3.7 to 6.7 3.3 \arrow < 2 mm > [ 0.25,0.75 ] from 6.3 1.7 to 6.7 1.3 \arrow < 2 mm > [ 0.25,0.75 ] from 0.3 1.7 to 0.7 1.3 \arrow < 2 mm > [ 0.25,0.75 ] from 2.3 1.7 to 2.7 1.3 \arrow < 2 mm > [ 0.25,0.75 ] from 0.3 3.7 to 0.7 3.3 \arrow < 2 mm > [ 0.25,0.75 ] from 4.3 3.7 to 4.7 3.3 \arrow < 2 mm > [ 0.25,0.75 ] from 5.3 2.7 to 5.7 2.3 \arrow < 2 mm > [ 0.25,0.75 ] from 0.4 5.6 to 0.7 5.3 \arrow < 2 mm > [ 0.25,0.75 ] from 2.4 5.6 to 2.7 5.3 \arrow < 2 mm > [ 0.25,0.75 ] from 4.4 5.6 to 4.7 5.3 \arrow < 2 mm > [ 0.25,0.75 ] from 6.4 5.6 to 6.7 5.3 \arrow < 2 mm > [ 0.25,0.75 ] from 1.3 1.3 to 1.7 1.7 \arrow < 2 mm > [ 0.25,0.75 ] from 0.3 2 to 0.7 2 \arrow < 2 mm > [ 0.25,0.75 ] from 1.2 2 to 1.6 2 \arrow < 2 mm > [ 0.25,0.75 ] from 2.4 2 to 2.7 2 \arrow < 2 mm > [ 0.25,0.75 ] from 6.4 2 to 6.6 2 \arrow < 2 mm > [ 0.25,0.75 ] from 7.4 2 to 7.7 2 \plot 8.4 5.6 8.7 5.3 / \plot 8.3 2 8.7 2 / \plot 1.3 5.3 1.6 5.6 / \plot 3.3 5.3 3.6 5.6 / \plot 5.3 5.3 5.6 5.6 / \plot 7.3 5.3 7.6 5.6 / \plot 8.3 4.3 8.6 4.6
/ \plot 8.3 2.3 8.6 2.6 / \plot 8.3 3.7 8.6 3.4 / \plot 8.3 1.7 8.6 1.4 / \setdots<1 mm > \plot 2.2 4 3.5 4 / \plot 0.2 1 0.8 1 / \plot 1.2 1 2.5 1 / \plot 7.5 1 8.8 1 / \setshadegrid span < 0.7 mm > \vshade 0 0.8 5.6 < , z , , > 1 0.8 5.6 / \vshade 1 3 5.6 < z , z , , > 2 4 5.6 < z , z , , > 4 4 5.6 < z , z , , > 7 0.8 5.6 < , z , , > 8.7 0.8 5.6 / \vshade 1 0.8 3 < z , z , , > 2.2 0.8 2.2 < z , z , , > 3.2 0.8 2 / \setdashes < 2 mm > \plot 2.3 3.7 4.9 1.1 / \plot 5.1 1.1 5.8 1.8 / \plot 2.3 2.3 3.8 3.8 / \plot 3.3 1.3 4.8 2.8 / \plot 3.4 2 5.7 2 / \endpicture}\ ] ]
consider the path algebra of the quiver @xmath802 from 0.8 0 to 0.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.8 0 to 1.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.2 0 to 2.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.8 0 to 3.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.8 0 to 4.2 0 \put{${\scriptstyle } 0 $ } at 0 0.3 \put{${\scriptstyle } 1'$ } at 1 0.3 \put{${\scriptstyle } 2'$ } at 2 0.3 \put{${\scriptstyle } 1''$ } at 3 0.3 \put{${\scriptstyle } 2''$ } at 4 0.3 \put{${\scriptstyle } 3 $ } at 5 0.3 \put{$\alpha$ } at 0.5 0.3 \put{$\beta'$ } at 1.5 0.3 \put{$\gamma$ } at 2.5 0.3 \put{$\beta''$ } at 3.5 0.3 \put{$\delta$ } at 4.5 0.3 \endpicture}\ ] ] and barify the arrows @xmath186 and @xmath187 .
the we obtain the quiver @xmath803 from 0.8 0 to 0.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.8 0 to 2.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.25 0.125 to 1.2 0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.25 -.125 to 1.2 -.1 \put{${\scriptstyle } 0 $ } at 0 0.3 \put{${\scriptstyle } 1 $ } at 1 0.3 \put{${\scriptstyle } 2 $ } at 2 0.3 \put{${\scriptstyle } 3 $ } at 3 0.3 \put{$\alpha$ } at 0.5 0.3 \put{$\beta$ } at 1.5 0.4 \put{$\gamma$ } at 1.5 -0.4 \put{$\delta$ } at 2.5 0.3 \setquadratic \plot 1.8 0.1 1.5 0.2 1.2 0.1 / \plot 1.8 -.1 1.5 -.2 1.2 -.1 / \setdots < .5 mm > \plot 0.6 -0.1 1 -0.13 1.3 -0.3 / \plot 2.6 -0.1 2 -0.13 1.7 -0.3 / \endpicture}\ ] ] here , starting with a representation - finite algebra , we obtain a tame one . similarly ,
if we start with the following tame quiver , the barification of @xmath562 and @xmath563 yields a wild algebra : @xmath804 from 0.8 0 to 0.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.8 0 to 1.2 0 \arrow < 1.5 mm >
[ 0.25,0.75 ] from 2.2 0 to 2.8 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.8 0 to 3.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.8 0 to 4.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 6.8 0.5 to 6.2 0.5 \put{${\scriptstyle } 0 $ } at 0 0.3 \put{${\scriptstyle } 1'$ } at 1 0.3 \put{${\scriptstyle } 2'$ } at 2 0.3 \put{${\scriptstyle } 1''$ } at 3 0.3 \put{${\scriptstyle } 2''$ } at 4 0.3 \put{${\scriptstyle } 3 $ } at 5 0.3 \put{${\scriptstyle } 4 $ } at 6 0.7 \put{${\scriptstyle } 6 $ } at 6 -0.3 \put{${\scriptstyle } 5 $ } at 7 0.7 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.8 0.4 to 5.2 0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 5.8 -.4 to 5.2 -.1 \put{$\alpha$ } at 0.5 0.3 \put{$\beta'$ } at 1.5 0.3 \put{$\gamma$ } at 2.5 0.3 \put{$\beta''$ } at 3.5 0.3 \put{$\delta$ } at 4.5 0.3 \endpicture}\ ] ] @xmath805 from 0.8 0 to 0.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 2.8 0 to 2.2 0 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.25 0.125 to 1.2 0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 1.25 -.125 to 1.2 -.1 \put{${\scriptstyle } 0 $ } at 0 0.3 \put{${\scriptstyle } 1 $ } at 1 0.3 \put{${\scriptstyle } 2 $ } at 2 0.3 \put{${\scriptstyle } 3 $ } at 3 0.3 \put{$\alpha$ } at 0.5 0.3 \put{$\beta$ } at 1.5 0.4 \put{$\gamma$ } at 1.5 -0.4 \put{$\delta$ } at 2.5 0.3 \setquadratic \plot 1.8 0.1 1.5 0.2 1.2 0.1 / \plot 1.8 -.1 1.5 -.2 1.2 -.1 / \setdots < .5 mm > \plot 0.6 -0.1 1 -0.13 1.3 -0.3 / \plot 2.6 -0.1 2 -0.13 1.7 -0.3 / \setsolid \arrow < 1.5 mm > [ 0.25,0.75 ] from 4.8 0.5 to 4.2 0.5 \put{${\scriptstyle } 4 $ } at 4 0.7 \put{${\scriptstyle } 6 $ } at 4 -0.3 \put{${\scriptstyle } 5 $ } at 5 0.7 \arrow < 1.5 mm
> [ 0.25,0.75 ] from 3.8 0.4 to 3.2 0.1 \arrow < 1.5 mm > [ 0.25,0.75 ] from 3.8 -.4 to 3.2 -.1 \endpicture}\ ] ]
we have mentioned in the introduction that the recent paper @xcite of bongartz has drawn the attention to the minimal representation - infinite algebras which have a good cover @xmath1 , such that all finite convex subcategories of @xmath1 are representation - finite .
as we show above , these algebras are special biserial and can be completely classified .
the title of the bongartz paper @xcite indicates that his main concern was to proof the following theorem : _ let @xmath18 be a finite dimensional @xmath0-algebra where @xmath0 is an algebraically closed field .
if there exists an indecomposable @xmath18-module of length @xmath806 , then there exists an indecomposable @xmath18-module of length @xmath807 .
_ unfortunately , the statement does not assert any relationship between the modules of length @xmath27 and those of length @xmath807 .
there is the following open problem : _ given an indecomposable @xmath18-module @xmath21 of length @xmath808 .
is there an indecomposable submodule or factor module of length @xmath809 _ the three subspace quiver shows that this may not be true in case the field @xmath0 is not algebraically closed , say if it is finite field with few elements .
in @xcite we slightly modified the arguments of bongartz in order to strengthen his assertion . using induction ,
one may define _ accessible _ modules : first , the simple modules are accessible .
second , a module of length @xmath311 is accessible provided it is indecomposable and there is a submodule or a factor module of length @xmath807 which is accessible .
the open problem mentioned above can be reformulated as follows : are all indecomposable representations of a @xmath0-algebra @xmath18 , where @xmath0 is algebraically closed , accessible ?
this is known to hold in case @xmath18 is representation - finite and the aim of @xcite was to show that any representation - infinite algebra over an algebraically closed field has at least accessible modules of arbitrarily large length . in dealing with special biserial algebras , we do not have to worry about the size of the base field @xmath0 .
the following assertion is vaild for @xmath0-algebras with @xmath0 an arbitrary field .
any indecomposable representation of a special biserial algebra is accessible .
it is obvious that string modules are accessible , thus we only have to consider band modules
. it will be sufficient to show the following : _ any band module has a maximal submodule which is a string module .
_ thus , let @xmath21 be a band module .
first , let us consider the special case of dealing with the kronecker algebra , thus @xmath810 with vector spaces @xmath811 and invertible linear maps @xmath812 let @xmath813 be a submodule of @xmath21 which is a band module and of smallest possible dimension .
note that @xmath813 is uniquely determined and is contained in any non - zero regular submodule of @xmath21 .
let @xmath814 and choose a direct complement @xmath815 for @xmath816 .
then @xmath817 is a submodule of @xmath21 , and of course a maximal one .
we claim that @xmath383 is a string module . as a submodule of a regular kronecker module ,
we can write @xmath818 with @xmath819 preprojective and @xmath820 regular .
but @xmath820 has to be zero , since otherwise @xmath821 , thus @xmath822 , a contradiction .
this shows that @xmath383 is a direct sum of say @xmath39 indecomposable preprojective kronecker modules . since @xmath823
, it follows that @xmath824 this shows that @xmath383 is an indecomposable preprojective kronecker module and thus a string module .
now consider an arbitrary special biserial algebra @xmath18 with quiver @xmath4 .
there is a primitive cyclic word @xmath825 and an indecomposable vector space automorphism @xmath826 such that @xmath827 let @xmath222 with letters @xmath71 ; we can assume that @xmath828 is a direct letter , whereas @xmath829 is an inverse letter .
denote by @xmath830 the terminal point of @xmath71 , for @xmath831 then @xmath21 is given by @xmath39 copies @xmath832 of @xmath101 , indexed by @xmath833 , such that the arrows of @xmath4 operate as follows : if @xmath834 is a direct letter ( thus an arrow ) , then @xmath7 is the identity map @xmath835 , if @xmath71 is an inverse letter , say @xmath836 for some arrow @xmath7 , then @xmath7 is the identity map @xmath837 for @xmath838 and the map @xmath839 for @xmath840 . note that @xmath841 is a band module for the kronecker quiver , thus , as we have seen already , it has a maximal submodule @xmath842 which is a string module . we obtain a submodule @xmath383 of @xmath843 by taking the subspace @xmath844 , where @xmath391 is considered as a subspace of @xmath845 .
since @xmath842 is a string module for the kronecker algebra , it follows that @xmath383 is a string @xmath18-module .
it should be mentioned that algebras defined by a quiver , commutativity relations and zero relations can be considered as factor algebras of a semigroup algebra @xmath846 $ ] modulo a one - dimensional ideal generated by a central idempotent @xmath847 , thus the paper may be seen as dealing with a class of minimal representation - infinite semigroups .
let @xmath19 be a semigroup ( a set with an associative binary operation ) .
an element @xmath355 of @xmath19 is called a _
zero _ element provided @xmath848 for all @xmath849 . of course
, if there is a zero element , then it is uniquely determined .
let @xmath19 be a semigroup with zero element @xmath355 , we want to consider the semigroup algebra @xmath846 $ ] . obviously , the element @xmath355 considered as an element of @xmath846 $ ] is a central idempotent and the ideal @xmath850 generated by @xmath355 is one - dimensional , thus @xmath355 is a primitive idempotent . with @xmath355
also @xmath851 is a central idempotent , and we obtain a direct decomposition of @xmath846 $ ] as a product of @xmath0-algebras @xmath852 = \langle z\rangle \times \langle 1-z\rangle = kz \times k[s](1-z).\ ] ] one may call @xmath853(1-z ) = k[s]/\langle z\rangle$ ] the _ reduced _ semigroup algebra of @xmath19 .
it follows that the modules for the reduced semigroup algebra of @xmath19 are precisely the @xmath846 $ ] modules @xmath21 with @xmath854 the product decomposition of the semigroup algebra @xmath846 $ ] shows that there is a unique simple ( one - dimensional ) @xmath846$]-module which is not annihilated by @xmath355 , all other indecomposable @xmath846$]-modules are annihilated by @xmath355 and thus are modules over the reduced semigroup - algebra . given a quiver @xmath4 ,
let @xmath855 be obtained from the set of all paths ( including the paths of length @xmath685 ) by adding an element @xmath355 ( it will become the zero element ) .
as in the definition of the path algebra @xmath856 of a quiver , define the product of two paths to be the concatenation , if it exist , and to be @xmath355 otherwise . in this way , @xmath855 becomes a semigroup with zero element @xmath355 , and _ the reduced semigroup algebra of @xmath855 can be identified with the path algebra @xmath857 $ ] of the quiver @xmath4 . _ of course
, if we deal with a set @xmath49 of commutativity relations and zero relations , then we may consider the factor semigroup @xmath858 , this is again a semigroup with zero , and its reduced semigroup algebra is just the algebra defined by the quiver @xmath4 and the relations @xmath49 . | let @xmath0 be a field .
a finite dimensional @xmath0-algebra is said to be minimal representation - infinite provided it is representation - infinite and all its proper factor algebras are representation - finite .
our aim is to classify the special biserial algebras which are minimal representation - infinite .
the second part describes the corresponding module categories .
primary 16g20 , 16g60 .
secondary 16d90 , 16g70 .
minimal representation - infinite algebras , special biserial algebras . quiver .
auslander - reiten quiver .
auslander - reiten quilt .
sectional paths .
irreducible maps .
gorenstein algebras .
semigroup algebras . |
following the first paper of this series @xcite , we address discrete systems and differential - difference equations , which have become topics of increasing physical and mathematical importance . the variety of physical applications where such models are relevant , and their significant differences from the mathematical theory of partial differential equations , contribute to recent interest in these topics .
the applicability of such models extends to areas as diverse as nonlinear optics , atomic and soft condensed - matter physics , as well as biophysics : specific details and references can be found in our first paper @xcite as well as in reviews @xcite .
the second paper is devoted to existence and stability of coherent structures in two - dimensional lattices , which include both discrete solitons @xcite and discrete vortices @xcite .
these two - dimensional coherent structures have emerged recently in studies of photorefractive crystals in nonlinear optics @xcite and droplets of optical lattices in bose - einstein condensates @xcite . a significant boost to this subject
was given by the experimental realization of two - dimensional photonic crystal lattices with periodic potentials based on the ideas of @xcite .
as a result , discrete solitons were observed in @xcite , while more complex structures such as dipoles , soliton trains and vector solitons were observed in @xcite .
most recently , observations of discrete vortices were reported by two independent groups @xcite where the fundamental vortices with topological charge _
one _ were experimentally created and detected in photorefractive crystals .
two main examples of charge - one discrete vortices include a _ vortex cross _ ( an on - site centered vortex ) and a _ vortex cell _ ( an off - site centered vortex ) .
these structures were also recently predicted in a continuous two - dimensional model with the periodic potential @xcite .
these discoveries have stimulated further theoretical work and numerical computations .
thus , while in @xcite , discrete vortices of charge _ two _ were shown to be unstable , recently in @xcite discrete vortices of charge _
three _ were found to be stable .
based on the theoretical predictions of @xcite , further experiments on localized structures were undertaken to unveil other interesting structures , such as the discrete soliton necklace which is more globally stable compared to the charge three vortex @xcite .
the discrete vortices have been also extended to three - dimensional discrete models @xcite .
furthermore , asymmetric vortices have been recently predicted in the two - dimensional lattices in @xcite .
the above activity clearly signals the importance and experimental relevance of discrete solitons and vortices in two - dimensional discrete lattices
. however , most of the above mentioned works are predominantly of experimental or numerical nature , while the mathematical theory of existence and stability of discrete localized structures has not been developed to a similar extent .
the aim of the present paper is to develop a categorization of discrete solitons and vortices in the discrete two - dimensional nonlinear schrdinger ( nls ) equations .
we start from a well - understood limit ( the so - called anti - continuum case of zero coupling between the lattice nodes ) and examining persistence of the limiting solutions for small coupling by means of the lyapunov - schmidt theory .
this method allows us to discuss persistence and stability of the localized structures by analyzing finite - dimensional linear eigenvalue problems .
the theoretical predictions agree well with full numerical computations of the discrete two - dimensional nls equation .
our main results are summarized for the simplest localized structures in table 1 .
these results corroborate and extend the previously reported experimental and numerical findings .
we quantify the stability of the charge - one vortex in accordance to @xcite , the instability of the charge - two vortex in accordance to @xcite and the stability of the charge - three vortex in accordance to @xcite .
we further demonstrate the instability of all asymmetric vortices proposed in @xcite .
furthermore , our results can be used to extract the spectral stability of the dipole mode considered in @xcite and of the soliton necklace of @xcite . [ cols="<,<,<,<,<",options="header " , ] * table 1 : * the numbers of small unstable , stable and zero eigenvalues , associated to vortices of the discrete two - dimensional nls equation with small coupling constant .
the paper is structured as follows .
abstract results on existence of discrete solitons and vortices are derived in section 2 .
persistence of localized modes for a particular square discrete contour is considered in section 3 .
stability of the persistent solutions is addressed in section 4 .
analytical results are compared to numerical computations in section 5 .
section 6 concludes the paper with a summary of our results and discussions of interesting directions for future study .
we consider the discrete nonlinear schrdinger ( nls ) equation in two space dimensions @xcite : @xmath0 where @xmath1 , @xmath2 , and @xmath3 is the inverse squared step size of the discrete two - dimensional nls lattice .
time - periodic localized modes of the discrete nls equation ( [ 2nls ] ) take the form : @xmath4 where @xmath5 and @xmath6 are parameters . since localized modes in the focusing nls lattice ( [ 2nls ] ) with @xmath7 may exist only for @xmath8 @xcite and the parameter @xmath9 is scaled out by the scaling transformation , @xmath10 the parameter @xmath11 will henceforth be set to @xmath12 . in this case , the complex - valued @xmath13 solve the nonlinear difference equations on @xmath2 : @xmath14 as @xmath15 , the localized modes of the difference equations ( [ 2difference ] ) are given by the limiting solution : @xmath16 where @xmath17 is a finite set of nodes on the lattice @xmath2 and @xmath18 are parameters for @xmath19 . since @xmath20 is arbitrary in the ansatz ( [ soliton - shape-2 ] ) , we can set @xmath21 for a particular node @xmath22 . using this convention ,
we define two special types of localized modes , called discrete solitons and vortices . [ definition - soliton ] the localized solution of the difference equations ( [ 2difference ] ) with @xmath7 , that has all real - valued amplitudes @xmath13 , @xmath2 and satisfies the limit ( [ 2soliton ] ) with all @xmath23 , @xmath19 , is called a discrete soliton .
[ definition - vortex ] let @xmath17 be a simply - connected discrete contour on the plane @xmath2 .
the localized solution of the difference equations ( [ 2difference ] ) with @xmath7 , that has complex - valued @xmath13 , @xmath2 and satisfies the limit ( [ 2soliton ] ) with @xmath24 $ ] , @xmath19 , is called a discrete vortex .
[ definition - symmetric ] let @xmath17 be a simply - connected discrete contour on the plane @xmath2 , such that each node @xmath25 has exactly two adjacent nodes in vertical or horizontal directions along @xmath17 .
if the phase difference between two adjacent @xmath18 for @xmath19 is constant in @xmath17 , the discrete vortex is called symmetric . otherwise , it is called asymmetric .
the total number of @xmath26 phase shifts across the closed contour @xmath17 is called the topological charge of the discrete vortex . in particular , we consider the ordered simply - connected discrete contour @xmath27 : @xmath28 where @xmath29 . according to definition [ definition - symmetric ] , the contour @xmath30 for a fixed @xmath31 could support symmetric and asymmetric vortices with some charge @xmath32 .
an example of the simplest vortices for @xmath33 is the symmetric charge - one vortex cell ( @xmath34 , @xmath35 , @xmath36 , @xmath37 ) @xcite and an asymmetric charge - one vortex ( @xmath34 , @xmath38 , @xmath39 , @xmath40 ) , where @xmath41 @xcite .
it follows from the general method @xcite that the discrete solitons of the two - dimensional nls lattice ( [ 2difference ] ) ( see definition [ definition - soliton ] ) can be continued to the domain @xmath42 for some @xmath43 .
it is more complicated to find a configuration of @xmath18 for @xmath44 that allows us to continue the discrete vortices ( see definition [ definition - vortex ] ) for @xmath7 .
the continuation of the discrete solitons and vortices is based on the implicit function theorem and lyapunov schmidt reduction theorem @xcite .
abstract results on existence of such continuations are formulated and proved below , after the introduction of some relevant notations .
let @xmath45 be a small neighborhood of @xmath15 , such that @xmath46 for some @xmath47 .
let @xmath48 and @xmath49 be the torus on @xmath50^n$ ] , such that @xmath18 for @xmath19 form a vector @xmath51 .
let @xmath52 be the hilbert space of square - summable complex - valued sequences @xmath53 , equipped with the inner product and the norm : @xmath54 let @xmath55 denote an infinite - dimensional vector in @xmath56 that consists of components @xmath57 for all @xmath2 .
[ proposition - existence ] there exists a unique ( discrete soliton ) solution of the difference equations ( [ 2difference ] ) in the domain @xmath58 that satisfies ( i ) @xmath59 , @xmath2 and ( ii ) @xmath60 , where @xmath61 is given by ( [ 2soliton ] ) with @xmath62 , @xmath19 . the solution @xmath63 is analytic in @xmath58 .
assume that @xmath64 for all @xmath2 .
the difference equations ( [ 2difference ] ) are rewritten as zeros of the nonlinear vector - valued function : @xmath65 the mapping @xmath66 is @xmath67 on @xmath68 and has a bounded continuous frchet derivative , given by : @xmath69 where @xmath70 is the shift operator , such that @xmath71 .
it is obvious that @xmath72 where @xmath73 is the discrete soliton of definition [ definition - soliton ] and @xmath74 is the operator @xmath75 computed at @xmath76 and @xmath15 .
it follows from ( [ eq1])([eq2 ] ) that @xmath77 has a bounded inverse . by the implicit function theorem ( *
* appendix 1 ) , there exists a local @xmath67 mapping @xmath78 , such that @xmath63 is continuous in @xmath79 and @xmath80 .
moreover , since @xmath81 is analytic in @xmath58 , then @xmath63 is analytic in @xmath79 ( * ? ?
* chapter 2.2 ) .
[ remark - complex ] proposition [ proposition - existence ] does not exclude a possibility of continuation of the limiting solution ( [ 2soliton ] ) with @xmath23 for all @xmath44 to the complex - valued solution @xmath82 in @xmath58 .
[ proposition - reductions ] there exists a vector - valued function @xmath83 , such that the limiting solution ( [ 2soliton ] ) is continued to the domain @xmath58 if and only if @xmath84 is a root of @xmath85 in @xmath58 .
moreover , the function @xmath86 is analytic in @xmath58 and @xmath87 for any @xmath88 . when @xmath89 for some @xmath2 , the difference equations ( [ 2difference ] ) are complemented by the complex conjugate equations in the abstract form : @xmath90 taking the frchet derivative of @xmath91 with respect to @xmath92 and @xmath93 , we compute the linearization operator @xmath94 for the difference equations ( [ 2difference ] ) : @xmath95 let @xmath96 .
it is clear that @xmath97 is a self - adjoint fredholm operator of index zero with @xmath98 .
moreover , eigenvectors of @xmath99 re - normalize the parameters @xmath18 for @xmath19 in the limiting solution ( [ 2soliton ] ) . by the lyapunov reduction theorem ( * ? ? ?
* chapter 7.1 ) , there exists a decomposition @xmath100 , such that @xmath101 is defined in terms of the projections to @xmath102 .
let @xmath103 be a set of @xmath104 linearly independent eigenvectors in the kernel of @xmath105 .
it follows from the representation , @xmath106 that each eigenvector @xmath107 in the set @xmath108 has the only non - zero element @xmath109 at the @xmath110-th position of @xmath111 . by projections of the nonlinear equations ( [ nonlinear - equations ] ) to @xmath99 , we derive an implicit representation for the functions @xmath112 : @xmath113 for @xmath19 , where the factor @xmath114 is introduced for convenient notations .
let @xmath115 for @xmath19 and @xmath116 for @xmath117 . since eigenvectors of @xmath99 are excluded from the solution @xmath92 in @xmath118 , we have @xmath119 for @xmath19 , such that @xmath120 and @xmath121 for any @xmath51 .
since @xmath122 is analytic in @xmath58 , then @xmath112 is analytic in @xmath79 ( * ? ? ?
* appendix 3 ) .
the function @xmath86 can be expanded into convergent taylor series in @xmath45 : @xmath123 if the root @xmath124 of @xmath125 is analytic in @xmath58 , then the solution @xmath82 is analytic in @xmath58 , such that @xmath126 where @xmath73 is given by ( [ 2soliton ] ) .
[ lemma - gauge ] let @xmath124 be a root of @xmath127 in @xmath58 .
an arbitrary shift @xmath128 , where @xmath5 and @xmath129 , gives a one - parameter family of roots of @xmath127 for the same @xmath130 .
the statement follows from the symmetry of the difference equations ( [ 2difference ] ) with respect to gauge transformation ( * ? ? ?
* chapter 7.3 ) .
[ proposition - sufficient ] let @xmath131 be the root of @xmath132 and @xmath133 be the jacobian matrix of @xmath134 at @xmath135 .
if the matrix @xmath133 has a simple zero eigenvalue , there exists a unique ( modulo gauge transformation ) analytic continuation of the limiting solution ( [ 2soliton ] ) to the domain @xmath58 . by lemma [ lemma - gauge ] , the matrix @xmath133 has always a non - empty kernel with the eigenvector @xmath136 due to gauge transformation .
let @xmath137 be the constrained subspace of @xmath138 : @xmath139 if the matrix @xmath133 is non - singular in the subspace @xmath137 , then there exists a unique ( modulo the shift ) analytic continuation of the root @xmath131 in @xmath140 by the implicit function theorem , applied to the nonlinear equation @xmath141 ( * ? ? ?
* appendix 1 ) .
[ lemma - nonexistence ] let @xmath131 be a @xmath142-parameter solution of @xmath143 and @xmath133 have a zero eigenvalue of multiplicity @xmath144 , where @xmath145 .
the limiting solution ( [ 2soliton ] ) can be continued in the domain @xmath146 only if @xmath147 is orthogonal to @xmath148 .
let @xmath149 and @xmath150 be eigenvectors of @xmath148 .
we define the constrained subspace of @xmath137 : @xmath151 if @xmath152 , the lyapunov - schmidt reduction theorem in finite dimensions ( * ? ? ?
* chapter 1.3 ) shows that the solution @xmath153 can not be continued in @xmath58 .
proposition [ proposition - reductions ] gives an abstract formulation of the continuation problem for the limiting solution ( [ 2soliton ] ) for @xmath154 .
proposition [ proposition - sufficient ] gives a sufficient condition of existence and uniqueness ( up to gauge invariance ) of such continuations .
proposition [ lemma - nonexistence ] gives a sufficient condition for termination of multi - parameter solutions .
particular applications of propositions [ proposition - reductions ] , [ proposition - sufficient ] and [ lemma - nonexistence ] are limited by the complexity of the set @xmath17 in the limiting solution ( [ 2soliton ] ) , since computations of the vector - valued function @xmath155 , @xmath156 , and the jacobian matrix @xmath133 could be technically involved .
we apply the abstract results of propositions [ proposition - reductions ] , [ proposition - sufficient ] and [ lemma - nonexistence ] to the simply - connected discrete contour @xmath30 , defined in ( [ family-3 ] ) .
we consider discrete solitons and vortices on the contour @xmath30 defined by ( [ family-3 ] ) .
let the set @xmath157 correspond to the ordered contour @xmath30 , starting at @xmath158 , @xmath159 and ending at @xmath160 , where @xmath161 . in what follows , we use the periodic boundary conditions for @xmath157 on the circle from @xmath162 to @xmath163 , such that @xmath164 , @xmath165 , and so on .
the discrete vortex has the charge @xmath32 if the phase difference changes on @xmath166 along the discrete contour @xmath30 . by gauge transformation
, we can always set @xmath167 for convenience .
we will also choose @xmath168 with @xmath169 for convenience , which corresponds to discrete vortices with @xmath170 .
substituting the limiting solution @xmath61 in the bifurcation function ( [ g - function ] ) , we find that @xmath134 in the taylor series ( [ taylor ] ) is non - zero for the contour @xmath30 and it takes the form : @xmath171 the bifurcation equations @xmath172 are rewritten as a system of @xmath104 nonlinear equations for @xmath104 parameters
@xmath173,@xmath174, ... ,@xmath175 as follows : @xmath176 we classify all solutions of the bifurcation equations ( [ sin3 ] ) and give explicit examples for @xmath33 and @xmath177 .
[ lemma - asymmetric ] let @xmath178 for @xmath179 , such that @xmath167 , @xmath168 , and @xmath180 , where @xmath161 , @xmath181 and @xmath32 is the vortex charge .
all solutions of the bifurcation equations ( [ sin3 ] ) reduce to the four families : + @xmath182 ( i ) discrete solitons with @xmath183 and @xmath184 such that the set @xmath185 includes @xmath186 coefficients @xmath187 and @xmath188 coefficients @xmath189 , where @xmath190 .
@xmath182 ( ii ) symmetric vortices of charge @xmath32 with @xmath191 , where @xmath192 , and @xmath193 such that all @xmath104 coefficients are the same : @xmath194 .
@xmath182 ( iii ) one - parameter families of asymmetric vortices of charge @xmath195 with @xmath196 and @xmath197 such that the set @xmath185 includes @xmath198 coefficients @xmath199 and @xmath198 coefficients @xmath200 .
@xmath182 ( iv ) zero - parameter asymmetric vortices of charge @xmath201 and @xmath202 such that the set @xmath185 includes @xmath203 coefficients @xmath204 and @xmath205 coefficients @xmath206 and the family ( iv ) does not reduce to any of the families ( i)(iii ) .
all solutions of the bifurcation equations ( [ sin3 ] ) are given by the binary choice ( [ asymmetric - vortices1 ] ) in the two roots of the sine - function on @xmath207 $ ] , where the first choice gives @xmath208 and the second choice gives @xmath209 .
let us assume that there are totally @xmath203 first choices and @xmath210 second choices , where @xmath211 .
then , we have @xmath212 where @xmath32 is the integer charge of the discrete vortex . there are only two solutions of the above equation . when @xmath213 is arbitrary parameter , we have @xmath214 and @xmath195 , which gives the one - parameter family ( iii ) . when @xmath215 is fixed , we have @xmath216 when @xmath217 , we have the family ( i ) with @xmath218 phases @xmath219 and @xmath220 phases @xmath221 . since discrete solitons do not have topological charge
, the parameter @xmath32 could be half - integer : @xmath222 , where @xmath190 .
when @xmath223 , we have the family ( ii ) for any @xmath192 .
other choices of @xmath203 , which are irreducible to the families ( i)(iii ) , produce the family ( iv ) .
[ remark - supersymmetric ] the one - parameter family ( iii ) connects special solutions of the families ( i ) and ( ii ) .
when @xmath224 and @xmath225 , the family ( iii ) reduces to the family ( i ) with @xmath226 . when @xmath227 , the family ( iii ) reduces to the family ( ii ) with @xmath195
we shall call the corresponding solutions of family ( i ) as the super - symmetric soliton and of family ( ii ) as the super - symmetric vortex .
there exist @xmath228 solutions of family ( i ) , @xmath229 solutions of family ( ii ) , and @xmath230 solutions of family ( iii ) , where @xmath231 the number @xmath232 of solutions of family ( iv ) can not be computed in general .
we consider such solutions only in the explicit examples of @xmath33 and @xmath177 .
* example @xmath33 and @xmath233 : * there are @xmath234 solutions of family ( i ) , @xmath235 solution of family ( ii ) , @xmath236 solutions of family ( iii ) , and no solutions of family ( iv ) . the three one - parameter asymmetric vortices are given explicitly by @xmath237 * example @xmath177 and @xmath238 : * there are @xmath239 solutions of family ( i ) , @xmath240 solutions of family ( ii ) , @xmath241 solutions of family ( iii ) , and @xmath242 solutions of family ( iv ) .
the three symmetric vortices have topological charge @xmath243 ( @xmath244 ) , @xmath245 ( @xmath227 ) , and @xmath246 ( @xmath247 ) .
the one - parameter asymmetric vortices include @xmath248 combinations of @xmath249 upper choices and @xmath249 lower choices in ( [ asymmetric - vortices1 ] ) , starting with the following three solutions : @xmath250 and so on .
the zero - parameter asymmetric vortices include @xmath251 combinations of vortices with @xmath243 for seven phase differences @xmath252 and one phase difference @xmath253 and @xmath251 combinations of vortices with @xmath246 for one phase difference @xmath252 and seven phase differences @xmath253 .
we compute the jacobian matrix @xmath133 from the bifurcation function @xmath254 , given in ( [ sin - bifurcation ] ) : @xmath255 subject to the periodic boundary conditions . the matrix @xmath133 is defined by the coefficients @xmath178 for @xmath179 .
it has the same structure as that in the perturbation theory of continuous multi - pulse solitons in coupled nls equations @xcite .
three technical results establish location of eigenvalues of the matrix @xmath133 .
[ lemma - periodic ] let @xmath256 , @xmath257 , and @xmath258 be the numbers of negative , zero and positive terms of @xmath259 , @xmath179 , such that @xmath260 .
let @xmath261 , @xmath262 , and @xmath263 be the numbers of negative , zero and positive eigenvalues of the matrix @xmath264 , defined by ( [ melements3 ] ) .
assume that @xmath265 and denote : @xmath266 if @xmath267 , then , @xmath268 , and either @xmath269 , @xmath270 or @xmath271 , @xmath272 .
moreover , @xmath261 is even if @xmath273 and is odd if @xmath274 .
if @xmath275 , then @xmath276 . the first statement follows from appendix a of @xcite .
let the determinant equation be @xmath277 . by induction arguments in @xcite , it can be found that @xmath278 and @xmath279 . on the other hand , @xmath280 , where @xmath281 ( which exists always with the eigenvector @xmath282 , see proposition [ proposition - sufficient ] ) .
then , it is clear that @xmath283 . when @xmath284 , at least one more eigenvalue is zero , such that @xmath276 .
[ lemma - constant ] let all coefficients @xmath285 , @xmath179 be the same : @xmath286 .
eigenvalues of the matrix @xmath133 are computed explicitly as follows : @xmath287 when @xmath286 , @xmath179 , the eigenvalue problem for the matrix @xmath133 takes the form of the linear difference equation with constant coefficients : @xmath288 the discrete fourier mode @xmath289 for @xmath290 results in the solution ( [ explicit - eigenvalue ] ) .
[ lemma - alternating ] let all coefficients @xmath291 , @xmath179 alternate the sign : @xmath292 , where @xmath161 .
eigenvalues of the matrix @xmath133 are computed explicitly as follows : @xmath293 such that @xmath294 , @xmath295 , and @xmath296 .
these numbers do not change if the set @xmath297 is obtained from the sign - alternating set by permutations . when @xmath292 , @xmath298 , the eigenvalue problem for the matrix @xmath133 takes the form of a coupled system of linear difference equation with constant coefficients : @xmath299 subject to the periodic boundary conditions : @xmath300 and @xmath301 .
the discrete fourier mode @xmath302 and @xmath303 for @xmath304 results in the solution ( [ explicit - eigenvalue - alternating ] ) . in this case
, we have @xmath294 , @xmath295 , and @xmath305 , such that @xmath306 in the determinant equation @xmath307 . in order to prove that @xmath295 remains invariant with respect to permutations of the sign - alternating set @xmath185
, we find from mathematica that @xmath308 where @xmath309 are numerical coefficients .
let @xmath310 denote the sign - alternating set @xmath185 , such that @xmath311 , and @xmath312 denote a set obtained from @xmath310 by permutations .
it is clear that @xmath313 and @xmath314 therefore , the expression in brackets in ( [ case2 ] ) can be estimated as follows : @xmath315 where the last inequality follows from the fact that @xmath316 for @xmath310 .
therefore , @xmath295 for @xmath312 . combining it with estimates from appendix
a in @xcite , we have @xmath317 .
using lemmas [ lemma - periodic ] , [ lemma - constant ] , and [ lemma - alternating ] , we classify the continuation of solutions of the first - order reductions , which are described in the families ( i)(iv ) of proposition [ lemma - asymmetric ] . for family ( i ) , excluding the case of super - symmetric solitons ( see remark [ remark - supersymmetric ] )
, the numbers of positive and negative signs of @xmath318 are different , such that the conditions @xmath319 and @xmath267 are satisfied in lemma [ lemma - periodic ] , and hence @xmath268 . by proposition
[ proposition - sufficient ] , the family ( i ) has a unique continuation to discrete solitons ( see definition [ definition - soliton ] ) .
continuations described in remark [ remark - complex ] are only possible for super - symmetric solitons . for family ( ii ) ,
all coefficients @xmath318 are the same : @xmath320 , @xmath179 . by lemma [ lemma - constant ]
, there is always a zero eigenvalue ( @xmath321 ) , while remaining ( @xmath322 ) eigenvalues are all positive for @xmath323 ( when @xmath324 ) , negative for @xmath325 ( when @xmath326 ) , and zero for @xmath327 ( when @xmath195 ) . by proposition
[ proposition - sufficient ] , the family ( ii ) has a unique continuation to symmetric vortices with charge @xmath32 , where @xmath328 and @xmath201 ( see definitions [ definition - vortex ] and [ definition - symmetric ] ) . for family ( iii ) , there are @xmath198 coefficients @xmath208 and @xmath198 coefficients @xmath329 , which are non - zero for @xmath330 . by lemma [ lemma - alternating ] , we have @xmath331 , @xmath295 , and @xmath332 .
the additional zero eigenvalue is related to the derivative of the family of the asymmetric discrete vortices ( [ asymmetric - vortices1 ] ) with respect to the parameter @xmath213 .
therefore , continuations of family ( iii ) of asymmetric vortices , including the particular cases of super - symmetric solitons of family ( i ) and super - symmetric vortices of family ( ii ) , must be considered beyond the first - order reductions . for family ( iv ) , since @xmath333 , the conditions @xmath265 and @xmath267 are satisfied in lemma [ lemma - periodic ] , and hence @xmath268 . by proposition [ proposition - sufficient ] , the family ( iv ) has a unique continuation to asymmetric vortices for @xmath154 .
results of the first - order reductions are insufficient to conclude persistence of the asymmetric vortices of family ( iii ) , including the super - symmetric soliton of family ( i ) and the super - symmetric vortex of family ( ii ) .
therefore , we continue the bifurcation function @xmath86 to the second order of @xmath130 in the taylor series ( [ taylor ] ) .
it follows from ( [ 2difference ] ) that the first - order correction of the taylor series ( [ taylorphi ] ) satisfies the inhomogeneous problem : @xmath334 we define solution of the inhomogeneous problem ( [ 2problem ] ) in @xmath335 , such that the homogeneous solutions in @xmath99 are removed from the solution @xmath336 .
this is equivalent to the constraint : @xmath337 , @xmath338 for all @xmath339 .
we develop computations for three distinct cases : @xmath33 , @xmath177 and @xmath340 .
this separation is due to the special structure of the discrete contours @xmath30 .
* case @xmath33 : * the inhomogeneous problem ( [ 2problem ] ) has a unique solution @xmath341 : @xmath342 e^{i \theta_j},\end{aligned}\ ] ] where the index @xmath343 enumerates the node @xmath110 on the contour @xmath30 , @xmath344 where the node @xmath110 is adjacent to the @xmath343-th node on the contour @xmath30 , while @xmath345 is empty for all remaining nodes . by substituting the first - order correction term @xmath345 into the bifurcation function ( [ g - function ] ) ,
we find the correction term @xmath156 in the taylor series ( [ taylor ] ) : @xmath346 \\ \label{sin - bifurcation2 } & + & \frac{1}{2 } \sin(\theta_{j-1 } - \theta_{j } ) \left [ \cos(\theta_j - \theta_{j-1 } ) + \cos(\theta_{j-2 } -\theta_{j-1 } ) \right ] , \qquad 1 \leq j \leq n.\end{aligned}\ ] ] we compute the vector @xmath347 at the asymmetric vortex solutions ( [ asyma])([asymc ] ) : @xmath348 the kernel of @xmath133 is two - dimensional with the eigenvectors @xmath149 and @xmath349 .
the second eigenvector @xmath349 is related to derivatives of the solutions ( [ asyma])([asymc ] ) in @xmath213 : @xmath350 the fredholm alternative @xmath351 is satisfied for the solution ( a ) but fails for the solutions ( b ) and ( c ) , unless @xmath352 . the latter cases are included in the definitions of super - symmetric discrete solitons and vortices ( see remark [ remark - supersymmetric ] ) . by proposition [ lemma - nonexistence ] , the solutions ( b ) and ( c ) can not be continued in @xmath154 , while the solution ( a ) can be continued up to the second - order reductions .
* case @xmath177 : * the solution @xmath353 of the inhomogeneous problem ( [ 2problem ] ) is given by ( [ solution31 ] ) and ( [ solution32 ] ) , except for the center node @xmath354 , where @xmath355 the correction term @xmath156 is given by ( [ sin - bifurcation2 ] ) but the even entries are modified as follows : @xmath356 the vector @xmath357 can be computed for each of 35 one - parameter asymmetric vortex solutions , starting with the first three solutions : @xmath358 the second eigenvector @xmath349 of the kernel of @xmath133 is related to derivatives of the family in @xmath213 , e.g. @xmath359 the fredholm alternative condition @xmath351 fails for all solutions of family ( iii ) but one , excluding the special values @xmath352 .
the only solution of family ( iii ) , where @xmath360 , is characterized by the alternating signs of coefficients @xmath259 for @xmath179 .
* case @xmath340 : * the solution @xmath361 of the inhomogeneous problem ( [ 2problem ] ) is given by ( [ solution31 ] ) and ( [ solution32 ] ) , except for the four interior corner nodes @xmath354,@xmath362,@xmath363 , and @xmath364 , where @xmath365 the correction term @xmath156 is given by ( [ sin - bifurcation2 ] ) , except for the adjacent entries to the four corner nodes on the contour @xmath30 : @xmath366 , @xmath367 , @xmath368 , and @xmath369 ) , which are modified by : @xmath370 again , there is only one solution of family ( iii ) , where @xmath371 , which is characterized by the alternating signs of coefficients @xmath372 for @xmath373 .
all other solutions of family ( iii ) do not satisfy the fredholm alternative condition @xmath351 . by using results of these computations , we classify continuations of solutions of the super - symmetric solitons of family ( i ) and asymmetric vortices of family ( iii ) .
let @xmath374 be the jacobian matrix computed from the bifurcation function @xmath375 , given in ( [ sin - bifurcation2 ] ) , ( [ sin - bifurcation2a ] ) and ( [ sin - bifurcation2b ] ) .
since @xmath376 for @xmath377 , except for the case of sign - alternating set @xmath185 with @xmath292 , it follows from regular perturbation theory that @xmath378 .
therefore , the second zero eigenvalue of @xmath133 bifurcates off zero for the matrix @xmath379 . by proposition [ proposition - sufficient ] ( which needs to be modified for the jacobian matrix @xmath379 ) ,
the super - symmetric solutions of family ( i ) , which are different from sign - alternating sets @xmath292 , are uniquely continued to discrete solitons ( see definition [ definition - soliton ] ) . by proposition
[ lemma - nonexistence ] , all asymmetric vortices of family ( iii ) , except for the sign - alternating set @xmath380 , @xmath381 , can not be continued to @xmath154 .
the only solution which can be continued up to the second - order reductions has the explicit form : @xmath382 where @xmath383 and @xmath169 .
this solution includes two particular cases of super - symmetric solitons of family ( i ) for @xmath224 and @xmath225 and super - symmetric vortices of family ( ii ) for @xmath227 .
continuation of the solution ( [ asymmetric - vortices ] ) must be considered beyond the second - order reductions .
the jacobian matrix @xmath133 of the first - order reductions is empty for super - symmetric vortices of family ( ii ) with @xmath195 . in order to study stability of super - symmetric vortices
, we need to compute the jacobian matrix @xmath374 from the second - order bifurcation function @xmath156 , given in ( [ sin - bifurcation2 ] ) , ( [ sin - bifurcation2a ] ) , and ( [ sin - bifurcation2b ] ) .
these computations are developed separately for three cases @xmath33 , @xmath177 , and @xmath340 . * case @xmath33 : * non - zero elements of @xmath374 are given by : @xmath384 or explicitly for @xmath233 : @xmath385 the matrix @xmath374 has four eigenvalues : @xmath386 and @xmath387 .
the two eigenvectors for the zero eigenvalue are @xmath388 and @xmath389 .
the eigenvector @xmath390 corresponds to the derivative of the asymmetric vortex ( [ asyma ] ) with respect to parameter @xmath213 , while the eigenvector @xmath391 corresponds to the shift due to gauge transformation .
* case @xmath177 : * the jacobian matrix @xmath374 is given in ( [ melements4 ] ) except for the even entries which are modified as follows : @xmath392 the matrix @xmath374 for @xmath238 takes the explicit form : @xmath393 the eigenvalue problem for @xmath374 decouples into two linear difference equations with constant coefficients : @xmath394 subject to the periodic boundary conditions for @xmath395 and @xmath396 . by the discrete fourier transform , see the proof of lemma [ lemma - alternating ] , the first problem has eigenvalues : @xmath397 , @xmath398 , @xmath399 , and @xmath400 , while the second problem has eigenvalues : @xmath401 , @xmath402 , @xmath403 , and @xmath404 .
the two eigenvectors for the zero eigenvalue are @xmath405 and @xmath406 , where the eigenvector @xmath407 corresponds to the derivative of the asymmetric vortex ( [ asymmetric - vortices ] ) with respect to parameter @xmath213 and the eigenvector @xmath408 corresponds to the shift due to gauge transformation .
* case @xmath340 : * the jacobian matrix @xmath374 is given in ( [ melements4 ] ) , except for the adjacent entries to the four corner nodes on the contours @xmath30 : @xmath366 , @xmath367 , @xmath368 , and @xmath369 ) , which are modified by @xmath409 the matrix @xmath374 for @xmath410 takes the explicit form : @xmath411 the eigenvalue problem for @xmath374 decouples into eigenvalue problems for two @xmath412-by-@xmath412 matrix , which are related by the toeplitz transformation . as a result ,
the spectra of these two matrices are identical with the eigenvalues , obtained with the use of matlab : @xmath413 we confirm that the matrix @xmath374 has exactly two zero eigenvalues , one of which is related to the derivative of the asymmetric vortex ( [ asymmetric - vortices ] ) in @xmath213 and the other one is related to the shift due to gauge transformation .
computations of the matrix @xmath374 for super - symmetric vortices of family ( ii ) confirm the results of the second - order reductions for asymmetric vortices of family ( iii ) .
although all @xmath230 solutions of family ( iii ) reduce to the super - symmetric vortex of family ( ii ) in the first - order reductions , it is the only family ( [ asymmetric - vortices ] ) that survives in the second - order reductions , such that the super - symmetric vortex of family ( ii ) with @xmath195 and @xmath227 can be deformed and continued up to the second - order reductions to the asymmetric vortex ( [ asymmetric - vortices ] ) .
all individual results on persistence of localized modes on the discrete contour @xmath30 are summarized as follows .
[ proposition - persistence ] consider the discrete soliton and vortices of the nonlinear equations ( [ 2difference ] ) that bifurcate from the limiting solution @xmath61 in ( [ 2soliton ] ) on the discrete contour @xmath30 in ( [ family-3 ] ) .
there exists a unique ( modulo gauge transformation ) continuation to the domain @xmath58 of discrete solitons of family ( i ) in ( [ discretesoliton ] ) , except for the case @xmath226 and @xmath292 , of symmetric vortices of family ( ii ) in ( [ symmetric - vortices ] ) , except for the case @xmath195 , and of zero - parameter asymmetric vortices of family ( iv ) in ( [ asymmetric - vortices2 ] ) .
asymmetric vortices of family ( iii ) in ( [ asymmetric - vortices1 ] ) can not be continued to the domain @xmath58 , except for the only solution ( [ asymmetric - vortices ] ) .
[ hypothesis - persistence ] there exists a one - parameter ( modulo gauge transformation ) continuation to the domain @xmath140 of the family of asymmetric vortices ( [ asymmetric - vortices ] ) with @xmath169 , which includes the one - parameter continuation of the two exceptions of proposition [ proposition - persistence ] .
we will not be proving hypothesis [ hypothesis - persistence ] beyond the second - order reductions .
instead , we use proposition [ proposition - persistence ] to study stability of persistent localized modes of the discrete nls equation ( [ 2difference ] ) .
the spectral stability of discrete vortices is studied with the standard linearization : @xmath414 where @xmath415 and @xmath416 solve the linear eigenvalue problem on @xmath2 : @xmath417 the stability problem ( [ eigenvalue ] ) can be formulated in the matrix - vector form : @xmath418 where @xmath419 consists of 2-blocks of @xmath420 , @xmath94 is defined by the linearization operator ( [ energy ] ) , and @xmath421 consists of @xmath422-by-@xmath422 blocks of @xmath423 the discrete vortex is called spectrally unstable if there exists @xmath424 and @xmath419 in the problem ( [ eigenvalue ] ) , such that @xmath425 .
otherwise , the discrete vortex is called weakly spectrally stable . by using the taylor series ( [ taylorphi ] ) ,
the linearized operator @xmath94 is expanded as follows : @xmath426 where @xmath427 is defined in ( [ energy-01 ] ) , while the first - order and second - order corrections are given by @xmath428 and @xmath429 it is clear from the explicit form ( [ energy-01 ] ) that the spectrum of @xmath430 has exactly @xmath104 negative eigenvalues @xmath431
, @xmath104 zero eigenvalues @xmath432 and infinitely many positive eigenvalues @xmath433 . the negative and zero eigenvalues @xmath431 and @xmath432 map to @xmath104 double zero eigenvalues @xmath434 in the eigenvalue problem @xmath435 .
the positive eigenvalues @xmath433 map to the infinitely many eigenvalues @xmath436 . since zero eigenvalues of @xmath437
are isolated from the rest of the spectrum of @xmath437 , their splitting can be studied through regular perturbation theory @xcite . on the other hand
, if the discrete vortex solutions @xmath13 for @xmath2 decays sufficiently fast as @xmath438 are located on the imaginary axis of @xmath424 near the points @xmath436 , similarly to the case @xmath439 for @xmath440 @xcite .
therefore , the infinite - dimensional part of the spectrum does not produce any unstable eigenvalues @xmath441 in the stability problem ( [ eigenvalue ] ) with small @xmath58 .
we shall consider how zero eigenvalues of @xmath427 and @xmath442 split as @xmath154 for solutions of the nonlinear equations ( [ 2difference ] ) , which are categorized by propositions [ lemma - asymmetric ] and [ proposition - persistence ] . the splitting of zero eigenvalues of @xmath94 is related to the lyapunov
schmidt reductions of the nonlinear equations ( [ nonlinear - equations ] ) .
we show that the same matrix @xmath264 , which gives the jacobian of the bifurcation functions @xmath134 , defines also small eigenvalues of @xmath94 that bifurcate from zero eigenvalues of @xmath105 in the first - order reductions .
[ lemma - splitting ] let the jacobian matrix @xmath133 have eigenvalues @xmath443 , @xmath179 .
the eigenvalue problem @xmath444 has @xmath104 eigenvalues @xmath445 in @xmath140 , such that @xmath446 we assume that there exists an analytical solution @xmath82 of the nonlinear equations ( [ nonlinear - equations ] ) .
the taylor series of @xmath82 is defined by ( [ taylorphi ] ) . by taking the derivative in @xmath130 ,
we rewrite the problem ( [ nonlinear - equations ] ) in the form : @xmath447 where the linearization operator ( [ energy ] ) is represented as @xmath448 . using the series ( [ taylorphi ] ) and ( [ series - h ] ) ,
we have the linear inhomogeneous equation : @xmath449 let @xmath450 , @xmath451 be eigenvectors of the kernel of @xmath427 .
each eigenvector @xmath450 contains the only non - zero block @xmath452 at the @xmath343-th position , which corresponds to the node @xmath110 on the contour @xmath30 .
it is clear that the eigenvectors are orthogonal as follows : @xmath453 let @xmath454 , @xmath451 be generalized eigenvectors , such that each eigenvector @xmath455 contains the only non - zero block @xmath456 at the @xmath343-th position .
direct computations show that @xmath457 the limiting solution ( [ 2soliton ] ) can be represented as follows : @xmath458 by comparing the inhomogeneous equation ( [ inhomog - equation-1 ] ) with the definition ( [ g - function ] ) of the bifurcation function @xmath459 and its taylor series ( [ taylor ] ) , we have the correspondence : @xmath460 consider a perturbation to a fixed point of @xmath461 in the form @xmath462 where @xmath463 .
it is clear that @xmath464 where @xmath465 , @xmath466 , and @xmath467 . by expanding the bifurcation function @xmath254 near @xmath468
, we define the jacobian matrix @xmath133 : @xmath469 where @xmath470 on the other hand , the regular perturbation series for small eigenvalues of the problem @xmath471 are defined as follows : @xmath472 where @xmath473 , according to the kernel of @xmath427 .
the first - order correction term @xmath474 satisfies the inhomogeneous equation : @xmath475 projection to the kernel of @xmath427 gives the eigenvalue problem for @xmath476 : @xmath477 by direct computations : @xmath478 such that the limiting relation ( [ limiting-1 ] ) follows from ( [ correction - term1 ] ) and ( [ eigenvalue - problem1 ] ) with @xmath479 .
we apply results of lemma [ lemma - splitting ] to the solutions of the first - order reductions , which are described in families ( i)(iv ) of propositions [ lemma - asymmetric ] and [ proposition - persistence ] .
the numbers of negative , zero and positive eigenvalues of @xmath133 are denoted as @xmath261 , @xmath262 and @xmath263 respectively .
these numbers determine the numbers of small negative and positive eigenvalues of @xmath94 for small positive @xmath130 .
they are predicted from lemmas [ lemma - periodic ] , [ lemma - constant ] and [ lemma - alternating ] . for family ( i )
, we compute the parameter @xmath480 in lemma [ lemma - periodic ] as @xmath481 , where @xmath186 is defined in proposition [ lemma - asymmetric ] . by lemma [ lemma - periodic ] , we have @xmath482 , @xmath483 , and @xmath484 for @xmath485 and @xmath486 , @xmath268 , and @xmath487 for @xmath488 . in the case of super - symmetric solitons for @xmath226 , by lemma
[ lemma - alternating ] , we have @xmath294 , @xmath295 , and @xmath305 . for family ( ii ) , by lemma [ lemma - constant ]
, we have @xmath489 , @xmath268 , and @xmath490 for @xmath491 and @xmath492 , @xmath268 , and @xmath493 for @xmath494 , where @xmath32 is defined in proposition [ lemma - asymmetric ] . the case of super - symmetric vortices
@xmath495 can only be studied in the second - order reductions , since @xmath496 .
the family ( iii ) is represented by the only one - parameter solution ( [ asymmetric - vortices ] ) for each @xmath195 . by lemma [ lemma - alternating ] , we have @xmath294 , @xmath497 , and @xmath498 for @xmath169 , excluding the case of super - symmetric vortices @xmath499 but including the case of super - symmetric solitons @xmath224 and @xmath225 . the family ( iv ) is characterized by the value of @xmath500 , @xmath201 , and @xmath501 , @xmath333 , specified in proposition [ lemma - asymmetric ] .
the parameter @xmath480 in lemma [ lemma - periodic ] is computed as @xmath502 , such that @xmath268 in all cases . in the case
@xmath503 , by lemma [ lemma - periodic ] , we have @xmath504 and @xmath505 for @xmath506 and @xmath507 and @xmath508 for @xmath509 . in the opposite case of @xmath510
, we have @xmath511 and @xmath512 for @xmath513 and @xmath514 and @xmath515 for @xmath509 .
the splitting of zero eigenvalue of @xmath94 is related to splitting of zero eigenvalues of @xmath516 in the stability problem ( [ eigenvalue ] ) .
[ lemma - splitting000 ] let the jacobian matrix @xmath133 have eigenvalues @xmath443 , @xmath179 .
the eigenvalue problem @xmath517 has @xmath104 pairs of eigenvalues @xmath518 in @xmath79 , such that @xmath519 the regular perturbation series for small eigenvalues of @xmath520 are defined as follows : @xmath521 where , due to the relations ( [ orthogonality - relations-1 ] ) and ( [ orthogonality - relations-2 ] ) , we have : @xmath522 according to the kernel and generalized kernel of @xmath523 .
the second - order correction term @xmath524 satisfies the inhomogeneous equation : @xmath525 projection to the kernel of @xmath427 gives the eigenvalue problem for @xmath526 : @xmath527 where @xmath528 and the matrix @xmath133 is the same as in the eigenvalue problem ( [ eigenvalue - problem1 ] ) . as a result , the relation ( [ limiting-000 ] )
is proved .
the numbers of negative , zero and positive eigenvalues of @xmath264 , denoted as @xmath261 , @xmath262 and @xmath529 , are computed above .
let @xmath530 , @xmath531 , and @xmath532 be the numbers of pairs of real , zero and imaginary eigenvalues of the reduced eigenvalue problem ( [ lyapunovreducedeigenvalue1 ] ) .
we consider stability of families ( i)(iv ) , described in propositions [ lemma - asymmetric ] and [ proposition - persistence ] . for family ( i )
, we have @xmath533 , @xmath534 , and @xmath535 for @xmath485 ; @xmath533 , @xmath536 , and @xmath537 for @xmath226 ; and @xmath538 , @xmath534 , and @xmath537 for @xmath539 , where @xmath186 is defined in proposition [ lemma - asymmetric ] . for family ( ii )
, we have @xmath540 , @xmath534 , and @xmath541 for @xmath324 ; @xmath540 , @xmath542 , and @xmath543 for @xmath544 ; and @xmath545 , @xmath534 , and @xmath546 for @xmath547 , where @xmath32 is defined in proposition [ lemma - asymmetric ] . for the only one - parameter solution ( [ asymmetric - vortices ] ) of family ( iii ) for each @xmath195
, we have @xmath548 , @xmath536 , and @xmath549 for @xmath169 and @xmath550 . for family ( iv ) with @xmath503
, we have @xmath551 , @xmath534 , and @xmath552 for @xmath553 and @xmath554 , @xmath534 , and @xmath555 for @xmath509 . in the opposite case of @xmath510
, we have @xmath556 , @xmath534 , and @xmath557 for @xmath553 and @xmath558 , @xmath559 , and @xmath560 for @xmath509 .
there are several features which are not captured in the first - order reductions . for super - symmetric solitons of family ( i ) , when @xmath561 but @xmath562 , the additional zero eigenvalue splits at the second - order reductions , which leads to an additional non - zero eigenvalues of the stability problem ( [ eigenvalue ] ) . for super - symmetric vortices of family ( ii ) ,
when @xmath195 , the matrix @xmath496 , such that non - zero eigenvalues occur only in the second - order reductions .
finally , for symmetric vortices of family ( ii ) , multiple real non - zero eigenvalues of the first - order reductions , according to the roots of @xmath563 in the explicit solution ( [ explicit - eigenvalue ] ) , split into the complex domain in the second - order reductions .
these questions are studied next in the reverse order .
we continue the regular perturbation series ( [ lyapunoveigenvector1 ] ) to the second - order reductions . by using the explicit first - order correction term ( [ solution31 ] ) ,
we compute the explicit solution : @xmath564 where the vectors @xmath565 are obtained from @xmath566 by shifts of non - zero elements of @xmath566 from the node @xmath567 to the adjacent nodes @xmath568 .
the third - order correction term @xmath569 satisfies the inhomogeneous equation : @xmath570 projection to the kernel of @xmath427 gives the extended eigenvalue problem for @xmath526 and @xmath571 : @xmath572 where the matrix @xmath573 is defined by @xmath574 subject to the periodic boundary conditions .
let @xmath445 be an eigenvalue of the symmetric matrix @xmath133 with the linearly independent eigenvector @xmath575 .
then , @xmath576 since the matrix @xmath573 is skew - symmetric , the second - order correction term @xmath571 is purely imaginary , unless @xmath577 . for discrete solitons of family ( i ) , we have @xmath578 , such that @xmath579 and @xmath580 . for symmetric vortices of family ( ii ) with @xmath201 ,
the matrix @xmath133 has double eigenvalues , according to the roots of @xmath563 in the explicit solution ( [ explicit - eigenvalue ] ) .
using the same discrete fourier transform as in the proof of lemma [ lemma - constant ] , one can find the values of @xmath526 and @xmath571 in this case .
[ lemma - constanta ] let all coefficients @xmath291 and @xmath581 , @xmath179 be the same : @xmath286 and @xmath582 .
eigenvalues of the reduced problem ( [ lyapunovreducedeigenvalue1a ] ) are given explicitly : @xmath583 according to lemma [ lemma - constanta ] , all double roots of @xmath526 for @xmath584 and @xmath585 split along the imaginary axis in @xmath571 .
when @xmath586 , the splitting occurs in the transverse directions to the real values of @xmath526 .
when @xmath587 , the splitting occurs in the longitudinal directions to the imaginary values of @xmath526 .
the simple roots at @xmath588 and @xmath589 are not affected , since @xmath580 for @xmath590 and @xmath589 . for asymmetric vortices of family ( iii )
, it follows from the explicit solution ( [ explicit - eigenvalue - alternating ] ) that @xmath580 for all roots , so that the splitting of eigenvalues does not occur in the second - order reductions . for asymmetric vortices of family ( iv ) ,
the value of @xmath571 can not be computed in general .
we extend results of the regular perturbation series ( [ regular - series1 ] ) and ( [ lyapunoveigenvector1 ] ) to the case @xmath496 , which occurs for super - symmetric vortices of family ( ii ) with charge @xmath195 .
it follows from the problem ( [ inhomogeneous - equation1 ] ) with @xmath496 that @xmath591 and the first - order correction term @xmath592 has the explicit form : @xmath593 where the vectors
@xmath565 are the same as in the formula ( [ inhomog - solution-1a ] ) .
the second - order correction term @xmath594 satisfies the inhomogeneous equation : @xmath595 projection to the kernel of @xmath427 gives the eigenvalue problem for @xmath596 : @xmath597 by direct computations in the three separate cases , one can show that the matrix on the left - hand - side of the eigenvalue problem ( [ eigenvalue - problem2 ] ) is nothing but the matrix @xmath374 , which is the jacobian of the nonlinear function @xmath375 , computed for the super - symmetric vortex of family ( ii ) .
let the number of negative , zero and positive eigenvalues of @xmath374 be denoted as @xmath598 , @xmath599 and @xmath600 respectively . for super - symmetric vortices of family ( ii ) , we have @xmath601 , @xmath602 and @xmath603 for @xmath33 ; @xmath604 , @xmath602 and @xmath605 for @xmath177 ; and @xmath606 , @xmath602 and @xmath607 for @xmath608 .
splitting of zero eigenvalues of @xmath516 is studied with the regular perturbation series ( [ lyapunoveigenvector1 ] ) .
when @xmath496 , it follows from the problem ( [ lyapunovlinearsystem1 ] ) that @xmath609 , such that the regular perturbation series ( [ lyapunoveigenvector1 ] ) can be re - ordered as follows : @xmath610 where @xmath611 and @xmath474 is given by ( [ correction - term2 ] ) .
the second - order correction term @xmath612 is found from the inhomogeneous equation : @xmath613 projection to the kernel of @xmath427 gives the eigenvalue problem for @xmath526 : @xmath614 where @xmath528 , the matrix @xmath374 is the same as in the eigenvalue problem ( [ eigenvalue - problem2 ] ) , and the matrix @xmath615 follows from the matrix @xmath573 in the form ( [ lelementsa ] ) with @xmath616 , or explicitly : @xmath617 subject to the periodic boundary conditions .
since @xmath374 is symmetric and @xmath615 is skew - symmetric , the eigenvalues of the problem ( [ lyapunovreducedeigenvalue2 ] ) occur in pairs @xmath618
. computations of eigenvalues of the reduced eigenvalue problem ( [ lyapunovreducedeigenvalue2 ] ) are reported in the three distinct cases : @xmath33 , @xmath177 and @xmath340 .
* case @xmath33 : * the reduced eigenvalue problem ( [ lyapunovreducedeigenvalue2 ] ) takes the form of the difference equation with constant coefficients : @xmath619 subject to the periodic boundary conditions . by the discrete fourier transform , the difference equation reduces to the characteristic equation : @xmath620 the reduced eigenvalue problem ( [ lyapunovreducedeigenvalue2 ] ) has two eigenvalues of algebraic multiplicity two at @xmath621 and @xmath622 and zero eigenvalue of algebraic multiplicity four .
* case @xmath177 : * the reduced eigenvalue problem ( [ lyapunovreducedeigenvalue2 ] ) takes the form of the system of difference equations with constant coefficients : @xmath623 where @xmath624 and @xmath625 subject to the periodic boundary conditions . the characteristic equation for the linear system takes the explicit form : @xmath626 the reduced eigenvalue problem ( [ lyapunovreducedeigenvalue2 ] ) has three eigenvalues of algebraic multiplicity four at @xmath627 , @xmath628 , and @xmath609 , and four simple eigenvalues at @xmath629 and @xmath630 .
* case @xmath340 : * eigenvalues of the reduced eigenvalue problem ( [ lyapunovreducedeigenvalue2 ] ) with @xmath608 are computed numerically by using mathematica .
the results are as follows : @xmath631 using these computations of eigenvalues , we summarize that the second - order reduced eigenvalue problem ( [ lyapunovreducedeigenvalue2 ] ) has no unstable eigenvalues @xmath424 when @xmath632 ; a simple unstable ( positive ) eigenvalue when @xmath633 ; two unstable real and two unstable complex eigenvalues when @xmath634 .
we note that destabilization of the super - symmetric vortex with @xmath177 occurs due to the center node @xmath354 , which couples the four even - numbered nodes of the contour @xmath635 in the second - order reductions . due to this coupling
, there exists a simple negative eigenvalue of the jacobian matrix @xmath636 and a simple positive eigenvalue in the reduced eigenvalue problem ( [ lyapunovreducedeigenvalue2 ] ) . similarly ,
destabilization of the super - symmetric vortex with @xmath608 occurs due to the coupling of eight nodes of the contour @xmath637 with four interior corner points @xmath354 , @xmath364 , @xmath363 , and @xmath362 , which result in the four negative eigenvalues of the matrix @xmath374 and four unstable eigenvalues in the reduced eigenvalue problem ( [ lyapunovreducedeigenvalue2 ] ) .
we note that if the matrix @xmath374 would be defined by the formula ( [ melements4 ] ) for any @xmath638 ( i.e. if all nodes inside the contour @xmath30 would be removed by drilling a hole ) , the eigenvalues of @xmath374 would be all positive and the eigenvalues of the reduced problem ( [ lyapunovreducedeigenvalue2 ] ) would be all purely imaginary , similarly to the case @xmath33 . for super - symmetric solitons of family ( i ) , when @xmath226 but @xmath639 , the jacobian matrix @xmath133 has two zero eigenvalues with eigenvectors @xmath149 and @xmath349 , but the jacobian matrix @xmath640 has only one zero eigenvalue with eigenvector @xmath149 .
therefore , the splitting of zero eigenvalue occurs in the second - order reduction . extending the regular perturbation series ( [ regular - series1 ] ) to the next order
, we find that @xmath641 for @xmath642 , and @xmath643 extending the regular perturbation series ( [ lyapunoveigenvector1 ] ) to the second order , we have find that @xmath644 for @xmath642 , @xmath579 , and @xmath645 thus , the splitting of zero eigenvalue in the second - order reduction is the same as the splitting of zero eigenvalues in the first - order reductions . a positive eigenvalue @xmath596 results in a pair of real eigenvalues @xmath571 , while a negative eigenvalue @xmath596 results in a pair of purely imaginary eigenvalues @xmath571 .
all individual results on stability of localized modes on the discrete contour @xmath30 are summarized as follows .
[ proposition - stability ] consider the stability problem ( [ eigenvalue ] ) in the domain @xmath58 , associated to the families of discrete solitons and vortices of propositions [ lemma - asymmetric ] and [ proposition - persistence ] .
the following solutions are spectrally stable in the domain @xmath58 : discrete solitons of family ( i ) with @xmath646 ; symmetric vortices of family ( ii ) with the charge @xmath647 ; and symmetric vortex of family ( ii ) with the charge @xmath632 .
all other solutions have at least one unstable eigenvalue with @xmath425 .
we note that stability of discrete solitons of family ( i ) with @xmath648 is equivalent to stability of discrete solitons in the one - dimensional nls lattice , which is proved in the first paper @xcite .
when @xmath646 , the limiting solution ( [ 2soliton ] ) consists of alternating up and down pulses along the contour @xmath30 , similar to theorem 3.6 in @xcite .
we also note that the purely imaginary eigenvalue of @xmath424 have negative krein signature , such that the hamiltonian hopf bifurcations may occur for larger values of @xmath130 beyond the neighborhood @xmath45 .
the number of eigenvalues of negative krein signature is related to the closure relation for the number of negative eigenvalues of the linearized hamiltonian @xmath94 ( see @xcite ) .
we can see from computations of families ( i)-(iv ) that the closure relation is satisfied in all cases , except for the super - symmetric vortices of family ( ii ) with @xmath195 .
indeed , in the case @xmath632 , the hamiltonian @xmath94 has four negative eigenvalues for @xmath58 , while it has two pairs of purely imaginary eigenvalues of negative krein signature and two pairs of zero eigenvalues , which exceeds the allowed negative index of @xmath94 .
derivation of a modified closure relation for vortex solutions of the discrete nls equations is beyond the scope of this manuscript .
we perform direct numerical simulations of the discrete nls equation ( [ 2nls ] ) in order to examine stability of the discrete vortices in the simplest cases @xmath33 and @xmath177 .
the results are shown on figures [ depf1]-[depf6 ] ( see also summary in table 1 ) . in computations of solutions of the problems ( [ 2difference ] ) and ( [ eigenvalue ] ) , we will use an equivalent renormalization of the problem with parameter : @xmath649 this renormalization is equivalent to keeping the diagonal term @xmath650 in the right - hand - side of the difference equations ( [ 2difference ] ) . as a result , the discrete solitons and vortices exist typically in the semi - infinite domain @xmath651 .
the anti - continuous limit is not affected by the renormalization since @xmath652 for small @xmath130 . on the figures [ depf1][depf6 ] ,
the top left panel shows the profile of the vortex solution for a specific value of @xmath653 by means of contour plots of the real ( top left ) , imaginary ( top right ) modulus ( bottom left ) and phase ( bottom right ) two - dimensional profiles .
the top right panel shows the complex plane @xmath654 for the linear eigenvalue problem ( [ eigenvalue ] ) for the same value of @xmath653 .
the bottom panel shows the dependence of small eigenvalues as a function of @xmath653 , obtained via continuation methods from the anti - continuum limit of @xmath655 .
the solid lines represent numerical results , while the dashed lines show results of the first - order and second - order reductions .
figure [ depf1 ] show results for the super - symmetric vortex of charge @xmath243 on the contour @xmath30 with @xmath33 . in the second - order reduction ,
the stability spectrum of the vortex solution has a zero eigenvalue of algebraic multiplicity 4 and two pairs of imaginary eigenvalues @xmath656 .
these pairs split along the imaginary axis beyond the second - order reductions and undertake hamiltonian hopf bifurcations for larger values of @xmath653 upon collision with the continuous spectrum ( only the first bifurcation at @xmath657 is shown on fig .
[ depf1 ] ) .
it follows from fig .
[ depf1 ] that the zero eigenvalue splits along the imaginary axis for larger values of @xmath653 , beyond the second - order reductions .
figure [ depf2 ] show results for the symmetric vortex of charge @xmath658 on the contour @xmath30 with @xmath659 .
there are three double and one simple real unstable eigenvalues in the first - order reductions , but all double eigenvalues split into the complex plane in the second - order reductions .
the asymptotic result ( [ explicit - eigenvaluea ] ) for eigenvalues @xmath660 with @xmath661 , @xmath662 and @xmath663 are shown on fig .
[ depf2 ] in perfect agreement with numerical results .
figure [ depf3 ] shows results for the super - symmetric vortex with @xmath664 .
the second - order reductions have a pair of simple real eigenvalues @xmath665 , a pair of simple imaginary eigenvalues @xmath666 , zero eigenvalue of algebraic multiplicity @xmath249 and a pair of imaginary eigenvalues of algebraic multiplicity @xmath249 at @xmath667 .
the bottom right panel of fig . [ depf3 ] shows splitting of multiple imaginary eigenvalues beyond the second - order reductions and also subsequent hamiltonian
hopf bifurcations for larger values of @xmath653 .
figure [ depf4 ] shows results for the symmetric vortex with @xmath668 and @xmath659 .
the first - order reductions predict three pairs of double imaginary eigenvalues , a pair of simple imaginary eigenvalues and a double zero eigenvalue .
the double eigenvalues split in the second - order reductions along the imaginary axis , given by ( [ explicit - eigenvaluea ] ) with @xmath661 , @xmath669 and @xmath670 .
the seven pairs of imaginary eigenvalues lead to a cascade of seven hamiltonian hopf bifurcations for larger values of @xmath653 due to their collisions with the continuous spectrum .
the first hamiltonian
hopf bifurcation when the symmetric vortex becomes unstable occurs for @xmath671 .
zero - parameter asymmetric vortices of family ( iv ) on the contour @xmath30 with @xmath177 are shown in fig .
[ depf5 ] for @xmath243 and in fig .
[ depf6 ] for @xmath246 . in the case of fig .
[ depf5 ] , all the phase differences between adjacent sites in the contour are @xmath672 , except for the last one which is @xmath673 , completing a phase trip of @xmath26 for a vortex of topological charge @xmath658 .
eigenvalues of the matrix @xmath133 in the first - order reductions can be computed numerically as follows : @xmath674 , @xmath675 , @xmath676 , @xmath677 , @xmath678 , @xmath679 , @xmath680 and @xmath681 . as a result
, the corresponding eigenvalues @xmath682 yield one pair of imaginary eigenvalues and six pairs of real eigenvalues , in agreement with our numerical results .
the bottom panel of fig .
[ depf5 ] shows that two pairs of real eigenvalues collide for @xmath683 and @xmath684 and lead to two quartets of eigenvalues . in the case of fig .
[ depf6 ] , all the phase differences in the contour are @xmath673 , except for the last one which is @xmath672 , resulting in a vortex of topological charge @xmath668 .
eigenvalues of the matrix @xmath133 are found numerically as follows : @xmath685 , @xmath686 , @xmath687 , @xmath688 , @xmath689 , @xmath690 , @xmath691 , and @xmath692 .
consequently , this solution has six pairs of imaginary eigenvalues and one pair of real eigenvalues . the first hamiltonian
hopf bifurcation occurs for @xmath693 .
we note that numerical results for asymmetric vortices of family ( iii ) are not shown .
both analytical and numerical analysis of such solutions are delicate problems which are open for future studies .
however , all such solutions ( if they persist ) are unstable in the first - order reductions .
in this series of two papers , we have developed the mathematical analysis of discrete soliton and vortex solutions of the discrete nls equations close to the anti - continuum limit .
these solutions are relevant to recent experimental and numerical studies in the context of nonlinear optics , photonic crystal lattices , soft condensed - matter physics , and bose - einstein condensates . in the present paper
, we have examined the persistence of discrete vortices starting from the anti - continuum limit of uncoupled oscillators and continuing towards a finite coupling constant @xmath130 .
we have found persistent families of such solutions that include symmetric and asymmetric vortices .
we have ruled out other non - persistent solutions by means of the lyapunov - schmidt method .
we have subsequently categorized the persisting solutions to discrete solitons , symmetric vortices and one - parameter and zero - parameter asymmetric vortices . for persistent solutions
, we have derived the leading - order asymptotic approximations for small unstable and neutrally stable eigenvalues of the stability problem , up to the first - order and second - order corrections .
we have applied the results to particular computations of discrete vortices of topological charge @xmath658 , @xmath245 , and @xmath246 on the discrete square contours @xmath30 with @xmath694 and @xmath659 . besides particular computations collected in table 1 and figs .
[ depf1][depf6 ] , these results offer a road map on stability predictions for larger contours , as well as predictions on how to stabilize the discrete vortices . for example ,
super - symmetric vortices of charge @xmath695 can be stabilized by excluding the inner nodes inside the discrete contour @xmath30 .
there are remaining open problems for future analytical work .
first , it would be nice to extend this analysis to three - dimensional structures , such as the discrete solitons , vortices , or vortex `` cubes '' ( see @xcite ) .
second , other types of contours can be studied in the two - dimensional lattice , such as the diagonal square contours which would include the `` vortex cross '' or the octagon ( see @xcite ) .
persistence of one - parameter asymmetric vortices including the super - symmetric vortices has not yet been proved in the present paper and it leaves space for future work .
finally , the closure relation for negative indices of linearized hamiltonian associated with the discrete vortex solutions must be derived and applied independently of the current studies . while conceptually , the methodologies and techniques presented here can be adapted to the problems mentioned above , actual computations of the higher - order lyapunov
schmidt reductions become technically involved .
z. chen , h. martin , e.d .
eugenieva , j. xu and a. bezryadina , `` anisotropic enhancement of discrete diffraction and formation of two - dimensional discrete - soliton trains '' , phys .
* 92 * ( 2004 ) 143902 .
neshev , t.j .
alexander , e.a .
ostrovskaya , yu.s .
kivshar , h. martin , i. makasyuk and z. chen , `` observation of discrete vortex solitons in optically induced photonic lattices '' , phys .
* 92 * ( 2004 ) 123903 .
fleischer , g. bartal , o. cohen , o. manela , m. segev , j. hudock and d.n .
christodoulides , `` observation of vortex - ring `` discrete '' solitons in 2d photonic lattices '' , phys .
* 92 * ( 2004 ) 123904 .
kevrekidis , b.a .
malomed , d.j .
frantzeskakis and r. carretero - gonzlez , `` three - dimensional solitary waves and vortices in a discrete nonlinear schrodinger lattice '' , phys .
* 93 * ( 2004 ) 080403 . | we study discrete vortices in the anti - continuum limit of the discrete two - dimensional nonlinear schrdinger ( nls ) equations .
the discrete vortices in the anti - continuum limit represent a finite set of excited nodes on a closed discrete contour with a non - zero topological charge . using the lyapunov schmidt reductions , we find sufficient conditions for continuation and termination of the discrete vortices for a small coupling constant in the discrete nls lattice .
an example of a closed discrete contour is considered that includes the vortex cell ( also known as the off - site vortex ) .
we classify the symmetric and asymmetric discrete vortices that bifurcate from the anti - continuum limit .
we predict analytically and confirm numerically the number of unstable eigenvalues associated with various families of such discrete vortices . |
intermetallic alloys of iron and aluminum have a high hardness with a much lower specific weight than steel . because of this , their low costs , and resistance against corrosion and oxidation ,
feal alloys are often used as lightweight structural materials .
most puzzling are the magnetic properties . here , experiments such as high - field mssbauer investigations @xcite indicate no magnetism for stoichiometric feal which forms a @xmath2 cscl - type of lattice ( two interpenetrating fe and al simple cubic lattices ) . especially the fact that feal does not show ferromagnetism in experiment , while electronic structure calculations within spin - polarized density functional theory ( dft ) predict a ferromagnetic ground state has drawn attention to the material : independently of the bandstructure code , dft orbital basis set and exchange correlation potential a ferromagnetic ground state with a magnetic moment at the fe site of about @xmath3{\mu_b}$ ] is found . @xcite even though the energy difference between the ferromagnetic and the nonmagnetic state is rather small , the ferromagnetic state is stable over a wide volume range .
in fact , only a reduction of the lattice constant by more than 10% would suppress ferromagnetism .
@xcite this high stability of the ferromagnetic phase in feal suggests that the deviation from experiment is not just a numerical inaccuracy , but requires a deeper understanding .
different approaches have been used hitherto to explain the deviation between spin - polarized dft and experiment .
one explanation is based on the fact that the processes used to prepare feal often `` freeze in '' chemical disorder .
that is , `` real '' feal is usually not fully ordered due to various lattice defects , such as vacancies and antisites , which in turn could have a significant effect on the magnetic properties of the material . against this background ,
there exist several studies concerning the effects of disorder on the magnetic properties of feal .
@xcite for example , in ref .
the disorder is included via the coherent potential approximation ( cpa ) @xcite in the korringa , kohn and rostoker ( kkr ) framework , @xcite with the paramagnetic phase described by the disordered local moment approximation ( dlm ) .
@xcite in agreement with previous dft calculations , it has been found that ideal feal is ferromagnetic .
however , even with a small degree of disorder the paramagnetic state , without net magnetization but nonzero local moments , becomes the stable configuration .
thus , disorder destroys the long - range ferromagnetic order in dft .
however , no ferromagnetism has ever been observed for stoichiometric `` real '' feal , even for samples with very low defect concentration .
therefore , it still remains the question if perfectly ordered feal would really be ferromagnetic as predicted by dft .
indeed , mssbauer experiments @xcite find magnetic moments only for fe antistructure atoms ( which means fe atoms sitting on an al lattice site ) and their eight fe neighbors .
another possible explanation has been given in ref .
using the dft+@xmath4 approach .
@xcite usually , one would expect dft+@xmath4 to yield larger magnetic moments and a stronger tendency towards ferromagnetism than dft . for @xmath4 values ranging from @xmath5 to @xmath6{ev}$ ] a nonmagnetic state
however coexists with the ferromagnetic one in dft+@xmath4 .
the ferromagnetic state even disappears for a rather large @xmath7{ev}$ ] , which offers another explanation of the non - magnetic nature of feal .
this rather unusual dft+@xmath4 result can be explained by the changes in the density of states ( dos ) : increasing @xmath4 reduces the dos at the fermi level so that according to the stoner criterion there is no ferromagnetism @xcite even though the effective exchange is increased by @xmath4 .
@xcite hence , in a narrow range of @xmath4 , there is no ferromagnetism in dft+@xmath4 .
@xcite in ref .
it has been argued that this dft+@xmath4 result has to be taken with a grain of salt and it has been proposed for the first time that dynamical spin fluctuations suppress ferromagnetism in feal .
this has been supported by a dynamical mean field theory ( dmft@xcite ) calculation.@xcite for @xmath8{ev}$ ] , feal is found@xcite to be paramagnetic in dft+dmft.@xcite however , ref . only shows a single dft+dmft result , the spectral function .
the proposed spin fluctuations , the magnetic properties and susceptibility have not been calculated .
considering these limited results as well as the improvements of dft+dmft in recent years , a more thorough analysis is in order . beyond the first dft+dmft spectrum of ref . , we study the local and bulk magnetic susceptibility , the magnetic moment and the @xmath9-resolved spectrum .
we also explicitly calculate the local interactions _ ab - initio _ by constrained random phase approximation ( crpa ) and beyond ref .
we include the calculated hund s exchange in dmft with its full @xmath10 symmetry , since it plays a pivotal role for the magnetic properties .
our results show that while there is a local moment of even @xmath1{\mu_b}$ ] on short time scales , it is screened ( suppressed ) on longer time scales .
this suppression of the local moment occurs on the fs time scale ( ev@xmath11 ) and explains why there is eventually no long - range ferromagnetic order . in section [ sec : escdft ] we present the dft bandstructure and dos as well as the wannier function projection .
section [ sec : dmftself ] is devoted to the one - particle properties as calculated in dft+dmft , i.e. , the self - energy as well as the local and @xmath9-resolved spectral function .
the dft+dmft magnetic properties are discussed in section [ sec : dmftmag ] , i.e. , the local and ( zero ) ferromagnetic moment as well as the time - dependent local susceptibility and bulk susceptibility .
finally , section [ sec : conclusion ] summarizes the results and puts them into context with experiment .
as a first step , we employ the vienna ab initio simulation package ( vasp ) @xcite with gga - pbe functional @xcite for calculating the bandstructure and density of states of feal .
[ fig : bands_vasp ] shows the bandstructure of feal around the fermi level .
the bands closest to the fermi level have mainly fe 3@xmath12 character and are split into @xmath13 and @xmath14 due to the cubic crystal field .
for these bands we will later include electronic correlations by dmft .
however , since the fe 3@xmath12 bands strongly hybridize with the al 3@xmath15 and 3@xmath16 states , we also include these al bands ( as non - interacting ) in our low energy hamiltonian .
the corresponding hamiltonian is obtained by a projection onto nine maximally localized wannier orbitals @xcite , which reproduce the dft bandstructure well , see fig .
[ fig : bands_vasp ] .
( red ) , fe @xmath14 ( green ) and al-@xmath17 ( blue ) orbital contribution.,width=302 ] fig .
[ fig : dos_vasp ] shows the orbital resolved density of states .
it can be seen that the central peak is mainly of @xmath13 character . as the fe @xmath13
orbitals form only weak bonds with the nearest - neighbor al atoms , they have a rather small energy dispersion .
the fe @xmath14 states instead point towards the neighboring fe atoms and hybridize more strongly .
hence they have a larger bandwidth and split into a bonding- and an antibonding - like part .
after obtaining the low energy hamiltonian in the basis of wannier functions , we perform dmft calculations including the five fe @xmath12 orbitals and the 4 al @xmath17 orbitals within a so - called @xmath18 model.@xcite we supplement the dft - based wannier hamiltonian in dmft by a local @xmath12-@xmath12 kanamori interaction , but disregard @xmath12-@xmath16 and @xmath16-@xmath16 interactions beyond what is already contained in dft . note that the hopping terms of the hamiltonian still contain the full information about the hybridization with the al @xmath17 states and charge transfer between @xmath12 and @xmath17 orbitals is allowed .
we calculate the screened many - body coulomb interactions @xmath4 , @xmath19 and @xmath20 by the constrained random phase approximation ( crpa),@xcite where we exclude only the fe @xmath12 states from the screening .
this is appropriate as interactions are also applied only to these @xmath12 states.@xcite for our dmft calculation , we use the average values for the intra - orbital coulomb interaction @xmath21{ev}$ ] , the inter - orbital coulomb interaction @xmath22{ev}$ ] and the hund s coupling @xmath23{ev}$ ] .
this yields a local , @xmath10-symmetric kanamori interaction : @xcite @xmath24\nonumber\\ & + & \sum_{m\neq m'}jc^\dagger_{m\uparrow}c^\dagger_{m'\downarrow}c_{m\downarrow}c_{m'\uparrow}\nonumber\\ & + & \sum_{m\neq m'}jc^\dagger_{m\uparrow}c^\dagger_{m\downarrow}c_{m'\downarrow}c_{m'\uparrow } .
\end{aligned}\ ] ] here , @xmath25 ( @xmath26 ) creates ( annihilates ) an electron with spin @xmath27 in the fe @xmath28 orbital @xmath29 ; @xmath30 .
we employ the double counting correction of the fully localized limit , @xcite and validate that a difference of @xmath31{ev}$ ] in the double counting does not change our findings ( not shown ) . for the solution of the dmft impurity problem we use a continuous - time quantum monte carlo ( ct - qmc ) algorithm in its hybridization expansion ( ct - hyb ) in the version of ref .
, for a review see ref . .
especially with regard to the magnetic properties that we will compute , it is important to employ the rotationally invariant form of the interaction term @xmath32 above , including a pair - hopping and a spin - flip term , and not only density - density contributions .
as for the ct - hyb , we note that it is essential to truncate the outer states for the evaluation of the local fermionic trace only at high energies , especially at high temperatures .
[ fig : siw ] shows the imaginary part of the dmft self energy @xmath33 on the matsubara axis for all five fe @xmath12 orbitals . in order to avoid all uncertainties related to an analytical continuation
, we calculate the quasiparticle weight @xmath34 directly from the self energy on the matsubara axis @xmath35 .
this yields a value of @xmath36{}$ ] , essentially the same for all @xmath28 orbitals .
this @xmath34 value would indicate a rather weakly correlated material .
orbitals ( at inverse temperature @xmath37{ev^{-1}}$ ] corresponding to @xmath38k ) .
the extracted quasiparticle weight is @xmath36{}$].,width=321 ] orbitals at @xmath37{ev^{-1}}$ ] , compared to dft ( @xmath39 corresponds to the fermi level).,width=321 ] the corresponding spectral function @xmath40 is shown in fig.[fig : maxent ] on the real frequency axis , for which an analytic continuation using a stochastic version of the maximum entropy method has been used.@xcite in comparison to the dft dos , both occupied and empty states are slightly shifted towards the fermi energy due to the fermi - liquid renormalization .
there is no evidence of pronounced upper and lower hubbard bands and one can only observe a weak increase of the spectral weight at high frequencies . in agreement with ref . , we find that the spectral function at the fermi level is essentially the same in dft+dmft as in dft .
we did not perform charge - self - consistent calculations since the difference in the occupation of the @xmath12-orbitals between the dft - derived hamiltonian and dmft is very small . in dft
, we have @xmath41{}$ ] electrons in the @xmath13 and @xmath31{}$ ] electrons in the @xmath14 states out of 11 electrons per unit cell , in dmft the @xmath13 orbitals are occupied with @xmath41{}$ ] and the @xmath14 orbitals with @xmath42{}$ ] electrons . also the changes in the one - particle spectrum are rather small .
note , only if dmft alters the spatial charge distribution @xmath43 considerably , charge self - consistency would have an effect .
thus , we expect changes by charge - self - consistency to be small . fig .
[ fig : bands_dmft ] presents the corresponding @xmath9-resolved spectrum which shows that also the dft+dmft bands essentially follow the dft bandstructure .
the most noteworthy effects are again a slight shift towards the fermi level , i.e. , a quasiparticle renormalization and a broadening of the bands , especially of the @xmath12 bands located around the fermi level .
hence , regarding only single - particle quantities , feal seems to exhibit only weak correlation effects .
however , this picture changes when considering also two - particle quantities , namely the magnetic susceptibility .
-resolved spectral function ( blue ) compared to the dft bandstructure ( black).,width=321 ]
in order to study the magnetic properties of feal within dft+dmft , we compute the local magnetic susceptibility , represented by the two - particle spin - spin correlation function @xmath44 with @xmath29 and @xmath45 being the orbital indices of the five fe @xmath12 orbitals , @xmath46 the imaginary time , and @xmath47 the gyromagnetic factor for the electronic spin .
@xmath48 is the @xmath49-component of the spin operator of orbital @xmath29 , expressed in terms of the corresponding density operators @xmath50 . technically speaking ,
@xmath51 is obtained by first measuring the generalized magnetic susceptibility @xmath52 of the converged dmft impurity model by means of ct - hyb quantum monte carlo sampling .
thereby , @xmath52 automatically contains all vertex corrections to the bare ( dmft ) bubble spin susceptibility .
the sum over the fermionic matsubara frequencies @xmath53 and @xmath54 and a fourier transform , @xmath55 , finally lead to @xmath51 . here , for the large frequency asymptotics , the bare bubble contribution , eq .
( [ eq:3 ] ) , which is known on a larger frequency grid and an additional fitting function of the form @xmath56 have been used .
the results for @xmath51 are shown in fig .
[ fig : chi_loc ] for @xmath37{ev^{-1}}$ ] ( for lower temperatures the numerically feasible frequency box becomes too small ) . as a function of ( imaginary )
time @xmath46 for @xmath37{ev^{-1}}$ ] .
besides the total susceptibility , also its diagonal and bare - bubble contribution are shown , as well as ( in the inset ) its @xmath14 and @xmath13 contribution .
have been checked to be stable over a temperature range from @xmath57{ev^{-1}}$ ] to @xmath58{ev^{-1}}$].],width=321 ] the solid , blue curve in fig .
[ fig : chi_loc ] corresponds to the total magnetic susceptibility @xmath51 of eq .
( [ chi_loc ] ) .
the dashed , orange curve instead represents the orbital - diagonal contribution @xmath59 .
the dotted , purple curve is the bare - bubble contribution @xmath60 , which neglects vertex corrections and is obtained by directly convoluting the dmft green functions @xmath61 : @xmath62 the significant difference between the bare - bubble contribution and the susceptibility including vertex corrections in fig .
[ fig : chi_loc ] reveals that electronic correlations actually play a major role in feal , more than it could be expected from the weak quasiparticle renormalization .
[ fig : chi_loc ] also shows that the enhancement of @xmath51 stems approximately in equal parts from an enhancement of the intra - orbital contribution ( diagonal part ) and additional inter - orbital ( off - diagonal ) contributions , which are not present in the bare - bubble susceptibility .
the local susceptibilities in fig .
[ fig : chi_loc ] show a rather fast and strong decay in @xmath46 . here
, the value of @xmath51 at @xmath63 can be interpreted in terms of the instantaneous , local magnetic moment .
the observed decay in @xmath46 reflects a dynamical screening of this local magnetic moment due to quantum fluctuations .
thus , we can conclude that dynamical quantum fluctuations significantly reduce the local magnetic moment in feal .
fitting @xmath51 to an exponential between @xmath64{ev^{-1}}$ ] and @xmath65{ev^{-1}}$ ] yields a time scale for the screening of @xmath66{ev^{-1}}=\unit[4.02]{fs}$ ] .
the inverse of @xmath67 is the energy scale associated with the screening which is essentially the bare bandwidth if we have a noninteracting system , the width of the central peak if we consider the interacting bubble , and the kondo temperature for the interacting system with vertex corrections .
this kondo temperature is smaller than the width of the central peak.@xcite hence the decay with vertex corrections should be slower . indeed , in fig .
[ fig : chi_loc ] the total @xmath51 decays slower than the bubble contribution . for a related analysis , how to interpret the susceptibility as a function of imaginary time and how the local , fluctuating magnetic moment reflects as a pronounced low - energy peak in the local neutron spectra , see . in the inset of fig .
[ fig : chi_loc ] , we separate the @xmath14 and @xmath13 contributions of the susceptibility .
these two contributions are rather independent as one clearly sees from the longer time scale on which the @xmath13 susceptibility decays .
this different decay rate can be explained by the considerably more narrow @xmath13 bandwidth and hence stronger correlations of the @xmath13 orbitals .
if hund s exchange was the major player , on the other hand , one would expect a stronger coupling of @xmath14 and @xmath13 susceptibility , and a decay on a similar time scale .
this all suggests that the hund s rule exchange @xmath20 , which mainly drives the inter - orbital contribution , is not exceedingly important in feal .
this is in contrast to other fe - based compounds such as the iron - based superconductor lafeaso.@xcite which have been classified as hund s metals .
@xcite from the local magnetic properties , we now turn to the bulk magnetic susceptibility and the long - range ordered ferromagnetic moment .
[ fig : magn ] shows the ordered magnetic moment , which has been obtained by breaking the spin symmetry in the first dmft iteration so that the system can either stabilize a para- or ferromagnetic solution . as fig .
[ fig : magn ] clearly shows , the ordered ferromagnetic moment is zero down to a temperature of @xmath0{k}$ ] .
thus , in the investigated temperature range , feal is paramagnetic in dft+dmft , in agreement with experiment but in contrast to dft .
this result is also supported by the calculation of the bulk ferromagnetic susceptibility in dft+dmft . to this end
, we have applied a small magnetic field of @xmath68{ev}$ ] , checked ( for some temperatures ) that this is still in the linear @xmath69 vs. @xmath70 regime ( which further confirms the paramagnetic phase ) and calculated @xmath71 at this @xmath70 . this way
all vertex corrections are included ; and this quantity allows to determine whether there is a second order phase transition towards a ferromagnetic phase or not .
prospectively competing phases with a different wave vector @xmath72 are however not accessible this way .
the full @xmath73-dependent susceptibility @xmath74 could in principle be obtained by solving the bethe - salpeter equation .
unfortunately , this is computationally too demanding for five orbitals at low temperatures . for the same reason the local susceptibility @xmath51 could only be calculated reliably down to @xmath37{ev^{-1}}$ ] . but to gain at least some insight whether ferromagnetism or magnetic phases with other @xmath72-vectors prevail , we study the bare bubble susceptibility @xmath75 , which does not include vertex corrections . the result shown in fig .
[ fig : chi0_q ] indicates that @xmath76 is the leading instability .
would predict ferromagnetism for @xmath77 or @xmath78 , it is known that this criterion largely overestimates the tendency towards ferromagnetism.@xcite ] thus , in the following we will focus on @xmath79 .
vs. @xmath80 and @xmath81 ( at @xmath82 ) calculated from the dmft @xmath12-electron green functions at @xmath83{ev^{-1}}$ ] .
the maximum at @xmath76 indicates that without vertex corrections ferromagnetism is the leading instability.,width=321 ] the temperature dependence of the susceptibility @xmath71 including vertex corrections is shown in fig .
[ fig : chi_q0 ] . upon decreasing temperature
, we first notice an increase of the susceptibility .
however , below @xmath84k , the susceptibility decreases again .
this clear trend of a _ reduction _ of the susceptibility by decreasing @xmath85 makes the onset of a ferromagnetic order at lower temperatures extremely unlikely .
we note that a marked low-@xmath85 reduction of @xmath79 has been also reported experimentally@xcite and theoretically@xcite in the iron - pnictide compound lafaaso . there , this behavior of @xmath79 coexists with an opposite ( increasing ) trend of the local magnetic susceptibility @xmath86 .
@xcite hence , the unusual low-@xmath85 reduction of @xmath79 has been attributed to specific features of the one - particle spectral function of lafeaso , displaying significant temperature variations near the fermi level . by performing the same analysis for feal
we find , however , that the low-@xmath85 behavior of @xmath86 ( inset of fig .
[ fig : chi_q0 ] ) and @xmath79 ( main panel ) is qualitatively very similar : both show a visible reduction for @xmath87k . in the very same temperature interval ,
a slight reduction of the instantaneous local moment ( @xmath88 , inset ) is also found , which is a typical behavior in the fermi liquid regime , as described by the dmft .
hence , in feal , the role played by emerging low - energy structures of the spectral function appears to be less important than in lafeaso .
rather , the trend of @xmath79 in feal may simply reflect the corresponding low-@xmath85 reduction of the local magnetic moment ( @xmath89 ) , in particular of the screened one .
the latter can be ascribed to the enhanced metallic coherence of the low - temperature region , which is a general effect of local correlations in the fermi - liquid regime . as a function of temperature .
the inset shows the temperature dependence of the corresponding local quantities : the local magnetic susceptibility ( @xmath90 , diamonds ) and the instantaneous one ( @xmath91 , circles ) , from which the local magnetic moment can be estimated ( @xmath89).,width=321 ]
in the present work , we studied feal by means of dft+dmft . in particular , we investigated the effects of electronic correlations on the magnetic properties since standard spin - polarized dft calculations yield a ferromagnetic ground state whereas in experiment it is a paramagnet . from dft
, we constructed a 9-band wannier hamiltonian with four al @xmath17 orbitals and five fe @xmath12 orbitals .
for the latter we include a local @xmath10-symmetric kanamori interaction with @xmath21{ev}$ ] , @xmath22{ev}$ ] and @xmath23{ev}$ ] as obtained from crpa and solve the many - body problem by dmft(ct - qmc ) . on the one particle level ,
the self energy and green function suggest rather weak electronic correlations with a quasiparticle renormalization of only @xmath92 and no evidence of pronounced upper and lower hubbard bands . in dft+dmft
we calculate from the spin - spin correlation function an equal - time local magnetic moment of @xmath1{\mu_b}$ ] which is twice as large as the magnetic moment in spin - polarized dft .
it is also much larger than the bubble contribution which demonstrates that electronic correlations are pivotal for the two - particle quantities in feal .
even more importantly , our results show that the moment is fluctuating in time and screened on the fs time scale .
this explains why we also do not find long - range ferromagnetic order .
according to our dft+dmft study , feal is paramagnetic with a maximum in the ferromagnetic susceptibility around room temperature and no tendency towards long range magnetic order in the temperature range studied .
previously , it has been proposed that disorder and a spin - glass behavior might explain the missing ferromagnetic moment in experiment .
our results show that if temporal fluctuations are taken into account the moment is actually screened on short time scales .
there is hence , even for a perfect lattice , neither ferromagnetism nor a local moment on longer time scales . if the magnetic moments were constant in time and spatially disordered , mssbauer experiments , which probe the _ local _ magnetic moment , are in principle able to reveal it .
the fluctuating local moment on the fs time scales , can however not be observed in mssbauer spectroscopy which can not resolve such short time scales .
hence , the mssbauer experiments , @xcite which show no local magnetic moment for stoichiometric feal , seem to better agree with a magnetic moment fluctuating in time as we find in dft+dmft than with a moment fluctuating in space .
we are grateful to p. mohn , s. khmelevskyi , r. podlucky , m. reissner , m. karolak and j. tomczak for valuable discussions .
we acknowledge financial support by the austrian science fund ( fwf ) through doctoral school solids4fun w1243 ( ag ) , sfb vicom f41 ( mw , gk , at ) , dfg research unit for 1346 ( mk , ct , gs ) , sfb 1170 tocotronics and by the european research council under the european union s seventh framework program ( fp/2007 - 2013)/erc through grant agreement n. 306447 ( ag , kh ) .
the computational results presented have been obtained using the vienna scientific cluster ( vsc ) . | while the stoichiometric intermetallic compound feal is found to be paramagnetic in experiment , standard band - theory approaches predict the material to be ferromagnetic .
we show that this discrepancy can be overcome by a better treatment of electronic correlations with density functional plus dynamical mean field theory .
our results show no ferromagnetism down to @xmath0{k}$ ] and since the susceptibility is decreasing at the lowest temperatures studied we also do not expect ferromagnetism at even lower temperatures . this behavior is found to originate from temporal quantum fluctuations that screen short - lived local magnetic moments of @xmath1{\mu_b}$ ] on fe . |
rayleigh - bnard convection in a thin horizontal fluid layer heated from below is a canonical example of pattern formation in a continuous dissipative system far from equilibrium @xcite . under various conditions ,
wavenumber selection mechanisms have been identified @xcite that reduce the band of stable wavenumbers , sometimes to a single value .
one such selection mechanism occurs when there is a one - dimensional spatial variation or ramping of the control parameter @xmath0 ; where @xmath1 , @xmath2 is the rayleigh number , and @xmath3 its critical value @xcite .
this can be accomplished by varying the plate separation @xmath4 such that @xmath0 goes from @xmath5 in the bulk of the layer ( i.e. , the unramped region ) to @xmath6 as a lateral boundary is approached .
it is expected that in the idealized case of an infinitely gradual one - dimensional ramp the wavenumber will equal @xmath7 at the position where the layer depth yields @xmath3 .
it has been shown under very general conditions , that as long as the layer becomes critical somewhere along the ramp this is sufficient to fix the wavenumber @xmath8 in the bulk and over the rest of the ramp @xcite .
for slightly supercritical conditions it is expected that the selected wavenumber in the bulk can be expressed as @xmath9 where @xmath10 @xcite and @xmath11 depends on the prandtl number , @xmath12 , and the specifics of the ramp .
recent rayleigh - bnard convection experiments @xcite in a cylindrical cell with a two - dimensional radial ramp in plate separation have generated intriguing results .
specifically , the plate separation as a function of radius used in experiment is , using the layer depth @xmath4 to nondimensionalize , @xmath13 , & \mbox{$r \ge r_0 $ } \end{array}\right .
\label{eq : platesep}\ ] ] where @xmath14 is a constant and the radius values @xmath15 and @xmath16 are the locations where the ramp begins and ends , respectively .
the ramp always extends to the sidewall .
a schematic of the cosine ramp is shown in fig .
[ fig : config ] ; note that @xmath15 and @xmath16 are geometric constants but @xmath17 , the location where the plate separation yields @xmath3 , is a function of the ramp shape and @xmath18 such that , for a ramp given by eq .
( [ eq : platesep ] ) , @xmath19 . defines where the ramp begins , @xmath16 defines where the ramp ends and @xmath17 is where the plate separation corresponds to the critical rayleigh number .
the ramp shown is a cosine ramp given by eq .
( [ eq : platesep ] ) . for presentation purposes
we show a steep ramp with @xmath20 ( this domain is not used in the simulations).,width=288 ] experiments using the cosine ramp defined by eq .
( [ eq : platesep ] ) have yielded unexpected results for the wavenumber @xcite .
mean pattern wavenumber measurements ( using the fourier methods discussed in @xcite ) yielded @xmath21 .
additionally , measurements of the local wavenumber defined at each position in space ( method discussed in @xcite ) displayed interesting variation as @xmath18 is increased .
for the time independent patterns the bulk region , @xmath22 , contains approximately straight parallel rolls . near threshold , @xmath23 ,
a centered egg - shaped domain of convection rolls with small wavenumber , @xmath24 , extends through the convection cell with the long - axis _
parallel _ to the roll axes . a dramatic roll expansion from @xmath25 at the edge where the ramp begins to @xmath26 in the center of the domain is observed .
as @xmath18 increases , the wavenumber field evolves into a domain characterized by large wavenumbers extending through the layer with the long - axis _ perpendicular _ to the roll axes ( see fig . 3 of @xcite ) .
similar experiments without a ramp do not exhibit this wavenumber behavior @xcite . in the experiments with a radial ramp in plate separation , for @xmath27 , time dependent states were found through the repeated formation of defects via an eckhaus mechanism consistent with the local wavenumbers crossing the eckhaus stability boundary for an ideal infinite layer of two - dimensional rolls .
also in the ramped experiments , for @xmath28 , defects were formed via a skewed varicose mechanism consistent with the local wavenumber exceeding the skewed varicose stability boundary for an ideal infinite layer of two - dimensional rolls similar to what has been observed in experiments in unramped cylindrical domains with rigid sidewalls @xcite .
it has been suggested that these features may be the result of the interaction of the convective roll pattern and weak large scale flows @xcite .
the visualization and quantification of these large scale flows is not possible in the current generation of experiments .
however , we are able to make these measurements by performing full numerical simulations with a new spectral element code ( discussed further in section [ section : numerical simulation ] ) .
we utilize the complete knowledge of the flow field together with analytical results valid near threshold to explore this further .
in this work the terminology large scale flows is used to describe flows that extend over distances larger than that of the convection roll scale .
we would like to distinguish between two different large scale flows : _ large scale counterflow _ and _ mean flow_. in the presence of a spatial ramp in plate separation a large scale counterflow is present for all values of the bulk control parameter @xmath18 , including @xmath29 .
warm fluid ascends the ramp eventually reaching the sidewall and is forced to flow back toward the center of the domain over the cold top wall causing it to descend resulting in a large zone of circulation in the vertical plane over the ramp .
the magnitude of the large scale counterflow depends upon the specifics of the ramp and is roughly independent of @xmath18 and @xmath12 @xcite .
figure [ fig : yslice ] illustrates this with a vertical slice from a three - dimensional numerical simulation where the entire convection layer is subcritical . as shown for this subcritical case the fluid motion of the large scale counterflow generates axisymmetric convection near the base of the ramp which extends a couple of roll widths toward the center of the domain . for the gradual ramp used in experiment
the large scale counterflow is small in magnitude and has not been measured . .
the ramp parameters are @xmath30 , @xmath31 , and @xmath32 .
a steep ramp is shown to clearly illustrate the flow .
solid vertical lines indicate the boundaries of the spectral element grid used in the simulation , only a portion of the layer is shown emphasizing the ramped region.,width=288 ] in discussing the mean flow it will be convenient to first present the governing heat and fluid equations .
the velocity @xmath33 , temperature @xmath34 , and pressure @xmath35 , evolve according to the boussinesq equations , @xmath36 where @xmath37 indicates time differentiation , and @xmath38 is a unit vector in the vertical direction opposite to gravity .
the equations are nondimensionalized in the standard manner using the layer depth @xmath4 , the vertical diffusion time for heat @xmath39 where @xmath40 is the thermal diffusivity , and @xmath41 the temperature difference between the top and bottom surfaces , as the length , time , and temperature scales , respectively .
the lower and upper surfaces are no - slip and are held at constant temperature .
the sidewalls are also no - slip and , unless otherwise noted , are insulating .
the mean flow field , @xmath42 , is the horizontal velocity integrated over the depth and originates from the reynolds stress induced by pattern distortions . as illustrated by the fluid equations , eqs .
( [ eq : mom ] ) and ( [ eq : mass ] ) , it is evident that the pressure is not an independent dynamic variable .
the pressure is determined implicitly to enforce incompressibility , @xmath43 + r \partial_z t.\ ] ] focussing on the nonlinear reynolds stress term and rewriting the pressure as @xmath44 yields , @xmath45 \right>_z , \ ] ] where @xmath46 represents an average in the z direction . in eq .
( [ eq : pressure_p0 ] ) the @xmath47 is not exact , in order to be more precise the finite system green s function would be required ; however the long range behavior persists .
this gives a contribution to the pressure that depends on distant parts of the convection pattern .
the poiseuille - like flow driven by this pressure field subtracts from the reynolds stress induced flow leading to a divergence free horizontal flow that can be described in terms of a vertical vorticity .
near threshold an explicit expression for the mean flow is @xcite @xmath48 where @xmath49 is a coupling constant given by @xmath50 , @xmath51 is the convection amplitude normalized so that the convective heat flow per unit area relative to the conducted heat flow at @xmath3 is @xmath52 , @xmath53 is a slowly varying pressure , see eq .
( [ eq : pressure_p0 ] ) , and @xmath54 is the horizontal gradient operator [ see @xcite for the complete analysis and more details ] .
the mean flow is important not because of its strength ; under most conditions the magnitude of the mean flow is substantially smaller than the magnitude of the roll flow making it extremely difficult to quantify experimentally .
the mean flow is important because it is a nonlocal effect acting over large distances ( many roll widths ) and changes important general predictions of the phase equation @xcite .
the mean flow is driven by roll curvature , roll compression and gradients in the convection amplitude .
the resulting mean flow advects the pattern giving an additional slow time dependence .
it is important to note , that unlike the long range counterflow , the magnitude of the mean flow vanishes when the convection layer becomes critical , @xmath55 for @xmath56 .
we have performed full numerical simulations of the governing fluid and heat equations , eqs .
( [ eq : mom])-([eq : mass ] ) , in a cylindrical geometry with a radial ramp in plate separation using a parallel spectral element algorithm ( described in detail elsewhere @xcite , see for related applications ) . for discussion purposes it will be convenient to define cartesian @xmath57 and polar @xmath58 coordinates centered on a mid - depth horizontal slice of a cylindrical convection layer containing a field of straight parallel x - rolls with wavevector @xmath59 as shown in fig .
[ fig : config_xy ] .
the x - axis is perpendicular to the roll axes , the y - axis is parallel to the roll axes and @xmath60 measures the angle from the positive x - axis .
and polar @xmath58 coordinates defined on a mid - depth horizontal cross - section of a cylindrical convection layer containing a field of x - rolls described by @xmath59.,width=192 ] we have investigated the results found in experiment by performing simulations on a ramped cylindrical convection layer for a variety of scenarios and initial conditions .
simulations were performed over the range @xmath61 and for simulation times of @xmath62 where @xmath63 is the time required for heat to diffuse horizontally across the bulk region of the layer which has been suggested as the earliest time scale for the flow field to reach equilibrium @xcite .
the mean flow present in the simulation flow fields , @xmath64 , is investigated by calculating the depth averaged horizontal velocity , @xmath65 where @xmath66 is the horizontal velocity field
. furthermore it will be convenient to work with the vorticity potential , @xmath67 , defined as @xmath68 where @xmath69 is the vertical vorticity and @xmath70 is the horizontal laplacian .
near threshold , assuming straight parallel rolls , it is possible to approximately determine @xmath69 analytically . it will be convenient to start from @xmath69 given by the vertical component of the curl of eq .
( [ eq : meanflow ] ) , latexmath:[\[{\omega}_z = \hat{z } { { \scriptscriptstyle \stackrel{\bullet}{{}}}}\left ( \vec{\nabla}_{\perp } \times \vec{u } \right ) = - \gamma \hat{z } { { \scriptscriptstyle \stackrel{\bullet}{{}}}}\vec{\nabla}_{\perp } \times \left [ \vec{k } \vec{\nabla}_{\perp } { { \scriptscriptstyle \stackrel{\bullet}{{}}}}\left ( \vec{k } convection layer with a radial ramp in plate separation containing a field of x - rolls given by @xmath59
. the amplitude can be represented for large @xmath18 , using an adiabatic approximation , as @xmath72 for @xmath73 and @xmath74 for @xmath75 , making the amplitude a function of radius only @xmath76 .
this approximation is good except for the kink at @xmath17 where @xmath77 . inserting @xmath76 into eq .
( [ eq : wz_general ] ) yields , after some manipulation , the following expression for the vertical vorticity , @xmath78 \sin 2 \theta .
\label{eq : wz}\ ] ] to correct for nonadiabaticity and to smooth @xmath79 near @xmath17 , the one - dimensional time independent amplitude equation is solved , @xmath80 where angular derivatives have been assumed small , @xmath81 , @xmath82 and @xmath83 is determined by @xmath84 where @xmath85 . in eq .
( [ eq : wz_general ] ) @xmath69 is dominated by radial derivatives so eq . ( [ eq : wz ] ) is still a good approximation for the case @xmath86 , @xmath87 . due to the angular dependence of the nonadiabaticity @xmath88 is now @xmath60 dependent which will induce higher angular harmonics in @xmath69 .
we will neglect these higher harmonics , assume a @xmath89 dependence and approximately evaluate the magnitude using eqs .
( [ eq : wz ] ) and ( [ eq : amp_dim ] ) at @xmath90 .
equation ( [ eq : amp_dim ] ) can be rewritten in a more convenient form as , @xmath91 where @xmath92 and @xmath93 @xcite ( since the amplitude goes to zero it can be shown that @xmath94 is now a real quantity @xcite ) . equation ( [ eq : amp_nondim ] ) is solved numerically using the boundary conditions @xmath95 at @xmath96 , and @xmath97 at @xmath98 .
these analytical results are now used to investigate the vertical vorticity generation in a large radially ramped cylindrical convection layer .
we start by looking at the configuration used in the recent experiments with input parameters : @xmath99 , @xmath100 , @xmath101 and @xmath102 .
when @xmath103 the amplitude @xmath104 is unable to adiabatically follow the ramp , this nonadiabaticity results in a considerable deviation from @xmath105 as shown in fig .
[ fig : amp_eps_g101_1715 ] .
however , when @xmath106 the amplitude @xmath104 follows @xmath105 adiabatically almost over the entire ramp except for the small kink at @xmath17 as shown in fig .
[ fig : amp_eps_g101_2000 ] .
the structure of @xmath69 depends upon this adiabaticity and is shown for various values of @xmath18 in fig .
[ fig : vort_var_sm_g101 ] where the @xmath107 dependence has been removed by choosing @xmath90 . if the @xmath60 dependence is included it is evident from fig .
[ fig : vort_var_sm_g101 ] that the vertical vorticity has a quadrupole angular structure for small @xmath18 , i.e. four lobes of alternating positive and negative vorticity with one lobe per quadrant , and makes a transition to an octupole angular dependence for larger @xmath18 , octupole in the sense of an inner , @xmath108 , and outer , @xmath109 , quadrupole .
in addition , since @xmath110 there is a radial shift of the vorticity curves as @xmath18 is increased . in all cases the amplitude @xmath104 decreases monotonically with @xmath111 and as a result @xmath112 thus generating only positive vorticity . however , the term @xmath113 can be of either sign and is responsible for the quadrupole and octupole angular structure in @xmath69 . )
plotted as @xmath104 for @xmath99 , @xmath114 , @xmath100 , @xmath115 , @xmath102 and @xmath116 .
also shown for comparison is @xmath105.,width=288 ] ) plotted as @xmath104 for @xmath99 , @xmath117 , @xmath100 , @xmath115 , @xmath102 and @xmath118 .
also shown for comparison is @xmath105.,width=288 ] with @xmath18 as determined analytically from eqs .
( [ eq : wz ] ) and ( [ eq : amp_nondim ] ) illustrating the evolution from a mean flow with a quadrupole dependence to an octupole dependence as @xmath18 increases . shown
explicitly are @xmath69 curves for @xmath119 , @xmath120 , and @xmath121 with the parameters @xmath99 , @xmath100 , @xmath115 and @xmath102 labelled ( a)-(c ) , respectively.,width=288 ] as @xmath18 approaches zero the nonadiabaticity of @xmath104 increases until @xmath104 exhibits a quadratic fall - off with @xmath111 for @xmath122 resulting in @xmath123 for all @xmath111 .
the mean flow generated by these vorticity distributions is determined by solving eq .
( [ eq : wz_general ] ) with the boundary condition @xmath124 .
the vorticity potential is related to the mean flow in polar coordinates by @xmath125 .
the vorticity potential is expanded radially in second order bessel functions while maintaining the @xmath89 angular dependence .
of particular interest is the mean flow perpendicular to the convection rolls , @xmath126 or equivalently @xmath127 , which is shown in fig .
[ fig : mf_var_sm_g101 ] . with @xmath18 as determined from eq .
( [ eq : vorticity_potential ] ) .
shown explicitly are curves for @xmath128 , @xmath129 , and @xmath121 labelled ( a)-(c ) , respectively with the parameters @xmath99 , @xmath100 , @xmath115 and @xmath102.,width=288 ] as expected , regions of negative and positive vorticity yield corresponding negative and positive values of the mean flow .
as @xmath18 vanishes @xmath130 for all @xmath111 providing a mechanism for roll expansion in the bulk . for larger @xmath18 the mean flow becomes larger in magnitude and increasingly negative for @xmath131 providing a mechanism for roll compression . to make the connection between mean flow and wavenumber quantitative it is noted that the wavenumber variation resulting from a mean flow across a field on x - rolls can be determined from the one - dimensional phase equation , @xmath132 where the wavenumber is the gradient of the phase , @xmath133 , @xmath134 , and @xmath135 @xcite . assuming that for @xmath56 the wavenumber is approximately @xmath136 everywhere and that the rolls are exposed to a constant mean flow the wavenumber change over the bulk can be expressed as @xmath137 for example , for curve ( c ) in fig .
[ fig : mf_var_sm_g101 ] the maximum value of the mean flow is @xmath138 which yields a small roll expansion of @xmath139 . if the mean flow were solely responsible for the dramatic roll expansion seen in experiment of @xmath140 ( @xmath141 see fig .
3(a ) of @xcite ) a mean flow of @xmath142 would be required , which is not found in the analytic results .
the large scale flows discussed in section [ section : lsf ] can not be measured in current experiments placing us in a unique position to use the complete flow field information from our full numerical simulations of eqs .
( [ eq : mom])-([eq : mass ] ) together with the analytical results of section [ section : analytical development ] to investigate how the mean flow and the large scale counterflow induce wavenumber distortions and the variation of this distortion with rayleigh number .
it is computationally expensive to perform full three - dimensional numerical simulations for the very large system used in experiment .
we have , however , performed a variety of simulations for radially ramped cylindrical convection layers .
the full three - dimensional simulations are of smaller spatial extent with the precise ramp defined by eq .
( [ eq : platesep ] ) and the specific input parameters : @xmath143 , @xmath144 , @xmath115 , and @xmath102 . two - dimensional simulations of a vertical slice of a three - dimensional domain ( see fig . [
fig : config ] ) were also performed for both the large experimental configuration and the smaller computational domain just described .
three - dimensional simulations were also conducted without a large scale counterflow by a specific choice of ramp parameters that will be discussed below .
initially we consider prescribed x - roll initial conditions given by @xmath145 .
other initial conditions such as random thermal perturbations or initial x - rolls of varying wavenumbers were also investigated and found not to affect the final pattern wavenumber or any of the conclusions drawn .
simulations were performed for @xmath146 , 0.054 , 0.113 , and 0.171 .
figure [ fig : wnall ] compares the wavenumbers found in these simulations with recent experiments and will be discussed in detail below . ) , @xmath147 , between simulation ( solid lines with symbols ) and experiment ( dashed - dotted line ) @xcite . unless otherwise noted @xmath143 , @xmath144 , @xmath115 , and @xmath102 .
the symbols represent : @xmath148 three - dimensional simulations , ( @xmath149 ) two - dimensional simulations , ( @xmath150 ) three - dimensional simulations for a specific ramp construction without a large scale counterflow , and ( @xmath151 ) two - dimensional simulations with @xmath99 , and @xmath100 .
dark solid lines denote the approximate location of the neutral ( n ) and eckhaus ( e ) stability boundaries for an ideal infinite layer of parallel rolls.,width=288 ] the final patterns in the simulations maintain the x - roll configuration imposed by the initial conditions .
figure [ fig : tp ] displays the final pattern observed for three - dimensional simulations with @xmath152 in panel ( a ) and @xmath153 in panel ( b ) .
figure [ fig : tp]a illustrates that near threshold the convection rolls exhibit very little curvature indicating that the assumption of straight parallel x - rolls in section [ section : analytical development ] is valid .
there is more roll curvature apparent in fig .
[ fig : tp]b as would be expected for larger @xmath18 .
figure [ fig : tp ] also illustrates the decreasing size of the subcritical region as the supercriticality of the bulk increases .
all simulations settled to a time independent state . and
@xmath154 are shown in panels ( a ) and ( b ) , respectively .
shaded contours of the thermal perturbation are shown with dark regions representing cool descending fluid and light regions warm ascending fluid .
the inner dotted circle indicates where the ramp begins , @xmath15 , and the outer dotted circle indicates where the convection layer becomes critical , @xmath17 .
simulation parameters , @xmath143 , @xmath144 , @xmath115 , @xmath102.,width=240 ] it is illustrative to compare the analytical results of section [ section : analytical development ] with the results of simulation .
figure [ fig : amp_mf_vort_eps_exp1750]a displays @xmath104 for the case @xmath152 , as determined by eq .
( [ eq : amp_nondim ] ) .
a significant nonadiabaticity is present for this case as shown by the deviation of @xmath104 from @xmath105 . for the ramped domain used in simulation
the distance @xmath155 is smaller than in the larger domain with a more shallow ramp used in experiment .
this results in the presence of more nonadiabaticity in the simulations when compared to experimental results at the same control parameter .
this is beneficial because this allows the exploration of highly nonadiabatic situations without having to perform the task of simulating near the convective threshold , which becomes computationally difficult because of the diverging time scales .
a comparison between theory and simulation of the vertical vorticity and the resulting mean flow is shown in fig .
[ fig : amp_mf_vort_eps_exp1750]b and c. the theoretical predictions are based on the amplitude variation caused when straight parallel convection rolls encounter a radial ramp in plate separation as discussed in section [ section : analytical development ] .
for both the vertical vorticity and the mean flow the comparison is made in the absence of any adjustable parameters . for the vertical vorticity calculated in simulation an angular average , weighted by @xmath89 ,
is used for the comparison .
the agreement between theory and simulation is quite good .
this illustrates quantitatively that the major source of vertical vorticity and mean flow is indeed the variation in the convective amplitude caused by the radial ramp in plate separation . over the bulk of the domain
the mean flow is negative and very small in magnitude with a maximum value of @xmath156 and by eq .
( [ eq : delta_k ] ) the wavenumber variation would be extremely small in agreement with the near constant bulk wavenumbers found in simulation . )
plotted as @xmath104 , shown for comparison is @xmath105 .
panel ( b ) compares the vertical vorticity found analytically from eq .
( [ eq : amp_nondim ] ) with an angular average , weighted by @xmath89 , of the vertical vorticity from simulation .
panel ( c ) compares the mean flow found analytically from eq .
( [ eq : vorticity_potential ] ) with the mean flow from simulation flowing along the x - axis at @xmath157 .
parameters are @xmath143 , @xmath158 , @xmath159 , @xmath115 , @xmath102 and @xmath146.,width=240 ] a similar comparison between theory and simulation is made in fig .
[ fig : amp_mf_vort_eps_exp2000 ] for @xmath153 . as shown in fig .
[ fig : amp_mf_vort_eps_exp2000]a @xmath104 is much more able to follow the ramp , @xmath105 , and exhibits very little nonadiabaticity except for the kink near @xmath17 .
this results in a much stronger negative vertical vorticity in the bulk which in turn yields a larger negative mean flow as shown if fig . [ fig : amp_mf_vort_eps_exp2000]b and c. the agreement between theory and simulation for the vertical vorticity is still quite good .
the discrepancy in the mean flow comparison may be due to the fact that as @xmath18 increases other mean flow sources such as roll curvature , see fig .
[ fig : tp]b , become important . ) plotted as @xmath104 , shown for comparison is @xmath105 .
panel b compares the vertical vorticity found analytically from eq .
( [ eq : amp_nondim ] ) with an angular average , weighted by @xmath89 , of the vertical vorticity from simulation .
panel ( c ) compares the mean flow found analytically from eq .
( [ eq : vorticity_potential ] ) with the mean flow from simulation flowing along the x - axis at @xmath157 .
parameters are @xmath143 , @xmath158 , @xmath159 , @xmath115 , @xmath102 and @xmath160.,width=240 ] figure [ fig : mf_wn ] illustrates the octupole structure in the vorticity potential in panel ( a ) and the roll compression occurring in the bulk by plotting contours of the local wavenumber in panel ( b ) for @xmath160 .
as illustrated in panel ( a ) the mean flow has significant structure over the ramped region as well as extending into the subcritical region of the layer , @xmath161 .
it has also been suggested that the mean flow extends into a subcritical region in related experiments implementing finned " boundaries @xcite . the vorticity potential displays an octupole structure containing a pair of counter - rotating vortices in each quadrant
. the inner quadrupole is localized around @xmath15 where gradients in the amplitude of convection occur as the ramp in plate separation begins .
the direction of rotation of the inner quadrupole causes a focussing of the mean flow into the bulk region of the domain and is responsible for the larger wavenumbers found as @xmath18 is increased as shown by the ( @xmath162 ) curve in fig .
[ fig : wnall ] .
, panel ( a ) , ( light indicates counterclockwise rotation drawn with solid contours and dark indicates clockwise rotation drawn with dashed contours ) and the corresponding local wavenumber distribution , @xmath163 , panel ( b ) .
the magnitude of the mean flow is approximately @xmath164 of the magnitude of the velocity field , @xmath165 .
local wavenumber distributions are shown only in the bulk , @xmath22 .
the inner dotted circle indicates where the ramp begins , @xmath15 , and the outer dotted circle indicates where the convection layer becomes critical , @xmath17 .
simulation parameters , @xmath143 , @xmath144 , @xmath115 , @xmath102 and @xmath153 ( @xmath166).,width=240 ] to make the connection between mean flow and wavenumber quantitative eq .
( [ eq : phase ] ) is applied to the simulation results in the form , @xmath167 figure [ fig : wn_mf_compare2000]a illustrates the wavenumber variation , @xmath168 for @xmath22 , found in simulation by simply measuring the distance between roll boundaries and makes evident the roll compression , @xmath169 .
figure [ fig : wn_mf_compare2000]b compares the mean flow calculated from simulation with the predicted value of the mean flow required to produce the wavenumber variation shown in fig .
[ fig : wn_mf_compare2000]a using eq .
( [ eq : phase_new ] ) .
the agreement is good and the discrepancy near @xmath15 , which is contained within one roll wavelength from where the ramp begins , is expected because the influence of the ramp was not included in eq .
( [ eq : phase ] ) .
this illustrates quantitatively that the mean flow compresses the rolls in the bulk of the domain . at @xmath170 .
panel ( b ) , a comparison of the mean flow from simulation ( solid line ) with the predicted value calculated from eq .
( [ eq : phase ] ) using the wavenumber variation from panel ( a ) .
simulation parameters , @xmath143 , @xmath144 , @xmath115 , @xmath102 and @xmath153 ( @xmath166).,width=240 ] as mentioned earlier , the mean flow vanishes as @xmath18 approaches critical whereas the large scale counterflow is present for all @xmath18 and therefore could play a role near threshold in the determination of the final convection pattern . in order to gain further insight into this possibility a radial ramp
was constructed that did not drive a large scale counterflow .
this was accomplished by setting the temperature of the ramped surface , @xmath171 , to the value of the linear conduction profile at that height , @xmath172 .
this ramp , therefore , does not bend the isotherms which is the source of the large scale counterflow .
the wavenumber variation for these simulations , see curve labelled with ( @xmath150 ) in fig .
[ fig : wnall ] , does not differ strongly from the simulations with a ramp producing large scale counterflow , see the curve labelled with ( @xmath162 ) .
the similarity in wavenumber results is strongest for small @xmath18 suggesting that the large scale counterflow is not responsible for the shift of the critical wavenumber to smaller values as seen in experiment . to study the large scale counterflow further two - dimensional simulations
were also performed , corresponding to a vertical slice of the domain considered thus far , and in addition to a more spatially extended domain as used in experiment . in two dimensions
the mean flow is absent however the large scale counterflow persists . as shown by the ( @xmath151 ) and ( @xmath149 ) curves in fig . [
fig : wnall ] the wavenumbers measured in the two - dimensional simulations are not compressed to the same extent as @xmath18 increases as in the three - dimensional simulations with both mean flow and large scale counterflow present .
as expected , the wavenumbers found in the two - dimensional simulations are also independent of aspect ratio .
additionally , for small @xmath18 the wavenumber found in simulation does not deviate markedly from its critical value suggesting that the large scale counterflow is not responsible for the wavenumber shift observed in experiment near threshold regardless of the spatial extent of the domain .
we also investigated the possibility that the vertical large scale counterflow could bifurcate into a horizontal flow similar to the mean flow in the presence of a slight spatial asymmetry .
this was accomplished by giving the ramped domain used in the full three - dimensional simulations an eccentricity of @xmath173 for a variety of ramps @xmath174 and simulating over a range of subcritical and supercritical conditions .
for all of the scenarios tested the large scale counterflow remained vertical and did not undergo any significant changes .
lastly , the possibility of wavenumber pinning was studied by varying the aspect ratio in increments of less than half of a roll width for both the two- and three - dimensional domains . in all of the scenarios tested the final pattern wavenumbers were not appreciably affected by these small changes in aspect ratio .
we have analytically and numerically investigated pattern formation in a cylindrical convection layer with a radial ramp in plate separation . in particular , we have studied quantitatively the effects of two large scale flows ; large scale counterflow and mean flow .
these large scale flows are important theoretically yet are extremely difficult to measure experimentally .
our results suggest that the mean flow plays an important role in the observed pattern wavenumber and is generated in a novel way , by the spatial variation of @xmath88 driven by the variation of @xmath0 rather than the more usual variations in roll curvature and wavenumber .
the mean flow sources are quantified analytically and agreement is found with numerical results . the geometric structure and magnitude of the mean flow
is used to explain quantitatively the wavenumber variation found in the simulations .
the large scale counterflow is investigated numerically and our results indicate a small roll compression effect away from threshold . in particular ,
although the large scale counterflow is present at and near threshold it does not appear responsible for the dramatic wavenumber shift to values less than critical as seen in experiment .
although it is too expensive computationally to simulate the very large systems used in the experiments we can use our quantitative understanding of the ramp - generated mean flow , validated by the simulations at smaller aspect ratio , to extrapolate our results to these larger systems .
furthermore our two - dimensional simulations in sizes equal to the experimental ones allow us to estimate the effect of the large scale counterflow on the wavenumber distribution . despite these exhaustive efforts ,
we are unable to reproduce the large shift to smaller wavenumbers observed near threshold in the experiments , and the physical origin of these results remains a mystery .
we are grateful to h. s. greenside and g. ahlers for helpful discussions .
this research was supported by the u.s .
department of energy , grant de - ft02 - 98er14892 , and the mathematical , information , and computational sciences division subprogram of the office of advanced scientific computing research , u.s .
department of energy , under contract w-31 - 109-eng-38 .
we also acknowledge the caltech center for advanced computing research and the north carolina supercomputing center . | pattern formation in rayleigh - bnard convection in a large - aspect - ratio cylinder with a radial ramp in the plate separation is studied analytically and numerically by performing numerical simulations of the boussinesq equations .
a horizontal mean flow and a vertical large scale counterflow are quantified and used to understand the pattern wavenumber .
our results suggest that the mean flow , generated by amplitude gradients , plays an important role in the roll compression observed as the control parameter is increased .
near threshold the mean flow has a quadrupole dependence with a single vortex in each quadrant while away from threshold the mean flow exhibits an octupole dependence with a counter - rotating pair of vortices in each quadrant .
this is confirmed analytically using the amplitude equation and cross - newell mean flow equation . by performing numerical experiments the large scale counterflow
is also found to aid in the roll compression away from threshold but to a much lesser degree .
our results yield an understanding of the pattern wavenumbers observed in experiment away from threshold and suggest that near threshold the mean flow and large scale counterflow are not responsible for the observed shift to smaller than critical wavenumbers . |
bose - einstein correlations ( bec ) , between identical bosons , have been extensively studied in hadronic final states produced in @xmath2 , @xmath3 , hadron - hadron and heavy - ion interactions @xcite .
the bosons studied are mainly charged pions @xcite . only rarely have neutral pions been studied @xcite , and never before in @xmath2 interactions . in this letter
, we report a study of bec of @xmath4 pairs in hadronic decays of the @xmath0 boson at lep , and compare them to bec of pairs of identically charged pions .
bec manifest themselves as an enhanced production of pairs of identical bosons which are close to one another in phase space .
this can be studied in terms of the two - particle correlation function @xmath5 @xcite : @xmath6 where @xmath7 is the two - particle number density for identical bosons with four - momenta @xmath8 and @xmath9 , subject to bose - einstein symmetry .
the reference distribution , @xmath10 , is the same density in the absence of bose - einstein symmetry . assuming a static spherical boson source with a gaussian density and a plane wave description of the bosons , @xmath11 is written as @xcite : @xmath12 where @xmath13 is the square of the four - momentum difference .
the parameter @xmath14 can be interpreted as the size of the boson source in the centre - of - mass system of the boson pair and a measurement of the correlation function @xmath5 gives access to the source size .
the parameter @xmath15 is introduced to describe the fraction of effectively interfering pion pairs . in this analysis
the normalization factor @xmath16 is added .
it takes into account possible long - range momentum correlations , as well as possible differences in pion multiplicity in the data and reference samples , over the four - momentum difference range studied .
the spherical shape of the boson source assumed here is a simplified picture .
high statistics charged pion data at lep revealed the source to be elongated @xcite .
the present measurement has , however , no sensitivity to the shape of the source of neutral pions , due to its limited statistics .
several theoretical predictions exist for differences in bec for pairs of bosons in the pion isospin triplet ( @xmath17 , @xmath18 , @xmath4 ) . from the string model @xcite a smaller spatial emission region , _
i.e. _ a wider momentum correlation distribution is expected for @xmath4@xmath4 , than for @xmath19@xmath19 .
this follows from the break - up of the string into @xmath20 pairs , which forbids two equally charged pions to lie next to each other on the string , whereas two neutral pions can .
the same effect is found when the probabilistic string break - up rule is interpreted as the square of a quantum mechanical amplitude @xcite . from a quantum statistical approach to bose - einstein symmetry @xcite , a small difference between @xmath19@xmath19 and @xmath4@xmath4 correlation is expected .
the size and shape of this difference is predicted to be similar to an expected bose - einstein correlation of @xmath19@xmath21 pairs .
it is theoretically uncertain whether bec between unlike sign pions are also to be expected on the ground of isospin invariance @xcite .
the main purpose of this letter is to measure the difference in size of the emission region of neutral and charged pions . in order to minimize systematic uncertainties on this difference
, the procedures followed for the charged pions are kept as close as possible to those for the neutral pions .
for this analysis , data collected with the l3 detector @xcite during the 1994 and 1995 lep runs are used .
the analysis is based mainly on measurements from the high resolution electromagnetic calorimeter and from the central tracking device .
the data sample corresponds to an integrated luminosity of about @xmath22 at centre - of - mass energies around @xmath23 . from this sample ,
about 2 million hadronic events are selected , using energy deposits in the electromagnetic and hadronic calorimeters @xcite . the jetset generator @xcite is used to study the detector response to hadronic events
. parameters of the generator are tuned to give a good description of event and jet shapes of hadronic events measured in l3 .
the effects of bose - einstein symmetry are simulated with the ` luboei ` routine @xcite .
the routine has two parameters , which have been chosen to obtain a reasonable description of l3 data .
this _ ad - hoc _ model shifts boson momenta after the hadronization phase in such a way that the correlation function @xmath24 for identical bosons is proportional to a constant plus a gaussian as in equation [ eq : c_gausexp ] .
the generated events are passed through a full detector simulation @xcite and are reconstructed and subjected to the same analysis procedure as the data .
this monte carlo sample ( jetset - be ) contains about 7 million events . a control sample ( jetset - nobe ) is also generated with jetset but it has ` luboei ` switched off and the generator parameters retuned .
significant differences are found in the tuned parameters in the two cases .
the number of events in the control sample is approximately 2 million .
unless stated otherwise , the jetset - be sample is used throughout this letter .
neutral pions in hadronic events are reconstructed from photon pairs .
photon candidates are identified in the electromagnetic calorimeter as a cluster of at least two adjacent crystals .
the clusters are required to be located in the central region of the detector , @xmath25 , and to be in the energy range @xmath26 .
above 6 , the two photons from a @xmath4 decay can no longer be distinguished as two separate clusters .
discrimination of clusters originating from photons or electrons from those due to other particles is based on the distribution of the energy over the crystals of the cluster .
a good photon discrimination is achieved with a neural network based on this energy distribution @xcite . to reject clusters due to charged particles ,
a minimum distance between the cluster and the extrapolation of the closest track of 30 is required .
this corresponds to 1.5 times the size of the front face of a crystal .
pairs of photon candidates within an event are used to reconstruct @xmath4 s .
the distribution of invariant mass of photon pairs shows a peak around the @xmath4 mass , above a smooth background .
these two components are extracted by a fit to the mass spectrum .
the background is described by a chebyshev polynomial of third order , @xmath27 .
the @xmath4 peak is parameterized by a gaussian function with exponential tails , which is continuous , and smooth in the first derivative : @xmath28 here , @xmath29 is the two - photon invariant mass , @xmath30 indicates the peak position , and @xmath31 is the width of the gaussian peak .
the parameters @xmath32 and @xmath33 determine the values of @xmath29 where the gaussian changes into an exponential .
the exponential tail at low mass is needed to describe the presence of converted photons in cluster pairs .
if a photon converts , _
e.g. _ , in the outer wall of the tracker , it can still be selected as a photon , although some of the original photon energy is lost .
the high mass tail accounts for an overestimate of the cluster energy due to another cluster nearby .
an example of the fit result is given in figure [ fig : minv1d ] .
the sum of the two functions @xmath27 and @xmath34 describes the distribution well .
also , the shape and the size of the @xmath4 peak in the data and monte carlo agree .
the mass resolution as determined from the fit is about @xmath35 . the photon pair is then selected as a @xmath4 candidate if it has an energy @xmath36 . for a mass window @xmath37 ,
a total of @xmath38 million @xmath4 s is selected in data .
the @xmath4 purity of the candidate sample is of the order of @xmath39 and the @xmath4 selection efficiency is approximately @xmath40 . about half of the background
is combinatorial , _ i.e. _ , photons from different @xmath4 s .
the other half consists of pairs where one or both of the photons do not come from a @xmath4 decay .
charged pions are detected as tracks in the central tracker .
they are selected in the same kinematic range as neutral pions : @xmath41 and @xmath42 , where the energy is calculated from the track momentum , assuming the @xmath19 mass . in addition , at least 35 out of 62 possible wire hits are required in the track fit , and the number of wires between the first and the last must be at least 50 . furthermore , the track must have at least one hit in the inner part of the tracker , and the distance of closest approach to the @xmath2 vertex in the plane transverse to the beam is required to be less than 5 .
finally a high resolution measurement of the polar angle is demanded .
charged pions are analyzed in the 1995 data only , in which @xmath43 million tracks are selected .
after the neutral pion selection , @xmath4 candidates within an event are paired , requiring that no cluster is common to the two candidates , and their four - momentum difference @xmath44 is calculated .
the @xmath4@xmath4 component of the @xmath44 distribution is estimated by a fit to the two - dimensional mass spectrum for every bin in @xmath44 .
an example of these mass distributions is shown in figure [ fig : minv2d]a , for the bin @xmath45 .
the various contributions are clearly visible : non-@xmath4 pairs give the smooth background , @xmath4 with non-@xmath4 pairs give the two `` ridges '' in the @xmath4 peak regions , @xmath4 pairs give part of the peak in the centre of the plot , the other part being caused by the sum of the @xmath4 with non-@xmath4 ridges .
this two - dimensional distribution is derived from the product of two one - dimensional mass distributions @xcite : @xmath46 \\ & + { \ensuremath{a_{\text{bgbg}}{\ensuremath{f_{\text{bgbg}}}}}}{(m_1,m_2 ) } , \label{eq : f2d } \end{split}\ ] ] where the first term describes the @xmath4@xmath4 part , the second term describes the @xmath4 with non-@xmath4 pair ridges and the third term is the non-@xmath4 pair background .
the number of @xmath4 pairs in the mass window @xmath47 is related to the parameter @xmath48 .
the functions @xmath34 and @xmath27 have the same functional form as described before , @xmath49 follows from the product of two chebyshev polynomials of third order and is required to be symmetric in the two masses .
all 18 parameters of @xmath50 are left free in the fit .
the result of a binned maximum likelihood fit for the mass distribution in figure [ fig : minv2d]a is shown in figures [ fig : minv2d]b d . in this representative example , the @xmath51 is 4737 for 4606 degrees of freedom , which corresponds to a @xmath52 confidence level . figure [ fig : pionpair_q]a presents the @xmath44 distribution for @xmath4 pairs in the mass window @xmath47 , obtained from the values of the parameter @xmath48 from two - dimensional mass fits of equation [ eq : f2d ] to both data and monte carlo .
some deviations between data and monte carlo are caused by the imperfect modeling of bec .
the efficiency to select a @xmath4 pair in an event ranges from about @xmath53 at @xmath54 to @xmath55 at @xmath56 .
bins in @xmath44 below 300 are not used for the rest of the analysis for two reasons .
first , the efficiency estimate depends strongly on the bec modeling in the generator , in the region of small @xmath44 .
this occurs because the bec modeling moves identical pions closer together , which lowers the detection efficiency .
secondly , the four - momentum difference of any pair of @xmath4 s from @xmath57 decays is kinematically constrained to have @xmath58 , and in that @xmath44-range , more than @xmath59 of all @xmath4 pairs originate from this decay .
a rejection of the small @xmath44 region thus avoids systematic uncertainties due to the simulation of the @xmath60 multiplicity .
the distribution of four - momentum difference of equally charged pion pairs is obtained by calculating @xmath44 for pairs of tracks selected within an event and with the same charge .
this raw @xmath19@xmath19 spectrum is corrected bin - by - bin for both pion purity and efficiency using the monte carlo simulation .
the uncorrected distribution is shown in figure [ fig : pionpair_q]b .
compared to the @xmath4@xmath4 case , smaller deviations are observed between the raw spectrum in data and monte carlo .
these deviations are due to the imperfect modeling of bec in the monte carlo .
to obtain the final correlation function @xmath24 , the @xmath44 distribution of @xmath4 pairs in the data , figure [ fig : pionpair_q]a , is corrected for selection efficiencies .
the efficiency is defined as the number of selected @xmath4 pairs in monte carlo events ( jetset - be ) divided by the number of generated @xmath4 pairs in the same events , where the generated pions have to be in the same kinematic range as the selected pions .
this definition includes an acceptance correction for those @xmath4 s which can not be selected kinematically . in this way
, the correlations of @xmath4 pairs can directly be compared to those of charged pion pairs .
the reference distribution @xmath61 is calculated from a jetset - nobe sample at generator level , where pions are taken in the same kinematic range as in the definition of the selection efficiency .
we choose this reference distribution rather than the distribution for @xmath19@xmath21 because of the uncertainty concerning bec between unlike sign pions mentioned in the introduction .
the correlation distribution @xmath24 is then the ratio of the corrected data spectrum to the reference spectrum .
the distribution of @xmath24 is displayed in figure [ fig : pion_be]a .
an enhancement at low @xmath44 values , expected from bose - einstein symmetry , is clearly visible .
the function @xmath24 from equation [ eq : c_gausexp ] is fitted to this ratio in the interval @xmath62 . extending the fit to lower values of @xmath44 results in consistent values of the parameters but with much larger systematic uncertainties .
the overall normalization @xmath63 is determined from the integrals of @xmath24 and the fit function ; the only free parameters are @xmath15 , @xmath14 and @xmath32 . in this fit
, the @xmath51 is 46.1 for 40 degrees of freedom , corresponding to a @xmath64 confidence level .
the systematic uncertainty on the result due to the @xmath4 selection is determined by varying the photon selection cuts and by changing the size of the @xmath4 mass window by @xmath65 .
the @xmath4@xmath4 mass fit of equation [ eq : f2d ] is tested by varying the fit range by @xmath66 .
the uncertainty due to the modeling of bose - einstein correlations in the monte carlo generator is taken into account by using the control sample jetset - nobe in the efficiency correction procedure .
in addition , the influence on the final result of the agreement between data and monte carlo of distributions relevant to the photon and @xmath4 selection , such as neural network output and energy and polar angle of @xmath4 s , is studied .
finally , the binning in @xmath44 is varied . the systematic uncertainty on the result due to each of the sources ,
is assigned as half the maximum deviation .
a summary is given in table [ tab : pi0syst ] .
the total systematic uncertainty is calculated as the quadratic sum of these uncertainties .
the final correlation function @xmath24 for @xmath19@xmath19 is calculated in a similar way as that for @xmath4@xmath4 .
the @xmath19@xmath19 data distribution , figure [ fig : pionpair_q]b , is corrected for purity and efficiency . as for @xmath4@xmath4 ,
the efficiency is calculated for generated pions in the same kinematic range as the selected pions .
the reference distribution @xmath61 and the correlation distribution @xmath24 are obtained in the same way as for @xmath4@xmath4 .
the correlation function for @xmath19@xmath19 is shown in figure [ fig : pion_be]b . due to the higher selection efficiency for charged pions as compared to neutral pions , the significance of the low @xmath44 value enhancement is much larger . as for @xmath4@xmath4 , the function defined in equation [ eq : c_gausexp ]
is fitted to the final distribution . in this fit
, the @xmath51 is 42.6 for 40 degrees of freedom , corresponding to a @xmath67 confidence level .
the systematic uncertainty on the result due to the track selection is determined by varying the requirements on number of hits , distance of closest approach and polar angle determination . as in the @xmath4@xmath4 case
, the uncertainty on the modeling of bose - einstein correlations in the monte carlo generator is obtained by using the control sample jetset - nobe in the analysis .
finally , the binning in @xmath44 is varied .
the systematic uncertainties are attributed as in the @xmath4@xmath4 case , and are summarized in table [ tab : cpisyst ] . the total systematic uncertainty is calculated as the quadratic sum of these uncertainties .
the final values for the strength of the correlation @xmath15 and the corresponding radii of the boson sources @xmath14 are given in table [ tab : final ] .
due to the lower efficiency of the @xmath4@xmath4 selection , the statistical uncertainty on the @xmath4@xmath4 result is larger than the statistical uncertainty on the @xmath19@xmath19 result . within these uncertainties ,
the data indicate both a weaker correlation and a smaller source radius for @xmath4@xmath4 .
the weakness of the @xmath4@xmath4 correlation can be partly explained by the bigger contribution of resonance decays to the @xmath44 spectrum .
the difference of the source sizes is @xmath68 where @xmath69 and @xmath70 indicate the value of @xmath14 for @xmath19@xmath19 and @xmath4@xmath4 , respectively . in this difference
, the systematic uncertainties due to the modeling of bose - einstein correlations and the binning in @xmath44 are taken to be correlated between the two samples .
the smaller radius found for @xmath4@xmath4 is in qualitative agreement with the predictions of the string model .
we wish to express our gratitude to the cern accelerator divisions for the excellent performance of the lep machine .
we acknowledge the contributions of the engineers and technicians who have participated in the construction and maintenance of this experiment .
g. goldhaber _
et al . _ ,
. lett . * 3 * ( 1959 ) 181 d.h .
boal , c.k .
gelbke , b.k .
jennings , rev .
* 62 * ( 1990 ) 553 g. baym , acta phys .
pol . * b 29 * ( 1998 ) 1839 , m. acciarri _
et al . _ ,
* b 458 * ( 1999 ) 517 , d. decamp _ et al . _ , z. phys . *
c 54 * ( 1992 ) 75 ; delphi collaboration , p. abreu _
et al . _ ,
* b 286 * ( 1992 ) 201 ; opal collaboration , g. alexander _
_ , z. phys .
* c 72 * ( 1996 ) 389 k. eskreys , acta phys . pol .
* 36 * ( 1969 ) 237 v.g .
et al . _ ,
47 * ( 1988 ) 278 g. goldhaber , s. goldhaber , w. lee , a. pais , phys . rev .
* 120 * ( 1960 ) 300 m. deutschmann _
et al . _ ,
* b 204 * ( 1982 ) 333 , p. abreu _ et al . _ , phys . lett . *
b 471 * ( 2000 ) 460 ; opal collaboration , g. abbiendi _
et al . _
, e. phys .
j. * c 16 * ( 2000 ) 423 b. andersson _
et al . _ ,
phys . rep .
* 97 * ( 1983 ) 31 b. andersson , w. hofmann , phys
* b 169 * ( 1986 ) 364 b. andersson , m. ringnr , nucl .
* b 513 * ( 1998 ) 627 i.v .
andreev , m. plmer , r.m .
weiner , phys .
* 67 * ( 1991 ) 3475 m. suzuki , phys
d 35 * ( 1987 ) 3359 ; m.g . bowler , phys .
* b 197 * ( 1987 ) 443 ; g. alexander , h.j .
lipkin , phys . lett .
* b 456 * ( 1999 ) 270 , b. adeva _ et al . _ , nucl .
. meth . * a 289 * ( 1990 ) 35 ; m. acciarri _
et al . _ ,
. meth . * a 351 * ( 1994 ) 300 ; a. adam _
et al . _ ,
. meth . * a 383 * ( 1996 ) 342 ; i.c .
. meth . * a 381 * ( 1996 ) 236 ; m. chemarin _
et al . _ ,
inst . meth . *
a 349 * ( 1994 ) 345 , m. acciarri _
et al . _
, e. phys .
j. * c 16 * ( 2000 ) 1 t. sjstrand , comp .
* 82 * ( 1994 ) 74 l. lnnblad , t. sjstrand , phys .
* b 351 * ( 1995 ) 293 the l3 detector simulation is based on geant version 3.15 .
+ see r. brun _
et al . _
, geant 3 , cern dd / ee/84 - 1 ( 1984 ) , revised 1987 .
+ the gheisha program ( h. fesefeldt , rwth aachen report pitha 85/02 ( 1985 ) ) is used to simulate hadronic interactions m.p .
sanders , ph.d .
thesis , university of nijmegen , 2002 p.achard o.adriani m.aguilar-benitez j.alcaraz g.alemannij.allabya.aloisio m.g.alviggih.anderhub v.p.andreevf.anselmoa.arefiev t.azemoon t.aziz m.baarmandp.bagnaiaa.bajo g.baksayl.baksays.v.baldew s.banerjee sw.banerjee a.barczyk r.barillre p.bartalini m.basilen.batalovar.battistona.bay f.becattiniu.beckerf.behnerl.bellucci r.berbeco j.berdugo p.berges b.bertuccib.l.betevm.biasinim.bigliettia.biland j.j.blaising s.c.blyth g.j.bobbink a.bhml.boldizsarb.borgia s.bottaid.bourilkovm.bourquins.braccinij.g.bransonf.brochu a.buijsj.d.burgerw.j.burgerx.d.cai m.capellg.cara romeog.carlinoa.cartacci j.casausf.cavallarin.cavallo c.cecchi m.cerradam.chamizoy.h.chang m.chemarina.chen g.chen g.m.chen h.f.chen h.s.cheng.chiefari l.cifarellif.cindoloi.clarer.clare g.coignet n.colino s.costantini b.de la cruzs.cucciarelli t.s.dai j.a.van dalen r.de asmundisp.dglon j.debreczenia.degr k.deiters d.della volpe e.delmeire p.denes f.denotaristefania.de salvo m.diemoz m.dierckxsens d.van dierendonckc.dionisi m.dittmara.doriam.t.dovad.duchesneau p.duinker b.echenarda.elineh.el mamounia.engler f.j.eppling a.ewersp.extermann m.a.falagans.falcianoa.favaraj.fay o.fedinm.felcinit.ferguson h.fesefeldt e.fiandrinij.h.field f.filthautp.h.fisherw.fisheri.fiskg.forconi k.freudenreichc.furettayu.galaktionovs.n.ganguli p.garcia-abiam.gataullins.gentiles.giaguz.f.gongg.grenier o.grimm m.w.gruenewald m.guida r.van gulikv.k.gupta a.gurtul.j.gutayd.haasd.hatzifotiadout.hebbekera.herv j.hirschfelderh.hofer g. holzner s.r.houy.hu b.n.jin l.w.jonesp.de jongi.josa-mutuberrad.kferm.kaurm.n.kienzle-focaccij.k.kimj.kirkbyw.kittela.klimentov a.c.knigm.kopalv.koutsenko m.krber r.w.kraemerw.krenz a.krger a.kunin p.ladron de guevarai.laktinehg.landim.lebeaua.lebedevp.lebrunp.lecomte p.lecoq p.le coultre h.j.leej.m.le goffr.leiste p.levtchenkoc.li s.likhoded c.h.linw.t.linf.l.lindel.listaz.a.liuw.lohmanne.longo y.s.lu k.lbelsmeyerc.luci d.luckeyl.luminariw.lustermannw.g.ma l.malgeria.malinin c.maad.mangeolj.mans
j.p.martin f.marzano k.mazumdarr.r.mcneil s.melel.merola m.meschini w.j.metzgera.mihulh.milcentg.mirabelli j.mnichg.b.mohanty g.s.muanzaa.j.m.muijsb.musicar m.musy s.nagym.napolitanof.nessi-tedaldih.newman t.niessena.nisatih.nowak r.ofierzynski g.organtinic.palomaresd.pandoulas p.paoluccir.paramatti g.passalevas.patricelli t.paulm.pauluzzic.pausf.paussm.pedaces.pensottid.perret-gallix b.petersend.piccolo f.pierella m.pioppip.a.pirou e.pistolesiv.plyaskin m.pohl v.pojidaevh.postemaj.pothierd.o.prokofiev d.prokofiev j.quartierig.rahal-callotm.a.rahaman p.raics n.rajar.ramelli p.g.rancoitar.ranieri a.raspereza p.razisd.ren m.rescignos.reucrofts.riemannk.rilesb.p.roel.romero a.rosca s.rosier-leess.rothc.rosenbleckb.rouxj.a.rubio g.ruggiero h.rykaczewski a.sakharovs.saremi s.sarkarj.salicio e.sanchezm.p.sandersc.schferv.schegelskys.schmidt-kaerstd.schmitz h.schopperd.j.schotanusg.schwering c.sciaccal.servolis.shevchenkon.shivarovv.shoutko e.shumilov a.shvorobt.siedenburgd.sonp.spillantini m.steuerd.p.stickland b.stoyanova.straessnerk.sudhakarg.sultanovl.z.suns.sushkovh.suter j.d.swainz.szillasix.w.tangp.tarjanl.tauscherl.taylorb.tellili d.teyssier c.timmermanssamuel c.c.ting s.m.ting s.c.tonwar j.tth c.tullyk.l.tungy.uchidaj.ulbricht e.valente r.t.van de wallev.veszpremig.vesztergombii.vetlitsky d.vicinanza g.viertel s.villam.vivargent s.vlachosi.vodopianov h.vogelh.vogt i.vorobiev a.a.vorobyov m.wadhwaw.wallraff m.wangx.l.wang z.m.wangm.weberp.wienemannh.wilkenss.x.wus.wynhoff l.xia z.z.xu j.yamamoto b.z.yang c.g.yang h.j.yangm.yangs.c.yeh an.zaliteyu.zalitez.p.zhang j.zhaog.y.zhur.y.zhuh.l.zhuanga.zichichig.zilizib.zimmermann m.zller .
pair candidates with a four - momentum difference in the range @xmath77 , in bins of @xmath78 .
result of the fit for ( b ) @xmath4 pairs , ( c ) @xmath4 with non-@xmath4 pairs , ( d ) non-@xmath4 pairs .
[ fig : minv2d ] ] for ( a ) @xmath4@xmath4 and ( b ) @xmath19@xmath19 , and results of the fits .
the points indicate the data , the full line corresponds to the fit result and the dashed line is the normalization factor @xmath79 .
[ fig : pion_be ] ] | bose - einstein correlations of both neutral and like - sign charged pion pairs are measured in a sample of 2 million hadronic @xmath0 decays collected with the l3 detector at lep .
the analysis is performed in the four - momentum difference range @xmath1 .
the radius of the neutral pion source is found to be smaller than that of charged pions .
this result is in qualitative agreement with the string fragmentation model . |
one - dimensional ( 1d ) waveguide - qed systems are currently generating increasing interest systems in which photons confined in one - dimension interact with one or several two - level systems ( qubits ) .
part of the motivation comes from the striking quantum optics effects that can be seen in these strongly coupled systems @xcite .
another motivating factor is the promise of waveguide - qed systems for quantum information processing @xcite .
finally , a key driver of the interest in waveguide - qed systems is the tremendous experimental progress that has been made recently in a number of systems @xcite .
perhaps the leading system for waveguide - qed investigations and applications is an open microwave transmission line coupled to superconducting qubits @xcite .
while much of the work to date has focused on systems in which there is a single qubit , and there is a growing literature on the case of two qubits @xcite , an important future direction for both fundamental effects and possible applications is to study a waveguide coupled to multiple ( or many ) qubits . as a step in this direction , here
we compare and contrast results for one , two , and three qubits coupled to a waveguide [ see fig.[fig : schematiccomparison](a ) ] , focusing in particular on the generation of photon - photon correlations .
correlations between photons are a key signature of non - classical light .
they are often characterized by the second - order correlation function ( photon - photon correlation function ) @xmath1 where @xmath2 is the observation time between the two photons ( see below for precise definition ) @xcite .
the uncorrelated , classical value is @xmath3 ( obtained , for example , for a coherent state ) .
bunching of photons , @xmath4 , often occurs due to the bosonic nature of photons but anti - bunching , @xmath5 , also occurs @xcite . in recent experiments ,
@xmath1 of microwave photons coupled to superconducting qubits was measured , and both bunching and anti - bunching were observed @xcite . in a multi - qubit situation , one expects to have interference between the various scattered partial waves ; interference effects in the photon - photon correlations @xmath1 are known as `` quantum beats '' @xcite . in this paper , we first present our method of calculation , which exploits a bosonic representation of the qubits in the rotating wave approximation . we obtain a complicated yet analytic result for @xmath6 in the markovian limit and show , by comparison with the full numerical result , that it is adequate for small , experimentally accessible separations between the qubits . in presenting results , we focus on an off - resonant case in which single photons have equal probability of being transmitted or reflected , and take the separation between qubits , denoted @xmath0 , to be either @xmath7 or @xmath8 where @xmath9 is the wavelength of a photon at the qubit resonant frequency .
we find several striking features in @xmath6 : first , for @xmath10 , the transmitted photons are largely bunched for all times and become more strongly bunched as the number of qubits increases , while the reflected photons oscillate between strong bunching and anti - bunching , showing particularly strong quantum beats in the three qubit case .
second , for @xmath11 , we find the surprising situation that both transmitted and reflected photons are bunched at @xmath12 but then become anti - bunched for a large time interval .
this suggests that the photons in this case become organized into bursts .
the hamiltonian describing @xmath13 identical qubits coupled to a 1d waveguide [ see fig . [
fig : schematiccomparison](a ) ] is , in the rotating wave approximation , @xmath14\nonumber\\ & \,\,+\sum_{i=1}^n\sum_{\alpha=\text{l , r } } \hbar v \int dx\ ; \delta(x - l_i)\left[a^\dagger_\alpha(x)\sigma^-_i+a_\alpha(x)\sigma^+_i\right],\end{aligned}\ ] ] where @xmath15 are the raising / lowering operators for @xmath16-th qubit , @xmath17 is its position which is fixed by @xmath18 , @xmath19 is the transition frequency of the qubit , and @xmath20 is the decay rate to channels other than the waveguide . the spontaneous decay rate to the waveguide continuum is given by @xmath21 . in the waveguide qed context , `` strong coupling '' signifies that the spontaneous decay rate to the waveguide is much faster than the decay to all other modes , namely that the purcell factor is large , @xmath22 . to find @xmath1
, we first obtain the two - photon eigenstate of @xmath23 . as discussed in ref .
@xcite , it is convenient to use a bosonic representation of the qubits that includes an on - site interaction , @xmath24 the raising / lowering operators @xmath15 in @xmath23 are replaced by the bosonic creation / annihilation operators @xmath25 and @xmath26 , respectively .
one then takes @xmath27 in the end to project out occupations greater than @xmath28 . in this bosonic representation
, the @xmath29 case corresponds to a non - interacting hamiltonian and can readily be solved . in terms of the non - interacting wavefunctions and green functions ,
a formal expression for the two - photon `` interacting '' wavefunction in the @xmath27 limit can be obtained ; this then is the solution to the waveguide qed problem in which we are interested . finally , the two - photon wavefunction together with the one - photon wavefunction yields @xmath1 for a weak incident coherent state .
more details of this procedure are given in the appendices .
the markovian approximation allows a considerable simplification of the final result @xcite . in the present context ,
the markovian approximation consists of an approximate treatment of certain interference terms valid for small separation between the qubits . in the formal expression for the two - photon wavefunction discussed above
, there is an integral over the non - interacting wavefunctions which generally must be performed numerically .
the non - interacting wavefunctions naturally involve interference factors @xmath30 that make this integral difficult . however , if the qubits are close enough , @xmath31 may be replaced by @xmath32 , allowing the integral to be performed analytically using contour integration ( the analytic expression of the final result is lengthy , so we just give the steps of the derivation in the appendices as well as the @xmath33 result as an example ) .
all of the results in this paper are obtained in the regime where this is valid .
an example of the checks we have made is shown in fig .
[ fig : schematiccomparison](b ) : the full numerical result is in good agreement with that from the small separation approximation .
we compare the one , two , and three qubit cases : @xmath34 , @xmath35 , or @xmath36 . in order to make a fair comparison
, the typical transmission through the system in the three cases should be the same ; otherwise , the lower probability of finding a photon in one case compared to another will affect @xmath37 .
we therefore consider off - resonance cases ( i.e. @xmath38 where @xmath39 is the incoming photon frequency ) in which the single - photon transmission probability @xmath40 is fixed .
because the single - photon transmission spectrum depends on the number of qubits , the frequency used is different in the three cases @xmath34@xmath36 . due to the asymmetry of the single - photon transmission spectrum in certain cases , the criterion used throughout
this work is to pick up the frequency closest to @xmath19 so that @xmath41 is the largest . in the following results , we consider @xmath34 , @xmath35 , or @xmath36 ; @xmath42 or @xmath43 ; and @xmath44 .
the single photon transmission curves used to choose the photon frequency @xmath39 are shown in fig.[fig : transmission - spectrum ] .
we use @xmath45 as our unit of frequency , take @xmath46 , and consider the lossless case , @xmath47 .
the results for a single qubit , shown in fig.[fig : piover2-t50 ] panels ( a ) and ( d ) , provide a point of comparison for the two and three qubit cases discussed below ; throughout we consider the response to an incident weak coherent state .
non - classical light in a waveguide produced by a single qubit has been extensively investigated theoretically @xcite as well as experimentally with microwave photons @xcite .
we see that for our chosen detuning such that @xmath44 , the transmitted field shows bunching while the reflected field is anti - bunched .
the correlation decays to its classical value ( namely , @xmath28 ) quickly and with little structure .
for this reason the single value @xmath41 is a good indication of the nature of the correlations overall .
note that in panel ( d ) , @xmath48 due to the inability of a single excited qubit to release two photons at the same time . for @xmath33 or @xmath36 , we start by considering the case @xmath49 , in which case the qubits are separated by @xmath7 ; the results are shown in fig.[fig : piover2-t50 ] . the presence of quantum beats coming from interference among the partial waves scattered by the qubits is clear , especially for three qubits . in the transmitted wave ,
photon bunching is considerably enhanced in magnitude and extends for a longer time ( compared to a single qubit ) . in reflection , @xmath1 develops a striking oscillation between strongly bunched and anti - bunched [ panel ( f ) ] .
such behavior in @xmath37 suggests that the photons become organized periodically in time and space .
turning now to the case @xmath50 , we see in fig.[fig : piover4-t50 ] that the behavior is completely different .
first , the quantum beats largely disappear in both transmission and reflection .
instead , for @xmath51 we see that both the reflected and transmitted photons are initially bunched , in the reflected case quite strongly bunched .
the initial bunching is followed in both cases by anti - bunching .
this anti - bunching is dramatic for the transmitted photons : strong anti - bunching persists for a time interval of several tens of @xmath52 ( the natural unit of time in our problem ) .
initial bunching followed by a long interval of anti - bunching suggests that the photons are organized into bursts .
the different behavior for @xmath50 compared to @xmath49 can be traced to a difference in the structure of the poles of the single photon green function ( see , e.g. , the discussion in ref .
for instance in the @xmath33 cases , for @xmath49 there are two dominant poles that have the same decay rate but different real frequencies , leading to maximum interference effects between those two processes .
in contrast , for @xmath50 , the poles have very different decay rates ; the one decaying most rapidly yields the sharp initial bunching , while the one with the slowest decay produces the long time anti - bunching . to study how the correlations depend on the frequency of the photons , we show the initial correlation , @xmath41 , in fig.[fig : g2(0 ) ] and [ fig : g2(0 ) vs frequency ] .
because of the oscillating structure in @xmath1 when there are multiple qubits , @xmath41 is not necessarily a good indication of the behavior at later times ; nevertheless , the degree of initial bunching or anti - bunching is a physically important and measurable quantity . for a fair comparison between the @xmath34 , @xmath35 , and @xmath36 cases , we first plot @xmath41 as a function of the single photon transmission , @xmath40 ; see fig .
[ fig : g2(0 ) ] . to match the desired @xmath40 with an off - resonant photon frequency
we follow the following procedure : starting near the resonant frequency @xmath19 ( where @xmath53 ) , we scan toward smaller frequencies until @xmath54 is reached .
we then use frequencies within the scanned range to calculate @xmath41 for both transmission and reflection as a function of @xmath40 at @xmath49 and @xmath55 .
another way of presenting the data is to simply plot @xmath41 directly as a function of frequency , as in fig .
[ fig : g2(0 ) vs frequency ] . by comparing with fig .
[ fig : transmission - spectrum ] , we see that the method above for selecting the range of frequencies to use in making fig .
[ fig : g2(0 ) ] selects the range with largest @xmath41 for a given value of @xmath40 .
finally , note that for reflection from one qubit , @xmath48 in all cases , as mentioned above , and so is not plotted in panels ( b ) and ( d ) of both figures .
several general trends are clear from fig.[fig : g2(0 ) ] .
bunching is favored over anti - bunching for both @xmath33 and @xmath36 . as the single photon transmission increases
, @xmath41 decreases for transmission but generally increases for reflection .
opposite trends for transmission and reflection are natural based on the simple argument that incoming uncorrelated photons divide between transmitted and reflected ones so that bunching in one implies anti - bunching in the other .
clearly , this simple argument does not apply here ; indeed , it is striking and surprising that for a broad range of parameters _ both transmitted and reflected photons are bunched .
_ trends as the number of qubits increases from @xmath28 to @xmath36 are also evident in fig .
[ fig : g2(0 ) ] . in panels ( a ) and ( d )
the trend is monotonic : for @xmath50 , the reflected photons become tremendously bunched [ panel ( d ) ] , whereas for @xmath49 and transmitted photons [ panel ( a ) ] , the curves cross at the same point indicating that the trend changes sign increasing bunching as @xmath13 increases for @xmath56 but decreasing bunching for smaller @xmath40 . in the other two cases , panels ( b ) and ( c ) ,
the trend as @xmath13 increases from @xmath28 to @xmath36 is not monotonic . for @xmath49 and @xmath57 ,
the reflected photons switch from being anti - bunched to bunched to anti - bunched as @xmath13 changes from @xmath28 to @xmath36 , but show increasing bunching for larger @xmath40 [ panel ( b ) ] .
finally , in panel ( c ) [ @xmath50 and transmitted photons ] , there is a monotonic trend toward less bunching for @xmath58 but non - monotonic behavior for larger transmission . from the explicit dependence on frequency shown in fig .
[ fig : g2(0 ) vs frequency ] , we see that bunching is generally favored even outside the frequency range chosen in fig .
[ fig : g2(0 ) ] ( which in the @xmath50 case is quite small ( @xmath59 ) ) . comparing to the single photon transmission spectrum
( see fig .
[ fig : transmission - spectrum ] ) , we point out two features : first , in the @xmath50 case , @xmath41 shows the asymmetry with respect to @xmath19 [ panel ( c ) and ( d ) ] seen in @xmath60 ; this again can be traced to the asymmetric pole structure of the green functions mentioned above .
@xmath41 is larger ( for reflection ) and varies more rapidly on the red - detuned side ( @xmath61 ) , which explains why we chose the frequency range use in fig .
[ fig : g2(0 ) ] .
in fact , on the blue - detuned side ( @xmath62 ) the structure in @xmath1 is less dramatic , and it returns to @xmath28 faster ( data not shown ) .
second , as also shown in fig .
[ fig : g2(0 ) ] , the peaks of @xmath41 for transmission are locatd where @xmath63 , while the peaks of reflected @xmath41 are located where @xmath64 .
note that the leftmost peak of @xmath51 in fig .
[ fig : g2(0 ) vs frequency](d ) is completely due to the small denominator ( @xmath65 ) at that point .
in this work , we have calculated the second - order correlation function , @xmath1 , for photons in a one - dimensional waveguide interacting with one , two or three qubits . by taking the separation between the qubits small , we are able to make a markovian approximation which then allows an analytic solution .
the small separation and small @xmath13 on which we focus means that these systems are within the range of current experimental capability @xcite .
the interference among the partial waves scattered from the qubits leads to a variety of behavior in @xmath37 that is sensitive to both the separation between the qubits ( @xmath0 ) and the frequency of the incoming photons . as examples of the rich variety accessible in these waveguide qed structures , we mention three here in conclusion : ( i ) for a wide range of parameters , both transmitted and reflected photons are initially bunched .
( ii ) for reflected photons with @xmath51 and @xmath49 , @xmath1 oscillates between bunching and anti - bunching [ fig.[fig : piover2-t50](f ) ] .
( iii ) for transmitted photons with @xmath51 and @xmath50 , initial strong bunching is followed by a long ( i.e. @xmath66 ) interval of antibunching [ fig.[fig : piover4-t50](c ) ] .
these last two observations suggest that some nascent organization of the photons may be occurring , providing an interesting direction for future research .
this work was supported by us nsfgrantno .
phy-10 - 68698 .
is supported by a john t. chambers fellowship from the fitzpatrick institute for photonics at duke university .
the single photon eigenstate @xmath67 with @xmath68 is by definition the eigenstate of @xmath23 , i.e. , @xmath69 , where @xmath70|0\rangle,\\ \phi_\text{r}^\text{r}(k , x ) & = \frac{e^{ikx}}{\sqrt{2\pi}}\left(\theta(l_1-x)+\sum_{i=1}^{n-1 } t_i(k)\theta(x - l_i)\theta(l_{i+1}-x ) + t_n(k)\theta(x - l_n)\right)\!,\\ \phi_\text{l}^\text{r}(k , x ) & = \frac{e^{-ikx}}{\sqrt{2\pi}}\left(r_1(k)\theta(l_1-x)+\sum_{i=2}^{n } r_i(k)\theta(x - l_{i-1})\theta(l_{i}-x)\right),\\ \phi_\text{r}^\text{l}(k , x ) & = \frac{e^{ikx}}{\sqrt{2\pi}}\left ( r_1(k)\theta(l_n - x)+\sum_{i=2}^{n } r_i(k)\theta(x - l_{n - i+1})\theta(l_{n - i+2}-x)\right),\\ \phi_\text{l}^\text{l}(k , x ) & = \frac{e^{-ikx}}{\sqrt{2\pi}}\bigg(\theta(x - l_n ) + \sum_{i=1}^{n-1 } t_i(k)\theta(x - l_{n - i})\theta(l_{n - i+1}-x ) + t_n(k)\theta(l_1-x)\bigg),\end{aligned}\ ] ] and the incoming photon travels in the @xmath71-direction with wavevector @xmath31 .
the single photon transmission amplitude is given by @xmath72 and the reflection amplitude by @xmath73 .
note that the positions of the qubits are chosen to be symmetric with respect to the origin , i.e. , @xmath74 , in order to take advantage of parity symmetry .
setting @xmath75 from now on , we have for @xmath33 @xcite @xmath76 r_1(k)&=\frac{\gamma \left(\gamma -e^{2 i k l } ( \gamma + 2 i k-2 i \omega_0 ) -2 i k+2 i \omega_0 \right)}{\gamma ^2 e^{3 i k l}+e^{i k l } ( i \gamma + 2 k-2 \omega_0 ) ^2}\\[1 mm ] e_1^\text{r}(k)&=-\frac{i \sqrt{\gamma } e^{-\frac{1}{2 } i k l } \left(\gamma \left(-1+e^{2 i k l}\right)+2 i k-2 i \omega_0 \right)}{\sqrt{\pi } \left((i \gamma + 2 k-2 \omega_0 ) ^2+\gamma ^2 e^{2
i k l}\right)}\\[1 mm ] e_2^\text{r}(k)&=\frac{2 \sqrt{\gamma } e^{\frac{i k l}{2 } } ( k-\omega_0 ) } { \sqrt{\pi } \left((i \gamma + 2 k-2 \omega_0 ) ^2+\gamma ^2 e^{2 i k l}\right)}.\end{aligned}\ ] ] the corresponding result for @xmath51 is @xmath77 with @xmath78 .
note that we do not need the other amplitudes for the rest of this section . for @xmath34 results
see , e.g. , ref .
@xcite .
we can now construct the two - photon `` non - interacting '' eigenstate @xmath79 as described in the supplementary material of ref .
@xcite , starting from the lippmann - schwinger equation @xmath80 where @xmath81 is given in eq .
and @xmath82 is the two photon energy , one can derive the two - photon interacting eigenstate in the coordinate representation in the @xmath83 limit : @xmath84 @xmath85 @xmath86 @xmath87 note that @xmath88 and @xmath89 here refer to the positions of the photons . by observing the structure of these green functions
, one would realize that given the following two pieces @xmath90 the whole prescription is complete and in principle one may numerically compute the two - photon interacting eigenstate eq . for any @xmath13 .
finally , to proceed with the markovian approximation , we explicitly write down the integrands in eqs . and , replace the factors @xmath91 by @xmath92 therein , and do the double integral by standard contour integral techniques enclosing the poles in the upper half complex plane ( for the @xmath51 case , for example , the denominator of each transmission amplitude is a cubic polynomial in @xmath31 , so there are three roots ) .
the @xmath33 case @xcite could serve as an illustrative example owing to its relatively simple polynomial structure : for @xmath93 with @xmath94 , we have @xmath95 \nonumber\\ & \quad \times \exp \left(i e
( t+x_1)-\frac{1}{2 } \gamma t \right)\biggr\}\biggl/8 \left(e^{2 i \pi a } \gamma ^2 \eta+ \eta^3\right)\\ g^\text{r , r}_2(x_1,x_2)&=i\biggl\{\gamma\bigl [ \beta_+ \left(-i \beta_- \gamma + e-2 \omega_0 \right ) \left(\left(-2+e^{i
\pi a}\right ) \beta_+ \gamma + 2 i e-4 i \omega_0 \right ) \nonumber\\ & \quad+ \beta_-
e^{e^{i \pi a } \gamma t } \left(-\left(e^{i \pi a}+e^{2 i \pi a}-2\right ) \gamma -2
i e+4 i \omega_0 \right ) \left(e+i \gamma\right)\bigr ] \nonumber\\ & \quad \times\exp \left(-\frac{1}{2 } \gamma t -i \pi a+i e ( t+x_1)\right)\biggr\}\biggl/8 \left(e^{2
i \pi a } \gamma ^2 \eta+ \eta^3\right),\end{aligned}\ ] ] where @xmath96 , @xmath97 , and @xmath98 . during the two contour integrations , @xmath99 and @xmath100 ( with @xmath101 )
are used . due to parity symmetry , @xmath102 , @xmath103 , @xmath104 and @xmath105 .
for a non - dispersive photonic field operator in the heisenberg picture which satisfies @xmath106 , the two - photon correlation function @xmath1 can be rewritten in the schrdinger picture as @xmath107 where @xmath108 is the asymptotic output state and @xmath109 for transmitted photons or @xmath110 for reflected photons .
the second equality holds if a weak incident coherent state ( mean photon number @xmath111 ) with right - going photons is assumed as is appropriate for comparison with an eventual experiment such that we consider only two - photon states in the numerator and one - photon states in the denominator .
the justification for the latter is twofold : ( i ) in the numerator , the 0- and 1- photon states are eliminated by the annihilation operators , leaving the 2-photon sector untouched which , then , can be described by @xmath112 . (
ii ) in the denominator , the probability of having only one photon is much larger then having two , so that the factors @xmath113 can be replaced by the single photon eigenstate @xmath114 given that @xmath115 ( i.e. two identical incident photons ) .
we are thus lead to an explicit expression for the photon - photon correlations in terms of the 1- and 2- photon states found using the method outlined in appendix a. | for a one - dimensional ( 1d ) waveguide coupled to two or three qubits , we show that the photon - photon correlations have a wide variety of behavior , with structure that depends sensitively on the frequency and on the qubit - qubit separation @xmath0 .
we study the correlations by calculating the second - order correlation function @xmath1 in which the interference among the photons multiply scattered from the qubits causes rich structure . in one case ,
for example , transmitted and reflected photons are both bunched initially , but then become strongly anti - bunched for a long time interval .
we first calculate the correlation function @xmath1 including non - markovian effects and then show that a much simpler markovian treatment , which can be solved analytically , is accurate for small qubit separation . as a result , the non - classical properties of microwaves in a 1d waveguide coupled to many superconducting qubits with experimentally accessible separation @xmath0 could be readily explored with our approach .
= 1 |
it is now well appreciated that in certain situations , the translational symmetry of an electron gas becomes broken at low temperatures giving rise to an electron solid state called the wigner crystal ( wc ) .
@xcite among other conditions , low density and low disorder are crucial for this novel state to be realized ; the inter - electron distance has to be large enough for the coulomb interaction to dominate the quantum zero - point fluctuations , and the disorder must be weak enough not to destroy the long- ( or quasi - long- ) range order .
one of the systems that satisfy these conditions best is the two - dimensional ( 2d ) electron gas in the gaas / algaas heterostructure , in which density and disorder can be controlled relatively easily .
furthermore , if a strong magnetic field is applied perpendicular to the 2d plane , the cyclotron motion quenches zero - point fluctuations and the low - density condition is even more easily achieved .
@xcite although direct observations of the positional order from scattering measurements , for example , have been unattainable , there have been considerable efforts to prove the existence of the wc through more or less indirect methods such as transport measurements @xcite and photoluminescence .
@xcite one of the most convincing evidences so far is that the 2d systems become insulating when the filling factor @xmath1 is low : @xmath2 for electron systems @xcite and @xmath3 for hole systems .
@xcite this insulating behavior is generally accepted as a result of pinning of the wc due to impurities .
the impurity potential is also supposed to provide a restoring force in the ac electromagnetic response of the wc .
if the frequency of the external driving force matches the natural frequency of the pinning mode , a resonance should occur .
recent experiments on 2d hole systems @xcite have revealed a resonance structure in the ac absorption spectrum in the low filling factor regime . however , some features of the spectrum were qualitatively different from the predictions of the previous calculations based on charge - density - wave ( cdw ) models ; @xcite namely , ( i ) the peak frequency @xmath4 increased with the magnetic field @xmath5 , and ( ii ) the quality factor @xmath6 was found to be as large as 30 in some data .
@xcite the cdw calculations had predicted that @xmath4 would decrease with @xmath5 and the peak would be much broader ( @xmath7 ) .
these puzzling experimental findings were later examined by the authors of refs .
[ ref : chitra98 ] and [ ref : fertig99 ] . based on weak disorder models at @xmath8 , they found that ( i ) was explained if the disorder potential was allowed to vary within a magnetic length @xmath9 .
this short - length - scale physics is missing in the previous cdw calculations since they assume that the disorder potential can vary only in a length scale longer than the lattice periodicity @xmath10 .
( in the experiments in refs . [ ref : li97 ] , [ ref : li98 ] , and [ ref : hennigan98 ] , @xmath11 . ) more specifically , in ref . [ ref : fertig99 ] ,
it was suggested that the roughness of the gaas / algaas interface could serve as a source of such a short - length - scale disorder potential .
a `` pit '' on the interface , which is typically several atoms wide , can trap an electron by allowing it to stay closer to the positively charged donor plane and gain electrostatic energy . in a strong magnetic field , one can project the many - body wave function onto the lowest landau level ( lll ) . assuming @xmath12 , one may ignore exchange energy and describe the many - body ground state as a hartree - type product of single electron wave packets of size @xmath13 , which are , in the absence of disorder , located at crystal lattice sites .
@xcite suppose an electron wave packet is situated right on a pit . since its size decreases with increasing @xmath5 , the probability for the electron to stay inside the pit subsequently increases , provided _ the pit size is smaller than @xmath14_. since @xmath4 is essentially given by the expectation value of the pinning potential strength , the above cartoon picture explains how @xmath4 grows with @xmath5 .
in addition , as described below , the long - range coulomb interaction suppresses low - energy collective excitations and leads to a sharp resonance . at finite temperatures ,
however , thermal fluctuations should broaden the resonance peak .
one important observation is that the peaks measured in experiments keep getting narrower down to the lowest @xmath15 measured .
this suggests that the main source of broadening is thermal fluctuations .
interestingly , the experiments also find that @xmath16 even though @xmath17 .
thus , how the thermal broadening comes about is a highly non - trivial question .
we address the question in this paper . as
the thermal broadening is the main issue , we will simplify the disorder model and assume that the pinning centers themselves form exactly the same lattice as the wc .
we also assume that the impurity potential has the same strength at all sites .
@xcite several different disorder potentials including the uniform commensurate model we will use in this work have been studied at @xmath8 in a previous study , @xcite but it was found that the details of disorder do not alter the resonance structure qualitatively .
we believe the thermal behavior is not sensitive to details of disorder , either . our result can be summarized as follows . in order to understand low , but finite temperature behavior , one needs to investigate the low - energy excitations and their interactions .
the longitudinal and the transverse phonon modes of the wc mix together in a strong magnetic field and two new normal modes , magnetophonons and magnetoplasmons , emerge .
@xcite we focus on the magnetophonons because the energy of the other mode , magnetoplasmons , is essentially given by the cyclotron energy @xmath18 , which is a few orders of magnitude greater than @xmath15 and @xmath4 . in a theory that treats displacement of electrons from lattice sites harmonically , magnetophonons are independent and produces only a delta - function peak in the ac absorption spectrum .
the interaction between magnetophonons comes about from anharmonicity of the hamiltonian .
there are two sources of anharmonicity in our model : coulomb interaction and pinning potential .
since the external ac field has a much larger wave length ( @xmath19 m ) than the size of the system ( typically @xmath20 m ) , @xcite it is safe to treat it as a uniform field that couples only to the center - of - mass degree of freedom .
if the pinning potential is completely harmonic , we find that the anharmonicity in the coulomb interaction does not contribute to the absorption spectrum , because the interaction depends only on the relative position of electrons .
this is analogous to the kohn s theorem , @xcite which states that the electron - electron interaction does not affect the cyclotron frequency in a disorder - free system .
the theorem remains true even in a harmonically pinned wc , therefore , anharmonicity of the pinning potential is crucial for understanding thermal broadening .
as will be shown below , magnetophonon modes are very closely related to angular momentum excitations .
it is therefore natural to exploit analogies to 2d spin lattice systems .
in fact , written in magnetophonon creation and annihilation operators , our model hamiltonian resembles that of an su(n ) schwinger boson formulation of a 2d quantum ( anti-)ferromagnet . @xcite further utilizing this analogy , we use a @xmath0 expansion technique that has proved useful in spin systems . not only
does it provide a systematic way of collecting important processes in feynman diagrams , but also its lowest - order mean - field ( mf ) solution already contains terms that are higher order in coupling constant in a standard diagrammatic many - body calculation .
importantly , the position of absorption resonance peak is mostly determined at the mf level .
the coulomb interaction and the pinning potential affect the spectrum of the magnetophonons significantly . without pinning ,
the dispersion relation is well known to take the form @xmath21 in the long wave length limit , calculated both classically @xcite and quantum mechanically @xcite . in the uniform commensurate pinning model ,
it is modified to @xmath22 , where @xmath23 represents the strength of the pinning potential and @xmath24 is the sound velocity of the magnetophonon mode .
the new spectrum is not only gapped , but also makes the density of states ( dos ) vanish linearly in the low - energy limit .
it turns out that the slope of the low energy dos is so small that there are extremely few low - energy excitation modes about @xmath4 . as a consequence ,
we find that the thermal fluctuations are substantially suppressed and the peak remains sharp .
this paper is organized as follows .
the magnetophonon creation and annihilation operators are constructed out of displacement operators in the lll in sec .
[ sec : magnetophonons ] .
we derive the ac conductivity in terms of magnetophonon green s functions in sec .
[ sec : conductivity ] .
we then introduce in sec .
[ sec : hamiltonian ] , the hamiltonian of the wc , and derive in sec .
[ sec : kohnstheorem ] , a generalized kohn s theorem that explains why the anharmonicity in the coulomb interaction does not contribute to thermal shift or broadening of the absorption spectrum in a harmonic pinning model .
[ sec:1_nexpansion ] is devoted to @xmath0 expansion calculations of the green s functions .
we present the results in sec .
[ sec : result ] , and finally conclude with a discussion in sec .
[ sec : conclusion ] .
details of some calculations are given in appendices .
before we write down the hamiltonian in a second - quantized form , it is helpful to see how the hilbert space is spanned by the eigenfunctions of magnetoplasmon and magnetophonon number operators . in the circular gauge , @xmath25 ,
a 2d single electron wavefunction in the @xmath26-th landau level with angular momentum quantum number @xmath27 takes the form @xmath28 where @xmath29 , @xmath30 , @xmath31 , and @xmath9 is the magnetic length .
this wavefunction spans the entire single particle hilbert space .
it is now useful to define ladder operators that raise or lower @xmath26 and @xmath27 .
we define the landau level raising operator @xmath32 and the angular momentum raising operator @xmath33 it is straightforward to show that @xmath34 = [ b , b^\dagger ] = 1 , \nonumber \\ & & [ a , b ] = [ a^\dagger , b^\dagger ] = [ a , b^\dagger ] = [ a^\dagger , b ] = 0.\end{aligned}\ ] ] in what follows , the above ladder operators will be defined at each wc lattice site . then , a fourier transformed operator @xmath35 will serve as a magnetoplasmon creation operator , and @xmath36 as a magnetophonon creation operator .
we will use the latter terms to refer to the @xmath37 and @xmath38 operators from this point on . in the low filling factor limit
, the landau gap @xmath39 is large , so one can safely truncate the hilbert space and work in the lll only .
furthermore , the ground state of the wc in this limit is believed to be well represented by a product of a collection of zero angular momentum single particle wave functions , each of which is centered at a lattice site .
@xcite this is analogous to a hartree wave function .
although the single particle wave functions are not orthogonal , their overlap is exponentially small if @xmath40 .
thus , one may ignore exchange energy correction and the wave functions do not have to be antisymmetrized .
in other words , electrons are distinguishable because each one of them is strongly localized at a lattice site far from the others . for higher angular momentum states ,
the size of wave packets grow with @xmath27
. however , the above argument should be still valid for the first few higher angular momentum states .
therefore , we assume that the hilbert space is given by a direct product of single particle hilbert spaces spanned by angular momentum states in the lll . due to the strong coulomb repulsion
, we assume that there is only one electron per site .
this allows us to label each many - body energy eigenstate by a collection of @xmath41 angular momentum quantum numbers or magnetophonon occupation numbers , where @xmath41 is the total number of sites .
as will be shown below , the above statement is not exactly true , because in reality , different angular momentum states mix together due to quantum fluctuations .
this is demonstrated in our hamiltonian as terms that do not conserve the magnetophonon number .
nonetheless , higher angular momentum mixing is small at low temperatures and the above argument should be a good qualitative description of the system .
the many - body wavefunction is written as @xcite @xmath42 where the single particle wavefunction @xmath43 is localized at a lattice site @xmath44 and has an angular momentum quantum number @xmath45 .
more specifically , it may be written as @xmath46 where @xmath47 .
note that the last exponential is a pure phase and is needed to keep the wavefunction in the lll .
we will assume that @xmath48 forms a triangular lattice in the ground state .
the ladder operators in eqs .
( [ eq : magnetoplasmon ] ) and ( [ eq : magnetophonon ] ) are modified at each lattice site and may be written as @xmath49 where @xmath50 is the complex notation of the displacement of the @xmath51th electron from its site . in terms of the above ladder operators ,
the original electron position is written as @xmath52 when an operator is applied only to states that are completely in the lll , it may be represented by a `` lll - projected '' operator . in the above expression , this means simply dropping the magnetoplasmon operators @xmath53 and @xmath54 that couples the lll states to higher landau levels .
the projected operators then take the form @xmath55 the above operators satisfy the usual lll commutation relation @xcite @xmath56 = l_0 ^ 2 \delta_{ij}. \label{eq : commutationxy}\ ] ]
the quantity we will eventually compute is the ac conductivity tensor @xmath57 , where @xmath58 . in this section
, we will show how @xmath57 is derived in terms of the magnetophonon green s functions .
the external perturbation due to a spatially uniform electric field with frequency @xmath59 is @xmath60 , where @xmath61 . in the linear response theory ,
the electric current is given by @xmath62 \rangle \\ & = & { \mathrm re}\ , \frac{i\omega e^2\rho_0}{\hbar } e^{-i\omega t } \chi_r^{\mu\nu}(\omega ) e_0^\nu , \label{eq : current}\end{aligned}\ ] ] where @xmath63 is the 2d number density of electrons .
the brackets @xmath64 denote a thermal average for the unperturbed hamiltonian and the zero wave vector response function is defined as @xmath65 \rangle .
\label{eq : chir1}\ ] ] the linear ac conductivity can be read off from eq .
( [ eq : current ] ) as @xmath66 in general , one has to include the polarization field in the calculation and solve for the conductivity self - consistently . in a system with uniform background charge ,
however , the @xmath67 component of the polarization is meaningful only on the boundaries . in 2d
, the boundary effect dies off inversely with the size of the system , and may be ignored in the thermodynamic limit .
the response function @xmath68 may be most easily obtained by computing the matsubara frequency response function @xmath69 and then analytically continuing @xmath70 where @xmath71 denotes time - ordering in the imaginary time @xmath72 . since @xmath73 is a linear combination of @xmath74 and @xmath75 as in eqs .
( [ eq : xfromb ] ) and ( [ eq : yfromb ] ) , @xmath76 can be written in terms of magnetophonon green s functions as @xmath77 \nonumber \\ & = & \frac{l_0 ^ 2}{2 } \left [ \begin{array}{cc } 1 & 1 \\ i & -i \end{array } \right ] \left
[ \begin{array}{cc } g^{-+ } & g^{-- } \\ g^{++ } & g^{+- } \end{array } \right ] \left [ \begin{array}{cc } 1 & -i \\ 1 & i \end{array } \right ] .
\label{eq : chi}\end{aligned}\ ] ] the green s functions are defined as @xmath78 = -\int d\tau e^{i\omega_n\tau } \left\langle t_\tau \left [ \begin{array}{cc } b b^\dagger & b b \\ b^\dagger b^\dagger & b^\dagger b \end{array } \right ] \right\rangle , \label{eq : green}\ ] ] where @xmath79 should be interpreted as @xmath80 , etc .
the discussions so far have been quite general and independent of the actual hamiltonian of the system . in this section , we will construct the hamiltonian in terms of magnetophonon operators .
let us first consider a wc without disorder .
the kinetic part of the hamiltonian is given by @xmath81 where @xmath54 is the magnetoplasmon creation operator in eq .
( [ eq : laddera ] ) .
the coulomb interaction takes the form @xmath82 where @xmath83 is the dielectric constant .
expanded in terms of the displacement @xmath84 , the above equation may be written as @xmath85 , \label{eq : hc}\ ] ] where @xmath58 , and @xmath10 is the lattice constant .
the dynamic matrix is given by @xmath86 where @xmath87 .
now we take the strong - magnetic - field limit and project the above hamiltonian to the lll .
technically , this means rewriting it in terms of ladder operators in eqs .
( [ eq : laddera ] ) and ( [ eq : ladderb ] ) , and abandoning any term that is normal - ordered in @xmath37 and @xmath10 , i.e. , any term in which all @xmath10 s are placed to the right of all @xmath37 s .
obviously , the kinetic term is constant after the projection , and may be ignored .
the coulomb interaction may be rewritten as @xmath88 \nonumber \\ & & + \sum_i \left ( \sum_{j(\neq i ) } \frac{e^2l_0 ^ 2}{2|{\mathbf r}_{ij}|^3 } \right ) b_i^\dagger b_i , \label{eq : hc2}\end{aligned}\ ] ] where @xmath89 .
the above expression may also be derived by making use of the analogy between the electron system and a spin - lattice system ( see ref .
[ ref : fertig99 ] ) . note that rewriting eq .
( [ eq : hc ] ) into eq .
( [ eq : hc2 ] ) , we have dropped not only a constant term , but also the last term that is anharmonic .
in fact , in order to investigate thermal broadening , it is necessary to keep at least one anharmonic term in the hamiltonian , because a completely harmonic hamiltonian would lead to a theory of free independent eigenmodes and all correlation functions would be temperature independent .
however , the anharmonic terms in the coulomb interaction do _ not _ contribute to thermal shift or broadening of the response function to a spatially uniform external perturbation in a harmonic pinning model .
this is a highly non - trivial statement , which we discuss in the next section . as a consequence of this ,
we are led to investigate anharmonicity in the other term in the hamiltonian : the pinning potential . for simplicity
, we consider only the uniform commensurate pinning model in this work . in reality
, of course , disorder is more complicated in several ways .
first , non - commensurate or non - uniform disorder will break the discrete translational symmetry and deform the lattice .
unless the impurity potential is exactly circular at each site , it will also mix different angular momentum states . however , it was shown in ref . [ ref : fertig99 ] that at @xmath8 , the uniform commensurate pinning model yields qualitatively the same result as more complicated disorder models .
furthermore , experimental data show that the resonance peaks keep getting narrower down to @xmath90 , suggesting that thermal fluctuation affects broadening more than disorder does , at least at the observed temperatures .
therefore , we believe this simple disorder model is sufficient for our purpose . fig .
[ fig : pinning ] shows a schematic picture of the probability function for the three lowest angular momentum states centered at an interface pit .
since states with @xmath91 have little probability to reside in the pit , they are not affected by it .
the pinning hamiltonian is thus appropriately described in terms of magnetophonon operators by @xmath92 where @xmath23 is the amount of energy an electron pays to move from @xmath93 state to a higher angular momentum state _ in the absence of coulomb interaction_. this potential is highly anharmonic , and the resulting phonon problem is exceedingly difficult to analyze . to overcome this
, we adopt an approximate form of the pinning potential . the simplest choice that includes anharmonicity is @xmath94 which takes the same value as @xmath95 at @xmath96 , 1 , and 2 .
at first , the above hamiltonian looks problematic because it has no lower bound . as will be shown below
, however , the @xmath0 expansion is well defined provided the thermal average @xmath97 .
it turns out that @xmath98 is indeed small and this approximation gives sensible results in the observed temperature range @xmath99 .
finally , the total hamiltonian of our model is given by @xmath100 a similar hamiltonian was derived in ref .
[ ref : fertig99 ] by limiting the hilbert space to two angular momentum states per site and writing the hamiltonian in a pseudospin representation .
although the higher - angular - momentum states should not affect low temperature behavior qualitatively , and the assumption that the many - body wave function takes a simple product form breaks down for large @xmath45 , the above hamiltonian has been derived without restricting the hilbert space to a fixed number of states per site .
in this section , we will derive a generalized kohn s theorem , which guarantees that the anharmonicity in the coulomb interaction does not alter the absorption spectrum of a uniform electromagnetic field for a harmonically and uniformly pinned model .
similar generalizations have been made in the context of an electron gas in a parabolic potential in refs .
[ ref : brey89 ] and [ ref : yip91 ] . before we present the proof ,
let us take a brief look at the original kohn s theorem . in a magnetic field ,
the energy spectrum of a single free electron is simply given by landau levels .
now , suppose that there are many electrons interacting with one another via coulomb or any other interaction .
kohn s theorem tells us that the electromagnetic absorption spectrum of such a system , as measured through a response to a spatially uniform external field , will be exactly the same as that of a single electron .
the proof of this theorem is disarmingly simple .
since the electron - electron interaction depends only on the relative degrees of freedom , it commutes with the center - of - mass motion .
therefore , the observed center - of - mass spectrum is not affected by the interaction at all .
the big difference in the problem considered in this paper is that there is pinning .
however , a similar argument can be established in a special case of harmonic uniform pinning model .
suppose that the pinning hamiltonian takes the form @xmath101 since this hamiltonian is separable in different momenta @xmath102 , its eigenstates may be written as a product of the form @xmath103 where @xmath104 is an eigenstate of the center - of - mass operator @xmath105 and @xmath106 of all the other relative degrees of freedom .
suppose that we calculate the ac conductivity of a _ non - interacting _ electron system using the expressions in sec .
[ sec : conductivity ] . in order to compute the thermal average
, one needs to compute the expectation values over the above eigenstates .
we make an important observation here that the conductivity depends only on @xmath104 , because the operators in eq .
( [ eq : chir1 ] ) are functions only of @xmath107 and @xmath108 .
now let us add the coulomb interaction as in eq .
( [ eq : hc0 ] ) to the hamiltonian . as
@xmath109 depends only on relative degrees of freedom , it obviously commutes with the center - of - mass displacement @xmath110 in terms of magnetophonon operators , it implies @xmath111 = 0.\ ] ] therefore , the coulomb interaction does not alter the spectrum of @xmath104 , although it may ( and in general , it does ) alter that of @xmath106 .
since the ac conductivity is determined by the spectrum of @xmath104 alone , it is unaltered by the presence of the coulomb interaction .
this completes the proof of the generalized kohn s theorem .
this theorem implies that if we use a harmonic pinning potential , the contribution from the anharmonic coulomb interaction terms in eq .
( [ eq : hc ] ) to the uniform response function will _ vanish in all orders _ of @xmath112 .
therefore , we will use a harmonic hamiltonian for coulomb interaction , but an anharmonic one for pinning potential , hence the total hamiltonian is given as in eq .
( [ eq : htotal ] ) .
the detailed analysis of the anharmonic pinning potential will be given below , using a @xmath0 expansion technique .
a @xmath0 expansion technique consists of three steps .
first , we make @xmath113 copies of the magnetophonon operators .
second , using a hubbard - stratonovich ( hs ) decoupling technique , the anharmonic terms are replaced by products of a harmonic term and an auxiliary hs field . finally , we expand the magnetophonon green s function in powers of @xmath0 . in principle , the expansion is exact if all orders of @xmath0 are kept . instead ,
as is conventionally done , we will stop at the first order to get an approximation . before going into details
, it should be noted that @xmath113 will be set equal to 1 at the end of the approximation , therefore it is not a perturbative expansion in any sense .
however , this technique usually does a better job in systematically picking out important processes than many other many - body techniques .
in fact , the zeroth order approximation , which is a harmonic theory with magnetophonons coupled to the mf value of the hs field , already contains the rpa contribution .
the first order approximation adds self - energy corrections to the green s functions and gives us a thermal broadening .
after replication , eq .
( [ eq : htotal ] ) becomes @xmath114 where @xmath115 , @xmath116 , and @xmath117 are species indices .
@xmath118 is the same as in eq .
( [ eq : hc2 ] ) with different magnetophonon operators substituted for each species .
there is an extra factor of @xmath0 in the last term because it involves two species summations .
note that the above equation has been normal - ordered . in this form
, we may recast the problem in a path - integral form simply by replacing @xmath119 ( @xmath38 ) by a classical field @xmath120 ( @xmath121 ) in the imaginary time @xmath122 .
the euclidean action is given by @xmath123 ^ 2 \biggr\}.\end{aligned}\ ] ] the first term is the usual berry phase in the imaginary time . in the second step ,
we introduce a real hs field @xmath124 to decouple the quartic term .
the procedure is formally expressed in a functional integral form as @xmath125 \nonumber \\ & & = c \int { \mathcal d}q_i\ , \exp \biggl [ -\int d\tau \nonumber \\ & & \qquad\qquad\qquad\qquad \frac{v}{2 } \biggl ( -2q_i\sum_l |z_{li}|^2 + nq_i^2 \biggr ) \biggr ] , \nonumber\end{aligned}\ ] ] where @xmath126 is a constant .
then the generating functional written in terms of both @xmath127 and @xmath128 takes the form @xmath129 where @xmath130 .
\label{eq : shs}\end{aligned}\ ] ] @xmath131 denotes functional integral over real and imaginary parts of @xmath127 field .
so far the action contains all orders of @xmath0 and is therefore exact . in the following subsections , we will taylor - series expand it in powers of @xmath0 . as is the usual practice , @xcite we assume that the mf solution of @xmath124 is independent of @xmath51 and @xmath72 ( ref .
[ ref : meanfield ] ) .
since the crystal momentum @xmath132 is still a good quantum number , it is useful to fourier - transform the action .
we define @xmath133 where @xmath134 is a shorthand notation for @xmath135 and @xmath136 is the inverse temperature . in terms of the above fourier components ,
the mf action is given by @xmath137 , \label{eq : smf}\end{aligned}\ ] ] where @xmath138 the mf green s functions take the form @xmath139 = \nonumber \\ & & \qquad \frac{1}{\omega_n^2+f_{\mathbf q}^2-|g_{\mathbf q}|^2 } \left [ \begin{array}{cc } -i\omega_n - f_{\mathbf q } & g_{\mathbf q}^ * \\
g_{\mathbf q } & i\omega_n - f_{\mathbf q } \end{array } \right ] .
\label{eq : mfgreensfunc}\end{aligned}\ ] ] we choose the mf value @xmath140 in such a way that the mf free energy @xmath141 has a saddle point . solving @xmath142 , we get @xmath143 note that @xmath140 above is equal to the thermally averaged value of the number of magnetophonons per site , @xmath144 .
@xmath140 has to be determined self - consistently and is a function of temperature .
note that @xmath145 for all @xmath132 if @xmath146 .
this implies that the mf theory is stable when there are not too many magnetophonons .
the instability in the many - magnetophonon regime signals the onset of a `` depinning transition . '' in fact , this instability is an artifact caused by the fact that our choice of the pinning potential @xmath147 in eq .
( [ eq : hp4 ] ) has no lower bound .
however , we believe there is indeed a depinning transition in real situations , and it is appropriate to regard @xmath148 as the critical number of magnetophonons at the transition in this model . as we go beyond the mf theory and include higher orders in @xmath0 , @xmath149 decreases due to fluctuations in @xmath150 . since the mf action @xmath151 is quadratic , it may be diagonalized using a bogoliubov transformation . it is more convenient to work in the hamiltonian representation .
the hamiltonian that reproduces @xmath151 is @xmath152.\end{aligned}\ ] ] we define a new boson annihilation operator @xmath153 in order to ensure the boson statistics , it has to satisfy @xmath154 = |u_{\mathbf q}|^2 - |v_{\mathbf q}|^2 = 1.\ ] ] by setting @xmath155 we can diagonalize @xmath156 and write it as @xmath157 where @xmath158 .
the energy eigenvalues are given by @xmath159 an example of the dispersion relation is shown in fig .
[ fig : dispersion ] .
substituting @xmath160 in the above expressions , the dispersion relation without pinning , @xmath21 , is recovered .
@xcite taylor - expanded in small @xmath161 , the parameters and the energy eigenvalue take the form @xmath162 , \\
g_{\mathbf q } & = & \frac{\nu e^2}{2\kappa } \frac{(q_x+iq_y)^2}{|{\mathbf q}| } + { \mathcal o}[|{\mathbf q}|^2 ] , \\
\varepsilon_{\mathbf q } & = & ( 1-\overline{q})v + \frac{\nu e^2}{2\kappa } |{\mathbf q}| + { \mathcal o}[|{\mathbf q}|^2 ] , \label{eq : smallq}\end{aligned}\ ] ] which are valid if @xmath163 one of the authors has obtained a similar dispersion relation at @xmath8 .
@xcite the main difference for this finite temperature model is that the zeroth order term contains the self - consistent mf value @xmath140 . as pointed out in ref .
[ ref : fertig99 ] , the above linear dispersion is a direct consequence of the fact that the coulomb interaction is long - ranged . with this dispersion relation ,
the magnetophonon density of states vanishes at the lower band edge , and thermal broadening will be substantially suppressed as will be shown below .
this is not the case for a short - range interaction .
for example , a screened interaction would have yielded a dispersion that is quadratic at small wave vectors . in that case
, the density of states jumps to a finite value at the band edge . in a separate test
, we have found that thermal broadening is indeed several orders of magnitude greater for a screened interaction than for an unscreened one .
since @xmath164 and @xmath165 are functions of @xmath140 , the saddle point solution for @xmath140 has to be determined self - consistently .
more specifically , it has to satisfy @xmath166,\end{aligned}\ ] ] where @xmath167 is the bose function . in the mf theory
, the uniform field absorption spectrum will have a delta function peak at @xmath168 .
since @xmath169 and @xmath140 increases with temperature , the thermal reduction of the peak frequency observed in experiments is explained qualitatively already at the mf level .
physical interpretation is also simple . at high temperatures , more
magnetophonons are thermally created .
this implies that the average number of electrons that stay in the pits decreases , leading to a drop in the spatially averaged pinning potential . in order to explain thermal broadening , however , one has to go beyond the mf theory .
we will compute the first order corrections in the @xmath0 expansion in the next subsection .
in this subsection , we will consider fluctuations of the hs field about its mf solution .
the fluctuation fields are defined at each site as @xmath170 and its fourier - transformed field is @xmath171 since @xmath172 is real , @xmath173 . in terms of @xmath174 , eq .
( [ eq : shs ] ) may be rewritten as @xmath175 ( { \mathsf g}_{\mathrm mf}^{-1 } + { \mathsf u})_{qq ' } \left [ \begin{array}{c } z_{lq ' } \\ z_{l ,- q'}^ * \end{array } \right ] \nonumber \\ & & + \frac{\beta nv}{2 } \left ( n_s\overline{q}^2 - 2\sqrt{n_s}\overline{q}r_{q=0 } + \sum_q |r_q|^2 \right),\end{aligned}\ ] ] where @xmath176 \delta_{qq ' } , \\
{ \mathsf u}_{qq ' } & = & - \frac{1}{\sqrt{n_s } } \left [ \begin{array}{cc } v r_{q - q ' } & 0 \\ 0 & v r_{q - q ' } \\
\end{array } \right].\end{aligned}\ ] ] @xmath177 is the mf magnetophonon green s function matrix and @xmath178 describes the coupling between a magnetophonon and the hs field @xmath179 , or a vertex in the feynman diagrams .
the `` bare '' hs propagator can be read off from @xmath180 and is given by @xmath181 fig .
[ fig : feynman](a ) shows the building blocks of the feynman diagrams that will be frequently used below .
= 2.5 in note that @xmath182 is proportional to @xmath0 , so each curly line raises the order of an overall diagram by one in the @xmath0 expansion . on the other hand , a magnetophonon loop
[ see fig .
[ fig : feynman](b ) ] contributes order @xmath113 due to a summation over an internal species - index . in order to find corrections to the hs propagator
, it is convenient to integrate out the magnetophonon degree of freedom .
the effective action for the hs field @xmath174 is given by @xmath183 & = & -\ln \int { \mathcal d}^2z\ , e^{-s_{\mathrm hs } } \\ & = & -\frac{n}{2 } \text{tr}\,\ln \left [ \beta({\mathsf g}_{\mathrm mf}^{-1 } + { \mathsf u } ) \right ] \nonumber \\ & & \quad + \frac{\beta nv}{2 } \left ( \sum_q |r_q|^2 - 2\overline{q}\sqrt{n_s}r_{q=0 } \right ) \label{eq : seff2 } \\ & = & \frac{n}{2 } \left [ \text{tr}\,\ln \frac{{\mathsf g}_{\mathrm mf}}{\beta } - \sum_{n=2}^\infty \frac{1}{n } \text{tr}\,({\mathsf g}_{\mathrm mf}{\mathsf u})^n \right ] \nonumber \\ & & \quad + \frac{\beta nv}{2 } \sum_q |r_q|^2 .
\label{eq : seff3}\end{aligned}\ ] ] going from eq .
( [ eq : seff2 ] ) to eq .
( [ eq : seff3 ] ) , we have taylor - series expanded the logarithmic function about @xmath184 .
unimportant constants were omitted at every step .
note that the last term in eq .
( [ eq : seff2 ] ) exactly cancels out the @xmath185 term that would have been in eq .
( [ eq : seff3 ] ) , because @xmath140 is chosen at a saddle point . @xmath177 and @xmath178 are considered as entities in a product space of a @xmath186 matrix and a matrix labeled by frequency and wave - number indices @xmath187 .
thus , a trace operator sums not only over the @xmath186 matrix indices , but also over @xmath134 .
let us take a closer look at the last expression ( [ eq : seff3 ] ) .
the first term is the mf free energy and the last term is the action of free independent hs fields @xmath174 .
each of the remaining terms is represented by a loop with @xmath26 vertices as in fig .
[ fig : feynman](b ) .
let us now search for the lowest order correction to the hs propagator .
there is no correction of the zeroth order or lower , and the only @xmath0 correction comes from the @xmath188 term of eq .
( [ eq : seff3 ] ) . as can be seen in the first feynman diagram in fig . [ fig : feynman](b ) , this term is nothing but the rpa correction to the hs propagator .
we explicitly write it as @xmath189 where @xmath190 is the mf magnetophonon density - density correlation function defined as @xmath191 where @xmath192 denotes an average in the mf theory
. detailed calculations of @xmath190 will be given in appendix a. the rpa `` dressed '' propagator is defined through the dyson equation @xmath193 since both the bare propagator @xmath182 and the rpa correction are of order @xmath0 , the above propagator is also of order @xmath0
. a thick curly line will be used for the feynman diagram of the dressed propagator as in fig .
[ fig : rpa ] .
= 2.5 in we now have all the ingredients to calculate the @xmath0 correction to the magnetophonon green s function . keeping in mind that each hs propagator is of order @xmath0 and each magnetophonon loop
is of order @xmath113 , it is straightforward to count the order of any feynman diagram .
there are two self - energy diagrams of order @xmath0 as shown in fig .
[ fig : selfenergy ] .
they are written in terms of mf green s functions @xmath194 and the hs propagator @xmath195 as @xmath196 \nonumber \\ & & \quad + \bigl [ g_{\mathrm mf}^{--}(q '' ) g_{\mathrm mf}^{++}(q ' ) g_{\mathrm mf}^{-+}(q ' ) + \text{c.c . }
\bigr ] \biggr\}. \label{eq : sigmab}\end{aligned}\ ] ] these expressions can be partially evaluated analytically via frequency summations .
@xcite more details may be found in appendix b. the self - energy correction is incorporated into the green s functions via the dyson equation @xmath197^{-1 } , \label{eq : gdyson}\ ] ] where @xmath198 , \qquad { \mathsf i } = \left [ \begin{array}{cc } 1 & 0 \\ 0 & 1 \end{array } \right].\ ] ] combining eq .
( [ eq : gdyson ] ) with eqs .
( [ eq : sigma ] ) , ( [ eq : chir ] ) , and ( [ eq : chi ] ) , we can finally compute the ac conductivity to the order @xmath0 . using the final expressions in appendix b , we have numerically computed the self - energy diagrams and obtained the conductivity as a function of @xmath15 .
the result is presented in the next section .
before we present the result , it is useful to review some important energy scales . for the purpose of comparison , we will use the data in ref .
[ ref : li98 ] .
we find @xmath199 the temperature range for which the line width was clearly discernible was @xmath200ev @xmath201ev . for efficiency in numerical computations ,
we have made the following isotropic approximations @xmath202 we did , however , keep the shape of the first brillouin zone to be hexagonal . according to our numerical tests ,
these approximations break down as the brillouin zone boundaries are approached , i.e. , when @xmath203 , but they are very accurate for @xmath204 .
for example , when @xmath205 , the energy @xmath165 is already an order of magnitude greater than @xmath206 at @xmath207 ( see fig .
[ fig : dispersion ] ) , but the error @xmath208 is still less than 4% .
therefore , this approximation should be quantitatively reliable if @xmath15 is not too much greater than @xmath206 . in order to compare our result with the experimental data , we have chosen the bare pinning potential @xmath23 in such a way that the resulting peak frequency @xmath4 agrees with the experiment at the lowest measured temperature .
the peak position data are shown in fig . [
fig : peakposition ] , along with experimental data from ref .
[ ref : li98 ] .
although our results are systematically overestimated , the qualitative agreement is rather good .
the results for line width @xmath209 are surprisingly small ( fig .
[ fig : peakwidth ] ) . compared to those from experimental data , our results are several orders of magnitude smaller at the lowest @xmath15 .
this is especially surprising because @xmath210 .
navely , one would expect that the line width should be of the same order of magnitude as @xmath15 , due to thermal broadening .
although much greater than our results , experiments also confirm that this is clearly not the case . in refs .
[ ref : li97 ] and [ ref : li98 ] , @xmath209 is almost one order of magnitude smaller than @xmath4 . in ref .
[ ref : hennigan98 ] , the quality factor @xmath6 even exceeds 30 .
there are also other indications that the peaks may be actually even sharper .
@xcite the unusually sharp peak may be rather well understood within our model . as discussed in sec .
[ subsec : mftheory ] , due to the long - range coulomb interaction , the dispersion starts out linearly in @xmath161 .
the density of states thus increases linearly from zero .
explicitly , the density of states per unit cell takes the form @xmath211 & \text{if}\ \varepsilon>(1-\overline{q})v , \\ 0 & \text{if}\ \varepsilon\leq ( 1-\overline{q})v .
\end{array } \right.\ ] ] a rough estimate of the total number of `` accessible states '' per site at temperature @xmath15 would be @xmath212 where @xmath213 is the unit cell area . using the parameters in ref .
[ ref : li98 ] , we find that there are merely @xmath214 accessible states per site at @xmath215 mk , which is the highest temperature for which the line width was measured . the physical interpretation of this may be given as the following .
at least in the semiclassical level , our model is essentially that of a crystal in which each electron is attached to a lattice site by a short - range binding potential .
the overall uniform motion ( @xmath216 ) determines the peak frequency .
broadening comes about from thermal and quantum fluctuations deforming the lattice and making the electrons move relative to one another .
however , the above analysis tells us that a wc in a strong magnetic field is so rigid that it is hardly deformed even at temperatures greater than the pinning potential . in reality
, the external ac field has a finite wave length of order @xmath217 m @xmath218 ( ref .
[ ref : li97 ] ) .
accordingly , we have performed a similar analysis for the response at the finite wave vector , but we found no qualitative change .
defects in a wc such as interstitials , vacancies , dislocations , etc . may also affect broadening .
it is certainly true that their presence will soften the lattice and subsequently enhance magnetophonon excitations . however , unless one is too close to the melting temperature of the wc , which is not the case in the experiments , their effects may be mostly taken into account through renormalization of the lattice stiffness , or dynamic matrix in eq .
( [ eq : dynamicmatrix ] ) .
this should not affect our results qualitatively
. there may be other sources of broadening , too .
other low energy modes , which are not included in this work , may be involved in the broadening .
one possible candidate is the edge states of the wc , @xcite the analysis of which will be given elsewhere .
another possible source of broadening is extrinsic low energy modes _ outside _ the wc .
when an experiment is performed , there are many external degrees of freedom that may be coupled to the sample .
if some of them are hard to remove and thus are left undetected to influence the data , the observed peak will be certainly broader than theoretically predicted .
it has also been suggested that higher landau level mixing might be important .
@xcite it is certainly true that there can be second or higher order scattering processes that use higher landau levels as virtual states .
however , these processes should be suppressed as a power @xmath219 with @xmath220 . since @xmath221 , @xmath209 as a result of such processes
would be still too small to explain currently available experimental data .
in this work , we studied the thermal broadening of the electromagnetic absorption resonance of a magnetically induced wc . in the strong magnetic field limit ,
the low - energy collective modes that are coupled to the spatially uniform ac electric field are magnetophonons . assuming that electrons are closely bound to lattice sites , magnetophonon creation and annihilation operators may be constructed out of displacements @xmath61 .
we showed that the hamiltonian could be divided into harmonic and anharmonic parts in terms of magnetophonon operators .
the harmonic part , when taken alone , describes independent magnetophonons and produces a delta function peak in the spectrum .
the anharmonic part introduces magnetophonon interactions .
it will not only renormalize the peak position , but also broaden it by mixing different magnetophonon modes .
anharmonicity in our model comes from two sources : the coulomb interaction and the pinning potential .
since the coulomb interaction depends only on the relative coordinates , it is completely decoupled from the center - of - mass degrees of freedom .
this leads to the derivation of a generalized kohn s theorem that asserts that the coulomb interaction can not shift or broaden an ac conductivity resonance peak even in the presence of a uniform harmonic pinning potential .
analysis of the magnetophonon interactions in the pinning potential was performed using a @xmath0 expansion technique .
this technique provides a systematic way of summing up important diagrams and it captures important pieces of physics in the low order solutions . for example , we found that the zeroth order mf solution could account for the decreasing peak frequency @xmath4 as a function of temperature .
thermal broadening appears in the @xmath0 self - energy corrections .
however , the line width @xmath209 is found to be many orders of magnitude smaller than @xmath4 and @xmath15 .
the reason for this lies in the magnetophonon dispersion that is linear and steep in the low energy limit .
this translates into a small density of states for collective excitations .
consequently the broadening is substantially suppressed .
our result of peak frequency agrees qualitatively with recent experiments .
@xcite the line width result is also qualitatively consistent in that it is much less than @xmath4 and @xmath15 .
quantitatively , our result of @xmath209 appears to be much smaller than published results , although there is continuing experimental work on sorting out just how narrow the intrinsic line width really is .
@xcite other possible sources of broadening , such as edge states and extrinsic low energy modes , may also be responsible for observed line width .
we are indebted to lloyd engel , chi - chun li , dan tsui , and jongsoo yoon for helpful discussions regarding their experiments .
carsten timm and steve girvin are gratefully acknowledged for useful discussions about the @xmath0 expansion technique .
we also thank institute for theoretical physics , uc santa barbara , where part of this work was done .
this work was supported by nsf grant dmr98 - 70681 and phy94 - 07194 and the research corporation .
in this appendix , we will present detailed calculations of the density - density correlation function @xmath190 .
( [ eq : densitycorr ] ) can be written as @xmath223 \nonumber \\ & = & \frac{1}{\beta n_s } \sum_{q ' } \bigl [ g_{\mathrm mf}^{-+}(q'-q ) g_{\mathrm mf}^{-+}(q ' ) + g_{\mathrm mf}^{--}(q'-q ) g_{\mathrm mf}^{++}(q ' ) , \bigr]\end{aligned}\ ] ] where we have used @xmath224 to get the second line .
diagrammatically , the last two terms are represented by two arrowed diagrams as in fig .
[ fig : densitycorrelation ] .
= 2.5 in using the green s functions in eq .
( [ eq : mfgreensfunc ] ) , we may rewrite the above equation more explicitly as @xmath225[(i\omega_n')^2-\varepsilon_{\mathbf q'}^2 ] } \\ & = & \frac{1}{2n_s } \sum_{\mathbf q ' } \left [ ( n_{\mathbf q'-q}-n_{\mathbf q ' } ) \frac{(\varepsilon_{\mathbf q'-q}-\varepsilon_{\mathbf q'})(1+\xi_{\mathbf q'-q ,
q'})}{(i\omega_n)^2-(\varepsilon_{\mathbf q'-q}-\varepsilon_{\mathbf q'})^2 } + ( n_{\mathbf q'-q}+n_{\mathbf q'}+1 ) \frac{(\varepsilon_{\mathbf q'-q}+\varepsilon_{\mathbf q'})(1-\xi_{\mathbf q'-q , q'})}{(i\omega_n)^2-(\varepsilon_{\mathbf q'-q}+\varepsilon_{\mathbf q'})^2 } \right],\end{aligned}\ ] ] where @xmath226 is a shorthand notation of the bose function @xmath227 . in the second line
, we have used the usual frequency summation technique .
@xcite @xmath228 has many simple poles in the complex @xmath59 plane on the real axis . in the thermodynamic limit ,
they become branch cuts .
therefore , summations involving @xmath228 must be performed with care , as will be demonstrated in the calculations of self energy corrections in the next appendix .
note that @xmath228 is analytic and real at @xmath229 , so that the branch cuts are divided into two : one on the positive and the other on the negative axis .
both branch cuts are bounded because @xmath230 is finite .
eventually , we will analytically continue @xmath231 to get a retarded correlation function .
since @xmath232 and @xmath228 is analytic except on the branch cuts , it has the following properties : @xmath233 which will become useful in the next appendix .
finally , @xmath234 was numerically computed using an isotropic approximation as explained in sec .
[ sec : result ] .
in this appendix , we will calculate the @xmath0 self energy corrections to the magnetophonon green s functions as in fig .
[ fig : selfenergy ] . since the bare hs propagator @xmath182 is a constant , it is convenient to separate it from the rest of the rpa dressed propagator @xmath195 .
the remaining part @xmath237}\end{aligned}\ ] ] vanishes as @xmath238 in the large @xmath59 limit , so all following frequency sums that contain @xmath239 converge without introducing cumbersome convergence factors such as @xmath240 .
the first self - energy correction @xmath241 is given in eq .
( [ eq : sigmaa ] ) .
the contribution of the bare propagator @xmath182 is simply @xmath242 the summation of mf green s function is defined with an appropriate convergence factor that ensures normal - ordering .
it is then straightforward to show @xmath243 where @xmath244 is the anomalous magnetophonon number per site . due to the six - fold symmetry of the lattice ,
however , @xmath245 .
we therefore get @xmath246 now let us turn our attention to the remaining part .
it needs to be computed with more care , since @xmath247 has branch cuts on the real @xmath59 axis , that arise from the density - density correlation function @xmath248 . using the green s functions in eq .
( [ eq : mfgreensfunc ] ) , we may write @xmath249,\end{aligned}\ ] ] where @xmath250 and @xmath251 are defined in eqs .
( [ eq : u ] ) and ( [ eq : v ] ) . as in a standard frequency summation technique ,
@xcite we may replace the sum over @xmath252 by a contour integral on the complex @xmath253 plane with simple poles on the imaginary axis .
after the contour is deformed as in fig .
[ fig : contour ] , the new expression takes the form @xmath254 n_b(\varepsilon_{\mathbf q ' } ) \nonumber \\ & & - \frac{v^2}{\pi nn_s } \sum_{\mathbf q ' } \int_{\omega'\neq 0 } d\omega ' [ { \mathrm im}\,\delta d(\omega'+i0^+,{\mathbf q - q ' } ) ] \left [ \frac{|u_{\mathbf q'}|^2 n_b(\omega')}{i\omega_n+\omega'-\varepsilon_{\mathbf q ' } } + \frac{|v_{\mathbf q'}|^2 n_b(-\omega')}{i\omega_n+\omega'+\varepsilon_{\mathbf q ' } } \right ] , \label{eq : contourintegral}\end{aligned}\ ] ] where @xmath255 is excluded from the domain of the integral @xmath256 although @xmath257 diverges as @xmath258 , the above integral is well defined , because @xmath259 vanishes as @xmath255 is approached .
finally , analytic continuation is performed to give @xmath260 n_b(\varepsilon_{\mathbf q ' } ) \nonumber \\ & & - \frac{v^2}{\pi nn_s } \sum_{\mathbf q ' } p\!\!\int d\omega ' { \mathrm im}\,\delta d(\omega'+i0^+,{\mathbf q - q ' } ) \left [ \frac{|u_{\mathbf q'}|^2 n_b(\omega')}{\omega+\omega'-\varepsilon_{\mathbf q ' } } + \frac{|v_{\mathbf q'}|^2 n_b(-\omega')}{\omega+\omega'+\varepsilon_{\mathbf q ' } } \right ] \nonumber \\ & & -i\frac{v^2}{nn_s } \sum_{\mathbf q ' } \bigl [ { \mathrm im}\,\delta d(\omega-\varepsilon_{\mathbf q'}+i0^+,{\mathbf q - q ' } ) |u_{\mathbf q'}|^2 n_b(\varepsilon_{\mathbf q'}-\omega ) \nonumber \\ & & \qquad\qquad\qquad + { \mathrm im}\,\delta d(\omega+\varepsilon_{\mathbf q'}+i0^+,{\mathbf q - q ' } ) |v_{\mathbf q'}|^2 n_b(\varepsilon_{\mathbf q'}+\omega ) \bigr ] , \label{eq : dsigmaa}\end{aligned}\ ] ] where @xmath261 denotes a cauchy principal integral .
( [ eq : rhoanalyticity ] ) was used in the last line .
the above quantity may be computed using the numerical solution of @xmath262 .
the other @xmath263 s can be calculated in a similar way .
the final form of @xmath264 is the same the above equation , except that @xmath265 and @xmath266 are switched .
@xmath267 ( @xmath268 ) is obtained by replacing both @xmath265 and @xmath266 by @xmath269 ( @xmath270 ) . in the isotropic approximation , the angular sum of @xmath271 vanishes ,
so @xmath272 the other @xmath0 correction , @xmath236 , may be calculated in a similar manner .
first , we compute the contribution from the bare hs propagator . substituting @xmath182 for @xmath273 , eq .
( [ eq : sigmab ] ) becomes @xmath274 + \bigl [ g_{\mathrm mf}^{--}(q '' ) g_{\mathrm mf}^{++}(q ' ) g_{\mathrm mf}^{-+}(q ' ) + \text{c.c .
} \bigr ] \biggr\}.\ ] ] using eqs .
( [ eq : sumg1 ] ) through ( [ eq : sumg4 ] ) along with @xmath275 , \\ \sum_{q ' } g_{\mathrm mf}^{--}(q ' ) g_{\mathrm mf}^{++}(q ' ) & = & \sum_{q ' } |u_{\mathbf q'}|^2 |v_{\mathbf q'}|^2 \left ( \frac{1}{i\omega_n-\varepsilon_{\mathbf q ' } } - \frac{1}{i\omega_n+\varepsilon_{\mathbf q ' } } \right)^2 \nonumber \\ & = & -\beta \sum_{\mathbf q ' } |u_{\mathbf q'}|^2 |v_{\mathbf q'}|^2 \left [ n_b'(\varepsilon_{\mathbf q ' } ) - \frac{2n_b(\varepsilon_{\mathbf q'})+1}{2\varepsilon_{\mathbf q ' } } \right],\end{aligned}\ ] ] we get @xmath276nn_s } \sum_{\mathbf q ' } \left\ { \frac{|g_{\mathbf q'}|^2}{\varepsilon_{\mathbf q'}^2 } \frac{2n_b(\varepsilon_{\mathbf q'})+1}{2\varepsilon_{\mathbf q ' } } - \frac{f_{\mathbf q'}^2}{\varepsilon_{\mathbf q'}^2 } n_b'(\varepsilon_{\mathbf q ' } ) \right\}. \label{eq : sigmab0}\ ] ] the derivative of the bose function satisfies @xmath277 $ ] . for the remaining part
, we substitute @xmath278 for @xmath273 .
we get @xmath279.\end{aligned}\ ] ] shorthand notations are defined as @xmath280 , @xmath281 , etc . in order to avoid the branch cuts in @xmath282 , we make a change of variables @xmath283 and perform the frequency sum over @xmath284 .
the result takes the form @xmath285 \nonumber \\ & & \qquad\qquad\qquad + |u'v''+v'u''|^2 \left(|u'|^2+|v'|^2\right ) \left [ \frac{n_b(\varepsilon'')+n_b(\varepsilon')+1}{(i\omega_n'+\varepsilon'+\varepsilon'')^2 } - \frac{n_b'(\varepsilon')}{i\omega_n'+\varepsilon'+\varepsilon '' } \right ] \nonumber \\ & & \qquad\qquad\qquad - \frac{(u'u''^*+v'v''^*)(u'v''+v'u'')u'^*v'^ * + \text{c.c.}}{\varepsilon ' } \left [ \frac{n_b(\varepsilon'')-n_b(\varepsilon')}{i\omega_n'-\varepsilon'+\varepsilon '' } - \frac{n_b(\varepsilon'')+n_b(\varepsilon')+1}{i\omega_n'+\varepsilon'+\varepsilon '' } \right ] \biggr\}. \label{eq : deltasigmab}\end{aligned}\ ] ] the last frequency sum @xmath286 is performed using a similar contour deformation technique as in fig .
[ fig : contour ] . using @xmath287 \frac{1}{\omega'-\varepsilon+i0^+},\end{aligned}\ ] ]
( [ eq : deltasigmab ] ) becomes @xmath288n_s^2 } \sum_{\mathbf q'q''}{\mathrm im}\int_{\omega'\neq 0 } d\omega'\ , \biggl\ { \delta d(\omega',{\mathbf -q'-q''})n_b(\omega ' ) \left ( \frac{c_1}{\omega'-\varepsilon'+\varepsilon''+i0^+ } + \frac{c_2}{\omega'+\varepsilon'+\varepsilon''+i0^+ } \right ) \nonumber \\ & & \qquad\qquad\qquad\qquad\qquad + \left [ \frac{\partial}{\partial\omega ' } \delta d(\omega',{\mathbf -q'-q''})n_b(\omega ' ) \right ] \left ( \frac{c_3}{\omega'-\varepsilon'+\varepsilon''+i0^+ } + \frac{c_4}{\omega'+\varepsilon'+\varepsilon''+i0^+ } \right ) , \label{eq : dsigmab}\end{aligned}\ ] ] where @xmath289 , \\ c_2 &
= & -|u'v''+v'u''|^2 \left(|u'|^2+|v'|^2\right ) n_b'(\varepsilon ' ) + \frac{(u'u''^*+v'v''^*)(u'v''+v'u'')u'^*v'^ * + \text{c.c.}}{\varepsilon ' } [ n_b(\varepsilon')+n_b(\varepsilon'')+1 ] , \\
c_3 & = & -|u'u''^*+v'v''^*|^2 \left(|u'|^2+|v'|^2\right ) [ n_b(\varepsilon')-n_b(\varepsilon '' ) ] , \\ c_4 & = & |u'v''+v'u''|^2 \left(|u'|^2+|v'|^2\right ) [ n_b(\varepsilon')+n_b(\varepsilon'')+1].\end{aligned}\ ] ] [ ref : wcgeneralfertig97inbook ] for more discussion on the subject , see , for example , h. a. fertig , in _ perspectives in quantum hall effects _ , edited by s. das sarma and a. pinczuk ( wiley , new york , 1997 ) . [ ref : hennigan98 ] p. f. hennigan , a. beya , c. j. mellor , r. gaal , f. i. b. williams , and m. hnini , proceedings of the 12th international conference on the electronic properties of two - dimensional systems , 1998 .
[ ref : meanfield ] it should be noted that a static homogeneous mf solution is not always the right choice .
for example , if there is nesting , including an oscillatory part may give us a mean - field state that is lower in energy .
[ see i. affleck and j. b. marston , phys .
b * 37 * , 3774 ( 1998 ) .
] however , there is no reason to believe such is the case in our problem .
[ ref : li99private ] recently , the authors of refs . [ ref : li97 ] and [ ref : li98 ] have found that @xmath209 is actually at least one order of magnitude smaller than previously reported , if a lower power external ac field is used as a probe .
li ( private commuication ) ; j. yoon ( private communication ) ; d. c. tsui ( private communication ) .
] as of this writing ,
the line width has not saturated yet and the final result is most likely to be further reduced . | we investigate the microwave absorption of a pinned , two - dimensional wigner crystal in a strong magnetic field at finite temperatures . using a model of a uniform commensurate pinning potential , we analyze thermal broadening of the electromagnetic absorption resonance .
surprisingly , we find that the pinning resonance peak should remain sharp _ even when the temperature is comparable or greater than the peak frequency .
_ this result agrees qualitatively with recent experimental observations of the ac conductivity in two - dimensional hole systems in a magnetically induced insulating state .
it is shown , in analogy with kohn s theorem , that the electron - electron interaction does not affect the response of a harmonically pinned wigner crystal to a spatially uniform external field at any temperature .
we thus focus on anharmonicity in the pinning potential as a source of broadening . using a @xmath0 expansion technique
, we show that the broadening is introduced through the self - energy corrections to the magnetophonon green s functions . |
modified gravity @xcite is one of the two main roads one can follow in order to provide an explanation for the early and late - time universe acceleration ( the second one in the introduction of the dark energy concept @xcite ) .
furthermore , apart from the cosmological motivation , modified gravity has a theoretical motivation too , namely to improve the renormalizability properties of standard general relativity @xcite . in constructing a gravitational modification ,
one usually starts from the einstein - hilbert action and extends it accordingly .
thus , he can obtain @xmath2 gravity @xcite , gauss - bonnet and @xmath3 gravity @xcite , gravity with higher - order curvature invariants @xcite , massive gravity @xcite etc .
nevertheless , he could start from the equivalent , torsional formulation of gravity , namely from the teleparallel equivalent of general relativity ( tegr ) @xcite , in which the gravitational lagrangian is the torsion scalar @xmath4 , and construct various modifications , such as @xmath0 gravity @xcite ( see @xcite for a review ) , teleparallel gauss - bonnet gravity @xcite , gravity with higher - order torsion invariants @xcite , etc .
an important question in the above gravitational modifications is what are the forms of the involved unknown functions , and what are the allowed values of the various parameters .
excluding forms and parameter regimes that lead to obvious contradictions and problems , the main tool we have in order to provide further constraints is to use observational data .
for the case of torsional gravity one can use solar system data @xcite , or cosmological observations from supernovae type ia , cosmic microwave background and baryonic acoustic oscillations @xcite . on the other hand , in some modified cosmological scenarios one can obtain a variation of the fundamental constants , such as the fine structure and the newton s constants .
such a possibility has been investigated in the literature since dirac @xcite and milne and jordan @xcite times .
later on , brans and dicke proposed the time variation of the newton s constant , driven by a dynamical scalar field coupled to curvature @xcite , while gamow triggered subsequent speculations on the possible variation of the fine structure constant @xcite .
similarly , in recent modified gravities , which involve extra degrees of freedom comparing to general relativity , one may obtain such a variation of the fundamental constants @xcite .
however , since experiments and observations give strict bounds on these variations @xcite , one can use them in order to constrain the theories at hand . in the present work we are interested in investigating the constraints on @xmath0 gravity by observations related to the variation of fundamental constants . in particular , since @xmath0 gravity predicts a variation of the fine - structure and newton s constants , we will use the recent observational bounds of these variations in order to constrain the @xmath0 forms as well as the range of the involved parameters .
the plan of the work is the following : in section [ ftcosmology ] we give a brief review of @xmath0 gravity and cosmology . in section [ alpha ]
we investigate the constraints on specific @xmath0 gravity models arising from the observational bounds of the fine - structure constant variation , while in section [ g ] we study the corresponding constraints that arise from the observational bounds of the newton s constant variation .
finally , in section [ conclusions ] we summarize our results .
in this section we provide a short review of @xmath0 gravity and cosmology .
we use the tetrad fields @xmath5 , which form an orthonormal base at each point of the tangent space of the underlying manifold @xmath6 , where @xmath7 is the metric tensor defined on this manifold ( we use greek indices for the coordinate space and latin indices for the tangent one ) .
furthermore , instead of the torsionless levi - civita connection which is used in the einstein - hilbert action , we use the curvatureless weitzenbck connection @xmath8 @xcite .
hence , the gravitational field in such a formalism is described by the following torsion tensor : @xmath9 subsequently , the lagrangian of teleparallel equivalent of general relativity , namely the torsion scalar @xmath4 , is constructed by contractions of the torsion tensor as @xcite @xmath10 one may consider generalized theories in which the lagrangian @xmath4 is extended to an arbitrary function @xmath0 , similarly to the @xmath2 extension of curvature - based gravity . in particular , such gravitational action will read as @xmath11 where @xmath12 , and @xmath13 is the newton s constant .
additionally , along the gravitational action ( [ actionbasic ] ) we consider the matter sector , and hence the total action writes as @xmath14 where @xmath15 is the total matter lagrangian including the electromagnetic field .
finally , variation in terms of the tetrad fields give rise to the field equations as @xmath16 where @xmath17 , @xmath18 , and with @xmath19 the total matter energy - momentum tensor . in the above equation we have inserted for convenience the `` super - potential '' tensor @xmath20 defined in terms of the co - torsion tensor @xmath21 . applying @xmath0 gravity in a cosmological framework we consider a spatially flat flrw universe with line element @xmath22 $ ] , which arises from the diagonal tetrad @xmath23 , with @xmath24 the scale factor . in this case
, the field equations ( [ field - equations ] ) become @xmath25,\label{f2 .
1}\end{aligned}\ ] ] where @xmath26 and @xmath27 are respectively the total matter energy density and pressure , and where @xmath28 , @xmath29 are the effective energy density and the pressure of gravitational origin , given by @xmath30,\label{f1.1a}\\ p_t & = \frac{1}{16 \pi g_n}\ , \left [ 4 \dot{h}\ , \left ( 2 t f_{tt } + f_t- 1\right ) \right ] - \rho_t~. \label{f2.1b}\end{aligned}\ ] ] in the above expressions we have used that @xmath31 which arises straightforwardly from ( [ tscalar ] ) in an frw universe . in the following we focus on two well - studied , viable @xmath0 models , which correspond to a small deviation from @xmath1cdm cosmology , and which according to @xcite are the ones that fit the observational data very efficiently . * the first scenario is the power - law model ( hereafter @xmath32cdm ) introduced in @xcite , with @xmath33 where @xmath34 , @xmath35 are the two free model parameters , out of which only one is independent . inserting this @xmath0 form into the first friedmann equation ( [ f1.1 ] ) at present time ,
i.e. at redshift @xmath36 ( we use the redshift @xmath37 as the independent variable , and for implicitly we set @xmath38 ) , one may derive that @xmath39 where @xmath40 is the corresponding density parameter at present .
hence , and using additionally that @xmath41 , equation ( [ f1.1 ] ) for this model can be written as @xmath42^b + \omega_{m0}(1+z)^3 . \label{e2model1}\ ] ] lastly , we mention that the above model for @xmath43 reduces to @xmath1cdm cosmology , while for @xmath44 it gives rise to the dvali - gabadadze - porrati ( dgp ) model @xcite .
* the second scenario is the square - root - exponential ( hereafter @xmath45cdm ) of @xcite , with @xmath46 in which @xmath47 and @xmath27 the two free model parameters out of which only one is independent . inserting this @xmath0 form into ( [ f1.1 ] ) at present time
, one obtains that @xmath48 finally , the first friedmann equation ( [ f1.1 ] ) for this model can be written as @xmath49 \
, e^{-\frac{ph(z)}{h_0 } } \!-
\!1\right\ } \nonumber\\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ , = \omega_{m0}\,(1+z)^3 . \label{e2model2}\end{aligned}\ ] ] lastly , note that this model reduces to @xmath1cdm cosmology for @xmath50 . hence ,
in the following , for this model it will be convenient to set @xmath51 , and hence @xmath1cdm cosmology is obtained for @xmath52 .
in this section we will use observational data of the variation of the fine - structure constant @xmath53 , in order to constrain @xmath0 gravity . let as first quantify the @xmath53-variation in the framework of @xmath0 cosmology . in general , in a given theory the fine - structure constant is obtained using the coefficient of the electromagnetic lagrangian . in the case of modified gravities ,
this coefficient generally depends on the new degrees of freedom of the theory @xcite . even if one starts from the jordan - frame formulation of a theory , with an uncoupled electromagnetic lagrangian , and
although the electromagnetic lagrangian is conformally invariant , and thus it is not affected by conformal transformations between the jordan and einstein frames , it will acquire a dependence on the extra degree(s ) of freedom due to quantum effects @xcite .
in particular , if @xmath54 is the extra degree of freedom that arises from the conformal transformation @xmath55 from the jordan to the einstein frame , then quantum effects such as the presence of heavy fermions will induce a coupling of @xmath54 to photons , namely @xcite @xmath56 where @xmath57 is the electromagnetic tensor , @xmath58 the bare coupling constant , and @xmath59 with @xmath60 the planck mass and @xmath61 a constant ( we have assumed that @xmath62 ) . hence , the scalar coupling to the electromagnetic field will imply a dependence of the fine structure constant of the form @xcite @xmath63 where the subscripts denote the einstein and jordan frames respectively , or equivalently @xmath64 the above factor is in general time- ( i.e redshift- ) dependent .
therefore , it proves convenient to normalize it in order to have @xmath65 at present ( @xmath66 ) , which in case where @xmath67 is obtained through a rescaling @xmath68 and @xmath69 .
thus , we result to @xmath70 although the above procedure is straightforward in cases where a conformal transformation from the jordan to the einstein frame exists , it becomes more complicated for theories where such a transformation is not known . in case of @xmath0 gravity
, it is well known that a conformal transformation does not exist in general , since transforming the metric as @xmath71 , with @xmath72 a smooth non - vanishing function of spacetime coordinates , one obtains einstein gravity plus a scalar field lagrangian , plus the transformed matter lagrangian , plus a non vanishing term @xmath73 @xcite .
this additional term forbids the complete transformation to the einstein frame , and hence in every application one has indeed to perform calculations in the more complicated jordan one . in order to avoid performing calculations in the jordan frame
we will make the reasonable assumption that @xmath74 , which has been shown to be the case according to observations @xcite , and holds for the two forms considered in this work , namely ( [ model1 ] ) and ( [ model2 ] ) , too .
hence , @xmath75 , and then @xmath76 is negligible , which implies that the above extra term can be neglected . in the end of our investigation , we will verify the validity of the above assumption .
thus , we can indeed obtain an approximate transformation to the einstein frame , and in particular the introduced degree of freedom reads as @xmath77 @xcite .
hence , inserting this into ( [ bf ] ) we acquire @xmath78 and thus inserting into relation ( [ finestuct1c ] ) , we can easily extract the variation of the fine - structure constant as @xmath79 where @xmath80 .
lastly , since @xmath61 , the above relation becomes @xmath81 hence , for a general @xmath0 , the ratio @xmath82 indeed depends on @xmath83 , through the @xmath84 function ( we remind that according to ( [ th62 ] ) @xmath85 ) , while in the case of standard @xmath1cdm cosmology , where @xmath86 , @xmath87 becomes 0 .
.compilation of recent measurements of the fine - structure constant obtained by different spectrographic methods . for details in each case , see the corresponding references .
[ cols="^,^,^,^",options="header " , ] we mention that while analyzing the model for the data set of @xmath88 of table [ tab_data ] , we have marginalized over @xmath89 , and thus the statistical information focuses only on the parameter @xmath35 . for the fittings @xmath90 and @xmath91 we have taken @xmath92 as a free parameter , and we note that @xmath93 ( for @xmath94 @xmath95 snia ) and @xmath96 ( for @xmath94 @xmath95 snia @xmath95 bao ) at 1@xmath97 confidence level . finally , in fig .
[ f2_delta_alpha ] we present the 68.27@xmath98 and 95.45@xmath98 confidence regions in the plane @xmath99 , considering the observational data @xmath94 @xmath95 snia @xmath95 bao ( we have taken @xmath100 in order to avoid divergences in the function @xmath101 at high redshifts ) .
note that these results are in qualitative agreement with those of different observational fittings @xcite , and show that @xmath1cdm cosmology ( which is obtained for @xmath102 ) is inside the obtained region .
furthermore , and similarly to the @xmath103cdm model , from table [ f2_1 ] we deduce that although the data from @xmath104 alone show a slightly deviating nature ( reduced @xmath105 ) from @xmath1cdm scenario , but for @xmath104 @xmath95 snia and @xmath94 @xmath95 snia @xmath95 bao data it is implied that the model is very close to @xmath1cdm cosmology .
this is also a self - consistent verification for the validity of our assumption that @xmath74 , which allowed us to work in the einstein frame .
_ 1@xmath97 and 2@xmath97 confidence regions for the @xmath106cdm square - root - exponential model of ( [ model2 ] ) , obtained from the joint analysis @xmath94 @xmath95 snia @xmath95 bao .
the cross marks the best - fit value .
_ , width=288 ]
in this section we will directly use the observational constraints imposed on @xmath0 models in order to examine the variation of the gravitational constant @xmath13 . let us first quantify the @xmath13-variation in the framework of @xmath0 cosmology . as is well known ,
a varying effective gravitational constant is one of the common features in many modified gravity theories @xcite . in case of @xmath0 gravity ,
the effective newton s constant @xmath107 can be straightforwardly extracted as @xcite @xmath108 hence , for the @xmath103cdm power - law model of ( [ model1 ] ) , and using ( [ fcdm1 ] ) , we obtain @xmath109^{(b-1)}},\ ] ] where the ratio @xmath110 is given by ( [ e2model1 ] ) ( clearly , for @xmath43 we have that @xmath111 ) .
similarly , for the @xmath106cdm square - root - exponential model of ( [ model2 ] ) , and using ( [ fcdm2 ] ) , we acquire @xmath112\ , \left [ \frac { h_0}{h(z ) } \right ] e^{-\frac{ph(z)}{h_0}}},\ ] ] where the ratio @xmath110 is given by ( [ e2model2 ] ) ( clearly , for @xmath113 we have that @xmath111 ) .
let us now use the above expressions for @xmath114 and confront them with the observational bounds of the newton s constant variation .
we use observational hubble parameter data in order to investigate the temporal evolution of the function @xmath114 , since such a compilation is usually used to constrain cosmological parameters , due to the fact that it is obtained from model - independent direct observations .
we adopt @xmath115 observational hubble parameter data in the redshift range @xmath116 , compiled in @xcite , out of which 27 data points are deduced from the differential age method , whereas 10 correspond to measures obtained from the radial baryonic acoustic oscillation method .
we apply the following methodology : firstly , we estimate the error in the measurements associated with the function @xmath117 , for both models of ( [ gneffmodel1 ] ) and ( [ gneffmodel2 ] ) , via the standard method of error propagation theory , namely @xmath118 and we fix the free parameters of the two models within the values obtained in the joint analysis of @xcite and of section [ alpha ] of the current work .
then , the measurements of @xmath117 are calculated directly for each redshift defined in the adopted compilation . in the left graph of fig .
[ ft1_geff ] we depict the 1@xmath97 confidence - level estimation of the function @xmath119 from @xmath115 hubble data points , in the case of the @xmath103cdm power - law model of ( [ model1 ] ) . additionally , in the right graph of fig .
[ ft1_geff ] we present the corresponding 1@xmath97 confidence - level reconstruction of @xmath119 for @xmath35 @xmath120 $ ] from the observational hubble parameter data . when we perform the analysis within the known range of the parameter @xmath35 for this model ( from @xcite as well as from fig .
[ f1_delta_alpha_2 ] above ) , we find that @xmath121 . nevertheless , a minor deviation is observed for the fixed value of @xmath122 ( see the left graph of fig .
[ ft1_geff ] ) .
for instance , note that @xmath123 and @xmath124 , for the first and the last data points of the redshift interval [ @xmath125 , @xmath126 , respectively . in fig .
[ ft2_geff ] we present the corresponding graphs for the @xmath106cdm square - root - exponential model of ( [ model2 ] ) .
when we perform the analysis within the known range of the parameter @xmath127 for this model ( from @xcite as well as from fig .
[ f2_delta_alpha ] above ) , we find that @xmath121 , similarly to the case of @xmath103cdm model . in summary , from the analysis of this section
, we verify the results of the previous section , namely that the parameter @xmath35 that quantifies the deviation of both @xmath103cdm and @xmath106cdm models from @xmath1cdm cosmology , is very close to zero .
these results are in qualitative agreement with previous observational constraints on @xmath0 gravity , according to which only small deviations are allowed , with @xmath1cdm paradigm being inside the allowed region @xcite .
in the present work we have used observations related to the variation of fundamental constants , in order to impose constraints on the viable and most used @xmath0 gravity models . in particular , since @xmath0 gravity predicts a variation of the fine - structure constant , we used the recent observational bounds of this variation , from direct measurements obtained by different spectrographic methods , along with standard probes such as supernovae type ia and baryonic acoustic oscillations , in order to constrain the involved model parameters of two viable and well - used @xmath0 models . for both the @xmath103cdm power - law model , as well as for the @xmath106cdm square - root - exponential model , we found that the parameter that quantifies the deviation from @xmath1cdm cosmology can be slightly different than its @xmath1cdm value , nevertheless the best - fit value is very close to the @xmath1cdm one .
additionally , since @xmath0 gravity predicts a varying effective gravitational constant , we quantified its temporal evolution with the use of the previously constrained model parameters . for both @xmath103cdm and @xmath106cdm models , we found that the deviation from @xmath1cdm cosmology , is very close to zero .
these results are in qualitative agreement with previous observational constraints on @xmath0 gravity @xcite , however they have been obtained through completely independent analysis . in summary ,
@xmath0 gravity is consistent with observations , and thus it can serve as a candidate for modified gravity , although , as every modified gravity , it may have only small deviation from @xmath1cdm cosmology , a feature that must be taken into account in any @xmath0 model - building .
the authors would like to thank p. brax for useful discussions .
the work of sp was supported by the national post - doctoral fellowship ( file no : pdf/2015/000640 ) under the science and engineering research board ( serb ) , govt . of india .
this article is based upon work from cost action `` cosmology and astrophysics network for theoretical advances and training actions '' , supported by cost ( european cooperation in science and technology ) .
s. capozziello and m. de laurentis , phys .
rept . * 509 * , 167 ( 2011 ) .
e. j. copeland , m. sami and s. tsujikawa , int .
j. mod .
d * 15 * , 1753 ( 2006 ) ; y. f. cai , e. n. saridakis , m. r. setare and j. q. xia , phys .
rept . * 493 * , 1 ( 2010 ) .
a. naruko , d. yoshida and s. mukohyama , class .
grav . * 33 * , no . 9 , 09lt01 ( 2016 ) ; e. n. saridakis and m. tsoukalas , phys .
d * 93 * , no . 12 , 124032 ( 2016 ) . c. de rham , living rev .
* 17 * , 7 ( 2014 )
. s. h. chen , j. b. dent , s. dutta and e. n. saridakis , phys . rev .
d * 83 * , 023508 ( 2011 ) r. j. yang , eur .
j. c * 71 * , 1797 ( 2011 ) ; m. li , r. x. miao and y. g. miao , jhep * 1107 * , 108 ( 2011 ) ; m. h. daouda , m. e. rodrigues and m. j. s. houndjo , eur .
j. c * 72 * , 1890 ( 2012 ) .
y. p. wu and c. q. geng , phys .
d * 86 * , 104058 ( 2012 ) ; k. bamba , r. myrzakulov , s. nojiri and s. d. odintsov , phys .
d * 85 * , 104036 ( 2012 ) ; k. karami and a. abdolmaleki , jcap 1204 ( 2012 ) 007 ; v. f. cardone , n. radicella and s. camera , phys . rev .
d * 85 * , 124007 ( 2012 ) ; g. otalora , jcap * 1307 * , 044 ( 2013 ) ; j. amoros , j. de haro and s. d. odintsov , phys .
d * 87 * , 104037 ( 2013 ) .
k. bamba , s. capozziello , m. de laurentis , s. i .
nojiri and d. sez - gmez , phys .
b * 727 * , 194 ( 2013 ) ; s. bahamonde , c. g. bhmer and m. wright , phys .
d * 92 * , 104042 ( 2015 ) ; j. de haro and j. amoros , phys .
lett . * 110 * , no . 7 , 071104 ( 2013 ) ; a. paliathanasis , j. d. barrow and p. g. l. leach , phys . rev .
d * 94 * , no . 2 , 023525 ( 2016 ) .
l. iorio and e. n. saridakis , mon . not .
( 2012 ) 1555 ; m. l. ruggiero and n. radicella , phys .
d * 91 * , 104014 ( 2015 ) ; l. iorio , n. radicella and m. l. ruggiero , jcap * 1508 * ( 2015 ) no.08 , 021 ; g. farrugia , j. l. said and m. l. ruggiero , phys .
d * 93 * , no .
10 , 104034 ( 2016 ) .
g. r. dvali and m. zaldarriaga , phys .
lett . * 88 * , 091303 ( 2002 ) ; t. chiba and k. kohri , prog .
* 107 * , 631 ( 2002 ) ; l. anchordoqui and h. goldberg , phys .
d * 68 * , 083513 ( 2003 ) ; c. wetterich , phys .
b * 561 * , 10 ( 2003 ) ; e. j. copeland , n. j. nunes and m. pospelov , phys .
d * 69 * , 023501 ( 2004 ) .
m. d. c. bento , o. bertolami and n. m. c. santos , phys .
d * 70 * , 107304 ( 2004 ) ; v. marra and f. rosati , jcap * 0505 * , 011 ( 2005 ) ; p. p. avelino , phys .
d * 78 * , 043516 ( 2008 ) ; h. b. sandvik , j. d. barrow and j. magueijo , phys . rev .
lett . * 88 * , 031302 ( 2002 ) ; d. f. mota and j. d. barrow , mon . not .
* 349 * ( 2004 ) 291 ; j. d. barrow , d. kimberly and j. magueijo , class .
* 21 * , 4289 ( 2004 ) ; h. wei , phys .
b * 682 * ( 2009 ) 98 .
avelino , c. j. a. p. martins , n. j. nunes and k. a. olive , phys .
d * 74 * , 083508 ( 2006 ) ; e. garcia - berro , j. isern and y. a. kubyshin , astron .
astrophys .
* 14 * , 113 ( 2007 ) ; j. d. barrow , annalen phys . * 19 * , 202 ( 2010 ) ; t. chiba , prog . theor
. phys . * 126 * , 993 ( 2011 ) ; j. k. webb , j. a. king , m. t. murphy , v. v. flambaum , r. f. carswell and m. b. bainbridge , phys .
lett . * 107 * , 191101 ( 2011 ) ; j. a. king , j. k. webb , m. t. murphy , v. v. flambaum , r. f. carswell , m. b. bainbridge , m. r. wilczynska and f. e. koch , mon . not .
soc . * 422 * , 3370 ( 2012 ) .
p. molaro , d. reimers , i. i. agafonova and s. a. levshakov , eur .
j. st * 163 * , 173 ( 2008 ) .
h. chand , r. srianand , p. petitjean , b. aracil , r. quast and d. reimers , astron .
astrophys . * 451 * , 45 ( 2006 ) .
molaro _ et al .
_ , astron .
astrophys .
* 555 * , a68 ( 2013 ) . | we use observations related to the variation of fundamental constants , in order to impose constraints on the viable and most used @xmath0 gravity models . in particular , for the fine - structure constant we use direct measurements obtained by different spectrographic methods , while for the effective newton s constant we use a model - dependent reconstruction , using direct observational hubble parameter data , in order to investigate its temporal evolution .
we consider two @xmath0 models and we quantify their deviation from @xmath1cdm cosmology through a sole parameter .
our analysis reveals that this parameter can be slightly different from its @xmath1cdm value , however the best - fit value is very close to the @xmath1cdm one .
hence , @xmath0 gravity is consistent with observations , nevertheless , as every modified gravity , it may exhibit only small deviations from @xmath1cdm cosmology , a feature that must be taken into account in any @xmath0 model - building . |
the alpha magnetic spectrometer ( ams ) is an international project devoted to search for new physics phenomena , such as primordial antimatter or dark matter , by means of direct detection of high - energy cosmic - ray ( cr ) particles in space . the final version of the detector , ams-02 , was successfully installed and activated on the international space station on may 19@xmath1 2011 .
ams-02 is capable of measuring cr leptons and nuclei , from hydrogen up to iron , from hundreds mev up to @xmath21tev of energy , with unprecedent precision and sensitivity .
the ams-02 data are expected to significantly improve our understanding of the cr acceleration and propagation processes in the galaxy .
the status of ams measurements of cr leptons and light - nuclei is reviewed in these proceedings .
the layout of the ams-02 instrument and its complete description can be found in earlier papers @xcite .
the detector is composed of a silicon tracker , a transition radiation detector ( trd ) , a time of flight system ( tof ) , a permanent magnet , an array of anticoincidence counters ( acc ) , a ring imaging cherenkov detector ( rich ) , and an electromagnetic calorimeter ( ecal ) .
redundant measurements ensure a complete particle identification and allow for a careful study of interactions inside the detector @xcite .
the trd @xcite allows for lepton / hadron separation .
it is made of 5248 proportional chambers , filled with xenon and carbon - dioxide , arranged in 20 layers of fiber fleece radiators .
four tof planes of scintillators @xcite , located above and below the magnet case , allow for velocity measurements , charge measurements , and trigger .
the tracker is build of nine layers of silicon micro - strip detectors @xcite , from l1 to l9 , distributed over the instrument .
layers from l2 to l8 are assembled with the permanent magnet , that has a magnetic field strength of 0.15 t .
l1 is located on top of the trd , and l9 is located between between rich and ecal .
the tracker measures the rigidity @xmath3 up to @xmath42 tv ( @xmath5 , momentum to charge ratio ) for charge one particles , and charge measurement up to @xmath6 .
the rich @xcite is made of a radiator layer , a conical mirror , and a photomultiplier plane which detects cherenkov light .
two radiators , with naf and aerogel ( refractive indices @xmath7=1.34 and @xmath7=1.04 ) , are used to provide high - precision measurement of the particle velocities and charge .
the ecal @xcite is made of lead and scintillating fibers arranged in 16 layers , providing 17 radiation lengths of detecting medium .
this allows for a precise reconstruction of the energy up to several tev .
a boosted decision tree classifier ( bdt ) is used to identify leptons based on the topology of their showers .
onboard the iss , ams-02 is orbiting the earth at an altitude of about 400 km and with inclination of 51.6@xmath8 .
the average trigger rate is about 600 hz , the average event size is 2 kbyte .
the minimal down - link bandwidth is 9 mbit / s .
ams-02 collects about 1.5 billion crs during each month of operation .
the detector is controlled from the ams-02 payload operations control center ( pocc ) at cern , geneva .
intensive time - dependent calibrations have been performed to all sub - detectors .
no significant degradation of the sub - systems has been observed during 3 years of operation in the iss .
the new results on cr leptons are based on the data collected during the initial 3 years of operations on the iss ( may 2011 may 2014 ) , which is about 16% of the total expected data sample . in the measurement of leptons ,
the tof is used to select relativistic particles traversing ams-02 in the downward direction .
the signals in trd and ecal are used to discriminate the leptonic component from the hadronic background . a track reconstructed from l2 to l8 and matching the trd and ecal signals
is used to select clean @xmath9 particles . for the lepton / hadron separation ,
three methods are used : a trd - likelihood estimator , an ecal - bdt estimator , and the @xmath10 ratio by combining ecal and tracker .
this redundancy allowed to study the performance of each discrimination method with the data .
various analyses with different combinations of cuts or template fits were also tested as a cross - check .
an important source of background comes from _ charge - confusion _ of electron events , arising from the finite rigidity resolution of the tracker , or from the emission of secondary particles .
the charge - confusion contribution is studied from data and monte carlo simulations @xcite .
a high - energy measurement of the positron fraction ( from 1 to 500 gev ) has been recently reported @xcite .
the results are shown in fig.[fig::ccpositronfraction2014 ] .
the left panel shows the fraction up to 35 gev of energy compared with the standard conventional calculations where cr electrons are emitted from supernova remnants ( snrs ) and secondary @xmath0 arise from collisions of cr nuclei with the interstellar matter ( ism ) .
this model predicts that the positron fraction decreases with energy . above 10gev ,
the ams data show a persistent rise up to 200 gev , in clear contrast with the conventional picture , followed by an intriguing flattening at higher energies ( right panel ) .
light nuclei with @xmath11 constitute @xmath21% of the total cr flux .
rare elements such as @xmath12h , @xmath13he and li - be - b are believed to be of secondary origin , i.e. produced by collisions primary crs ( p , @xmath14he , c - n - o , or fe ) with the gas nuclei of the ism .
secondary to primary ratios such as @xmath13he/@xmath14he , b / c , or f / ne are used to discriminate among astrophysical models of cr propagation in the galaxy @xcite .
furthermore , the b / c ratio at @xmath15100gev / n represents a crucial tool to understand open problems in cr physics such as the observed structures in primary cr spectra @xcite or possible snr origin for the positron excess @xcite .
the measurement of the b / c ratio is a prime physics goal of ams-02 .
the analysis of the first 24 months of data has been presented in @xcite . with ams-02 ,
high purity samples of b and c events have been obtained using the tof and tracker charge estimators , with a charge selection efficiency of 98% and a charge mis - identification probability less than 10@xmath16 .
monte - carlo simulated events are produced using package and @xcite , including a description of detector geometry , electromagnetic and hadronic physics processes , as well as the full digitization chain in order to reproduce the signals of all sub - detector .
the several charge detectors above and below the inner core of the spectrometer have been used to improve the overall charge separation capability as well as to select clean data samples of non - interacting cr particles .
in particular , they allows for a data - driven characterization of nuclear interactions occurring at different levels in the spectrometers .
data and monte - carlo show an agreement at @xmath17 level .
figure[fig::ccbcratioicrc2013 ] ( left ) shows the l1 charge distribution obtained after selecting pure b samples in inner - tracker ( l2-l8 ) and tof .
the colored curves are the single - layer charge responses for known elements .
the data ( black markers ) are described by a combination of these templates using boron ( 94% ) , carbon ( 4% ) , and heavier elements ( @xmath21% ) .
these numbers represent the probability for top - of - instrument processes of charge - changing fragmentation , obtained with flight data . in this figure
, no cuts are applied to the trd signal .
simple cuts on the trd or l1 charge allows to highly - efficiently suppress these events . the measured b / c ratio is shown in fig.[fig::ccbcratioicrc2013 ] ( right ) as a function of the kinetic energy obtained from the tracker rigidity .
effect of bin - to - bin energy migration due to the finite resolution of the instrument has been accounted using an unfolding procedure .
the ratio decreases smoothly with energy .
the b / c behavior at tev energies will be become more transpaent with more data .
new ams results on the positron fraction show persistent rise up at @xmath210 200 gev of energy , in clear contrast with conventional model predictions , followed by an intriguing flattening at higher energies . from the measurement of the single spectra of @xmath18 and @xmath19 @xcite ,
the data show that both components @xmath19 and @xmath18 harden smoothly at @xmath1530gev .
these data are also consistent with the measured `` all - electron '' spectrum , @xmath20 , which has been independently determined up to @xmath211tev @xcite .
all these data indicate the existence of a new high - energy source of @xmath18 and @xmath19 , which can be interpreted either in terms of dark matter ( dm ) particles annihilation or in terms of nearby astrophysical sources of @xmath0 .
the dm interpretation require annihilating particles with mass of the order of @xmath21tev , and may predict possible signals in the @xmath22 ratio , depending on the specific dm
dm annihilation channels @xcite .
the existing antiproton data show no clear excess within the uncertainties in the data and in the model predictions . in the second class , a recent work demonstrated that the ams-02 data are described well in terms of @xmath0 pairs emission from pulsar wind nebulae ( pwne ) @xcite .
the pwn scenario gives no signatures in the antiproton channel .
it was also proposed that high - energy @xmath0 can be produced inside old - snrs via interactions of cr protons with the background medium @xcite .
this mechanism predicts remarkable signatures in both the @xmath22 ratio @xcite and the @xmath23 ratio @xcite , that may be detectable by ams-02 @xcite .
the data show no prominent signatures in the @xmath23 ratio @xcite .
the present models are affected by large astrophysical uncertainties that may be dramatically reduced with new data on cr nuclei @xcite .
this work is supported by acknowledged persons and institutions in @xcite and by the labex grant .
+ 00 aguilar , m. , et alii , tomassetti , n. , & oliva , a. , 2013 , 33@xmath1 icrc , 896 , rio de janeiro [ arxiv:1510.09215 ] saouter , p. , et alii , 2013 , 33@xmath1 icrc , 789 , rio de janeiro kirn , t. , et alii , basili , a. , et alii , alpat , b. , et alii , aguilar , m. , et alii , adloff , c. , et alii , accardo , l. , et alii , oliva , a. , 2013 , 33@xmath1 icrc , 1266 , rio de janeiro strong , a. w. , et alii , tomassetti , n. , [ arxiv:1204.4492 ] mertsch , p. , & sarkar , s. , tomassetti , n. , & donato , f. , [ arxiv:1203.6094 ] tomassetti , n. , & donato , f. , [ arxiv:1502.06150 ] agostinelli , s. , et alii , ranft , j. , aguilar , m. , et alii , aguilar , m. , et alii , boudad , m. , 2015 , a&a 575 , a67 [ arxiv:1410.3799 ] geng , c. q. , et alii , 2015 , prd 91 095006 [ arxiv:1411.4450 ] di mauro , et alii , blasi , p. , blasi , p. , & serpico , p. d. , tomassetti , n. , [ arxiv:1210.7355 ] maurin , d. , et alii , | the alpha magnetic spectrometer ( ams-02 ) is a particle physics experiment designed to study origin and nature of galactic cosmic rays ( crs ) up to tev energies from space . with its high sensitivity , long exposure and excellent identification capabilities
, ams is conducting a unique mission of fundamental physics research in space . to date , more than 60 billion cr events have been collected by ams .
the new results on cr leptons and the analysis and light - nuclei are presented and discussed .
the new leptonic data indicate the existence of new sources of high - energy cr leptons , that may arise either by dark - matter particles annihilation or by nearby astrophysical sources of @xmath0 pairs .
future data at higher energies and forthcoming measurements on the antiproton spectrum and the boron - to - carbon ratio will be crucial in providing the discrimination among the different scenario . |
the transition metal compound @xmath6 ( x@xmath11/3 , 3/4 and y@xmath10 , 2/3 , 4/3 ) has an appreciable , quasi - metallic conductivity and still shows aspects of strong electronic correlations .
there is broader consensus that these correlations play an important role for the recently discovered superconductivity with @xmath7=4.6 k @xcite , the very large thermopower , and other anomalous transport properties @xcite . the two - dimensional structure of @xmath8 is given by an incoherent coupling of @xmath9 layers with na layers stacked along the c axis @xcite .
the resulting misfit and pronounced strain effects allow , similar to other misfit - layered oxides , for a considerable non - stoichiometry both on the cation and the oxygen sites and an inhomogeneity in charge distribution along the stacking direction of the compound @xcite . in @xmath8 , with [email protected] , two partly occupied na sites on a honeycomb lattice alternate along the c axis with a @xmath9 layer of edge sharing @xmath10 octahedra @xcite .
ionic conduction based on high na mobility exists at room temperature .
the octahedra in the @xmath9 layers are tilted and have only two oxygen coordinates per layer along the c axis .
these oxygen sites contain the single co site . in the ab plane the co sites form a planar triangular lattice , from geometrical point of view the perfect base for competing magnetic interactions . in figure 1
a ) and b ) two projected views on the layered structure of @xmath8 are shown .
the na sites in a ) are only partially occupied .
c ) electronic level scheme of @xmath11 with a sketch of two distorted @xmath10 octahedra @xcite .
the two sites na1 and na2 have an approximate occupation 1/4 and 1/2 , respectively .
the line gives the unit cell .
the dashed line marks a distortion of the octahedron.,height=377 ] samples of @xmath8 with a na content 0.3@[email protected] have been reported .
this stoichiometry range corresponds to an average co valency between @xmath13 and @xmath14 .
the @xmath15 and @xmath16 configurations of @xmath5 and @xmath17 occupy @xmath4 levels in a low spin state with s=1/2 and s=0 , respectively .
therefore , the above mentioned triangular lattice of co sites with s=1/2 is far from half - filled , i.e. the spins are considerably diluted with s=0 states . in figure 1
c ) the corresponding cef level schemes are shown together with the local coordination of two distorted @xmath10 octahedra with the two na sites , na1 and na2 .
relevant states at the fermi level are given by @xmath3 states of @xmath15 in @xmath5 that are split off from the other @xmath18 levels @xcite .
this split off is due to a combination of a trigonal distortion of the octahedra and kinetic effects @xcite .
the electronic state of @xmath8 shows some analogy to strongly doped high temperature superconductors with the additional very attractive aspect of frustration @xcite .
such a triangular lattice is expected to stabilize a resonance valence bond state better than the well - studied @xmath19 square lattice .
it also leads to three nesting vectors that could have important implications for the superconducting order parameter @xcite .
single band hubbard models have been proposed as a reasonable simplification taking the split off @xmath3 states as an underlying basis @xcite . from experimental point of view
the role of electronic correlations and the minimal low - energy model is not clear @xcite , as the related mott - hubbard insulating phase of @xmath8 with 1/2 doping has never been prepared or observed and important electronic parameters could not be determined unambiguously .
the magnetic susceptibility @xmath20 of @xmath8 shows a curie - weiss behavior with a negative curie constant @xmath21=-170 k for x=0.75 , while @xmath20 is less temperature dependent for smaller x , i.e. @xmath21 is decreasing with increasing @xmath22 .
this composition dependence might be related either to spin frustration and/or additional ferromagnetic correlations . only in samples with highest na content ( x=0.75 ) long - range magnetic ordering with ferrimagnetic or commensurate sdw ground state is observed with a @xmath7=22 k @xcite . in systems with smaller x long - range ordering
can only be induced by defects , as cu substitution on the co site @xcite .
interestingly , the flat magnetic susceptibility for x=1/3 has been compared with tiocl , a two - dimensional quantum spin system with spin gap formation due to orbital ordering and strong spin - lattice coupled fluctuations @xcite .
@xmath8 has an enormous thermopower and excellent thermoelectric figure of merit @xcite that are comparable to the best broad - band semiconductors @xcite . these effects origin from the @xmath9 layers as these layers are common structural elements for other known thermoelectric cobaltites , as @xmath23 and @xmath24 .
the thermopower strongly depends on the ratio @xmath25 .
it increases with increasing x , i.e. for smaller @xmath25 and closer proximity to magnetic ordering @xcite . applying a moderate longitudinal magnetic field in the ab plane
it can be strongly depressed @xcite .
the large magnetic field dependence rules out its origin in fermi surface anomalies . instead , the combination of a large spin / electronic degeneracy and hopping on a triangular , frustrated network has been suggested to be the origin of the large thermopower @xcite .
furthermore , the linear temperature dependence of the hall coefficient at high temperatures and the linear metallic - like conductivity conductivity at low temperatures are both proposed to be related to electronic correlations on such a triangular lattice @xcite .
if the na content is strongly reduced to x=0.33 - 0.35 the resulting samples are extremely hygroscopic and loosely bind @xmath2 into two intercalated layers between the na and the @xmath9 layers .
the result is @xmath26 with a nearly doubled c axis parameter @xcite .
this systems shows superconductivity with @xmath7=4.6 k and considerable experimental and theoretical evidence for an unconventional superconducting order parameter exist @xcite .
initial investigations show a complex defect chemistry that allows reversible changes of the hydration level even at room temperature @xcite . on the other side
it can safely be assumed that strain related to incoherent coupling of @xmath9 layers to the partly occupied na square planes is partly released by hydration .
therefore it might also have an effect on the electronic properties of @xmath27 . at this point more explicit experimental information on the local site symmetries is highly desirable .
as function of na content x and hydration y at 10k and 295 k in the upper and lower curves , respectively .
the curves are shifted for clarity.,height=491 ] the existence range of the superconducting phase is very narrow 0.25@[email protected] @xcite or has proposed to even be a single point of composition ( x=1/3,y=4/3 ) @xcite .
the former interval motivated theoretical investigations about possible competing charge ordered instabilities at the phase boundaries @xcite .
these phases would correspond to a honeycomb and a kagome lattice of s=1/2 , respectively . until now no experimental evidence for such phases or the corresponding transitions exists .
it should also be noted , that the electronic structure of @xmath8 and the hydrated variant can be considered as kagome - like even without charge ordering @xcite .
the superconducting properties of @xmath27 resemble those of high temperature superconductors .
the system is a type ii superconductor with the critical fields @xmath28=28 oe and @xmath29=61 t. the coherence lengths are @xmath30=2.3 nm and @xmath31=570 nm with @xmath32=250 @xcite .
the large critical field @xmath29=61 t is exceptional if compared to the small transition temperature .
the anisotropy in transport increases with hydration from @xmath33=200 to 1000 .
the specific heat shows a maximum at the transition @xmath34 and an electronic term @xmath35 leading to a reduced jump of @xmath36/\gamma = 0.85 $ ] considerably smaller than the bcs value ( 1.43 ) @xcite . in another specific heat study a @xmath37 contribution to the specific heat is observed and attributed to line nodes of the superconducting order parameter @xcite .
here , the other superconducting parameters differ considerably from above : @xmath38 , @xmath39 and @xmath36/\gamma = 1.31 $ ] . in these data
a larger non - superconducting volume fraction has been taken into account and @xmath40 corrected accordingly .
recent nmr / nqr data show a similar lack of consistency . in one study
the co - knight is t - independent above @xmath7 and decreases moderately for smaller temperatures . together with the coherence peak in nmr
this would signal a singlet order parameter with a full superconducting gap @xcite .
another experimental study highlights the negligible temperature dependence of the knight - shift that is taken as evidence for spin - triplet state of superconductivity in @xmath27 @xcite . these and other experimental studies are strongly hampered by the easy degradation of the sample and disorder on the na site .
the resulting small superconducting volume fraction does not allow a sample assembly into a vacuum setup at rt .
in the present raman scattering study special care has been taken to avoid these problems using a setup that allows rapid cooling down in a helium contact gas or performing the experiments at room temperature in a sample cell with humidified atmosphere .
thereby we searched for possible effects of charge ordering or other instabilities and used phonon scattering due to oxygen vibrations as a sensitive probe of the local electronic and structural configuration .
we have performed raman scattering experiments on powder and single crystal samples of @xmath6 as function of na content x and hydration y. ( xx ) light polarization has been investigated with the incident and scattered electric field vector x in the ab plane of the crystallographic structure .
powder samples were prepared as described elsewhere @xcite and cold pressed into tablets .
single crystals of @xmath41 were prepared using a travelling solvent floating zone optical furnace . in a following
preparation step na was deintercalated using a bromine or an electrochemical preparation step @xcite .
the final hydration and/or equilibration has been performed in humidified air @xcite .
a ) freshly cleaved single crystal , b ) after 2 days in humidified air , c ) after 6 hours in helium exchange gas , d ) after 3 hours and e ) after 6 hours in vacuum .
the spectra are normalized in intensity to the mode at 470 @xmath42 . , height=377 ]
a symmetry analysis taking into account the @xmath43 ( 194 ) point group for @xmath41 @xcite and @xmath6 @xcite leads to the following raman - active modes @xmath44 and the infrared - active modes @xmath45 each of the @xmath46 , @xmath47 or @xmath48 modes are doubly degenerate . in this analysis we skipped modes due to hydration of the samples , as no experimental evidence for such modes exist . in the following we will also not discuss anymore the @xmath49 modes related to na and oxygen .
these modes would be observable in ( xz ) polarization @xcite .
the two na related modes are expected to be smeared out due to disorder and the small occupation of the na sites . for the considered modes at the @xmath50-point the displacement symmetries are @xmath51 for a displacement in the @xmath52 direction and @xmath46 for in - plane , diagonal @xmath53 displacements .
the @xmath51 and @xmath46 modes involve vibrations from only oxygen atoms . due to full point - group symmetry
the co sites do not contribute to raman scattering .
we have performed lattice shell model calculations of the phonon frequencies in @xmath54 @xcite . to accommodate for the partial occupation of the two na sites , na1(0.24 ) and na2(0.5 ) , the phonon frequencies for either a fully occupied site na1 ( or a fully occupied na2 ) site have been calculated .
the frequency of the e@xmath55 and a@xmath56 modes are 458 @xmath42 ( 457 @xmath42 ) and 586 @xmath42 ( 574 @xmath42 ) , respectively , i.e. the higher energy a@xmath55 mode strongly depends on the na site occupancy . as mentioned above the e@xmath56 mode is an in - plane oxygen mode with diagonal displacements while the a@xmath55 mode is an out - of - plane mode .
this explains the strong sensitivity of this excitation to the na content in the layers that divide the oxygen octahedra in c axis direction . in the following
we will show that the pronounced frequency dependence of the a@xmath55 mode can be used as a very susceptible sensor of na distribution / ordering . in ( xx ) polarization at room temperature .
the inset displays spectra on an enlarged scale together with fits . for x
, y=0.7/0.0 three individual modes are resolved that overlap and form a broad band .
, height=340 ] raman spectra of polycrystalline samples as given in figure 2 show two sharp modes at 480 @xmath42 and 598 @xmath42 .
these modes correspond to the e@xmath55 and the a@xmath55 eigenmodes , respectively .
the e@xmath56 mode hardens due to anharmonicity with decreasing temperature .
the linewidth of the a@xmath55 mode is substantially broadened at higher temperature .
this must be related to pronounced na diffusion and disorder . at higher energies
two very broad maxima exist at 800 and 1000 @xmath57 .
these signals are understood as two - phonon scattering .
it is noteworthy , that the temperature dependence of this signal is most pronounced for samples that show superconductivity . reducing the hydration to zero , drastic effects happen at lower frequency .
a broadened , quasi - elastic scattering is suppressed .
this continuum might be related to quasi - diffusive excitations of @xmath2 molecules . a second possibility are local paramagnetic fluctuations of the spin system .
such paramagnon scattering has been discussed for high temperature superconductors or other low - dimensional quantum spin systems @xcite .
the continuum observed in @xmath6 resembles to early observations in @xmath58 @xcite and would correspond to an exchange coupling constant of approximately 130 k. its highest intensity is observed for y=0.7 , i.e. in the sample with one @xmath59 layer .
other sharp modes that do not show a consistent stoichiometry dependence , e.g. at 240 and 700 @xmath42 , are attributed to defect modes .
raman scattering experiments on single crystal surfaces do in general provide a better and more consistent view on the excitation spectrum of a compound . in the case of @xmath6 , however , due to the complex defect chemistry given by easy hydration loss , high mobility of na and its hydroxide formation special precautions have to be taken . to investigate such effects results of a helium gas and vacuum exposure study are shown in figure 3 .
a similar study on @xmath41 can be found in ref .
the initial state of our @xmath26 single crystal ( curve a ) has been prepared by a freshly cleaved surface .
the sample has then been stored and investigated in humid air for two days ( b ) and then mounted into a cryostat with helium contact gas .
consecutive spectra have been taken after a storage at rt in dry helium gas ( c ) and later under vacuum ( d and e ) .
these steps were interrupted by a cooling down to 200 k or below with cooling rates of 2 - 4 k / min .
it is evident from these data that the intrinsic phonon linewidth of the fully hydrated , single crystalline @xmath6 is very small .
however , surface properties of the crystals show a degradation even under humidified air . therefore , the linewidth of the c - axis @xmath51 mode broadens .
the following reduction of hydration in dry helium gas leads to a moderate decrease of the in - plane @xmath46 mode by about 3 @xmath57 .
most drastic is the effect of a vacuum treatment .
the observed shift by 9 @xmath42 is exceptionally large and as the related mode is an in - plane oxygen vibration can not fully be accounted by the loss of hydration .
therefore we propose a loss of oxygen as the most probable origin of the phonon frequency shift in vacuum . to figure out the effect of hydration on the `` parent compound '' ( xx ) polarized raman spectra of single crystals at t=295 k
are given in figure 4 .
in addition to two - phonon scattering with a maximum at 1170 @xmath42 and modes in the anti - stokes regime at -470 and -590 @xmath42 for x=0.3 and 0.7 , respectively , there are remarkable changes of the phonon modes in the frequency regime from 450 - 650 @xmath42 .
@xmath41 shows a strongly broadened band - like scattering intensity that has a maximum at 588 @xmath42 , very close to the frequency calculated in a lattice shell model with fully occupied na1 site . in @xmath60 , in stead , two modes are observed at 582 and 469 @xmath42 .
the inset shows a fit using lorentzians to these intensities .
three modes with different frequencies can be fitted to the broad band of @xmath41 .
we interpret the broad band of scattering in @xmath41 as due to a considerable strain and possible sublattice formation of the not fully occupied na1 and na2 sites . with decreasing temperature a part of this broadening
is reduced , as e.g. also seen in the experiments on polycrystalline samples in figure 2 .
ordering processes have been evidenced from @xmath61 in the temperature regime 250 - 300 k and attributed to a charge disproportionation of the effective @xmath62 into @xmath63 @xmath5 @xcite . the sensitivity of the c - axis @xmath64 mode frequency on na site occupation and the high mobility of na at room temperatures , however , do not support this interpretation . in the superconducting , hydrated system the phonon spectrum is not broadened
although the na sublattice is even more diluted . therefore we conclude , that the intercalated and well - ordered @xmath2 layers @xcite shield the related disorder .
as the observed phonon frequencies correspond to in - plane and out - of - plane eigenmodes of the @xmath10 octahedra and their trigonal distortion determines the above discussed co @xmath3 level energy , an effect of disorder on this level and the density of states at @xmath65 @xcite is unquestionable .
the observed broadenings and frequency shifts therefore suggest that the narrow superconducting phase space as function of x @xcite and the reduced superconducting volume observed in specific heat measurements @xcite are rather connected with na disorder than with the proposed charge ordering instabilities that compete with superconductivity @xcite .
further well - controlled studies as function of stoichiometry , disorder and temperature are underway .
our raman scattering investigations on the superconducting cobaltites @xmath0 as function of na content and hydration show pronounced effects related to partial occupation and disorder on the na site .
in contrast , no evidence for charge ordering has been found .
the strong sensitivity of the transition temperature on local lattice distortions is proposed to origin from a local modulation of the @xmath66 split off from the remaining co @xmath4 levels . these effects support models of superconductivity based on electronic correlations .
the authors acknowledge fruitful discussions with r. kremer , i. nekrasov , v.i .
anisimov , r. zeyer , p. horsch , and g. khaliullin .
this work was supported by dfg spp1073 , nato pst.clg.9777766 , intas 01 - 278 , crest of jst ( japanese science and technology corporation ) and the mrsec program of the national science foundation under award number dmr 02 - 13282 .
10 takada k , sakurai h , takayama - muromachi e , izumi f , dilanian r a and sasaki t 2003 _ nature _ * 422 * , 53 | we report on a raman scattering study on the superconducting cobaltite @xmath0 as function of na content and hydration ( x@xmath11/3 , 3/4 and y@xmath10 , 2/3 , 4/3 ) .
the observed phonon scattering and scattering continua are analyzed in terms of lattice strain due to the structural misfit and disorder .
hydration , due to the intercalation of one or two @xmath2 layers , releases a part of this strain .
our raman data suggest a connection between disorder on the partly occupied na sites , the split off of the @xmath3 level from the other @xmath4 states of @xmath5 and superconductivity . |
the celebrated ads / cft correspondence @xcite has promoted the solution of string theory in anti - desitter ( ads ) spaces to one of the central problems of modern mathematical physics .
progress in this direction requires to construct new types of quantum field theories with internal lie superalgebra symmetries .
the precise model to be considered depends on the particular approach that is employed .
recent investigations have been based on certain gauge fixed versions of the green - schwarz superstring @xcite , the pure spinor formalism @xcite and the hybrid formalism @xcite . without much further comment on the precise relation with string theory ( see some remarks below , however )
, we shall turn our attention to a particular class of quantum theories with internal supersymmetries , namely to non - linear sigma models on supergroups .
they are characterized by the following simple action @xmath3\ = \
-\frac{1}{2\pi f^2 } \int_{\sigma } \!d^2z \;{\text{str}}\left(s^{-1}\partial s s^{-1 } \bar{\partial}s \right ) -\frac{k}{12\pi } \int_{\sigma}\ ; d^{-1}{\text{str}}\left(\left(s^{-1}ds \right)^3 \right)\ ] ] with a suitably normalized supertrace @xmath4 . here
, @xmath5 is a map from the world - sheet @xmath6 to some supergroup @xmath7 .
we have weighted the standard kinetic term with a coupling constant @xmath8 and also added a topological wess - zumino ( wz ) term with coefficient @xmath9 . for sigma models on bosonic groups , quantum conformal
invariance requires @xmath10 .
once we have adjusted the coupling constants in this way , we are dealing with a wess - zumino - novikov - witten ( wznw ) theory which can be solved using the algebraic techniques of 2-dimensional conformal field theory , exploiting the infinite dimensional current algebra symmetry of the wznw model .
it is one of the intriguing features of certain supersymmetric target spaces that the requirement of quantum conformal invariance may not impose any restriction on @xmath11 , see e.g.@xcite .
this happens whenever the supergroup @xmath7 has vanishing dual coxeter number .
the latter condition is satisfied e.g. for the superconformal groups psl(n@xmath12n ) that appear in the ads / cft correspondence , but also for osp(2n+2@xmath12n ) and @xmath13 . in these cases ,
the action gives rise to a continuous family of conformal quantum field theories .
all models share the same global target space symmetries . on the other hand ,
the wz point with @xmath10 is still distinguished by an enhancement of world - sheet symmetries . for generic values of @xmath14 ,
one only expects to find a few chiral higher spin fields in addition to the virasoro symmetry that comes with conformal invariance ( see @xcite for details ) .
whatever the precise chiral symmetry is , it will almost certainly not suffice for a full algebraic solution of generic supergroup sigma models .
this insight has lead many scientists working in the field to discard conformal field theory techniques and to turn to other methods in integrable systems , such as the bethe - ansatz and generalizations thereof .
though ultimately , computations in superspace sigma models may involve a variety of integrable techniques ( see e.g.@xcite for an incomplete collection of recent relevant ideas , a few results and many further references , in particular to the earlier literature ) , it seems to us that the real potential of conformal field theory methods has not been explored with sufficient care .
in fact , we shall see below that a combination of algebraic techniques with conformal perturbation theory can provide powerful new results going far beyond the wz point . to be more precise ,
we propose to consider the sigma models as deformations of a wznw model , @xmath15 \ = \ { \mathcal}{s}_{k}^{{{\text{wznw}}}}[s ] \ - \
\frac{\lambda}{2\pi } \int_{\cal h } \!d^2z \;{\text{str}}\left(s^{-1}\partial s s^{-1 } \bar{\partial}s \right ) \ = \ { \mathcal}{s}_{k}^{{{\text{wznw}}}}[s ] + { \mathcal}{s}_{\lambda}[s]\ \ .\ ] ] the deformation parameter @xmath16 is related to @xmath9 and @xmath14 through @xmath17 . for reasons to be explained below
, we shall often refer to this deformation of the wznw model as a `` rr - deformation '' .
note , however , that on the level of sigma models it simply changes the overall scale factor of the metric while leaving the magnetic background field invariant .
our approach is then to study the sigma model through conformal perturbation theory around the wz point . in this note
we restrict our attention to the simplest objects , namely to partition functions , leaving investigations of correlators etc . as an interesting problem for future research . in order to explain our strategy ,
let us briefly look at simple torus compactifications .
suppose we are interested e.g. in the spectrum of strings on a 1-dimensional circle with arbitrary compactification radius @xmath18 . at generic points in the 1-dimensional moduli space ,
the chiral symmetry of the model is generated by the u(1 ) current @xmath19 and its anti - holomorphic counterpart .
with respect to these currents , the theory is not rational .
but there exist some distinguished points in the moduli space at which the chiral symmetry is enhanced and the theory becomes rational once the additional chiral fields are taken into account . in particular
, the moduli space contains one point , known as the self - dual radius @xmath20 , where the symmetry gets enhanced to an @xmath21 current algebra at level @xmath22 . at this special radius
, all spectra can be composed from a finite number of sectors . with later generalizations in mind ,
we consider the partition function on a strip or half - plane with neumann boundary conditions which is simply given by the vacuum character of the @xmath21 current algebra @xmath23 other points in the moduli space may be reached through a deformation with the perturbation @xmath24 .
the perturbation series for the conformal dimensions of boundary fields can be summed up to all orders in perturbation theory .
our partition function gets deformed to @xmath25 the result corresponds to the spectrum of a point - like brane on a circle with radius @xmath26 . in the perturbative treatment ,
the factor @xmath27 arises from a geometric series as explained e.g. in @xcite .
bulk spectra can also be computed , either directly or through modular transformation of the boundary partition function .
let us point out that the perturbative analysis is insensitive to the fact that the theory ceases to be rational once we move away from the self - dual radius .
of course , in this particular case the u(1 ) current algebra symmetry is sufficiently large for an algebraic construction of the theory at generic radii and such a construction is about as difficult as it is at the self - dual point .
hence , there is no good motivation to pass through a perturbative construction .
but there exists a better example to illustrate the enormous potential conformal perturbation theory may possess .
it is provided by the 1-dimensional boundary sine - gordon theory . in this model , a periodic potential is switched on along the boundary of a free field theory . as a consequence , the spectrum of boundary dimension develops gaps which can grow with the strength @xmath16 of the perturbation .
eventually , only a point - like spectrum remains .
given the complexity of the spectrum at intermediate values of @xmath16 , one might suspect that its precise form is very difficult to determine . yet
, the boundary partition function can be calculated rather easily in perturbation theory @xcite , for any value of the deformation parameter @xmath16 .
in this example , the boundary potential reduces the chiral symmetry to the virasoro algebra . in principle , the latter is still sufficiently large to allow for a standard cft construction of the boundary sine - gordon theory , but such an analysis is of the same level of difficulty as the solution of liouville theory and it has never been carried out .
hence , the example of boundary sine - gordon theory supports our claim that in some situations , conformal perturbation theory provides an easy route to complicated results that seem ( almost ) inaccessible through the usual algebraic methods . a similar picture will emerge from our study of boundary spectra on supergroup @xmath0-models .
even though most of the ideas and technical steps we are about to explain hold quite generally , we shall carry them out in a particular example , namely for the supergroup @xmath1 .
this allows our presentation to be very concrete .
furthermore , our results apply to string theory in @xmath28 whose solution has been reduced to the construction of sigma models on the supergroup @xmath1 through the hybrid approach developed by berkovits , vafa and witten @xcite . in this context , the wznw model corresponds to a background with pure nsns 3-form flux .
switching on an additional rr field is modelled by the marginal perturbation with @xmath29 which is why we often refer to this term as rr - deformation .
sigma models on @xmath1 and closely related target superspaces have been investigated by several groups @xcite . for our analysis , the studies by bershadsky et al .
have been particularly useful . with the example of strings in @xmath2 in mind , we may re - evaluate our optimistic hopes to compute exact spectra through perturbation theory .
let us think of the target space as a 3-dimensional solid cylinder . since @xmath2 is curved , the corresponding sigma model is interacting . at the wz point
, the interaction falls off exponentially towards the boundary of the cylinder .
this has several effects on the bulk spectrum .
in particular , the spectrum is continuous and there exist so - called long string states that can stretch along the boundary @xcite
. the rr - deformation now adds another term to the interaction which increases exponentially near the boundary .
obviously , such a new term must have drastic effects on the spectrum .
certainly , long string states disappear .
in addition , the spectrum is expected to become discrete since closed strings are now moving in a box between the two exponential walls .
the dramatic effects of the rr - deformation may raise doubts that perturbative computations could be successfully performed . and
indeed , it is most likely true that the bulk spectrum of the theory is not amenable to a perturbative expansion in @xmath16 .
but the situation changes if we consider the boundary spectrum @xcite on a d - instanton instead .
suppose , the instanton has been placed at the center of the solid cylinder .
open strings that end on such a d - instanton must be very highly excited in order to penetrate into the region close to the boundary where the rr - background flux can be felt .
therefore a d - instanton spectrum might be accessible through a perturbative computation .
below we shall see that this intuition is indeed correct .
in fact , we are able to determine the exact spectrum of a d - instanton for any value of @xmath16 .
the same calculation fails at one step when we try to apply it to the bulk or to spectra on non - compact branes .
our main new result is a complete computation of the boundary spectrum for maximally symmetric , point - like branes in sigma models on the supergroup @xmath1 .
the partition function of such a system was argued in @xcite to be of the general form @xmath30 & = & \sum_{j_1\neq j_2}\ , a^\lambda_{j_1j_2}(q ) \ , \chi_{[j_1,j_2]}(z_1,z_2 ) + \ \sum_{j}\ , b^ \lambda_j ( q ) \,\chi_{{\cal p}[j]}(z_1,z_2)\ \ .
\nonumber\end{aligned}\ ] ] here , @xmath31 and @xmath32 are two cartan elements in the bosonic subalgebra @xmath33 of @xmath34 and we have denoted the characters of the contributing finite dimensional @xmath34 representations by @xmath35 ( see appendix a for explicit formulas ) .
the branching functions @xmath36 and @xmath37 at the wz point @xmath38 were also determined in @xcite .
our aim in this work is to show that the branching functions @xmath36 are independent of the deformation parameter @xmath16 while @xmath39 let us already point out that the dependence of the conformal weights on the deformation parameter @xmath16 is very similar to the one found in free field theory ( see eq . ) .
we shall see that this is due to some peculiar features of the lie superalgebra @xmath34 .
our formulas and contain a surprising wealth of information .
let us unravel some of that through a few selected cases .
consider , for example , the boundary current @xmath40 where @xmath41 runs through some @xmath42-dimensional basis of @xmath34 .
under the action of the global @xmath34 , the currents transform in the adjoint representation which is part of the atypical module @xmath43 $ ] ( see appendix a ) .
since the branching functions @xmath36 are independent of @xmath16 , states transforming in any of the @xmath44 $ ] do not receive corrections .
hence , the currents @xmath45 continue to possess dimension @xmath46 , as expected .
things become more interesting once we proceed to products @xmath47 of currents .
these form a @xmath48-dimensional subspace of fields transforming in the @xmath49-dimensional representations @xmath50 , [ 1,0]$ ] and various subspaces of @xmath51 $ ] .
hence , under the deformation , the weight of @xmath52 fields gets lifted while @xmath53 fields remain at conformal weight @xmath54 .
formula passes a few interesting test . to begin with
, we observe that the energy shift is positive for states with sufficiently large momentum @xmath55 in the radial direction of @xmath2 .
this is in line with our geometric intuition : only states that are highly excited in the radial direction can penetrate to the region near the boundary of @xmath2 where their energy gets lifted due to the rr perturbation .
it is also interesting to evaluate our formula in the semi - classical regime , i.e. for large values of the level @xmath9 .
inserting the relation @xmath56 in and sending @xmath9 to infinity , the spectrum of boundary conformal weights is seen to coincide with the spectrum of @xmath57 up to the usual integer shifts .
the eigenvalues of @xmath58 may be interpreted as energies for a particle moving on @xmath1 .
hence , at large level @xmath9 and modulo integers , the spectrum of the sigma model on @xmath1 agrees with the minisuperspace approximation , as it is supposed to . a much more detailed investigation of the approach to minisuperspace spectra for supersphere models is included in a very interesting upcoming paper by candu and saleur @xcite .
the plan of this work is as follows . in the next section
we collect some background material , partly from our earlier paper @xcite .
this includes a careful discussion of maximally symmetric , point - like branes in the wznw model on @xmath1 .
the ones that are relevant for our analysis are located at the group unit @xmath59 of the bosonic base and they extend in all eight fermionic directions .
the associated boundary partition function is discussed in section 2.2 along with more details on the casimir decomposition at the wz point .
section 2.3 contains a construction of the perturbing field in terms of currents and a new proof of its exact marginality .
most of our novel results are obtained in section 3 which begins with a few comments on 2-point functions .
section 3.2 lists several observations concerning the perturbative series generated by @xmath29 .
we shall show that the rr - deformation , while being non - abelian and non - constant on @xmath1 in general , simplifies drastically in the evaluation of @xmath34 invariant quantities , such as conformal weights .
in fact , the rr - deformation turns out to be quasi - abelian , i.e. its combinatorics is no more complex than it is for constant shifts of the closed string background fields in a flat target space .
there remains a mixing problem , however , that we can only overcome when the general results are applied to boundary conformal weights of a point - like d - instanton .
this is explained in section 3.3 before we combine all our results into an exact computation of boundary weights , following closely the steps of a similar computation in @xcite .
our concluding section includes extensive comments on possible generalizations , applications and consequences .
the purpose of the following section is mainly to provide the background material that our subsequent perturbative evaluation of boundary partition functions is based upon . in the first part we gear up to explain the structure of the boundary partition function we are about to deform .
we start with a few comments on brane geometries in wznw models on @xmath1 , extending our previous analysis of branes in the gl(1@xmath121 ) wznw model @xcite .
one of the instantonic d - branes we find , possesses exactly the spectrum that was anticipated in @xcite .
the full field theory partition function and its so - called casimir decomposition is reviewed in the second subsection .
we then turn to a more detailed analysis of the perturbing field , mostly following our previous discussion in @xcite . on this occasion
, we propose a new argument for the exact marginality of @xmath60 which differs a bit in character from the original derivation @xcite . rather than using the vanishing of the dual coxeter number of @xmath34
, our reasoning exploits the position of @xmath60 at the bottom of a logarithmic multiplet ( see also @xcite ) .
most of the results we describe below are not new and the impatient or experienced reader may skip forward to section 3 , at least on first reading . as we shall explain in great detail below , the success of our subsequent exact computation of a boundary partition function for the sigma model on @xmath1 hinges on three key properties of the imposed boundary condition . to begin with , it ( i ) must preserve some combination of left and right regular @xmath34 transformations . at the wz point ,
maximally symmetric boundary conditions are associated with so - called twisted conjugacy classes ( see @xcite and @xcite for the supersymmetric case ) .
explicit formulas for the deformation of the partition function can only be found if ( ii ) the corresponding twisted conjugacy class is point - like localized on the bosonic base and ( iii ) it is delocalized in all the fermionic directions .
later we shall rephrase these conditions as inherent features of the boundary conformal field theory .
our aim here is to describe a boundary condition which meets all these requirements and to determine the relevant boundary partition function at the wz point . in the wznw model ,
the global symmetries of the @xmath1 sigma model are generated by the zero modes of chiral currents @xmath61 a boundary wznw model is scale invariant if the sugawara stress tensor obeys @xmath62 all along the boundary @xmath63 .
such a conformal boundary theory preserves a global @xmath34 symmetry provided that the currents satisfy the following gluing condition @xmath64 here , @xmath65 is a metric preserving automorphism of the lie superalgebra .
it determines the precise combination @xmath66 of global @xmath34 charges that remains unbroken by the boundary condition . in the case of bosonic groups ,
the geometry underlying maximally symmetric boundary conditions in wznw models was unravelled in @xcite ( see also @xcite for various generalizations and @xcite for a review ) .
there it was shown that a boundary condition in which left and right moving currents are identified with a trivial gluing automorphism @xmath67 correspond to branes whose world - volume is localized along conjugacy classes .
when @xmath65 is nontrivial , the relevant geometric objects are twisted conjugacy classes @xmath68 where @xmath69 is an element in @xmath7 and we have lifted the automorphism @xmath65 from the lie algebra to the group . as explained in @xcite , the derivation of @xcite carries over to wznw models on supergroups ( see also @xcite for a general analysis ) .
having outlined the link between boundary conditions and conjugacy classes we are now searching for a pair @xmath70 such that @xmath71 meets the requirements ( ii ) and ( iii ) we have listed in the introductory paragraph to this subsection .
we shall not conduct our search systematically .
instead , let us simply argue that the choice @xmath72 and @xmath73 does the job
. the corresponding twisted conjugacy class @xmath74 is localized at the unit element @xmath59 of the bosonic group and it extends in all fermionic directions , i.e. along those tangent vectors @xmath75 which have degree @xmath76 .
it is easy to see that @xmath73 is consistent with the lie superalgebra structure and the metric .
hence , it extends to a gluing automorphism on the entire current algebra .
moreover , parametrizing elements @xmath77 of the supergroup in the form @xmath78 where @xmath79 is any linear combination of odd ( even ) elements , we find @xmath80 indeed , the bosonic coordinates have dropped out and we remain with a superconjugacy class of superdimension @xmath81 which extends merely along the @xmath82 fermionic directions . we conclude that the space of functions on the corresponding brane is given by @xmath83 where @xmath84 and their bared counterparts are four fermionic coordinates that parametrize the odd generators @xmath85 .
the relevant action of @xmath34 on this @xmath86-dimensional space is obtained by restricting the @xmath65-twisted adjoint action of @xmath34 on the supergroup @xmath1 to a point in the bosonic submanifold , or more explicitly , @xmath87
a^{ab } & = \
-i\eta^a\del^b+i\eta^b\del^a- i\bar{\eta}^a\bar{\del}^b+i\bar{\eta}^b\bar{\del}^a \ \ \ \ \ , \ \ \ \ \ a^{a}_{2 } \ = \ \del^a \ .
\\[2 mm ] \end{split}\ ] ] the curious reader can find explicit formulas for the generators of the left and right regular action @xmath88 and @xmath89 in @xcite . when combined as @xmath90 , they result in the twisted adjoint action that is relevant for the minisuperspace description of the symmetry of our brane . under the twisted adjoint action @xmath91 , the @xmath86-dimensional space of ground states
may be seen to transform according to the representation @xmath92\ \oplus\ [ 0,1 ] \ \ .\ ] ] here , @xmath93 denotes the bosonic subalgebra of the lie superalgebra @xmath94 and we introduced @xmath95 for the trivial 1-dimensional representation of @xmath93 . according to general mathematical results , the module @xmath96 is projective .
hence , it is guaranteed to decompose into a direct sum of projective modules .
the corresponding decomposition is spelled out on the right hand side . here , the symbols @xmath50 $ ] and @xmath97 $ ] denote 48-dimensional irreducible typical representations ( long multiplets ) of @xmath34 .
these are generated from the two 3-dimensional representations of @xmath33 by the application of four fermionic generators .
in addition , there appears the 160-dimensional projective cover @xmath43 $ ] of the trivial representation @xmath98 $ ] .
it is an indecomposable representation that is built up from irreducible atypicals ( short multiplets ) of @xmath34 according to the following diagram @xmath99 : \ [ 0 ] \ \longrightarrow \
3[1/2 ] \ \longrightarrow \
2[1 ] \oplus 6 [ 0 ] \ \longrightarrow
\ 3[1/2 ] \ \longrightarrow \ [ 0 ] \ \ .\ ] ] this so - called composition series tells us that @xmath43 $ ] contains the trivial representation @xmath98 $ ] as a true subrepresentation .
its representation space is spanned by the unique invariant element in @xmath43 $ ] .
we call this subrepresentation @xmath98 $ ] the socle of @xmath43 $ ] . at the other end of the diagram ,
i.e. in the so - called head of @xmath43 $ ] , we find another copy of @xmath98 $ ] .
it is associated with the factor space of @xmath43 $ ] which is obtained if we divide the projective cover by its maximal non - trivial subrepresentation .
a brief summary of the representation theory of @xmath34 is provided in appendix a. many more details can be found in @xcite .
we advise readers who are unfamiliar with indecomposable representations of lie superalgebras to consult those references or other mathematical literature . after this brief discussion of brane geometry and the space of ground states ,
let us analyze the excited states which arise through application of current algebra modes . by construction ,
these states transform in representations that emerge from a product of a projective module with some power of the adjoint and which , by abstract mathematical results , can be decomposed into projectives .
explicit formulas for the involved characters were provided in @xcite .
since we do not need the details below , we refrain from reproducing these formulas here . in @xcite
we also explained how sectors erected over projective modules can be decomposed into representations of the lie superalgebra @xmath34 .
the result can be expressed in the form @xmath100 } ( z_1,z_2;q ) + \chi_{[1,0]}(z_1,z_2;q ) + \chi_{[0,1 ] } ( z_1,z_2;q ) \\[5 mm ] & = & \sum_{j_1\neq j_2}\ , \left ( a^{{\cal p}[0]}_{j_1j_2}(q ) + a^{[1,0]}_{j_1j_2}(q)+ a^{[0,1]}_{j_1j_2}(q)\right ) \ , \chi_{[j_1,j_2]}(z_1,z_2 ) \label{eq : projdeco } \\[2 mm ] & & + \ \sum_{j}\ , \left(b^{{\cal p}[0]}_j ( q ) + b^{[1,0]}_j(q)+ b^{[0,1]}_j(q)\right ) \,\chi_{{\cal p}[j]}(z_1,z_2 ) \nonumber\end{aligned}\ ] ] where @xmath101}$ ] and @xmath102}$ ] are supercharacters of the lie superalgebra @xmath34 ( see appendix a for explicit formulas ) .
formula is known as the casimir decomposition of the partition function .
the various branching coefficients @xmath103 and @xmath36 count how many times a projective @xmath34 multiplet appears on a given energy level .
these numbers may be determined with the help of the racah - speiser algorithm .
a detailed explanation can be found in @xcite along with a few explicit expressions for the branching of the representation @xmath104 $ ] .
here it suffices to recall that the lowest conformal weight @xmath105 among all the multiplets @xmath106 $ ] that are generated out of ground states in the representations @xmath107 , [ 0,1 ] , [ 1,0]$ ] satisfies @xmath108 where we denoted the eigenvalue of the quadratic casimir element in the representation @xmath109 by @xmath110.$ ] .
its generalized eigenfunctions possess vanishing eigenvalue . ]
the same formula with @xmath111 applies to the projective covers @xmath44 $ ] . note that at the wz point the spectrum has huge degeneracies because many different representations of @xmath34 can appear on the same level of the state space .
we shall see how the rr - deformation partially removes this degeneracy .
the most important actress of this work certainly is the perturbing field @xmath60 that generates the deformation away from the wz point .
so , it is important to fully appreciate its structure and properties .
the following discussion is mostly borrowed from our paper @xcite which in turn was based upon @xcite .
the deformation we are interested in is generated by the field @xmath112 the second formulation involves the left and right invariant ( anti-)holomorphic currents @xmath113 and @xmath114 along with some degenerate primary fields @xmath115 that transform in the ( atypical ) adjoint representation @xmath116 $ ] of @xmath34 , i.e. @xmath117 \bar j^\mu(\bar z ) \ , \phi_{\nu\rho}(w,\bar w ) & = & \frac{i{f^{\mu{{\sigma}}}}_\rho } { \bar{z}-\bar{w } } \
\phi_{\nu{{\sigma}}}(w,\bar w ) + \dots\ \ .\end{aligned}\ ] ] let us stress that the vertex operators @xmath118 possess zero conformal weight , as all vertex operators that are associated with the atypical sector of the theory . according to @xcite , the field @xmath60 generates a truly marginal perturbation @xmath119 of the wznw model . by construction , the field @xmath60 has conformal weights @xmath120 but in principle its dimension could change when we perturb the theory , i.e. @xmath60 could be marginally relevant .
this is not the case .
we shall establish true marginality of @xmath60 through a new argument , simpler and of a somewhat different character than the one used in @xcite .
our key observation is that all @xmath121-point functions of @xmath60 vanish identically .
recall from @xcite that the entire bulk spectrum of the @xmath1 wznw model is organized in projective modules with respect to global @xmath1 ( left or right ) transformations .
hence , our perturbing field @xmath60 is part of an indecomposable @xmath1 multiplet @xmath43 $ ] . since @xmath60 is invariant , i.e.@xmath122 \
= \ 0 \ \ \ \ \text { for all } \ \ x \
\in\ { { \text{psl}(2|2)}}\ , \ ] ] it is associated with the bottom ( socle ) of the projective cover @xmath43 $ ] .
consequently , there must exist another bulk field @xmath123 along with a ( fermionic ) symmetry generator @xmath124 such that @xmath125\ \ .\ ] ] the rest of the argument is now rather standard .
let us consider the @xmath121-point function of @xmath60 .
by the previous comment we can represent one of the @xmath121 fields as @xmath126 and obtain @xmath127 & = & \sum_{\nu=2}^n \bigl\langle \psi(z_1,\bar z_1 ) \phi(z_2,\bar z_2 ) \dots q\phi(z_\nu,\bar z_\nu ) \dots \phi(z_n,\bar z_n ) \bigr\rangle\ = 0 \ \ .\end{aligned}\ ] ] we have used the @xmath34 invariance of the expectation value to re - shuffle @xmath124 from @xmath128 to the other fields .
the resulting @xmath129 terms in the second row all vanish because @xmath124 now acts on one of the invariant field insertions @xmath60 .
since all @xmath121-point functions of @xmath60 vanish , there is no need to ever renormalize the perturbing field .
hence , its scaling dimension remains unaltered .
with the proper preparation from the previous section we now come to the central aim of this work : to compute the conformal weights of boundary fields on our point - like brane as we go beyond the wz point .
after a few remarks on the general structure of 2-point functions we shall discuss several remarkable features of the rr - deformation for conformal weights .
these lead to drastic simplifications of the relevant perturbative expansions .
in fact , their combinatorics is no more complex than the combinatorics of radius deformations in torus compactifications ! there remains a mixing problem , however , that we can only overcome for the boundary spectra of point - like localized branes .
the relevant argument is presented in the third subsection .
finally , all the pieces are collected and the conformal weights of boundary fields are computed explicitly , following a closely related computation in @xcite . a boundary partition function stores all information about the conformal weights of boundary fields .
the latter are also encoded in the boundary 2-point functions which is the place from which we are going to read them off . in logarithmic conformal field theories , such as the wznw model on @xmath1 ,
the 2-point functions contains additional data that we are not interested in and , in fact , can not compute perturbatively .
since the reader may not be so familiar with these issues , we shall briefly discuss the general structure of 2-point functions in rr - deformations of the wznw model on @xmath1 .
let us recall that our boundary conditions were chosen such that they preserve a global @xmath34 symmetry .
this remains unbroken by the rr - deformation and hence all quantities in the deformed theory are organized in @xmath34 multiplets .
we shall label the boundary fields by @xmath130 with a superscript @xmath131 that refers to the @xmath34 representation the field transforms in . as we have reviewed above , boundary fields on our instantonic brane
can only transform in projective modules @xmath131 of @xmath34 .
these can be either typical long multiplets or the projective covers of atypical short multiplets . in the following discussion we do not have to distinguish between these two possibilities .
the form of the 2-point functions is strongly constrained by the usual ward identities expressing conformal invariance and global @xmath34 symmetry , @xmath132 here , the symbol @xmath133 denotes an intertwiner from the carrier space of the tensor product @xmath134 to the trivial representation .
let us note in passing that the space of such intertwiners may be multi - dimensional .
the objects @xmath135 act on the carrier space of the representation @xmath131 .
they describe the action of @xmath136 on the field multiplets @xmath137 .
therefore , they clearly commute with the action of @xmath34 .
we may split @xmath138 into a term that is proportional to the identity and a nil - potent contribution , @xmath139 where some finite power of @xmath140 vanishes .
if the nilpotent part @xmath140 is non - zero for one of the fields @xmath141 or @xmath142 then the 2-point function contains logarithmic singularities .
it is important to stress that all the quantities we have introduced , namely the constants @xmath143 and the maps @xmath144 depend on the deformation parameter @xmath16 . for reasons that will soon become clear , we are not able to say anything useful about the @xmath16-dependence of @xmath140 and @xmath145 . on the other hand
, we shall compute @xmath146 exactly , to all orders in perturbation theory . for a field @xmath147 ,
the results is @xmath148 here , @xmath149 is the ( generalized ) eigenvalue of the quadratic casimir in the representation @xmath131 , i.e. @xmath150 for @xmath151 $ ] and @xmath152 whenever @xmath131 is one of the projective covers @xmath44 $ ] .
note that the shift of the conformal weight only depends on the transformation behaviour of @xmath147 under the action of @xmath34 .
the simple result is rather remarkable .
let us stress again that the numbers @xmath146 provide exactly the information that is encoded in the boundary partition function .
in particular , the trace over state space is blind to any nilpotent terms @xmath153 so that our ignorance concerning their @xmath16 dependence does not really matter as long as we do nt attempt to go beyond computing partition functions .
there is one more comment that might be worth adding . as we have seen in section 2.3 already , logarithmic conformal field theories contain many vanishing correlators .
in particular , suppose that @xmath154 and @xmath155 are two fields that are associated with states in the socle of a projective cover . then their 2-point function is bound to vanish by the same arguments we explained in section 2.3 .
a related observation was made by bershadsky et al.in @xcite .
the authors of that work then went on to conclude that the conformal weights of fields in atypical representations could not be read off from their 2-point functions .
we see now that this conclusion is incorrect . for each field in an atypical multiplet there exists some field such that the associated 2-point function is non - zero . if we pick @xmath154 from the socle of a projective cover , for instance , then we can find an appropriate field @xmath155 in the head of the dual projective cover . the perturbative computation of @xmath156 may seem like a daunting task at first , yet alone because of the very involved combinatorics of perturbation theory in curved backgrounds . in this subsection
we shall list three observations that will allow us to drop most of the terms in the expansion for conformal weights .
in fact , the terms that can safely be ignored are precisely the ones that arise from the curvature of @xmath1 . such simplifications , however , only apply to computations of @xmath34 invariant quantities such as conformal weights etc . the reader is warned never to use the rules we are about to derive for computations of other structure constants .
all observations made in this subsection are based on a simple mathematical result that was first formulated and exploited in the work of bershadsky et .
al . @xcite .
consider some @xmath34 invariant @xmath157 and suppose that @xmath157 may be written as @xmath158 where @xmath159 are the structure constants of @xmath34 and @xmath160 are some numbers .
then @xmath157 can be shown to vanish , i.e. @xmath161 .
since the supporting argument provided in @xcite lacks a bit of mathematical precision , we have included a full proof and further discussion in appendix b of this paper .
bershadsky and collaborators applied the vanishing of @xmath157 to a perturbative construction of the @xmath34 invariant @xmath162-function .
we shall exploit the same result in our computation of the numbers @xmath163 which are @xmath34 invariants as well .
a similar vanishing criterion is not satisfied for intertwiners @xmath138 between two indecomposables or for maps @xmath164 from the tensor product of indecomposables to the trivial representation ( see also further comments in appendix b ) . therefore , we are not able to compute the full 2-point function of boundary fields , as mentioned before .
let us now apply this mathematical lemma to our computation of conformal weights .
the perturbative treatment we have in mind requires to evaluate correlators with insertions of the perturbing field @xmath60 .
recall that @xmath60 was composed from the vertex operators @xmath118 and currents @xmath165 .
an initial step is to remove all the current insertions through current algebra ward identities . in the process
, pairs of currents get contracted using @xmath166 the first equality is the usual operator product for @xmath34 currents .
since we are only interested in computing the invariants @xmath163 , we can drop all terms that involve the structure constants @xmath14 of the lie superalgebra @xmath34 .
this applies to the first term in the above operator product which distinguishes the non - abelian currents from the abelian algebra of flat target spaces . here and in the following
we shall use the symbol @xmath167 to mark equalities that are true up to terms involving structure constants . in conclusion , we have seen that , as far as the computation of conformal dimensions is concerned , we may neglect the non - abelian nature of the currents @xmath45 .
obviously , this leads to first drastic simplifications of the perturbative expansion .
currents are not only contracted with other currents .
they can also act on the vertex operators @xmath118 .
the relevant operator product expansions have already been displayed in eq .
when we first introduced @xmath118 . with our new sensitivity for the appearance of structure constants we notice immediately that these operator products are proportional to @xmath14
hence , we conclude @xmath168 consequently , we can simply ignore all terms in which a current acts on one of the vertex operators @xmath118 . in this respect , @xmath118 does no longer behave like a vertex operator , but rather mimics the behavior of a constant background field . of course , @xmath118 still is a non - trivial field and it therefore has possibly singular operator products with other fields in the theory .
such non - trivial operator products of the fields @xmath118 could threaten a successful computation of conformal dimension . here is where a third observation comes to our rescue .
note that shifts of the insertion point of the field @xmath118 are controlled by the following operator version of the knizhnik - zamolodchikov equation @xmath169 this means that in computations of invariants we can treat @xmath118 as a _ function _ of conformal weight zero .
let us stress again that the operator products of @xmath118 can certainly contain singularities .
relation only asserts that all singular terms may be dropped in computations of conformal dimensions . the rules to are the main results of this subsection .
they will be employed at the end of this section when we compute boundary conformal weights .
related observations for the background field expansion of sigma models on psl(n@xmath12n ) were formulated in @xcite .
a successful computation of conformal weights requires one more important ingredient , though , that is novel to our analysis .
this is what we are going to address next .
our arguments up to this point have made no use of the fact that we were setting off to compute conformal dimensions of boundary fields for a very particular boundary condition .
in fact , everything we have stated applies to whatever conformal dimension we would like to compute , bulk or boundary .
but there remains an issue that we can not overcome in such a general context .
according to the results of the previous subsection our vertex operators @xmath118 behave like a matrix of functions rather than fields .
this simplifies things immensely .
but even multiplication with a set of functions can be a rather involved operation which we would have to diagonalize explicitly on field space before we could spell out conformal dimensions .
in other words , there still exists a potentially complicated mixing problem to be solved .
here is where our special choice of boundary conditions comes in .
as we shall see , it is chosen such that we can effectively replace @xmath118 by a constant .
thereby , the mixing problem disappears .
while the reasoning to be detailed below is somewhat technical , the basic idea is rather simple : before the bulk field @xmath118 can act on boundary fields , it must be restricted to the world - volume of the brane .
since our brane is point - like localized at the group unit of the bosonic base , the restriction of @xmath118 contains no further dependence on the bosonic coordinates and hence should have a rather simple action on boundary fields . in order to make this geometric intuition more precise ,
let us look at the bulk - boundary operator product expansion of the vertex operator @xmath170 .
as the world - sheet coordinate approaches the point @xmath171 on the boundary of the upper half - plane , we can re - expand the bulk field in terms of operators @xmath172 on the boundary .
the leading terms of this expansion read @xmath173}\ , \psi^{[1,0]}(x ) + c^{{\cal p}[0]}\ , \psi^{{\cal p}[0]}(x ) \
+ \dots \ .\ ] ] on the boundary , the field with smallest conformal weight is the multiplet @xmath174}$ ] that is associated with the ground states in the 48-dimensional typical representation @xmath97 $ ] . in addition , there is one multiplet @xmath175}$ ] of fields with vanishing conformal weight .
all other fields possess positive scaling dimension and we have not displayed them in the expansion .
the structure constants @xmath176}$ ] and @xmath177}$ ] are largely determined by @xmath34 symmetry . under the action of the unbroken global @xmath34
, the bulk multiplet @xmath118 transforms in the 2-fold twisted where the second term is twisted by the gluing automorphism @xmath65 ] tensor product @xmath116\otimes_\omega [ 1/2]$ ] of the adjoint representation .
consequently , @xmath177}$ ] intertwines between @xmath116 \otimes_\omega [ 1/2]$ ] and the projective cover @xmath43 $ ] etc .
let us recall from the previous subsection that , in all computations of conformal dimensions , the bulk field @xmath118 behaves like a set of functions on target space .
thereby , we are allowed to drop all terms from the bulk boundary operator product which contain a non - trivial dependence on world - sheet coordinates , i.e.@xmath179}\ , \psi^{{\cal p}[0]}(x)\ \ .\ ] ] here , @xmath167 has the same meaning as before , warning us that the relation should only be used in computations of conformal weights .
further progress now requires to turn attention to the intertwiner @xmath177}$ ] from the twisted tensor product @xmath116 \otimes_\omega [ 1/2]$ ] to the projective cover @xmath43 $ ] .
the precise structure of @xmath116\otimes[1/2 ] \cong [ 1/2 ] \otimes _ \omega [ 1/2]$ ] has been determined in @xcite . there
, the tensor product was shown to decompose into four indecomposable representations .
these include the typical multiplets @xmath97 $ ] and @xmath50 $ ] along with the trivial representations @xmath98 $ ] and one atypical indecomposable whose socle consists of a single adjoint @xmath116 $ ] .
the result implies that the space of intertwiners from @xmath116 \otimes [ 1/2]$ ] to the projective cover @xmath43 $ ] is 1-dimensional .
in fact , the only non - trivial intertwiner @xmath180}$ ] maps the invariant @xmath98 $ ] in @xmath116 \otimes_\omega [ 1/2]$ ] to the socle of @xmath43 $ ] .
transferred back to our bulk boundary operator product we conclude that only the socle of the boundary multiplet @xmath181}$ ] can arise .
since the corresponding boundary operator is the identity field , we conclude @xmath182 here , we have used that every intertwiner from @xmath116\otimes_\omega[1/2]$ ] to the trivial representation @xmath98 $ ] is related to the metric by @xmath183 with @xmath184 as before .
since the field @xmath185 is a quantum analogue of the representation matrix @xmath186 and since we are evaluating the latter at the unit element , @xmath187 , we obviously have @xmath188 . consequently , in all computations of boundary conformal weights we are allowed to set @xmath189 .
let us stress that our arguments rely heavily on the fact that we analyze the boundary fields on point - like branes . in particular , we used that there was no boundary field that transforms in the atypical @xmath116 $ ] representation .
let us now finally harvest the results of our careful analysis in the previous two subsections .
as we have shown in the second subsection , the perturbation series for conformal dimensions is identical to the one that appears in an abelian theory with constant background fields .
put differently , the currents @xmath45 and @xmath190 behave like @xmath191 and @xmath192 in a theory of @xmath42 free fields @xmath193 . moreover , the matrix @xmath118 can be treated as if it was a constant , similar to the parameter @xmath194 we introduced in our brief discussion of circle compactifications around eq . . including our choices of normalization , the precise relation is read off from @xmath195 here , we have used a lower case @xmath171 in the argument of @xmath196 in order to stress that it behaves like a function on target space . on the other hand , there is no dependence on the fields @xmath193 . for our special choice of @xmath65 ,
the gluing condition mimics dirichlet boundary conditions for the bosons and neumann boundary conditions for the fermions in free field theory , @xmath197 putting things together , our setup is essentially identical to the starting point of the perturbative analysis in @xcite , hence , we can carry over all results from that paper and conclude that the change of boundary conformal dimensions can be determined from an effective perturbing bulk field of the form @xmath198 where @xmath199 is the upper half - plane and we are no longer allowed to contract currents among each other or with the matrix valued fields @xmath200 .
the matrix @xmath201 is defined by @xmath202 . to leading order
, the effective perturbation agrees with the original perturbing term .
higher order contributions are encoded in a factor @xmath203 that resembles the familiar @xmath204 in the circle compactification ( see discussion after eq . [ sdpf ] ) .
the signs in the denominator take care of the gluing condition we imposed .
there are a few remarks we would like to add . to begin with ,
note that there is no need for any normal ordering in the previous formula , just as in free field theory with constant background fields .
our effective perturbation has rather limit validity , though .
while in @xcite the effective perturbation was used to compute both the change of conformal weights and of 3-point couplings , our entire derivation here was restricted to conformal weights !
so , the formula for the effective interaction should never be used in computations of structure constants .
let us finally point out that for the time being we only assumed that the left and right moving currents satisfy the gluing condition .
therefore , our result holds for all branes of this gluing type , including those cases in which the brane extends along some of the bosonic directions . in the final step
we specialize now to the instantonic brane that is located at the unit element @xmath59 of the bosonic base . using our results from the previous subsection
we may then replace the functions @xmath118 by constants , i.e. we insert @xmath205 into the formula , @xmath206 the change of the boundary conformal weights is determined by the logarithmic divergence in the regulated 2-point function which in turn arises from the simple poles of the operator products between the effective perturbing field and the boundary fields @xmath137 . with the usual normalizations , the resulting shift @xmath207 of conformal weights becomes @xmath208 note that the factor @xmath209 in the effective perturbation is absorbed when we relate the anti - holomorphic current @xmath210 with the boundary value of the holomorphic current @xmath45 . as a result , we have established the anticipated formula .
in this note we computed the full spectrum on a point - like brane in sigma models with target space @xmath1 .
the result was obtained by summing explicitly the perturbation series that is generated by the rr - deformation @xmath29 .
a non - vanishing topological wz term was required in our analysis to guarantee that we could construct the spectrum directly at one point of the moduli space .
we believe that this is merely a technical condition that can be overcome , at least in many examples ( see next paragraph ) .
a very decisive element was to focus on invariants of a lie superalgebra to which the vanishing lemma ( see appendix b ) applies .
this leaves ample room for generalization to other supergroup and coset spaces with psl(n@xmath12n ) or osp(2n+2@xmath122n ) symmetry .
as explained in section 3.2 , the vanishing lemma renders the perturbation series for conformal dimensions quasi - abelian . on the other hand ,
the effective perturbing operator requires additional diagonalization whenever @xmath118 is non - trivial . here
, we circumvented the issue with our special choice of instantonic boundary conditions which allowed us to replace @xmath118 by a constant .
finally , to have sufficient control over the boundary partition function , a casimir decomposition of the spectrum had to be performed .
such a decomposition is not always possible - it needs the brane to stretch out in all fermionic directions .
since branes in generic positions are fully delocalized along fermionic coordinates , no serious limitations should arise for generalizations to other backgrounds . in the following few paragraphs
we shall go through all our assumptions in more detail , with an emphasis on general structures rather than the specific model we dealt with above . to get our perturbative expansion started , we need the exact form of the boundary partition function at one point of the moduli space . in many cases ,
such an _ initial condition _ may come from a wznw model .
the solution of wznw models on type i supergroups has been addressed in @xcite , based on similar studies of several concrete examples @xcite .
it may be interesting to stress that a point with non - abelian current algebra symmetry may exist in the moduli space even if no topological term appears in the action of the model under consideration .
the simplest example is once more provided by circle compactification whose world - sheet symmetry gets enhanced to an su(2 ) current algebra at the self - dual radius .
similar phenomena are very likely to occur for many other principal chiral models on supergroups or cosets .
for example , according to an intriguing conjecture of candu and saleur @xcite , there exists a particular choice of the coupling at which the principal chiral model on the supersphere @xmath211 coincides with a osp(4@xmath122 ) wznw model at level @xmath212 . in general , such special points in moduli space and their exact properties are difficult to detect . but even if no points with current algebra symmetries are known to exist , exact spectra may still be accessible with different techniques , such as the use of lattice constructions etc .
( see e.g. @xcite ) .
once the wz point ( or any other explicitly solvable point ) is under control , we would like to deform the model .
in most cases , summing up an entire perturbation series is a hopeless enterprise .
still , we have seen that explicit summation is possible for the rr - deformation of the @xmath1 sigma model , at least once we focus on appropriate quantities such as conformal weights of boundary fields .
drastic simplifications in the combinatorics of the perturbative expansion resulted from three observations , to , in section 3.2 .
none of them is specific to a target space with @xmath34 symmetry .
in fact , the underlying technical lemma ( reviewed in appendix b ) is closely related to the vanishing dual coxeter number of @xmath34 , a property @xmath34 shares with three families of lie superalgebras , namely psl(n@xmath12n ) , osp(2n+2@xmath12n ) and @xmath13 .
these describe the global symmetries of many interesting superspaces , ranging from odd dimensional superspheres @xmath213 to the coset spaces that are involved in the ads / cft correspondence .
we wish to stress that a vanishing @xmath162 function of the deformation and the quasi - abelianness of the perturbative expansion for conformal dimensions appear as two sides of the same coin .
indeed , they can both be traced back to the vanishing lemma .
let us also point out once more that , even though the perturbation series simplifies for all spectra , we were only able to exploit this fact in the case of point - like branes .
it seems to us that the absence of bosonic zero modes might be an important feature for the success of the computation , but whether it is decisive remains an interesting open problem . in particular
, our brief discussion of bulk spectra in @xmath2 ( see introduction ) suggests that the remaining diagonalization for closed string modes could be more than a mere technical issue . in case the direct perturbative computation of
bulk spectra turns out to be impossible , one might still be able to find bulk conformal dimensions indirectly through modular transformation of boundary partition functions .
approaching the bulk spectrum through open closed string duality would certainly require explicit formulas for the branching functions @xmath214 , going somewhat beyond their mere algorithmic construction @xcite .
another potential hurdle to overcome are the modular properties of the branching functions @xmath214 which might be difficult to control .
even if this is not possible in general , the branching functions might well combine into simpler objects for specific values of the deformation parameter @xmath16 . at points with an enhanced world - sheet symmetry
one would expect an infinite number of branching functions to align such that they build the characters of a larger chiral algebra .
the latter could well possess simpler modular properties .
a systematic detection of points with enhanced symmetry along the line of deformations and the reconstruction of the bulk spectrum is a promising path for future research .
two further comments concern the degeneracies we found in our d - instanton spectra . according to the results in @xcite , the chiral symmetry of sigma models on @xmath1 is generated by the @xmath34 casimir fields , and hence is much smaller than the full casimir algebra , see @xcite for more explanation .
here , we found that the degeneracies of the boundary spectrum are determined by the casimir decomposition .
hence , they are larger than one would have expected based on the chiral symmetry alone .
this is a remarkable result which points towards the existence of some enhanced ( possibly non - local ) symmetry , at least for the boundary spectra we were concerned with in our work
. it would certainly be very rewarding to uncover this symmetry .
a second enhancement of degeneracies is found in the atypical sector of the model .
in fact , the conformal weight of fields transforming in an atypical representation of @xmath34 do not receive any corrections .
therefore such fields are guaranteed to possess an integer conformal weight .
similar phenomena have been encountered in recent work of read and saleur @xcite .
following their analysis we believe that the large degeneracy in the atypical sector may be explained by the combined action of two commuting symmetries .
one of them is the lie superalgebra @xmath34 of global transformations .
the second should be closely related to the algebra of casimir fields or some extension thereof .
results on non - linear sigma models with super target spaces are currently not directly applicable to strings in ads geometries other than via the hybrid approach for @xmath2 .
nevertheless we believe that two rather general lessons can be inferred from our studies .
first of all , conformal field theory techniques , and in particular conformal perturbation theory , can be rather powerful even in cases when the chiral symmetry is not sufficient to carry out a full - fledged algebraic construction of the model .
furthermore , models with a @xmath34 symmetry can be much better behaved than one would expect after looking at any of their subsectors .
in fact , supposedly simpler subsectors , such as e.g.those based on the bosonic @xmath21 , can lead to technical problems that are much more difficult and never encountered in the full @xmath34 model . in this sense
, subsector theories may turn out to be inappropriate as toy models for the kind of theories we are ultimately interested in .
* acknowledgment : * we wish to thank constantin candu , gerhard gtz , andreas ludwig , sylvain ribault , hubert saleur , jrg teschner and kay wiese for comments or discussions during the various stages of this work .
the research of t.q . is funded by a marie curie intra - european fellowship , contract number meif - ct-2007 - 041765 .
we furthermore acknowledge partial support from the eu research training network _ superstring theory _ , mrtn - ct-2004 - 512194 and from _ forcesuniverse _ , mrtn - ct-2004 - 005104 .
the lie superalgebra @xmath34 possesses six bosonic generators @xmath216 with @xmath217 .
they form the lie algebra @xmath218 which is isomorphic to @xmath219 . in addition , there are eight fermionic generators that we denote by @xmath220 .
they split into two sets ( @xmath221 ) each of which transform in the vector representation of @xmath218 ( @xmath222 ) which is the @xmath223 of @xmath219 .
the relations of @xmath34 are then given by @xmath224 & \ = \
i \left [ \d^{ac}k^{bd}-\d^{bc}k^{ad}- \d^{ad}k^{bc}+\d^{bd}k^{ac}\right ] \\[2 mm ] [ k^{ab},s^c_\c ] & \ = \
i \left [ \d^{ac } s^b_\c - \d^{bc } s^a_\c \right ] \\[2 mm ] [ s^a_\a , s^b_\b ] & \ = \
i \,{\epsilon}_{\a\b}\ , { \epsilon}^{abcd } k^{cd } \ \ .
\end{split}\ ] ] here , @xmath225 and @xmath226 denote the usual completely antisymmetric @xmath227-symbols with @xmath228 and @xmath229 , respectively .
an invariant metric is given by @xmath230 it is unique up to a scalar factor .
the signs have been chosen in view of the real form @xmath231 which is considered in the main text . in order to define a root space decomposition of @xmath34 we split the fermions into two sets of four generators @xmath232 as indicated by the subscripts @xmath233
, we shall think of the fermionic generators @xmath234 as annihilation operators and of @xmath235 as creation operators .
finite dimensional projective representations of @xmath34 fall into two classes .
the first one consists of all the long multiplets .
these are labelled by two spins @xmath236 with @xmath237 and their supercharacters read @xmath238}(z_1,z_2 ) \ = \ { \text{tr}}\bigl[(-1)^fz_1^{k_1 ^ 0}\,z_2^{k_2 ^ 0}\bigr ] \ = \ \chi_{j_1}(z_1)\,\chi_{j_2}(z_2)\,\chi_f(z_1,z_2)\ \ .\ ] ] where @xmath239 are the standard characters for finite dimensional representations of the lie algebra @xmath21 and the fermionic factor @xmath240 is given by @xmath241 let us also note in passing that the value @xmath242)$ ] of the quadratic casimir in typical representations may be expressed as @xmath243\bigr ) \ =
\ j_2(j_2 + 1 ) - j_1 ( j_1 + 1 ) \ \ \ .\ ] ] there exists a second class of projective representations @xmath51 $ ] whose members are labelled by a single spin @xmath244 .
they are built up from short multiplets such that their supercharacter becomes @xmath245 } \ = \ \bigl[2\chi_j(z_1)\chi_j(z_2)-\chi_{j+\frac12}(z_1 )
\chi_{j+\frac12}(z_2 ) -\chi_{|j-\frac12|}(z_1 ) \chi_{|j-\frac12|}(z_2)\bigr]\,\chi_f(z_1,z_2)\ .\ ] ] the quadratic casimir is non - diagonalizable in the projective covers , with jordan cells up to rank five
. generalized eigenvalues of @xmath246 in @xmath44 $ ] are well known to vanish for all spins @xmath244 . in this sense we shall write @xmath247\bigr )
= 0 $ ] .
the characters and are important ingredients in the racah - speiser algorithm that furnishes the casimir decomposition for the partition function of a point - like brane , see @xcite for details .
our evaluation of the perturbative expansion for conformal weights is based on the fact that a @xmath34-invariant @xmath157 vanishes whenever it is of the form @xmath158 . in order to make our presentation
self - contained the vanishing lemma is derived below .
we use this opportunity to clarify a few unsatisfactory issues in the original argument @xcite . for the following discussion it is useful to consider @xmath248 and @xmath14 as intertwiners rather than a bunch of numbers . by definition ,
an invariant @xmath157 is an intertwiner from the trivial representation to itself .
similarly , the structure constants @xmath159 may be considered as an intertwiner from the 3-fold tensor product of the adjoint @xmath249}}$ ] to the trivial representation .
the possible form of @xmath249}}^{\otimes3}$ ] can be severely constrained using results from @xcite .
the 2-fold tensor product @xmath116\otimes[1/2]$ ] contains three irreducible representations @xmath250\oplus[1,0]\oplus[0,1]$ ] as well as a more complicated indecomposable @xmath251 .
the tensor product of @xmath252 with @xmath249}}$ ] can easily be evaluated .
furthermore , the typical contributions to @xmath253 $ ] do not present any obstacle .
this results in the decomposition @xmath254}}^{\otimes3 } & \
= \ [ 1/2]\oplus2{\mathcal}{p}_{1/2}\oplus3\bigl([1,0]\oplus[0,1]\bigr ) \oplus4\bigl([3/2,1/2]\oplus[1/2,3/2]\bigr)\\[2 mm ] & \qquad\oplus\bigl([2,0]\oplus[0,2]\bigr ) \oplus\cdots \end{split } \label{eq : ad3}\ ] ] the remaining terms `` @xmath255 '' are the atypical parts in the tensor product @xmath253 $ ] .
they are built by combining the following constituents @xmath256_1,2[0]_3,5[1/2]_1,2[1/2]_3,4[1]_2,[3/2]_1,[3/2]_3\bigr\}\ ] ] into a bunch of indecomposable representations .
multiplicity , see @xcite . ]
the precise form of these indecomposables is currently not known to us .
nevertheless one can derive analytically that their socles can only contain the representations @xmath98_1 $ ] and @xmath257_1 $ ] . due to the self - duality of @xmath249}}^{\otimes3}$ ] , the same statement holds for the heads .
one can also check that there is no true invariant in @xmath249}}^{\otimes3}$ ] , i.e. that the head and the socle are formed by two different @xmath98 $ ] s .
the argument rests on an explicit construction of the unique invariant state and the subsequent proof that it , in fact , lies in the image of the quadratic casimir operator .
hence the unique invariant state has to be the socle of a larger indecomposable multiplet .
given any representation of @xmath34 , the number of independent interwiners to the trivial representation may be obtained by counting the number of times @xmath98 $ ] appears as the head of an indecomposable sub - representation . in the case of @xmath249}}^{\otimes3}$ ] , there is only one such occurrence of @xmath98 $ ] , as we have just argued .
hence , the intertwiner to the trivial representation is unique up to normalization .
this map is what we denote by @xmath14 .
bershadsky et al . now continued to argue that the constants @xmath160 that are contracted with @xmath159 to form the invariant @xmath157 must be proportional to @xmath258 ( indices lowered with the metric ) because of the uniqueness of @xmath14 .
@xmath157 then vanishes because of the numerical identity @xmath259 .
we arrive at the same conclusion if we employ that @xmath14 and @xmath164 combine into an invariant @xmath157 provided that @xmath164 is a co - invariant , i.e. an intertwiner from the trivial representation to the 3-fold tensor product of the adjoint .
such co - invariants are in one to one correspondence with representations @xmath98 $ ] in the socle of @xmath249}}^{\otimes3}$ ] .
a glance back onto our argument above shows that there is a single such representation and hence @xmath164 is unique .
the reason for the vanishing of any invariant @xmath260 is that the image @xmath261 of @xmath164 , given by the socle of @xmath249}}^{\otimes3}$ ] , is in the kernel of @xmath14 , i.e. @xmath261 has no component in the head of @xmath249}}^{\otimes3}$ ] .
the outcome of this analysis , namely the vanishing of an invariant @xmath262 , is the crucial ingredient in our observations to .
j. m. maldacena , _ the large n limit of superconformal field theories and supergravity _ , adv .
theor . math .
* 2 * ( 1998 ) 231252 [ http://arxiv.org/abs/hep-th/9711200[hep-th/9711200 ] ] .
o. aharony , s. s. gubser , j. m. maldacena , h. ooguri and y. oz , _ large n field theories , string theory and gravity _ , phys
* 323 * ( 2000 ) 183386 [ http://arxiv.org/abs/hep-th/9905111 [ hep - th/9905111 ] ] .
r. r. metsaev and a. a. tseytlin , _ type iib superstring action in @xmath263 background _ , nucl
b533 * ( 1998 ) 109126 [ http://arxiv.org/abs/hep-th/9805028[hep-th/9805028 ] ] .
i. pesando , _ a kappa gauge fixed type iib superstring action on @xmath263 _ , jhep * 11 * ( 1998 ) 002 [ http://arxiv.org/abs/hep-th/9808020[hep-th/9808020 ] ] .
r. kallosh and j. rahmfeld , _ the gs string action on @xmath264 _ , phys . lett .
* b443 * ( 1998 ) 143146 [ http://arxiv.org/abs/hep-th/9808038[hep-th/9808038 ] ] .
i. pesando , _ the gs type iib superstring action on @xmath265 _ , jhep * 02 * ( 1999 ) 007 [ http://arxiv.org/abs/hep-th/9809145[hep-th/9809145 ] ] .
j. rahmfeld and a. rajaraman , _ the gs string action on @xmath266 with ramond - ramond charge _ ,
* d60 * ( 1999 ) 064014 [ http://arxiv.org/abs/hep-th/9809164[hep-th/9809164 ] ] .
j. park and s .- j .
rey , _ green - schwarz superstring on @xmath266 _ , jhep * 01 * ( 1999 ) 001 [ http://arxiv.org/abs/hep-th/9812062[hep-th/9812062 ] ] .
n. berkovits , _ super - poincare covariant quantization of the superstring _
, jhep * 04 * ( 2000 ) 018 [ http://arxiv.org/abs/hep-th/0001035[hep-th/0001035 ] ] .
n. berkovits and o. chandia , _ superstring vertex operators in an @xmath263 background _ , nucl .
b596 * ( 2001 ) 185196 [ http://arxiv.org/abs/hep-th/0009168[hep-th/0009168 ] ] .
n. berkovits , _ conformal field theory for the superstring in a ramond - ramond plane wave background _ , jhep * 04 * ( 2002 ) 037 [ http://arxiv.org/abs/hep-th/0203248[hep-th/0203248 ] ] .
n. berkovits , _ quantum consistency of the superstring in @xmath264 background _ ,
jhep * 03 * ( 2005 ) 041 [ http://arxiv.org/abs/hep-th/0411170[hep-th/0411170 ] ] .
n. berkovits , c. vafa and e. witten , _ conformal field theory of ads background with ramond - ramond flux _ ,
jhep * 03 * ( 1999 ) 018 [ http://arxiv.org/abs/hep-th/9902098[hep-th/9902098 ] ] .
n. berkovits , m. bershadsky , t. hauer , s. zhukov and b. zwiebach , _ superstring theory on @xmath267 as a coset supermanifold _ ,
b567 * ( 2000 ) 6186 [ http://arxiv.org/abs/hep-th/9907200[hep-th/9907200 ] ]
. m. bershadsky , s. zhukov and a. vaintrob , _
@xmath268 sigma model as a conformal field theory _ , nucl .
b559 * ( 1999 ) 205234 [ http://arxiv.org/abs/hep-th/9902180[hep-th/9902180 ] ] .
d. kagan and c. a. s. young , _ conformal sigma - models on supercoset targets _ , nucl .
* b745 * ( 2006 ) 109122 [ http://arxiv.org/abs/hep-th/0512250[hep-th/0512250 ] ] .
a. babichenko , _ conformal invariance and quantum integrability of sigma models on symmetric superspaces _ ,
* b648 * ( 2007 ) 254261 [ http://arxiv.org/abs/hep-th/0611214[hep-th/0611214 ] ] .
n. read and h. saleur , _ exact spectra of conformal supersymmetric nonlinear sigma models in two dimensions _ , nucl .
* b613 * ( 2001 ) 409 [ http://arxiv.org/abs/hep-th/0106124[hep-th/0106124 ] ] .
h. saleur and b. wehefritz - kaufmann , _ integrable quantum field theories with @xmath269 symmetries _ , nucl
b628 * ( 2002 ) 407441 [ http://arxiv.org/abs/hep-th/0112095[hep-th/0112095 ] ] .
f. h. l. essler , h. frahm and h. saleur , _ continuum limit of the integrable @xmath270 @xmath271 superspin chain _ , nucl .
* b712 * ( 2005 ) 513572 [ http://arxiv.org/abs/cond-mat/0501197 [ cond - mat/0501197 ] ] .
i. bena , j. polchinski and r. roiban , _ hidden symmetries of the @xmath263 superstring _ , phys . rev . *
d69 * ( 2004 ) 046002 [ http://arxiv.org/abs/hep-th/0305116[hep-th/0305116 ] ] .
n. j. mackay , _ introduction to yangian symmetry in integrable field theory _ , http://arxiv.org/abs/hep-th/0409183[hep-th/0409183 ] .
n. mann and j. polchinski , _ finite density states in integrable conformal field theories _ , http://arxiv.org/abs/hep-th/0408162 [ hep - th/0408162 ] .
n. mann and j. polchinski , _ bethe ansatz for a quantum supercoset sigma model _ ,
phys . rev .
* d72 * ( 2005 ) 086002 [ http://arxiv.org/abs/hep-th/0508232[hep-th/0508232 ] ] .
a. m. polyakov , _ supermagnets and sigma models _ , http://arxiv.org/abs/hep-th/0512310[hep-th/0512310 ] .
n. beisert , _ the su(2@xmath122 ) dynamic s - matrix _ , http://arxiv.org/abs/hep-th/0511082[hep-th/0511082 ] .
a. g. bytsko and j. teschner , _ quantization of models with non - compact quantum group symmetry : modular xxz magnet and lattice sinh - gordon model _ , j. phys .
* a39 * ( 2006 ) 1292712981 [ http://arxiv.org/abs/hep-th/0602093[hep-th/0602093 ] ] .
j. teschner , _ on the spectrum of the sinh - gordon model in finite volume _ , http://arxiv.org/abs/hep-th/0702214[hep-th/0702214 ] .
c. candu and h. saleur , _ a lattice approach to the conformal osp(2s+2@xmath122s ) supercoset sigma model .
part i : algebraic structures in the spin chain .
the brauer algebra .
_ , http://arxiv.org/abs/to appear [ to appear ] .
a. recknagel and v. schomerus , _ boundary deformation theory and moduli spaces of d - branes _ ,
b545 * ( 1999 ) 233 [ http://arxiv.org/abs/hep-th/9811237[hep-th/9811237 ] ] .
v. schomerus , _ d - branes and deformation quantization _ , jhep * 06 * ( 1999 ) 030 [ http://arxiv.org/abs/hep-th/9903205 [ hep - th/9903205 ] ] .
j. callan , curtis g. , i. r. klebanov , a. w. w. ludwig and j. m. maldacena , _ exact solution of a boundary conformal field theory _
b422 * ( 1994 ) 417448 [ http://arxiv.org/abs/hep-th/9402113 [ hep - th/9402113 ] ] .
m. r. gaberdiel and a. recknagel , _ conformal boundary states for free bosons and fermions _
, jhep * 11 * ( 2001 ) 016 [ http://arxiv.org/abs/hep-th/0108238[hep-th/0108238 ] ] . m. r. zirnbauer , _ conformal field theory of the integer quantum hall plateau transition _ , http://arxiv.org/abs/hep-th/9905054 [ hep - th/9905054 ] .
s. guruswamy , a. leclair and a. w. w. ludwig , _ @xmath272 super - current algebras for disordered dirac fermions in two dimensions _ , nucl
b583 * ( 2000 ) 475512 [ http://arxiv.org/abs/cond-mat/9909143[cond-mat/9909143 ] ] . m. j. bhaseen , i. i. kogan , o. a. solovev , n. tanigichi and a. m. tsvelik , _ towards a field theory of the plateau transitions in the integer quantum hall effect _
, nucl . phys .
* b580 * ( 2000 ) 688720 [ http://arxiv.org/abs/cond-mat/9912060[cond-mat/9912060 ] ] .
j. m. maldacena and h. ooguri , _ strings in @xmath2 and @xmath273 wzw model .
, j. math .
* 42 * ( 2001 ) 29292960 [ http://arxiv.org/abs/hep-th/0001053[hep-th/0001053 ] ] .
a. giveon , d. kutasov and a. schwimmer , _
comments on d - branes in @xmath274 _ , nucl .
b615 * ( 2001 ) 133168 [ http://arxiv.org/abs/hep-th/0106005[hep-th/0106005 ] ] .
b. ponsot , v. schomerus and j. teschner , _ branes in the euclidean @xmath2 _ , jhep * 02 * ( 2002 ) 016 [ http://arxiv.org/abs/hep-th/0112198[hep-th/0112198 ] ] . g. gtz , t. quella and v. schomerus , _ the wznw model on psu(1,1@xmath122 ) _ , jhep * 03 * ( 2007 ) 003 [ http://arxiv.org/abs/hep-th/0610070[hep-th/0610070 ] ] . c. candu and h. saleur , _ a lattice approach to the conformal osp(2s+2@xmath122s ) supercoset sigma model .
part ii : the boundary spectrum .
_ , http://arxiv.org/abs/to appear[to appear ] .
t. creutzig , t. quella and v. schomerus , _ branes in the gl(1@xmath121 ) wznw - model _ , http://arxiv.org/abs/arxiv:0708.0583 [ hep - th ] [ arxiv:0708.0583 [ hep - th ] ] .
a. leclair , _ the gl(1@xmath121 )
super - current algebra : the role of twist and logarithmic fields _ , http://arxiv.org/abs/arxiv:0710.2906 [ hep - th][arxiv:0710.2906 [ hep - th ] ] .
a. yu . alekseev and v. schomerus , _ d - branes in the wzw model _ , phys
* d60 * ( 1999 ) 061901 [ http://arxiv.org/abs/hep-th/9812193[hep-th/9812193 ] ] .
k. gawedzki , _ conformal field theory : a case study _ , http://arxiv.org/abs/hep-th/9904145[hep-th/9904145 ] .
g. felder , j. frhlich , j. fuchs and c. schweigert , _ the geometry of wzw branes _
, j. geom . phys .
* 34 * ( 2000 ) 162190 [ http://arxiv.org/abs/hep-th/9909030[hep-th/9909030 ] ]
. v. schomerus , _ lectures on branes in curved backgrounds _ , class .
* 19 * ( 2002 ) 57815847 [ http://arxiv.org/abs/hep-th/0209241[hep-th/0209241 ] ] .
t. creutzig , _ branes in wznw - models on supergroups _ , unpublished .
g. gtz , t. quella and v. schomerus , _ tensor products of @xmath275 representations _ , http://arxiv.org/abs/hep-th/0506072 [ hep - th/0506072 ] . t. quella and v. schomerus , _ free fermion resolution of supergroup wznw models _ , jhep * 09 * ( 2007 ) 085 [ http://arxiv.org/abs/arxiv:0706.0744 [ hep - th][arxiv:0706.0744 [ hep - th ] ] ] . v. schomerus and h. saleur , _ the @xmath276 wzw model : from supergeometry to logarithmic cft _ , http://arxiv.org/abs/hep-th/0510032[hep-th/0510032 ] .
h. saleur and v. schomerus , _ on the su(2@xmath121 ) wznw model and its statistical mechanics applications _ ,
* b775 * ( 2007 ) 312340 [ http://arxiv.org/abs/hep-th/0611147[hep-th/0611147 ] ] .
n. read and h. saleur , _ associative - algebraic approach to logarithmic conformal field theories _ , nucl
b777 * ( 2007 ) 316351 [ http://arxiv.org/abs/hep-th/0701117[hep-th/0701117 ] ] . | in this note we compute exact boundary spectra for d - instantons in @xmath0-models on the supergroup @xmath1 .
our results are obtained through an explicit summation of the perturbative expansion for conformal dimensions to all orders in the curvature radius .
the analysis exploits several remarkable properties of the perturbation series that arises from rescalings of the metric on @xmath1 relative to a fixed wess - zumino term . according to berkovits , vafa and witten ,
the models are relevant in the context of string theory on @xmath2 with non - vanishing rr - flux .
the note concludes with a number of comments on various possible generalizations to other supergroups and higher dimensional supercoset theories . |
the discovery of a metal - insulator transition ( mit ) in the two - dimensional electron gas in si - mosfets @xcite has induced renewed attention to the transport properties of disordered electrons .
this mit is in conflict with the theory of localization for _ non - interacting _ electrons which predicts that all states are localized in 2d .
the electron density in the si - mosfets is very low which makes the electron - electron interaction particularly important .
thus it is generally assumed that some type of interaction effect is responsible for this mit .
one of the most remarkable findings about the mit in si - mosfets is that an in - plane magnetic field ( which does not couple to the orbital motion of the electrons ) strongly suppresses the conducting phase @xcite .
this suggests that the spin degrees of freedom play an important role for the transition .
a complete understanding has , however , not yet been obtained .
there have been attempts to explain the experiments based on the perturbative renormalization group @xcite , non - perturbative effects @xcite , or the transition being a superconductor - insulator transition rather than a mit @xcite . in order to attack the problem of disordered interacting electrons
numerically we have developed @xcite an efficient method , the hartree - fock based diagonalization ( hfd ) which is related to the quantum - chemical configuration interaction approach .
we have used this method to study the influence of interactions on the conductance in one @xcite , two @xcite , and three @xcite dimensions .
we found a delocalizing tendency of the interactions for strong disorder but a localizing one for weak disorder .
similar results have been obtained by means of the density - matrix renormalization group @xcite in one dimension and exact diagonalization in two dimensions @xcite . since in most of the numerical work in the literature spinless electrons
were considered , there are not many results about the importance of the spin degrees of freedom . in this work
we address this question by generalizing the hfd method to spinful electrons .
we then use it study the influence of the spin degrees of freedom on the kubo - greenwood conductance .
the generic model for _ spinless _ interacting disordered electrons is the quantum coulomb glass @xcite . in this paper
we use a straight - forward generalization of the quantum coulomb glass to spinful electrons .
it is defined on a regular hypercubic lattice with @xmath0 ( @xmath1 is the spatial dimensionality ) sites occupied by @xmath2 electrons ( @xmath3 ) . to ensure charge neutrality
each lattice site carries a compensating positive charge of @xmath4 .
the hamiltonian is given by @xmath5 where @xmath6 and @xmath7 are the creation and annihilation operators at site @xmath8 and spin @xmath9 , and @xmath10 denotes all pairs of nearest - neighbor sites .
@xmath11 is the strength of the hopping term , i.e. , the kinetic energy , and @xmath12 is the occupation number of spin state @xmath9 at site @xmath8 .
we parametrize the interaction @xmath13 by its value @xmath14 between nearest - neighbor sites .
the coulomb repulsion between two electrons at the same site is described by the hubbard interaction @xmath15 the random potential values @xmath16 are chosen independently from a box distribution of width @xmath17 and zero mean .
the boundary conditions are periodic and the coulomb interaction is treated in the minimum image convention ( which implies a cut - off at a distance of @xmath18 ) . a numerically exact solution of a quantum many - particle system requires the diagonalization of a matrix whose dimension increases exponentially with system size .
this severely limits the possible sample sizes . in order to overcome this problem
we have developed the hfd method .
the basic idea is to work in a truncated hilbert space consisting of the corresponding hartree - fock ( slater ) ground state and the low - lying excited slater states . for each disorder configuration
three steps have to be performed : ( i ) find the hartree - fock solution of the problem , ( ii ) determine the @xmath19 slater states with the lowest energies , and ( iii ) calculate and diagonalize the hamilton matrix in the subspace spanned by these states .
the number @xmath19 of new basis states determines the quality of the approximation , reasonable values have to be found empirically .
the conductance of a quantum many - particle system can be obtained from linear - response theory .
it is essentially determined by the current - current correlation function of the ground state .
the real ( dissipative ) part of the conductance ( in units of @xmath20 ) is given by the kubo - greenwood formula @xcite , @xmath21 where @xmath22 denotes the frequency .
@xmath23 is the @xmath24 component of the current operator and @xmath25 denotes the eigenstates of the hamiltonian .
( [ eq : kubo ] ) describes an isolated system while in a real d.c .
transport experiment the sample is connected to contacts and leads .
this results in a finite life time @xmath26 of the eigenstates leading to an inhomogeneous broadening @xmath27 of the @xmath28 functions in ( [ eq : kubo ] ) @xcite . to suppress the discreteness of the spectrum of a finite system
, @xmath29 should be at least of the order of the single - particle level spacing . for our systems
this requires a comparatively large @xmath30 .
we tested different @xmath29 and found that the conductance _ values _
depend on @xmath29 but the qualitative results do not .
the main results of this paper are summarized in fig.[fig : conductance ] which shows the disorder and interaction dependence of the typical conductance for both spinless and spinful electrons on a two - dimensional lattice of @xmath31 sites .
the qualitative behavior in both cases is similar : in the strongly localized regime ( small @xmath11 ) a moderate interaction delocalizes the electrons while a sufficiently strong interaction always strongly suppress the conductance .
this is the precursor of a wigner crystal or wigner glass . with decreasing disorder ( increasing @xmath11 )
the interaction - induced enhancement of the conductance also decreases and eventually vanishes .
the behavior of the conductance can be attributed to the competition of two effects : first , the interactions destroy the phase of the electrons and thus the interference necessary for localization .
this is particularly effective if the localization length is small to begin with .
second , the interactions introduce an additional source of randomness which tends to increase the localization .
a comparison of the cases of spinless and spinful electrons shows that the interaction induced delocalization is significantly larger for spinful electrons .
moreover , the enhancement seems to vanish at a larger kinetic energy ( which we did not reach in the simulations ) .
a systematic investigation of the dependence of the conductance on @xmath14 and @xmath15 will be published elsewhere . in summary , we have studied the influence of electron - electron interactions on anderson localization for spinless and spinful electrons in two dimensions . for strong disorder moderate interactions significantly enhance the transport .
this enhancement is much stronger for spinful than for spinless electrons .
identifying a real phase transition and thus establishing a connection between these findings and the experiments on si - mosfets requires a finite - size scaling analysis of the conductance .
this remains a task for the future .
99 s. v. kravchenko , d. simonian , and m. p. sarachik , phys .
* 77 * , 4938 ( 1996 ) .
d. popovich , a. b. fowler , and s. washburn , phys .
79 * , 1543 ( 1997 ) ; d. simonian , s. v. kravchenko , and m. p. sarachik , _ ibid . _
* 79 * , 2304 ( 1997 ) . c. castellani , c. dicastro , and p. a. lee , phys .
b * 57 * , r9381 ( 1998 ) . s. chakravarty , s. kivelson , c. nayak , and k. vlker , cond - mat/9805383 .
d. belitz and t. r. kirkpatrick , phys .
b * 58 * , 8214 ( 1998 ) ; f. c. zhang and t. m. rice , cond - mat/9708050 ; p. phillips _
et al . _ ,
nature * 395 * , 253 ( 1998 ) .
t. vojta , f. epperlein , and m. schreiber , phys .
* 81 * , 4212 ( 1998 ) .
t. vojta , f. epperlein , and m. schreiber , accepted for publication in comp .
, cond - mat/9809171 . m. schreiber , f. epperlein , and t. vojta , physica a * 266 * , 443 ( 1999 ) .
t. vojta and f. epperlein , ann .
( leipzig ) * 7 * , 493 ( 1998 ) .
p. schmitteckert , r.a .
jalabert , d. weinmann , and j .- l .
pichard , phys .
lett . * 81 * , 2308 ( 1998 ) .
g. benenti , x. waintal , and j .- l .
pichard , cond - mat/9904096 .
a. l. efros and f. g. pikus , solid state commun .
* 96 * , 183 ( 1995 ) ; j. talamantes , m. pollak , and l. elam , europhys .
lett . * 35 * , 511 ( 1996 ) ; f. epperlein , m. schreiber , and t. vojta , phys . rev .
b * 56 * , 5890 ( 1997 ) .
k. kubo , j. phys .
* 12 * , 576 ( 1957 ) ; d. a. greenwood , proc .
. soc . * 71 * , 585 ( 1958 ) .
see , e.g. , s. datta , _ electronic transport in mesoscopic systems _ , cambridge university press , cambridge ( 1997 ) . | we numerically investigate how electron - electron interactions influence the transport properties of disordered electrons in two dimensions .
our study is based on the quantum coulomb glass model appropriately generalized to include the spin degrees of freedom . in order to obtain the low - energy properties of this model we employ the hartree - fock based diagonalization , an efficient numerical method similar to the configuration interaction approach in quantum chemistry .
we calculate the d.c .
conductance by means of the kubo - greenwood formula and pay particular attention to the spin degrees of freedom . in agreement with earlier results
we find that electron - electron interactions can cause delocalization . for spinful electrons
this delocalization is significantly larger than for spinless electrons . |
since shor and grover algorithms @xcite were proposed with various following significant developments , e.g. , @xcite , quantum computing has been displaying its more and more amazing charm against classical computing .
as more progress has been made in this area , it is urgent to discover various robust , controllable and scalable two - level systems - qubits as the basic elements for the future architecture of quantum computers .
generally speaking , electron spins are a natural qubit , especially the single electron spin confined in a quantum dot for its well separation and easy addressability . in ref .
@xcite , electron spins in quantum dots were employed as qubits and two - qubit operations were performed by pulsing the electrostatic barrier between neighboring spins .
thereafter , kane s model made use of the nuclear spins of @xmath0p donor impurities in silicon as qubits @xcite .
it combined the long decoherence time of nuclear spins and the advantage of the well developed modern semiconductor industry . in practice
, it seems difficult to control the coupling between qubits because the coupling is based on the overlap of two adjacent spin wave functions @xcite ; the coupling is given to be fixed once photolithography of the chip has been finished .
another feasible way to induce the controllable inter - spin interaction is to couple two spins by spin - orbit interaction @xcite .
however , because of the weakness of spin - orbit coupling , it is very crucial to find a scheme to manipulate two spin coupling in the strong interaction regime . in the present paper ,
we consider the possibility of creating quantum entanglement of two electron spins by making them pass through a 2d quantum well containing many `` cooled '' nuclear spins ( see fig .
[ sketch ] ) .
it was discovered that the effective coupling intensity is increased by a factor of @xmath1 when an electron spin is coupled with an ensemble of @xmath2 nuclear spins @xcite .
such an electron spin coupled to the nuclei has been considered for cooling the nuclear ensemble @xcite .
moreover , people have proposed a quantum computing scheme using a scanning tunneling microscopy with a moving tip as a commuter to perform the control - not gate between two qubits on the silicon surface @xcite . here , the tip plays the role of the quantum data bus to coherently link the qubits .
enlightened by these works , we suggest a new scheme to entangle two electron spins by a tip since it can be modeled as an ensemble of many spins @xcite .
indeed , when the couplings of the spin to the nuclear ensemble are quasi - homogeneous , the interaction between the electron spin and the collective excitation of nuclear spins can be well described in terms of artificial cavity qed @xcite . here , the collective excitation can behave as a single mode boson to realize a quantum data bus , while the electron spin acts as a two - level artificial atom . with the frequency selection due to the resonance effect
, there is only one mode of collective excitations interacting with the two spins .
especially , when the zeeman splits of all nuclear spins are the same , the single mode excitation can decouple with other modes @xcite .
then , the coupling system with two qubit spins and nuclear ensemble just acts as a typical cavity qed system or spin - boson system . to coherently manipulate the indirect interaction between the two spins , which is induced by
the above mentioned collective excitation , we need to let two electrons go through the quantum well to realize a two qubit logical gate operation . since the moving of electrons leads to a time - dependent coupling , we need to use some new method to derive the effective hamiltonian for the inter - spin coupling .
fortunately , a recent paper suggested such a time - dependent approach @xcite .
the rest of our paper is organized as follows . in sec.ii , in the low excitation limit , we simplify the total system we considered above as two spins interacting with a single mode of the collective excitation of the nuclei , which forms a cavity - qed under the quasihomogeneous condition . in sec .
iii , we derive the effective hamiltonian between the two electron spins by the time - dependent frhlich transformation ( tdft ) method developed recently in ref.@xcite .
we remark that the tdft method can be used to derive an effective hamiltonian for a class of cavity qed systems with time - dependent perturbations . here
, we use this transformation for the case with time - dependent couplings of two spins to a many spin ensemble .
section iv contains the discussion of entanglement induced by the effective hamiltonian and collective excitation s effect on the entanglement . in sec .
v we review most of the significant results . finally , technical details are given in appendices [ app : appendix1]-[app : appendix2 ] .
move with a uniform speed @xmath3.,width=7 ] we consider a system illustrated in fig.[sketch ] .
two electrons go through a quantum well one after the other .
the electron spins described by gaussian packets with width @xmath4 are initially located at @xmath5 ( @xmath6 ) and move along the z - direction with a uniform speed @xmath3 .
the 2d quantum well consists of many polarized nuclear spins located in position @xmath7 with @xmath8 , @xmath9 , @xmath10 .
when a static magnetic field is applied to the total system , the hamiltonian reads as @xmath11 where @xmath12 and @xmath13 ( @xmath14 ) ( @xmath15 ) are the spin operators for the @xmath16th electron spin , @xmath17 and @xmath18 ( @xmath19 ) ( @xmath20 ) the spin operators for the @xmath21th nuclear spin , @xmath22 ( @xmath23 , @xmath24 ) the hyperfine coupling constants between @xmath16th electron and @xmath21th nuclear spin .
the first and second terms of hamiltonian ( originalh ) are the zeeman energies for the electron spins and the nuclear spins respectively , and the terms besides them are the hyperfine interaction between the electrons and nuclear spins . in our setup ,
the nuclear spins are restricted in a flat square box with @xmath25 , @xmath26 , @xmath27 .
we have ( for the necessary details please refer to the appendix [ app : appendix1 ] ) @xmath28 in ref .
@xcite , the collective excitation of an ensemble of polarized nuclei fixed in a quantum dot was studied . under the approximately homogeneous condition
the many - particle system behaves as a single - mode boson interacting with the spin of a single conduction - band electron confined in this quantum dot .
likewise , we introduce a collective operator @xmath29^{2}}}\]]and its conjugate @xmath30 to depict the collective excitations in the ensemble of nuclei with spin @xmath31 from its polarized initial state @xmath32which is the saturated ferromagnetic state of nuclear ensemble . in our model ,
the nuclear spins are fixed in gaas crystal lattice with @xmath33 , @xmath34 and the density of nuclei @xmath35 . for simplicity , we assume the nuclei are located in a simple cubic lattice .
thus , we have the average distance between two neighboring nuclear spins @xmath36 , @xmath37 , @xmath38 , where @xmath39 is the maximum of @xmath40 and @xmath41 . on condition that the number of excitations in the system @xmath42 , we have @xmath43\rightarrow 1 $ ] . in other words ,
the collective excitation described by @xmath44 can behave as a boson mode in the large @xmath2 limit with an initial polarization of all spins in ground ( spin down ) state .
in addition to the basic mode denoted by @xmath44 and @xmath30 , there exist auxiliary modes @xmath45}i_{-}^{(i)}}{\sqrt{2i_{0}(\vec{% h}^{[k]})^{2}}}\ ] ] for @xmath46 . here
, @xmath47}=(h_{1}^{[k]},h_{2}^{[k]},\ldots , h_{n}^{[k]})\ ] ] are @xmath2 orthogonal vectors in @xmath2-dimensional space @xmath48 , which can be systematically constructed by making use of the gram - schmidt orthogonalization method @xcite starting from @xmath49}=(g_0^{(1)},g_0^{(2)},\ldots , g_0^{(n)})\in \mathbf{r^{n}}.\ ] ] therefore , the hamiltonian ( [ originalh ] ) is rewritten as @xmath50 here , the effective rabi frequency @xmath51 ^{2}}\text{.}\]]describes the enhanced coupling of the electron spin to the collective excitations . and the single particle excitation term @xmath52 can be treated as a perturbation term in the low - excitation limit , which originates from the inhomogeneity of the couplings .
as shown in hamiltonian ( [ hs ] ) , there are only couplings of two electron spins with the single mode boson respectively . by making use of the canonical transformation @xcite
, we can eliminate the boson operator and obtain the effective interaction between the two electron spins .
former research mainly focused on the case where @xmath53 and @xmath54 are time independent @xcite .
however , the time - independent approach may not work well in practice . now , because of the motion of the electrons , we take the time - dependence of interaction into consideration , namely , @xmath55 and @xmath54 depend on time .
let us first summarize the main idea of the time - dependent frhlich transformation @xcite so that our paper is self consistent for reading . generally speaking ,
frhlich transformation @xcite is frequently used in condensed matter physics to obtain effective interaction between two electrons by exchanging virtual phonons . for a quantum system described by hamiltonian @xmath56 , where @xmath57 is time independent and @xmath58
, we can make a canonical transformation @xmath59where @xmath60 is an anti - hermitian operator and @xmath61 the state of the system .
when @xmath60 is appropriately chosen to make the first - order term of the effective hamiltonian vanishing , i.e. , @xmath62-i\partial _
{ t}f=0 $ ] , we obtain an effective hamiltonian to the second order @xmath63/2 $ ] in principle , and the above equation explicitly determines @xmath64 . in this section , the canonical transformations are made to obtain the effective hamiltonian . in the interaction picture with respect to @xmath65
the hamiltonian @xmath66 contains three parts @xmath67here,@xmath68 are the detunings of electron spin and nuclear spin and hyperfine interaction .
it can be observed from the hamiltonian @xmath69 that the time - dependent term @xmath70 can be considered as first - order perturbation with respect to the zeroth - order term @xmath71 ( disregarding @xmath72 ) .
then , we perform a transformation @xmath73 to the hamiltonian @xmath69 to eliminate the time - dependent term @xmath70 , that is , the condition @xmath74-i\partial
_ { t}f=0 \label{condition}\]]should be fulfilled , where tdft operator is@xmath75it follows from eq .
( [ condition ] ) that the corresponding coefficients of @xmath76 , @xmath77 , @xmath78 and @xmath79 at the left hand side of eq .
( [ condition ] ) vanish , i.e. , @xmath80 in case that the electrons go through the nuclear spins with a uniform speed @xmath3 , the solutions to the above equations are @xmath81 ( @xmath82 ) , where @xmath83 is used to replace @xmath84 and @xmath85 since @xmath86 in the realistic parameters ( for
the necessary details please refer to appendix [ app : appendix2 ] ) .
then , the effective hamiltonian is obtained approximately as follows @xmath87+h_{p}^{i}+[h_{p}^{i},f ] \notag \\ \simeq & [ \delta _ { 1}-\frac{f_1 ^ 2\omega^2}{\delta}(2<b^+b>+1)]s_{z}^{(1 ) } \notag \\ & + [ \delta _ { 2}-\frac{f_2 ^ 2\omega^2}{\delta}(2<b^+b>+1)]s_{z}^{(2 ) } \\ & -\frac{f_1 f_2\omega^2 } { \delta}% [ s_{+}^{(1)}s_{-}^{(2)}+s_{+}^{(2)}s_{-}^{(1)}]+h_{p}^{i}+[h_{p}^{i},f]\text{% , } \notag\end{aligned}\ ] ] where @xmath88 denotes the average number of nuclear excitation , and the fast - oscillating terms including the factor @xmath89 have been dropped off . when almost all nuclear spins are in their ground state , the system is in the low collective excitation limit , i.e. , @xmath90 . by using@xmath91
we have @xmath92thus , @xmath93\text{.}\end{aligned}\ ] ] in the following calculation , it will be shown that the complex term @xmath94 $ ] will be dropped in the interaction picture . with respect to @xmath95
, the effective interaction hamiltonian is @xmath96where @xmath97e^{-ih_{0}t } \notag \\ & = -\frac{f_1 f_2\omega^2}{\delta}% ( s_{+}^{(1)}s_{-}^{(2)}+s_{+}^{(2)}s_{-}^{(1)})\text { , } \label{v1 } \\
v_{2 } & = e^{ith_{0}}h_{p}^{i}e^{-ith_{0}}=h_p^s \text{.}\end{aligned}\ ] ] in the above calculation , we have dropped the high - frequency terms @xmath98\exp ( -ih_{0}t)$ ] including the factors @xmath99 .
it is a reasonable approximation which is frequently used in the jaynes - cummings model . now , we study the time evolution driven by the above effective hamiltonian .
first of all , we study a special case that the total system is initially prepared without the collective excitations of the bus spins . in this case
, the effective interaction @xmath100 does not play a role . in a hilbert space spanned by the two electron states
@xmath101 , @xmath102 , @xmath103 and @xmath104 , it is clear that there exists an invariant subspace spanned by @xmath105 and @xmath103 . if the system starts from @xmath106 , at time @xmath107 it would definitely evolve into @xmath108 where @xmath109 in comparison with the result in ref .
@xcite , where @xmath110 was considered as a perturbation in the low excitation approximation , we examine the system evolving under total hamiltonian containing @xmath111 and @xmath100 .
then we can get the equations for the coefficients as follow @xmath112where @xmath113 and @xmath114 .
similarly , there s an invariant subspace @xmath115 .
in the above sections , we have obtained a typical spin - spin coupling in the effective hamiltonian , which is induced by the collective excitations . driven by this hamiltonian
, two electron spins can be entangled dynamically . to characterize the extent of entanglement , we use concurrence to measure the induced entanglement . for an arbitrary state of two - qubit system described by the density operator @xmath116 , a measure of entanglement can be defined as the concurrence @xcite , @xmath117where the @xmath118 s are the square roots of the eigenvalues of the non - hermitian matrix @xmath119 in decreasing order .
and@xmath120where @xmath121 is the complex conjugate of @xmath116 , @xmath122 the pauli operator .
actually , even from the original hamiltonian ( 1 ) we can also prove that the corresponding reduced density matrix for two spin is of the form @xmath123 to prove the above result , we consider that , in the original hamiltonian the interaction terms @xmath124conserves the total spin z - component @xmath125i.e . , @xmath126=0 $ ] . for such conserved system
we express the concurrence characterizing quantum entanglement in terms of observables , such as correlation functions .
the complete basis vectors of the total system are denoted by @xmath127 denote the electronic spins@xmath128 respectively .
the fact that @xmath129 is conserved can be reflected by the vanishing of some matrix elements of the density operator @xmath130 on the above basis for any state of the total system , that is , @xmath131 where @xmath132 the functional @xmath133 of the hamiltonian may be a ground state or thermal equilibrium states .
the reduced density matrix @xmath134 $ ] for two spins , e.g. , @xmath135 and @xmath136 are obtained as @xmath137_{s_{1},s_{2};}^{s_{1}^{\prime } , s_{2}^{\prime } ; } = \delta ( s , s^{\prime } , 0,0)\sum_{\{i_{j}\}}\rho _ { s_{1},s_{2};i_{1}, .. ,i_{n}}^{s_{1}^{\prime } , s_{2}^{\prime } ; i_{1}, .. ,i_{n}}\]]by tracing over all nuclear variables . the corresponding reduced density matrix for two spins @xmath138 and
@xmath139 is of the form in eq.([reduced ] ) . using the observable quantities ,
the quantum correlation@xmath140the concurrence is rewritten as a computable form @xmath141we note that this formula for the concurrence of two electron spins in the coupled system is the same as that for a spin-1/2 coupling system modeled by the effective hamiltonian @xcite .
this general form is consistent with that obtained straightforwardly from the effective hamiltonian given in the last section . and velocity @xmath3 , distance @xmath142 , on condition that @xmath143 nm , @xmath144 hz , @xmath33 , @xmath145 . , width=7 ] with @xmath146 nm.,width=7 ] with the above general consideration , we now study quantitatively the concurrence for the quantum entanglement of the two electrons passing the nuclear spins at a uniform speed @xmath3 .
in fig . [ v1_total ] , the concurrence is plotted while the speed @xmath3 and the initial distance between the two electrons @xmath147 are varied .
it is obvious that the concurrence fluctuates from @xmath148 to @xmath138 in the low speed region ( see also fig .
[ v1_sectional ] ) .
when the electrons move with a relative low speed , the concurrence oscillates rapidly since a lower speed means more time for evolution from a direct product state towards an entangled state . as the speed increases ,
the concurrence falls monotonously if it is bigger than a certain value .
according to eq .
( [ theta ] ) , the maximum entangled state can be obtained when @xmath149however , the general relation between the concurrence and @xmath150 is a little more complicated .
the further the two electrons separate from each other , the longer time both of them need to pass through the nuclear ensemble . on the other hand , the matrix elements of the effective interaction @xmath111 in eq .
( [ v1 ] ) , i.e. , @xmath151 , h.c . ,
drop dramatically as the inter - spin distance increases .
this observation is obviously correct from an intuitively physical consideration . for a longer inter - spin distance ,
the spatial wave function of two spins has a smaller overlap , and then the effective coupling is weak . in the last section ,
we have only taken @xmath111 into consideration . however , due to the nuclear excitation , i.e. , @xmath152 , @xmath100 will lead to decoherence . according to ref .
@xcite , under the quasihomogeneous condition , we have @xmath153 , where @xmath154 , @xmath155 is the average number of collective excitation .
thus , the single particle perturbation can be approximated as @xmath156 in fig .
[ v1_v1+v2 ] , we plot the concurrence evolution under @xmath157 . as shown in the figure , the two spins evolve into a maximum entangled state with appropriate parameters when @xmath145 . on the contrary ,
the concurrence is suppressed when there exists collective excitation in the nuclei . moreover , as more nuclei are excited , the concurrence falls dramatically . on condition that @xmath158 nm , @xmath159 hz , @xmath33 , @xmath160 mm / s.,width=7 ]
in summary , we have proposed a scheme to entangle two electron spins via an ensemble of nuclei .
we also explore the influence of its collective excitation on the concurrence characterizing two spin entanglement .
theoretically , the maximum entangled state can be obtained if the electrons move in a suitable way .
furthermore , with the optimized experimental parameters , the operation time is within the relaxation time of electron spins in solid state systems , i.e. , the order of ms @xcite .
however , this scheme may encounter some difficulties in practice since it is based on the low excitation requirement of nuclear ensemble . in recent experiments ,
the number of excitation @xmath161 is of the order of @xmath162 bracker .
thus , further progress in experiment , i.e. , lowering the temperature or optical excitation , are expected to prepare all nuclear spins in their ground states in order to put this scheme into practice .
actually , there are only the collective excitations considered as the quantum data bus to coherently link two spins so that the inter - spin entanglement is induced .
if there exists noncollective excitations , then extra decoherence will be induced to break our scheme presented in this paper .
further investigations are needed for these questions . however ,
if we can cool the nuclear ensemble via some new mechanism , e.g. , similar to ref.@xcite , our scheme will probably work well .
thanks very much for helpful discussion with peng zhang , nan zhao , zhangqi yin , zhensheng dai , yansong li . this work is supported partially by the 973 program grant nos .
2006cb921106 , 2006cb921206 and 2005cb724508 , national natural science foundation of china , grant nos .
10325521 , 60433050 , 60635040 , 10474104 , and 90503003 .
according to ref.@xcite , the hyperfine interaction constant is mainly proportional to the electron spin density located at the nucleus .
thus , @xmath165 where @xmath166 is vacuum permeability , @xmath167 total nuclear spin quantum number , @xmath168 the bohr magneton , @xmath169 the nuclear magnetic moment , @xmath170 the wavefunction for electron 1 located at the @xmath171th nuclear spin . in a semiconductor crystal ,
the wavefunction is given by the product of the bloch amplitude @xmath172 and an envelope function @xmath173 , i.e. , @xmath174 . in a realistic crystal @xmath175
reaches a climax at the lattice positions .
it was discovered that @xmath176 and @xmath177 @xcite . on account of isotope abundance and their different nuclear magnetic moment @xcite , @xmath178 . in our simulation , we assume that the wavefunction is gaussian wave packet with initial location @xmath179 and a uniform speed @xmath3 , that is @xmath180/a^2}\text{,}\ ] ] where @xmath181 is the position of the @xmath182th nuclear spin with respect to the center of nuclei . since the nuclei are distributed in a flat box with @xmath183 , @xmath184 , @xmath185 .
therefore , @xmath186where @xmath187 here , we have neglected the term @xmath188 based on the following consideration . on the one hand
, the above approximation tends to be exact as the quasi-2d quantum well becomes narrower in the z - direction , e.g. , @xmath189 . on the other hand ,
the effective coupling intensity is increased as more nuclear spins are included when @xmath190 get larger .
thus , optimal value is chosen for the valid approximation .
similarly , we have @xmath191 where@xmath192
according to eq.([x1 ] ) , one has @xmath194 \notag \\ & \simeq \frac{-\omega } { \delta } e^{-i\delta t}(f_{1}e^{i\delta t}-\frac{8v}{a% } \int_{-\infty } ^{t}f_{1}e^{i\delta t^{\prime } } dt^{\prime } ) \text{. } \notag\end{aligned}\ ] ] here , we have replaced @xmath195 by @xmath83 since@xmath196 furthermore , we have replaced @xmath197dt^{\prime } \ ] ] by @xmath198since one notices the fact that @xmath199 is the effective integration range and the change of @xmath200 ( also change of @xmath201 ) is much slower than that of @xmath202 .
thus , @xmath203 generally speaking , @xmath204 depends on the applied magnetic field . in case that @xmath205 hz and @xmath206 nm , we have @xmath207 for all @xmath208 m / s .
thus , we have @xmath209 similarly , @xmath210 d. kleckner and d. bouwmeester , nature * 444 * , 75 ( 2006 ) ; a. naik , o. buu , m. d. lahaye , a. d. armour , a. a. clerk , m. p. blencowe , and k. c. schwab , nature * 443 * , 193 ( 2006 ) ; s. gigan , h. r. bhm , m. paternostro , f. blaser , g. langer , j. b. hertzberg , k. c. schwab , d. buerle , m. aspelmeyer , and a. zeilinger , nature * 444 * , 67 ( 2006 ) ; o. arcizet , p .- f .
cohadon , t. briant , m. pinard , and a. heidmann , nature * 444 * , 71 ( 2006 ) ; m. poggio , c. l. degen , h. j. mamin , and d. rugar , cond - mat/0702446 . | we theoretically explore the possibility of creating spin entanglement by simultaneously coupling two electronic spins to a nuclear ensemble . by microscopically modeling the spin ensemble with a single mode boson field
, we use the time - dependent frhlich transformation ( tdft ) method developed most recently [ yong li , c. bruder , and c. p. sun , phys .
rev . a * 75 * , 032302 ( 2007 ) ] to calculate the effective coupling between the two spins .
our investigation shows that the total system realizes a solid state based architecture for cavity qed .
exchanging such kind effective boson in a virtual process can result in an effective interaction between two spins .
it is discovered that a maximum entangled state can be obtained when the velocity of the electrons matches the initial distance between them in a suitable way .
moreover , we also study how the number of collective excitations influences the entanglement .
it is shown that the larger the number of excitation is , the less the two spins entangle each other . |
the inner parts of some spiral galaxies show higher star formation rates than usual and this star formation is frequently arranged in a ring or pseudo - ring pattern around their nuclei . in general ,
circumnuclear star forming regions ( cnsfr ) , also referred to as `` hotspots '' , are alike luminous and large disk hii regions , but look more compact and show higher peak surphace brightness ( kennicut et al .
1989 ) . in many cases
they contribute substantially to the uv emission of the entire nuclear region ( _ e.g. _ colina et al .
their h@xmath1 luminosities overlap with those of hii galaxies being typically higher than 10@xmath2 erg s@xmath3 which points to relatively massive ionizing star clusters .
these regions are expected to show a high metallicity as corresponds to their position near the galactic bulge .
they have considerable weight in the determination of abundance gradients , which in turn are widely used to constrain chemical evolution models , and constitute excellent laboratories to study how star formation proceeds in high metallicity environments .
12 cnsfrs were observed with the 4.2 m wht at the roque de los muchachos observatory using the isis double spectrograph , with the eev12 and tek4 detectors in the blue and red arm respectively with the dichroic at @xmath47500 .
gratings r300b in the blue arm and r600r in the red arm were used , covering the spectral ranges @xmath43650 - @xmath47000 in the blue and ( @xmath48850 -@xmath49650 ) in the near ir , yielding spectral dispersions of 1.73 pixel@xmath3 in the blue arm and 0.79 pixel@xmath3 in the red arm .
a slit width of 1 arcesec was used .
the nominal spatial sampling was 0.4 arcsec pixel@xmath3 in each frame and the average seeing for this night was @xmath51.2 arcsec .
the data were reduced using the iraf ( image reduction and analysis facility ) package following standard procedures .
the high spectral dispersion used in the near infrared allowed the almost complete elimination of the night - sky oh emission lines and , in fact , the observed @xmath49532/@xmath49069 ratio is close to the theoretical value of 2.48 in all cases .
telluric absorptions have been removed from the spectra of the regions by dividing by a relatively featureless continuum of a subdwarf star observed in the same night .
examples of blue and red spectra are shown in figure 1 .
emission line fluxes were measured using the iraf splot software package .
the presence of a conspicuous underlying stellar population , more evident in the blue spectra , in most observed regions complicates the measurements .
an example of underlying absorption can be seen in the inset to the left panel of figure 1 . a two - component ( emission and absorption )
gaussian fit was performed in order to correct the balmer lines for underlying absorption .
an example of this procedure is shown in figure 2 .
the low excitation of the regions , as evidenced by the weakness of the [ oiii ] @xmath4 5007 line ( see left panel of figure 1 ) , precludes the detection and measurement of the auroral [ oiii ] @xmath4 3463 necessary for the derivation of the electron temperature .
it is therefore impossible to obtain a direct determination of the oxygen abundances .
empirical calibrations have to be used instead . in the left panel of figure 3
we show the calibration of oxygen abundance by the commonly used @xmath6 parameter defined as @xmath7\lambda 3727,29 + [ oiii]\lambda 4959,5007)/h_\beta$ ] ( pagel et al .
data on hii galaxies , disc hii regions and cnsfr are shown . the hii region sample (
prez - montero & daz 2005 ) has been divided in under - solar ( open triangles ) and over - solar ( filled triangles ) according to the empirical criterion given by daz & prez - montero ( 2000 ) , _
@xmath8 0.47 and -0.5@xmath9 0.28 .
is defined as @xmath10 + [ siii])/h_\beta$ ] + ] the hii galaxy data ( filled squares ) come from prez - montero & daz ( 2003 ) .
cnsfr are represented by circles .
solid ones for our observed objects and open ones for regions in ngc 3310 and ngc 7714 , known to have under - solar abundances ( pastoriza et .
1993 ; gonzlez - delgado et al .
as can be seen the calibration is two - folded , shows a considerable scatter and its high abundance end is not well sampled .
the position of the observed cnsfr is indicated .
their observed @xmath6 values , lower than those of the lowest abundance galaxy known ( izw18 ) , indicates that cnsfr belong to the high abundance branch of the calibration showing possibly the highest metallcities shown by hii region like objects . in the right panel of figure 3
the position of the regions is indicated in the @xmath11 ( [ nii]/ h@xmath12 ) abundance calibration diagram ( denicol et al .
2002 ) , which shows a linear behaviour .
again cnsfr appear to show the highest oxygen abundances .
the left panel of figure 4 shows the [ nii]/[oii ] ratio _ versus _ the @xmath11 abundance parameter .
since a good correlation has been found to exist between the [ nii]/[oii ] ratio and the n@xmath13/o@xmath13 ionic abundance ratio , which in turn can be assumed to measure the n / o ratio , this graph is the observational equivalent of the n / o _ vs _ o / h diagram .
we can see that a very tight correlation exists for all represented objects : hii galaxies , low and high abundance hii regions and cnsfr .
again our observed regions show the highest n / o ratios of the sample .
abundance parameter as a funtion of excitation , as measured by the [ sii]/[siii ] ratio . ,
title="fig:",width=244,height=192 ] abundance parameter as a funtion of excitation , as measured by the [ sii]/[siii ] ratio . ,
title="fig:",width=244,height=192 ] the right panel of figure 4 shows the run of the excitation degree with metallicity for the observed regions through the [ sii]/[siii ] ratio , which has been shown to be a good ionization parameter indicator ( daz et al .
1991 ) , and the @xmath11 parameter .
it can be seen that the observed cnsfr show the lowest excitation of the sample .
a hint on the ionizing temperature of the regions can be obtained through the use of the @xmath0 parameter , which is a measure of the softness of the ionizing radiation ( see vlchez & pagel 1988 ) and increases with decreasing ionizing temperature .
the left panel of figure 5 shows the run of @xmath0 with metallicity as parametrized by @xmath11 .
unexpectedly , cnsfr show values of @xmath0 higher than those of over - solar disc hii regions .
this is better appreciated in the right panel of the figure where cnsfr are seen to segregate from disc hii regions in the [ oii]/[oiii ] _ vs _ [ sii]/[siii ] diagram .
the former cluster around the value of log@xmath0 = 0.0 ( t@xmath14 40,000 k ) while the latter cluster around log @xmath0 = 0.8 ( t@xmath14 35,000 k ) .
_ vs _ @xmath0 relation .
_ right panel : _ logarithmic relation between the [ oii]/[oiii ] and [ sii]/[siii ] line ratios , title="fig:",width=244,height=192 ] _ vs _ @xmath0 relation .
_ right panel : _ logarithmic relation between the [ oii]/[oiii ] and [ sii]/[siii ] line ratios , title="fig:",width=244,height=192 ] | we present a spectrophotometric study of circumnuclear star forming regions ( cnsfr ) in the early type spiral galaxies : ngc 2903 , ngc 3351 and ngc 3504 , all of them of over solar metallicity according to standard empirical calibrations .
a detailed determination of their abundances is made after careful subtraction of the very prominent underlying stellar absorption .
it is found that most regions show the highest abundances in hii region - like objects .
the relative n / o and s / o abundances are discussed . in is also shown that cnsfr , as a class , segregate from the disk hii region family , clustering around smaller softness parameter "
@xmath0 ' values , and therefore higher ionizing temperatures . [ 1996/06/01 ] |
systems with many degrees of freedom which are pinned in some external potential are very common in condensed matter .
examples are fluid - fluid interfaces in porous media @xcite , flux - lattices in type - ii superconductors @xcite , and charge density waves @xcite to mention only a few . also dry friction ( i.e. , solid - solid friction ) belongs to this class of systems because the asperities of the surfaces interlock . a common feature of all these systems is a strongly nonlinear mobility .
if one applies some field or force @xmath0 on the system , the mobility is zero below some usually well - defined threshold @xmath1 . above this threshold
the mobility is nonzero . in general
it is some nonlinear function of the applied force @xmath0 .
the transition from a pinned system with zero mobility to a depinned one with some finite , nonzero mobility is called the _ pinning - depinning transition_. it can be understood as a kind of `` melting '' which happens far from thermal equilibrium .
the process is a typical nonequilibrium one because of two reasons .
first , there is no ground state for @xmath2 and the pinned system has to be in some metastable state . due to thermal fluctuations
the system can overcome the barrier of the metastable state and move into another metastable state with less energy .
this phenomenon leads to _ creeping _ with a very low mobility .
second , beyond the pinning - depinning transition energy flows through the system at a constant rate which is given by the mobility times @xmath3 .
this flow is usually not small .
thus it can not be deduced from linear response theory which works only near thermal equilibrium .
the mobility of the sliding state strongly depends on the kind of energy dissipation .
the inverse process of this nonequilibrium melting is the _ depinning - pinning transition _ which is a kind of nonequilibrium `` freezing '' .
both kind of transitions do not have to occur at the same value of the applied force @xmath0 .
the behavior depends strongly on whether the degrees of freedom ( i.e. , flux lines , atoms etc . ) have inertia or inertia is negligible compared to dissipative forces . in the case of strong dissipation
the motion is overdamped .
the pinning - depinning transition is in most cases of second order and indistinguishable from the depinning - pinning transition .
typical examples for such systems are flux lines in type - ii superconductors and charge - density waves . if the motion is underdamped , hysteresis is possible because the inertia can overcome a pinning center .
this is intuitively clear if one imagines the simplest model system of this kind , namely a particle in a spatially periodic potential @xcite .
another important aspect of the collective behavior of pinned systems is whether the potential caused by the pinning centers is regular or irregular ( quenched randomness ) .
often the pinning landscape is random .
this case together with a purely dissipative diffusion - like dynamics has been studied extensively in the literature @xcite .
the aim of this paper is to study the opposite case in a fairly simple model , namely , the frenkel - kontorova ( fk ) model @xcite .
there is no quenched randomness .
all pinning centers are identical forming a regular array .
furthermore , all pinned objects are identical and have a mass .
the damping is assumed to be weak .
we will see that weak damping is responsible for randomness that is caused by chaotic motion .
important physical applications of the fk model are arrays of identical josephson junctions @xcite and adsorbate layers on clean crystal surfaces @xcite . in this paper we consider the one - dimensional fk model .
the equation of motion ( in dimensionless units ) reads @xmath4 where @xmath5 is the position of particle @xmath6 , @xmath7 is the damping constant , @xmath8 is the strength of the external potential , and @xmath0 is the external force .
the time derivative is denoted by a dot . in order to avoid effects due to the boundary layers we choose periodic boundary conditions ,
i.e. , @xmath9 where @xmath10 is the number of particles and @xmath11 is an arbitrary integer .
the periodic boundary conditions fixes the average particle distance @xmath12 to @xmath13 .
because of symmetry @xmath12 can be restricted to interval @xmath14 $ ] without loss of generality .. together with the previous paper @xcite in which we have already investigated periodic and quasiperiodic solutions , the aim is to give a detailed investigation of the dynamical behavior in the weakly damped case for long chains ( i.e. , @xmath15 ) . in this paper
we deal with spatio - temporal chaotic solutions ( called _ fluid - sliding states _ ) and the transition between these solutions and the stationary states .
the typical behavior is summarized in figure [ f.vf.chaos ] .
[ f.vf.chaos](a ) shows the velocity - force characteristic .
we see hysteresis loops between three different branches which belong to different types of solutions .
the states with the largest average sliding velocities @xmath16 are the _ solid - sliding states _ which are characterized by a chain with nearly no internal vibrations .
these states have the maximum possible mobility , i.e. , @xmath17 .
the solid - sliding state becomes unstable due to first - order parametric resonance if @xmath16 is below some critical value@xcite .
the second type of sliding state is the fluid - sliding state . in general
one can summarize all sliding states having not the maximum mobility into this category . but strictly speaking the name makes only sense if these states are spatio - temporally chaotic . for larger values of the damping constant
this is not the case as we have seen in the previous paper .
the chaotic vibration in the fluid - sliding state can be characterized by an _
effective temperature_. the temperature of the fluid - sliding states of fig .
[ f.vf.chaos](a ) is shown in figure [ f.vf.chaos](b ) . even though the distribution of the particle velocity @xmath18 is gaussian
[ see inset of fig .
[ f.vf.chaos](b ) ] we will show that the fluid - sliding state is a _
nonequilibrium _ state . a very specific test to show this is the violation of the equipartition theorem for the phonon modes ( see sec .
[ stc ] ) .
the third type of states are the stationary ones .
their mobility is zero . in order to model the creeping due to thermal activation
one has to add a white - noise term to the equation of motion ( [ int : eqm ] ) .
we have not done this because the qualitative behavior does not change very much as long as the thermal energy is much smaller than the amplitude of the external potential .
this is confirmed in numerical simulations of similar models @xcite .
for example , the hysteresis seen in fig . [
f.vf.chaos](a ) still exist for nonzero but small temperatures @xcite . for this behavior it seems to be important that the system has many degrees of freedom because in the case of @xmath19 the hysteresis disappears even for infinitesimal small noise amplitude @xcite .
[ f.vf.chaos](b ) clearly shows that nonequilibrium melting and freezing is accompanied by an abrupt change of the temperature of the chain .
the transition is like a first - order one in thermal equilibrium but the pinning - depinning transition point is larger than the depinning - pinning transition point .
thus hysteresis occurs .
the transition points fluctuate [ see inset of fig .
[ f.vf.chaos](a ) ] , especially the pinning - depinning transition point .
the paper is organized as follows : in section [ cs ] we investigate in detail the fluid - sliding state .
we show that it is indeed spatio - temporally chaotic . for chains with @xmath20 near an integer value we found a pronounced transition from a kink - dominated sliding state and the fluid - sliding state .
this transition is relatively sharp even though there is no hysteresis .
but it becomes hysteretic for small @xmath10 .
the depinning - pinning transition and the pinning - depinning transition are discussed in section [ mf ] .
we show that nonequilibrium freezing is similar to ordinary freezing whereas melting is clearly different .
the pinning - depinning transition point depends on the stationary state .
furthermore local rearrangements ( micro - slips ) of the chain may occur before the transition . in section [ con ]
we compare our results with results of similar models .
decreasing the damping constant @xmath7 increases the complexity of the sliding state from periodic motion via quasi - periodic motion ( which is usually spatially chaotic , see the preceding paper ) to spatio - temporal chaos .
an example of the latter one is shown in figure [ f.chaos ] .
the aim of this section is to investigate and to characterize the chaotic - sliding state which we call the _ fluid - sliding state_. first of all we see in fig .
[ f.vf.chaos](a ) that the velocity - force characteristic of this state is nearly structureless .
this has to be compared with the case of periodic and quasiperiodic motion where a multitude of hysteresis loops appear ( see the preceding paper ) . here
there occur only hysteresis loops between the solid - sliding state ( where the particles are shaken so fast that they nearly do not feel the external potential ) , the fluid - sliding state and the stationary states .
fig [ f.chaos ] is of course not a proof that the chain slides chaotically .
it is well - known that chaotic motion is characterized by the sensitivity on the initial conditions .
it is measured by the largest lyapunov exponent @xmath21 , which is the rate of divergence ( or convergence , if it is negative ) of trajectories in phase space that start out infinitely close to each other @xcite .
figure [ f.lmax ] shows that the fluid - sliding states in figs .
[ f.chaos ] and [ f.vf.chaos ] are indeed temporally chaotic . but
is it also spatially chaotic ? in order to answer this question we have calculated the normalized velocity correlation function @xmath22 defined by @xmath23 where @xmath24 for the same parameters as in fig .
[ f.chaos ] the result is shown in figure [ f.cf ] .
one clearly sees that @xmath22 is a rapidly decaying oscillatory function .
the envelope seems to be proportional to @xmath25 with a correlation length @xmath26 . because of @xmath27 and @xmath28 the fluid - sliding state is spatio - temporally chaotic .
a very strong criterion for spatio - temporal chaos is that the number of positive lyapunov exponents is proportional to @xmath10 for large @xmath10 .
we have calculated the lyapunov spectrum with the method described in ref . for various values of @xmath10 .
figure [ f.lspect ] shows the cumulative density @xmath29 of lyapunov exponents , i.e. , the probability to find a lyapunov exponent larger than @xmath30 .
the result is typical for spatio - temporal chaos @xcite . in the thermodynamic limit
( i.e. , @xmath31 ) the sequence of cumulative densities @xmath32 converges uniformly to @xmath33 .
thus , for large @xmath10 the number of positive lyapunov exponents is indeed proportional to @xmath10 .
[ f.lspect ] shows clearly that the spatio - temporally chaotic nature of the fluid - sliding state does _ not _ depend on the commensurability of the chain .
because the fluid - sliding state is spatio - temporally chaotic it makes sense to introduce an effective _ temperature_. in the dimensionless units of the equation of motion ( [ int : eqm ] ) it is the average kinetic energy in the frame co - moving with the center of mass , i.e. , @xmath34 where @xmath16 is the average sliding velocity @xmath35 in the previous paper we have derived a formula for the applied force @xmath0 in terms of the first and second moments of the particle velocity [ eq .
( ) in ref . ] . with the help of this formula
the temperature can be expressed in terms of the applied force and the average sliding velocity : @xmath36 figure [ f.vf.chaos](b ) shows the temperatures of the fluid - sliding states of the velocity - force characteristic in fig .
[ f.vf.chaos](a ) . even though the temperature of the solid - sliding states and the stationary states is formally zero in accordance with ( [ stc : tf ] )
, it does not make sense to call ( [ stc : t ] ) a `` temperature '' in regular , non - chaotic sliding states or stationary states .
the periodic and quasi - periodic domain - like states , for example , investigated in the previous paper have also non - zero `` temperature '' . in the frame co - moving with the center of mass , the chain is shaken by the washboard wave ( i.e. , the external potential ) and moves in a spatio - temporally chaotic way . the gaussian distributed velocities might suggest that the chain is in thermal equilibrium .
but is it true ?
this raises the following question of general interest : _ can we replace the spatio - temporally chaotic chain by an equivalent system which is in thermal equilibrium _ ? or more general , is it possible to describe the chaotic attractor of a weakly damped and strongly driven hamiltonian system with many degrees of freedom ( infinitely many in the thermodynamic limit ) by an equivalent undriven and undamped system ? as a consequence of a positive answer one would expect that the equipartition theorem from thermodynamics holds , i.e. the ensemble averages of @xmath37 and @xmath38 are independent of @xmath6 [ @xmath39 is the hamilton function , the @xmath40 s are the generalized coordinates , and the @xmath41 s are the corresponding canonical momenta ] . in numerical simulations
one usually replaces the ensemble average by the temporal average assuming that the ergodicity hypothesis holds .
an obvious candidate for a test of the equipartition theorem would be the particle momentum in the co - moving frame , i.e. , @xmath42 . but this is not a wise choice : because of symmetry the result has to independent of @xmath6 .
a better choice is its spatial fourier transform , i.e. , @xmath43 that is , we want to check whether the average kinetic energy @xmath44 of the phonon modes is equipartitioned or not . figure [ f.phmod ] shows that the equipartition theorem is not fulfilled @xcite .
this is a clear signature for the fact that the _ fluid - sliding state is a state far away from thermal equilibrium_. therefore it is not possible to develop a theory for this state based on equilibrium thermodynamic .
the violation of the equipartition theorem is equivalent with non - zero velocity correlations for @xmath45 because @xmath46 is the modulus of the fourier transform of @xmath22 . when @xmath20 approaches an integer value the velocity - force characteristic of the fluid - sliding state develops a relatively sharp transition step at a characteristic value of the applied force @xmath0 .
an example for @xmath47 is shown in figure [ f.vf.chao2 ] . for long enough chains
no hysteresis is observable . for small chains we get bistability between different types of sliding states .
similar results has been found in a generalized fk model by braun _
the aim of this section is to answer the following obvious questions : what is the nature of the different sliding states ?
why does the bistability depends on @xmath10
? can we understand this transition and where does it occur ? first we take a more detailed look at the dynamics below and above the transition ( see fig .
[ f.sst ] ) .
the motion far below the threshold is almost regular .
it corresponds to one of the multi - domain states we have discussed in the previous paper .
there are two domain types , a stationary one with @xmath48 [ it is responsible for the tilted lines in fig .
[ f.sst](a ) ] and a sliding one .
often the sliding domains are so small that they are actually @xmath49-kinks , and larger sliding domains can be interpreted as clusters of @xmath49-kinks @xcite .
that is , multi - domain states like the example of fig .
[ f.sst](a ) are nonuniform distributions of @xmath49-kinks .
thus , we call this state the _ kink - dominated sliding state_. note , that there are @xmath11 kinks but no antikink .
it is well - known that kinks ( and antikinks ) can not travel faster than the sound velocity ( which is equal to one in our case ) .
each kink or antikink needs therefore at least @xmath10 time units to travel through the whole chain . after that time
a chain with @xmath11 kinks will be shifted by @xmath50 .
therefore , the average sliding velocity of a state like the one of fig .
[ f.sst](a ) has to be less than @xmath51 .
[ f.vf.chao2 ] shows that it is actually much below the sound velocity .
if the transition point is approached from below , the behavior depends on whether the chain is long or short . for long chains with many kinks ,
the average sliding velocity @xmath16 starts to increase with @xmath0 faster and faster .
later on the increase slows down .
we define the transition point @xmath52 as the value of @xmath0 where the slope of @xmath53 has a maximum .
after the transition point the system is in a fluid - sliding state [ see fig .
[ f.sst](b ) ] .
all kinks ( and antikinks ) have disappeared , and the system is completely spatio - temporally chaotic .
a short chain with only a few kinks still stays in the kink - dominated regime beyond @xmath52 .
eventually , it jumps to the fluid - sliding state or to the solid - sliding state ( see fig .
[ f.vf.chao2 ] ) .
a hysteresis occurs , and the chain goes back to the kink - dominated state at @xmath54 . at the transition point the sliding velocity strongly fluctuates .
these fluctuations set in already much below the transition point .
they lead to a larger value of the average sliding velocity compared to the value for small chains .
a typical example is shown in figure [ f.fluct ] .
one sees bursts of activity above a level given by the value of @xmath16 for small chains ( see fig .
[ f.vf.chao2 ] ) .
a detailed look onto the dynamics of the chain reveals that an increase of @xmath16 is caused by the _ production of kink - antikink pairs _
these pairs usually appear behind a @xmath49-kink cluster
. this may be the reason why for small chains the kink - dominated states survive beyond the transition point because the probability for a @xmath49-kink cluster is too small .
each kink and antikink contributes to the sliding velocity of the chain .
that is , @xmath16 is given by @xmath55 where @xmath56 is the velocity of the kinks and antikinks and @xmath57 is the number of kink - antikink pairs . in the quiescent state without kink - antikink pairs ,
the average sliding velocity is given by @xmath58 .
thus the ratio @xmath59 gives directly the total number of kinks and antikinks ( i.e. , @xmath60 ) .
figure [ f.fluctd ] depicts the distribution of the sliding velocity shown in fig .
[ f.fluct ] .
in addition to the large peak at the quiescent state one sees clearly small peaks for @xmath61 and @xmath62 . for @xmath12 approaching zero ( or any other integer multiple of @xmath49 ) the average sliding velocity of the kink - dominated sliding state also approaches zero .
it disappears at @xmath48 .
thus , the system goes from the fluid - sliding state directly to a stationary state when @xmath0 is decreased below the transition point @xmath52 .
if the value of @xmath8 is not too large [ i.e. , @xmath63 , the stationary state will be the ground state @xcite .
the transition is often accompanied by transients where a few kink - antikink pairs survive for a considerably long time .
but in all simulations these pairs eventually disappeared .
we found that the transition point @xmath52 is nearly independent of @xmath12 ( as long as @xmath20 is near an integer value ) .
table [ t.fsst ] shows that @xmath52 increases with @xmath8 weaker than linearly . in order to understand the transition between the fluid - sliding state and the kink - dominated sliding state we compare the fk model with a simpler model , namely , one particle in a tilted spatially periodic potential plus additive white noise
this model was studied in details by risken and vollmer @xcite .
the single particle and the noise correspond to the center of mass of the fk model and the chaotic motion of the internal degrees of freedom , respectively .
of course the noise is neither additive nor white .
its strength depends on the state .
it is obvious that the solid - sliding state and the stationary state of the fk model correspond to the running state and the locked state of the simple model in the absence of noise .
we suggest that the fluid - sliding state and the kink - dominated state of the fk model correspond also to the running and locked states of the simple model but now with noise .
risken and vollmer showed that the bistability between the running state and the locked state disappears even for infinitesimally weak noise .
there is a well - defined transition point @xmath64 which is smeared out for increasing noise strength .
for @xmath65 it is given by @xmath66 thus we expect @xmath67 .
table [ t.fsst ] shows that this is indeed the case especially for small values of @xmath8 .
ccc @xmath8&@xmath52&@xmath64 + + 0.25 & 0.096 & 0.0840 + 0.5 & 0.130 & 0.1187 + 1 & 0.187 & 0.1679 + 2 & 0.276 & 0.2374 + 3 & 0.362 & 0.2908 + 4 & 0.444 & 0.3358 + 5 & 0.519 & 0.3754 +
the aim of this section is to take a closer look at the hysteresis loop between stationary states and the fluid - sliding state ( see fig .
[ f.vf.chaos ] ) .
there are two different types of transitions .
a depinning - pinning transition where the ( chaotically ) sliding chain turns into a stationary one .
it is a kind of nonequilibrium freezing and , as in usual first - order phase transitions , the chain can be `` supercooled '' below the transition point @xmath68 .
that is , for @xmath0 slightly below @xmath68 the chain does not freeze immediately .
it takes a while until a critical nucleus has appeared .
the second transition is the pinning - depinning transition .
the transition point @xmath69 depends on the stationary state and therefore on the history of the system . in an actual experiment where one sweeps through the hysteresis loop at a finite rate
, one will therefore not get well - defined transitions points but more or less broad distributions .
an example of such a ( numerical ) experiment is shown in the inset of fig .
[ f.vf.chaos](a ) .
it is typical that the distribution of @xmath68 is narrower than the distribution of @xmath69 . for a quasistatic sweep the distribution of @xmath70 becomes sharp , whereas the width of the distribution of @xmath71 is nearly independent on the sweeping rate . from equilibrium thermodynamics it is well - known that melting and freezing occur at the same temperature and
the transition is of first - order . in our case
far from thermal equilibrium the situation is different . in the overdamped
limit the pinning - depinning transition point @xmath69 is identical with the depinning - pinning transition @xmath68 but the transition is of second order @xcite . in the underdamped case we have bistability between sliding states and stationary states ,
i.e. , @xmath72 .
a further characteristic feature of nonequilibrium melting and freezing is the change of the effective temperature [ see e.g. fig .
[ f.vf.chaos](b ) ] .
this is a general property which is also found in similar models with nonzero temperature of the environment @xcite .
that is , the fluid - sliding state has an effective temperature which is _ independent _ on the temperature of the environment as long as the latter one is considerably less than the former one .
the effective temperature is a measure of the enhanced energy flow due to the excitation of phonons .
we have studied in detail the depinning - pinning transition from a fluid - sliding state to a stationary state .
we have chosen a value of @xmath12 where the fluid - sliding state is completely spatio - temporally chaotic .
like in an ordinary first - order phase transition at thermal equilibrium the transition is caused by a _
nucleation process_. that is , a small portion of the chain becomes stationary , and the fronts between this stationary nucleus and the sliding chain propagate into the chain .
an example of such a nucleation process is shown in figure [ f.nuc ] .
one sees that states appear behind the fronts which can be characterized by an average particle distance @xmath12 ( or density @xmath73 ) which is roughly constant .
note that the values of @xmath12 behind both fronts have to be different .
this can be understood by the following argument . in the previous paper
we have seen that because of the conservation of the number of particles the velocity of a front traveling from one particle to the next is given by @xmath74 , where the average particle distance and the average sliding velocity on both sides of the front are given by @xmath75 and @xmath76 , respectively .
the chaotic sliding state is characterized by @xmath77 and @xmath78 . for the stationary state @xmath79 holds . in order to have fronts traveling in opposite directions ( see fig .
[ f.nuc ] ) the average particle distances of the stationary states have to be different ( one larger than @xmath12 and one smaller than @xmath12 ) .
we found that they are always the nearest integer multiples of @xmath80 , that is , for @xmath81 they are zero and @xmath80 . from this consideration
it is clear that after the depinning - pinning transition the system can not be in the ground state @xcite which is characterized by a single domain with a uniform @xmath12 .
instead the chain will be in a stationary _ two - domain state _ which is not quite perfect because each domain contains defects at low density .
a similar depinning - pinning transition was found in a seemingly completely different model , namely , the quenched kadar - parisi - zhang equation with negative nonlinearity @xcite .
it is a nonlinear diffusion equation which models the surface growth in disordered media . here
the dissipation is strong and the randomness of the pinning landscape is important contrary to the weakly damped fk model .
nevertheless the pinning - depinning transition is of first order and the state which freezes out of the sliding state is a two - domain state , as in the fk model ( compare fig . 2 of ref . with fig .
[ f.nuc ] ) .
the duration of freezing @xmath82 is the sum of the nucleation time @xmath83 and the growth time @xmath84 .
the nucleation time is the time that evolves until a critical nucleus appears .
a critical nucleus is a nucleus which is large enough to grow into the fluid - sliding state .
the nucleation time @xmath83 will be a poisson distribution if the probability for the appearance of a critical nucleus in a short time - step is small , i.e. , @xmath85 it is well - known that the average and the standard deviation of a poisson - distributed value are identical , i.e. , @xmath86 . for systems much larger than the critical nucleus , the probability for nucleation increases linearly with the system size , i.e. , @xmath87 .
the time @xmath84 for a critical nucleus to grow up to the stationary state does not fluctuate as strongly as @xmath83 because it is roughly given by the system size divided by the sum of front velocities .
thus @xmath88 .
figure [ f.poi ] shows that the nucleation process is indeed characterized by a poisson distribution with @xmath87 . from fits of the curves shown in fig .
[ f.poi ] we found @xmath89 .
one can also see a shift of the distribution for increasing @xmath10 to larger values of @xmath82 .
this reflects the fact that @xmath90 increases with @xmath10 . for values of @xmath12 between 2.45 and 3.09 ( and @xmath91 , @xmath92 , @xmath93 , and @xmath54 ) we found that the fluid - sliding state changes its character .
a domain appears where the particles are stationary ( with two particles per potential well ) .
this domain which is surrounded by spatio - temporal chaos has a constant size and travels through the chain with a constant velocity .
the transition from this so - called _ traffic - jam state _
@xcite to a stationary one occurs also via nucleation .
the critical nucleus seems to appear always at the back of the stationary domain .
it reverses the propagation direction of the front .
that is , the chaotic state stops traveling into the stationary one . instead a front propagates into the chaotic state leaving behind a stationary state different from the already existing one .
thus the result is again a stationary two - domain state where one domain already exists at least partially before the transition .
the nucleation probability is roughly independent of the system size @xmath10 because the nucleation site is predetermined .
the transition from a stationary chain to a sliding one is the pinning - depinning transition .
the transition point @xmath69 depends on the stationary state .
it corresponds to a saddle - node bifurcation where the particular stationary state annihilates with an unstable stationary state . in the overdamped limit @xmath69
is uniquely defined by the saddle - node bifurcation of the last stable stationary state which is usually the state which develops adiabatically out of the ground state for @xmath94 .
the disappearances of the other stationary states in a saddle - node bifurcation lead only to more or less local rearrangements of the chain .
we call such a rearrangements a _ micro - slip_. a micro - slip changes the center of mass of the chain but does not lead to sliding .
an example is shown in figure [ f.mslip ] .
this behavior and the statistics of micro - slips has been studied mainly in models with random external potential @xcite . in the underdamped regime
the energy gained from a local rearrangement is not dissipated immediately .
this may lead to an avalanche which turns the whole system into a sliding state . whether a saddle - node bifurcation leads to a micro - slip or to a transition to sliding depends on the stationary state and on the damping constant @xmath7 . for
each stationary state one can presumably find a critical value @xmath95 which distinguishes both cases : for @xmath96 we get a micro - slip , otherwise a transition to sliding . in our simulations we found only a few ( not more than three ) micro - slips for @xmath92 and @xmath91 .
the pinning - depinning transition point depends strongly on the history of the system because the actual value of @xmath69 depends on the stationary state .
sweeping several times through the pinning - depinning hysteresis loops , we find a broad distribution of @xmath69 even for very small sweep rates .
in this paper we have shown that in the _ weakly _ damped and strongly driven fk model spatio - temporal chaos appears .
this chaotic state , called _ fluid - sliding state _ , has an effective temperature which reflects the fact that phonons are excited .
this has to be contrasted by the solid - sliding state where no phonons are excited @xcite .
the excitation of phonons opens up additional channels of dissipation .
thus friction in the fluid - sliding state is larger than in the solid - sliding state .
the velocity - force characteristic shows a pronounced transition if the average particle distance @xmath12 is near but not identical to an integer multiple of the period of the external potential .
it is a transition between the kink - dominated sliding state and the fluid - sliding state . below the transition point @xmath52 sliding
is caused by the propagation of @xmath49 kinks .
approaching the transition point from below leads to an increasing production of kink - antikink pairs . above the transition point
all kinks and antikinks disappear and the chain is in the fluid - sliding state with an average sliding velocity which depends only weakly on the average particle density @xmath73 . in the kink - dominated regime
the average sliding velocity is proportional to @xmath97 .
the transition point @xmath52 is roughly independent of @xmath12 . for small values of @xmath8
it is very well approximated by the transition point between the running and locked solution of an ( @xmath19)-fk model with infinitesimally small additive white noise [ i.e. eq . ( [ sst : f2 ] ) ] .
this raises the question whether the dynamics can be reduced to a center - of - mass motion plus some nontrivial ( e.g. colored , state - dependent ) noise term .
the nonequilibrium freezing ( i.e. , the transition from sliding to stationarity ) of the fluid - sliding state is a nucleation process .
the resulting stationary state has two domains of different particle densities . the nonequilibrium melting ( i.e. , the transition from a stationary state to sliding ) of such a stationary state corresponds to a saddle - node bifurcation where the stationary and , of course , the stable state annihilates with its unstable counterpart . in the case of nonzero environmental temperature
the transition occurs a bit earlier because thermal activation overcomes the barrier which decreases to zero at the saddle - node bifurcation .
not all saddle - node bifurcations lead to melting .
they may lead to a local rearrangement of the particle configuration called micro - slip .
the pinning - depinning transition point depends on the history because each stationary state has another bifurcation point .
our results are qualitatively very similar to results of similar models .
persson studied a two - dimensional lennard - jones liquid on a corrugated potential with square symmetry at a nonzero temperature @xcite .
he varied the temperature between @xmath98 and @xmath99 of the melting temperature ( the thermal energy was roughly a tenth of activation energy for single particle diffusion ) .
he obtained velocity - force characteristics and effective temperature plots similar to fig [ f.vf.chaos ] . because of the relatively high temperature of the environment he did not find hysteresis between the solid - sliding state and the fluid - sliding state . as in the fk model , the transition from the fluid - sliding state to a stationary state is also a nucleation process .
persson found that it occurs when the effective temperature is equal to the melting temperature at thermal equilibrium . because the velocity distribution of the fluid - sliding state was gaussian
, he argued that the fluid - sliding state is in a kind of thermal equilibrium .
therefore it has to freeze below the melting temperature .
but as we have seen in sec .
[ stc ] , gaussian distributed velocities do not imply a quasi - thermal equilibrium .
it would be interesting to look whether in persson s model the equipartition theorem of the phonon modes is fulfilled or not .
this is clearly a better test on thermal equilibrium .
granato _ et al .
_ studied a two - dimensional frenkel - kontorova model at nonzero temperatures @xcite .
the external potential is corrugated only in the direction of the applied force .
also the particles can move only in this direction .
they did the simulations at a temperature which corresponds to @xmath100 of the activation energy and which is roughly @xmath101 of the melting temperature .
they also found a behavior like in fig .
[ f.vf.chaos ] again without a hysteresis between the solid - sliding state and fluid - sliding state .
they also confirmed the observation of persson that the transition from sliding to stationarity occurs at the melting temperature .
braun _ et al .
_ investigated a generalized fk model that is quite similar to the model studied by persson @xcite .
the main difference is that the substrate potential is anharmonic and the interaction potential is an exponential repulsion .
they did the simulations mainly for two different temperatures which correspond to @xmath102 and @xmath103 of the activation energy for single particle diffusion .
they report results for one - dimensional chains of atoms as well as for two - dimensional layers .
there seems to be no qualitative differences between one and two dimension . again
velocity - force characteristics and effective temperature plots are similar to our results .
they found a hysteresis between solid sliding and fluid sliding .
its width decreases with the environmental temperature and disappears eventually ( at a temperature which is roughly @xmath101 of the activation barrier for hopping of uncoupled single particles ) .
near fully commensurate particle densities they found a transition ( denoted by @xmath104 ) which is similar to the transition from the kink - dominated sliding regime to the fluid - sliding regime of the fk model . because their chain was short ( i.e. , @xmath105 ) this transition shows hysteresis for very low temperatures . instead of a fully chaotic fluid - sliding regime they always found two - domain states with alternatively running and locked particles ( traffic - jam regime ) .
they also measured the transition point @xmath104 as function of the damping constant @xcite .
it increases roughly linear with the damping constant but the authors seemed not to be aware that ( [ sst : f2 ] ) is again a remarkable good approximation ( i.e. errors are less than 10% ) for @xmath104 .
we thank h. thomas for critical reading of the manuscript .
we also acknowledge the possibility to do simulations on the nec sx-3 and sx-4 at the centro svizzero di calcolo scientifico at manno , switzerland .
this work was supported by the swiss national science foundation . by the way , we checked in microcanonical simulations of the thermal equlibrium ( i.e. , no damping , @xmath106 , no driving , @xmath94 ) , that for the same values of @xmath8 and @xmath107 as in fig .
[ f.phmod ] the kinetic energy @xmath46 of the phonon modes are indeed equipartitioned . | the dynamical behavior of a weakly damped harmonic chain in a spatially periodic potential ( frenkel - kontorova model ) under the subject of an external force is investigated .
we show that the chain can be in a spatio - temporally chaotic state called fluid - sliding state .
this is proven by calculating correlation functions and lyapunov spectra .
an effective temperature is attributed to the fluid - sliding state . even though the velocity fluctuations are gaussian distributed , the fluid - sliding state is clearly not in equilibrium because the equipartition theorem is violated .
we also study the transition between frozen states ( stationary solutions ) and molten states ( fluid - sliding states ) .
the transition is similar to a first - order phase transition , and it shows hysteresis . the depinning - pinning transition ( freezing ) is a nucleation process .
the frozen state contains usually two domains of different particle densities .
the pinning - depinning transition ( melting ) is caused by saddle - node bifurcations of the stationary states .
it depends on the history .
melting is accompanied by precursors , called micro - slips , which reconfigurate the chain locally .
even though we investigate the dynamics at zero temperature , the behavior of the frenkel - kontorova model is qualitatively similar to the behavior of similar models at nonzero temperature . |
the reported enhancement of strange particle production in heavy ion collisions relative to proton - proton and proton - nucleus collisions @xcite tells us that we are dealing with a state of matter which is _ i ) _ very dense or _ ii ) _ relatively long - lived , or in which _ iii ) _ strangeness production cross sections are enhanced , or which _ iv ) _ possesses some combination of these three factors . in order to be able to be more specific about the nature of the dense matter , the relative production abundances of strange and multi - strange baryons and anti - baryons were studied .
they turned out to be particularly sensitive probes of the thermal conditions of their source @xcite .
studies of the properties of a hadron resonance gas fireball including strange particles have shown @xcite that the constraint of vanishing net strangeness of the emitting source leads to similar thermal and chemical conditions for both a hadronic gas ( hg ) and a deconfined quark - gluon plasma ( qgp ) fireball , if the temperature is taken to be near @xmath0 mev .
this value was extracted @xcite from the @xmath1-spectra from 200 geva s w and s pb collisions under the assumption that their slope is dominated by the thermal motion of the particles at freeze - out and does not contain a collective flow component @xcite . we will discuss that the observed smallness of the strange quark chemical potential appears to be inconsistent with the presence of a strong flow component in the spectra .
it has already been pointed out @xcite that near @xmath2 mev further information , such as the particle multiplicity per participant nucleon , may help to distinguish qgp from the hg . following this route , we will show here in detail @xcite that the data of emulsion experiment emu05 @xcite indeed appear to be incompatible on these grounds with the hadronic gas fireball model .
further support for a new physical state such as the quark - gluon plasma as the source of these anti - baryons may be obtained by studying the response of the measured parameters to changes in beam energy or size of the colliding nuclei .
our work has been in part motivated by the imminent extension of the experimental program at cern and bnl to heavy nuclear projectiles ( pb and au , respectively ) , and by the prospect of higher ( and lower ) beam energies becoming available , in particular towards the end of the decade at brookhaven with rhic and possibly at cern with lhc . therefore , in addition to @xmath3 mev appropriate for the cern sps energy range , we also include in our discussion the parameter range ( @xmath4 mev ) suitable for the current bnl heavy - ion experiments at 1015 geva and for potential future low beam energy runs at cern , as well as @xmath5 mev , which we judge appropriate for the future rhic and lhc facilities @xcite .
our expectation is that , as the conditions of the experiments are changed , we will be able to study the response of the statistical parameters such as the strange quark chemical potential @xmath6 or the strangeness saturation factor @xmath7 .
these parameters show characteristic differences in a hadron gas and a qgp : @xmath7 , for example , is expected to rapidly approach unity for increasing four - volume of the qgp fireball , while remaining small in hg fireballs @xcite .
the strange quark chemical potential @xmath6 , on the other hand , has the characteristic property that it is exactly zero in a strangeness - zero qgp fireball , independent of its baryon density .
in contrast to this , @xmath6 is generally different from zero in conventional hadronic gas reactions , except for the special case of baryon - free matter with @xmath8 ; however , if @xmath9 is not too large ( @xmath10 mev ) , for each value of the temperature a certain value of baryochemical potential @xmath11 can be found for which also @xmath12 .
this value generally depends on the detailed form of the equation of state ( eos ) , and for the case of a conventional ( hagedorn type ) hg it has been discussed in refs .
@xcite . while the large values of @xmath7 which are characteristic of the fast chemical kinetics in a qgp are largely preserved during hadronization ,
the value of @xmath6 is usually subject to modifications , in particular if the phase transformation occurs slowly .
experimentally , the values of @xmath6 and @xmath7 are reflected in the observed strange hadron ratios ( where , as we will discuss , in the case of qgp certain model assumptions for the hadronization process are necessary in order to quantify this relationship ) , which allows for a rather straightforward experimental test of the fireball picture , of the concept of ( absolute or relative ) chemical equilibrium , and of the nature ( hg or qgp ) of the collision fireball . in order to distinguish between the two phases by studying particle production as the conditions of the experiments
are changed we need to understand also in the hg phase how the strange chemical potential @xmath6 relates to the baryochemical potential @xmath13 in the collision fireball .
such a relationship is required by the fact that the total strangeness in the initial state is zero , and that strangeness is conserved by strong interactions .
thus , up to ( presumably small ) pre - equilibrium emission effects which can introduce a small strangeness asymmetry @xmath14 , the number of strange and anti - strange quarks in the fireball are equal ( @xmath15 ) .
we work out in detail how the strange chemical potential is determined at fixed temperature by the baryochemical potential and a given ( small ) fraction @xmath16 of net strangeness in the fireball .
we find that at a fireball temperature around @xmath17 mev , there is a peculiar behavior of @xmath6 which causes a particular insensitivity of the strange anti - baryon ratios alone to the structure of the source . in order to be able to reach a conclusion about the internal structure of the source
, we then study the entropy content of the particle source .
we also show that the kaon to hyperon ratio is a poor signature for the nature of the source , but could be a good measure of the baryochemical potential @xcite .
the hg part of our work is similar in spirit , but different in detail from the parallel effort of cleymans and satz @xcite of the ( anti-)hyperon number , while we prefer to work directly with the strange quark chemical potential @xmath6 .
the relationship between their @xmath18 and our @xmath6 is given by @xmath19.[foot1 ] ] .
we study particle ratios at fixed transverse mass @xmath20 gev .
we allow for the strangeness phase space to be only partially saturated and show how the degree of saturation @xmath7 can be experimentally measured in order to later serve as a test for the kinetic theory of strangeness production .
we allow the strangeness to be slightly unbalanced ( up to 10% ) and discuss the corresponding variations in the observables .
furthermore we compare the entropy content for the two different scenarios ( hg and qgp ) and confront the findings with the observed particle multiplicity .
in particular we consider the ratio of net charge to total charged multiplicity , @xmath21 , which can be measured ( without identifying particles ) with any tracking device within a magnetic field ; we show that this ratio is directly related to the specific entropy @xmath22 of the fireball , which should be considerably larger for a qgp than for a hg @xcite . since , due to the different stopping , the experimental results from s s and s w collisions at 200 geva are not in the same class , we avoid combining these results because this would weaken the conclusions from our analysis .
we do not combine in our analysis the global kaon data with anti - hyperon data , since low @xmath23 kaons , unlike strange anti - baryons , can also arise from peripheral , spectator related processes and other rescattering processes and are therefore not necessarily messengers from the same stages of the collision .
instead we consider the @xmath24 and @xmath25 as well as @xmath26 and @xmath27 ratios which , as we will explain , provide independent tests of the degree of strangeness saturation @xmath7 . at this point
it is worth recording that if a qgp state was formed ( as may be indicated by the relatively large value of @xmath7 , see below ) , the strangeness pair abundance per participating baryon in this state would be about a factor 2.5 larger compared to hadronic gas interactions . on the other hand , there would be a similar enhancement in the produced entropy ( which will manifest itself in the total multiplicity ) , resulting in a difficulty to identify the absolute enhancement of both quantities : the simplest observable , the @xmath28 ratio , is therefore not a very good discriminator in this context @xcite .
as we will show , measurement of the specific entropy @xmath29 , either globally via the total particle multiplicity per participating nucleon , or ( better , see below ) in the same kinematic range where the produced strange particles are observed via the ratio @xmath30 , can remove this ambiguity .
one of the important objectives we pursue is to present a large number of predictions for ( strange ) particle production which are capable to test the internal consistency of the model developed here and can shed new light on the problem of distinguishing the hg and qgp fireballs .
these predictions are quantitative for the s
w collision system at 200 geva , but are of a more qualitative nature for the yet unexplored pb
pb collisions where we can not predict precisely the governing statistical parameters ( temperature , chemical potentials ) .
we also discuss the systematic behavior of the different statistical model observables in changing environments .
perhaps the most notable conclusion we reach concerns the model of hadronization of the fireball : we conclude that the only consistent model of the reaction involves explosive disintegration of a qgp as the source of the observed strange anti - baryons .
this allows the chemical potentials of the quarks in the qgp phase to get transferred without modification to the observed high-@xmath31 strange baryons and anti - baryons , via quark coalescense , and explains naturally the observed vanishing of the strange chemical potential .
since only high @xmath31 domain of strange anti - baryon spectra has been experimentally explored , the simple picture of a complete explosion is not absolutely necessary .
it is conceivable that at low @xmath31 , in particular for low-@xmath32 pions , another slower hadronization mechanism also plays a role .
we do see the need for some mechanism that produces additional pions , in excess over those obtained from the explosive quark coalescence picture , in order to save the entropy balance and to account for the relatively large observed multiplicity densities .
these excess pions ( plus perhaps some other light mesons ) could easily arise from the gluons in the qgp which carry a large fraction of its entropy .
a quantitative formulation of such a hadronization model , however , is not yet available and would require additional data on the detailed origin of the excess multiplicity and its distribution in momentum space .
naturally , these conclusions rest on the assumed validity of the thermal fireball model .
while at present the number of parameters just equals the number of independent observables , the model has considerable predictive power and , with the help of the predictions presented in this paper , its validity will be ( hopefully soon ) tested .
other pictures for the source of the observed hadrons , such as an equilibrated hg or a qgp fireball which evolves slowly and in full equilibrium into a hg fireball , can not account consistently for certain important experimental features of strange anti - baryon production .
in section [ sect2 ] we present our thermal model with special attention given to the parameters describing the chemical equilibrium .
we take into account the @xmath33-@xmath34 asymmetry and study the condition of stangeness balance . in section [ sect3 ]
we determine the thermodynamical parameters using the wa85 data on strange baryon and anti - baryon production ( with and without taking into account resonance decays ) . in section [ sect4 ]
the implication of these parameters on the entropy content of the qgp and hg fireballs and on the multiplicity are discussed .
we then confront our model with emu05 and na35 data . in section [ sect5 ]
we discuss the implications of our results regarding the nature of the initial fireball and the dynamic of its evolution .
we give some concluding remarks concerning the distinction we draw between qgp and hg .
the use of thermal models to interpret data on particle abundances and spectra from nuclear collisions is motivated by the hope and expectation that in collisions between sufficiently large nuclei at sufficiently high energies a state of excited nuclear matter close to local thermodynamic equilibrium can be formed , allowing us to study the thermodynamics of qcd and the possible phase transition from a hg to a qgp at a critical energy density .
since we do not yet reliably know how big the collision system and how large the beam energy has to be for this to occur ( if at all ) , thermal models should be considered as a phenomenological tool to test for such a behavior .
even if successful , the validity of such an approach must be checked later by a more detailed theoretical analysis of the kinetic evolution of the collision fireball .
it has been noted repeatedly that , especially in collisions involving heavy nuclear targets , the observed particle spectra , after correcting for resonance decay effects @xcite , resemble thermal distributions with a common temperature @xmath9 ( although a flow component can not presently be excluded and is actually quite likely @xcite ) , and that so far the interpretation in terms of a generalized thermal model , which assumes equilibration only of the momentum distributions and the light quark abundances , but allows for deviations in particle abundances from _ absolute chemical _ equilibrium , in particular in the strange sector , appears to be consistent with the observed particle ratios @xcite . in this section
we will discuss this model as far as it is relevant for our analysis .
let us assume that in collisions between two nuclei a region with nearly thermal equilibrium conditions is formed near `` central '' rapidity , _
i.e. _ at rest in the center of momentum frame of the projectile and a target tube with the diameter of the projectile . this central fireball from which the observed particles emerge is described by its temperature @xmath9 and by the chemical potentials @xmath35 of the different conserved quark flavors @xmath36 .
since the conserved quantum numbers of hadrons are simply the sum of the corresponding quantum numbers of their quark constituents , the hadronic chemical potentials are given as the sum of constituent quark chemical potentials .
since the strong and electromagnetic interactions do not mix quark flavors , @xmath37 and @xmath38 quarks are separately conserved on the time scale of hadronic collisions , and they can only be produced ( or annihilated ) in pairs : @xmath39 .
this implies that in _ absolute chemical equilibrium _
the chemical potentials for particle and anti - particle flavors are opposite to each other .
this observation can be restated in terms of the particle fugacities which are convenient for counting particles and are related to the chemical potentials by : @xmath40 since the quantum numbers of a hadron are simply the sum of the corresponding quantum numbers of their constituent quarks , the fugacity of each hg species is simply the product of the quark fugacities , _ viz .
_ @xmath41 , etc .
consequently the relation between particle and anti - particle fugacity is in general : @xmath42 if absolute equilibrium is not reached , this relationship can still be maintained by introducing an additional parameter characterizing the approach to equilibrium . for strange particles ,
specifically , absolute chemical equilibrium is not easy to obtain : due to the higher mass threshold , the production of @xmath43 pairs usually proceeds much more slowly than that of @xmath44 and @xmath45 pairs .
this is true both in hadronic interactions and in the quark - gluon plasma @xcite .
in fact , the @xmath46 equilibration time scale is estimated to be of the same order of magnitude as the total lifetime of the collision fireball in the qgp phase @xcite , and much longer than that in the hadronic phase @xcite .
on the other hand , once a certain number of strange - anti - strange pairs has been created in such a hadron gas , their redistribution among the various strange hadronic species is no longer hindered by any thresholds and occurs relatively fast .
it thus makes sense to introduce the concept of _ relative chemical equilibrium _ , in which the strange phase space is not fully saturated , but whatever strangeness has already been created is distributed among the available strange hadron channels according to the law of maximum entropy . parametrizing the approach to strange phase space saturation by a factor @xmath47 ,
the individual hadron fugacities are then given by @xmath48 where the @xmath49 still satisfy the chemical equilibrium condition ( [ lam ] ) , and where the power @xmath50 counts the total number of strange valence quarks _ plus _ anti - quarks in hadron species @xmath51 .
altogether , within the thermal model the fireball is thus described by 5 parameters : its temperature @xmath9 , its baryon chemical potential @xmath52 the strangeness chemical potential @xmath6 , the light quark asymmetry potential @xmath53 ( see subsection [ 2.1 ] ) and the strangeness saturation factor @xmath7 .
the value of the temperature can be extracted from the observed transverse momentum spectra whose exponential slope is governed by a combination of thermal motion and collective expansion of the fireball @xcite .
the separation of these two contributions , _
i.e. _ the isolation of @xmath9 , is not straightforward , but has been studied extensively in @xcite .
given @xmath9 , the number of independent variables is further reduced by studying the @xmath33@xmath34 asymmetry in the fireball and relating it to the asymmetry potential @xmath53 , and by the condition of strangeness conservation .
this leaves us with two independent variables , say @xmath13 and @xmath7 , which have to be determined from measured particle ratios .
the various parameters described above enter the thermodynamic description through the partition function from which all observables are derived .
for the hadronic gas state we follow here the conventional hadronic gas model first developed by hagedorn @xcite in which the equation of state of any hadronic system is obtained from a partition function which sums over all hadronic resonances .
it is believed that by including a complete set of intermediate scattering states in the form of ( usually zero - width ) hadronic resonances the dominant interactions within the hadronic system are automatically taken into account .
however , an important but tacit approximation is made in this approach , concerning the parameters such as mass ( and sometimes widths ) of the resonances which are inserted with their free space values .
this neglects the possibility that in particular at high temperature and particle or baryon density , there may occur significant shifts from the free space values
one often speaks in this context of the melting of hadronic masses as one approaches extreme conditions .
while we do our computations entirely within the conservative framework of the conventional model , our general physics discussion will be shaped in such a way that our conclusions should not critically depend on the exact validity of the model used .
the standard hadronic gas model is certainly not sufficiently precise for our purpose if we were to ask questions concerning the absolute values of energy , entropy , baryon density , etc . which depend either on the ever increasing mass spectrum of particles or on the proper volume occupied by the particles @xcite .
however , the condition of zero strangeness and quantities such as the entropy per baryon , which involve ratios of extensive variables , are independent of the absolute normalization of the volume and of the renormalization introduced by the diverging particle spectrum and hence can safely be considered in our approach .
we now turn to the discussion of the constraints which limit the number of the free thermal parameters in the fireball . in our calculations
we will distinguish between the @xmath33 and @xmath34 quarks by introducing separate fugacities , @xmath54 and @xmath55 , instead of @xmath56 in the partition function .
since even in the heaviest nuclei there is only a small @xmath33@xmath34 asymmetry due to the neutron excess , we find it convenient to work with the variables : @xmath57 here , @xmath58 is the `` light quark '' chemical potential ( @xmath59 in terms of the baryochemical potential ) and @xmath53 describes the asymmetry in the number of up and down quarks due to the neutron excess in the heavy ion collision .
the ratio of the net number of down and up quarks in the fireball is @xmath60 in a central s
w collisions , where a tube with the transverse area of the s projectile is swept out from the w target and participates in the fireball @xmath61 ; in pb
pb collisions @xmath62 .
the value of @xmath53 is at each fixed @xmath9 determined by the value of @xmath63 , but depends on the assumed structure of the source , _ i.e. _ the eos .
we have investigated this relation for the two cases considered here : the conventional hg and the qgp .
@xmath64 in the hg phase we include in the partition function all non - strange mesons up to a mass of 1690 mev , all nucleons up to 1675 mev and all @xmath65 s up to 1900 mev .
[ we note that higher resonances would matter only if their number were divergent ( as is the case in the bootstrap approach of hagedorn @xcite ) and if the hg was sufficiently long lived to populate all high mass resonances ] .
in addition strange hadrons are included as described in eq.([4a ] ) below , which we expand in calculations to allow to distinguish between @xmath33 and @xmath34 quarks within the various strange hadron multiplets . in fig.[f1 ]
we show how in the conventional hadron gas the difference @xmath66 induced by the isospin asymmetry of the fireball depends on the temperature , for two selected values @xmath67 and @xmath68 ( appropriate for s w and pb pb collisions , respectively ) , and for three fixed values of @xmath69 and 1.05 .
@xmath64 @xmath64 the curves shown correspond to non - trivial solutions ( _ i.e. _ solutions with @xmath70 ) of the strangeness balance equation with the given values for @xmath71 ; as we will explain below , these solutions cease to exist ( since we require @xmath72 ) above a certain critical temperature , which is given by @xmath73 mev for @xmath74 and smaller ( larger ) for smaller ( larger ) values of @xmath71 .
we see that , while for @xmath75 the ratio @xmath66 vanishes in the hadron gas , it is of the same order as the deviation from unity of @xmath63 once @xmath9 reaches values around 200 mev which are of interest here . for the qgp
the ratio @xmath66 is independent of @xmath71 , due to the decoupling of the strange and non - strange chemical potentials in the partition function . for @xmath76
one finds @xcite @xmath77 such that @xmath66 is exactly equal to @xmath78 .
it is remarkable that towards the maximum temperature consistent with strangeness conservation the hg and qgp based results for @xmath53 at a given @xmath63 nearly agree .
we thus find that , although small , the difference between the chemical potentials of @xmath33 and @xmath34 quarks is not always negligible .
we note for later applications that in the region of interest to us here ( @xmath79 mev ) we have irrespective of the nature of the fireball : @xmath80 in the boltzmann approximation , it is easy to write down the partition function for the strange particle fraction of the hadronic gas , @xmath81 , and we follow the notation of ref.@xcite .
however , given the recently measured values of the thermal parameters ( which allow in particular for a rather high temperature @xmath79 mev ) it is necessary to sum over some more strange hadronic particles than was done in eq.([eq4 ] ) of ref .
@xcite . including the possibility of only partially saturated strange phase space through the factor @xmath7 above , but neglecting for the moment the isospin asymmetry @xmath82
, we have : @xmath83 \label{4a } \end{aligned}\ ] ] where the kaon ( @xmath84 ) , hyperon ( @xmath85 ) , cascade ( @xmath86 ) and omega ( @xmath87 ) degrees of freedom in the hadronic gas are included successively .
the phase space factors @xmath88 of the various strange particles are : @xmath89 where the @xmath90 are the spin - isospin degeneracy factors , @xmath91 , and @xmath92 is the modified bessel function . for the quark - gluon plasma ,
on the other hand , the strange contribution to the partition function is much simpler , because the strange quarks and anti - quarks are isolated and do not occur in clusters with non - strange quarks as in the hadron gas : @xmath93\ , .
\label{lnzs}\ ] ] with the strange quark spin - color degeneracy factor @xmath94 for the value of the strange quark mass we take @xmath95 150180 mev ; in eq.([lnzs ] ) @xmath7 and @xmath71 do not factorize as has been the case for the hadron gas , because @xmath9 is of order @xmath96 and the boltzmann approximation is not always satisfactory and can produce errors of the order of 20% . recall that in general the strange particle density in a qgp is larger by a factor 25 than in a hg , due to the lower threshold ( @xmath97 ) and the presence of the color degeneracy factor in eq.([gs ] ) .
we will now discuss the reduction of the number of free parameters caused by the condition of strangeness balance .
since the net strangeness of the colliding nuclei in the initial state is zero , and strangeness is conserved by strong interactions , the net strangeness of the collision fireball will stay close to zero throughout the collision , except for corrections due to possible strangeness asymmetry in pre - equilibrium and asymmetric surface radiation of strange hadrons @xcite .
if these processes can be neglected , the resulting condition of vanishing total strangeness takes the form @xmath98 where @xmath99 is the net strangeness density , @xmath100 with @xmath101 being the strangeness of particle species @xmath51 having density @xmath102 , and the sum going over all particle species @xmath51 existing in the respective phase ( qgp or hg ) .
while for the qgp this relation leads always to the result @xmath103 or @xmath104 ( as one easily verifies by inserting eq.([lnzs ] ) ) , for the hadron gas it is an implicit equation relating @xmath71 to @xmath56 in a way which depends on the temperature @xmath9 @xcite . in order to also be able to account for pre - equilibrium and surface emission fluctuations ,
we introduce the net strangeness fraction @xmath105 and will also consider solutions for the analogue of eq.([eq6 ] ) , where the left hand side is given by @xmath106 . in this work
we restrict ourselves to asymmetries @xmath107 which might be accessible by purely statistical fluctuations caused by surface emission , without strangeness distillation " during qgp hadronization @xcite which arises when chemical equilibrium is enforced during a slow qgp hadronization process .
an analysis of the phase diagram and the strangeness balance conditions for systems with very large net strangeness is presented elsewhere @xcite .
we note briefly the practical consequences of @xmath108 : for the fireball created in central s
w collisions one has about 108 baryons in the interaction tube .
if we consider a hadron gas with the thermal parameters ( without flow component ) as determined for these s w collision data @xcite , we find a strange pair abundance of about 0.4 per baryon or about 40 strange pairs in total .
the maximum asymmetry considered here in the hg phase at @xmath3 mev is thus about @xmath109 strange quarks .
fluctuations of this magnitude are expected to occur in about half of the collision systems .
since the vanishing of @xmath6 for all values of @xmath9 and @xmath13 in a strangeness neutral qgp appears to be such a unique feature of this state , it is interesting to ask to what extent this condition could arise in a hg .
we will therefore first study the mechanisms which lead in a hg model to a small strange quark chemical potential in a restricted region of temperatures . of course , for a baryon - free system ( @xmath8 ) even in the hadron gas @xmath110
is a natural consequence of strangeness neutrality . for systems with non - vanishing baryon number ( like the central rapidity fireballs created at ags and sps energies ) this is no longer true .
we therefore ask ourselves under which conditions @xmath110 , _ viz .
_ @xmath103 , allows for non - zero values of @xmath111 to be consistent with strangeness neutrality .
this question can be answered analytically : inserting the partition function ( [ 4a ] ) into the strangeness neutrality condition ( [ eq6 ] ) , we see that at @xmath103 the phase space for @xmath87 and @xmath112 cancels out , giving the exact result @xmath113 there is a real solution only when the argument on the right hand side is larger than unity .
it turns out that this condition is a sensitive function of the temperature @xmath9 and of the hadronic resonances included ( and therefore of the masses of the resonances ) .
in particular , for any given resonance mass spectrum used to compute the phase space factors @xmath88 , there is a temperature @xmath114 beyond which no such solution is possible this occurs since @xmath115 is a monotonically decreasing function of @xmath9 ( see fig.1 in ref.@xcite ) .
the case of exact strangeness neutrality ( @xmath116 ) in a hg in absolute chemical equilibrium ( @xmath117 ) was considered in @xcite , and for this choice of parameters our following theoretical analysis essentially agrees with that work . an important aspect to consider is the value of @xmath6 observed and how it is compatible with the picture of the reaction .
the reason that @xmath6 is of such importance is the expected value @xmath104 associated with qgp .
we first will address the question under what circumstances this value is accessible to a hg fireball , and then turn to discuss more generally how different dynamical scenarios can be distinguished .
we expect that there is a domain of temperatures for which even a considerable change of @xmath13 between @xmath118 and @xmath111 does not induce a significant change of @xmath6 . for the extensive set of resonances included here
this quite peculiar behavior occurs at @xmath119 mev for @xmath116 and at @xmath120 mev for @xmath121 . in figs.[f2]a , [ f2]b , [ f2]c we present the constraint between @xmath6 and @xmath13 at fixed @xmath122 mev ( solid curves ) , 150 mev ( long - dashed curves ) , 300 mev ( short - dashed curves ) and 1000 ( @xmath123 ) mev ( dotted curves ) for @xmath116 ( fig.[f2]a ) , @xmath124 ( fig.[f2]b ) , and @xmath125 ( fig.[f2]c ) , all for @xmath126 .
the choice @xmath117 would not change our results significantly .
the flat behavior of @xmath6 as a function of @xmath13 for @xmath116 and @xmath3 mev makes it hard to distinguish the hadron gas from a qgp ( which has @xmath127 independent of @xmath13 ) in this temperature range .
this was pointed out in ref.@xcite , but we think that eq.([mub ] ) and the corresponding fig.[f2]a , with our definition of @xmath6 ( see footnote [ foot1 ] ) , exhibits this effect more clearly and allows us to understand the accidental nature of this behavior . for @xmath128 ( fig.[f2]a ) all curves pass through @xmath110 at @xmath8 , and as @xmath129 one approaches a limiting curve which is rather well represented by the dotted curve corresponding to 1000 mev . for @xmath130
we note that in the qgp phase the strange quark chemical potential does not vanish ( see horizontal lines in figures[f2]b , [ f2]c ) , and that there is a common crossing region of the hg and qgp results at a finite value of @xmath6 , close to @xmath8 . for @xmath131 for a fixed positive ( negative ) net strangeness fraction @xmath16 , @xmath6 continues to increase ( decrease ) at fixed @xmath13 as a function of @xmath9 , and only the slope as a function of @xmath13 approaches a definite limit as @xmath132 .
@xmath64 + @xmath64
in this section we describe , using as an example the experiment with the most detailed information of strange baryon and anti - baryon production , wa85 @xcite , how to extract estimates of the relevant thermodynamic parameters from particle production data . this procedure is straightforward if one can focus on the ground state baryons of a given flavor content , _
i.e. _ if the production and decay of heavier resonances can be neglected .
such an analysis was presented in @xcite , and its essential steps will be shortly repeated in subsections 3.1 through 3.3 .
if @xmath9 is indeed in the region of 210 mev as the @xmath1-slopes seem to indicate , resonance production and decays can , however , not be neglected , and therefore such a procedure is not fully consistent . in subsection
[ 3.4 ] we therefore redo this analysis including resonance decays , which is then much more involved .
the main result is that resonance decays mainly affect the extraction of the strangeness saturation factor @xmath7 ( inclusion of resonance decays drives the extracted @xmath7 closer to 1 , _
i.e. _ towards full saturation of hadronic gas phase space ) , while the striking conclusion that the data point towards a vanishing strange chemical potential @xmath133 is remarkably stable against these corrections
. it was argued in @xcite that within the thermal model it is quite easy to extract the fugacities from a given particle ratio _ at fixed @xmath1 and @xmath134 _ , thus eliminating the need to extrapolate the experimental data to full phase space .
specifically , for a stationary fireball one finds for the ratio of particles @xmath51 and @xmath135 , for each of which the invariant cross section is experimentally known at a certain rapidity @xmath136 in the region @xmath137 : @xmath138 here we used the identity @xmath139 for the particle energy ( with @xmath134 measured relative to the center of mass of the collision fireball ) to rewrite the transverse mass integral over the boltzmann distribution @xmath140 ( the factor @xmath141 results from starting with the lorentz - invariant cross section @xmath142 ) .
if the fireball is not stationary , but expands in the longitudinal direction in a boost - invariant way , @xmath143 is independent of @xmath134 , and the factors @xmath140 under the integrals on the r. h. s. of eq .
( [ ninj ] ) have to be replaced by a bessel function @xcite : @xmath144 where we neglected a common normalization constant containing the fireball volume which drops out from the ratio ) is integrated over rapidity , _
i.e. _ if one wants to describe with a stationary fireball data obtained over a large rapidity window . ] .
the important observation is that obviously in both cases the boltzmann factors ( and with them all dependence on the particle rest masses ) drop out from the ratio , due to the common lower cutoff on @xmath31 ( which , of course , must be larger than either rest mass ) .
the same remains true if @xmath31 is not integrated over , but the ratio is evaluated at fixed @xmath31 . since the degeneracy factors @xmath90 are known , this immediately yields the ratio of the generalized fugacities @xmath145 ( see eq.([gammai ] ) ) .
thus , comparing the _ spectra _ of particles with those of their anti - particles within overlapping regions of @xmath1 , the boltzmann and _ all _ other statistical factors cancel from their relative normalization , which is thus a function of the fugacities only .
[ in subsection[3.4 ] we show that this statement , although no longer strictly true , still remains approximately valid when feed - down by resonance decays is included . ] for the currently available two such ratios one has specifically @xmath146 these ratios can easily be related to each other , in a way which shows explicitly the respective isospin asymmetry factors and strangeness fugacity dependence .
eqs.([ratiol ] , [ ratio ] ) imply : @xmath147 eqs.([r1 ] , [ r2 ] ) are generally valid , irrespective of the state of the system ( hg or qgp ) , as long as the momentum spectra of the radiated particles are thermal with a common temperature .
we see that once the left hand side is known experimentally , it determines rather accurately the values of @xmath148 which enter on the right hand side with a dominating factor 6 , while the ( small ) isospin asymmetry @xmath53 plays only a minor role .
this explains how , by applying these identities to the wa85 data @xcite , it has been possible to determine the chemical potentials with considerable precision in spite of the still pretty large experimental errors on the measured values of @xmath149 , @xmath150 .
turning to the pertinent experimental results we first recall the recently released data on the rapidity dependence of @xmath151 and @xmath152 production rates measured by the experiment na36 @xcite .
there is a clear indication that the production of these particles in nuclear s
pb collisions is concentrated in the central rapidity domain it does not follow in its behavior the fritiof simulations which , on the other hand , are consistent with the production rates from p pb collisions .
this supports the notion of a central fireball reaction picture for s w collisions studied by wa85 @xcite .
the most striking result of the latter experiment is the fact that the abundance of charged anti - cascades ( @xmath153 ) is unusually enhanced , in particular when compared to the abundance of cascades ( @xmath154 ) and anti - lambdas ( @xmath155 ) .
we now interpret the data in terms of the local equilibrium model , thus fixing the values of @xmath6 and @xmath58 .
the @xmath156 ratio has been reported as : @xmath157 note that , in p
w reactions in the same @xmath158 region , a smaller value for the @xmath150 ratio , namely @xmath159 , is found .
the @xmath160 ratio is : @xmath161 in eq.([lambda ] ) , corrections were applied to eliminate hyperons originating from cascade decays , but not those originating from decays of @xmath162 or @xmath163 which are of little significance for the high @xmath31 considered here .
the ratio @xmath149 for s w collisions is slightly smaller than for p w collisions in the same kinematic range . from these two results , together with eqs.([r1 ] , [ r2 ] , [ deltamu ] ) and the value @xmath164 mentioned after eq .
( [ eq4 ] ) , we obtain the following values of the chemical potentials for s w central collisions at 200 gev a : @xmath165 a study of the temperature observed in these collisions
was recently made @xcite . as noted in @xcite and discussed below in section 4
, the possibility of a hg interpretation of these data only poses itself if the transverse mass slope of the produced particles is entirely due to the thermal motion . under this assumption
the temperature is @xmath166 mev , and therefore we have : @xmath167 the last result translates into the value @xmath168 for the strange quark fugacity .
much of what we do later in this paper will rest on this stunning result that the strange chemical potential is very small and perfectly compatible with zero . in our calculations below we
will thus consider in particular a hg constrained to this range of values ( as well as to near strangeness neutrality ) .
a complete cancellation of the fugacity factors occurs when we form the product of the abundances of baryons and anti - baryons .
we further can take advantage of the cancellation of the boltzmann factors by comparing this product for two different particle kinds , _
e.g. _ we consider : @xmath169 if the phase space of strangeness , like that of the light flavors , were fully saturated , the fireball model would imply @xmath170 .
however , any deviation from absolute chemical equilibrium as expressed by the factor @xmath7 will change the value of @xmath171 . ignoring as before any feed - down effects from higher resonances we find @xmath172 in principle
the measurement of @xmath173 can be done with other particle ratios ; in the absence of resonance feed - down we have @xmath174 where in the last relation the factors 2 in the denominator correct for the spin-3/2 nature of the @xmath87 . as we will see in subsection 3.5 , eq .
( [ gam4 ] ) , these double ratios are considerably affected by resonance decays .
we note that in the kinematic domain of eqs.([cascade ] , [ lambda ] ) the experimental results reported by the wa85 collaboration are : @xmath175 if the mass difference between @xmath152 and @xmath176 is neglected , this implies in the framework of the thermal model that an equal number of @xmath152 s and @xmath176 s are produced , such that @xmath177 the fact that the more massive and stranger anti - cascade exceeds at fixed @xmath1 the abundance of the anti - lambda is most striking .
these results are inexplicable in terms of simple cascade models for the heavy - ion collision @xcite .
the relative yield of @xmath178 appears to be 5 times greater than seen in the @xmath179@xmath179 isr experiment @xcite and all other values reported in the literature @xcite . combining the experimental result eq.([new1 ] ) with eqs .
( [ gam1 ] , [ gam2 ] ) , we find the value @xmath180 the data from the wa85 collaboration @xcite correspond to an apparent temperature ( from the slope of the @xmath31-spectra ) of @xmath181 mev @xcite . in our numerical studies we therefore concentrated on two extreme cases : * @xmath2 mev , @xmath182 ( no transverse flow ) , which we call the thermal " scenario , * @xmath183 mev , @xmath184 , which we call the flow " scenario for which the freeze - out temperature has been estimated following the results of @xcite .
if the slope of the experimental spectra is not entirely due to thermal motion , but contains a flow component due to collective expansion of the emitting source @xcite , then flow spectra " ( _ i.e. _ boosted boltzmann spectra ) have to be used in the numerator and denominator of eq .
( [ ninj ] ) .
if the flow is azimuthally symmetric and boost - invariant in the longitudinal direction , the thermal spectrum ( [ boost ] ) has to be replaced by ( again neglecting a common normalization constant ) @xmath185 where @xmath186 and @xmath187 are the modified bessel functions and @xmath188 is the transverse flow rapidity .
this spectrum has an asymptotic slope corresponding to an apparent blue - shifted " temperature @xmath189 the same result as in eq.([flow ] ) is obtained if the spectrum from a spherically expanding fireball is integrated over rapidity .
since some longitudinal flow is likely to be present in s
w collisions , and also the kinematic window of the wa85 experiment is nearly one unit of rapidity , we think that the form ( [ flow ] ) , in which integration over a large rapidity window is implied , is most appropriate for an analysis of the experimental spectra .
( the large window " combines the one unit of @xmath134-acceptance in wa85 with about 1.5 units of flow expected @xcite for this system . ) obviously ,
through the separate dependence on @xmath31 and @xmath32 eq . ( [ flow ] ) depends on the particle rest mass explicitly .
this spoils the cancellation of the spectral shape from the @xmath31-integrated particle ratios as well as their independence from the @xmath31-cut , and thus also renders the extraction of the fugacities , etc .
more complicated .
in order to assess the importance of these effects we devote next subsection to a more complex reanalysis of the experimentally measured strange baryon and anti - baryon ratios which takes all these complications into account . anticipating this development
we show in fig.[f5new ] the @xmath31-spectra of the various particle species , for both the thermal " scenario ( figs.[f5new]a , b ) and the flow " scenario ( figs.[f5new]c , d ) . dashed lines indicate the direct thermal contribution to the spectra , while the solid lines include all resonance decay contributions ( see following subsection ) . these results allow to directly assess the importance of resonance decays on the @xmath31-slopes of the observed particles and on the particle ratios . for the @xmath87
there are no resonance contributions , due to the imposed mass cutoff in the resonance spectrum .
while the overall normalization of the spectra is arbitrary , their relative normalization is according to the fugacities given in table 1 .
the presence of flow affects the spectra mostly at low @xmath31 near the mass threshold @xcite , where it leads to a flattening , _
i.e. _ to a higher apparent temperature .
the effect is proportional to the particle mass ( since for small @xmath32 the effect of flow can be estimated by @xmath190 ) and thus most clearly seen in the @xmath191 and @xmath87 spectra .
as already mentioned , eq.([ninj ] ) is only correct if feed - down to the observed particles by strong decays of higher resonances is neglected . at temperatures of order 200 mev
these resonance decays can no longer be totally ignored .
this has two annoying consequences : first , since the decay products have a steeper @xmath31-spectrum than the original resonances @xcite , the various contributions to numerator and denominator in ( [ ninj ] ) have different slopes , and the integrals over the boltzmann factors no longer cancel exactly .
second , since different sets of resonances contribute to @xmath51 and @xmath135 , the ratio is modified by an a priori unknown factor which turns out to be hard to calculate and in general _ does _ depend on the experimental @xmath31-cutoff as well as on the temperature .
since the decay spectra drop more steeply as a function of @xmath31 than the thermal ones , the role of resonance feed - down can in principle be suppressed by going to sufficiently large @xmath31 . for pions and kaons , for example
, the thermal contribution dominates for @xmath192 11.5 gev , in spite of the huge contribution of decay pions to the total yield @xcite ; the latter all end up at lower @xmath31 . for baryons due to their larger masses ,
the difference in slope between the thermal and decay baryons is less drastic , and one has to go to still larger @xmath31 before the resonance decay contributions are sufficiently suppressed . including resonance decays , the numerator and denominator of eq .
( [ ninj ] ) take the form @xmath193 with the direct thermal or flow contribution given by ( [ boost ] ) or ( [ flow ] ) , respectively , and the contribution from decays of resonances @xmath194 ( with branching ratio @xmath195 into the observed channel @xmath51 ) calculated according to @xcite @xmath196 ^ 2 } } } \left ( { { dn_r}\over{dy \,dm_\perp^2 } } \right)\ , .
\end{aligned}\ ] ] capital letters indicate variables associated with the resonance @xmath197 , @xmath198 is the invariant mass of the unobserved decay products @xmath199 , and the kinematic limits are given by @xmath200 in ( [ eq : reso ] ) @xmath201 is the decay phase space for the @xmath202-body decay @xmath203 , and for the dominant 2-body decay ( assuming isotropic decay in the resonance rest frame ) it is given by @xmath204 for @xmath205 see ref .
@xcite . the resonance spectrum @xmath206 entering the r. h. s. of eq .
( [ eq : reso ] ) is in general itself a sum of a thermal or flow spectrum like ( [ boost ] ) or ( [ flow ] ) and of decay spectra from still higher lying resonances ; for example , most decays of high - lying nucleon resonances into protons or anti - protons proceed through the @xmath207 or its anti - particle ( _ e.g. _ @xmath208 ) .
this has been taken into account in our calculations . extracting the degeneracy factors and fugacities of the decaying resonances
, we write shortly @xmath209 with the sum now including only resonances with identical quantum numbers ; if these quantum numbers agree with those of species @xmath51 , the sum is meant to also include the thermal contribution . between particles and anti - particles we have the relation @xmath210 the three independent particle ratios measured by wa85 @xcite are thus given by @xmath211 @xmath212 contains also ( in fact as its most important contribution ) the electromagnetic decay @xmath213 , and @xmath214 includes the weak decay @xmath215 ( since @xmath87 s have not been reconstructed so far by wa85 ) . if the right hand sides are evaluated using eqs .
( [ reso ] , [ eq : reso ] , [ ntilde ] ) with @xmath216 gev , for the left hand sides the experimental values ( [ cascade ] , [ lambda ] , [ newa ] ) can be inserted , and eqs .
( [ rxi ] , [ rlam ] , [ rs ] ) can be solved for @xmath217 , @xmath71 , and @xmath56 .
note that we do not distinguish here between @xmath54 and @xmath218 since this would require the calculation of a separate decay cascade for each member of an isospin multiplet .
the smallness of @xmath53 in ( [ dmu ] ) justifies this approximation .
the results of our analysis are given in tables 1 and 2 and figures [ f5new ] and [ f6new ] .
comparing the results for the thermal parameters in table 1 with eqs .
( [ muq ] , [ mus ] , [ gams ] ) which contained only an estimate of the radiative @xmath176 decay but no contributions from higher lying resonances , we can draw the following conclusions : * the extraction of @xmath56 and @xmath71 according to eqs .
( [ ratiol ] , [ ratio ] ) is only very weakly affected by resonance decays and by the origin of the slope of the @xmath31-spectrum ( thermal or flow ) .
( the absolute value of the associated chemical potentials does , of course , depend on the choice of the freeze - out temperature . )
the reason is that in eqs .
( [ rxi ] ) and ( [ rlam ] ) in each case the first term in the sums occurring in the numerator and denominator completely dominates .
( for @xmath191 s the effects from @xmath87-decays are small , and for @xmath152 s the strong @xmath176 contribution overwhelms the effect from @xmath219 decays . ) taking the results for the thermal " scenario from table 1 and evaluating with them the @xmath220 , @xmath221 , and @xmath222 ratios excluding the contributions from higher resonances changes them to 0.383 , 0.121 , and 0.262 , respectively ( instead of the input values 0.39 , 0.13 , and 0.20 , respectively ) . leaving out
also the @xmath223 contribution changes only the @xmath222 ratio which now becomes 0.430 .
obviously this ratio is the one which is most strongly affected by the resonance decays .
* since the @xmath222 ratio is the crucial ingredient for the determination of the strangeness saturation factor @xmath224 , the effects from resonance decays and flow can be clearly seen in this number . from eq .
( [ new1 ] ) we had extracted @xmath126 ; improving the rough factor 2 estimate for the @xmath176 contribution by its correct thermal weight at @xmath2 mev would have lowered @xmath7 to 0.57 .
the decay of higher - lying resonances raises the value of @xmath7 up to 0.76 at @xmath2 mev .
if due the presence of flow the @xmath31-slope does not give the true + @xmath64 + + @xmath64 + + @xmath64 + + + + + temperature , the estimate for @xmath7 goes up even higher : for @xmath183 mev , @xmath184 ( _ i.e. _ values which are consistent with the kinetic freeze - out criterion developed in @xcite ) we find @xmath225 , _ c.f .
_ table 1 , corresponding to nearly complete saturation of the strange phase space ! * the conclusion that the wa85 data on strange baryon and anti - baryon production in 200 geva s w collisions seem to indicate a vanishing strange quark chemical potential is extremely stable with respect to resonance decay contributions and the possibility of a flow component in the spectra .
it will serve as the cornerstone of the following studies in section4 .
@xmath64 it is clear that the extraction of three thermodynamic parameters ( @xmath56 , @xmath71 , and @xmath7 ) from three independent particle ratios can not be considered a very convincing test of the thermal picture , in particular since the extracted values depend on the choice of the temperature whose independent derivation from the shape of the @xmath1-spectrum is not quite unique ( unless one accepts the theoretical consistency arguments of ref .
@xcite which prefer the lower temperature of around 150 mev ) .
it is therefore mandatory that the picture receives further experimental tests in the form of other independent particle ratios . in table 2 and figure [ f6new ]
we have therefore included predictions for several additional particle ratios , in particular involving @xmath226 and @xmath227 which are expected to become soon available from the wa94 collaboration @xcite in a kinematic region compatible with the wa85 results .
our calculations explore the consequences of both the thermal " and flow " scenarios above , and anticipate also a possible improvement of the experimental acceptance towards lower transverse momenta , by calculating in table 2 some of the ratios also with an @xmath1 cut of only 1.2 gev , or plotting them ( in fig.[f6new ] ) as a function of this cut .
comparing @xmath1-integrated particle ratios with different @xmath1 cuts gives an idea of the effects coming from the somewhat different slopes of thermally emitted particles and resonance decay products . from fig.[f6new ] we see that these effects are generally small for ratios involving only baryons : due to their heavy mass , the slope difference of parent and daughter particles in resonance decays is small @xcite , resulting in a slight steepening of the spectra at low and intermediate @xmath1 .
for kaons , on the other hand , the decay contributions are considerably steeper than the thermal one , leading to an appreciable upward curvature of the spectrum at low @xmath1 , while flattening out to the asymptotic apparent temperature for @xmath228 11.5 gev , where the thermal component dominates . in consequence , as seen in fig.[f6new ] , the @xmath229 and @xmath230 ratios are proportionally more sensitive to the @xmath1 cut
. it should be kept in mind that the large @xmath231 and ( in particular ) @xmath232 ratios are partly due to the additional spin degeneracy of the spin-3/2 @xmath87 baryon .
the apparent strong @xmath233-dependence of all ratios involving @xmath87 or @xmath112 baryons ( especially in the flow scenario ) should be taken with some caution : it results mostly from the absence of resonance decay contributions to the @xmath87 spectrum in our calculation and might be modified if ( so far unknown ) @xmath234 resonances existed and should be included .
when comparing different ratios involving @xmath152 s ( _ e.g. _ in order to estimate the systematic of these ratios with increasing strangeness ) , one should always keep in mind the considerable @xmath176 contribution which , at @xmath2 mev , effectively changes the degeneracy factor associated with the @xmath152 by about a factor 1.6 . once this is corrected for , _
i.e. _ the @xmath176 contribution is accounted for , one finds the inequalities @xmath235 paradoxically , eq.([gam4 ] ) suggests an increasing importance of resonance decay corrections with increasing strangeness content , as compared to the estimate , eq.([gam3 ] ) .
this is due to the fact that the stranger baryons are also heavier and thus receive less resonance contributions ( because the corresponding excited states are strongly suppressed by the mass term in the boltzmann factor ) than the lighter ones ( whose resonances are easier to excite ) . in part
this tendency may be somewhat exaggerated by our cut in the resonance spectrum at about 2 gev ; taking into account even higher resonances ( although it is not clear whether can indeed be excited with statistical probabilities in nuclear collisions ) might modify in particular the @xmath26 ratios .
we regard the ratios , the assumption of chemical equilibrium may break down for kaons if they are created by hadronization of a qgp . ] predicted in table 2 and fig.[f6new ] as a qualitative test of the idea that the emitter is a fireball which is in thermodynamic equilibrium with the given thermodynamic parameters .
however , even if this test is eventually passed , the fact that we can successfully describe all observed particle ratios in this way does not yet tell us immediately the nature of the emitting source ( hg or qgp ) .
that information resides in supplementary information concerning the global properties of the source and their _ relationship _ to the extracted thermodynamic parameters , as well as in the systematic behavior of these parameters under varying experimental conditions ( collision energy , projectile and target size , collision centrality etc . ) .
these important issues will be discussed in more detail in the following sections [ sect4 ] and [ sect5 ] . since thermal particle ratios are functions of only @xmath13 and @xmath6 , the accidental vanishing of @xmath6 at a temperature @xmath236 200220 mev due to the constraint to strangeness neutrality
removes the opportunity to simply distinguish the two phases a hg phase of strongly interacting matter , at this level of discussion , is _
indistinguishable _ from qgp as long as a measurement of the relative multiplicities of strange particles is considered at one , and only one collision energy . even in this case
many theoretical questions arise , such as why is @xmath110 or why is @xmath7 near to unity ? however , as figs.[f2]a , [ f2]b , [ f2]c clearly show , this will not remain a problem at higher or lower temperature . to make this point more quantitative , we now study strange baryon ratios arising at different temperatures from thermal fireballs _ constrained _ to a ( nearly ) vanishing strangeness . in fig.[f3 ]
we show , for the case of exactly vanishing strangeness and for various temperatures , the resulting relation between @xmath150 and with @xmath149 .
in addition to the hg results for temperatures @xmath122 mev ( solid line ) , @xmath183 mev ( dashed line ) and @xmath237 mev ( dotted line ) we show the case @xmath110 corresponding to a qgp source hadronizing rapidly by quark recombination ( dashed - dotted line ) .
the cross corresponds to the result reported by the wa85 experiment @xcite . as can be seen , the @xmath110 curve nearly coincides with the @xmath2 mev curve , as noted before @xcite , but fireballs decoupling at significantly lower or higher temperatures ( as expected at lower and higher collision energies , respectively ) would lead to significantly different particle ratios if the thermal model is correct .
[ f3 ] [ f4 ] in addition to the baryon ratios we can also consider the ratio of kaons to hyperons , again at fixed @xmath1 . because of the experimental procedures used , which rely on the observation of the disintegration of neutral strange particles into two charged decay products , a comparison of the @xmath238 with the @xmath152 ( which includes the @xmath152 s from the electromagnetic decay of @xmath176 s ) is the most useful next step .
we therefore introduce : @xmath239 where the second identity is again , of course , only valid if resonance decay contributions can be neglected .
for the rather light kaons there exist , however , many different possibilities for secondary production through resonance decays , which limits somewhat the practical usefulness of this ratio . for example , if we use the thermodynamic parameters extracted in subsection [ 3.4 ] from the wa85 data , we obtain for the thermal " interpretation of the @xmath1-slope ( @xmath2 mev , no flow , @xmath240 gev ) a value @xmath241 , if all resonance contributions are included , while the ground state contributions alone according to eq .
( [ rk ] ) would give @xmath242 ( @xmath243 if the @xmath152@xmath176 mass difference is taken into account ) .
since the size of these corrections depends on the fireball temperature and the @xmath1-cut , a full quantitative study does presently not appear practical .
given these uncertainties , and in order to demonstrate the insensitivity of the @xmath229-ratio to the nature of the fireball , we show in fig.[f4 ] ( as a matter of principle ) the uncorrected ratio @xmath244 as a function of the uncorrected ratio @xmath149 : as the figure clearly shows , this observable is rather insensitive to both the temperature of the source and to its intrinsic structure ( @xmath110 or @xmath245 ) .
the lines are as in fig.[f3 ] , and we note again the near coincidence of the @xmath122 mev hg result with the qgp result . but even as @xmath9 is changed to 150 or 300 mev , the variation of @xmath149 at fixed @xmath246 stays well within the present experimental error bars and will thus not permit to identify the properties of the source .
we now return to discuss in detail a very important observation which is related to the particle multiplicity and charge flow associated with strange baryon production @xcite . even though no data are yet available on multiplicity densities in coincidence with strange particle production , the properties of the hg and qgp fireballs with regard to their entropy content are so drastically different that we can draw at least qualitative conclusions by comparing multiplicity data from s pb collisions with s w strangeness production , despite the fact that the two targets ( @xmath247pb and @xmath248w ) differ slightly in mass .
qgp and hg states are easily distinguishable in the regime of values @xmath249 of interest here .
the entropy per baryon in the hg is @xmath250 .
consequently , the pion multiplicity which can be expected from such a hg fireball is at best @xmath251 .
this is less than half of the qgp based expectation , which we can easily estimate by considering the perturbative qgp equation of state : up to corrections of order @xmath252 , the leading term of the specific entropy from light quarks and gluons is : @xmath253 \nonumber \\ & & \hspace{-0.6cm}= 17\ , ( t/\mu_q)\quad \mbox { for } \alpha_s=0.6\quad .
\end{aligned}\ ] ] adding to this the considerable entropy content of the strange quarks ( neglecting for technical reasons the @xmath254-corrections in this case ) , we find for the qgp a specific entropy of about 45 units at the same value of @xmath255 .
poor knowledge of the value of @xmath254 and the strange quark mass in the qgp phase make this estimate uncertain by about @xmath256% .
still , the difference to the hg result is considerable in terms of experimental sensitivity .
checking the theoretical sensitivity we note that the point at which the entropy of hg and qgp coincide _ and _ strangeness vanishes _ and _ @xmath257 is at @xmath258 mev , @xmath259 mev , quite different from the region of interest here .
we therefore consider now an observable which is capable to distinguish between the hg and qgp via their entropy content .
total charged particle multiplicities ( excluding target / projectile fragments ) _ above _ 600 in the central region have been observed by emu05 @xcite in 200 geva s pb collisions , corresponding possibly to a total particle multiplicity of as large as 1000 .
we believe that such large multiplicities are implied by a qgp scenario for the central fireball , while being hardly compatible with the corresponding hg interpretation .
the data of the emu05 collaboration @xcite were obtained with a high multiplicity `` trigger '' which makes them useful in comparison with the wa85 results which are also triggered on high multiplicity .
some of these charged multiplicity data from emulsions are shown as a function of rapidity in fig.[f5 ] .
we depict the quantity @xmath260 , the difference in the number of positively and negatively charged particles normalized by their sum @xcite : @xmath261 the data points correspond to 15 fully scanned central " events of 200 geva s pb interactions , with the trigger requirement being a total charged multiplicity @xmath262 .
reaction spectators ( target fragments ) are not observed in this experiment . at central rapidity
a value of 0.080.09 is found for @xmath263 .
we note the rise of @xmath263 in the projectile and target pseudorapidity regions , indicating a more rapid decrease of the particle multiplicity ( mostly pions ) than of the positive particle excess ( protons ) .
obviously this feature proves the partial transparency at cern energies of the pb nucleus to the incoming sulphur projectile ; it will be quite interesting to see if this feature persists for heavier projectiles and in particular how the shape is in pb pb collisions .
@xmath64 [ f5 ] [ f6 ] + [ f7 ] @xmath64 in our theoretical work we address primarily the central region , in which the smallest values of @xmath263 indicate the highest multiplicity production . the numerical evaluation of @xmath263 in a hg at @xmath264 , where we fix the temperature for each @xmath13 such that strangeness is conserved , is shown in fig.[f6 ] .
we see that approximately @xmath265 this hg result is extremely simple , considering the complexity of the calculation .
we thus see that in the only permissible hg scenario for the strangeness source which ( according to eq .
( [ chempot ] ) ) _ has to have _
@xmath266 gev , eq.([dq ] ) would predict a value of @xmath267 ; this is incompatible with the emu05 results , see fig.[f5 ] , where the central value is @xmath268 . at this value of @xmath263
the hg phase would have @xmath269 mev , rather than the @xmath270 mev we have extracted from the strange baryon data . despite the fact that both measurements use different targets we believe that we can draw the conclusion that a hg fireball is incompatible with the combined emu05 and wa85 data .
physically the variable @xmath263 represents essentially the inverse of the specific entropy .
the measured very small value of @xmath263 thus implies a very large value for @xmath271 . in fig.[f7 ] we show the specific entropy as a function of @xmath272 for a zero - strangeness hadron gas at @xmath273 .
we see the very strong direct correlation between these two quantities .
we supplement this observation in figs.[f8]a , [ f8]b , and [ f8]c by also showing @xmath271 as a function of @xmath272 at fixed @xmath274 150 , 200 , 300 mev , varying @xmath275 . we observe that , within the range of these parameters , the small observed value of @xmath263 always requires a rather large specific entropy @xmath276
. thus @xmath272 effectively measures the entropy per baryon , independent of the specific thermal conditions of the source .
please note that in this connection it is essential to correctly account for the strange particles . to illustrate this point and to understand the value of @xmath263 we show in fig.[f9 ] the product @xmath277 as a function of @xmath13 for fixed @xmath278 ,
where akin to fig.[f7 ] , the lower set of curves was obtained without strange particles .
we see that strange particles have considerable impact on the result .
both the limiting value and near constancy of the result obtained without strange particles can be easily understood since in this case @xmath277 is effectively the entropy per pion . to see this note that in absence of strange particles we have ( neglecting the small @xmath33@xmath34 asymmetry and assuming pion symmetry @xmath279 ) : @xmath280 where the last term in the denominator involves the sum over all initial charged particles other than pions , in particular heavy mesons , protons and anti - protons , etc . considering the product of @xmath263 with @xmath29 , the baryon content cancels and the result is effectively entropy per pion renormalized ( as it turns out numerically with a factor 2 ) by the degeneracy factors and resonance contributions .
this also explains why the product is rather constant as @xmath13 ( and along with it @xmath9 ) changes : there is a slow increase in the entropy per pion with increasing @xmath13 since the resulting decrease in @xmath9 enhances the relative importance of the heavier baryon resonances with the associated larger entropy per particle .
the question arises whether the qgp fireball model would be in better agreement with the combined wa85 and emu05 data , or if the disagreement of the hg picture with the data is rather an indication for the failure of the fireball model in general . in order to be able to answer this question in detail we need in principle to have a sophisticated model for the hadronization of a qgp .
however , as long the final state is one with thermal momentum distributions , the relationship between @xmath263 and @xmath271 is likely to be similar to that of fig.[f7 ] even if the observed hadrons result from a hadronizing qgp .
the reason is that the product @xmath281 ( in other words , the entropy per pion ) will not deviate drastically from its equilibrium hadron gas value even in a state which is considerably out of chemical equilibrium . noting that there is little entropy production in the qgp hadronization process @xcite we can @xmath64 + @xmath64
thus subsume that after hadronization the result would be a point close to the curve shown in fig.[f7 ] .
we see that the value @xmath282 implies @xmath283 50 , which is indeed of the order required by a hadronizing qgp with the thermal parameters extracted in section 3 .
we note that entropy per baryon is a dimensionless function of the dimensionless quantity @xmath284 and hence different qgp scenarios with the same @xmath284 all have the same value of the entropy per baryon . as the result of this study
we conclude that the charge flow ratio with the charged particle multiplicity is an excellent experimental measure of the entropy per baryon in the source , and that the current results are suggestive of a much more entropy - rich emitter at central rapidity than can be expected in hadronic gas models .
@xmath64 even though the use of thermal model is less convincing for the small collision system s s , we will shortly consider the results obtained by the na35 collaboration @xcite in s
s interactions at 200 geva . in this case
we can base our discussion on the reported charged particle rapidity densities @xcite and the measured @xmath221 ratio @xcite .
s s interactions show at central rapidity visibly less baryon number stopping than s
w collisions ; from the data on @xmath285 and @xmath286 in @xcite we read off a central rapidity value of @xmath287 . the raw @xmath221 ratio ( uncorrected for @xmath86-decays )
is strongly peaked at central rapidity and rises there to a value @xmath288 ( with error bars of order 50% ) @xcite . the observed transverse mass spectra can again be interpreted in terms of a thermal , longitudinally expanding central [ f9 ] @xmath64 fireball with @xmath289 200220 mev @xcite or , including a transverse flow component , by @xmath290 mev combined with a flow velocity @xmath291 @xcite .
following eq .
( [ ratio ] ) , the asymmetric @xmath292 ratio implies a non - zero combination of chemical potentials , @xmath293 .
we note that the experimental ratio decreases when averaged over a larger rapidity interval , implying even larger chemical potentials .
if @xmath294 mev , both the hg and qgp model imply @xmath295 for strangeness neutral systems , and we thus obtain @xmath296 mev . for the flow "
scenario with @xmath183 mev , strangeness conservation in a hg allows the above relation to be satisfied with @xmath297 mev at @xmath298 mev , again with rather large experimental errors . for the thermal " case we can use fig . [ f6 ] to translate @xmath13 into @xmath263 ; with @xmath299
mev we obtain @xmath300 . although the central value of @xmath263 extracted in this way is twice the measured value for this ratio , due to the large experimental uncertainty of the @xmath221 ratio the two numbers still agree within one standard deviation .
we note that a hg with @xmath301 possesses a specific entropy @xmath22 = 35 , while @xmath263 = 0.065 requires @xmath22 = 6575 . for the flow "
scenario , on the other hand , we compute with the conventional hg - eos ( with @xmath302 mev , @xmath303 mev ) a value @xmath304 , consistent with the measured value 0.065 .
as this discussion demonstrates , the current na35 strange particle data are neither precise enough nor do they provide sufficiently complete information to allow for a similar analysis as done for the wa85 data above . in particular the lack of data on @xmath191 and @xmath305 production in s
s collisions prevents us from determining the values of @xmath7 and @xmath6 . in the absence of such information ,
the validity of a hg description for the na35 s s data can at the present moment not be excluded .
so far we have extracted , with the help of thermodynamical concepts , from various sets of nuclear collision data the characteristic statistical properties of the source of the emitted particles . in the following
we will comment on what our findings imply for the possible formation of a quark - gluon plasma , for its evolution and ultimately its dissociation into the final state hadrons .
we have seen that a hg interpretation of the wa85 data on strange baryon production is not consistent with the large multiplicity densities observed by emu05 which indicate a much larger specific entropy than implied by the hg picture . in this context
we have discussed both a purely thermal " interpretation of the data ( which identifies the slope of the transverse mass spectra directly as the fireball temperature ) and a flow scenario " , where part of the observed slope is due to a blueshift factor caused by transverse collective expansion of the fireball .
independent of which version of the hg picture we used , we found from the wa85 data a surprisingly large degree of strangeness saturation @xmath7 , ranging from 75% in the purely
purely thermal " interpretation to 90% in the flow scenario " .
these are hard to understand in a hg scenario , because the time scale for strangeness producing processes is usually believed to be much too slow to achieve that kind of saturation during the lifetime of the collision , unless a qgp is formed . within the hg picture ,
the dynamical aspects connected with the flow seem to be required for consistency with the freeze - out kinetics @xcite .
for this reason the flow scenario " has been highly favored before .
however , we have shown here that within the hg picture the value @xmath110 constrains the liberty to interpret the slope of the transverse mass spectra in terms of a temperature which is blueshifted by flow .
the observed values @xmath110 and @xmath306 are inconsistent with strangeness neutrality of the observed system of hadrons at the rather low value of @xmath307 mev resulting from the `` flow '' picture .
to recover consistency with the strangeness neutrality of the fireball under these conditions , the hg equation of state would have to be very different from the conventional hagedorn resonance gas which we employed here .
while such a major modification of the physics of the hadron phase can not at present be completely excluded , it would invalidate a large body of successful phenomenological work based on the conventional hg picture .
thus , if the experimental result @xmath110 should be confirmed via other particle ratios , in our opinion it practically eliminates within the framework of a conventional hg - eos the flow interpretation of the spectra as a viable alternative to the purely thermal " interpretation .
we therefore conclude that within a hg picture the interpretation of the @xmath1-slope of @xmath308 mev as the true temperature of the fireball appears to be the only possibility consistent with the constraint of ( approximate ) strangeness neutrality .
this excludes the possibility of transverse flow , however , and thus causes consistency problems for the freeze - out process .
in addition , it does not lead to the observed value of @xmath263 .
therefore a hg interpretation of the data appears to be excluded .
if one tries instead to interpret the observations in terms of qgp formation , one faces the problem to understand how the hadronization process can lead to a final state with the observed values for the thermodynamic parameters .
we have seen that the conditions of strangeness neutrality and of a fixed up - down asymmetry in general lead , at the same temperature , to different values for the chemical potentials @xmath309 , @xmath310 , @xmath6 in the two phases . while a vanishing value of @xmath6 is a natural consequence of strangeness neutrality in the qgp , it is not at all natural in the hadron gas , where @xmath110 solves the strangeness neutrality condition only for very specific pairs of @xmath9 and @xmath13 , and thus in general it is not a value expected to arise in an interpretation of the experiments involving hg fireballs . on the other hand , while the values of the chemical potentials assumed in a hadron gas can be directly determined from the measured abundance ratios between the various hadronic species emitted by the gas , this is not automatically true for the qgp chemical potentials .
how @xmath309 , @xmath310 , and @xmath6 inside the qgp are related to the abundance ratios of the finally observed hadrons depends on the dynamical and kinetic details of the hadronization process . to illustrate this problem ,
let us consider the limiting case of a very long - lived and slowly hadronizing quark - gluon plasma , where hadronization proceeds through a mixed phase providing intimate contact between the two sub - phases .
in this case chemical equilibrium not only within the hadronic sub - phase ( with respect to processes which transform the hadrons among each other ) and within the quark sub - phase ( with respect to processes transforming quarks and gluons into each other ) , but also between the two phases ( with respect to processes which create hadrons out of quarks and gluons or dissociate them again into quarks and gluons ) can be maintained . as a consequence ,
the chemical potentials change smoothly from their values originally assumed in the qgp phase into the chemical equilibrium values of the hadron gas , and only the latter ( and not the former ) would be observable by studying particle ratios .
in such a scenario , the value of the strange quark chemical potential @xmath6 extracted from a suitable combination of strange hadron abundances @xcite would in general not be zero ( although it started out as zero in the original qgp phase ) , but would be given by the solution to the strangeness balance constraint ( [ eq6 ] ) in an equilibrated hadron gas .
however , as demonstrated in this paper , the strange particle yields from the wa85 collaboration @xcite indeed do point to a value of @xmath110 , and in view of the naturalness of this value in the context of a qgp we are inclined to believe that this result may not be entirely accidental .
therefore we ask under which conditions @xmath110 could have survived the qgp hadronization process after all . one possibility to transfer the qgp values for the chemical potentials and temperature directly to the final state hadrons
is to assume rapid break - up of the qgp combined with statistical quark recombination @xcite .
a necessary condition for the preservation of the qgp thermodynamic parameters is that sequentially produced hadrons escape sufficiently rapidly such that there is no formation of a intermediate hg phase in which the chemical potentials and/or momentum distributions could re - equilibrate .
one can view this process as the peeling of the qgp surface into the vacuum since there is no barrier , the surface regions of the qgp fireball are peeled off in form of rapidly escaping hadrons , carrying in their abundances the information about the thermal and chemical properties of the qgp .
one problem with such a picture is that the system of radiated hadrons , described by thermal distributions with chemical potentials and temperature as given originally in the qgp phase , by definition is characterized by the same particle abundance ratios and by the same value of the specific entropy as an equilibrium hg at the same temperature .
the radiation obtained from this quark recombination picture is in general neither strangeness neutral , nor does it carry away the correct amount of entropy per baryon ( as fixed by the original qgp ) it is well known that the entropy balance can be saved by considering the entropy contained in the gluons of the qgp , letting them materialize as hadrons via gluon fragmentation into additional quark - anti - quark pairs . in @xcite the additional quarks from gluon fragmentation where hadronized according to the same laws of combinatorics as the thermal quarks from the qgp ; as a result the qgp chemical potentials were not preserved , but the quark chemical potentials were modified during hadronization in the direction of their hg equilibrium values .
the minimal dynamical picture could consist of a primordial explosive production of the few high @xmath1 strange anti - baryons , followed by equilibrium hadronization of the remaining fireball at nearly constant entropy . to test such a picture requires the simultaneous measurement of the abundances and momentum spectra of ( strange and non - strange ) baryons and anti - baryons and of ( fully identified ) pions and kaons . only with such a complete set of data on produced hadrons
can a full understanding of the evolution of the primordial dense phase ( qgp ? ) be achieved .
of course , our contention that the presently available data are inconsistent with a hg picture for the collision and rather point to a very entropy - rich source of particle emission , possibly a quark - gluon plasma , must appear provocative at the present stage .
our findings show that it is necessary to widen the investigation of strange anti - baryon ratios to higher _ and _ lower nuclear beam energies and to study simultaneously the particle multiplicity and charge flow at high and low @xmath1 , in order to confirm the now very suggestive conclusion of qgp formation in the collisions of s ions at 200 gev a with heavy nuclear targets at near zero impact parameter . in view of the ambiguity between the hg and qgp picture , as far as the consistency of a vanishing strange quark chemical potential with the condition of strangeness neutrality is concerned , the presently available beam energies at cern turn out to be somewhat unfortunate , due to the observed particular value of around 210 mev for the @xmath31-slope ; this ambiguity could be easily resolved if @xmath104 were still observed in cases where the @xmath31-slope ( apparent temperature ) is lower or higher .
it is natural to expect such a variation in the slope parameters ( combined by an inverse variation of the baryon chemical potential ) once collisions at lower or higher beam energies are studied ( for example with a 50 gev a pb - beam at the cern sps or with 100 + 100 gev a au - beams at rhic ) .
we are also looking forward with anticipation to the forthcoming pb pb collisions at about 170 gev a projectile energy , which will allow to check and extend the results of our analysis by going to the largest possible thermodynamic systems using nuclear collisions .
* acknowledgements : * j.r .
thanks his colleagues in paris and regensburg for their warm and personal hospitality . the work of u.h . and j.s . was supported by dfg , bmft and gsi .
the collaboration between u.h . and
j.r . was supported by nato collaborative research grant 910991 .
j.r . acknowledges funding of his research by doe , grant de - fg02 - 92er40733 .
j. rafelski and a. schnabel , in : _ intersections between particle and nuclear physics _ , g. m. bunce ( ed . ) , aip conference proceedings * 176 * , american institute of physics , new york , 1988 , p. 1086
( see in particular fig.1 ) .
j. letessier , a. tounsi and j. rafelski , _ phys .
lett . _ * b292 * ( 1992 ) 417 + j. letessier , a. tounsi and j. rafelski , _ strange anti - baryons
qgp versus hg _ , paris preprint par / lpthe/92 - 32 , to appear in the proceedings of the international high energy physics conference , dallas , tx , august 1992 , american .
inst . of physics , 1993 .
k. s. lee , u. heinz , and e. schnedermann , _ z. phys .
_ * c48 * ( 1990 ) 525 ; + e. schnedermann and u. heinz , _ phys
lett . _ * 69 * ( 1992 ) 2908 ; + e. schnedermann , j. sollfrank , and u. heinz , _ fireball spectra _ , in : _ particle production in highly excited matter _ , h. h. gutbrod ( ed . ) , plenum , new york , in press . y. takahashi et al . ,
cern - emu05 collaboration , private communication .
u. heinz , p. koch , and b. friman , in : _ proceedings of the large hadron collider workshop _ , aachen , 1990 , g. jarlskog and d. rein ( eds . ) , cern report cern 90 - 10 , vol .
ii , p. 1079 .
p. koch and j. rafelski , _ nucl .
phys . _ * a444 * ( 1985 ) 678 .
r. hagedorn , i. montvay , and j. rafelski , `` thermodynamics of nuclear matter from the statistical bootstrap model '' , in : _ hadronic matter at extreme density _
, n. cabbibo and l. sertorio ( eds . ) , plenum , new york , 1978 .
j. rafelski and m. danos , _ phys .
_ * b192 * ( 1987 ) 432 .
b. l. friman , g. baym , and j .-
blaizot , _ phys .
* b132 * ( 1983 ) 291 ; + h. w. barz , b. friman , j. knoll , and h. schulz , _ nucl .
phys . _ * a484 * ( 1989 ) 661 . | a constraint between thermal fireball parameters arises from the requirement that the balance of strangeness in a fireball is ( nearly ) zero .
we study the impact of this constraint on ( multi-)strange ( anti-)baryon multiplicities and compare the hadron gas and quark - gluon plasma predictions .
we explore the relation between the entropy content and particle multiplicities and show that the data are compatible with the quark - gluon plasma hypothesis , but appear to be inconsistent with the picture of an equilibrated hadron gas fireball .
we consider the implications of the results on the dynamics of evolution and decay of the particle source .
-0.5 cm = -0.2 cm = -0.2 cm = 10000 _ submitted to physical review d _
* par / lpthe/9227 * + * tpr-92 - 28 * + * azph - th/92 - 23 * |
` unavoidable complexes ' , introduced by blagojevi , frick , and ziegler in @xcite , play the fundamental role in their ` constraint method ' ( gromov - blagojevi - frick - ziegler reduction ) .
inspired by @xcite , and in part developing some ideas from the unpublished preprint @xcite , here we begin a systematic study of ` unavoidable complexes ' , as combinatorial objects that may have some independent interest and which may deserve to be studied in their own right .
the paper is focused on developing some new ideas and constructions which may provide additional insight into the general structure of ` unavoidable simplicial complexes ' .
we begin with the observation that all known ` key examples ' of @xmath7-unavoidable complexes , used in the seminal paper @xcite ( lemma 4.2 ) , are ` linearly realizable ' ( threshold ) complexes , or contain such a complex as a subcomplex , see section [ sec : key - examples ] ( and definitions [ def : linear ] and [ def : simple ] in section [ sec : threshold ] ) .
for this reason one of the central themes in our paper is the construction of unavoidable complexes which are ` non - linear ' in some , precisely defined , sense ( see sections [ sec : intrinsic ] and [ sec : lin - realiz ] ) .
we introduce two natural and possibly new invariants of a simplicial complex @xmath9}$ ] , the _ partition number _ @xmath0 ( definition [ def : pi - invariant ] ) and the _ threshold characteristic _ @xmath10 ( definition [ def : threshold - char ] ) .
the ` linearity ' or ` non - linearity ' of a complex @xmath6 is then rephrased as a statement about mutual relationship of @xmath0 and @xmath10 ( as exemplified by proposition [ prop : vazna - implikacija ] and example [ eqn : tight ] in section [ sec : threshold - char ] ) .
the fundamental relation between these two invariants is the inequality , @xmath11 ( proposition [ prop : vazna - implikacija ] ) where the equality holds precisely if @xmath6 is an ` intrinsically linear complex ' ( see definition [ def : simple ] and corollary [ cor : char - simple ] ) .
we say that @xmath6 is ` intrinsically non - linear ' ( definition [ def : simple ] ) if the inequality in ( [ eqn : uvod - fund - ineq ] ) is strict , and one of our central objectives is to generate and study these complexes . especially interesting
are ` strongly non - linear complexes ' ( definition [ def : strong - intrinsic ] ) which are both intrinsically non - linear and minimal in the sense that @xmath12 for each proper subcomplex @xmath13 of @xmath6 . theorem [ thm : vazna-2 ] in section [ sec : intrinsic ] describes a large class of strongly non - linear complexes , emphasizing the role of ( highly symmetric ) minimal triangulations of manifolds that ` look like a projective plane ' ( section [ sec : look - alikes ] ) .
one of the highlights of the paper is an algorithmic procedure for evaluating @xmath10 ( see propositions [ prop : thr - char - min ] and [ prop : vazna - simetrija ] in section [ sec : threshold - char ] ) , based on the ideas from linear programming , which is quite efficient if @xmath6 has a reasonably large symmetry group .
this connection with polyhedral combinatorics turns out to be very fruitful and a more complete report of ( computer experiments and other results ) will appear elsewhere . in order to evaluate ( estimate ) the partition invariant @xmath0 , we rely on the assumption that @xmath14 has a join decomposition , where @xmath15 is @xmath16-unavoidable for some @xmath17
. one of our central results ( theorem [ theo : join ] ) says that @xmath6 , as a subcomplex of @xmath18 where @xmath19 , is @xmath7-unavoidable provided @xmath20 .
combined together , these ingredients ( the join decomposition @xmath14 of carefully chosen complexes @xmath21 with known partition number and a large group of automorphisms ) lead to the construction of ` intrinsically non - linear ' unavoidable complexes and ` strongly non - linear complexes ' ( section [ sec : intrinsic ] ) . the observation that the large symmetry group may be instrumental in detecting non - linear examples of unavoidable complexes , shifted our attention and focus to known examples of @xmath22-unavoidable ( self - dual ! ) complexes with large group of automorphisms .
an immediate choice were the unique ( @xmath23-vertex ) minimal triangulation @xmath24 of the real projective plane , and the unique ( @xmath25-vertex ) minimal triangulation @xmath26 of the complex projective plane , together with the three element @xmath27-dimensional complex @xmath28 $ ] and the known non - isomorphic simplicial complexes associated with the quaternionic projective plane @xmath29 @xcite .
the threshold characteristic of these combinatorial ` projective planes ' is calculated in section [ sec : look - alikes ] .
the paper is organized as follows . the partition number @xmath0 , together with its relatives and more general versions @xmath30 and @xmath31 , is introduced in section [ sec : partition ] . in section
[ sec : reduction ] we briefly review the ` constraint method ' ( gromov - blagojevi - frick - ziegler reduction ) and the reduction based on ` sarkaria s inequality ' .
these two reduction procedures have been successfully applied to theorems of tverberg - van kampen - flores type , and they serve as our primary motivation for studying unavoidable complexes . fundamental constructions ( pushdown , join , etc . ) and basic properties of @xmath7-unavoidable complexes , including the inequality @xmath32 ( corollary [ cor : pi - join ] ) , are presented in section [ sec : fundamental ] .
threshold complexes and their manyfold relations with unavoidable complexes are studied in sections [ sec : threshold ] and [ sec : threshold - char ] .
the focus is on the ` threshold characteristic ' @xmath10 of a simplicial complex @xmath6 and its relationship with the partition number @xmath0 .
section [ sec : intrinsic ] provides a summary and a list of our main examples of _ intrinsically non - linear _ and _ strongly non - linear _ simplicial complexes . finally in section
[ sec : class ] we offer a glimpse into some future developments , including a promising connection with cooperative game theory . in the focus of our study
are so called @xmath7-unavoidable complexes @xmath9}$ ] , characterized by the property that for each partition @xmath4 = a_1\uplus\dots\uplus a_r$ ] of @xmath4 $ ] into @xmath7 pieces , there exists @xmath33 $ ] such that @xmath34 ( see ( * ? ? ?
* definition 4.1 ) and ( * ? ? ?
* section 2 ) as well as the definition [ def : neizbezno ] in section [ sec : partition ] ) . as a preview and illustration of properties analyzed in our paper , here is an example of an @xmath7-unavoidable simplicial complex which is _ strongly non - linear _ ( definition [ def : strong - intrinsic ] in section [ sec : intrinsic ] ) , in the sense that it is _ not _
@xmath7-unavoidable for ` obvious reasons ' ( it does not contain an @xmath7-unavoidable linear threshold complex ) , and which is minimal in the sense that if @xmath13 is a proper subcomplex of @xmath6 than @xmath35 is _ not _ @xmath7-unavoidable .
-copies of @xmath24 . ]
let @xmath24 be the ` hemi - icosahedron ' ( * ? ? ?
* example 5.8.5 ) , the simplicial complex on @xmath23 vertices , arising as the unique , minimal triangulation of the real projective plane .
let @xmath36 be the join of @xmath37 copies of @xmath24 where @xmath38 .
as a complex on the set @xmath39\times [ r-1]$ ] of cardinality @xmath40 , the complex @xmath41 is @xmath7-unavoidable ( corollary [ cor : spec - cases ] to theorem [ theo : join ] in section [ sec : join ] ) .
the complex @xmath24 has a very large symmetry group inherited from the group of symmetries of the icosahedron . as a consequence the complex @xmath41 also has a very large group of symmetries
this fact alone ( theorem [ thm : vazna-2 ] in section [ sec : intrinsic ] ) is the main reason why @xmath41 is _ intrinsically non - linear _
, i.e. why there does not exist an @xmath7-unavoidable threshold complex @xmath42 ( section [ sec : threshold ] ) such that @xmath43 .
one way of relating topological and combinatorial complexity of an @xmath7-unavoidable simplicial complex @xmath44}$ ] is via the inequality ( corollary [ thm : izvor ] ) , @xmath45 where @xmath7 is a prime number , @xmath46 is the @xmath7-fold deleted join of @xmath6 and @xmath47 is the @xmath48-equivariant index function described in ( * ? ? ?
* section 6 ) .
as a consequence we obtain the inequality , @xmath49 which in turn implies a borsuk - ulam type statement saying that there does not exist a @xmath50-equivariant map , @xmath51 where @xmath52 is a free @xmath50-complex of dimension @xmath53 .
* acknowledgements : * two of the authors ( m. jeli and r. ivaljevi ) participated ( at the final stage of the project ) in the program ` topology in motion ' at the institute for computational and experimental research in mathematics ( icerm , brown university ) . with great pleasure
we acknowledge the support , hospitality and excellent working condition at icerm .
in this section we collect the definitions of the key ` partition invariants ' of simplicial complexes .
the central are definitions [ def : pi - invariant ] and [ def : neizbezno ] .
[ def : simpl - compl ] a simplicial complex @xmath55 on a set @xmath56 is a pair @xmath57 of sets such that @xmath58 and , @xmath59 in general the set @xmath60 of vertices of @xmath6 is allowed to be a proper subset of the ambient set @xmath56 . the set @xmath61 is not properly defined as a simplicial complex until we make explicit the ambient set @xmath56 .
similarly , a simplicial complex @xmath6 ( in the usual sense ) is turned into a simplicial complex in the sense of definition [ def : simpl - compl ] by specifying a set @xmath56 such that @xmath62 or equivalently @xmath58 .
* caveat : * following the usual practise we will sometimes write @xmath6 instead of @xmath55 when the ambient set @xmath56 is clear from the context .
the partition number @xmath63 of a simplicial complex @xmath64 was introduced in @xcite as a combinatorial invariant relevant for the study of results of tverberg - van kampen - flores type .
[ def : pi - invariant ] the partition number or @xmath65-invariant @xmath66 of a simplicial complex @xmath67 is the minimum integer @xmath2 such that for each partition @xmath68 of @xmath56 at least one of the sets @xmath5 is in @xmath6 . more generally , if @xmath69 is an arbitrary set ( hypergraf ) then the associated @xmath70-partition number @xmath71 of @xmath64 is defined as the minimum integer @xmath2 such that for each subpartition @xmath72 in @xmath56 , where @xmath73 for each @xmath74 $ ] , at least one of the sets @xmath5 is in @xmath6 . _
( @xmath7-unavoidable complexes , @xcite ) _ [ def : neizbezno ] let @xmath38 be an integer .
suppose that @xmath64 is a simplicial complex with vertices in @xmath56 .
if @xmath75 we say that the complex @xmath55 is @xmath7-unavoidable , or equivalently that @xmath6 is @xmath7-unavoidable on @xmath56 .
more explicitly @xmath6 is @xmath7-unavoidable on @xmath56 if , @xmath76 more generally @xmath58 is @xmath7-unavoidable relative to a hypergraph @xmath77 if @xmath78 .
if @xmath79 and the complex @xmath6 is @xmath7-unavoidable then @xmath35 is @xmath7-unavoidable as well .
this is the reason why it may be sometimes useful to focus on ` minimal @xmath7-unavoidable complexes ' .
[ def : minimal ] a complex @xmath67 is _
minimal @xmath7-unavoidable on @xmath56 _ if it is @xmath7-unavoidable on @xmath56 and if @xmath80 is a proper subcomplex of @xmath6 then @xmath35 is not @xmath7-unavoidable on @xmath56
. here we collect , for completeness and further reference , some auxiliary definitions .
the following natural extension of definition [ def : pi - invariant ] is also useful .
recall that the _ covering multiplicity _ of a finite family @xmath81 of sets is the cardinality of the largest subset @xmath82 such that @xmath83 .
[ def : pi - extension ] suppose that @xmath84 and @xmath85 are integers and let @xmath69 .
the extended partition number @xmath86 of @xmath64 is defined as the minimum integer @xmath2 such that for each collection @xmath87 of size @xmath88 and covering multiplicity at most @xmath85 , the set @xmath89 has at least @xmath84 elements .
in particular , @xmath90 .
the property of being @xmath7-unavoidable is an intrinsic , combinatorial property of a simplicial complex @xmath6 .
here is a natural generalization . _
( @xmath91-unavoidable complexes ) _ [ def : r - and - s ] choose integers @xmath92 and @xmath85 and let @xmath93}$ ] .
a simplicial complex @xmath44}$ ] is called _
@xmath91-unavoidable _ on @xmath94 $ ] ( relative to @xmath70 ) if @xmath95 .
more explicitly , @xmath6 is @xmath91-unavoidable on @xmath94 $ ] ( relative to @xmath96 ) if for each collection @xmath97}$ ] of size @xmath98 and covering multiplicity at most @xmath85 , where @xmath99 for each @xmath100 , the cardinality of the set @xmath89 is at least @xmath84 .
we say that a complex is @xmath101-unavoidable ( relative to @xmath70 ) if it is @xmath102-unavoidable ( relative to @xmath70 ) .
the ultimate reason for studying various classes of _ unavoidable complexes _ is their central role in reduction methods for proving results of tverberg - van kampen - flores type .
there exist two main reduction procedures , the _ ` constraint method ' _ or the gromov - blagojevi - frick - ziegler reduction , and the method based on _
inequality_. the gromov - blagojevi - frick - ziegler reduction , or the _ ` constraint method ' _ , is an elegant and powerful method for proving results of tverberg - van kampen - flores type . in its basic form
the method can be summarized as follows .
@xmath103 suppose that @xmath104}$ ] is a pair of simplicial complexes .
assume that an analogue of the continuous tverberg theorem holds for the triple @xmath105 in the sense that for each continuous map @xmath106 there exists a collection of @xmath7 disjoint faces @xmath107 of @xmath35 such that @xmath108 .
suppose that @xmath6 is _
@xmath7-unavoidable relative to @xmath35 _ in the sense that @xmath109 ( definitions [ def : pi - invariant ] and [ def : neizbezno ] ) .
more explicitly we assume that for each subpartition @xmath110 $ ] , where @xmath111 for each @xmath33 $ ] , at least one of the simplices @xmath5 is in @xmath6 . under these assumptions
the conclusion is that @xmath112 is also a tverberg triple in the sense that for each continuous map @xmath113 there exist disjoint simplices @xmath114 such that @xmath115 .
indeed , let @xmath116 be an extension ( @xmath117 ) of the map @xmath118 to @xmath35 .
suppose that @xmath119 is the function @xmath120 , measuring the distance of the point @xmath121 from @xmath6 .
define @xmath122 and assume that @xmath123 is the associated family of vertex disjoint faces of @xmath35 , such that @xmath124 .
more explicitly suppose that @xmath125 such that @xmath126 for each @xmath127 . since @xmath6 is @xmath7-unavoidable relative to @xmath35 , @xmath128 for some @xmath100 . as a consequence @xmath129 , and in turn @xmath130 for each @xmath131 .
if @xmath132 is the minimal face in @xmath35 containing @xmath133 then @xmath134 for each @xmath135 and @xmath136 . for a more complete exposition and numerous examples of applications of the ` constraint method '
the reader is referred to @xcite , see also ( * ? ? ? *
section 2.9(c ) ) and ( * ? ? ?
* proposition 2.5 ) .
the preprint @xcite offers a different reduction procedure , based on sarkaria s index inequality @xcite and @xcite , illustrating a different use of the @xmath7-unavoidability of the complex @xmath6 . as in section [ sec : gbfz - reduction ] we assume that @xmath137 is a tverberg triple in the sense that for each continuous map @xmath138 there exists a collection of @xmath7 disjoint faces @xmath107 of @xmath35 such that @xmath108 .
assume that @xmath139 and suppose that @xmath6 is _
@xmath7-unavoidable relative to @xmath35 _
( definition [ def : neizbezno ] ) .
we want to show that @xmath112 is a tverberg triple as well .
so far the set up is the same as in section [ sec : gbfz - reduction ] and in particular we have the following commutative diagram ( where @xmath140 is replaced by @xmath141 ) .
@xmath142 following the standard ` configuration space - test map ' reduction scheme @xcite , a map @xmath113 induces an @xmath143-equivariant map @xmath144 , where @xmath46 is the @xmath7-fold deleted join of @xmath6 @xcite and @xmath145 is the diagonal subspace .
the problem is to show that this map must have a zero , which is directly related to the problem of estimating the equivariant index @xmath146 where @xmath147 is a cyclic group and @xmath7 is a prime number .
this transition is functorial and in particular one obtains from ( [ eqn : cd-2 ] ) the following commutative diagram , @xmath148 where @xmath149 ( or more generally some other integer @xmath139 ) . in this setting the constraint method , based on the diagram ( [ eqn : cd-1 ] ) , can be replaced by the index calculation ( sarkaria s inequality ) ( based on the diagram ( [ eqn : cd-3 ] ) ) .
( in diagram ( [ eqn : cd-3 ] ) we tacitly assume that the target spaces are subspaces of the corresponding ( quotient ) vector spaces . ) recall that the sarkaria s inequality ( * ? ? ?
* proposition 6.2.4 ) is the index inequality , @xmath150 where @xmath151 is a free @xmath152-complex , @xmath153 is @xmath152-invariant , and @xmath154 is the order complex of the poset @xmath155 . by specializing @xmath156 and
@xmath157 one obtains a tool for estimating the equivariant index of @xmath46 .
the @xmath7-unavoidability of @xmath6 relative to @xmath35 , emerges in the form of the following simple lemma ( see also the proof of theorem 3.6 . in @xcite ) .
[ lema : index ] let @xmath38 be a prime .
suppose that @xmath158}$ ] and assume that @xmath6 is @xmath7-unavoidable relative to @xmath35 ( definition [ def : neizbezno ] ) .
let @xmath156 and @xmath157 and let @xmath154 be the ordered complex of the set @xmath159 ordered by the inclusion .
then , @xmath160
* proof : * by construction a simplex in @xmath156 is an @xmath7-tuple @xmath161 of disjoint simplices @xmath111 . by definition @xmath162 if and only if @xmath163 \mid a_i\notin k\}$ ] is a non - empty set . since @xmath6 is @xmath7-unavoidable relative to @xmath35 , we observe that @xmath164 $ ] for each @xmath162 .
the map @xmath165}\setminus\{\emptyset , [ r]\}$ ] is monotone so it induces a @xmath143-equivariant map of the associated order complexes , @xmath166}\setminus\{\emptyset , [ r]\})\cong s^{r-2}.\ ] ] this completes the proof of the lemma in light of the monotonicity property of the index function @xmath167 , ( * ? ? ?
* section 6 ) .
@xmath168 [ thm : izvor ] suppose that @xmath44}$ ] is an @xmath7-unavoidable complex with vertices in @xmath94 $ ] .
assume that @xmath169 is a prime and let @xmath170 be the associated cyclic group .
let @xmath46 be the @xmath7-fold deleted join of @xmath6 .
then , @xmath171
* proof : * by assumption @xmath6 is @xmath7-unavoidable relative to @xmath172 ) = 2^{[m]}$ ] . in this case
@xmath173^{\ast m}$ ] is a @xmath143-free , @xmath174-dimensional , @xmath175-connected simplicial complex .
it follows that @xmath176 which , together with lemma [ lema : index ] and inequality ( [ eqn : sarkaria ] ) , implies the inequality ( [ eqn : izvor ] ) . @xmath168
in this section we collect some general procedures for constructing new unavoidable complexes from the old ones . we start with very simple and basic constructions which pave the way for discovering less obvious examples of unavoidable complexes , including the construction of _ intrinsically non - linear _ ( definition [ def : simple ] ) and _ strongly non - linear _ unavoidable complexes ( section [ sec : intrinsic ] ) .
[ def : induced ] a map @xmath177 of simplicial complexes @xmath178 and @xmath179 is a function @xmath180 such that @xmath181 .
if @xmath178 is a simplicial complex and @xmath182 an arbitrary map then @xmath183 is the simplicial complex on @xmath184 defined by @xmath185 .
the complex @xmath179 is referred to as the _ pushforward _ of @xmath186 or the complex _ induced on @xmath184 by the map @xmath118 _ .
[ prop : induced - r - unavoid ] let @xmath187 be an arbitrary map and let @xmath188 be a simplicial complex on @xmath56 . then , @xmath189 where @xmath190 is the complex on @xmath184 induced by @xmath118 ( definition [ def : induced ] ) . as a consequence if @xmath6 is @xmath7-unavoidable on @xmath56 then @xmath191 is @xmath7-unavoidable on @xmath184 . in the following examples @xmath118 is an inclusion map
. the complex @xmath61 is @xmath7-unavoidable on @xmath192 $ ] . as a consequence the complex @xmath193 \mid
a \cap [ r-1]=\emptyset\}$ ] is @xmath7-unavoidable on @xmath94 $ ] where @xmath194 .
the complex @xmath195}\choose{\leq 1 } } = \{a\subset [ 2r-1]\mid \vert a\vert\leq 1\}$ ] is @xmath7-unavoidable on @xmath196 $ ] .
it follows that for each @xmath197 the complex @xmath198\mid \vert a\cap [ 2r-1]\vert \leq 1\}$ ] is @xmath7-unavoidable on @xmath94 $ ] .
the central result in this section is theorem [ theo : join ] which allows us to construct new unavoidable complexes as joins of known unavoidable complexes ( typically with smaller @xmath65-invariant ) .
we begin with the observation that the classes of @xmath199-unavoidable and @xmath7-unavoidable complexes coincide if @xmath200 .
more generally this is true for @xmath201-closed hypergraphs , and in particular this holds for hypergraphs @xmath202 where @xmath203 is an order ideal .
a hypergraph @xmath204 is @xmath201-closed if , * @xmath205 ; * if @xmath206 then @xmath207 .
recall ( definition [ def : r - and - s ] ) that a complex @xmath208 is @xmath209-unavoidable , relative to @xmath204 , if for each collection @xmath210}\subset \mathcal{h}$ ] of pairwise disjoint sets , the cardinality of the set @xmath211\mid a_j\in k\}$ ] is at least @xmath84 .
[ prop : coincide ] suppose that @xmath70 is a @xmath201-closed hypergraph on @xmath212 .
a complex @xmath208 is @xmath199-unavoidable ( relative @xmath70 ) if and only if it is @xmath7-unavoidable ( relative @xmath70 ) .
* proof : * since @xmath205 , each disjoint @xmath7-family @xmath213 can be completed to a disjoint family @xmath214 where @xmath215 for @xmath216 .
from here it immediately follows that if @xmath6 is @xmath199-unavoidable it must be @xmath7-unavoidable as well . for the opposite implication
let us assume that @xmath217 is an @xmath7-unavoidable simplicial complex ( relative @xmath70 ) .
let @xmath218 ( hence @xmath219 ) and let @xmath220 be a ( pairwise ) disjoint family of sets in @xmath70 .
we claim that the cardinality of the set @xmath211\mid
k\}$ ] is at least @xmath221 . for a given
set @xmath222 $ ] of cardinality @xmath223 let @xmath224 where @xmath225 . since @xmath70 is @xmath201-closed the family @xmath226 consists of @xmath7 pairwise disjoint sets which are all elements in @xmath70 .
there are two possibilities : 1 .
there exists some @xmath222 $ ] of cardinality @xmath223 such that @xmath227 . in this case
@xmath34 for each @xmath228 which concludes the proof .
2 . in the opposite case
we observe that for each set @xmath222 $ ] of cardinality @xmath223 there exists @xmath229\setminus i$ ] such that @xmath34 .
it follows that there can be at most @xmath230 elements @xmath231 $ ] such that @xmath232 , hence @xmath34 for at least @xmath233 distinct indices @xmath229 $ ] .
@xmath168 [ theo : join ] let @xmath234 be a collection of simplicial complexes where @xmath235 .
assume that @xmath236 is @xmath237-unavoidable on @xmath238 for some integers @xmath239 and @xmath240 .
then the join @xmath241 is an @xmath7-unavoidable complex on @xmath19 provided , @xmath242 * proof : * by definition @xmath243 , where @xmath244 .
let @xmath245 be an arbitrary family of pairwise disjoint sets in @xmath212 where by assumption @xmath246 .
each @xmath247 can be uniquely represented as a disjoint union @xmath248 where @xmath249 for each @xmath33 $ ] and @xmath231 $ ] .
for each @xmath231 $ ] let @xmath250 \mid a^{(i)}_j \in k_j\}$ ] . in order to show that @xmath251 for some @xmath33 $ ] we are supposed to prove that , @xmath252.\ ] ] since the complex @xmath236 is by assumption @xmath237-unavoidable , we know by proposition [ prop : coincide ] that it is also @xmath253-unavoidable .
it follows that the set @xmath254 has cardinality at least @xmath255 which implies that , @xmath256 and ( [ eqn : presek ] ) is an immediate consequence .
@xmath168 [ cor : pi - join ] if @xmath14 is a join of @xmath257 simplicial complexes then , @xmath258 here are some special cases and consequences of corollary [ cor : pi - join ] .
[ cor : spec - cases ] 1 .
let @xmath6 be an @xmath7-unavoidable complex .
then the join of @xmath259 copies of @xmath6 is @xmath260-unavoidable .
2 . let @xmath6 be a @xmath22-unavoidable complex .
then the join of @xmath259 copies of @xmath6 is @xmath261-unavoidable .
3 . the complex @xmath262 is self - dual ( minimal @xmath22-unavoidable ) .
it follows that the complex @xmath263 is @xmath7-unavoidable . _
( minimal @xmath7-unavoidable complex ) _ [ def : min - unavoidable ] let @xmath243 be an @xmath7-unavoidable simplicial complex .
we say that @xmath6 is a _
minimal @xmath7-unavoidable complex _
if for every maximal simplex @xmath264 , the complex @xmath265 is not @xmath7-unavoidable .
[ lema : deletion_max ] let @xmath243 be an @xmath7-unavoidable simplicial complex on @xmath212 .
let @xmath264 be a maximal simplex in @xmath6 .
then , @xmath266 is @xmath7-unavoidable on @xmath212 if and only if @xmath267 is @xmath37-unavoidable on @xmath268 , where @xmath269 is the restriction of @xmath6 to the set of vertices @xmath270 .
* proof : * @xmath271 suppose that @xmath266 is an @xmath7-unavoidable complex on @xmath212 .
if @xmath272 is a partition of @xmath268 then @xmath273 is a partition of @xmath212 . by assumption @xmath274 for some @xmath100 and as a consequence @xmath275 .
@xmath276 suppose that @xmath267 is @xmath37-unavoidable on @xmath268 .
let @xmath277 be a partition of @xmath212 .
since @xmath6 is @xmath7-unavoidable then @xmath278 for some @xmath33 $ ] . if @xmath279 then @xmath274 and we are done . if @xmath280 then @xmath281 is a partition of @xmath268 . by assumption @xmath282 for some @xmath283 which implies @xmath284
. @xmath168 [ cor : main - min ] _ ( criterion for minimal @xmath7-unavoidability ) _ let @xmath243 be an @xmath7-unavoidable complex .
then @xmath6 is minimal @xmath7-unavoidable if and only if for every maximal simplex @xmath264 , the complex @xmath267 is not @xmath37-unavoidable on @xmath268 .
[ cor : pi - equality ] if @xmath9}$ ] is a minimal @xmath7-unavoidable complex then @xmath285 . *
proof : * since @xmath6 is @xmath7-unavoidable we know that @xmath8 . by minimality of @xmath6 for a given facet @xmath286
there exists a partition @xmath287\setminus a$ ] such that @xmath288 for each @xmath289 $ ] . in turn
@xmath290 $ ] is a partition of @xmath4 $ ] into @xmath37 sets proving that @xmath291 .
@xmath168 [ prop : join - min ] let @xmath234 be a family of complexes such that for each @xmath231 $ ] the complex @xmath235 is a minimal @xmath237-unavoidable complex on @xmath238 . then the join @xmath292 is a minimal @xmath7-unavoidable complex on @xmath19 where @xmath293 .
* proof : * by theorem [ theo : join ] we know the @xmath6 is @xmath7-unavoidable on @xmath212 . in order to prove its minimality , it is sufficient ( by lemma [ lema : deletion_max ] and corollary [ cor : main - min ] ) to prove that for every maximal simplex @xmath264 the complex @xmath267 is not @xmath37-unavoidable on @xmath268 .
the simplex @xmath294 , as a maximal simplex in @xmath6 , is the join of maximal simplices @xmath295 , where @xmath5 is a maximal simplex in @xmath21 for each @xmath100 . by lemma [ lema : deletion_max ] we know that for each @xmath296 , @xmath297 is not @xmath298-unavoidable , which implies that there exists a partition , @xmath299 such that for each @xmath100 and @xmath85 , @xmath300 .
from here we obtain a partition of @xmath301 by taking the union of all partitions ( [ eqn : partition ] ) , @xmath302 this way we obtain a partition of @xmath268 into @xmath303 subsets such that none of the pieces is a simplex in @xmath267 .
it follows that the complex @xmath267 is _ not _
@xmath37-unavoidable , which completes the proof of the theorem . @xmath168
for a given function @xmath304 and a ` threshold ' @xmath305 the associated sub - level set is @xmath306 .
suppose that @xmath307 is a boolean partially ordered set ( poset ) and let @xmath308 be a monotone function which satisfies the property @xmath309 .
then the associated sub - level set @xmath310 is a simplicial complex .
this simplicial complex is referred to as the @xmath311-threshold complex and often denoted by @xmath312 . in this section
we collect some of the most fundamental properties of @xmath7-unavoidable threshold complexes .
[ def : weigts - and - measures ] a ( positive ) weight distribution on @xmath94 $ ] is a vector @xmath313 where @xmath314 for each @xmath100 and @xmath315 .
the associated positive measure on @xmath316}$ ] ( also denoted by @xmath317 ) is defined by @xmath318 . if @xmath319)=1 $ ] then the measure @xmath317 is referred to as a probability measure .
more generally a _ superadditive measure _
@xmath2 on @xmath316}$ ] is a non - negative function @xmath320 } \rightarrow \mathbb{r}_+$ ] such that @xmath321)>0 $ ] , @xmath322 and , @xmath323 [ exam : superadditive]suppose that @xmath324 is a finite collection of positive measures on @xmath94 $ ] .
then the function defined by , @xmath325 is a ( positive ) superadditive measure on @xmath94 $ ] . for the reasons explained in section [ sec : geometric ] the function @xmath2 is referred to as a _ geometric superadditive measure_. [ def : mu - kompleksi ] suppose that @xmath326 is a positive measure on @xmath94 $ ] and let @xmath327 .
the associated sub - level simplicial complexes @xmath328 and @xmath329 , are defined by , @xmath330 \mid \mu(a)\leq \beta\ } , \qquad k_{\mu < \beta } = \{a\subset [ m ] \mid \mu(a ) < \beta\}.\ ] ] more generally , if @xmath2 is a positive superadditive measure on @xmath94 $ ] then the associated sub - level sets , @xmath331 \mid \nu(a)\leq \beta\ } , \qquad k_{\nu < \beta } = \{a\subset [ m ] \mid \nu(a ) < \beta\}\ ] ] are also simplicial complexes .
[ prop : concave ] suppose that @xmath2 is a positive superadditive measure on @xmath94 $ ] and let @xmath332 ) = \alpha$ ] .
then the sub - level complex @xmath333 is @xmath7-unavoidable .
* proof : * suppose that @xmath94 = a_1\cup\ldots\cup a_r$ ] is a partition
. then at least one of the sets @xmath5 is in @xmath334 , otherwise we would have a contradiction , @xmath335 ) = \nu(a_1\cup\ldots\cup a_r)\geq \nu(a_1)+\ldots + \nu(a_r ) > r\frac{\alpha}{r } = \alpha.\ ] ] suppose that @xmath2 is a positive superadditive measure on @xmath94 $ ] and let @xmath332 ) = \alpha$ ] .
then for each @xmath336 , @xmath337 [ def : linear ] an @xmath7-unavoidable simplicial complex @xmath44}$ ] is called _ ` linear ' _ or _ ` linearly realizable ' _ if @xmath338 for some probability measure @xmath317 on @xmath94 $ ] .
the @xmath7-unavoidable complexes described in proposition [ prop : concave ] are for similar reasons sometimes referred to as _ ` concave ' _ or _ ` concavely realizable ' _ complexes .
the map @xmath339}\rightarrow \{0,1\}^m$ ] , identifying a subset @xmath340 $ ] with its characteristic function @xmath341 , allows us to give a simple and useful geometric interpretation of threshold complexes @xmath342 in the situation when the threshold function @xmath343}\rightarrow \mathbb{r}_+$ ] has a natural extension @xmath344 ( or @xmath345^{m}\rightarrow \mathbb{r}_+$ ] ) .
for example if @xmath346 is the linear threshold complex associated to a measure @xmath326 then , @xmath347 where @xmath348 is the negative half - space associated to @xmath349 . a similar interpretation can be given to the ` concave ' threshold complex @xmath350 where , in agreement with ( [ eqn : min ] ) , the concave function @xmath2 is defined on @xmath351 by the formula , @xmath352 the question when a general superadditive function @xmath2 has a concave extension to @xmath351 is more subtle .
this is one of the reasons why we pay a special attention to ` geometric superadditive functions ' ( example [ exam : key-(iv ) ] and proposition [ prop : concave-2-linear ] ) .
if @xmath44}$ ] is an arbitrary simplicial complex , identified with its geometric image @xmath353^m$ ] , we can check if it is a linear threshold complex as follows .
let @xmath354}\setminus k$ ] be the complementary set or the associated ` simple game ' ( in the language of cooperative game theory @xcite ) .
we also identify @xmath35 with its image @xmath355^m$ ] . then @xmath6 is a linear threshold complex if and only if @xmath356 ( figure [ fig : separacija-2 ] ) .
moreover the size ( volume , number of integer points , etc . ) of the convex set @xmath357 can be used as a measure of its geometric non - linearity .
[ prop : concave-2-linear ] suppose that @xmath2 is a geometric superadditive measure on @xmath94 $ ] ( example [ exam : superadditive ] ) .
suppose that @xmath332 ) = \alpha$ ] and let @xmath358 be the associated sub - level @xmath7-unavoidable complex .
then @xmath6 contains a linearly realizable , @xmath7-unavoidable complex , @xmath359 * proof : * let @xmath360 and @xmath361 . by assumption
@xmath2 admits an extension ( [ eqn : min - min ] ) to a concave function @xmath362 , which in turn implies the convexity of the set , @xmath363 if @xmath364 then @xmath365 .
indeed , if @xmath366 then @xmath367 while @xmath368 implies @xmath369 . note that @xmath370 is closed while @xmath141 is an open convex set in @xmath351 .
let @xmath371 be a linear function that separates convex sets @xmath370 and @xmath141 , @xmath372 \quad \mbox { { \rm and } } \quad d\subset f^{-1}(c , + \infty).\ ] ] from the first inclusion in ( [ eqn : separation ] ) we deduce that @xmath373 for each @xmath374 . moreover if @xmath375 ( where @xmath376 ) then @xmath377 . from the second inclusion in ( [ eqn : separation ] )
we deduce that @xmath378 which implies @xmath379 . by normalizing
we observe that the probability measure @xmath313 defined by @xmath380 has the properties claimed by the proposition .
[ fig : separacija-1 ] [ def : simple ] suppose that @xmath381 and @xmath382 are @xmath7-unavoidable complexes on the same set @xmath56 of vertices .
if @xmath383 we say that @xmath6 is dominated by @xmath35 ( or that @xmath35 dominates @xmath6 ) .
we also say that a probability measure @xmath317 on @xmath56 is dominated by @xmath6 if @xmath384 . by definition
an @xmath7-unavoidable complex @xmath6 is _ intrinsically linear _ if @xmath384 for some probability measure @xmath317 on @xmath56 , otherwise we call it _
intrinsically non - linear_. [ prop : simetrija ] let @xmath9}$ ] be an @xmath7-unavoidable complex on @xmath4 $ ] .
then the set of all probability measures on @xmath4 $ ] dominated by @xmath6 , @xmath385)=1 \mbox{\rm { \ , and } } k_{\mu \leq 1/r } \subseteq k\}\ ] ] is a convex subset of the simplex @xmath386 of all probability measures on @xmath4 $ ] .
moreover , the group @xmath152 of symmetries of @xmath6 acts on @xmath387 and in particular if @xmath388 then there exists a @xmath152-invariant measure @xmath317 dominated by @xmath6 . *
proof : * a probability measure @xmath389 is in @xmath387 if and only if @xmath390 for each non - simplex @xmath391 .
the set @xmath392 is an open half - space in @xmath393 and @xmath387 is convex since it can be represented as the intersection , @xmath394 the set @xmath387 is clearly invariant with respect to the ( permutation ) group @xmath152 which acts on @xmath387 from the right .
a @xmath152-invariant measure is obtained from any initial measure @xmath395 as the average @xmath396 . @xmath168
it may be interesting to study how large can be a threshold complex which is dominated by a simplicial complex @xmath6 ( which is not necessarily assumed to be @xmath7-unavoidable ) . in this section
we introduce the _ threshold characteristic _ of the complex @xmath6 as the unique number @xmath10 such that the equivalence , @xmath397 holds for some probability measure @xmath317 .
[ def : threshold - char ] let @xmath9}$ ] be a simplicial complex and let @xmath398 be the simplex of probability measures on @xmath4 $ ] .
the threshold characteristic @xmath10 of @xmath6 is defined by , @xmath399 \mid ( \exists \mu\in \delta_{n-1})~ k_{\mu\leq\alpha } \subset k\ } \\
\nonumber & = & \mbox{\rm max}\{\alpha\in [ 0,+\infty ] \mid ( \exists \mu\in \delta_{n-1})~ k_{\mu < \alpha } \subset k\}. \end{aligned}\ ] ] the equivalence of two definitions of @xmath10 in ( [ eqn : thr-2 ] ) can be checked directly , or deduced ( more geometrically ) from ( the proof of ) proposition [ prop : thr - char - min ] . by definition [ def : threshold - char ] @xmath400 if and only if @xmath401}$ ] .
the reason the supremum ( maximum ) is evaluated in the interval @xmath402 $ ] ( rather than in @xmath403 $ ] ) is the inequality ( [ eqn : vazna - nejednakost ] ) which ( by our definition ) reduces to an equality in the case @xmath401}$ ] ( see also proposition [ prop : kad - je - jednakost ] ) .
the following consequence of ( [ eqn : char - ineq ] ) is a fundamental relation between invariants @xmath0 and @xmath10 that holds unconditionally for all simplicial complexes @xmath6 .
[ prop : vazna - implikacija ] if @xmath9}$ ] is a simplicial complex on @xmath4 $ ] and @xmath404 is an integer then , @xmath405 the implication ( [ eqn : vazna - implikacija ] ) is equivalent to either of the following equivalent inequalities , @xmath406 [ eqn : tight ] let @xmath407 where @xmath408 are integers .
let @xmath409\choose \leq k}$ ] . then @xmath410 and @xmath411 ( see example [ exam : full - symmetry ] ) .
it follows that the inequalities ( [ eqn : vazna - nejednakost ] ) can not be improved in general .
observe that if @xmath412}$ ] then @xmath413 .
note also that if @xmath9}$ ] then @xmath400 if and only if @xmath401}$ ] . since @xmath414 we could have restricted thresholds @xmath415 to the interval @xmath403 $ ] in definition [ def : threshold - char ] .
our choice was motivated by the inequality ( [ eqn : vazna - nejednakost ] ) and the relation @xmath416 .
the following geometric interpretation of the invariant @xmath10 is also useful . for a given non - simplex @xmath391 and @xmath417 let ,
@xmath418 observe that @xmath419 and @xmath420 where @xmath421 ( respectively @xmath422 .
the convex set , @xmath423 is referred to as the _ anti - dual polyhedron _ of @xmath6 .
[ prop : thr - char - min ] let @xmath424 be the minimum value on the anti - dual polyhedron @xmath425 of @xmath6 of the functional @xmath426 , defined by @xmath427 . then , @xmath428 * proof : * by definition [ def : threshold - char ] @xmath10 is the largest @xmath415 such that , @xmath429 or equivalently it is equal to the reciprocal of the smallest @xmath424 such that , @xmath430 it immediately follows that @xmath431 .
@xmath168 the following proposition is easily deduced from proposition [ prop : thr - char - min ] ( cf .
the proof of proposition [ prop : simetrija ] ) .
[ prop : vazna - simetrija ] let @xmath152 be the group of all permutations of @xmath4 $ ] which keep @xmath6 invariant and let @xmath432 be the associated invariant subspace of @xmath393 . then
, @xmath433 our first application of proposition [ prop : vazna - simetrija ] is to complexes @xmath9}$ ] which are invariant with respect to the full symmetry group @xmath434 .
[ exam : full - symmetry ] if @xmath409\choose \leq k}\subset 2^{[n]}$ ] then @xmath435 .
indeed , the invariant subspace @xmath436 is the diagonal @xmath437 .
the intersection @xmath438 is identified as the interval @xmath439 while the restriction of @xmath440 on @xmath141 is given by the formula @xmath441 .
it follows that @xmath442 and @xmath435 .
the first three ` key examples ' from ( * ? ? ?
* lemma 4.2 ) are easily seen to be linearly realizable , @xmath7-unavoidable ( threshold ) complexes .
the last example in this lemma ( example ( iv ) ) is in general not a linear threshold complex . as an application of proposition [ prop : vazna - simetrija ] ( and in agreement with proposition [ prop : concave-2-linear ] )
we show ( proposition [ prop : intri - iv ] ) that the complex constructed in this example always contains a linearly realizable @xmath7-unavoidable threshold complex as a subcomplex . according to our classification ( definition [ def : simple ] and corollary [ cor : char - simple ] to proposition [ prop : kad - je - jednakost ] ) this
complex is _ intrinsically linear_. [ exam : key-(iv ) ] the key example ( iv ) from _ ( * ? ? ?
* lemma 4.2 ) _ says that if @xmath443 and @xmath84 are integers such that @xmath444 with @xmath445 , then the complex , @xmath446\choose{\leq k}}\cup { [ n+1-(r - s)]\choose{\leq k+1}}\subset 2^{[n+1]}\ ] ] is @xmath7-unavoidable on @xmath447 $ ] .
[ prop : intri - iv ] the complex @xmath448 described in example [ exam : key-(iv ) ] is intrinsically linear , @xmath7-unavoidable complex .
* proof : * let @xmath449 and @xmath450 where @xmath451 and @xmath452 . following the algorithm described in proposition [ prop : vazna - simetrija ]
we calculate @xmath10 where ( in agreement with ( [ eqn : binomi ] ) ) , @xmath453\choose{\leq k}}\cup { [ p]\choose{\leq k+1}}\subset 2^{[n]}.\ ] ] the complex @xmath6 is invariant with respect to the action of the group @xmath454 of all permutations of @xmath4 $ ] which preserve the partition @xmath4 = [ p]\uplus ( [ n]\setminus [ p])$ ] .
it follows that @xmath455 so the restriction of the linear function @xmath440 ( proposition [ prop : vazna - simetrija ] ) to @xmath212 has the form @xmath456 .
the polyhedron @xmath457 is described by specializing the inequalities defining the anti - dual polyhedron ( [ eqn : antidual ] ) to the plane @xmath212 . after some reflection ( assuming @xmath458 ) we observe that the inequalities describing @xmath457 ( as a polyhedral subset of @xmath459 ) are the following , @xmath460 where @xmath461 . the intersection of lines @xmath462 is the point a(1/(k+2 ) , 2/(k+2 ) ) ( figure [ fig : lin - program ] ) .
for the complex @xmath409\choose{\leq k}}\cup { [ p]\choose{\leq k+1}}$ ] . ]
it is not difficult to check that @xmath294 is a feasible point for the polyhedron @xmath457 ( i.e. it satisfies all the inequalities ( [ eqn : reflection ] ) ) . as a consequence
the minimum @xmath424 of the function @xmath440 on @xmath457 satisfies the inequality @xmath463 . from here
we immediately deduce ( proposition [ prop : vazna - simetrija ] ) that @xmath464 .
finally , the condition that @xmath465 for some @xmath389 is equivalent to the initial inequality ( on the left hand side ) , @xmath466 while the last one is precisely the inequality @xmath467 .
it is interesting to check when the minimum @xmath424 of @xmath440 is attained precisely at the point @xmath468 .
for example if @xmath469 then the defining equations for @xmath457 are the inequalities @xmath470 . in this case @xmath471
if and only if ( figure [ fig : lin - program ] ) @xmath472 which is equivalent to the inequality @xmath473 .
the significance of propositions [ prop : thr - char - min ] and [ prop : vazna - simetrija ] ( as demonstrated by proposition [ prop : intri - iv ] ) is that they provide a quite efficient algorithm for calculating @xmath10 for a complex @xmath9}$ ] with a reasonably large group of automorphisms . in light of the implication @xmath474 ( proposition [ prop : vazna - implikacija ] ) ,
they provide a method for a systematic generation of examples of @xmath7-unavoidable ( threshold ) complexes .
for example if @xmath475 one can try this method on the complex , @xmath476\choose
\leq k}\cup { [ q]\choose \leq k+1}\cup { [ p]\choose \leq k+2}\ ] ] or on some other , more general ( ` rank selected ' ) simplicial complex .
a more detailed presentation of computer experiments and other related results will be documented in a separate publication .
in this section we evaluate ( or estimate ) the threshold characteristic of joins of simplicial complexes .
before we formulate more general statements , let us test proposition [ prop : vazna - simetrija ] on some important special cases .
[ exam:2 ] if @xmath477^{\ast n } = [ 2 ] \ast \dots \ast [ 2 ] = \partial ( \diamondsuit_n)$ ] is the boundary complex of the @xmath259-dimensional cross - polytope @xmath478 , then @xmath413 .
indeed , the group of symmetries of @xmath6 is transitive on its vertices so @xmath10 is attained if @xmath317 is the uniform measure where @xmath479 for each @xmath229 $ ] .
the minimal non - simplices @xmath480 are all of the same cardinality and @xmath481 .
this example shows that a complex @xmath6 can have a small @xmath482-characteristic , and at the same time non - trivial homology in the top dimension .
[ exam : k ] generalizing example [ exam:2 ] it is easily checked that @xmath483^{\ast n})= \frac{2}{kn}$ ] for each @xmath484 and @xmath257 .
the case @xmath485 is especially interesting since @xmath28 $ ] ( as a complex on @xmath28 $ ] as a set of vertices ) is self - dual ( minimal @xmath22-unavoidable ) . in light of proposition [ prop : join - min ] @xmath486^{\ast n } ) = n+1 $ ] and @xmath487 for each proper subcomplex @xmath488^{\ast n}$ ] .
the associated @xmath482-invariant is @xmath489^{\ast n } ) = \frac{2}{3n}$ ] .
it follows that in the case @xmath490^{\ast n}$ ] the difference , @xmath491 can be as large as desired .
hence , for @xmath492 we obtain examples of _ strongly non - linear _ @xmath261-unavoidable complexes in the sense of definition [ def : strong - intrinsic ] .
the following proposition allows us to estimate the threshold characteristic of a join of complexes even in the case when the constituent complexes are asymmetrical .
[ prop : join - weights ] let @xmath234 be a family of complexes @xmath493 and let @xmath494 $ ] .
let @xmath495 be the associated join where @xmath244 . if there is a positive probability measure @xmath317 on @xmath212 such that @xmath496 , then there exists some @xmath231 $ ] , and a positive probability measure @xmath2 on @xmath238 , such that @xmath497 .
* proof : * let @xmath498 be the restriction of the measure @xmath317 on @xmath499 and let @xmath500 . since @xmath501 there exists @xmath231 $ ] such that @xmath502
let @xmath503 be the probability measure on @xmath238 induced by @xmath504 . by assumption
we known that for each subset @xmath505 if @xmath506 then @xmath507 .
since @xmath508 we observe that for each @xmath505 if @xmath509 then @xmath510 .
since @xmath511 we have the implication , @xmath512 which proves that @xmath513 . @xmath168 [ cor : join - threshold - ineq ] under the assumptions of proposition [ prop : join - weights ] there is an inequality , @xmath514 * proof : * as a consequence of proposition [ prop : join - weights ] there is an implication , @xmath515 [ cor : join - weights ] assume that @xmath243 is a simplicial complex and let @xmath494 $ ] .
let @xmath516 be the @xmath259-fold join of @xmath6 which is assumed to be a simplicial complex on @xmath517 $ ] .
if there is a positive probability measure @xmath317 on @xmath518 $ ] such that @xmath519 , then there exists a positive probability measure @xmath2 on @xmath212 , such that @xmath520 . as a consequence
there is an implication , @xmath521 and an inequality , @xmath522 in this section we compute the threshold characteristic of complexes which are minimal triangulations of classical projective planes @xmath523 and @xmath29 .
together with the @xmath27-dimensional ` projective plane ' @xmath28=\{1,2,3\}$ ] , these are ( following @xcite ) manifolds which are ` like a projective plane ' in the sense that they allow a morse function with exactly three critical points . recall that all these complexes are self - dual ( minimal @xmath22-unavoidable ) in the sense that @xmath286 if and only if @xmath524 , see @xcite . following the algorithm described in section [ sec : threshold - char ] ( proposition [ prop : vazna - simetrija ] )
we need the information about ( a ) the number and size of orbits of the associated symmetry groups and ( b ) the minimum size of non - simplices in these complexes .
@xmath24 the minimum triangulation of the real projective plane has six elements ( figure [ fig : poly - iko ] ) .
the group of simplicial automorphisms of @xmath24 is vertex transitive .
the size of the smallest non - simplex is @xmath525 .
it follows that @xmath526 .
@xmath26 the minimum triangulation of the complex projective plane has @xmath25 elements .
the size of the smallest non - simplex is @xmath527 .
the group of simplicial automorphisms of @xmath26 is also vertex transitive .
this can be easily deduced from the description of this group given in @xcite ( see also ( * ? ? ?
* section 2 ) ) .
it follows that @xmath528 .
@xmath529 brehm and khnel constructed in @xcite @xmath530-isomorphic simplicial complexes @xmath531 on @xmath532 vertices and conjectured that they triangulate the quaternionic projective plane @xmath29 ( this conjecture was recently confirmed by gorodkov in @xcite ) .
the most symmetric among them is @xmath533 , which is invariant under _ vertex transitive _
action of the group @xmath534 ( see the theorem on page 169 in @xcite ) .
moreover the size of the minimum size non - simplex in @xmath535 is @xmath23 .
from here we deduce that @xmath536 .
the inequality ( [ eqn : vazna - nejednakost ] ) , which provides a fundamental relation between invariants @xmath0 and @xmath10 , can be rewritten as follows , @xmath537 the following proposition clarifies when the inequality in ( [ eqn : fundam - ineq ] ) on the right is actually an equality .
[ prop : kad - je - jednakost ] let @xmath9}$ ] be a simplicial complex
. then @xmath538 if and only if @xmath465 for some probability measure @xmath389 . *
proof : * if @xmath285 and @xmath465 then @xmath539 and in ( [ eqn : vazna - nejednakost ] ) we have the equality @xmath540 .
conversely , if @xmath541 then @xmath542 and , @xmath543 which implies @xmath465 for some probability measure @xmath544 .
@xmath168 the following corollary of proposition [ prop : kad - je - jednakost ] gives a characterization of _ intrinsically linear _ and _ intrinsically non - linear _ complexes introduced in definition [ def : simple ] .
[ cor : char - simple ] a simplicial complex @xmath9}$ ] is intrinsically linear ( respectively intrinsically non - linear ) if and only if @xmath540 ( respectively @xmath545 ) .
[ def : strong - intrinsic ] a simplicial complex @xmath208 is _ strongly non - linear _ if it is both intrinsically non - linear and minimal in the sense of definition [ def : min - unavoidable ] .
the following theorem collects some of the central and the most interesting examples of strongly non - linear unavoidable complexes which can be studied by the methods developed in previous sections .
[ thm : vazna-2 ] let @xmath6 be one of the following complexes : 1 . the @xmath525-element , @xmath27-dimensional complex @xmath28\subset 2^{[3]}$ ] ; 2 . the unique , minimal @xmath23-vertex triangulation @xmath546}$ ] of the real projective plane ; 3 . the minimal @xmath25-vertex triangulation @xmath547}$ ] of the complex projective plane ; 4 . the @xmath532-element triangulation @xmath533 of the quaternionic projective plane @xmath548 @xcite .
let @xmath549 be the join of @xmath259 copies of @xmath6 .
given a complex @xmath550 , let @xmath551 and @xmath552 ( the number of vertices of @xmath35 ) .
the following table summarizes our knowledge about the invariants @xmath553 and @xmath2 of the complex @xmath554 .
@xmath555 & \mathbb{r}p^2_6 & \mathbb{c}p^2_9 & \mathbb{h}p^2_{15 } \\\hline \pi(k^{\ast n } ) & n+1 & n+1 & n+1 & n+1 \\ \hline \rho(k^{\ast n } ) & 2/3n & 1/2n & 4/9n & 6/15n \\
\hline \epsilon(k^{\ast n } ) & \lfloor n/2\rfloor & n & \lfloor 5n/4\rfloor & \lfloor 3n/2\rfloor \\ \hline \nu(k^{\ast n } ) & 3n & 6n & 9n & 15n \\
\hline { \epsilon}/{\nu } & \approx 1/6 & 1/6 & \approx 5/36 & \approx 1/10 \\\hline \end{tabu}\ ] ] * proof : * all complexes @xmath6 listed in the theorem are self - dual @xcite which implies that in all these cases @xmath556 . in follows from the inequality ( [ eqn : pi - join ] ) ( corollary [ cor : pi - join ] ) that @xmath557 . as a consequence of proposition [ prop : join - min ] the complex @xmath549 is minimal @xmath261-unavoidable . from here and corollary [ cor : pi - equality ] we immediately deduce that @xmath558 .
( note that the existence of the partition @xmath559 provides a more direct proof of the inequality @xmath560 . )
the second row in the table ( evaluating the invariant @xmath482 ) is computed by the method developed in section [ sec : threshold - char ] which uses proposition [ prop : vazna - simetrija ] as the key ingredient ( see example [ exam : k ] for a more detailed calculation in the case @xmath490 $ ] ) . indeed ,
if @xmath152 acts transitively on the vertices of the complex @xmath6 then the wreath product @xmath561 acts transitively of the set of vertices of @xmath549 .
in particular this holds for the triangulations of ` projective planes ' @xmath562 and @xmath563 .
so the numbers in the second row are obtained as the quotients of a the size of a minimal non - simplex in @xmath564 ( which is the same as the size of the minimal non - simplex in @xmath6 ) and the number @xmath565 of vertices in @xmath564 . the rest of the table is completed by simple calculation .
@xmath168 the join operation @xmath566 was in previous sections our key tool for constructing intrinsically non - linear complexes . here
we show that it may be applied even to complexes which do not have a large symmetry group with a similar effect .
it turns out that even if the complex @xmath6 is intrinsically linear ( @xmath567 ) than the resulting complex is by a rule intrinsically non - linear , @xmath568 , provided @xmath492 .
let @xmath569}$ ] be the @xmath259-fold join of a complex @xmath208
. then , @xmath570
* proof : * by corollary [ cor : pi - join ] and corollary [ cor : join - weights ] , @xmath571 from the second inequality in ( [ eqn:2-nejednakosti ] ) we deduce , @xmath572 and the inequality ( [ eqn : by - a - rule ] ) follows as an easy consequence . @xmath168
in this section we collect some ideas and facts not directly connected with the main themes and central results of the paper . in section [ sec : classification ]
we briefly describe a classification of unavoidable complexes based on weighted hypergraphs .
section [ sec : ramsey ] establishes a link of unavoidable complexes with ramsey theory , while in concluding remarks we point to some promising future developments . _
` weighted hypergraph ' _ realizable or @xmath573-realizable @xmath7-unavoidable complexes is a class of complexes considerably larger then the class of linearly realizable complexes . here
again we use a weight distribution but the weights may be associated to sets which are not necessarily singletons .
a hypergraph @xmath574}$ ] is by convention an arbitrary , non - empty family of sets of not necessarily the same cardinality .
the @xmath575-uniform hypergraph on @xmath94 $ ] is the sets @xmath94^{(k ) } = { { [ m]}\choose{k } } = \{a\subset [ m ] \mid \vert a\vert = k\}$ ] .
[ def : f - super ] let @xmath576}$ ] and let @xmath577 be a distribution of non - negative weights on @xmath578 . define the associated superadditive measure on @xmath94 $ ] by , @xmath579 if @xmath580 $ ] does not admit a ( sub)partition @xmath581 by elements from @xmath578 we set @xmath582 which is in agreement with @xmath583 . [ def : f - r - unavoidable ] an @xmath7-unavoidable complex @xmath44}$ ] is @xmath573-realizable or ` weighted hypergraph realizable ' if for some hypergraph @xmath574}$ ] , an associated weight distribution @xmath588 , and the corresponding superadditive measure @xmath584 , @xmath589 where @xmath590)$ ] .
the pair @xmath591 is referred to as the @xmath573-realization of the @xmath7-unavoidable complex @xmath6 . [
exam : lin - je - f]suppose that @xmath592\}$ ] is the collection of all one - element subsets ( singletons ) in @xmath94 $ ] .
then for each weight distribution @xmath593 the associated superadditive measure @xmath584 is additive and we recover the class of linearly realizable , @xmath7-unavoidable complexes introduced in section [ sec : lin - realiz ] .
in light of example [ exam : lin - je - f ] it is quite natural to ask for examples of @xmath573-realizable , @xmath7-unavoidable complexes which are _ not _ linearly realizable .
it is quite clear that this class is considerably bigger than the class of linearly realizable complexes .
the following proposition shows that this is true already in the case of @xmath22-unavoidable complexes .
* proof : * let @xmath594}$ ] be the collection of all subsets of @xmath94 $ ] . define the weight distribution @xmath595}\rightarrow \mathbb{r}^+$ ] by , @xmath596 since @xmath6 is self - dual , for each @xmath597 $ ] either @xmath598 or @xmath599 ( but not both ) .
it follows from here that @xmath600 and in turn @xmath601 .
@xmath168 proposition [ prop : nula ] is elementary however it has some important consequences .
the following corollary suggests that there is an increasing hierarchy of @xmath573-realizable , @xmath7-unavoidable complexes . in particular , for a given @xmath7-realizable complex @xmath6 , it may be desirable to find its @xmath591-realization with @xmath578 of the smallest size ( or minimizing some other complexity ) . here
we describe a procedure for generating intrinsically non - linear , @xmath7-unavoidable simplicial complexes based on ramsey s theorem and its relatives .
the method is quite general and links in an unusual way topological combinatorics with ramsey theory .
the following example explains ` in a nutshell ' the central idea of the construction .
we describe an intrinsically non - linear , @xmath22-unavoidable complex @xmath35 on the set @xmath613}\choose{2}}$ ] of all @xmath22-element subsets of @xmath39 $ ] . by definition @xmath614
is in @xmath35 if and only if there is a triangle in the complement of @xmath56 .
it is clear that @xmath35 is a simplicial complex .
moreover , ramsey s theorem @xmath615 implies that @xmath35 is @xmath22-unavoidable ( @xmath616 ) .
let us compute the threshold characteristic @xmath617 of @xmath35 .
since the group of automorphisms of @xmath35 acts transitively on the set @xmath618\choose 2}$ ] it is sufficient to determine the size of the minimal non - simplex in @xmath35
. figure [ fig : ramsey-1 ] on the left displays a graph without triangles which means that the complementary graph ( on the right ) is not in @xmath35 .
the size of this graph is @xmath23 and it is not difficult to see that this is precisely the minimum size of a non - simplex in @xmath35 . as an immediate consequence
we obtain that @xmath619 .
it follows that @xmath35 is intrinsically non - linear since @xmath620 .
note that the ratio @xmath621 is relatively small ( compared to the examples listed in theorem [ thm : vazna-2 ] ) .
suppose that @xmath622 is a graph property which is monotone in the sense that if @xmath623 and @xmath624 then @xmath625 .
for example ` having triangles ' ( or cliques of given size ) is such a property .
if the set of vertices is fixed , we frequently identify a graph with its set of edges .
[ def : stevo-1 ] choose @xmath38 and denote by @xmath626}\choose{2}}$ ] the set of all @xmath22-element subsets of @xmath4 $ ] .
we say that an integer @xmath259 is @xmath627-admissible if for each partition ( coloring ) @xmath628}\choose{2 } } = a_1\uplus a_2\uplus\ldots\uplus a_r\ ] ] of @xmath626}\choose{2}}$ ] into @xmath7 disjoint sets , for some @xmath100 the graph @xmath629 has the property @xmath622 .
for example if @xmath622 is the property that the graph has a clique of size @xmath575 then @xmath259 is @xmath630-admissible if for each coloring ( [ eqn : nebo-1 ] ) there exists a set @xmath631 $ ] of size @xmath575 such that @xmath632 is colored by at most @xmath633 colors ( at least one of the colors is missed ) .
[ prop : stevo - r ] suppose that @xmath259 is @xmath627-admissible .
let @xmath35 be the simplicial complex defined on the set @xmath634}\choose{2}}$ ] ( as the set of vertices ) where @xmath635 if and only if the complement @xmath636}\choose{2}}\setminus s$ ] of @xmath56 has the property @xmath622 .
then the simplicial complex @xmath35 is @xmath7-unavoidable .
* proof : * if ( [ eqn : nebo-1 ] ) is a partition of the set of vertices @xmath626}\choose{2}}$ ] of @xmath35 then at least one of the sets @xmath629 has the property @xmath622 ( definition [ def : stevo-1 ] ) which means that @xmath111 .
@xmath168 [ rem : program]it is expected that the @xmath7-unavoidable complexes described in proposition [ prop : stevo - r ] are often intrinsically non - linear .
indeed , in the opposite case we should be able ( in principle ) to prove ramsey type statements by a simple counting argument ( using a system of non - negative weights @xmath637 on edges @xmath638 ) . both _ threshold complexes _ and
_ @xmath22-unavoidable complexes _ can be directly linked to some classes of simplicial complexes which ( often independently ) emerged and attracted attention of researchers in _ game theory , combinatorial topology , social choice theory , reliability theory _ , _ geometry of moduli spaces of polygonal linkages _ , and other areas . in topology
@xcite they appear as the alexander self - dual complexes , and provide key examples of @xmath259-dimensional complexes non - embeddable in @xmath639 . in social choice theory ( reliability theory ) @xcite they arise as the complexes describing the winning ( losing ) coalitions in game theory ( ` simple games ' of von neumann and morgenstern @xcite ) . in geometry of configuration spaces
@xcite they arise as the complexes of ` short sets ' , characterizing the configuration spaces of polygonal linkages in euclidean spaces .
it may be expected that the invariant @xmath0 and the associated class of @xmath7-unavoidable complexes are as important and interesting for other fields as their @xmath22-unavoidable counterparts .
. chapter 14 in _ handbook of discrete and computational geometry _ , j.e.goodman , j. orourke , eds , chapman & hall / crc 2004 , 305330 .
( new edition under preparation . ) | the partition number @xmath0 of a simplicial complex @xmath1}$ ] is the minimum integer @xmath2 such that for each partition @xmath3 $ ] of @xmath4 $ ] at least one of the sets @xmath5 is in @xmath6 .
a complex @xmath6 is @xmath7-unavoidable if @xmath8 .
motivated by the problems of tverberg - van kampen - flores type , and inspired by the ` constrained method ' of blagojevi , frick , and ziegler @xcite , we study the combinatorics of @xmath7-unavoidable complexes . |
the solar ultraviolet measurements of emitted radiation spectrometer ( sumer ; * ? ? ?
* ) on the esa / nasa _ solar and heliospheric observatory _
( _ soho _ ; * ? ? ?
* ) has been observing the sun with high spectral and spatial resolution since 1996 january .
a considerable number of sumer observations are in the form of long - duration slit - scan `` rasters '' ( spectro - heliograms ) of polar and equatorial coronal hole regions .
these coronal hole rasters have been used to investigate the spatial correspondence between supergranular network cells , visible in the photosphere and chromosphere , and their counterpart emission and plasma flows observed in the solar transition region and lower corona .
the analysis of these sumer rasters has placed particular emphasis on the study of the 770 emission line , identifying and investigating regions of significant plasma outflow ( @xmath05 km s@xmath1 ) in the 600,000 k plasma as a physical link to the origin of solar wind outflow ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in this letter
, we discuss and demonstrate two new and very simple diagnostics of the magnetic environment encompassing a specific set of sumer coronal hole rasters : the relative unbalance of magnetic polarities on the supergranule spatial scale , and the radial distance required to balance local flux concentrations with those of opposite polarity .
we demonstrate that the partitioning ( i.e. , how the doppler image can be broken into regions of blue- and red - shift ) of the 770 doppler shifts and emission in the upper transition region of the coronal hole can be qualitatively understood by imposing a simple model of magnetic flux balance .
further , we observe that the magnitude of the net magnetic unbalance in the coronal hole appears to have a critical influence on the doppler shift patterning , and on the blue - shift in particular .
we focus our discussion on two individual sumer coronal hole rasters from 1999 november 6 and 7 that , when combined , form a synoptic scan of a large equatorial coronal hole ( ech ) in the 1530 to 1555 range .
this spectral range includes four strong emission lines with formation temperatures representative of the upper chromosphere (; 1533.4- 30,000 k ) , transition region (; 1548.2 and 1550.8 - 100,000 k ) and the upper transition region / low solar corona (;
1540.8 in second order - 770.4 in first order - 600,000 k ) . in this letter
, we concentrate only on the doppler patterning of the emission line and its connection to the observed global morphology of the ech region .
a recent paper @xcite has developed an improved method to correct small ( persistent ) electronic imperfections of sumer s detector a , which can cause systematic shifting of the resulting solar uv spectra when acquired in raster mode . by fitting the gaussian profiles of ten accurately measured ( in the laboratory ) emission lines in the 1530 - 1555 range at each spatial row of detector pixels and mapping their locations onto the plane of the detector ,
they are able to build a map of the detector distortion that is caused by the small electronic defects . using this corrective offset to `` co -
align '' the emission lines on each detector image in the raster , they are able to compose a self - consistent wavelength calibration for the entire data set .
the net result of this wavelength calibration is a set of doppler velocities in the emission line to an accuracy of about @xmath21 km / s .
further , following the discussion presented in sect .
4.1 of @xcite , we have chosen to use a rest wavelength of 770.420 for this line , consistent with the laboratory measurement of @xcite .
details of the sumer raster observations of november 6 and 7 are summarized in table tab : one and are pictorially represented in the panels of fig . fig1 .
as discussed in @xcite , we have rotatedzarro / idl / maps.html . ] both raster `` maps '' to their coordinates at the start time of the november 6 raster , as this facilitates co - alignment with the other _ soho _ instruments .
the left panel of fig .
fig1 shows the _ soho _ extreme - ultraviolet imaging telescope ( eit ; * ? ? ?
* ) 195 passband coronal context image from 1999 november 6 taken closest to the start of the first raster .
each eit 195 passband image contains broadband spectral ( multi - thermal ) emission that is strongly peaked at a little over 1.5 mk ( see , e.g. , figs . 8b and 9b of * ? ? ?
the right panel of fig .
fig1 shows the _ soho _ michelson doppler imager ( mdi ; * ? ? ?
* ) line - of - sight magnetogram taken at 12:47 utc , rotated to the time of the eit image at 14:05 utc . as well as
the sumer raster regions in the panels we show the ( estimated ) boundary of the ech , which we will set at an eit iso - intensity contour of 150 data numbers ( dn ) .
we note that the global morphology of the corona hole region does not change over the twenty hours of observation required to build the two sumer rasters and so the step of comparing them to the context coronal images , implied ech boundary , and photospheric magnetogram at the same reference time is assumed to be valid .
figure fig2 shows the coronal hole region of 1999 november 6 as determined by a numerical adaptation of the harvey method for coronal hole identification @xcite .
the automated coronal hole identification tool @xcite , like its counterpart , uses the equivalent width of 10830 line determined from daily synoptic kitt peak vacuum telescope ( kpvt ) spectroheliograms and magnetograms to morphologically isolate regions of reduced equivalent width .
the red- or blue - colored regions indicate the net magnetic polarity of the kpvt coronal hole regions identified by the code as positive or negative respectively .
we note that the three - pronged eit 195 contour ( centered on @xmath3200 , @xmath3400 ) that is clearly not identified as part of the coronal hole from the kpvt boundary is a filament channel and is quite distinct from the coronal hole in other euv wavelengths ( j.b .
gurman - private communication ) .
other than the filament channel , we see that the heuristic eit coronal hole boundary and that numerically determined from kpvt observations agree well . we will therefore continue to use the simple eit boundary of the ech in the discussion that follows without any loss of generality . the top row of panels in fig .
fig3 show the intensity ( left ) and doppler velocity ( right ) maps from the regions rastered by sumer on november 6 ( upper ) and
november 7 ( lower ) , respectively .
again , we show the eit 195 150 dn iso - intensity contour for reference as the ech boundary .
we notice that while the line intensity raster map from november 6 ( north - east portion ) is almost uniformly dark in the coronal hole , with emission rising at the boundary , the november 7 raster map shows a `` break '' in the emission structure ( centered on 100,@xmath3250 - running from seven to two oclock ) , where the intensity increases by a factor of two , to the quiet - sun level ( i.e. , that outside the hole boundary ) and drops again in the lower right ( south - west ) portion of the coronal hole .
very close inspection of the eit image shows some ( very low ) enhanced emission contrast in the break region of the ech , but it does not reflect the factor of two change in the emission .
precisely the same spatial break , or partition , exists in the doppler shift rasters shown in the top right panel of the figure .
the raster from november 6 shows a large , complex pattern of blue - shifts inside the eit contour , while that of november 7 shows a more divided picture with the break region showing a mottled , but net - postive ( red ) , doppler patterning that is reminiscent of the quiet - sun data shown to the east of the eit contour on the sixth and those published in the literature ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in the south - west portion of the november 7 raster
the doppler - shift pattern again resembles the dark blue - shift structure of the north - eastern coronal hole region of the previous day .
@xcite queried the strange emission and doppler velocity partitioning of the november 7 1999 raster within the uniformly dark euv emission that forms the body of the coronal hole .
it is the ( qualitative ) understanding of the partitioning of the emission and velocity structure and their connection to the global morphology of the ech that provides the motivation for this letter . the bottom row of panels in fig .
fig3 show the 20 mm radially averaged line - of - sight magnetic field ( @xmath4 ; left ) and the `` magnetic range of influence '' ( mroi ; right ) .
the first is a relatively simple diagnostic which highlights the net magnetic field variation at the spatial scale of supergranular cells ( 15 - 30 mm ; * ? ? ?
it shows that the bulk of the region in the north - east of the coronal hole has a dominant and non - negligible net positive magnetic flux , as does the strongly blue - shifted region in the south - west of the november 7 raster , while the central `` break '' region is near - zero , mixed magnetic polarity and threaded by the neutral line , or where @xmath4 is identically zero .
the net positive magnetic polarity of the entire coronal hole can be easily compared to the kpvt coronal hole boundary determination in fig .
the mroi is another simple realization of the magnetic environment and reflects the ( radial ) distance required to balance the integrated magnetic field contained in a 25 mm@xmath5 circular area centered at any pixel in the mdi magnetogram .
it is computed by advancing radially outward from the central region one pixel at a time and summing the contained field to seek the distance ( the number of mdi pixels times their spatial scale , @xmath61.9 mm ) where the total summed field returns to zero ( @xmath71 g ) .
while the mroi contains _ no _ directional information , it can be thought of as a measure of magnetic partitioning of the plasma into open and closed regions .
when the mroi is small , the magnetic field is `` closed '' locally , while , if the mroi is large , the magnetic field at the center is largely unbalanced and the magnetic environment is effectively `` open '' .
considering again our choice of 770.420 as the rest wavelength , we point to the line intensity pattern of both rasters and its connection to the global magnetic environment of the region surveyed . while we know that the integrated line intensity is completely insensitive to the choice of rest wavelength
, the observed doppler patterning _ must _ reflect the variation of the intensity and magnetic environment because they are physically coupled .
we have seen that the darkest regions of emission are predominantly blue - shifted and correlate to regions of large mroi and unbalanced magnetic fields .
conversely , the brighter regions are correlated to small mroi and balanced magnetic fields .
the spatially averaged doppler velocity in the magnetically closed regions is slightly red - shifted ( 1 - 3 km / s * ? ? ?
* ) , which , when combined with the @xmath21 km / s accuracy of the wavelength calibration , point to a rest wavelength of 770.420 and an error in the range of @xmath21 - 4 m .
we argue that the results presented in this letter add weight to our choice of 770.420 and are entirely consistent with the statistical justification of @xcite .
we should declare that while this remotely sensed inference may be the most self - consistent measurement that we have from the solar plasma , it in no way replaces the need for new theoretical calculations or ( technically difficult ) controlled laboratory experiments to determine what is possibly the most physically important , and contentious , rest wavelength in the solar ultraviolet spectrum .
a detailed investigation into the sumer measurements of the rest wavelength is beyond the scope of this letter and will be revisited in a forthcoming paper ( davey et al .
2006 , in preparation ) .
from the top row of panels of fig .
fig3 , we are immediately drawn to three features : the locations of strong ( @xmath0 5 km s@xmath1 ) blue shift in the north - east and south - west of the eit defined coronal hole region and the brighter emission region in its central belt . while the quantitative comparison of the doppler shift in the sumer rasters with the @xmath4 and mroi maps does not yield direct physical correlations , we can clearly see that the regions of strongest blue - shift are co - spatial with regions of large - scale , locally unbalanced , net positive ( @xmath010 g ) magnetic field .
the mixed velocity regions are those where the @xmath4 and mroi are both small , @xmath60 g and @xmath8 20 mm , respectively .
we notice that the portion of the coronal hole where the emission is enhanced is co - spatial with the region where the mroi is @xmath820 mm .
we also see that this region is where the magnetic field strength averaged over supergranular spatial scales is , on balance , zero and is threaded by the neutral line . this portion of the coronal hole displays doppler velocity patterning that is ( qualitatively ) closest to that of the quiet sun ( e.g. , fig
. 12 of * ? ? ?
* ) , or at least to the raster portions that are outside the coronal hole .
we should note that there is no obvious enhancement in the eit emission in the same location ( see fig .
fig1 ) . while it might be magnetically open in the hot solar corona ( the uniformly dark eit emission )
, it would appear that the region is locally closed , at least to the formation height of in the transition region .
the large - scale complexity of the emission and doppler velocity structure points to a considerably more complex coronal hole boundary than might simply be implied considering _ only _ the eit 195 emission , or the kpvt boundary for that matter .
this quandary does raise an important , interesting and timely question to the community : how do we observationally determine the boundary of a coronal hole in a robust manner ?
we have observed that the doppler partitioning and simple diagnostics of the magnetic environment imply that there are transition region structures on supergranular length scales ( or larger ) that are topologically closed ( small @xmath4 and mroi ) or open ( larger @xmath4 and mroi ) . in general , the closed regions show enhanced line emission while the open regions show decreased emission and strong blue - shifts .
we have also observed that the emission structure in the 1.5 mk corona does not always reflect that of the 600,000 k transition region .
this strange feature implies the existence of `` nesting '' in the different scales of magnetic topology , such that the lower region can be almost entirely closed , while that above is open .
while space limits any explanation of how the apparent nesting of the plasma arises , we can suggest one simple situation that could produce such an effect : two distinct coronal holes are present ( delimited by the ne and sw doppler - shift regions ) , with the magnetically closed region in the transition region underneath the place where the divergent magnetic field lines from each merge in the corona .
however , we note that many models of varying complexity can satisfy the global aspects of these observations ; discussion is therefore left to later work . in summary ,
the apparently different plasma loading , heating and acceleration that occurs between the open and closed magnetic environments in the transition region ( and possibly into the hotter corona ) is a subject requiring a detailed investigation of the local and global magnetic topologies . as such
, it can not be dealt with in this short letter .
however , based on the observational evidence , we can speculate that the magnetic environment leads to a preferential energy balance in the transition region plasma ; the magnetically closed regions are dominated by plasma heating ( enhanced thermal emission ) while that of the open regions would appear to be dominated by kinetic energy ( enhanced plasma outflow ) .
further , we can infer that , if the blue shifts in open field regions are indeed the origins of solar wind outflow ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , it is possible that the presence of the unbalanced magnetic fields enables the plasma to create , load mass , heat and sustain the lowest tributaries of the solar wind through a magneto - convection driven mechanism like `` magnetic exchange reconnection '' @xcite .
this is a subject of ongoing research by the authors ( mcintosh et al .
2006 , in preparation ) .
we would like to thank the anonymous referee , werner curdt , stein vidar haugan , john mariska , klaus wilhelm and meredith wills - davey for carefully reading , commenting on and enhancing the manuscript .
similarly , we would like to thank carl henney for discussing his method of automatic coronal hole boundary determination .
this material is based upon work carried out at the southwest research institute that is supported in part by the national aeronautics and space administration under grants issued under the living with a star and sun - earth connection guest investigator programs .
specifically grants nag5 - 13450 and nag5 - 11594 to dmh and nng05gm75 g to swm .
the sumer project is financially supported by dlr , cnes , nasa and the esa prodex program ( swiss contribution ) .
sumer is part of soho , the _ solar and heliospheric observatory _ , of esa and nasa .
cccccc observation & observation & & & & solar y + start & end & & & & ( ) + 06-nov-99 14:00:13 & 07-nov-99 00:15:50 & 10.260 & -535 & 199 & 45 + 07-nov-99 00:35:23 & 07-nov-99 10:16:18 & 9.682 & -4 & 708 & -249 + | we present a novel investigation into the relationship between simple estimates of magnetic flux balance and the doppler velocity partitioning of a large equatorial coronal hole observed by the solar ultraviolet measurements of emitted radiation spectrometer ( sumer ) on the _ solar and heliospheric observatory _ ( soho ) in november 1999 .
we demonstrate that a considerable fraction of the large scale doppler velocity pattern in the coronal hole can be qualitatively described by simple measures of the local magnetic field conditions , i.e. , the relative unbalance of magnetic polarities and the radial distance required to balance local flux concentrations with those of opposite polarity . |
the model we use is based on the gilad et al .
vegetation model@xcite , which captures three different mechanisms of vegetation pattern formation@xcite .
however , applying the model to the nfc ecosystem ( sandy soil , confined root zones ) results in a simplified model describing the dynamics of aboveground biomass ( @xmath0 ) and soil - water ( @xmath1 ) areal densities .
the simplified model captures a single pattern - forming feedback associated with the high rate of water uptake by the perennial grasses and the fast soil - water diffusion ( relative to biomass expansion ) in sandy soils .
this mechanism leads to higher soil - water content in the bare fairy circles as compared with the vegetation matrix@xcite , in agreement with reported observations@xcite .
the model equations read : @xmath2 in the biomass equation , @xmath3 is the biomass growth rate coefficient , @xmath4 is the maximal standing biomass , @xmath5 is a measure for the root - to - shoot ratio , @xmath6 is the mortality rate , and @xmath7 represents the seed - dispersal or clonal growth rate . in the soil - water equation
, @xmath8 is the precipitation rate , @xmath9 is the evaporation rate , @xmath10 is a dimensionless factor representing a reduction of the evaporation rate due to shading , @xmath11 is the water - uptake rate coefficient , and @xmath12 is the effective soil - water diffusivity in the lateral ( @xmath13 ) directions .
the model parameters used in this study are different than those used by getzin et al .
@xcite , and were estimated from published data .
we refer the reader to the si appendix for more details on the derivation of the model and to the methods and materials section for details on the parameter estimations and values .
we first study the stationary solutions , in one spatial dimension ( 1d ) , along the precipitation axis .
fig.1a shows uniform solutions representing bare soil and uniform vegetation and periodic solutions representing gap patterns . within the bistability range of uniform vegetation and gap patterns
, there is a subrange with many stable hybrid states , consisting of confined domains of uniform vegetation in an otherwise periodic vegetation pattern ( and vice versa ) , ( inset b ) .
the figure further shows 1d spatial profiles of a periodic gap pattern ( panel c ) and of a hybrid state that describes a periodic pattern with one missing gap ( panel d ) .
the corresponding 2d patterns are shown in panels e and f , and similar fairy circle patterns ( obtained from satellite images ) are shown in panels g and h. similarly to other models of dryland vegetation @xcite , the model also exhibits spotted and striped patterns ( see the si appendix ) , for lower values of precipitation .
norm of the biomass versus the precipitation rate .
the steady states shown include two uniform solutions of bare soil and uniform vegetation ( black ) , two periodic solutions of gap patterns with different wavelengths ( red and green ) spanning the bistability range of uniform vegetation and patterned states , and hybrid solutions representing uniform vegetation with an increasing ( as the branch snakes down ) number of gaps ( blue ) .
the solid ( dashed ) lines in the diagrams represent stable ( unstable ) solutions . *
b : * blowup of the hybrid - state range .
* c : * the periodic solution corresponding to the red branch in * a*. * d : * a hybrid solution corresponding to the lowest blue branch in * b * ( a single missing gap ) obtained with @xmath14 $ ] . * e , f : * the 2d versions of the solutions shown in * c , d*. * g , h : * corresponding fairy circle patterns , obtained from the 2013 satellite image . ]
next , we use the model , along with high - resolution satellite images of the namibrand area from 2004 to 2013 , to suggest that fairy circle `` birth '' and `` death '' processes ( appearance and disappearance of gaps)@xcite correspond to transitions between hybrid states .
figure 2a shows the birth of a fairy circle following a drought in 2007 , while fig .
2c shows the death of a fairy circle following a spate ( a period of excess precipitation ) in 2008 .
figure 2b and 2d show the same behaviors in model simulations , starting with initial conditions that mimic the initial fairy circle patterns in 2004 and applying precipitation downshift and upshift to mimic the drought and spate .
the precipitation shifts were chosen to be strong enough to take the system out of the subrange of hybrid states ( in opposite directions ) , thus inducing hybrid - state transitions .
$ ] /@xmath15 $ ] ) , after which the precipitation was set back to normal ( @xmath14 $ ] ) .
the time shown at the top of the satellite image ( simulation ) denotes the year it was captured ( number of years elapsed from the beginning of the drought or the spate ) .
the time intervals between the simulation snapshots were chosen to emphasize the short - time - scale changes in gap size .
all panels show the dynamics of a region of @xmath16\times30[m]$ ] . ] in addition to the appearance and disappearance of gaps , fig .
2 also shows changes in gap sizes ; a moderate drought expands the gaps , while a spate contracts the gaps . to gain a deeper understanding of these dynamics , we correlated fairy circle data that were extracted from satellite images with rainfall data from several meteorological stations in the namibrand region , and compared the observed correlations with those predicted by the model . figure 3 shows the cross - correlations between the accumulated rainfall over different time periods ( integration time ) and the average fairy circle size ( panel a ) and the fairy circle number ( panel b ) .
analogous cross - correlations using model simulations ( panels c and d ) show a qualitative agreement .
the simulation results are based on averaging over 100 regions and long time series , and therefore , they show smooth curves .
panels e and f show the same quantities for the model simulations but averaged over fewer regions and using shorter time series .
the curves in these panels resemble the curves for the real data ( panels a and b ) .
the stronger correlations in the simulations ( compared with those in the field ) between the gap size and the precipitation is due to the fact that the simulated system is homogeneous , while the field is heterogeneous and also because the simulated precipitation is not intermittent as it is in the field .
the results highlight the different time scales for changes in the size and in the number of fairy circles . for fairy circle size ,
the correlation magnitude is maximal for a relatively short integration time ( of a few years ) , while for fairy circle number , the correlation magnitude does not decrease with time throughout the time range of available data .
this suggests that the timescale for changes in the fairy circle size is much shorter than the time scale for changes in the fairy circle number .
we turn now to the question of gradual regime shifts and ask whether climate variability in the form of repeated short droughts can induce a gradual regime shift . to this end , we simulated the model equations , starting with a uniform vegetation state that is locally disturbed to form a few gaps , and using precipitation downshifts that take the system outside the hybrid - state subrange periodically in time .
4a shows , such a precipitation regime drives the system towards a nearly periodic patterned state ( rightmost panel ) by successive shifts to hybrid states of larger gap numbers .
the opposite transition is shown in fig .
4b , in which a series of short spates drives the system from a nearly periodic gap pattern , through a multitude of hybrid states , to a uniform vegetation state .
systems initially in a uniform vegetation state ( see the si appendix ) or a perfectly periodic patterned state ( fig .
4c ) , exposed to the same series of droughts or spates , show only minor fluctuations in the biomass density and remain practically unchanged .
$ ] , followed by 15 years of @xmath14 $ ] . *
b : * the reversed shift , namely a gradual shift from a nearly periodic pattern to uniform vegetation , driven by one - year spates with @xmath17 $ ] followed by 15 years with @xmath14 $ ] . the droughts / spates take the system out of the hybrid - state range ( see fig .
1a , b ) and induce hybrid - state transitions in the course of which gaps appear / disappear , leading to a gradual regime shift .
* c : * when the initial pattern is strictly periodic , no regime shift takes place even though the system is subjected to the same rainfall regime as in * b*. the simulated domains are @xmath18\times56 [ m]$ ] . ]
we regard the model results described above and their agreement with field observations as compelling evidence for the view of fairy circles as a pattern - formation phenomenon and for the existence of hybrid states in the nfc ecosystem .
this evidence is based on the following grounds : _ spatial properties _ the model , adjusted to the particular context of the nfc ecosystem , predicts gap patterns that resemble the spatial distribution of fairy circles in the field , as fig .
moreover , the model predictions agree with recent field observations of higher soil - water content within the fairy circles , relative to vegetated domains@xcite .
_ dynamics _ fairy - circle birth and death processes can be reproduced with model simulations by temporal escapes outside the existence range of hybrid states ( fig .
2 ) , supporting the interpretation of these processes as hybrid - state transitions . _
correlations with precipitation _
significant correlations between the rainfall and the number and size of fairy circles indicate the importance of biomass - water interactions in the dynamics of the nfc ecosystem . these correlations and the time scales associated with them ,
short for size variations and long for number variations , are in agreement with the model predictions ( fig .
the time - scale difference can be understood using pattern formation theory ; gap - size changes are related to the convergence to a steady state , while gap - number changes are related to the dynamics of fronts that separate domains of uniform vegetation and gap patterns@xcite , structural and wavelength changes of the patterned state , and other nonlinear processes .
this evidence does not rule out alternative hypotheses@xcite , but these hypotheses can not be considered as satisfactory explanations of the fairy circle phenomenon unless they uncover the local feedbacks that are responsible for the emergence of large - scale order and account for the correlation of fairy circle dynamics with rainfall variability .
the model suggested here requires additional evidence in order to fully uncover the origin and dynamics of the nfc .
first , measurements of the soil - water diffusivity and the dependence of the soil - water extraction rate on the vegetation density are required in order to support the mechanism suggested in our model .
second , the precipitation fluctuations in the field are very large and , according to the simplified model , are expected to drive the system towards one of the uniform states . we believe that accounting for the spatial heterogeneity of the ecosystem , i. e. , topography , soil heterogeneity and other factors , would result in more robust patterns in the model .
third , the exact biological origin of hybrid states in the nfc system remains unclear .
controlled field experiments of the fcs dynamics , combined with the mathematical framework suggested here , will be able to provide conclusive evidence for the hypothesis proposed here and reveal the relevant biological mechanisms .
in particular , experiments in areas with different soil textures , topography , and plant species , together with manipulations of the precipitation , are expected to better clarify the relative importance of specific plant physiology and ecosystem characteristics in shaping the observed nfc dynamics .
the results presented here are not limited to the nfc ecosystem and may be observed in other pattern - forming systems .
the appearance of hybrid states in systems exhibiting a bistability of uniform and patterned states implies that transitions , from uniform vegetation to a gap pattern and vice versa , may take place through successive hybrid - state transitions driven by periodic escapes outside the existence range of hybrid states .
dynamics of this type allow for spatially extended ecosystems to exhibit gradual rather than abrupt regime shifts .
vegetation pattern formation in this simplified model results from a finite wavenumber ( turing ) instability of the uniform vegetation state .
this instability requires strong water uptake , quantified by the parameters @xmath11 and @xmath5 , and fast soil - water transport , characterized by @xmath12 , relative to biomass dispersal , quantified by @xmath7@xcite .
the high hydraulic conductivity , typical of sandy soils , acts in favor of this condition , but field observations point towards the possible role of an additional factor the large underground termite populations found in the nfc ecosystem@xcite .
the underground network of channels that the termites form may increase soil - moisture transport in the lateral directions , leading to higher effective values of @xmath12 . studying this possible termite effect
is not only significant for understanding fairy circle formation , but also for understanding regime shifts , as the presence of termites may affect the existence and stability ranges of the uniform and patterned vegetation states , as well as those of the hybrid states .
the bifurcation diagram in fig .
1 a , b was calculated using a numerical continuation method ( auto software ) .
the dynamics of the system was simulated using the pseudo - spectral method with periodic boundary conditions in matlab , in either one or two dimensions .
the simulations in fig .
2 were made using initial conditions of steady states of the system , each derived from a region of the 2004 satellite image .
this was done by taking the relevant region in the original image , segmenting it into vegetation and bare soil areas , and setting the values of the biomass and soil - water areal densities to those corresponding to the uniform solutions ( uniform vegetation and bare soil , respectively ) . the system was then integrated forward in time , with a constant value of precipitation rate , @xmath8 , until a steady state was reached .
in all model simulations , we used the following parameter values : @xmath19 $ ] , @xmath20 $ ] , @xmath21 $ ] , @xmath22 $ ] , @xmath23 $ ] , @xmath24 $ ] , @xmath25 , @xmath26 $ ] , @xmath27 $ ] .
these values were either taken from data published in the literature or estimated using such data , as described below .
the precipitation rate , @xmath8 , was varied within the range @xmath28 $ ] , which is similar to the range of annual rainfall in the nfc ecosystem .
the maximum standing biomass , @xmath4 , was estimated based on maximal biomass densities reported for semi - arid savannah grasses@xcite .
the mortality rate , @xmath6 , was estimated based on the mortality data of _ stipagrostis uniplumis _ as a representative example of an abundant perennial grass in the nfc ecosystem@xcite .
the reported mortality rate was measured under normal conditions ( i.e. , including all the growth factors ) .
the mortality rate in the model represents the natural mortality alone , and therefore , we used a higher mortality rate .
the biomass diffusion parameter , @xmath7 , was estimated based on published clonal growth and seed dispersal data@xcite .
the value for the shading factor , @xmath10 , was inferred from data on reduced evaporation by tree canopies@xcite .
no relevant literature was found for the root - to - shoot parameter , @xmath5 , for the perennial grass ; the chosen value is similar to that used in an earlier study@xcite for woody dryland species .
the evaporation rate , @xmath9 , was estimated using the relation @xmath29 for a uniform bare soil state , where the soil - water density , @xmath1 , was determined from water - content sampling data inside a fairy circle@xcite .
the water - uptake rate coefficient , @xmath11 , was inferred from differences in the levels of soil water within and outside the fairy circles , and from biomass measurements@xcite .
the biomass growth parameter , @xmath3 , was determined from estimates of the ratio of transpiration to dry biomass for tropical grass@xcite and the previously estimated value of @xmath11 .
finally , the value of the soil - water diffusivity , @xmath12 , was estimated using the van genuchten model for water retention curves @xcite . using parameters corresponding to sandy soil@xcite and the values of soil - water content that correspond to those found in the field@xcite
, we estimate @xmath12 to be in the range @xmath30 $ ] .
we used a value that slightly exceeds the upper limit of this range to account for the possible enhancement of lateral water transport due to the network of underground channels formed by termite populations@xcite .
we emphasize that the model results reported here are robust and can be obtained with many different sets of parameters ( see the si appendix ( sensitivity analysis ) ) .
we analyzed high resolution satellite images taken over the period of 2004 - 2013 in the nambirand nature reserve .
the snapshots shown in figs . 1 and 2 were processed from four images , captured in 2004 , 2008 , 2010 and 2013 , whose spatial resolution is as high as @xmath31 $ ] .
the rgb data shown were derived using pan - sharpening with erdas software when rgb was available ( 2004 , 2010 and 2013 ) , and false - color was used for the 2008 image .
the data used for fig .
3 were taken from monthly precipitation data and from measurements of fairy circles features using satellite images .
the precipitation data were derived from interpolations of both the readings of meteorological stations within the nambirand nature reserve ( for the years 2000 - 2012 ) and of the noaa data center ( for the years 1990 - 1999 ) for two locations in namibia .
the numbers and sizes of the fairy circles were estimated from eight satellite images , the four previously described ( from 2004 , 2008 , 2010 and 2013 ) , and four others ( from 2005 , 2009 , 2011 and 2012 ) taken from google earth data . from each of the eight images , four regions of @xmath32x180[m]$ ] were chosen and analyzed . a proprietary segmentation algorithm , involving morphological operations and thresholding with a clahe algorithm , was used to identify each fairy circle .
the identification results were then reviewed manually to minimize errors caused by the automatic process , and in total , about 800 fairy circles were identified and measured for each time point .
the cross - correlation measures shown in fig .
3 were calculated by comparing the accumulated precipitation , @xmath33 , with either the number of fairy circles , @xmath34 , or the average size of a fairy circle , @xmath35 .
the accumulated precipitation is defined as : @xmath36 here , @xmath37 is the precipitation at time @xmath38 , @xmath39 is the lag time and @xmath40 is the integration time , with all times given in number of years . in order to fit the rainy season in the field
, we considered the annual precipitation between november 1 and the following october 31 .
the lag time , @xmath39 , was set to zero ( i.e. , considering the accumulated rain during a period @xmath40 ending at the end of the previous year ) for the fairy circle size and @xmath41 for the fairy circle number ; these values yield the most significant correlations between the precipitation and the fairy circle dynamics .
the cross - correlation for the fairy circle number @xmath42 is defined as : @xmath43 @xmath44 stands for the average of @xmath45 ( @xmath46 ) , and @xmath9 is the number of time points , with the value of @xmath47 for the observation data , and either @xmath47 or @xmath48 for the simulation data .
the cross - correlation for the fairy circle size , @xmath49 , was calculated in the same manner .
we wish to thank nils odendaal , the chief executive officer of the namibrand nature reserve , for providing us the precipitation measurements of various meteorological stations within the reserve , and we also thank elad eizner and noa levi - ohana for their help with the segmentation and analysis of the satellite data .
the research leading to these results has received funding from the european union seventh framework programme ( fp7/2007 - 2013 ) under grant number [ 293825 ] , and from the israel science foundation under grant number 305/13 .
we present the derivation of the simplified model that we use and analyze in the main text .
our starting point is the vegetation model of gilad et al .
@xcite that includes three pattern - forming feedbacks . using the characteristics of the fairy circles ecosystem
, we explain the assumptions and approximations that allow us to replace the integrals of the root - augmentation feedback with simpler algebraic terms and to decouple the aboveground water dynamics from the soil - water and vegetation dynamics . the dryland vegetation model of gilad et al .
@xcite describes the coupled dynamics of the areal densities of vegetation biomass ( _ b _ ) , soil water ( _ w _ ) and surface water ( _ h _ ) , all having the dimension of mass per unit area .
the dynamics is described with a temporal resolution that is smaller than the typical time scale for changes in the biomass ( dictated by the biomass growth rate , mortality and dispersion rate ) but is large enough to allow averaging the intermittent nature of the precipitation .
the spatial scale resolution is larger than the scale of a single plant ( in order to allow the description of a continuous vegetation density ) but smaller than the typical patch size . restricting our interest to flat terrains
, the model reads : [ eq : modeldm ] @xmath50 where @xmath51 \label{eq : g } \\
i & = a \frac{b(\textbf{x},t)+qf}{b(\textbf{x},t)+q } \end{aligned}\ ] ] the terms @xmath52 and @xmath53 involve integration over a root kernel ( @xmath54 ) that represents the spatial extent of the root zone in the lateral directions .
applying the model to perennial grasses , e.g. _ stipagrostis ciliata _
( common in the fairy circle ecosystem ) , which form patterns with a typical length scale of @xmath55 @xcite and have a root girth of approximately @xmath56 @xcite , allows us to assume that the kernel function is much narrower than the biomass and soil - water distributions . under this condition , we can approximate the kernel by a dirac delta function .
formally , this is done by taking the limit @xmath57 in eq .
[ eq : g ] , where @xmath58 represents the lateral root - zone size of a seedling . following this approximation
, we may replace the integrals of eqs . by the algebraic forms : @xmath59 the fairy circle ecosystem consists of sandy soil .
this soil type is characterized by a high rate of surface water infiltration that is comparable to the infiltration rate in vegetated soil . to account for the absence of a significant infiltration contrast between bare and vegetated soil , we set @xmath60 .
the infiltration rate then becomes a constant , @xmath61 , independent of the biomass @xmath0 , and the equation for the surface water variable @xmath62 decouples from those for @xmath0 and @xmath1 .
this equation has a single stationary uniform solution , @xmath63 , which is always linearly stable .
since @xmath62 is the fastest variable , we can assume that , on the much slower time scales over which @xmath0 and @xmath1 significantly change , it has already equilibrated at @xmath64 .
inserting the solution @xmath65 into the equation for @xmath1 , we obtain the two - variable model : [ eq : modeldmvsg ] @xmath66 in the biomass equation , @xmath3 is the biomass growth rate coefficient , @xmath4 is the maximal standing biomass , @xmath5 is a measure for the root - to - shoot ratio , @xmath6 is the mortality rate , and @xmath7 represents the seed dispersal or clonal growth rate . in the soil - water equation
, @xmath8 is the precipitation rate , @xmath9 is the evaporation rate , @xmath10 is a dimensionless factor representing a reduction of the evaporation rate due to shading , @xmath11 is the water - uptake rate coefficient , and @xmath12 is the effective soil - water diffusivity in the lateral ( @xmath13 ) directions , assumed to be a constant , independent of the state variables , that represents linear diffusion@xcite .
we refer the reader to an earlier publication@xcite for additional information about the original model .
the model considered here ( eqs . ) predicts , in addition to the states shown and discussed in the main paper , the appearance of spotted and patterned states under low precipitation rates as shown in fig .
[ fig : patternslowp ] .
in order to test the sensitivity of the results ( presented in the main text ) to the set of parameters , we first write the model in a dimensionless form . thereby , we reduce the number of parameters and simplify the analysis . .
relative change in the existence range of the hybrid states ( snaking range ) , due to varied parameters compared to those used in the main text . in each case , one parameter is changed by either @xmath67 or @xmath68 , and the relative existence range is shown . [ cols="^,^,^",options="header " , ] , for which stable hybrid states exist , as a function of either @xmath12 ( left ) or @xmath5 ( right ) . ]
was changed to @xmath69 [ @xmath70/@xmath71 , while on the right panel , @xmath5 was changed to @xmath72 [ @xmath70/@xmath73 .
uniform states ( bare soil and uniform vegetation ) are shown in black , two periodic states are shown in red and green , and the hybrid states are shown in blue . ]
equations describe the dynamics of the aboveground biomass ( @xmath0 ) and the soil water ( @xmath1 ) in a dimensional form .
translation of the model equations to a dimensionless form is achieved by rescaling the state variables @xmath74 and the space and time coordinates as follows : @xmath75 in terms of these dimensionless quantities , the model reads : @xmath76 @xmath77 the dimensionless parameters here are related to their dimensional counterparts by the following relations : @xmath78 we are thus left with only six parameters , for which we use @xmath79 as the main bifurcation parameter , and as such , its value changes throughout our analysis .
the basis for the main results is the existence of stable hybrid states ; therefore , we tested the conditions under which they occur and their existence range ( range of precipitation rate , @xmath79 ) . we changed the other five parameters , one at a time , within @xmath80 of their value .
we found that in all cases , the hybrid states exist , with their existence range slightly changing .
the relative changes in their existence range are shown in table i. as can be seen , the parameter @xmath81 has the strongest effect on the snaking range . the parameters @xmath82 and @xmath83 control the pattern - forming feedback in our model @xcite .
therefore , we present a more thorough analysis of their effect on the snaking range . in fig .
[ fig : eredw ] , we show the snaking range as a function of either @xmath12 ( left panel ) or @xmath5 ( right panel ) .
[ fig : bifexamples ] shows two examples of the bifurcation diagram for a relatively big change in one of these parameters .
in addition to the sensitivity of the snaking range to the parameters the size of the gaps and the distance between neighboring gaps are also affected by the parameters .
the histogram of the fc sizes , in the regions for which the satellite images were analyzed , is shown in fig .
[ fig : fcsizehist ] .
( left panel ) and @xmath5 ( right panel ) as predicted by the model for constant @xmath14 $ ] . ] in fig . [
fig : modelsize ] we show the average size of the fcs and the average distance between them as predicted by our model for constant precipitation rate of @xmath14 $ ] and for a range of the feedback control parameters @xmath84 and @xmath5 .
it is shown that there is a range of values of these parameters that correspond to the fcs size observed in the field .
the range of possible values for these parameters , which have not been measured directly in the field , may be estimated from these graphs .
in the main paper , we showed that the response of vegetation to a series of droughts or spates depends on the initial condition . to complement the information we present in fig .
[ fig : unresponse ] the response of a uniform vegetation to a series of droughts . | large responses of ecosystems to small changes in the conditions regime shifts are of great interest and importance . in spatially extended ecosystems , these shifts may be local or global . using empirical data and mathematical modeling , we investigated the dynamics of the namibian fairy circle ecosystem as a case study of regime shifts in a pattern - forming ecosystem .
our results provide new support , based on the dynamics of the ecosystem , for the view of fairy circles as a self - organization phenomenon driven by water - vegetation interactions .
the study further suggests that fairy circle birth and death processes correspond to spatially confined transitions between alternative stable states .
cascades of such transitions , possible in various pattern - forming systems , result in gradual rather than abrupt regime shifts .
the response of ecosystems to climate variability and anthropogenic disturbances is a fundamental aspect of ecology .
much attention has been devoted recently to large responses of ecosystems to small environmental changes or disturbances .
such responses , often termed `` catastrophic regime shifts , '' are conceived of as abrupt transitions between two alternative stable states that occur uniformly across the ecosystem@xcite .
spatially extended systems can respond in different ways to varying conditions , including pattern formation@xcite and spatially confined transitions to alternative stable states@xcite .
when one of the alternative stable states is spatially patterned , a multitude of additional stable states can appear , each a mosaic of fixed domains that alternate between the uniform and the patterned states @xcite .
the existence of such hybrid states can strongly affect the dynamics of state transitions in fluctuating environments@xcite , and may lead to gradual rather than abrupt shifts@xcite .
this prediction of pattern formation theory has never been tested either in a real ecosystem or in a model describing the dynamics of a specific ecosystem .
pattern formation is widespread in natural ecosystems@xcite , but good case models for studying gradual regime shifts are not abundant .
an outstanding candidate is the namibian fairy circle ( nfc ) ecosystem , which is fairly homogeneous and undisturbed , and recently has been the subject of intense research @xcite .
the nfc ecosystem consists of a uniform matrix of perennial grass , punctured by circular gaps of sandy bare soil the fairy circles that on landscape scales , form nearly periodic patterns@xcite .
various explanations for the formation of fairy circles have been suggested , including the release of poisonous gas and the feeding habits of ants and termites@xcite .
these explanations , however , have not uncovered , as of yet , the small - scale feedbacks needed to account for the emergence of the large - scale order@xcite . on the other hand , it is well established by model studies and confirmed by empirical observations that patch - scale biomass - water feedbacks can lead to regular landscape - scale vegetation patterns , and that periodic gap patterns , which highly resemble fairy circle patterns , can appear from uniform vegetation in response to water stress@xcite . indeed , recent detailed comparisons of fairy circle remote sensing data and model gap patterns show high similarity in static properties , such as hexagonal symmetry and spatial correlations of fairy circle size and distance@xcite . in this paper , we combine a theoretical analysis of a vegetation model , fitted to the biotic and abiotic conditions of the nfc ecosystem , with an empirical data analysis , to account for fairy circle dynamics in the namibrand nature reserve from the years 2004 to 2013 , thereby accomplishing two goals .
the first goal is to substantiate the view of fairy circles as a pattern - formation phenomenon by complementing the current statistical evidence@xcite with evidence based on the dynamics of the nfc ecosystem .
the second goal is to demonstrate the feasibility of gradual regime shifts in the nfc ecosystem as cascades of unidirectional transitions across hybrid states . |
we say that a graph is intrinsically knotted or ik if every tame embedding of the graph in @xmath1 contains a non - trivially knotted cycle .
blain , bowlin , et al .
@xcite and ozawa and tsutsumi @xcite independently discovered an important criterion for intrinsic knotting .
let @xmath2 denote the join of the graph @xmath3 and the complete graph on two vertices , @xmath4 .
a graph of the form @xmath2 is ik if and only if @xmath3 is non planar .
a graph is called @xmath5apex if it becomes planar after the deletion of at most @xmath5 vertices ( and their edges ) .
the proposition shows that @xmath0apex graphs are not ik .
it s known that many non ik graphs are @xmath0apex . as part of their proof that intrinsic knotting requires 21 edges , johnson , kidwell , and michael @xcite showed that every triangle - free graph on 20 or fewer edges is @xmath0apex and , therefore , not knotted . in the current paper , we show [ thmain]all graphs on 20 or fewer edges are @xmath0apex
. this amounts to a new proof that an ik graph has at least 21 edges .
moreover , we also show [ prop821]every non ik graph on eight or fewer vertices is @xmath0apex .
this suggests the following : is every non ik graph @xmath0apex ? we answer the question in the negative by giving an example of a graph , @xmath6 , having nine vertices and 21 edges that is neither ik nor @xmath0apex .
( we thank ramin naimi @xcite for providing an unknotted embedding of @xmath6 , which appears as figure [ fige9 ] in section 3 . )
further , we show that no graph on fewer than 21 edges , no graph on fewer than nine vertices , and no other graph on 21 edges and nine vertices has this property . in this sense , @xmath6 is the simplest possible counterexample to our question .
the notation @xmath6 is meant to suggest that this graph is a `` cousin '' to the set of 14 graphs derived from @xmath7 by triangle
y moves ( see @xcite ) .
indeed , @xmath6 arises from a y triangle move on the graph @xmath8 in the @xmath7 family .
although intrinsic knotting is preserved under triangle y moves @xcite , it is not , in general , preserved under y triangle moves .
for example , although @xmath8 is derived from @xmath7 by triangle y moves and ,
therefore , intrinsically knotted , the graph @xmath6 , obtained by a y triangle move on @xmath8 , has an unknotted embedding .
our analysis includes a classification of ik and @xmath0apex graphs on nine vertices and at most 21 edges .
such a graph is @xmath0apex unless it is @xmath6 , or , up to addition of degree zero vertices , one of four graphs derived from @xmath7 by triangle y moves @xcite .
( here @xmath9 denotes the number of vertices in the graph @xmath10 and @xmath11 is the number of edges . )
[ prop921@xmath0apex]let @xmath10 be a graph with @xmath12 and @xmath13 .
if @xmath10 is not @xmath0apex , then @xmath10 is either @xmath6 or else one of the following ik graphs : @xmath14 , @xmath15 , @xmath16 , or @xmath17 .
the knotted graphs are exactly those four descendants of @xmath7 : [ prop921ik]let @xmath10 be a graph with @xmath12 and @xmath13 .
then @xmath10 is ik iff it is @xmath14 , @xmath15 , @xmath16 , or @xmath17 .
in particular , we find that there are no new minor minimal ik graphs in the set of graphs of nine vertices and 21 edges .
we remark that a result of sachs @xcite suggests a similar analysis of @xmath18apex graphs .
a graph is intrinsically linked ( il ) if every tame embedding includes a pair of non - trivally linked cycles .
a graph of the form @xmath19 is intrinsically linked if and only if @xmath3 is non planar .
it follows that @xmath18apex graphs are not il and one can ask about the converse .
a computer search suggests that the simplest counterexample ( a graph that is neither il nor @xmath18apex ) in terms of vertex count is a graph on eight vertices and 21 edges whose complement is the disjoint union of @xmath4 and a six cycle .
bhme also gave this example as graph @xmath20 in @xcite . in terms of the number of edges ,
the disjoint union of two @xmath21 s is a counterexample of eighteen edges .
it s straightforward to verify , using methods similar to those presented in this paper , that a counterexample must have at least eight vertices and at least 15 edges . beyond these observations , we leave open the what is the simplest example of a graph that is neither il nor @xmath18apex ? the paper is organized into two sections following this introduction . in the first we prove theorem [ thmain ] . in the second we prove propositions [ prop821 ]
, [ prop921@xmath0apex ] , and [ prop921ik ] .
in this section we will prove theorem [ thmain ] , a graph of twenty or fewer edges is @xmath0apex . we will use induction and break the argument down as a series of six propositions that , in turn , treat graphs with eight or fewer vertices , nine vertices , ten vertices , eleven vertices , twelve vertices , and thirteen or more vertices . following a first subsection where we introduce some useful definitions and lemmas , we devote one subsection to each of the six propositions . in this subsection we introduce several definitions and three lemmas .
the first lemma and the definitions that precede it are based on the observation that , in terms of topological properties such as planarity , @xmath0apex , or ik , vertices of degree less than three can be ignored .
let @xmath22 denote the neighbourhood of the vertex @xmath23 .
let @xmath23 be a degree two vertex of graph @xmath10 .
let @xmath24 .
* smoothing * @xmath23 means replacing the vertex @xmath23 and edges @xmath25 and @xmath26 with the edge @xmath27 to obtain a new ( multi)graph @xmath28 .
if @xmath27 was already an edge of @xmath10 , we can remove one of the copies of @xmath27 to form the simple graph @xmath29 .
we will say @xmath29 is obtained from @xmath10 by * smoothing and simplifying * at @xmath23 .
we will use @xmath30 to denote the minimal degree of @xmath10 , i.e. , the least degree among the vertices of @xmath10 . let @xmath10 be a graph .
the multigraph @xmath3 is the * topological simplification * of @xmath10 if @xmath31 and @xmath3 is obtained from @xmath10 by a sequence of the following three moves : delete a degree zero vertex ; delete a degree one vertex and its edge ; and smooth a degree two vertex .
graphs @xmath32 and @xmath33 are * topologically equivalent * if their topological simplifications are isomorphic .
the following lemma demonstrates that in our induction it will be enough to consider graphs of minimal degree at least three , @xmath34 . for @xmath35 a vertex of graph @xmath10 , let @xmath36 denote the induced subgraph on the vertices other than @xmath35 : @xmath37
similarly , @xmath38 and @xmath39 will denote induced subgraphs on @xmath40 and @xmath41 .
[ lemd12]suppose that every graph with @xmath42 vertices and at most @xmath43 edges is @xmath0apex .
then the same is true for every graph with @xmath44 vertices , at most @xmath45 ( respectively , @xmath43 ) edges , and a vertex of degree one or two ( respectively , zero ) .
let @xmath10 have @xmath44 vertices and @xmath46 edges where @xmath47 and @xmath48 . if @xmath10 has a degree zero vertex , @xmath23 , we assume further that @xmath49 . in this case , deleting @xmath23 results in a @xmath0apex graph @xmath50 , i.e. , there are vertices @xmath35 and @xmath51 such that @xmath52 is planar
this implies @xmath38 is also planar so that @xmath10 is @xmath0apex .
if @xmath10 has a vertex @xmath23 of degree one , we may delete it ( and its edge ) to obtain a graph , @xmath53 on @xmath54 vertices with @xmath55 edges .
again , by hypothesis , @xmath50 is @xmath0apex , so that @xmath52 is planar for an appropriate choice of @xmath35 and @xmath51 .
this means @xmath38 is also planar so that @xmath10 is @xmath0apex . if @xmath10 has a vertex @xmath23 of degree two , smooth and simplify that vertex to obtain the graph @xmath28 on @xmath54 vertices and @xmath55 or @xmath56 edges . by assumption , there are vertices @xmath57 in @xmath58 such that @xmath59 is planar .
since @xmath60 , @xmath35 and @xmath51 are also vertices in @xmath10 .
notice that @xmath38 is again planar so that @xmath10 is @xmath0apex . in showing that all graphs of 20 or fewer edges are @xmath0apex
, we will frequently investigate a graph @xmath10 of 20 edges and delete two vertices to obtain @xmath61 which we may assume to be non planar . by the previous lemma
, we can assume @xmath10 has no vertices of degree less than three ( i.e. , @xmath34 ) .
it follows that @xmath62 the following lemma characterises the graphs @xmath38 of this form .
in the proof we will make use of the euler characteristic @xmath63 , where @xmath64 is the number of vertices and @xmath65 the number of edges .
[ lemnonpl]let @xmath10 be a non planar graph on @xmath54 vertices , where @xmath66 , with @xmath67 .
then @xmath10 has at least @xmath68 edges .
first observe that if @xmath10 is connected , @xmath10 will have at least @xmath69 edges . indeed , by kuratowski s theorem , @xmath10 must have @xmath70 or @xmath21 as a minor .
if there is a @xmath21 minor , then we can construct @xmath21 from @xmath10 by a sequence of edge deletions and contractions .
since both @xmath10 and @xmath21 are connected , we can arrange for the sequence to pass through a sequence of connected graphs .
we will delete any multiple edges that result from an edge contraction so that the intermediate graphs are also simple . to complete the argument notice that an edge deletion or contraction can only increase the euler characteristic .
as @xmath71 , we conclude that @xmath72 , whence @xmath73 . if , instead @xmath10 has a @xmath70 minor , then , since @xmath74 , a similar argument shows that @xmath75 . if @xmath10 is not planar , then it must have a connected component @xmath28 for which @xmath76 .
additional components will increase @xmath77 only if they are trees , i.e. , @xmath78 where @xmath79 denotes the number of tree components of @xmath10 .
if @xmath28 has at least six vertices , then , as a tree component requires at least two vertices ( recall that @xmath67 ) , we see that @xmath80 .
thus @xmath81 , as required .
if @xmath28 does nt have six vertices , then @xmath82 and @xmath83 . in this case , a similar argument shows that @xmath84 .
c|cccc & 9 & 10 & 11 & 12 + 6 & 1 & 1 + 7 & 0 & 2 & 9 + 8 & 0 & 1 & 11 + 9 & 0 & 0 & 3 + 10 & 0 & 0 & 1 & 15 + 11 & 0 & 0 & 0 & 3 + + [ rmknonpl ] table [ tblnonpl ] gives the number of graphs satisfying the hypotheses of the lemma .
moreover , using the reasoning outlined in the proof of the lemma , we can characterise such a non planar graph @xmath10 according to the number of vertices as follows . if @xmath10 has six vertices and nine edges , then @xmath85 . if @xmath86 and @xmath87 , then @xmath88 with one additional edge .
if @xmath10 has seven vertices and ten edges , it is one of the two graphs illustrated in figure [ figg710 ] obtained from @xmath21 by splitting a vertex . as for @xmath89 and @xmath90 , there are nine such graphs obtained by splitting a vertex of a non planar graph on six vertices or else by adding an edge to a graph on ten edges , see figure [ figg711 ] .
the disjoint union @xmath91 is the only graph @xmath10 with eight vertices and ten edges .
the 11 graphs @xmath10 with @xmath92 and @xmath93 are illustrated in figure [ figg811 ] .
two of the three graphs with @xmath12 and @xmath94 are formed by the union of @xmath4 with the two graphs having seven vertices and ten edges .
the third is the union of @xmath21 and the tree of two edges .
the unique graph with @xmath95 and @xmath96 is @xmath97 . of the 15 graphs with @xmath98 and @xmath99 ,
11 are formed by the union of @xmath4 with one of the non planar graphs on eight vertices and eleven edges , two are the union of the tree of two edges with a non planar graph on seven vertices and ten edges , and the remaining two are formed by the union of @xmath21 with the two trees of three edges .
the graphs with @xmath100 and @xmath99 are formed by the union of @xmath4 with a non planar graph on nine vertices and 11 edges .
if @xmath10 has @xmath101 vertices and @xmath102 edges , then , it is either @xmath103 , or else it has exactly one tree component , the rest of the graph having a @xmath21 minor .
almost all of the graphs mentioned in the remark have @xmath21 minors .
the following definition seeks to take advantage of this .
let @xmath10 be a graph with vertex @xmath104 and let @xmath105 and @xmath106 denote the vertices in the two parts of @xmath21 .
the pair @xmath107 is a * generalised @xmath108 * if the induced subgraph @xmath109 is topologically equivalent to @xmath110 .
it follows that the vertices of @xmath109 can be partitioned into five disjoint sets @xmath111 , @xmath112 , @xmath113 , @xmath114 , and @xmath115 , where each of these five sets induces a tree as a subgraph of @xmath109 , such that when each of these trees is contracted down to a single vertex , the tree on @xmath116 becomes the vertex @xmath117 in @xmath110 and similarly for the @xmath118 .
when there is a choice of partitions , a partition of a generalised @xmath21 will be one for which @xmath111 and @xmath112 are minimal .
we next observe that when @xmath38 is a generalised @xmath21 this will have implications for @xmath119 and @xmath120 , under the assumption that @xmath10 is not @xmath0apex .
[ lemgenk33]suppose that @xmath10 is not @xmath0apex and that @xmath121 is a generalised @xmath21 .
then @xmath119 and @xmath120 each include at least one vertex from each of @xmath113 , @xmath114 , and @xmath115 .
let @xmath111 , @xmath112 , @xmath113 , @xmath114 , and @xmath115 be the partition of the vertices of @xmath52 as in the definition of a generalised @xmath21 .
suppose @xmath35 has no neighbour in @xmath113 .
note that by contracting the subgraphs of @xmath122 induced by @xmath111 , @xmath112 , @xmath113 , @xmath114 , and @xmath115 , we obtain a ( multi)graph @xmath123 formed by adding a vertex @xmath35 to @xmath110 .
as @xmath35 is not adjacent to @xmath124 , it follows that @xmath125 has a planar embedding .
now , reversing the contractions performed earlier , this results in a planar embedding of @xmath122 , a contradiction .
therefore , @xmath35 has a neighbour in @xmath113 .
similarly , @xmath51 also has a neighbour in @xmath113 , and both @xmath35 and @xmath51 have neighbours in @xmath114 and @xmath115 .
[ rmkgenk33]lemma [ lemgenk33 ] also applies ( with obvious modifications ) when @xmath38 has a generalised @xmath21 component with the remaining components being trees .
we are now in a position to prove theorem [ thmain ] .
we begin with graphs of eight or fewer vertices . in what follows
, we will often make use of the following strategy . to argue that a graph @xmath10 is @xmath0apex , proceed by contradiction .
assume @xmath10 is not @xmath0apex .
this means that every subgraph of the form @xmath38 is non planar . using this assumption
we eventually deduce that a particular @xmath38 is planar .
although we wo nt always say it explicitly , in demonstrating a planar @xmath38 , we have in fact derived a contradiction that shows that @xmath10 is @xmath0apex .
[ prop820]a graph @xmath10 with @xmath126 and @xmath127 is @xmath0apex .
we can assume @xmath128 as otherwise @xmath10 is planar and _ a fortiori _ @xmath0apex .
if @xmath129 , then @xmath10 is a proper subgraph of @xmath7 .
so , with an appropriate choice of vertices @xmath35 and @xmath51 , @xmath38 is a proper subgraph of @xmath70 and therefore planar .
thus , @xmath10 is @xmath0apex .
so , we may assume @xmath92 and we will also take @xmath130 .
we will investigate induced subgraphs @xmath38 formed by deleting two vertices @xmath35 and @xmath51 .
notice that @xmath35 and @xmath51 may be chosen so that @xmath131 . indeed , the maximum degree of @xmath10 is at most seven , while the pigeonhole principle implies the maximum degree is at least five : @xmath132 . by lemma [ lemd12 ] ,
the minimum degree is at least three : @xmath34 .
since @xmath133 , the sum of the vertex degrees is @xmath134 and it follows that there are vertices @xmath35 and @xmath51 such that @xmath38 has at most ten edges .
assume @xmath10 is not @xmath0apex .
then for each pair of vertices @xmath35 and @xmath51 , @xmath38 is not planar . by lemma [ lemnonpl ] such a non
planar @xmath38 has at least nine edges .
thus , it will suffice to consider the cases where @xmath10 has a non planar subgraph @xmath38 of nine or ten vertices .
we may assume @xmath135 .
suppose first that @xmath38 is non planar and has nine edges . by remark [ rmknonpl ] , @xmath136 .
let @xmath137 be the vertices in one part of @xmath21 and @xmath138 those in the other .
since @xmath130 , @xmath139 , and @xmath135 , then @xmath140 is seven or six . in either case , @xmath141
, so we can assume @xmath142 and @xmath143 , say , are in the intersection .
if @xmath144 , it follows that @xmath145 is planar and @xmath10 is @xmath0apex . if @xmath146 , by lemma [ lemgenk33 ] , @xmath147 .
but then , since @xmath148 , we can assume @xmath149 ( i.e. , @xmath150 is an edge of @xmath10 ) and it follows that @xmath151 is planar whence @xmath10 is @xmath0apex .
next suppose @xmath38 is non planar and has ten edges .
that is , by remark [ rmknonpl ] , @xmath38 is @xmath21 with an extra edge .
again , @xmath117 and @xmath152 , ( @xmath153 ) will denote the vertices in the two parts of @xmath21 and let @xmath154 be the additional edge .
suppose first that @xmath155 .
this implies @xmath156 , @xmath157 , and there are four or five elements in @xmath158 .
if five , then @xmath10 has @xmath21 as an induced subgraph after deleting two vertices , a case we considered earlier .
so , we can assume there are four vertices in the intersection , including at least one of the vertices @xmath137 , call it @xmath104 and at least one @xmath152 vertex , say @xmath124 . then , @xmath159 is planar and @xmath10 is @xmath0apex .
so , we can assume @xmath160 .
by lemma [ lemgenk33 ] , @xmath147 . in that case , without loss of generality , @xmath149 .
then @xmath151 is planar and @xmath10 is @xmath0apex .
this completes the argument when @xmath38 has ten edges .
we have shown that when @xmath92 and @xmath161 , @xmath10 is @xmath0apex .
it follows that graphs having @xmath92 and @xmath162 are also @xmath0apex . in this subsection
we prove theorem [ thmain ] in the case of graphs of nine vertices .
we begin with a lemma .
[ lem91]let @xmath10 be a graph with @xmath163 , @xmath133 , @xmath34 , @xmath164 , and such that all degree five vertices are mutually adjacent .
then @xmath10 is @xmath0apex .
the degree bounds imply that @xmath10 has four , five , or six degree five vertices .
if @xmath10 has six degree five vertices , then , as they are mutually adjacent , @xmath10 has a @xmath165 component .
this implies the other component , on three vertices , has at most three edges and the graph has at most 18 edges in total , which is a contradiction .
so , in fact , @xmath10 can not have six degree five vertices . if @xmath10 has five degree five vertices , then the induced subgraph on the other four vertices has five edges , so it is @xmath166 ( @xmath167 with a single edge deleted ) .
let @xmath23 be a degree four vertex that has degree two in the induced subgraph @xmath166 and let @xmath35 and @xmath51 be the two degree five neighbours of @xmath23
. then @xmath38 is planar and @xmath10 is @xmath0apex . if @xmath10 has four degree five vertices , then the induced subgraph on the other five vertices has six edges , so it is one of the six graphs in figure [ fig9f1 ] .
for graphs i , ii , and iii , the argument is similar to the previous case .
that is , let @xmath23 be a degree four vertex of @xmath10 that has degree two in the induced subgraph and let @xmath35 and @xmath51 be the degree five neighbours of @xmath23 .
then @xmath38 is planar and @xmath10 is @xmath0apex . for graph
iv , if @xmath35 and @xmath51 are any of the degree five vertices , @xmath38 is planar and @xmath10 is @xmath0apex . for graphs v and vi , the argument is a little more involved , but , again , there are vertices @xmath35 and @xmath51 such that @xmath38 is planar and @xmath10 is @xmath0apex .
we are now ready to prove theorem [ thmain ] in the case of nine vertices .
[ prop920]a graph @xmath10 with @xmath12 and @xmath127 is @xmath0apex .
first , we ll assume @xmath161 .
then , @xmath168 and , by lemma [ lemd12 ] , @xmath34 .
if @xmath169 , by appropriate choice of vertices @xmath35 and @xmath51 , @xmath38 has at most ten edges .
this is also true when @xmath164 unless all degree 5 vertices are mutually adjacent . as lemma [ lem91 ] treats that case , we may assume that there is a @xmath38 of at most ten edges .
moreover , we ll take @xmath135 .
assuming @xmath10 is not @xmath0apex , then that @xmath38 is non planar . by remark [ rmknonpl ] , @xmath38 is one of the two graphs in figure [ figg710 ] .
suppose first that it is the graph at left in the figure . as @xmath170 has degree three or more in @xmath10 ,
both @xmath35 and @xmath51 are adjacent to @xmath170 . by lemma [ lemgenk33 ] , @xmath171 . without loss of generality
, we can assume @xmath149 .
then @xmath151 is planar and @xmath10 is @xmath0apex .
suppose , then , that @xmath38 is the graph at right in figure [ figg710 ] . by lemma [ lemgenk33 ] , @xmath172 and
at least one of @xmath124 or @xmath170 is a neighbour of each @xmath35 and @xmath51 .
now , as @xmath10 is not @xmath0apex , @xmath173 is non planar and it is also a graph on seven vertices and ten edges with either @xmath170 or @xmath124 of degree at least four . in other words , @xmath174 is the graph on the left of figure [ figg710 ] , a case we considered earlier .
we have shown that if @xmath161 , then @xmath10 is @xmath0apex .
it follows that the same is true for graphs with @xmath175 . in this subsection
we prove theorem [ thmain ] for graphs of ten vertices .
we begin with a lemma that treats the case of a graph of degree four .
[ lem104]suppose @xmath10 is a graph with @xmath98 , @xmath130 , and such that every vertex has degree four .
then @xmath10 is @xmath0apex .
we can assume that @xmath10 has at least three vertices @xmath35 , @xmath51 , and @xmath23 that are pairwise non adjacent for otherwise @xmath10 must be @xmath176 and is @xmath0apex .
then @xmath177 as @xmath23 will retain its full degree in @xmath38 . also , @xmath178 ; since @xmath179 , @xmath35 and @xmath51 must share at least one neighbour in the remaining seven vertices .
this will become a degree two vertex in @xmath38 .
now , @xmath38 is a graph on eight vertices and 12 edges with at least one degree two vertex .
smoothing that vertex , we arrive at @xmath28 , a multigraph on seven vertices and 11 edges that we can take to be non planar ( otherwise @xmath10 is @xmath0apex ) .
in other words , @xmath28 is either one of the graphs in figure [ figg711 ] or else one of the two graphs in figure [ figg710 ] with an edge doubled .
moreover , @xmath180 and @xmath181 . examining these candidates for @xmath28 , we see that @xmath38 has degree sequence @xmath182 or @xmath183 . .
] the six non planar graphs with degree sequence @xmath184 ( see figure [ figg842 ] ) are obtained by either doubling an edge at @xmath170 in the graph on the right of figure [ figg710 ] or else by adding a degree two vertex to graph v of figure [ figg711 ] .
if @xmath38 is one of the graphs ii , iii , or iv in figure [ figg842 ] , then we argue that @xmath10 is @xmath0apex as follows . by applying lemma [ lemgenk33 ] to @xmath185
, we find @xmath186 .
but then @xmath187 , contradicting our hypothesis that all vertices have degree four .
a similar argument ( using @xmath188 and @xmath189 in place of @xmath190 ) applies when @xmath38 is graph i. for graphs v and vi , the same approach shows that at least one of @xmath191 and @xmath192 has degree five .
the contradiction shows that @xmath10 is @xmath0apex in case @xmath38 has degree sequence @xmath184 .
so , we may assume @xmath38 has degree sequence @xmath183 .
then @xmath38 is either obtained by doubling an edge of the graph at right in figure [ figg710 ] or else by adding a degree two vertex to graph iii , iv , vi , or viii of figure [ figg711 ] .
suppose first that @xmath38 comes from doubling an edge of the right graph of figure [ figg710 ] ( and adding a degree two vertex to one of the two edges in the double ) .
up to symmetry , the doubled edge is either @xmath193 or else @xmath194 . in either case , ( @xmath38 ; @xmath191 )
is a generalised @xmath21 , whence @xmath195 . but then @xmath196 in contradiction to our hypotheses .
so @xmath10 is @xmath0apex in this case .
finally , to complete the proof , suppose @xmath38 is graph iii , iv , vi , or viii of figure [ figg710 ] .
the strategy here is similar to the previous case .
we identify a degree four vertex , @xmath23 , of @xmath38 , ( @xmath23 is @xmath143 , except for graph viii in which case @xmath23 is @xmath142 ) and observe that @xmath121 is a generalised @xmath21 .
we then find a vertex @xmath197 ( either @xmath191 or @xmath192 depending on the placement of the degree two vertex ) that must lie in @xmath158 .
consequently @xmath198 , a contradiction .
the contradiction shows that @xmath10 is @xmath0apex
. we can now prove theorem [ thmain ] for graphs of ten vertices .
a graph @xmath10 with @xmath98 and @xmath175 is @xmath0apex .
suppose @xmath95 and @xmath161
. then @xmath199 . by lemma [ lemd12 ] , we can take @xmath34 and by lemma [ lemnonpl ] , if @xmath38 is non planar , it has at least ten edges .
so , we may assume @xmath200 as , otherwise , there are vertices @xmath35 and @xmath51 so that @xmath201 whence @xmath10 is @xmath0apex . if @xmath202 , then @xmath10 is @xmath0apex unless every subgraph @xmath38 has at least ten edges .
so , we can assume @xmath10 has degree sequence @xmath203 with each of the degree four vertices adjacent to the vertex , @xmath35 , of degree seven . for almost all choices of @xmath51 ,
@xmath204 so that , by remark [ rmknonpl ] , @xmath205 .
then @xmath38 has two degree one vertices which must arise from degree three vertices of @xmath10 from which two edges have been deleted .
this implies @xmath35 is adjacent to at least two degree three vertices in @xmath10 .
this is a contradiction as @xmath119 includes only one degree three vertex , the remaining six vertices being those of degree four .
the contradiction shows that @xmath10 is @xmath0apex in case @xmath202 . if @xmath206 , then , in fact every vertex of @xmath10 has degree four .
this case is treated in lemma [ lem104 ] .
thus , the remainder of this proof treats the case where @xmath207 or @xmath208 .
then there are vertices @xmath35 and @xmath51 such that @xmath209 . by remark [ rmknonpl ]
we may assume @xmath38 is either @xmath91 or else one of the graphs in figure [ figg811 ] .
further , we will assume @xmath210 .
suppose @xmath38 is @xmath91 and let @xmath137 and @xmath106 be the vertices in the two parts of @xmath21 while @xmath211 will denote the vertices of @xmath4 . by remark [ rmkgenk33 ]
, @xmath212 shows @xmath147 .
similarly , @xmath213 implies @xmath214 . finally , as @xmath215 and @xmath216 have degree one in @xmath38 , both must be adjacent to @xmath51 in @xmath10 .
this implies @xmath217 which contradicts our assumption that @xmath218 .
the contradiction shows that @xmath10 is @xmath0apex in case it has a subgraph of the form @xmath91 .
we may now assume that @xmath219 and that for any other pair @xmath220 , @xmath221 , @xmath222 .
this allows us to dismiss the case where @xmath223 .
indeed , the condition @xmath224 then implies that the other vertices of @xmath10 have degree at most four and each degree four vertex is adjacent to @xmath35 .
but then @xmath10 would have degree sequence @xmath225 and there are too many degree four vertices for them all to be adjacent to @xmath35 .
the contradiction shows that @xmath10 is @xmath0apex in this case .
suppose then that @xmath164 , @xmath34 , and that for every choice of @xmath220 and @xmath221 , @xmath226 .
further , let @xmath35 and @xmath51 be vertices such that @xmath219 .
then @xmath38 is one of the graphs in figure [ figg811 ] and we can assume that @xmath155 .
the following argument applies to all but the last two graphs in the figure . by lemma
[ lemgenk33 ] ( or remark [ rmkgenk33 ] ) , @xmath227 .
however , either this is already a contradiction because @xmath191 or @xmath192 now has degree greater than @xmath164 , or else , @xmath187 . in the latter case , as @xmath228 then @xmath229 , contradicting our assumption that @xmath230 .
the contradiction shows that @xmath10 is @xmath0apex .
similar considerations show that if @xmath38 is graph x or xi of figure [ figg811 ] , then , again , @xmath10 must be @xmath0apex .
this completes the argument in the case that @xmath161 .
we have shown that if @xmath161 , then @xmath10 is @xmath0apex .
it follows that the same is true for graphs with @xmath175 . in this subsection , we prove theorem [ thmain ] for graphs of 11 vertices . we begin with a lemma that handles the case where @xmath206 .
[ lem114]let @xmath10 have @xmath100 , @xmath130 , and @xmath206 .
then @xmath10 is @xmath0apex . by lemma [ lemd12 ] , we can take @xmath34 so that @xmath10 has degree sequence @xmath231 .
let @xmath35 and @xmath51 be two non adjacent vertices of degree four .
then @xmath38 has nine vertices and 12 edges . since @xmath232 and @xmath233
, we see that @xmath38 has at least two vertices of degree less than two . deleting or smoothing those two , we arrive at a multigraph @xmath28 with seven vertices and ten edges .
we can assume @xmath28 is non planar as otherwise @xmath38 is planar and @xmath10 is @xmath0apex .
thus @xmath28 is either one of the two graphs in figure [ figg710 ] , @xmath234 where @xmath235 is a loop on a single vertex , @xmath236 , or else the union of @xmath237 and @xmath21 with an extra edge .
we will consider these five possibilities in turn . if @xmath28 is @xmath236 , then @xmath238 . in order to bring the four degree one vertices of @xmath38 up to degree three in @xmath10 ,
each must be adjacent to both @xmath35 and @xmath51 .
then the induced subgraph on @xmath35 , @xmath51 , and the vertices of the two @xmath4 s is planar so that @xmath10 is not only @xmath0apex , it s actually @xmath18apex .
suppose next that @xmath28 is the union of @xmath237 and @xmath21 with an extra edge .
let @xmath105 and @xmath106 be the vertices in the two parts of @xmath21 . without loss of generality ,
the extra edge of @xmath21 is either @xmath239 ( doubling an existing edge ) or else @xmath240 . by remark [ rmkgenk33 ] ,
@xmath35 and @xmath51 both have neighbours in the three sets @xmath113 , @xmath114 , and @xmath115 .
moreover , at least one of these three sets consists of a single vertex @xmath241 .
but then @xmath242 , a contradiction . the contradiction shows that @xmath10 is @xmath0apex in this case .
if @xmath243 or @xmath28 is the graph at the left of figure [ figg710 ] , the same argument applies and we conclude @xmath10 is @xmath0apex . now , if @xmath28 is the graph at the right of figure [ figg710 ] , then @xmath170 is a degree two vertex near @xmath124 ( so that @xmath113 includes at least those two vertices ) and the additional two degree one and two vertices might lie near @xmath191 and @xmath192 so that in the generalised @xmath21 , @xmath244 , none of the @xmath118 s is a single vertex . for example
, @xmath38 may be graph i of figure [ figg933222 ] below .
actually , we can conclude that @xmath38 must be graph i. for otherwise , examining @xmath245 in turn for all choices of vertex @xmath104 , we will discover at least one @xmath117 or @xmath152 vertex , call it @xmath241 , that must lie in @xmath158 which leads to the contradiction that @xmath242 .
thus , we are left to consider the case where @xmath38 is graph i of figure [ figg933222 ] below .
each of the three vertices @xmath215 , @xmath216 , and @xmath246 is adjacent to at least one of @xmath35 and @xmath51 as the @xmath247 s must have degree at least three in @xmath10 .
without loss of generality , we can assume @xmath215 and @xmath216 are neighbours of @xmath35 .
also , @xmath119 must include at least one vertex from the six @xmath117 and @xmath152 vertices .
up to symmetry , this gives two cases : @xmath248 and @xmath249 .
suppose first that @xmath248 .
then in the generalised @xmath21 , @xmath250 , @xmath251 and @xmath252 .
but , if @xmath253 , then @xmath254 is planar .
so we can assume that @xmath255 .
note that @xmath256 for otherwise @xmath257 , contradicting our assumption about the maximum degree of @xmath10 . also , we ve assumed that @xmath157 . then @xmath258 is planar unless @xmath259 .
similarly , @xmath260 and @xmath261 show that we can assume @xmath262 .
now , up to symmetry , we can assume that the fourth vertex of @xmath120 is either @xmath143 , @xmath124 , or @xmath191 , so we consider those three cases . if @xmath263 then @xmath264 is planar and @xmath10 is @xmath0apex . if @xmath265 then @xmath264 is planar and @xmath10 is @xmath0apex . if @xmath266 then @xmath267 is planar and @xmath10 is @xmath0apex .
the argument in the case that @xmath268 is similar .
having treated the case where @xmath206 , we are ready to prove theorem [ thmain ] for graphs of 11 vertices .
a graph @xmath10 with @xmath100 and @xmath175 is @xmath0apex .
suppose @xmath269 and @xmath161
. then @xmath270 . by lemma [ lemd12 ] , we can take @xmath34 and by lemma [ lemnonpl ] , if @xmath38 is non planar , it has at least 11 edges .
so , we may assume @xmath218 as , otherwise , there are vertices @xmath35 and @xmath51 so that @xmath271 whence @xmath10 is @xmath0apex .
lemma [ lem114 ] deals with graphs having @xmath206 and we treat the case of @xmath207 in the following paragraph .
suppose @xmath272 and let @xmath35 be a vertex of maximum degree . if @xmath10 is not @xmath0apex , then , to meet the requirement that @xmath273 for every choice of @xmath51 , the remaining vertices have degree three or four with all degree four vertices adjacent to @xmath35 .
it follows that @xmath10 has degree sequence @xmath274
. then @xmath35 is adjacent to exactly two of the degree three vertices , call them @xmath23 and @xmath275 .
thus @xmath276 consists of at most four other vertices beside @xmath35 .
let @xmath51 be a vertex not in @xmath276 .
then @xmath38 has 11 edges and no degree one vertex . by remark [ rmknonpl ] , @xmath38 is planar and @xmath10 is @xmath0apex .
so , for the remainder of the proof , we assume @xmath164 . if @xmath10 is not @xmath0apex , then , the condition @xmath273 implies all degree five vertices are mutually adjacent .
moreover , either there are vertices @xmath35 and @xmath51 with @xmath277 and @xmath219 , or else @xmath10 has degree sequence @xmath278 .
suppose , first , that @xmath219 with @xmath277 .
assuming @xmath10 is not @xmath0apex , by remark [ rmknonpl ] , @xmath38 is one of three graphs .
if @xmath38 is the union of the graph at the left of figure [ figg710 ] and @xmath4 , then @xmath35 must be adjacent to each of the three degree one vertices of @xmath38 as otherwise they will have degree at most two in @xmath10 . by remark [ rmkgenk33 ] ,
@xmath279 which implies @xmath280 , a contradiction .
so @xmath10 is @xmath0apex in this case .
if @xmath38 is either the union of the graph at the right of the figure and @xmath4 or else the union of @xmath21 and a tree on three vertices , again , @xmath35 must be adjacent to the two degree one vertices in the tree .
but , by remark [ rmkgenk33 ] , @xmath281 .
this again gives the contradiction @xmath280 , which shows that @xmath10 is @xmath0apex in this case as well .
thus , we can assume that @xmath10 has degree sequence @xmath282 .
further , we can assume all the degree four vertices are adjacent to @xmath35 , the vertex of degree five .
for otherwise , let @xmath51 be a degree four vertex not adjacent to @xmath35 .
then @xmath219 so it is one of the three graphs mentioned in remark [ rmknonpl ] , each of which has two degree one vertices .
as @xmath51 is adjacent to all the degree one vertex , it has at most two neighbours in @xmath283 . that would imply @xmath38 is planar , a contradiction .
so , let @xmath35 be adjacent to all the degree four vertices
. then @xmath36 has all vertices of degree three and , for any vertex @xmath51 , @xmath38 has degree sequence @xmath284 .
smoothing one of the degree two vertices , we have the multigraph @xmath28 with @xmath285 and @xmath286 . if @xmath10 is not @xmath0apex , then @xmath28 is non planar and , by remark [ rmknonpl ] , is either @xmath91 with one edge doubled or else it is one of the graphs of figure [ figg811 ] with an additional degree two vertex
. then @xmath38 is either @xmath287 , where @xmath288 is the cycle of three vertices , or else @xmath38 is @xmath21 with the addition of three degree two vertices .
however , if @xmath38 is @xmath287 we deduce that @xmath36 is @xmath289 .
let @xmath142 be one of the vertices of @xmath21 , then @xmath290 is planar and @xmath10 is @xmath0apex . . ]
so , we can assume @xmath38 is @xmath21 with the addition of three degree two vertices .
let @xmath137 and @xmath138 denote the vertices in the two parts of @xmath21 as well as the corresponding vertices in @xmath38 .
suppose the degree two vertices are all on the edges , @xmath239 , @xmath291 , and @xmath292 of @xmath21 .
then @xmath293 is planar so that @xmath10 is @xmath0apex .
thus , we can assume @xmath38 is one of the three graphs in figure [ figg933222 ] .
now , if @xmath38 is graph ii or iii , then @xmath294 is planar and @xmath10 is @xmath0apex .
so , the remainder of the proof treats the case of graph i. assume then that @xmath38 is graph i of figure [ figg933222 ] and that @xmath10 is not @xmath0apex .
further , let @xmath157 .
since @xmath295 is non planar , then @xmath296 and by removing the pairs @xmath297 and @xmath298 in turn , we see that we can assume that @xmath35 is adjacent to @xmath215 , @xmath216 , and @xmath246
. then @xmath35 is adjacent to exactly two vertices of @xmath21 , without loss of generality , either @xmath299 ; @xmath300 ; or @xmath301 .
let us examine these three subcases in turn . if @xmath302 , then @xmath303 is planar and @xmath10 is @xmath0apex . if @xmath304 , then @xmath305 is planar and @xmath10 is @xmath0apex . if @xmath306 , then @xmath307 is planar and @xmath10 is @xmath0apex .
this completes the argument in case @xmath38 is graph i of figure [ figg933222 ] and with it the case of a graph @xmath10 of twenty edges .
we have shown that if @xmath161 , then @xmath10 is @xmath0apex .
it follows that the same is true for graphs with @xmath175 . in this subsection
we prove theorem [ thmain ] in the case of a graph of 12 vertices .
a graph @xmath10 with @xmath308 and @xmath175 is @xmath0apex .
suppose @xmath309 and @xmath161
. then @xmath310 . by lemma [ lemd12 ] , we can take @xmath34 and by lemma [ lemnonpl ] , if @xmath38 is non planar , it has at least 11 edges .
so , we may assume @xmath218 as , otherwise , there are vertices @xmath35 and @xmath51 so that @xmath271 whence @xmath10 is @xmath0apex .
in fact , we can assume @xmath311 . indeed , suppose instead @xmath312 with @xmath35 a vertex of maximum degree .
as there are only twenty edges in all , there must be a degree three vertex @xmath51 not adjacent to @xmath35 . then @xmath219 . if @xmath10 is not @xmath0apex , then , by remark [ rmknonpl ] , @xmath313 .
however , as @xmath314 , @xmath38 can have at most three degree one vertices .
the contradiction shows that @xmath10 is @xmath0apex when @xmath207 .
let @xmath164 and suppose that @xmath10 has two degree five vertices @xmath35 and @xmath51 .
assuming @xmath10 is not @xmath0apex , then @xmath38 is non planar . by remark
[ rmknonpl ] , @xmath35 and @xmath51 are adjacent and @xmath313 .
it follows that each of @xmath35 and @xmath51 is adjacent to each of the four degree one vertices in @xmath38 as these vertices come to have degree three in @xmath10 .
in particular , the induced subgraph on @xmath35 , @xmath51 , and the vertices of the two @xmath4 s is planar . if @xmath142 is a vertex in the @xmath21 component of @xmath38 , then @xmath315 is planar so that @xmath10 is @xmath18apex and , therefore , also @xmath0apex .
so , we can assume @xmath10 has exactly one degree five vertex @xmath35 .
it follows that @xmath10 has exactly two degree four vertices with the remaining vertices of degree three .
we can assume that both degree four vertices are adjacent to @xmath35 as otherwise a similar argument to that of the last paragraph shows that @xmath10 is @xmath18apex .
let @xmath51 be one of the degree four vertices
. then @xmath316 . assuming @xmath10 is not @xmath0apex , then @xmath38 is non planar and therefore one of the 15 graphs described in remark [ rmknonpl ]
however , as @xmath35 is adjacent to the two degree four vertices , we see that @xmath317 which leaves seven candidate graphs : the union of @xmath4 with graph viii , ix , x , or xi of figure [ figg811 ] ; the union of the tree on two edges with the graph to the right in figure [ figg710 ] ; or @xmath21 union a tree on three edges .
( there are two such trees . )
we will consider each possibility in turn .
if @xmath38 is @xmath318 where @xmath3 is graph ix , x , or xi of figure [ figg811 ] , then we deduce that @xmath35 is adjacent to one of the degree three vertices of @xmath3 , call it @xmath104 , as that is the only way to produce a second degree four vertex in @xmath10 ( besides @xmath51 ) .
we claim that @xmath319 is planar .
indeed , @xmath320 . but @xmath321 is connected , so it is not one of the non planar graphs described in remark [ rmknonpl ] . as @xmath321 is planar , @xmath10 is @xmath0apex .
if @xmath38 is @xmath318 where @xmath3 is graph viii of figure [ figg811 ] , again , @xmath35 is adjacent to a degree three vertex of @xmath3 .
if that vertex is one of the six @xmath117 or @xmath152 vertices , the argument proceeds as above .
so assume instead @xmath35 is adjacent to the seventh degree three vertex .
in this case @xmath315 is planar so @xmath10 is @xmath18apex , hence @xmath0apex .
if @xmath38 is the union of the right graph of figure [ figg710 ] , call it @xmath3 , with a tree @xmath79 of two edges , we again conclude that if @xmath35 is adjacent to @xmath104 , a degree three vertex , of @xmath3 then @xmath321 is planar whence @xmath10 is @xmath0apex . the only other way to produce a degree four vertex for @xmath10 is if @xmath35 and @xmath51 are both adjacent to all three vertices of @xmath79 .
however , in this case we find that the subgraph induced by @xmath35 , @xmath51 , and the vertices of the tree is planar so that @xmath10 is @xmath18apex and , therefore , also @xmath0apex .
similar arguments apply when @xmath38 is the union of @xmath21 and the tree @xmath322 , the path of three edges : either @xmath35 is adjacent to a vertex @xmath104 of @xmath21 , which means that @xmath321 is planar , or else the graph induced by @xmath35 , @xmath51 and @xmath322 is planar so that @xmath10 is , in fact , @xmath18apex , hence @xmath0apex . as for @xmath323 , where @xmath324 is the star of three edges , again @xmath321 is planar where @xmath104 is the vertex of @xmath21 adjacent to @xmath35 if there is such and otherwise @xmath104 is an arbitrary vertex of @xmath21 .
this completes the argument when @xmath164 .
finally , suppose @xmath206 .
then there are four degree four vertices with the remaining vertices of degree three .
if there are non adjacent degree four vertices @xmath35 and @xmath51 , then @xmath232 and the analysis is much as the one just completed in the @xmath164 case .
that is , we can assume @xmath38 is one of the fifteen graphs described in remark [ rmknonpl ] with the additional condition that @xmath325 .
so , to complete the proof , let s assume the four degree four vertices , call them @xmath35 , @xmath51 , @xmath23 , and @xmath275 , are mutually adjacent . then @xmath23 and @xmath275 become two adjacent degree two vertices in @xmath38 .
smoothing these we arrive at @xmath28 where @xmath285 and @xmath286 .
we can assume that @xmath28 is non planar ( otherwise @xmath38 is planar and @xmath10 is @xmath0apex ) so that it is one of the graphs of figure [ figg811 ] , @xmath91 with an edge doubled , or else the union of one of the graphs of figure [ figg710 ] with @xmath235 , a loop on one vertex .
in addition , @xmath326 , which leaves six possibilities : graph viii , ix , x , or xi of figure [ figg811 ] , @xmath327 , where @xmath328 is the cycle on two vertices , or else the union of @xmath235 and the graph at the right of figure [ figg710 ] . we ll consider these in turn .
if @xmath28 is graph viii , ix , x , or xi of figure [ figg811 ] , let @xmath329 be the edge of @xmath28 that contained @xmath23 and @xmath275 before smoothing .
that is , @xmath197 and @xmath330 are the vertices in @xmath38 such that @xmath331 , @xmath25 , and @xmath332 is a path .
then @xmath333 is planar and @xmath10 is @xmath0apex .
if @xmath28 is @xmath327 , then @xmath38 is @xmath334 with @xmath23 and @xmath275 two of the vertices in the @xmath335cycle @xmath336 .
then @xmath145 is planar where @xmath142 is a vertex of @xmath21
. finally , if @xmath28 is the union of @xmath235 and the right graph of figure [ figg710 ] , call it @xmath337 , then @xmath38 is @xmath338 where @xmath23 and @xmath275 are two of the vertices in the @xmath339cycle @xmath288 .
it follows that @xmath315 is planar so that @xmath10 is @xmath18apex , hence @xmath0apex .
this completes the case where @xmath206 , and with it the proof for @xmath161 . as usual , since all graphs with @xmath133 are @xmath0apex , the same is true for graphs with @xmath175 . in this subsection
, we complete the proof of theorem [ thmain ] by examining graphs with 13 or more vertices .
a graph @xmath10 with @xmath340 and @xmath175 is @xmath0apex .
suppose @xmath341 and @xmath161 . by lemma [ lemd12 ] , we can assume @xmath34 so that @xmath10 has a single vertex @xmath35 of degree four with all other vertices of degree three .
let @xmath51 be a vertex that is not adjacent to @xmath35 so that @xmath342 .
assume @xmath10 is not @xmath0apex .
then @xmath38 is non planar .
now , @xmath317 , so @xmath38 has no @xmath70 component . by remark [ rmknonpl ]
, @xmath38 has exactly one tree component @xmath79 , with the rest of the graph @xmath343 having a @xmath21 minor . as @xmath233 , there are no isolated degree zero vertices , so @xmath344 and we have four cases . if @xmath345 , then @xmath79 is @xmath4 and @xmath343 is a non - planar graph on nine vertices with 12 edges . as @xmath326 and @xmath346 , @xmath28 has a vertex of degree two . by smoothing that vertex , we have either a multigraph obtained by doubling an edge of the graph @xmath91 or else one of the graphs of figure [ figg811 ] .
moreover , as @xmath326 , of the graphs in the figure , only viii , ix , x , and xi are possibilities .
suppose then that , after smoothing and simplifying , @xmath28 is @xmath91 .
then , as @xmath326 , the doubled edge is that of the @xmath4 and @xmath347 , where @xmath288 denotes the cycle on three vertices .
thus , @xmath348 . let @xmath23 be one of the vertices in the @xmath21 component .
then @xmath349 is planar and @xmath10 is @xmath0apex . if , after smoothing a degree two vertex , @xmath28 becomes graph viii , ix , x , or xi of figure [ figg811 ] , then @xmath350 is planar and @xmath10 is @xmath0apex .
next suppose @xmath351 .
as , @xmath285 , @xmath352 , and @xmath326 , we conclude that @xmath28 is graph viii , ix , x , or xi of figure [ figg811 ] .
whichever it is , @xmath350 will be a planar subgraph of @xmath10 so that @xmath10 is @xmath0apex .
similarly , if @xmath353 , then @xmath354 , @xmath355 . as @xmath326 , we conclude that @xmath28 is the graph to the right of figure [ figg710 ] .
then @xmath290 is planar and @xmath10 is @xmath0apex . finally , if @xmath356 , then @xmath357 and @xmath358 so that @xmath28 is @xmath21 .
again , @xmath290 is planar and @xmath10 is @xmath0apex .
we have shown that a graph with @xmath341 and @xmath130 is @xmath0apex .
it follows that the same is true for graphs having @xmath341 and @xmath175 .
now , suppose @xmath359 and @xmath130 . if @xmath34 , then the degree sum is at least @xmath360 , a contradiction .
so , we may assume @xmath361 which implies @xmath10 is @xmath0apex by lemma [ lemd12 ] .
it follows that any graph of 14 or more vertices with fewer than 20 edges is also @xmath0apex .
in this section we prove propositions [ prop821 ] ( in the first subsection ) and propositions [ prop921@xmath0apex ] and [ prop921ik ] ( in the second subsection ) . in this subsection
we prove proposition [ prop821 ] , a non - ik graph of eight or fewer vertices is @xmath0apex .
this implies that for these graphs @xmath0apex is equivalent to not ik and the classification of @xmath0apex graphs on eight or fewer vertices follows from the ik classification due to @xcite and @xcite .
every non ik graph on eight or fewer vertices is @xmath0apex . by proposition [ prop820 ]
, the theorem holds if @xmath362 , so we may assume that @xmath363 .
the only graph with 21 edges and fewer than eight vertices is @xmath7 , which is ik .
so we may assume @xmath92 . and
@xmath33.[fig8g1g2 ] ] knotting of graphs on eight vertices was classified independently by @xcite and @xcite . using the classification , the non ik graphs with 21 or more edges are all subgraphs of two graphs on 25 edges , @xmath32 and @xmath33 , whose complements appear in figure [ fig8g1g2 ] .
each of these two graphs has at least two vertices of degree seven and , for both graphs , deleting two such vertices leaves a planar subgraph of @xmath165 .
thus , both @xmath32 and @xmath33 are @xmath0apex and the same is true of any subgraph of @xmath32 and @xmath33 . in this subsection
we prove propositions [ prop921@xmath0apex ] and [ prop921ik ] , which classify the graphs of nine vertices and at most 21 edges with respect to @xmath0apex and ik .
we begin with proposition [ prop921@xmath0apex ] : among these graphs , all but @xmath6 ( see figure [ fige9 ] ) and four graphs derived from @xmath7 by triangle y moves ( @xmath14 , @xmath15 , @xmath16 , and @xmath17 , see @xcite ) are @xmath0apex .
we first present four lemmas that show this is the case when there is a subgraph @xmath38 of the form shown in figure [ figg711 ] .
the first lemma shows that we can assume @xmath364 .
the next three treat the five graphs ( iii , iv , v , vi and viii ) of figure [ figg711 ] that meet this condition .
[ lem9.1]let @xmath10 be a graph with @xmath12 , @xmath365 and @xmath34 .
suppose that , for each pair of vertices @xmath220 and @xmath221 , @xmath222 with equality for at least one pair @xmath35 , @xmath51 .
then , @xmath35 and @xmath51 can be chosen so that one of the following two holds . * the vertices @xmath35 and @xmath51 have degrees six and five , respectively
, @xmath10 has one of the following degree sequences : @xmath366 , @xmath367 , or @xmath368 , and @xmath35 is adjacent to each degree five vertex ( including @xmath51 ) . * the vertices @xmath35 and
@xmath51 both have degree five , @xmath10 has one of the following degree sequences : @xmath369 or @xmath370 , and @xmath35 and @xmath51 are not neighbours .
moreover , @xmath35 and @xmath51 can be chosen so that @xmath364 .
we can assume @xmath210 . since @xmath222 for every pair of vertices @xmath220 ,
@xmath221 , we must have @xmath146 or @xmath155 . if @xmath146 , the condition @xmath226 implies that there is exactly one degree six vertex , @xmath35 , and every degree five vertex is adjacent to @xmath35 . as @xmath371 , the degree sum is 42 and , therefore , there are only three possibilities for the degree sequence .
in particular , there is always a vertex of degree five @xmath51 which is adjacent to @xmath35 so that @xmath219 .
similarly , if @xmath155 , then the condition @xmath371 leaves two possible degree sequences .
there must be two degree five vertices @xmath35 and @xmath51 that are not adjacent so that @xmath219 .
this is clear for the degree sequence with seven degree five vertices . in the case of six
degree five vertices , if they were all mutually adjacent , they would constitute a @xmath165 component of 15 edges .
the other component has three vertices and , at most , three edges . in total
, @xmath10 would have at most 18 edges , a contradiction .
finally , we argue that it is always possible to choose @xmath35 and @xmath51 so that @xmath364 . indeed , this is obvious when @xmath372 as deleting @xmath35 and @xmath51 can reduce the degree of the other vertices by at most two . for the sequence @xmath366 ,
the degree six vertex @xmath35 is adjacent to each degree five vertex and is , therefore , not adjacent to either of the degree three vertices .
hence in @xmath38 these degree three vertices have degree at least two and @xmath364 . for @xmath367 ,
the degree three vertex is adjacent to at most three of the degree five vertices . by choosing @xmath51 as one of the other degree five vertices , we will have @xmath364 . similarly , for @xmath369 , the degree three vertex is adjacent to at most three of the degree five vertices , call them @xmath142 @xmath143 , and @xmath373 .
we can find degree five vertices @xmath35 and @xmath51 that are not adjacent and not both neighbours of the degree three vertex ( so that @xmath364 ) . for , if not , then the remaining four degree five vertices , @xmath374 , @xmath375 , @xmath376 , and @xmath377 are all mutually adjacent and also all adjacent to @xmath142 , @xmath143 , and @xmath373 .
but this is not possible , e.g. , @xmath378 , so it can not have all of the other degree five vertices as neighbours .
[ lem9iii]let @xmath10 be a graph with @xmath12 , @xmath365 , and @xmath34 .
suppose that there are vertices @xmath35 and @xmath51 such that @xmath38 is graph iii of figure [ figg711 ] and for any pair of vertices @xmath220 and @xmath221 , @xmath379 .
if @xmath10 is not @xmath0apex , then @xmath10 is @xmath17 . by lemma [ lem9.1 ] , @xmath146 or @xmath208 . also , since @xmath170 has degree three or more in @xmath10 , at least one of @xmath35 and @xmath51 is adjacent to @xmath170 .
assuming @xmath10 is not @xmath0apex , by lemma [ lemgenk33 ] , @xmath380 .
then @xmath381 is planar ( and @xmath10 is @xmath0apex ) in the case @xmath155 .
so , we can assume @xmath146 and we are in the first case of lemma [ lem9.1 ] .
as above @xmath382 .
then , since @xmath151 is non planar , we deduce that @xmath383 .
finally , since @xmath384 is non planar , @xmath142 and @xmath143 are also neighbours of @xmath35 , i.e. , @xmath385 .
but these choices of @xmath120 and @xmath119 result in the graph @xmath17 .
so , if @xmath10 is not @xmath0apex , it is @xmath17 .
[ lem9ivtovi]let @xmath10 be a graph with @xmath12 , @xmath365 , and @xmath34 .
suppose that there are vertices @xmath35 and @xmath51 such that @xmath38 is graph iv , v , or vi of figure [ figg711 ] and for any pair of vertices @xmath220 and @xmath221 , @xmath379 .
then @xmath10 is @xmath0apex . by lemma [ lem9.1 ] , @xmath146 or @xmath208 .
if @xmath155 , note that @xmath386 is a cycle . by placing @xmath35 inside the cycle and @xmath51 outside ,
@xmath387 is planar and @xmath10 is @xmath0apex .
so , we may assume @xmath388 and we are in the first case of lemma [ lem9.1 ] .
assume @xmath10 is not @xmath0apex and apply lemma [ lemgenk33 ] to @xmath185 , for which @xmath389 and @xmath390 , @xmath391 .
if @xmath10 is graph iv or v , then @xmath173 is planar and @xmath10 is @xmath0apex .
so , let @xmath10 be graph vi . then , since @xmath173 is non planar , either @xmath142 or @xmath143 , say @xmath142 , is a neighbour of @xmath51 .
but , then the degree of @xmath142 in @xmath10 is at least five .
we deduce that @xmath392 , for otherwise , @xmath257 and , by lemma [ lem9.1 ] , @xmath142 is adjacent to @xmath35 , a contradiction .
however , as @xmath35 is adjacent to @xmath142 , @xmath393 which again contradicts lemma [ lem9.1 ] as @xmath35 is the unique vertex of degree six .
the contradiction shows that @xmath10 is @xmath0apex .
.[fige9 ] ] [ lem9viii]let @xmath10 be a graph with @xmath12 , @xmath365 , and @xmath34 .
suppose that there are vertices @xmath35 and @xmath51 such that @xmath38 is graph viii of figure [ figg711 ] and for any pair of vertices @xmath220 and @xmath221 , @xmath379 .
if @xmath10 is not @xmath0apex , then @xmath10 is @xmath6 . by lemma [ lem9.1 ] , @xmath146 or @xmath208 .
also , since @xmath170 has degree three or more in @xmath10 , at least one of @xmath35 and @xmath51 is adjacent to @xmath170 .
assume @xmath10 is not @xmath0apex and apply lemma [ lemgenk33 ] using @xmath389 and @xmath390 , @xmath394 , to see that @xmath395 and that @xmath119 and @xmath120 both intersect @xmath113 .
similarly @xmath188 shows @xmath396 and both @xmath35 and @xmath51 have a neighbour in @xmath397 .
if @xmath146 , then @xmath398 , which contradicts what we already know about @xmath120 .
so , it must be that @xmath155 , from which it follows that @xmath399 and that @xmath400 .
having treated graphs containing an induced subgraph as in figure [ figg711 ] , we are ready to prove proposition [ prop921@xmath0apex ] .
let @xmath10 be a graph with @xmath12 and @xmath13 .
if @xmath10 is not @xmath0apex , then @xmath10 is either @xmath6 or else one of the following ik graphs : @xmath14 , @xmath15 , @xmath16 , or @xmath17 . by theorem [ thmain ]
, we can assume @xmath365 .
as in the proof of lemma [ lemd12 ] , @xmath34 unless @xmath10 has a vertex of degree lower than three whose deletion ( or smoothing in the case of a degree two vertex ) results in a graph that is not @xmath0apex . as all graphs of 20 edges are @xmath0apex , this is possible only in the case that @xmath10 has a degree zero vertex ; deleting that vertex must result in a graph on eight vertices with 21 edges that is not @xmath0apex . by proposition [ prop821 ]
such a graph is ik and , by the classification of knotting of eight vertex graphs , we conclude that @xmath10 is either the union of @xmath7 with two degree zero vertices , @xmath14 , or else @xmath10 is @xmath15 , where @xmath401 is the graph obtained by a single triangle
y move on @xmath7 ( see @xcite ) .
in other words , so long as @xmath402 and @xmath403 , we can assume @xmath34 . also , @xmath168 .
now , if @xmath164 , there are at least six degree five vertices and , therefore , there must be a pair of non adjacent degree five vertices .
thus , whatever the maximum degree @xmath404 , by appropriate choice of vertices @xmath35 and @xmath51 , we may assume @xmath38 has at most 11 edges . if @xmath405 , then @xmath10 is @xmath0apex . indeed ,
if @xmath35 has degree eight , then , as @xmath406 , there is a vertex @xmath51 with @xmath407 .
this means @xmath38 has at most nine edges and , by lemma [ lemnonpl ] , is planar .
so we ll assume @xmath38 has at most 11 edges and that @xmath408 assume @xmath10 is not @xmath0apex ; then @xmath38 is non planar . by remark [ rmknonpl ] ,
@xmath38 is one of the two graphs in figure [ figg710 ] or one of the nine in figure [ figg711 ] .
suppose first that @xmath38 is the graph at left in figure [ figg710 ] .
since @xmath204 , we can assume that @xmath144 or @xmath409 and @xmath410 .
as @xmath170 has degree three in @xmath10 , both @xmath35 and @xmath51 are adjacent to @xmath170 . by lemma [ lemgenk33 ] , @xmath171 . without loss of generality
, we can assume @xmath149 .
then @xmath151 is planar and @xmath10 is @xmath0apex .
if @xmath38 is the graph at right in figure [ figg710 ] , then , as @xmath170 has degree three or more in @xmath10 , at least one of @xmath35 and @xmath51 is a neighbour of @xmath170 .
again , @xmath411 so @xmath144 or @xmath409 and @xmath410 .
applying lemma [ lemgenk33 ] with @xmath412 and @xmath390 , @xmath394 , we see that @xmath119 and @xmath120 each include at least one vertex from each @xmath118 .
similarly , @xmath413 shows @xmath119 and @xmath120 each include at least one vertex from each of @xmath414 and @xmath415 .
@xmath394 . in particular , we conclude that @xmath416 .
then @xmath173 is a non planar graph on seven vertices and eleven edges , i.e. , one of the graphs in figure [ figg711 ] . in particular , if @xmath144 , then the degree of @xmath35 in @xmath173 is five .
the only graph of figure [ figg711 ] with a degree five vertex is i. however , in that graph , the degree five vertex is adjacent to a degree one vertex which is not a possibility for @xmath35 .
so , we conclude @xmath173 is planar and @xmath10 is @xmath0apex if @xmath144 .
thus , we can assume @xmath146 and @xmath156 or @xmath409 .
if @xmath156 , then the discussion above shows that @xmath417 and @xmath387 is planar ( so that @xmath10 is @xmath0apex ) . if @xmath418 , then @xmath419 which implies @xmath420 and @xmath421 .
we can now assume that there is a pair of vertices @xmath35 and @xmath51 such that @xmath219 and @xmath38 is one of the nine graphs in figure [ figg711 ] .
moreover , we can also posit that for any pair of vertices @xmath220 ,
@xmath221 , @xmath422 , for otherwise the subgraph has ten vertices ( in order to ensure it is non planar , see lemma [ lemnonpl ] ) , which is the case we just treated above .
lemma [ lem9.1 ] describes the possible degrees for such a graph .
in particular , @xmath364 and since @xmath10 is not @xmath0apex , @xmath38 must be graph iii , iv , v , vi , or viii in figure [ figg711 ] .
lemmas [ lem9iii ] through [ lem9viii ] show that in those cases , if @xmath10 is not @xmath0apex , then @xmath10 is @xmath6 or @xmath17 .
this completes the proof . finally , we prove proposition [ prop921ik ] .
let @xmath10 be a graph with @xmath12 and @xmath13 .
then @xmath10 is ik iff it is @xmath14 , @xmath15 , @xmath16 , or @xmath17 .
it follows from @xcite that , if @xmath10 is one of the four listed graphs , then it is ik . by proposition [ prop920 ] ,
@xmath10 is @xmath0apex and , therefore , not ik when @xmath175 . in case @xmath371 ,
proposition [ prop921@xmath0apex ] shows @xmath10 is @xmath0apex and not ik unless @xmath10 is one of the four listed graphs or @xmath6 .
however , figure [ fige9 ] is an unknotted embedding of @xmath6 .
so , if @xmath10 is ik , it must be one of the four listed graphs .
it is straightforward to verify that figure [ fige9 ] is an unknotted embedding .
for example , here s a strategy for making such a verification .
number the crossings as shown .
it is easy to check that there are 16 possible crossing combinations for a cycle in this graph : 1236 , 134 , 1457 , 1459 , 1479 , 16789 , 234689 , 23479 , 236789 , 25689 , 2578 , 345689 , 3457 , 3578 , 36789 , and 4578 . that is
, any cycle in the graph will have crossings that are a subset of one of those 16 sets .
for example , a cycle that includes crossings 1 , 2 , 3 , and 6 must include the edges @xmath423 and @xmath424 and therefore , can not have the edge @xmath425 required for crossing 4 . indeed , a cycle that includes 1 , 2 , 3 , and 6 can have none of the other five crossings . to show that there are no knots , consider cycles that correspond to each subset of the 16 sets and check that each such cycle ( if such exists ) is not knotted in the embedding of figure [ fige9 ] .
we thank ramin naimi for encouragement and for sharing an unknotted embedding of the graph @xmath6 and joel foisy for helpful conversations .
this study was inspired by the master s thesis of chris morris .
mrs p. blain , g. bowlin , t. fleming , j. foisy , j. hendricks , and j. lacombe , ` some results on intrinsically knotted graphs , ' _
j.knot theory ramifications _ , * 16 * ( 2007 ) , 749760 .
t. bhme , ` on spatial representations of graphs , ' _ contemporary methods in graph theory _ , bibliographisches inst . ,
mannheim , ( 1990 ) , 151167 .
j. campbell , t.w .
mattman , r. ottman , j. pyzer , m. rodrigues , and s. williams , ` intrinsic knotting and linking of almost complete graphs , ' _
kobe j. math _ , * 25 * ( 2008 ) , 3958 .
math.gt/0701422 b. johnson , m.e .
kidwell , and t.s .
michael , ` intrinsically knotted graphs have at least 21 edges , ' ( to appear in _ j. knot theory ramifications _ ) .
t. kohara and s. suzuki , ` some remarks on knots and links in spatial graphs ' , in _ knots 90 , osaka , 1990 _ , de gruyter ( 1992 ) 435445 .
r. motwani , a. raghunathan , and h. saran , ` constructive results from graph minors : linkless embeddings , ' _ 29th annual symposium on foundations of computer science , ieee _ ( 1998 ) , 298409 .
r. naimi , private communication .
m. ozawa and y. tsutsumi , ` primitive spatial graphs and graph minors , ' _ rev .
complut . _ * 20 * ( 2007 ) , 391406 .
h. sachs , ` on spatial representation of finite graphs ' , in : a. hajnal , l. lovasz , v.t .
ss ( eds . ) , _ colloq .
math.soc .
jnos bolyai _ , north - holland , amsterdam , ( 1984 ) , 649662 . | a graph is @xmath0apex if it is planar after the deletion of at most two vertices .
such graphs are not intrinsically knotted , ik .
we investigate the converse , does not ik imply @xmath0apex ? we determine the simplest possible counterexample , a graph on nine vertices and 21 edges that is neither ik nor @xmath0apex . in the process , we show that every graph of 20 or fewer edges is @xmath0apex .
this provides a new proof that an ik graph must have at least 21 edges .
we also classify ik graphs on nine vertices and 21 edges and find no new examples of minor minimal ik graphs in this set . |
in our ongoing araucaria project , we are improving on the usefulness of a number of stellar distance indicators by determining their environmental dependences from a study of these objects in a number of nearby galaxies with largely different environmental parameters .
we have described our approach in a number of previous papers ( pietrzyski et al .
2002 , 2004 ; gieren et al .
cepheid variables are among the most important distance indicators we are studying . in spite of recent claims about a possible non - universality of the period - luminosity ( pl ) relation ( sandage et al .
2004 , marconi et al .
2005 ) , or a break in the relation at a period of about 10 days ( ngeow et al .
2005 , marconi et al . 2005 ) , previous results of the araucaria project seem to indicate that the pl relation is a superb instrument for distance measurement , particularly in the near - infrared where reddening is not a significant source of systematic error ( gieren et al .
2005b , 2005c ; pietrzyski et al .
2004 , 2006 ) .
we have been conducting wide - field optical photometric surveys for cepheid variables in all target galaxies of the project ; so far , results were reported for ngc 300 ( pietrzyski et al .
2002 ; gieren et al . 2004 ) and ngc 6822 ( pietrzyski et al .
we have also been carrying out near - ir ( jk ) follow - up observations of cepheids in all araucaria target galaxies ; results were reported in gieren et al .
2005c ( ngc 300 ) , and pietrzyski et al .
2006 ( ic 1613 ) .
these studies have shown that a combined optical / near - infrared analysis of cepheid variables in nearby , well resolved galaxies can yield distances accurate to 3 percent . in this paper
, we are extending our cepheid work to the local group irregular galaxy ngc 3109 , with the aim to provide a sample of cepheids with well - observed light curves in v and i bands which is near - complete over the whole spatial extent of the galaxy , and down to a pulsation period of about 4 days .
this survey complements , and enlarges previous surveys which have covered only small areas in ngc 3109 , or which lacked the sensitivity to discover the population of short - period and relatively faint variables .
our new cepheid observations give us a superb database to improve on the previous cepheid - based distance determinations of ngc 3109 of sandage & carlson ( 1988 ) , capaccioli et al .
( 1992 ) , and musella et al .
( 1997 ) , which we exploit in section 4 of this paper . in the first , photographic study of ngc 3109
sandage & carlson ( 1988 ) had discovered 29 cepheids . in this study
, we recovered 28 of these 29 variables , which are among the brightest cepheids in the galaxy .
the sandage & carlson cepheid variable v12 was not recovered in our survey because it fell in the gap between the two ccds which we used to image the galaxy ( fig .
capaccioli , piotto and bresolin ( 1992 ) did not discover additional cepheids but rather corrected the b - band zero point of sandage s photometry and improved on the distance determination .
later musella , piotto and capaccioli ( 1997 ) obtained bvri light curves for 24 cepheids in a small field of ngc 3109 , 16 of which were newly discovered , and bringing the total number of known cepheid variables in the galaxy to 45 .
we did not detect 6 out of 16 new cepheids reported by musella , piotto and capaccioli ( 1997 ) .
three of them , p4 , p8 and p11 have periods of about 2 days and are clearly too faint on our images to be able to derive reliable photometry for them . the cepheids designated as p7 and p9
are heavily blended and we suspect that they may have been spurious detections . finally , p15 is not a stellar object - most probably it is a background galaxy .
our current survey enlarges the known cepheid population in ngc 3109 to 113 objects and is the first comprehensive survey of cepheids in this galaxy which is near - complete to rather short periods ( about 4 days ) , and over the complete spatial extent of ngc 3109 .
the paper is organized as follows . in section 2
, we desribe the observations , calibrations and data reduction procedures . in section 3
, we present the catalog of cepheid variables in ngc 3109 obtained from our observations , and in section 4 we discuss the pl relations obtained from our data , and use them to determine the distance to ngc 3109 .
we discuss our results in section 5 and summarize the main conclusions of the paper in section 6 .
all the data presented in this paper were collected with the warsaw 1.3 m telescope at las campanas observatory .
the telescope was equipped with a mosaic 8 chip detector , with a field of view of about 35 x 35 arcmin and a scale of 0.25 arcsec / pixel . for more details on the instrumental system ,
the reader is referred to the ogle web site ( www.astrouw.edu.pl/ogle ) .
the observations were secured on 74 different nights during two consecutive observing seasons . on each night
, 900 sec integrations were performed through v and i filters .
after preliminary reductions with the iraf package ( e.g. debiasing , flatfielding ) the data were reduced with the ogle iii pipeline on the basis of the image - subtraction technique ( udalski 2003 , woniak 2000 ) . in order to calibrate our photometry onto the standard system , about 25 landolt standards were observed together with our target field during three photometric nights .
standard stars were observed on each of the chips , and transformation coefficients were derived independently for each chip , and on each of the three photometric nights . in order to account for small variations of the zero point in both filters over the mosaic field
, we adopted the correction maps elaborated by pietrzyski et al .
we estimated the accuracy of the zero points of our photometry to be better than 0.03 mag in both filters .
2 shows a comparison of our photometry with the photometry obtained by hidalgo et al .
( in preparation ) with the 2.5 m du pont telescope at las campanas .
it turns out that in both filters the zero points are consistent at the 0.03 mag level ( e.g. within the quoted errors ) .
all stars observed in ngc 3109 were searched for photometric variations with periods between 0.2 and 100 days , using the analysis of variance algorithm ( schwarzenberg - czerny 1989 ) . in order to distinguish cepheids from other types of variable stars
, we used the same criteria as in pietrzyski et al .
all light curves of those variables identified as cepheid candidates were approximated by fourier series of order 4 .
we then rejected those objects with v amplitudes smaller than 0.4 mag , which is approximately the lower amplitude limit for normal classical cepheids .
for the stars passing our selection criteria , mean v and i magnitudes were derived by integrating their light curves , which had been previously converted onto an intensity scale , and converting the results back to the magnitude scale . because of the very good quality and phase coverage of our light curves , the statistical accuracy of the derived mean magnitudes was typically 0.01 mag for the brighter variables , and 0.02 - 0.03 mag for the faintest cepheids in our catalog .
we ended up with a list of 113 bona fide classical cepheids with periods between 31.4 and 3.4 days .
there are evidently no truly long - period cepheids in ngc 3109 , with periods up to 50 days or more , as in most other galaxies of the araucaria project we have studied so far .
table 1 gives the individual photometric v and i observations for all cepheid variables we found .
the full table 1 is available in electronic form . in table 2 , we present their identifications , equatorial coordinates , periods , times of maximum light in v , mean v- and i - band intensity magnitudes , and cross - identifications with previous surveys .
the accuracy of the periods can be judged from the last digits given in column 4 of table 2 ; it is typically a few units of the last digit .
we note here that for 14 cepheids previously discovered in ngc 3109 , the reported periods were wrong . for the remaining variables with first - epoch observations
, our new data helped to improve the periods significantly . in fig .
3 , we show phased v- and i - band light curves for some of the variables in our catalog which are typical for other cepheids of similar periods , in order to demonstrate the quality of our data .
for the purpose of studying the pl relations defined by the cepheids in ngc 3109 , and determine the distance of the galaxy , we excluded all variables with periods less than logp ( days)=0.75 ( 5.6 days ) .
this period cutoff was chosen because shortward of it the scatter of the data points in the pl plane starts to increase significantly , as a consequence of the increasingly lower signal - to - noise ratio in the photometry .
also , below a period of 5.6 days contamination of the sample with first - overtone pulsators is expected to become a problem as impressively demonstrated in the lmc pl relations provided by the different microlensing projects ( e.g. udalski et al .
1999 ) .
4 shows the location of all cepheids in our catalog on the period - magnitude planes in v and i. stars with periods shorter than the adopted cutoff period , as well as a very few outliers , are marked with open symbols in these plots . our final cepheid sample used for the analysis of the pl relations in ngc 3109 , excluding the stars marked with open circles in fig .
4 , contains 83 objects .
this subsample of cepheids should not contain peculiar , or strongly blended stars , and it should be free of contaminating overtone pulsators whose inclusion in the sample would tend to make the derived cepheid distance to ngc 3109 too short . in fig .
4 , an indication of a possible change of the slope at the log(p ) = 1.2 can be noticed , but it can not be clearly verified due to the relatively small number of long period cepheids in ngc 3109 . however , since the slopes of the p - l relations ( see below ) agree relatively well with the corresponding slopes obtained for the lmc cepheids this effect should not affect our distance determination significantly . a plot of the location of these 83 variables on the v , v - i color - magnitude diagram in fig .
5 shows that all these stars , with the sole possible exception of one rather blue object , are indeed located in the cepheid instability zone , giving support to our conclusion that we have isolated a very nearly uncontaminated sample of classical cepheids in ngc 3109 for our study of the pl relation , and its application for the determination of the distance to the galaxy . in agreement with our usual procedure ( e.g. gieren et al .
2004 , pietrzyski et al .
2004 ) , we performed fits to a straight line to the 83 datapoints in the pl planes adopting for this purpose the slopes of the lmc cepheid pl relations established by the ogle - ii project ( udalski 2000 ) .
free fits to the datapoints which do not pre - specify the slopes , yield slopes which agree with the adopted lmc slopes within the combined 2 @xmath1 uncertainties , yielding no evidence for any significant departure of the ngc 3109 pl relation slopes in v and i from the fiducial lmc pl relations , in line with our previous findings for ngc 300 and ngc 6822 ( gieren et al .
2004 , pietrzyski et al .
2004 ) . fitting the ogle lmc slopes to our data
yields the following relations : + v = -2.775 log p + ( 24.418 @xmath0 0.048 ) + i = -2.977 log p + ( 23.833 @xmath0 0.042 ) + wi = -3.000 log p + ( 22.908 @xmath0 0.051 ) + the linear regressions to our v , i , and wi cepheid pl relations are shown in fig . 6 and 7 . in the case of the wi index we excluded four more widely oustanding objects : cep033 , cep057 , cep078 and cep082 .
all of them have either very blue or very red colors , what coused their abnormal locations on the wi pl relation . assuming the schlegel et al .
( 1998 ) reddening law , in agreement with our previous work in the araucaria project , a foreground reddening towards ngc 3109 of e(v - i ) = 0.05 mag ( minniti , zijlstra and alonso 1999 ) , and adopting a true lmc distance modulus of 18.5 , we derive from the above zero points true distance moduli of ngc 3109 of 25.72 @xmath0 0.05 mag in the v band , and 25.66 @xmath0 0.04 in the i band .
we adopt the distance modulus of 25.54 @xmath0 0.05 mag , derived from the reddening - free wesenheit magnitudes as our best result
. as one can see it is significantly shorter than the corresponding distance moduli obtained based on the v and i band magnitudes .
a straightforward explanation is that in addition to our adopted extinction ( e(b - v ) = 0.05 mag ) , we do have some contribution to the e(b - v ) related to the intrinsic ( to ngc 3109 ) extinction . indeed assuming an additional intrinsic reddening of 0.05 mag the distances derived for the v and i band agree within 0.02 mag with our adopted distance modulus from the wesenheit index .
we would like to note that we found an identical situation studying cepheid p - l relations in the optical v and i bands in ngc 300 ( gieren et al .
2004 ) and also postulated an additional intrinsic reddening as the most reasonable explanation .
recently gieren et al . ( 2005c ) measured the total extinction in this galaxy from combined optical and infrared data of cepheids and indeed confirmed this suspition .
we will precisely measure the total extinction of ngc 3109 once we will have analyzed follow - up near - infrared photometry of a subsample of the cepheids in this galaxy .
our adopted cepheid distance to ngc 3109 derived from optical photometry in the v and i bands is 25.54 @xmath0 0.05 mag ( statistical error ) .
this distance value compares very well to previous distance determinations of ngc 3109 from cepheid variables , and the trgb method .
the most recent determination from cepheids is the one by musella , piotto and capaccioli ( 1997 ) who analyzed bvri photometry of 24 cepheids and obtained a true distance modulus of 25.67 mag .
capaccioli et al .
( 1992 ) had obtained a distance of 25.50 @xmath0 0.16 mag , from b - band photographic photometry of 21 cepheids .
lee ( 1993 ) obtained a true distance modulus of 25.55 @xmath0 0.1 mag from the location of the i - band trgb in ngc 3109 , very similar to the more modern value of 25.61 @xmath0 0.1 mag of minniti , zijlstra and alonso ( 1999 ) .
another very recent determination from the trgb method gives 25.66 @xmath0 0.1 mag ( hidalgo et al .
, in preparation ) . as detailed in several of our previous papers in this series , for the scientific purposes of the araucaria project we are interested in obtaining a set of galaxy distances of the highest possible accuracy .
we have shown ( gieren et al . 2005c , bresolin et al .
2005 , pietrzyski et al .
2006 ) that by combining high - quality optical and near - infrared photometry of cepheids , distance determinations accurate to 3 percent seem possible , minimizing systematic problems with blending with unresolved companion stars , and most importantly with reddening , which seems to be the most critical single systematic factor which can bias distance determinations to nearby galaxies based on optical photometry alone .
we therefore prefer to postpone an exhaustive discussion of the various systematic uncertainties affecting our current distance result for ngc 3109 to a follow - up paper which will present near - infrared photometry of a subsample of the cepheids studies in the optical spectral range in this paper , and which will derive the ultimate , most accurate distance to ngc 3109 from its population of cepheid variables .
the principal value of the present paper is the presentation of a catalog of cepheid variables in this rather distant local group galaxy which increases the size of the known cepheid sample by almost a factor of 3 , presents high - quality light curves and periods for all these variables .
we have conducted a wide - field survey for cepheid variables in the local group galaxy ngc 3109 which covers the entire spatial content of the galaxy , and which is complete to cepheid periods down to about 4 days .
the survey was conducted in the v and i bands on 70 nights distributed over an interval of approximately one year .
we have discovered 113 cepheid variables with periods in the range from 31.4 to 3.4 days .
most of the previously known ( 45 ) cepheids were re - discovered in our survey ; for a number of these variables with first - epoch data we were able to improve very significantly on the periods with our data , and show that for 14 objects the previously determined periods were wrong .
we have used our data to construct period - luminosity relations in the v and i bands which are of very good quality . adopting a period - cutoff of log p(days ) = 0.75 , and the slopes of the lmc cepheid pl relations in v and i measured by udalski ( 2000 )
, we obtain a reddening - corrected distance modulus of 25.54 @xmath0 0.05 mag ( statistical error ) to ngc 3109 , which is in good agreement with previous distance results from optical work on smaller cepheid samples in this galaxy , and with determinations of the i - band magnitude of the tip of the red giant branch . as in the former papers of the araucaria project ,
this distance is tied to an assumed lmc distance modulus of 18.50 .
we are currently engaged in analyzing a subsample of the cepheids in ngc 3109 at near - infrared ( j and k ) wavelengths . with the advent of pl relations in j and k
, we expect to improve on the accuracy of our current , optical distance determination , mainly by improving on the value of the total reddening appropriate to ngc 3109 , and to achieve a distance determination whose accuracy is comparable to our previous work on ngc 300 ( gieren et al .
2005c ) and ic 1613 ( pietrzyski et al .
2006 ) .
we would like to thank the anonymous referee for his interesting suggestions .
wg and gp gratefully acknowledge financial support for this work from the chilean center for astrophysics fondap 15010003 .
support from the dst and bw grants for warsaw university observatory is also acknowledged .
gieren , w. , pietrzyski , g. , bresolin , f. , kudritzki , r .-
p . , minniti , d. , urbaneja , m. , soszyski , i. , storm , j. , fouqu , p. , bono , g. , walker , a.r . , and garca , j.a . , 2005a , the eso messenger , 121 , 23 ccccc cep001 & v & 2452343.545710 & 19.929 & 0.015 + cep001 & v & 2452344.548600 & 19.884 & 0.013 + cep001 & v & 2452345.561710 & 20.005 & 0.018 + cep001 & v & 2452346.558740 & 20.072 & 0.018 + cep001 & v & 2452347.554180 & 20.100 & 0.014 + cep001 & v & 2452349.556220 & 20.136 & 0.018 + cep001 & v & 2452350.539630 & 20.167 & 0.017 + cep001 & v & 2452351.544490 & 20.220 & 0.017 + cep001 & v & 2452353.541320 & 20.367 & 0.026 + cep001 & v & 2452354.545400 & 20.423 & 0.029 + c c c l c c c c cep001 & 10:03:21.84 & -26:09:52.9 & 31.4793 & 2337.913830 & 20.463 & 19.512 & v14 + cep002 & 10:03:07.83 & -26:09:54.5 & 31.270 & 2321.649470 & 20.174 & 19.373 & + cep003 & 10:03:14.66 & -26:09:23.0 & 29.110 & 2337.636330 & 20.677 & 19.753 & + cep004 & 10:03:19.50 & -26:10:12.8 & 27.389 & 2319.596160 & 20.449 & 19.644 & + cep005 & 10:03:10.99 & -26:09:25.8 & 26.8274 & 2337.755930 & 20.643 & 19.680 & v42 + cep006 & 10:02:27.01 & -26:10:13.1 & 22.3795 & 2331.567570 & 20.780 & 19.936 & v80 + cep007 & 10:03:01.79 & -26:09:28.8 & 20.388 & 2331.509800 & 21.727 & 20.552 & + cep008 & 10:03:43.38 & -26:10:45.4 & 19.9645 & 2343.055610 & 21.006 & 20.107 & v81 + cep009 & 10:02:57.57 & -26:09:55.5 & 19.5759 & 2342.458090 & 21.098 & 20.150 & v64 + cep010 & 10:03:28.20 & -26:10:20.5 & 17.889 & 2325.195440 & 20.809 & 20.043 & + cep011 & 10:03:00.16 & -26:10:36.1 & 17.2293 & 2340.398770 & 21.384 & 20.462 & v34 + cep012 & 10:02:53.53 & -26:09:37.0 & 14.750 & 2332.001570 & 21.357 & 20.351 & + cep013 & 10:02:46.36 & -26:09:56.9 & 14.6980 & 2336.741810 & 21.266 & 20.521 & v5 + cep014 & 10:03:02.97 & -26:09:13.0 & 14.062 & 2335.413040 & 21.126 & 99.999 & + cep015 & 10:03:18.39 & -26:10:31.5 & 14.030 & 2337.973110 & 21.305 & 20.532 & + cep016 & 10:03:10.58 & -26:11:07.6 & 13.9047 & 2335.674300 & 20.967 & 20.178 & v8 + cep017 & 10:02:59.18 & -26:09:29.1 & 13.6250 & 2330.266540 & 21.372 & 20.597 & p3 + cep018 & 10:03:10.86 & -26:09:38.6 & 13.364 & 2336.635230 & 21.385 & 20.583 & + cep019 & 10:03:41.49 & -26:08:45.9 & 13.3126 & 2335.877720 & 21.790 & 20.534 & v92 + cep020 & 10:02:55.42 & -26:09:13.7 & 13.011 & 2341.364340 & 21.740 & 20.869 & + cep021 & 10:03:05.67 & -26:08:38.5 & 12.6891 & 2340.566180 & 21.590 & 20.696 & v65 + cep022 & 10:02:58.65 & -26:09:11.0 & 11.864 & 2333.079460 & 20.996 & 20.291 & + cep023 & 10:02:54.37 & -26:09:19.2 & 11.707 & 2333.657310 & 21.248 & 20.412 & + cep024 & 10:02:52.98 & -26:08:24.5 & 11.6577 & 2335.671760 & 21.540 & 20.821 & v69 + cep025 & 10:02:55.00 & -26:08:56.2 & 11.596 & 2335.052600 & 21.184 & 19.824 & + cep026 & 10:03:05.25 & -26:08:58.7 & 11.534 & 2338.752600 & 21.959 & 21.228 & + cep027 & 10:02:58.24 & -26:08:48.2 & 11.238 & 2333.192590 & 21.687 & 20.807 & + cep028 & 10:03:16.97 & -26:09:04.7 & 11.0663 & 2341.410650 & 21.547 & 20.754 & v47 + cep029 & 10:03:08.68 & -26:10:04.2 & 10.903 & 2339.496070 & 21.071 & 20.462 & + cep030 & 10:03:11.39 & -26:09:49.2 & 10.887 & 2340.595580 & 21.825 & 21.070 & + cep031 & 10:02:57.81 & -26:09:49.8 & 10.851 & 2343.875650 & 21.479 & 20.659 & + cep032 & 10:03:08.44 & -26:09:40.3 & 10.826 & 2342.752810 & 21.108 & 20.200 & + cep033 & 10:02:51.05 & -26:09:29.6 & 10.625 & 2336.145430 & 21.294 & 21.219 & + cep034 & 10:02:43.74 & -26:09:06.0 & 10.618 & 2333.751730 & 21.746 & 20.915 & + cep035 & 10:03:21.40 & -26:08:43.3 & 9.7501 & 2336.288160 & 21.652 & 20.704 & + cep036 & 10:03:10.75 & -26:09:38.6 & 9.4084 & 2335.443150 & 21.324 & 20.770 & + cep037 & 10:03:51.99 & -26:10:54.0 & 9.3920 & 2334.622060 & 21.545 & 20.923 & + c c c l c c c c cep038 & 10:03:18.31 & -26:09:05.9 & 9.3242 & 2340.354910 & 21.809 & 20.959 & + cep039 & 10:03:02.24 & -26:09:20.2 & 9.11136 & 2337.674660 & 21.842 & 21.146 & v72 + cep040 & 10:02:41.23 & -26:09:14.7 & 9.0687 & 2337.266440 & 21.831 & 21.239 & + cep041 & 10:02:40.38 & -26:08:57.8 & 9.0686 & 2341.630130 & 21.850 & 21.340 & + cep042 & 10:03:13.70 & -26:08:38.5 & 8.87692 & 2334.110570 & 21.740 & 21.122 & v44 + cep043 & 10:03:23.30 & -26:10:10.8 & 8.7970 & 2342.435330 & 21.778 & 20.652 & + cep044 & 10:02:33.50 & -26:08:48.9 & 8.7726 & 2334.151600 & 21.891 & 21.161 & v2 + cep045 & 10:03:16.16 & -26:07:57.8 & 8.6479 & 2341.346220 & 21.388 & 20.522 & + cep046 & 10:03:37.95 & -26:10:39.7 & 8.55646 & 2342.433730 & 21.692 & 21.001 & v18 + cep047 & 10:03:15.01 & -26:10:10.3 & 8.55708 & 2337.332320 & 21.417 & 20.905 & v45 + cep048 & 10:03:08.96 & -26:10:07.2 & 8.4791 & 2339.496390 & 21.644 & 21.190 & + cep049 & 10:03:41.88 & -26:09:43.2 & 8.26911 & 2343.627340 & 21.969 & 21.198 & v20 + cep050 & 10:03:18.62 & -26:10:40.5 & 8.24330 & 2338.772320 & 21.772 & 21.075 & v79 + cep051 & 10:03:00.55 & -26:10:42.1 & 8.19173 & 2337.470680 & 21.925 & 21.099 & v35 + cep052 & 10:03:14.48 & -26:08:55.3 & 8.10451 & 2343.965380 & 21.991 & 21.214 & v11 + cep053 & 10:03:15.59 & -26:10:18.2 & 7.9351 & 2342.029060 & 21.944 & 21.269 & v9 + cep054 & 10:03:20.08 & -26:10:27.1 & 7.9038 & 2342.277150 & 21.679 & 20.959 & + cep055 & 10:03:16.56 & -26:08:58.7 & 7.7960 & 2339.853750 & 21.924 & 21.031 & + cep056 & 10:02:57.41 & -26:09:11.6 & 7.77376 & 2340.412360 & 21.712 & 20.772 & p5 + cep057 & 10:03:08.82 & -26:09:33.8 & 7.5389 & 2338.974770 & 21.850 & 20.554 & + cep058 & 10:03:00.67 & -26:09:10.5 & 7.4089 & 2340.271440 & 21.736 & 21.102 & + cep059 & 10:02:55.95 & -26:10:20.4 & 7.3858 & 2336.028380 & 22.293 & 21.521 & + cep060 & 10:03:21.39 & -26:10:42.6 & 7.2814 & 2340.586460 & 22.007 & 21.432 & + cep061 & 10:03:15.41 & -26:09:41.9 & 7.25995 & 2337.447310 & 21.174 & 20.370 & v57 + cep062 & 10:02:52.62 & -26:09:34.2 & 7.2186 & 2337.754860 & 22.445 & 21.760 & + cep063 & 10:02:59.25 & -26:09:06.7 & 7.15940 & 2340.268610 & 21.654 & 20.760 & v6 + cep064 & 10:03:05.80 & -26:10:08.7 & 7.13061 & 2339.841040 & 21.917 & 21.343 & v36 + cep065 & 10:03:24.84 & -26:10:24.4 & 7.0137 & 2337.913530 & 22.213 & 21.288 & + cep066 & 10:03:03.88 & -26:10:08.9 & 6.8511 & 2337.446440 & 21.849 & 21.148 & + cep067 & 10:03:04.20 & -26:09:40.8 & 6.8294 & 2341.764590 & 22.410 & 21.359 & + cep068 & 10:03:10.06 & -26:10:28.4 & 6.6932 & 2338.281130 & 21.894 & 21.093 & + cep069 & 10:02:59.33 & -26:08:52.4 & 6.6735 & 2340.994290 & 22.180 & 21.029 & p13 + cep070 & 10:02:44.35 & -26:07:28.2 & 6.5140 & 2341.722300 & 22.355 & 21.649 & + cep071 & 10:02:50.10 & -26:09:04.2 & 6.4676 & 2340.656380 & 22.132 & 21.423 & + cep072 & 10:03:01.45 & -26:08:55.7 & 6.4250 & 2336.815350 & 22.013 & 21.309 & p12 + cep073 & 10:02:55.83 & -26:08:53.4 & 6.3065 & 2341.222110 & 22.064 & 21.474 & + cep074 & 10:03:09.17 & -26:09:40.5 & 6.2719 & 2337.864370 & 22.102 & 21.403 & + cep075 & 10:03:04.54 & -26:09:46.9 & 6.2713 & 2340.553600 & 22.747 & 21.615 & + c c c l c c c c cep076 & 10:03:18.76 & -26:09:08.8 & 6.16368 & 2336.988860 & 22.166 & 21.407 & v46 + cep077 & 10:03:05.10 & -26:09:05.7 & 6.1329 & 2338.014500 & 21.809 & 21.222 & + cep078 & 10:03:01.33 & -26:09:03.5 & 6.1098 & 2336.930330 & 22.277 & 21.866 & p10
+ cep079 & 10:03:11.02 & -26:10:11.9 & 6.0795 & 2338.452730 & 22.015 & 21.243 & + cep080 & 10:03:03.39 & -26:08:54.4 & 6.0749 & 2342.638130 & 22.398 & 21.637 & + cep081 & 10:03:35.64 & -26:10:01.7 & 6.0363 & 2341.558610 & 22.336 & 21.449 & + cep082 & 10:03:02.23 & -26:08:28.4 & 5.9876 & 2345.016240 & 23.222 & 21.324 & + cep083 & 10:03:02.20 & -26:08:29.1 & 5.98812 & 2337.986850 & 22.139 & 21.348 & v7 + cep084 & 10:03:13.02 & -26:08:53.6 & 5.9343 & 2337.868200 & 22.681 & 21.606 & + cep085 & 10:03:34.92 & -26:09:38.6 & 5.90434 & 2342.150450 & 22.027 & 21.267 & v77 + cep086 & 10:03:51.68 & -26:08:08.1 & 5.9029 & 2337.458290 & 22.557 & 21.684 & + cep087 & 10:03:12.09 & -26:08:29.9 & 5.8967 & 2343.837790 & 22.330 & 21.486 & + cep088 & 10:03:05.58 & -26:09:39.7 & 5.5369 & 2341.694860 & 22.206 & 20.822 & + cep089 & 10:03:24.52 & -26:08:54.9 & 5.4940 & 2342.016820 & 22.895 & 22.223 & + cep090 & 10:03:07.53 & -26:10:17.1 & 5.4844 & 2338.056080 & 22.135 & 21.456 & + cep091 & 10:02:56.16 & -26:08:36.1 & 5.4031 & 2338.409700 & 22.487 & 21.661 & + cep092 & 10:02:56.00 & -26:10:37.2 & 5.39909 & 2339.090920 & 22.428 & 21.787 & p1 + cep093 & 10:03:26.68 & -26:10:27.2 & 5.2624 & 2341.892040 & 22.419 & 21.640 & + cep094 & 10:03:09.40 & -26:10:17.6 & 5.2442 & 2338.532080 & 22.748 & 21.910 & + cep095 & 10:03:18.92 & -26:09:14.8 & 5.2420 & 2339.558630 & 22.046 & 21.144 & + cep096 & 10:03:20.83 & -26:09:33.1 & 5.2163 & 2340.797600 & 22.080 & 21.278 & + cep097 & 10:02:55.50 & -26:10:05.5 & 5.17264 & 2342.796630 & 22.216 & 21.592 & p2 + cep098 & 10:03:17.06 & -26:09:05.8 & 5.0493 & 2343.287550 & 21.938 & 21.223 & + cep099 & 10:02:58.38 & -26:09:10.5 & 5.03650 & 2339.717280 & 21.733 & 21.164 & p6 + cep100 & 10:02:58.59 & -26:09:49.7 & 4.9883 & 2341.693950 & 21.918 & 21.861 & + cep101 & 10:03:05.41 & -26:08:59.6 & 4.9221 & 2337.756730 & 21.983 & 21.436 & + cep102 & 10:03:15.00 & -26:08:59.1 & 4.7765 & 2342.741240 & 22.182 & 21.432 & + cep103 & 10:02:54.09 & -26:08:57.7 & 4.5808 & 2339.387750 & 22.165 & 21.255 & + cep104 & 10:02:53.62 & -26:09:54.1 & 4.5658 & 2341.226270 & 22.203 & 21.065 & + cep105 & 10:03:14.83 & -26:10:03.1 & 4.5120 & 2340.292010 & 22.575 & 21.742 & + cep106 & 10:03:08.03 & -26:08:52.2 & 4.4919 & 2341.467330 & 21.787 & 20.590 & + cep107 & 10:02:44.12 & -26:09:07.4 & 4.4613 & 2343.882180 & 21.854 & 21.312 & + cep108 & 10:02:55.76 & -26:09:34.4 & 4.0877 & 2341.161440 & 22.755 & 22.258 & + cep109 & 10:03:46.95 & -26:10:25.0 & 4.036 & 2344.285480 & 22.463 & 22.065 & + cep110 & 10:03:18.92 & -26:09:10.7 & 3.9641 & 2341.545940 & 22.177 & 21.372 & + cep111 & 10:03:02.45 & -26:08:32.8 & 3.9604 & 2343.848810 & 21.936 & 20.898 & + cep112 & 10:02:46.74 & -26:09:33.6 & 3.8455 & 2341.021000 & 22.421 & 21.192 & + cep113 & 10:02:59.50 & -26:08:36.7 & 3.4016 & 2340.045330 & 23.049 & 21.946 & p14 + | we have obtained mosaic images of ngc 3109 in the v and i bands on 74 nights , spanning approximately one year . from these data
, we have conducted an extensive search for cepheid variables over the entire field of the galaxy , resulting in the discovery of 113 variables with periods ranging from 3.4 to 31.4 days . in this sample , 76 cepheids , including many long - period variables , were not known before . for the previously known 45 cepheids in this galaxy , our data proved that reported periods were wrong for 14 objects ; for nearly all other previously known cepheid variables we were able to significantly improve on the periods .
we construct period - luminosity relations from our data and obtain reddening - corrected distance moduli of 25.72 @xmath0 0.05 mag in v , and 25.66 @xmath0 0.04 mag in i. the distance modulus derived form the reddening - independent v - i wesenheit index turns out to be significantly shorter ( 25.54 @xmath0 0.05 mag ) , which indicates that in addition to the foreground extinction of e(b - v ) = 0.05 mag , there is an intrinsic to ngc 3109 redening of about 0.05 mag .
our distance obtained based on the reddening - free wesenheit magnitudes is consistent with earlier distance determinations of ngc 3109 from cepheids , and the tip of the red giant branch .
we will improve on our distance and extinction determination combining our optical data with the follow - up near - infrared observations of a subsample of ngc 3109 cepheids . |
forecast of various time series has been widely using in many fields of science and practice and many methods have been advanced for prediction of time series .
but common problem of each method is a priori estimate of its accuracy .
common practice is to take a truncated series ( reference series ) ended in the past , investigate its statistical parameters , build prediction and compare it with existing continuation of series under investigation .
using moving shift of reference series one can collect needed statistics and obtain estimate of accuracy of used method depending on length of prediction .
after that obtained accuracy is assigned to real predictions . in most of predicted time series last observed point ( epoch ) preceding the first predicted one has its final value and is not subject of refinement in future ( e.g. , number of sunspots for some epoch ) .
it is not the case for eop .
all real eop predictions are based on operational solution that may differ substantially from final eop values that comes usually in one - two months . as an illustration of the foregoing ,
let s consider fig . 1 . in this figure @xmath0 the epoch of beginning of prediction , @xmath1 operational eop series used for computation of prediction , @xmath2 prediction computed in real time ( at date @xmath0 ) , @xmath3 final eop series , @xmath4 prediction that would be computed from the final series if it were available at date @xmath0
. upon this circumstance , accuracy and other statistical parameters of operational eop series may differ substantially from ones for final series that are commonly used for a priori estimate of prediction accuracy .
it means that estimates of prediction accuracy obtained by `` standard '' method may be far from reality and some its modification ( or at least investigation ) is desirable .
evidently , the most simple way to make a priory estimate of accuracy of prediction is a fictive disturbances of one or more last points of reference interval to investigate reaction of given method of prediction on errors at the last observed epoch(s ) .
this test was realized in ( malkin & skurikhina 1996 ) .
two kind of fictive errors was applied to real observed points : _ test 1 _ :
the value of 1 mas was added to ( or subtract from ) the c04 value corresponding to the last observed epoch . _
test 2 _ : the values of 0.5 , 1.0 , 1.5 mas were added to ( or subtract from ) the c04 value corresponding to the three last observed epoch .
this test was used only for arima method because its influence on the extrapolation of trend - harmonics model ( e.g. , mccarthy & luzum 1991 ) can be easily foreseen without special calculations .
typical differences between predictions of real and distorted c04 series are presented in table [ tab : pred_xy_dist ] .
one can see that serious degradation of accuracy may occur when arima method is used for erroneous observed eop values .
it should be mentioned that this effect practically linearly depends on the value of error .
analogous results were obtained for prediction of ut ( malkin & skurikhina 1996 ) .
.influence of errors in the last values on prediction results .
[ cols="^ , > , > , > , > , > , > , > , > " , ] [ tab : pred2 m ]
in this paper we have attempted to estimate real accuracy of predictions of eop using real - time predictions made at the iaa and the usno has been also made .
although collected statistics is too poor to make more or less final conclusions , we can state that : * estimate of accuracy of prediction based on use of old data is not adequate to accuracy of real - time prediction , especially for short - time prediction .
a modification of commonly used method of a priori estimate of accuracy , e.g. , proposed in ( malkin & skurikhina 1996 ) can give more realistic estimates .
* accuracy of methods of prediction of eop used at the iaa and the usno is approximately the same .
more detailed conclusion can be made only after collecting supplement statistics .
* estimate of both rms and maximal errors in prediction is very useful for potential users .
it seems reasonable to provide such estimates for iers and other prediction series . | to estimate real accuracy of eop prediction real - time predictions made by the iers subbureau for rapid service and prediction ( usno ) and at the institute of applied astronomy ( iaa ) eop service are analyzed .
methods of a priory estimate of accuracy of prediction are discussed . |
the study of axially symmetric solutions of the einstein and einstein - maxwell field equations corresponding to disklike configurations of matter , apart from its purely mathematical interest , has a clear astrophysical relevance .
indeed , thin disks can be used to model accretion disks , galaxies in thermodynamical equilibrium and the superposition of a black hole and a galaxy .
disk sources for stationary axially symmetric spacetimes with magnetic fields are also of astrophysical importance mainly in the study of neutron stars , white dwarfs and galaxy formation .
exact solutions that have relativistic static thin disks as their sources were first studied by bonnor and sackfield @xcite and morgan and morgan @xcite .
subsequently , several classes of exact solutions corresponding to static @xcite and stationary @xcite thin disks have been obtained by different authors , and the superposition of a static or stationary thin disk with a black hole has been considered@xcite .
relativistic disks embedded in an expanding frw universe have been studied in @xcite , perfect fluid disks with halos in @xcite , and the stability of thin disks models has been investigated using a first order perturbation of the energy - momentum tensor in @xcite . on the other hand ,
thin disks have been discussed as sources for kerr - newman fields @xcite , magnetostatic axisymmetric fields @xcite , and conformastatic and conformastationary metrics @xcite . also , models of electrovacuum static counterrotating dust disks were presented in @xcite , charged perfect fluid disks were studied in @xcite , and charged perfect fluid disks as sources of static and taub - nut - type spacetimes in @xcite .
now , the thin disks with magnetic fields presented at references @xcite were obtained by means of the well known ` displace , cut and reflect ' method in order to introduce a discontinuity at the first derivative of one otherwise smooth solution .
the result is a solution with a singularity of the delta function type in all the @xmath0 hypersurface and so can be interpreted as an infinite thi disk . on the other hand ,
solutions that can be interpreted as thin disks of finite extension can be obtained if a proper coordinate system is introduced . a coordinate system that adapts naturally to a finite source and presents the required discontinuous behavior is given by the oblate spheroidal coordinates .
some examples of finite thin disks from vacuum solutions expressed in these coordinates can be found in references @xcite , and from electrovacuum solutions in reference @xcite . according to the above considerations , the purpose of our paper is to present a new infinite family of exact solutions of the einstein - maxwell equations for axially symmetric spacetimes .
the solutions are obtained by the use of the horsk - mitskievitch conjecture @xcite , which prescribes a quite close connection between isometries of vacuum spacetimes ( seed metrics ) and the electromagnetic four - potential of a generated einstein - maxwell fields .
we take the morgan and morgan metric disk @xcite as the seed solution .
the generated solutions describe a family of magnetized finite thin disks , which is the magnetized version of the family of relativistic static morgan and morgan disks .
the plan of our paper is as follows .
first , in section [ sec : einm ] , the weyl - lewis - papapetrou line element is considered and einstein - maxwell equations in cylindrical coordinates are introduced . the procedure to obtain magnetovacuum static , axially symmetric relativistic thin disks without radial pressure is also summarized in this section .
section [ sec : hm ] introduces the horsk - mitskievitch generating conjecture used to obtain a general kind of weyl - lewis - papapetrou spacetime with magnetic field . in section [ sec : mmm ] , we put forward a solution of the einstein - maxwell equations describing an infinite family of finite static magnetized thin disks by the use of the horsk - mitskievitch ( hm ) conjecture .
for this purpose , the well known morgan and morgan metric disk @xcite is employed as seed solution @xmath1 .
the analysis of the physical behavior of solutions is presented in section [ sec : beh ] where the asymptotic behavior of the solutions is examined .
then , we study the behavior of the corresponding energy density , current density , circular velocity and azimuthal pressure
. finally , in section [ sec : conc ] , we conclude with a discussion of our results .
the vacuum einstein - maxwell equations , in geometrized units such that @xmath2 , can be written as @xmath3 with the electromagnetic energy - momentum tensor given by @xmath4 where @xmath5 is the electromagnetic field tensor and @xmath6 is the electromagnetic four potential .
the line element of a static vacuum spacetime can be written in the weyl - lewis - papapetrou ( wlp ) form@xcite @xmath7 , \label{eq : met0 } \end{aligned}\ ] ] where @xmath8 are the usual cylindrical coordinates : @xmath9 , @xmath10 , @xmath11 , the metric functions @xmath12 and @xmath13 depend only on @xmath14 and @xmath15 .
the solutions of the einstein - maxwell equations corresponding to a disklike source are even functions of the @xmath15 coordinate .
therefore , they are continuous functions everywhere but their first @xmath15-derivatives discontinuous at the disk surface . consequently , in order to obtain the energy - momentum tensor and the current density of the source , the jump across the disk of the first @xmath15-derivatives of the metric tensor is expressed as @xmath16 = 2 { g_{ab , z}}|_{_{z = 0^+}},\end{aligned}\ ] ] and the jump across the disk of the electromagnetic field tensor is expressed as @xmath17 = [ a_{a , z } ] = 2 { a_{a , z}}|_{_{z = 0^+}},\end{aligned}\ ] ] where the reflection symmetry of the functions with respect to @xmath0 has been used . then , using the distributional approach @xcite or the junction conditions on the extrinsic curvature of thin shells @xcite , the einstein - maxwell equations yield an energy - momentum tensor as @xmath18 + q^{ab } \delta(z ) , \label{eq : emtot}\ ] ] and the current density as @xmath19 where @xmath20 and @xmath21 are , respectively , the heaviside and dirac distributions with support on @xmath0 . here
@xmath22 are the electromagnetic energy - momentum tensors as they are defined by ( [ eq : emtensor ] ) in the @xmath23 and @xmath24 regions , respectively , whereas @xmath25 gives the part of the energy - momentum tensor corresponding to the disk source , and @xmath26 \label{eq : currenti}\ ] ] is the contribution of the disk source to the current density .
the surface energy - momentum tensor , @xmath27 , and the surface current density , @xmath28 , of the disk , can be obtained through the relations @xmath29 and @xmath30 where @xmath31 is the physical measurement of length in the normal direction of the disk .
now , we choose the magnetic potential as @xmath32 so that we have a pure axially symmetric magnetic field @xmath33 thus , for the metric ( [ eq : met0 ] ) , the only non - zero components of @xmath34 and @xmath35 are @xmath36 where all the quantities are evaluated at @xmath37 . with the orthonormal tetrad @xmath38 where @xmath39 we can write the surface energy - momentum tensor of the disk in the canonical form as @xmath40 here @xmath41 and @xmath42 are the energy density and the azimuthal pressure of the disk , respectively .
we also consider the mass density on the surface of the disk , defined as @xmath43 in the same way , the current density on the disk can be written as @xmath44 as we can see the charge density on the disk surface is identically equal to zero .
this can be checked by using eqs .
( [ eq : currenti ] ) , ( [ eq : currentj ] ) , ( [ eq : empot ] ) and ( [ eq : currentet ] ) .
the electromagnetic field is of a magnetic type , which can be demonstrated by the electromagnetic invariant @xmath45 \geq 0 .
\label{eq : hp2}\end{aligned}\ ] ] we consider now , on the basis of refs .
@xcite and @xcite , the possibility that the energy - momentum tensor @xmath46 and the current density @xmath47 can be interpreted as the superposition of two counterrotating fluids .
in order to do this , we cast @xmath48 where @xmath49 are the energy densities and charge densities of the two counterrotating fluids .
the counterrotating velocity vectors are given by @xmath50 where @xmath51 is the counterrotating tangential velocity .
therefore , we have two counterrotating charged fluids with equal energy densities and equal but opposite charge densities .
the wlp line element of a static vacuum spacetime can be written as @xcite @xmath52 , \label{eq : met}\end{aligned}\ ] ] where @xmath53 are the usual cylindrical coordinates : @xmath54 , @xmath10 , @xmath11 .
the metric functions @xmath1 and @xmath55 depend on @xmath14 and @xmath15 .
this metric admits two killing vectors : @xmath56 which , in the vectorial space @xmath57 are , @xmath58 or , in terms of their components are @xmath59 where @xmath60 is the usual kronecker delta tensor . in terms of the one - forms @xmath61 and @xmath62
we have @xmath63 where all the orthonormal bases employed are always chosen as a generalization of the set @xmath64 the simplest tetrad that can be used for the weyl solution .
the present section is devoted to the derivation of solutions of the einstein - maxwell equations via hm conjecture .
following the method used in @xcite , we modify the line element ( [ eq : met ] ) to the form @xmath65 that admits two killing vectors @xmath66 with @xmath67 an arbitrary function .
the killing vectors @xmath68 and the electromagnetic four - potential @xmath69 satisfy @xmath70 where the @xmath71 is the usual hodge ( star ) operation .
then , the killing vector @xmath72 induces the four - potential @xmath73 or , @xmath74 it can be verified through standard calculations that the sourceless einstein - maxwell equations are fulfilled if @xmath75 where the function @xmath76 must be a solution of the differential equation @xmath77 given the fact , for electrovacuum spacetimes , the einstein tensor , @xmath78 , is traceless . as in @xcite
we can choose @xmath79 and @xmath80 , and then we have the four - potential @xmath81 the @xmath82 parameter characterizing the strenght of the electromagnetic field . under the conditions described above the metric set by ( [ eq : met2 ] )
is reduced to @xmath83 , \label{eq : met3}\end{aligned}\ ] ] where,@xmath84 and @xmath85 . using ( [ eq : emt22 ] ) ,
( [ eq : tet ] ) and ( [ eq : emt2 ] ) we find out the surface energy density @xmath86 the azimuthal pressure @xmath87\phi_{s , z } } { ( e^{2\phi_s } + q^2r^2)^3}\label{eq : pressure},\end{aligned}\ ] ] the surface mass density @xmath88 the surface current density @xmath89 and the circular velocity @xmath90 \label{eq : velocity},\end{aligned}\ ] ] where @xmath91 , and @xmath92 is the circular velocity of the unmagnetized source , i.e. @xmath93 .
as we can see from ( [ eq : kil1 ] ) , ( [ eq : kil2 ] ) , ( [ eq : kil1a ] ) and ( [ eq : kil2a ] ) , the method is a solution generating technique that changes the norm of the two killing vectors without adding twist . besides
, the possible connection of the hm conjecture with inner symmetries of einstein - maxwell equations would connect the conjecture with a set of generating methods elaborated by ernst and other authors for axisymmetric fields @xcite .
let s restrict the previous general model to obtain a solution of the einstein - maxwell equations describing an infinite family of finite static magnetized thin disks . for this purpose ,
we use , as a seed solution @xmath94 , the well known morgan and morgan metric disk @xmath95 @xcite , @xmath96 where @xmath97 are arbitrary constants that must be properly specified so that a particular solution may be set . @xmath98
are the usual legendre polynomials and @xmath99 .
@xmath100 being the legendre functions of second kind ( see @xcite and , for the legendre functions of imaginary argument , @xcite pag .
the @xmath101 and @xmath102 are the oblate spheroidal coordinates related with the cylindrical coordinates by the relations @xcite
@xmath103 where @xmath104 and @xmath105 .
the disk has coordinates @xmath106 , @xmath107 and , when the disk is crossed , the sign of @xmath102 changes , but not its absolute value .
we use the constants @xmath97 as they have been determined by gonzlez and reina in @xcite @xmath108 , \label{recu}\end{aligned}\ ] ] for @xmath109 and @xmath110 for @xmath111 .
consequently , these solutions correspond to the magnetized version of the well known morgan - morgan disk @xcite .
for instance , the first three members of the family of the seed metric functions are given by @xmath112 , \label{eq:4.22 } \\
\phi_2(x , y ) & = & - \frac{mg}{a } [ \cot^{-1 } x + \frac{10 a}{7 } ( 3y^{2 } - 1 ) + \ b ( 35 y^{4 } - 30 y^{2 } + 3 ) ] , \label{eq:4.23 } \\
\phi_3(x , y ) & = & - \frac{mg}{a } [ \cot^{-1 } x + \frac{10 a}{6 } ( 3 y^{2 } - 1 ) + \label{eq:4.24 } \\ & & \frac{21 b}{11 } ( 35 y^{4 } - 30 y^{2 } + 3 ) + \ c ( 231 y^{6 } - 315 y^{4 } + 105 y^{2 } - 5 ) ] , \nonumber\end{aligned}\ ] ] where @xmath113 , \\ & & b = \frac{3}{448 } [ ( 35 x^{4 } + 30 x^{2 } + 3 ) \cot^{-1 } x - 35 x^{3 } - \frac{55}{3 } x ] , \\ & & c = \frac{5}{8448 } [ ( 231 x^{6 } + 315 x^{4 } + 105 x^{2 } + 5 ) \cot^{-1 } x - 231 x^{5 } - 238 x^{3 } - \frac{231}{5 } x ] , \end{aligned}\ ] ] which similar , but more involved , expressions for greater values of @xmath114 . on the other hand , according to ( [ eq : energy ] ) , the energy density of the magnetized disk can be written as @xmath115\phi_{n , x } } { ay^2[e^{2\phi_n } + q^2a^2(1 -y^2)]^3},\end{aligned}\ ] ] with @xmath116 while , according to ( [ eq : pressure ] ) and ( [ eq : mass ] ) the pressure , @xmath42 , and mass density , @xmath117 , in the surface of the disk are @xmath118 } { ay^2[e^{2\phi_n } + q^2a^2(1 -y^2)]^3},\end{aligned}\ ] ] and @xmath119 ^ 2},\end{aligned}\ ] ] respectively , and , according to ( [ eq : current ] ) and ( [ eq : velocity ] ) , we have the following surface current density and circular velocity @xmath120 ^ 2},\end{aligned}\ ] ] and @xmath121,\end{aligned}\ ] ] respectively , where @xmath122 is the circular velocity of the unmagnetized source , i.e. @xmath93 . the non zero components of the magnetic field on the surface of the disk , @xmath123 and @xmath124 , are @xmath125 } { y[e^{2\phi_n } + q^2a^2(1 - y^2)]^2},\end{aligned}\ ] ] and @xmath126 ^ 2},\end{aligned}\ ] ] where we have used the magnetic potential of ( [ eq : magpot ] ) , and all the quantities are evaluated on the disk .
in order to elucidate the behavior of the different particular models , firstly we introduce the dimensionless energy surface density on the disks , defined for @xmath127 as @xmath128 where the dimensionless radial variable @xmath129 has been introduced . in figure
[ fig : energy ] the dimensionless surface energy densities @xmath130 for the models corresponding to @xmath131 and @xmath132 are depicted . in each case , @xmath133 for @xmath134 with @xmath135 for different values of the parameter @xmath136 are shown .
first , we set @xmath137 , the bottom curve in each graphic , and then @xmath138 and @xmath139 ( dotted curve ) , the vacuum case . in all the cases
, it can be seen that the energy density is everywhere positive and that it vanishes at the edge of the disk .
the disks with higher values of @xmath140 and @xmath141 show an energy distribution that is more concentrated in the center and lower at the edge . it can be observed that the energy density in the central region of the disk decreases with the presence of a magnetic field .
the figure [ fig : pressure ] shows the dimensionless azimuthal pressure , @xmath142 for the models corresponding to @xmath131 and @xmath132 . as we see , in all the cases , the pressure is everywhere positive and rapidly increases as we move away from the center of the disk .
it reaches a maximum and , later , it rapidly decreases .
it can be observed that the presence of a magnetic field increases the pressure everywhere on the disk . otherwise , the maximum increases and moves away from the disk edge as @xmath140 increases .
it can be also observed that the surface current density , represented in figure [ fig : current ] , has a behavior similar to that of the pressure . in order to illustrate graphically the behavior of the circular velocities or rotation curves , we introduce the dimensionless quantity @xmath143 in figure [ fig : velocity ] we plot the dimensionless rotation curves for the models corresponding to @xmath131 and @xmath132 .
the circular velocity corresponding to @xmath144 is a monotonously increasing function of the radius . on the other hand , for @xmath145 and @xmath146 , the circular velocity increases from a value of zero at the center of the disks until it attains a maximum at a critical radius and then , it decreases to acquire a finite value at the edge of the disk .
it can also be seen that the value of the critical radius increases as the values of @xmath140 decreases and @xmath141 increases .
from which , we can conclude that the magnetic field decreases the circular velocity everywhere on the disk .
we study these solutions with other values of the parameters as well , but , in all the cases , we find a similar behavior .
we see that the magnitude of @xmath147 and @xmath148 are limited uniquely by the condition that the magnitude of the velocities of the disk can not exceed the speed of light .
the explicit calculation of the kretschmann invariant @xmath149 is too lengthy to be reproduced here .
a direct computation by the use of reduce shows that the first member of the family of the magnetized morgan - morgan disk , @xmath144 , has a singularity at the rim of the disk where the kretschmann invariant becomes infinite although its mass density is finite everywhere , when all the @xmath150 disks are regular .
obviously , this property in the curvature is , of course , inherited from the seed morgan and morgan metric disk ( see @xcite ) .
the solution of ( [ eq : met3 ] ) in terms of the seed solution ( morgan and morgan metric disk , @xmath151 ) can be written in the form @xmath152 - \phi_s \label{eq : hp}. \end{aligned}\ ] ] therefore , when we take @xmath153 clearly we see that , unlike the morgan and morgan seed solution , generated spacetimes are not asymptotically flat . consequently , although the technique described here produces a new class of solution of einstein - maxwell s equations and all the physical quantities exhibit an excellent behavior , the method offers very little insight into the physical implications of the metrics .
the reason of this non - flatness far away from the disk should be that the magnetic field itself does not decay at infinity as it appears to be easily expressible from ( [ eq : hp2 ] ) .
on the other hand , this still appears to be a viable solution of the einstein - maxwell system but , apparently , corresponds to magnetized disks immersed in a melvin magnetic universe .
dimensionless energy density @xmath130 as a function of @xmath154 for the first three disk models with @xmath155 . in each case , we plot @xmath133 for @xmath156 with @xmath135 for different values of the parameter @xmath141 .
first , we take @xmath137 , the bottom curve in each plot , and then @xmath157 and @xmath139 ( curve with dots ) , the vacuum case.,title="fig:",width=307 ] dimensionless energy density @xmath130 as a function of @xmath154 for the first three disk models with @xmath155 . in each case , we plot @xmath133 for @xmath156 with @xmath135 for different values of the parameter @xmath141 .
first , we take @xmath137 , the bottom curve in each plot , and then @xmath157 and @xmath139 ( curve with dots ) , the vacuum case.,title="fig:",width=307 ] dimensionless energy density @xmath130 as a function of @xmath154 for the first three disk models with @xmath155 . in each case , we plot @xmath133 for @xmath156 with @xmath135 for different values of the parameter @xmath141 .
first , we take @xmath137 , the bottom curve in each plot , and then @xmath157 and @xmath139 ( curve with dots ) , the vacuum case.,title="fig:",width=307 ] dimensionless azimuthal pressure @xmath158 as a function of @xmath154 for the first three disk models with @xmath155 . in each case , we plot @xmath159 for @xmath160 with @xmath135 for different values of the parameter @xmath148 .
first , we take @xmath137 , the top curve in each plot , and then @xmath157 and @xmath139 ( curve with dots ) , the vacuum case.,title="fig:",width=307 ] dimensionless azimuthal pressure @xmath158 as a function of @xmath154 for the first three disk models with @xmath155 . in each case , we plot @xmath159 for @xmath160 with @xmath135 for different values of the parameter @xmath148 .
first , we take @xmath137 , the top curve in each plot , and then @xmath157 and @xmath139 ( curve with dots ) , the vacuum case.,title="fig:",width=307 ] dimensionless azimuthal pressure @xmath158 as a function of @xmath154 for the first three disk models with @xmath155 . in each case , we plot @xmath159 for @xmath160 with @xmath135 for different values of the parameter @xmath148 .
first , we take @xmath137 , the top curve in each plot , and then @xmath157 and @xmath139 ( curve with dots ) , the vacuum case.,title="fig:",width=307 ] dimensionless circular velocity @xmath161 as a function of @xmath154 for the three first disk models with @xmath155 . in each case , we plot @xmath162 for @xmath160 with @xmath135 for different values of the parameter @xmath148 .
we first take @xmath137 , the top curve in each plot , and then @xmath157 and @xmath139 ( curve with dots ) , the vacuum case.,title="fig:",width=307 ] dimensionless circular velocity @xmath161 as a function of @xmath154 for the three first disk models with @xmath155 . in each case , we plot @xmath162 for @xmath160 with @xmath135 for different values of the parameter @xmath148 .
we first take @xmath137 , the top curve in each plot , and then @xmath157 and @xmath139 ( curve with dots ) , the vacuum case.,title="fig:",width=307 ] dimensionless circular velocity @xmath161 as a function of @xmath154 for the three first disk models with @xmath155 . in each case , we plot @xmath162 for @xmath160 with @xmath135 for different values of the parameter @xmath148 .
we first take @xmath137 , the top curve in each plot , and then @xmath157 and @xmath139 ( curve with dots ) , the vacuum case.,title="fig:",width=307 ] dimensionless surface current density @xmath163 as a function of @xmath154 for the first three disk models with @xmath155 . in each case , we plot @xmath164 for @xmath160 with @xmath135 for different values of the parameter @xmath148 .
first , we take @xmath137 , the top curve in each plot , and then @xmath165 and @xmath166,title="fig:",width=307 ] dimensionless surface current density @xmath163 as a function of @xmath154 for the first three disk models with @xmath155 . in each case , we plot @xmath164 for @xmath160 with @xmath135 for different values of the parameter @xmath148 .
first , we take @xmath137 , the top curve in each plot , and then @xmath165 and @xmath166,title="fig:",width=307 ] dimensionless surface current density @xmath163 as a function of @xmath154 for the first three disk models with @xmath155 . in each case , we plot @xmath164 for @xmath160 with @xmath135 for different values of the parameter @xmath148 .
first , we take @xmath137 , the top curve in each plot , and then @xmath165 and @xmath166,title="fig:",width=307 ]
we have presented an infinite family of new exact solutions of the vacuum einstein - maxwell equations for static and axially symmetric spacetimes .
the solutions describe an infinite family of magnetized finite thin disks , the magnetized version of the family of morgan and morgan relativistic thin disks @xcite .
the first member of the family of the magnetized morgan - morgan disk , @xmath144 , has a singularity at the rim of the disk where the kretschmann invariant becomes infinite , although its mass density is finite everywhere . whereas all the @xmath167 disks are regular . obviously , this property in the curvature is , of course , inherited from the seed morgan and morgan metric disk . unlike the morgan and morgan seed solution ,
the generated spacetimes are not asymptotically flat .
consequently , although the technique described here produces a new class of solution of einstein - maxwell s equations and all the physical quantities exhibit an excellent behavior and the energy - momentum tensor is in fully agreement with all the energy conditions , the method offers very little insight into the physical implications of the metrics . however , these are the first fully integrated exact solutions for this kind of magnetized thin disk sources .
they are important in so far as they represent a new family of exact solutions of the einstein - maxwell vacuum equations .
moreover , the outlined method may serve as a guideline to find more physically acceptable solutions in future works .
we pass on to the definition of the hodge @xmath71 ( star ) operation .
let s us @xmath168 a @xmath169dimensional manifold endowed with a metric @xmath170 .
the hodge @xmath71 ( star ) operation is a map : @xmath171 whose action , on the basis of the vector @xmath172 , is defined by @xcite @xmath173 for a @xmath174form @xmath175 we have @xmath176 where the totally anti - symetric tensor @xmath177 is @xmath178 and `` @xmath179 '' is the usual exterior product or wedge product . | applying the horsk - mitskievitch conjecture to the empty space solutions of morgan and morgan due to the gravitational field of a finite disk , we have obtained the corresponding solutions of the einstein - maxwell equations .
the resulting expressions are simply written in terms of oblate spheroidal coordinates and the solutions represent fields due to magnetized static thin disk of finite extension .
now , although the solutions are not asymptotically flat , the masses of the disks are finite and the energy - momentum tensor agrees with the energy conditions .
furthermore , the magnetic field and the circular velocity show an acceptable physical behavior . |
low dimensional systems are platforms for research into a variety of quantum behavior . however , the broader application of such devices in an engineering context is limited due to unpredictable local variations in the electrostatic landscape in the device s active area .
when considering their practical application it is important to study the statistical spread in device characteristics .
we have previously studied identical devices fabricated on a modulation doped heterostructure which show variations in their conductance properties @xcite . in this paper
we investigate the effect of gate size on two important quantum properties of split gate devices the conductance quantization @xcite and the occurrence of 0.7 anomaly @xcite .
the split gate is the simplest mesoscopic device that can be used to study how device behavior is affected by gate size .
the conductance through a split gate @xcite is quantized in multiples of @xmath0 as a function of the voltage applied to the gates @xcite , due to the formation of one - dimensional ( 1d ) subbands .
for an ideal 1d conductor , this does not depend on the length as long as transport remains ballistic .
the effect of split gate size can be investigated either by varying lithographic dimensions @xcite , or fabricating several split gates in close - proximity which act in series to modify the potential of a single 1d channel @xcite .
so far it has been shown that the split - gate voltage ( @xmath1 ) at which the conductance through the 1d channel is pinched off occurs closer to zero for longer and narrower devices @xcite .
additionally , the quality of conductance quantization degrades as the gate length increases @xcite .
this latter effect is related to the higher probability of encountering an impurity in the channel with longer / wider split gates , and fluctuations in the background disorder potential which modifies the potential landscape in the channel area .
some studies of the effect of split gate size have focused on the 0.7 anomaly @xcite , a conductance feature which occurs near @xmath2 @xcite which arises from enhanced electron interactions at low conductance .
a reduction in the conductance of the 0.7 anomaly for longer split gates has been reported @xcite .
another study using split gates in series has shown periodic modulations in the value of the 0.7 anomaly as a function of 1d channel length @xcite .
the origin of the 0.7 anomaly is currently debated , theories proposed for its occurrence include spontaneous spin polarization @xcite , the kondo effect @xcite , wigner crystallization @xcite , and inelastic scattering plus the local enhancement ( smeared van hove singularity ) of the 1d density of states @xcite .
we use a multiplexing scheme @xcite to measure 95 split gates of 7 different sizes on a gaas / algaas modulation - doped heterostructure .
this is the first study of the impact of split gate dimensions using data obtained from a large number of split gates fabricated on a single chip and measured during one cool down .
the array contains groups of devices , where the dimensions of each split gate in the group are the same , and dimensions are changed between groups .
measuring groups in this way provides enough statistical information to determine typical device properties .
our technique allows us to compare incremental changes in gate design .
this is important in the context of device developement , for example if it is necessary to design a device with a specific set of operating parameters , or find ways of improving a particular design .
the traditional approach to nanostructure measurement require many cool downs to build up the statistical information that we obtain in a single cool down .
we systematically compare the accuracy of conductance quantization between split gates of different dimension , and find a rapid reduction in quality of quantization as gate length increases .
we obtain average values of pinch off voltage and definition voltage as a function of gate length and width which show how the pinch off voltage occurs closer to zero for longer and narrower split gates , as expected from electrostatics @xcite . for longer gates
the voltage at which the 1d channel is defined also occurs closer to zero , and the 1d subband spacing reduces .
a key finding of this paper is that the background disorder potential is at least as important as the split gate dimension in governing the potential landscape in the 1d channel .
this is shown in three separate ways .
firstly , the spread in both the measured 1d subband spacing and the estimated 1d barrier curvature for split gates at fixed gate lengths exceed variations in the mean value of these parameters between devices of different length .
secondly , the strength of coupling between the split gates and the 1d channel does not monotonically increase with device length .
a monotonic increase is expected if one only considers the electric field generated by the gates themselves .
thirdly , changes in the 1d barrier curvature from device to device which indicates the length of the 1d channel do not depend on split gate length .
we also directly compare the 0.7 anomaly in devices of different gate length .
this is possible since for a non - interacting system the shape of the conductance trace depends on @xmath3 .
we use a technique developed in ref .
@xcite to remove the trivial geometric dependence from the conductance data , leaving differences that are only due to electron interactions .
our data are consistent with ref .
@xcite in that the 0.7 anomaly occurs at lower conductance values for devices with a shallower longitudinal barrier .
however , the present dataset is gathered from devices of different dimensions ( data in ref .
@xcite are from lithographically identical gates ) . by obtaining the same result from devices with various dimensions
we show that the electrostatic length of the 1d channel is the decisive factor governing the conductance of the 0.7 anomaly .
our data are obtained from 95 split gates of various dimensions located within a single array ( of total area @xmath4 mm@xmath5 ) .
two widths were chosen , @xmath6 and @xmath7 @xmath8 m [ the dimensions are indicated on a schematic diagram of a split gate in fig .
[ fig1](a ) ] . for both widths , 15 devices of each length
@xmath9 were measured , where @xmath10 , @xmath11 , and @xmath12 @xmath8 m .
one of the devices did not define a 1d channel ( @xmath13 @xmath8 m ) .
a further 6 devices with @xmath14 @xmath8 m were also measured .
the sample is fabricated on a gaas / algaas heterostructure in which the 2deg forms 90 nm below the surface of the wafer , separated from a 40 nm wide si - doped algaas layer by 40 nm of undoped algaas .
the carrier density and mobility are @xmath15 @xmath16 and @xmath17 cm@xmath5v@xmath19s@xmath19 , respectively , determined by measuring a hall bar on a nearby piece of the wafer .
the electron mean free path is 22 @xmath8 m .
the split gates are defined by electron - beam lithography .
other surface gates are defined by optical lithography , and all gates are metallized by thermally evaporating ti / au . a two - terminal
, constant voltage technique is used to measure the differential conductance through the split gates as a function of gate voltage , using an ac excitation voltage of 100 @xmath8v at 17 hz . all measurements are carried out at 1.4 k and @xmath20 t. the device measured here is different to that used in our previous work on multiplexed split gates @xcite .
the earlier work used an array of identical split gates , where both the length and width were 0.4 @xmath8 m .
the multiplexed array measured here contains split gates of 7 different length / width combinations and is fabricated on a higher mobility heterostructure .
this paper presents a self - contained story of the influence ( or lack thereof ) of gate length on conductance in 1d devices .
we investigate the effect of split gate dimensions on device yield , 1d conductance properties , the electrostatic potential profile and the 0.7 anomaly .
such a systematic study on many split gates with different dimensions and measured under identical conditions is the first of its kind .
( a ) typical trace of conductance @xmath21 as a function of the voltage applied to the split gate @xmath1 .
the vertical and horizontal dotted lines indicate the 1d definition voltage ( @xmath22 ) and conductance ( @xmath23 ) , respectively .
the left - hand arrow indicates pinch - off voltage @xmath24 .
the inset shows a schematic split gate device with dimensions @xmath9 and @xmath25 labeled .
source and drain ohmic contacts are marked s and d , respectively .
( b ) scatter plot of pinch - off voltage as a function of split - gate length @xmath9 .
the triangles ( circles ) represent data from devices of width @xmath6 ( @xmath7 ) @xmath8 m .
the diamonds and error bars show the mean and standard deviation for each @xmath9 , respectively , offset horizontally by @xmath26 @xmath8 m for clarity .
panels ( c ) and ( d ) show @xmath22 as a function of @xmath9 , for @xmath6 and @xmath7 @xmath8 m , respectively .
, width=302,height=755 ] percentage of devices showing clean quantization as a function of split - gate length @xmath9 .
panels ( a ) and ( b ) show data for widths @xmath6 and @xmath7 @xmath8 m , respectively .
the triangles ( circles ) indicate devices for which the first and second ( first , second , and third ) plateaus occur within @xmath27 , after correcting for series resistance .
, width=340,height=226 ]
we begin by investigating the effect of split gate length and width on the definition voltage @xmath22 and pinch - off voltage @xmath24 .
figure [ fig1](a ) shows a typical plot of the conductance as a function of the voltage applied to the split gate @xmath1 .
a 1d channel is formed when the 2deg beneath the gates is fully depleted , indicated by a sudden change in the gradient of the conductance trace [ corresponding to voltage @xmath22 and conductance @xmath23 in fig .
[ fig1](a ) ] .
as @xmath1 decreases further the conductance reduces ( showing a series of plateaus ) until the channel is completely pinched off ( marked by @xmath24 ) .
figure [ fig1](b ) shows a scatter plot of @xmath24 against @xmath9 , in which the triangles ( circles ) represent data for @xmath6 ( @xmath7 ) @xmath8 m .
the diamonds / error bars show average values / standard deviation for each @xmath9 .
both sets of data show @xmath24 becoming closer to zero with increasing @xmath9 @xcite .
the trend is more pronounced for @xmath28 @xmath8 m , where @xmath29 is nearly double that of @xmath6 @xmath8 m . these trends arise from simple electrostatics , since for a given @xmath1 , the electric field is stronger in the center of the channel for longer and narrower split gates @xcite .
an additional effect also occurs : a longer wire is more likely to be affected by fluctuations in the background disorder potential due to ionized donors .
this can modify the confining potential and therefore @xmath24 . the larger spread in @xmath24 for @xmath28 @xmath8 m may reflect the increased role of disorder .
however , the spread as a percentage of the mean is similar for both @xmath25 . figures [ fig1](c ) and ( d ) show @xmath22 against @xmath9 for @xmath6 and @xmath7 @xmath8 m devices , respectively . in both cases
the magnitude of @xmath22 reduces as @xmath9 increases , which can be attributed to the higher electric field strength in the center of the channel for longer devices . for each @xmath9 ,
the range and average values of @xmath22 are similar for both widths .
we investigate the role of disorder as a function of device length / width .
the values of conductance plateaus are used to define a yield criterion , since when disorder affects the transmission through a 1d channel the conductance plateaus deviate from expected values . for systematic analysis
the data are first corrected for series resistance ( @xmath30 ) using @xmath31 at @xmath32 v ( i.e. the open - channel resistance ) .
two cases ( a and b ) are considered . case a follows ref .
@xcite and requires the first two conductance plateaus to occur within @xmath33 .
case b extends this to include the third plateau .
figures [ fig2](a ) and [ fig2](b ) show the number of devices which pass the yield criterion as a function of @xmath9 , for @xmath6 and @xmath7 @xmath8 m , respectively .
triangles and circles indicate devices which passed for cases a and b , respectively .
fewer devices show accurate conductance quantization as @xmath25 or @xmath9 increases @xcite , due to the increased likelihood of encountering impurities in the 1d channel and greater variation in the background potential . including the third plateau in the analysis
gives a stricter yield criterion and leads to a lower yield .
the specific form of the relationship between yield and length in fig .
[ fig2 ] is not clear : it is necessary to measure more than 15 devices of each @xmath9 to obtain this information .
this is an interesting avenue for future research since finding the exact nature of the correlation between yield and length may provide information about disorder correlation lengths or dominant disorder effects in the 1d channel .
( a ) grayscale diagram of the transconductance @xmath34 as a function of @xmath1 and source - drain bias @xmath35 , from a typical device with dimensions @xmath36 .
black ( white ) regions correspond to high ( low ) transconductance , i.e. transitions between plateaus ( the plateaus themselves ) .
the 1d subband spacings are estimated by the crossings of peaks in the transconductance @xcite , highlighted by dashed lines .
subband spacing @xmath37 is labeled for illustrative purposes .
( b ) cumulative 1d subband spacings @xmath38 as a function of pinch - off voltage @xmath24 .
the blue , red , and green symbols correspond to @xmath39 , @xmath40 , and @xmath41 , respectively .
unique symbols represent devices of each size , as described in the legend .
the error bounds show the cumulative error in the estimate .
( c)-(e ) scatter plots of @xmath42 against @xmath9 for @xmath43 , @xmath44 and @xmath45 , respectively ( to avoid confusion the error bars on individual data points are not shown ) .
the diamonds show the average for each @xmath9 ( offset horizontally by 0.1 @xmath8 m for clarity ) , and error bounds indicate the average error .
, width=604,height=453 ]
having characterized the length- and width - dependence of two electrical properties of the 1d conductance trace ( @xmath24 and @xmath22 ) , we now consider the length - dependence of the potential profile for a subset of devices .
we first investigate the transverse confining potential by extracting the 1d subband spacing @xcite .
we then study the longitudinal potential by estimating the curvature of the barrier in the direction of electron transport @xcite .
the following measurements and analysis are performed using devices with @xmath6 @xmath8 m , since a higher percentage of split gates with @xmath28 @xmath8 m were affected by disorder .
dc bias spectroscopy is used to measure the 1d subband spacing @xcite for 18 split gates .
figure [ fig3](a ) shows an example grayscale plot of the transconductance @xmath34 as a function of source - drain bias @xmath35 and @xmath1 ( dark / light regions correspond to high / low transconductance ) .
the data are corrected for series resistance following the procedure outlined in the supplementary material of ref .
@xcite . as @xmath46 increases ,
dark regions representing peaks in the transconductance diverge into two ( highlighted by the dashed lines ) , corresponding to the bottom of the 1d subband reaching either the source or drain chemical potential @xcite .
two lines cross when @xmath35 is equal to the energy difference between consecutive 1d subbands , giving the 1d subband spacing @xmath42 ( @xmath47 is the subband index ) .
for example , @xmath37 is marked in fig .
[ fig3](a ) .
figure [ fig3](b ) shows cumulative 1d subband spacings @xmath38 as a function of @xmath24 for 18 devices .
the spacing between the first and second ( @xmath39 ) , first and third ( @xmath40 ) , and first and fourth ( @xmath41 ) subbands are shown ( unique symbols represent devices of different @xmath9 , described in the legend ) . in each case
@xmath38 reduces as @xmath24 becomes closer to zero .
we plot the cumulative data to better accentuate this trend , although the spacing between consecutive subbands [ shown in figs .
[ fig3](c)(e ) ] , all show a downward trend with @xmath24 .
the trend can be attributed to the weakening of the electric field at pinch off for smaller @xmath29 , since this leads to shallower confinement and therefore closer subband spacings .
figures [ fig3](c)-(e ) show individual 1d subband spacings as a function of @xmath9 . panels
( c ) , ( d ) , and ( e ) show @xmath39 , @xmath37 , and @xmath48 , respectively .
the diamonds represent average values for each @xmath9 ( offset horizontally for clarity ) , and error bounds show the average error for each @xmath9 @xcite .
the average subband spacing reduces with @xmath9 in agreement with electrostatic modeling of a saddle point potential @xcite . however , the spread in @xmath42 for individual values of @xmath9 is larger than or similar to the average change in @xmath42 across all values of @xmath9 . since variations in device characteristics likely arise from fluctuations in the background potential , these data highlight the importance of unique electrostatic environment near each split gate .
they suggest that backgound variations are as significant as the lithographic dimensions in governing the potential landscape in the device active area .
( a ) , ( b ) lever arm @xmath49 as a function of lithographic length @xmath9 and 1d subband spacing @xmath39 , respectively .
the diamonds in panel ( a ) represent the mean for each @xmath9 ( offset by 0.1 @xmath8 m for clarity ) , and the error bars indicate the average error .
( c ) experimentally measured conductance @xmath21 as a function of split - gate voltage @xmath1 , for an example device ( solid line ) .
the dashed line shows a fit to the data using a modified saddle - point model .
the transition between plateaus gives an estimate of barrier curvature @xmath50 .
( d ) characteristic length of the potential barrier @xmath51 as a function of gate length @xmath9 .
the diamonds and error bars represent the average value and the average error , respectively ( offset by 0.1 @xmath8 m for clarity ) .
( e ) curvature of the potential barrier @xmath52 as a function of @xmath39 .
, width=321,height=755 ] the degree of coupling between the split gate and the 1d channel is given by lever arm @xmath53 , where @xmath54 ( estimated from dc bias spectroscopy measurements ) .
figure [ fig4](a ) shows @xmath49 ( the lever arm for the first subband ) , as a function of @xmath9 for 18 split gates .
the diamonds indicate the average value , offset horizontally for clarity .
the error bounds indicate the average error @xcite .
the lever arm @xmath49 almost doubles between @xmath55 and @xmath11 @xmath8 m .
this trend is expected since @xmath56 scales with @xmath57 , which , for a given @xmath58 , will be larger in the center of a channel formed by a longer gate . however , as @xmath9 increases further , @xmath49 reduces , deviating from the expected trend .
this also occurs for subbands 2 and 3 .
a length of @xmath11 @xmath8 m is consistent with the correlation length of fluctuations in the background potential . in ref .
@xcite , calculations of the effect of potential fluctuations in the 2deg on split gate devices are performed , for a heterostructure with a similar donor density to the one measured here @xcite .
the fluctuations were assumed to arise from the random positioning of donors in the doped layer .
the correlation length of these fluctuations was found to be comparable with the device active area for split gate lengths of 0.2 and 0.6 @xmath8 m .
our data show @xmath56 depends on split gate length up to @xmath59 @xmath8 m , which may suggest the correlation length of potential fluctuations is slightly greater than this value for the heterostructure used in our experiment .
figure [ fig4](b ) shows @xmath49 as a function of @xmath39 .
data from each length split gate are represented using different symbols , described in the legend .
overall , no trend is apparent .
data from the same length devices appear to be grouped , and a weak positive correlation exists for the @xmath55 @xmath8 m data .
this does not occur for longer devices , consistent with the increased effect of disorder .
we now estimate the curvature of the potential barrier in the transport direction @xmath60 by assuming the confining potential is described by a saddle - point model @xcite .
we achieve this by fitting the measured data with a conductance calculated using the landauer - bttiker formalism @xcite , simulating a system of non - interacting electrons traversing a saddle - point potential with transmission probability @xmath61^{-1}$ ] , where @xmath62 is the energy of the bottom of subband @xmath47 .
the saddle - point approximation can be used for the first few subbands even for devices with a large length - to - width ratio , since @xmath21 is governed by transmission through the narrowest part of the channel .
figure [ fig4](c ) shows measured conductance ( solid line ) and the fit ( dashed line ) for an example device , as a function of @xmath1 .
the fit is achieved as follows : transmission probability @xmath63 is calculated for the first three subbands individually as function of @xmath64 , using an initial input of @xmath65 .
subband - dependent lever arms @xmath66 measured for each device individually are used to convert @xmath64 to a voltage scale .
a minimization routine then optimizes @xmath50 to find the best fit between the calculated and measured traces .
the use of subband - dependent values of @xmath50 reflects how the barrier profile for higher subbands is modified by the increased presence of electrons .
the sum of @xmath63 for @xmath43 , 2 and 3 gives the final trace shown by the dashed line in fig .
[ fig4](c ) .
the transition between @xmath67 and @xmath68 is almost independent of temperature up to at least @xmath69 k @xcite .
therefore , for @xmath43 the fit is performed with @xmath70 , such that the calculated conductance @xmath71 . for higher subbands a temperature dependence
is observed experimentally , therefore for @xmath72 and 3 we calculate @xmath73 at @xmath74 k using @xmath75 where @xmath76 is the fermi - dirac distribution .
we find that using either @xmath70 or @xmath74 for @xmath43 does not affect the trends observed @xcite .
good fits are obtained for 11 of the 18 split gates . for the other seven devices , conductance plateaus
are weakened or suppressed due to strong disorder effects , therefore these data are discarded in the following analysis .
the saddle - point model assumes a parabolic potential barrier , which has a characteristic length @xmath77 , where @xmath78 is the effective mass of the electrons .
barrier length @xmath79 is therefore the distance over which the potential changes by @xmath80 .
figure [ fig4](d ) shows @xmath81 as a function of @xmath9 .
diamonds show the average @xmath82 for each @xmath9 , and error bounds indicate the average error , offset horizontally for clarity
. we would expect the barrier length to scale with the gate length .
however , our data show the opposite trend with @xmath82 initially reducing for @xmath9 from @xmath83 to @xmath11 @xmath8 m .
this change is very small compared to the change of @xmath9 : @xmath84 nm ( a decrease of @xmath85 ) , as @xmath9 increases by @xmath86 .
as @xmath9 increases further , @xmath87 then increases slightly . for our devices therefore , the split gate length is not a good indicator of the length of the potential barrier . for @xmath55 , 0.7 and 1 @xmath8 m , the spread of @xmath81 at fixed @xmath9 is larger than or similar to the overall change in average @xmath81 as @xmath9 varies .
this highlights the crucial importance of the background potential in determining the electrostatic landscape in the active device area , even between devices with nominally identical length .
this is further supported by fig .
[ fig4](e ) , which shows @xmath52 as a function of @xmath39 . for a given @xmath9
, there is no correlation between these parameters .
the trends for subbands 2 and 3 are similar .
( a ) fitted conductance @xmath88 for the first 1d subband from 11 devices .
data are collapsed onto a universal curve by aligning @xmath89 with @xmath32 , then scaling @xmath1 by @xmath90 .
( b ) corresponding experimentally - measured conductance @xmath91 , where the data are offset and scaled using the same parameters as ( a ) . for @xmath92
the traces collapse onto a similar curve . above @xmath68 , variations arise due to the differences in the 0.7 anomaly .
( c ) @xmath91 as a function of barrier curvature @xmath52 , at fixed values of the scaled voltage axis @xmath93 . from bottom - to - top
, @xmath93 increases from @xmath94 to @xmath95 in steps of @xmath96 [ corresponding to vertical dashed lines , left - to - right , in panel ( b ) ] .
the dashed lines show a linear least squares fit to @xmath91 for each value of @xmath93 , as a guide to eye .
( d ) conductance @xmath91 as a function of spacing between the first and second 1d subbands @xmath39 . in panels ( c)(d ) , data points for each length split gate are indicated by the symbols defined in the legend .
, width=604,height=434 ]
in this section 0.7 anomalies from devices of different gate length are compared . following bauer _
et al . _
@xcite , we refer to the conductance between @xmath68 and @xmath97 as the ` sub - open ' regime .
additionally , we refer to experimentally measured ( fitted ) conductance data as @xmath98 ( @xmath99 )
. direct comparison of the 0.7 anomaly between devices is possible by removing the trivial that which can be accounted for in a non - interacting scenario @xcite dependence of the conductance transition between @xmath67 and @xmath97 on barrier curvature .
this is achieved by offseting the conductance traces horizontally to align @xmath89 to @xmath32 and scaling each @xmath1 axis by @xmath90 @xcite .
differences in conductance that remain are only due to electron interactions .
the strength of these interactions still depends on barrier shape @xcite .
figure [ fig5](a ) shows the fitted conductance @xmath99 for the first subband as a function of the scaled voltage axis @xmath100 .
data from all 11 split gates are plotted .
the traces collapse onto a universal curve since these data are obtained using a a non - interacting model .
figure [ fig5](b ) shows the corresponding @xmath91 data after applying the same scaling procedure . for @xmath101 , the traces collapse onto a very similar curve . for @xmath102
, differences occur due to the 0.7 anomaly ( a spread in @xmath103 arises due to the different @xmath39 ) .
the variation in the 0.7 anomaly between traces is related to the varying barrier curvature from device to device . in order to compare the 0.7 anomaly between devices ,
@xmath91 is plotted as a function of @xmath52 in fig .
[ fig5](c ) , for 6 fixed values of @xmath93 . from bottom - to - top
, @xmath93 increases from -0.5 to 0.75 in steps of 0.25 [ corresponding to the vertical lines in fig .
[ fig5](b ) ] . for @xmath104 , @xmath105 and @xmath106
, @xmath91 is independent of @xmath52 .
this is expected since @xmath91 is below or close to @xmath107 .
unique symbols are used to represent data from split gates of each length .
the dashed lines are linear fits as a guide to the eye @xcite
. in the sub - open regime @xmath91 reduces with decreasing @xmath3 .
this is consistent with our previous work @xcite .
however , in ref .
@xcite an array of identical length split gates was measured in which differences in the electrostatic profile between devices arose only from variations in the background potential . in fig .
[ fig5](c ) data from split gates of different length all follow the same trend line for a given @xmath93 .
this implies that the profile of the longitudinal barrier rather than the gate length is the most significant factor governing the conductance value of the 0.7 anomaly .
the data in fig .
[ fig5 ] are presented in a similar way to fig . 2 of ref .
@xcite for ease of comparison .
we emphasize that these are entirely different datasets , from separate devices .
the similarity arises because the same technique developed in ref .
@xcite to analyze the 0.7 anomaly is applied here .
the importance of device - specific confining potential has been highlighted in ref .
@xcite , which summarizes numerous studies of the 0.7 anomaly as a function of carrier density .
the conductance of the 0.7 anomaly has been seen to both increase and decrease as a function of carrier density .
these conflicting trends likely arise due to the extreme sensitivity of the 0.7 anomaly to differences in the electrostatic potential between devices @xcite .
the trend in fig .
[ fig5](c ) is also consistent with the split - gate length dependence reported in ref .
the 0.7 anomaly occurred at lower values as gate length increased ( three devices @xmath108 0 , 0.5 and 2 @xmath8 m long were measured ) .
a stronger link between the longitudinal profile and gate length may exist for devices measured in ref .
@xcite because they are fabricated on an undoped heterostructure @xcite , where the absence of dopants may lead to smaller variations in the background potential .
calculations of the conductance transition between zero and @xmath97 using the inelastic scattering model plus the local density of states enhancement @xcite predict a lowering of the conductance in the sub open regime as @xmath109 decreases .
these calculations are shown in fig .
s14(b ) in supplementary material of ref .
@xcite , and are in agreement with our data .
unfortunately , our data do not allow us to distinguish between theories for the occurrence of the 0.7 anomaly since the same trend is predicted in both the spontaneous spin polarization @xcite and kondo scenarios @xcite ( this is discussed in more detail in ref .
@xcite ) . however , the inelastic scattering scenario is the only theory for which detailed calculations have been performed as a function of @xmath109 .
figure [ fig5](d ) shows @xmath91 as a function of 1d subband spacing @xmath39 .
there is a slight reduction of @xmath98 with increasing @xmath39 in the sub - open regime , although the trend is weak . as seen in fig .
[ fig2](b ) , a larger @xmath39 occurs for devices with a more negative pinch off voltage .
this suggests a possible explanation for the trend in fig .
[ fig5](d ) : the strength of transverse confinement is stronger for the devices with larger @xmath39 , leading to an increase of strength of electron interactions , thus affecting the conductance value of the 0.7 anomaly .
this may be understood within the framework of the inelastic scattering model , which makes predictions of the effect of electron interaction strength on the conductance of the 0.7 anomaly .
for example , fig .
s14(c ) of supplementary material for ref .
@xcite shows a reduction of the conductance in the sub - open regime with increasing interaction strength , for a constant barrier curvature . as a final point of interest ,
the trends in figs .
[ fig5](c ) and [ fig5](d ) do not depend on the split gate length , compatible with the inelastic scattering model which is based purely on the shape of the potential barrier .
we have systematically studied the effect of changing split gate size on device behavior .
this is achieved using a multiplexing technique which allows many nanostructure devices to be compared on a single cooldown .
multiple devices of each size are measured , providing statistical information on the variance of conductance properties between individual designs . in total
, we measured 95 split gates with 7 different length / width combinations .
increasing the gate length and/or reducing the width moves the average voltage required to define a 1d channel and the pinch off voltage closer to zero .
the 1d subband spacing also reduces for devices with longer gates .
further , the percentage of devices displaying accurate quantization of conductance reduces dramatically as the area of the channel increases .
the electrostatic environment in the 2deg is very influential on the 1d potential profile .
the significance of the background potential is highlighted by three key results reported here .
firstly , the spread in values of the 1d subband spacing and the 1d barrier curvature for a given gate length are as large as the overall variation in the average values of these properties over the range of gate lengths measured .
secondly , the lever arm @xmath56 which depends on the coupling between the gate and the 1d channel does not continue to increase as a function of gate length beyond @xmath110 @xmath8 m , a deviation from the expected result .
thirdly , the curvature of the longitudinal potential barrier estimated using a saddle - point model is not strongly related to gate length .
these results imply that _ i _ ) gate size can not be relied upon as a good indicator of the length of the 1d channel , and _ ii _ ) the background disorder potential is at least as significant as gate size in determining the potential landscape in the 1d channel . the 0.7 anomaly is compared between split gates of different length . the conductance value of the 0.7 anomaly reduces as the barrier curvature becomes shallower , rather than depending specifically on the split gate length .
the particular confining potential in each device , and principally the barrier curvature , may be the primary factor governing the conductance of the 0.7 anomaly at a given temperature and magnetic field .
this study has characterized split gates on a modulation doped gaas / algaas heterostructure . a potential avenue for future work is to perform measurements on undoped heterostructures where the absence of si donors results in a reduction of background disorder @xcite , leading to more reproducible behavior both between split gate devices @xcite and after thermal cycling @xcite .
measuring an array of devices on an undoped structure may also give insight on the degree to which disorder affects nanostructures fabricated on undoped heterostructures @xcite .
this work was supported by the engineering and physical sciences research council grant no .
ep / i014268/1 .
the dataset for this article is available at www.repository.cam.ac.uk/handle/1810/249174 .
the authors thank a. r. hamilton , j. von delft , s. ludwig , and e. t. owen for helpful discussions , and r. d. hall for e - beam exposure .
mm l. w. smith , h. al - taie , f. sfigakis , p. see , a. a. j. lesage , b. xu , j. p. griffiths , h. e. beere , g. a. c. jones , d. a. ritchie , m. j. kelly , and c. g. smith , phys .
b * 90 * , 045426 ( 2014 ) .
d. a. wharam , t. j. thornton , r. newbury , m. pepper , j. ahmed , j. e. f. frost , d. g. hasko , d. a. ritchie , and g. a. c. jones , j. phys .
c * 21 * , l209 ( 1988 ) .
b. j. van wees , h. van houten , c. w. j. beenakker , j. g. williamson , l. p. kouwenhoven , d. van der marel , and c. t. foxon , phys .
* 60 * , 848 ( 1988 ) .
k. j. thomas , j. t. nicholls , m. y. simmons , m. pepper , d. r. mace , and d. a. ritchie , phys .
* 77 * , 135 ( 1996 ) . t. j. thornton , m. pepper , h. ahmed , d. andrews , and g. j. davies , phys .
. lett . * 56 * , 1198 ( 1986 ) .
lee , k. murakia , e. y. chang , and y. hirayama , j. appl .
phys . * 100 * , 043701 ( 2006 ) .
e. j. koop , a. i lerescu , j. liu , b. j. van wees , d. reuter , a. d. wieck , and c. h. van der wal , j. supercond . nov .
magn * 20 * , 433 ( 2007 ) .
d. j. reilly , g. r. facer , a. s. dzurak , b. e. kane , r. g. clark , p. j. stiles , a. r. hamilton , j. l. obrien , n. e. lumpkin , l. n. pfeiffer , and k. w. west , phys . rev .
b * 63 * , 121311 ( 2001 ) . s. j. koester , b. brar , c. r. bolognesi , e. j. caine , a. patlach , e. l. hu , h. kroemer , and m. j. rooks , phys . rev .
b * 53 * 13063 ( 1996 ) .
k. j. thomas , d. l. sawkey , m. pepper , w. r. tribe , i. farrer , m. y. simmons and d. a. ritchie , j. phys . : condens .
matter * 16 * , l279-l286 ( 2004 ) .
m. j. iqbal , r. levy , e. j. koop , j. b. dekker , j. p. de jong , j. h. m. van der velde , d. reuter , a. d. wieck , r. aguado , y. meir and c. h. van der wal , nature ( london ) * 501 * , 79 ( 2013 ) .
m. j. iqbal , j. p. de .
jong , d. reuter , a. d. wieck , and c. h. van der wal , j. appl
. phys . * 113 * , 024507 ( 2013 ) j. heyder , f. bauer , e. schubert , d. borowsky , d. schuh , w. wegscheider , j. von delft , and s. ludwig , arxiv:1409.3415 g. timp , r. behringer , s. sampere , j. e. cunningham , and r. e. howard , in _ proceedings of the international symposium on nanostructure physics and fabrication _ , edited by m. a. reed and w. p. kirk ( academic , new york , 1989 ) , p. 331 .
k. j. thomas , j. t. nicholls , n. j. appleyard , m. y. simmons , m. pepper , d. r. mace , w. r. tribe , and d. a. ritchie , phys .
b * 58 * , 4846 ( 1998 ) .
a. p. micolich , j. phys .
matter * 23 * , 443201 ( 2011 ) .
wang and k .- f .
berggren , phys .
b * 54 * , r14257 ( 1996 )
. s. m. cronenwett , h. j. lynch , d. goldhaber - gordon , l. p. kouwenhoven , c. m. marcus , k. hirose , n. s. wingreen , and v. umansky , phys .
* 88 * , 226805 ( 2002 ) .
y. meir , k. hirose , and n. s. wingreen , phys .
lett . * 89 * , 196802 ( 2002 ) .
t. rejec and y. meir , nature ( london ) * 442 * , 900 ( 2006 ) .
k. a. matveev , phys .
b * 70 * , 245319 ( 2004 ) .
b. brun , f. martins , s. faniel , b. hackens , g. bachelier , a. cavanna , c. ulysse , a. ouerghi , u. gennser , d. mailly , s. huant , v. bayot , m. sanquer , and h. sellier , nat commun . * 5 * , 4290 ( 2014 ) . c. sloggett , a. i. milstein , and o. p. sushkov , eur .
j. b * 61 * , 427432 ( 2008 ) .
f. bauer , j. heyder , e. schubert , d. borowsky , d. taubert , b. bruognolo , d. schuh , w. wegscheider , j. von delft , and s. ludwig , nature ( london ) * 501 * , 73 ( 2013 ) .
m. bttiker , phys .
b * 41 * , 7906 ( 1990 ) .
h. al - taie , l. w. smith , b. xu , p. see , j. p. griffiths , h. e. beere , g. a. c. jones , d. a. ritchie , m. j. kelly , and c. g. smith , appl .
. lett . * 102 * , 243102 ( 2013 ) . h. al - taie , l. w. smith , a. a. j. lesage , p. see , j. p. griffiths , h. e. beere , g. a. c. jones , d. a. ritchie , m. j. kelly , and c. g. smith , j. appl .
* 118 * , 075703 ( 2015 ) .
a. a. j. lesage , l. w. smith , h. al - taie , p. see , j. p. griffiths , i. farrer , g. a. c. jones , d. a. ritchie , m. j. kelly , and c. g. smith , j. appl . phys .
015704 ( 2015 ) . l. w. smith , h. al - taie , a. a. j. lesage , f. sfigakis , p. see , j. p. griffiths , h. e. beere , g. a. c. jones , d. a. ritchie , a. r. hamilton , m. j. kelly , and c. g. smith , phys .
b * 91 * , 235402 ( 2015 ) .
r. k. puddy , l. w smith , h. al - taie , c. h. chong , i. farrer , j. p. griffiths , d. a. ritchie , m. j. kelly , m. pepper , and c. g. smith , appl .
. lett . * 107 * , 143501 ( 2015 ) .
j. h. davies , i. a. larkin and e. v. sukhorukov , j. appl .
phys . * 77 * , 4504 ( 1995 ) . c. w. j. beenakker and h. van houten , solid state phys . ,
* 44 * , 1 ( 1991 ) .
n. k. patel , j. t. nicholls , l. martn - moreno , m. pepper , j. e. f frost , d. a. ritchie , and g. a. c. jones , phys .
b * 44 * , 13549 ( 1991 ) . a. srinivasan , l. a. yeoh , o. klochan , t. p. martin , j. c. h. chen , a. p. micolich , a. r. hamilton , d. reuter , and a. d. wieck , nano lett .
* 13 * , 148 ( 2013 ) .
k. j. thomas , m. y. simmons , j. t. nicholls , d. r. mace , m. pepper , and d. a. ritchie , appl .
. lett . * 67 * , 109 ( 1995 ) . the error associated with the mean value
is given by the square root of the averaged square of individual errors divided by the number of data points .
j. a. nixon , j. h. davies , and h. u. baranger , phys . rev .
b * 43 * , 12638 ( 1991 ) .
the heterostructure used in our experiment contains a 40 nm si doped algaas layer with a doping density @xmath115 @xmath116 , separated from the 2deg by 40 nm of undoped algaas . in ref
@xcite calculations are performed for a @xmath117-doped heterostructure with @xmath118 @xmath16 , separated from the 2deg by 42 nm of undoped algaas . for comparison
, we can estimate a 2d density for our heterostructure by assuming the doping layer is infinitesimally thick , which gives @xmath119 @xmath16 . in ref .
@xcite the 2d electron density is calculated to be @xmath120 @xmath16 , very similar to ours .
calculations were performed for split gates of length 0.2 and 0.6 @xmath8 m , both 0.3 @xmath8 m wide .
data which are included are those which can be fitted with a parabolic potential .
this is a self - selecting process , and may explain why the range of @xmath52 is within that which was found for the lithographically identical devices .
p. jaksch , i. yakimenko , and k .- f .
berggren , phys .
b * 74 * , 235320 ( 2006 ) .
k. hirose , y. meir , and n. s. wingreen , phys .
* 90 * , 026804 ( 2003 ) .
a. m. burke , o. klochan , i. farrer , d. a. ritchie , a. r. hamilton , and a. p. micolich , nano lett .
* 12 * , 4495 ( 2012 ) .
a linear fit is plotted to highlight the upward trend .
this is simply a guide to the eye .
the range of @xmath3 is too small to determine the exact nature of the relationship between @xmath98 and @xmath3 .
b. e. kane , g. r. facer , a. s. dzurak , n. e. lumpkin , r. g. clark , l. n. pfeiffer and k. w. west , appl .
72 * , 3506 ( 1998 ) .
b. e. kane , l. n. pfeiffer , k. w. west , and c. k. harnett , appl .
. lett . * 63 * , 2132 ( 1993 ) .
f. sfigakis , k. das gupta , s. sarkozy , i. farrer , d. a. ritchie , m. pepper , and g. a. c. jones , physica e * 42 * , 1200 ( 2010 ) .
s. sarkozy , f. sfigakis , k. das gupta , i. farrer , d. a. ritchie , g. a. c. jones , and m. pepper , phys .
b * 79 * , 161307(r ) ( 2009 ) . a. m. see , i. pilgrim , b. c. scannell , r. d. montgomery , o. klochan , a. m. burke , m. aagesen , p. e. lindelof , i. farrer , d. a. ritchie , r. p. taylor , a. r. hamilton , and a. p. micolich , phys .
lett . * 108 * , 196807 ( 2012 ) .
a. m. see , a. r. hamilton , a. p. micolich , m. aagesen , and p. e. lindelof , phys .
b * 91 * , 085417 ( 2015 ) . | we study 95 split gates of different size on a single chip using a multiplexing technique .
each split gate defines a one - dimensional channel on a modulation - doped gaas / algaas heterostructure , through which the conductance is quantized .
the yield of devices showing good quantization decreases rapidly as the length of the split gates increases .
however , for the subset of devices showing good quantization , there is no correlation between the electrostatic length of the one dimensional channel ( estimated using a saddle point model ) , and the gate length .
the variation in electrostatic length and the one - dimensional subband spacing for devices of the same gate length exceeds the variation in the average values between devices of different length .
there is a clear correlation between the curvature of the potential barrier in the transport direction and the strength of the `` 0.7 anomaly '' : the conductance value of the 0.7 anomaly reduces as the barrier curvature becomes shallower .
these results highlight the key role of the electrostatic environment in one - dimensional systems . even in devices with clean conductance plateaus , random fluctuations in the background potential are crucial in determining the potential landscape in the active device area such that nominally identical gate structures have different characteristics . |
in this section , we review basic facts about systems of conservation laws and standardize our notation .
a system of conservation laws ( or simply `` conservation law '' ) is a partial differential equation ( pde ) of the form @xmath0 where @xmath1 ( for @xmath2 ) , @xmath3 , is the unknown , and @xmath4 is a given _
flux function_. the conserved quantity @xmath5 we refer to as `` mass '' in the following discussion . in the case where @xmath5 is scalar valued ( @xmath6 ) , we refer to eq .
( [ scl ] ) as a scalar conservation law .
our working example of a conservation law will be the hopf equation ( a.k.a . the `` inviscid burgers equation '' ) , @xmath7 which is a scalar conservation law in one spatial dimension ( @xmath8 ) .
the hopf equation is one of the simplest examples of a pde which develops discontinuities in finite time@xcite .
such discontinuities and other non - smooth features appearing in solutions to pde we refer to collectively as
_ defects_. in the presence of discontinuities in @xmath5 , the derivatives appearing in eqs .
( [ scl]-[hopf ] ) are no longer well defined , and so we must generalize our notion of `` solution '' to a pde in order to make sense of the conservation law at later times .
such a generalization is referred to as a _ weak solution_. several distinct notions of weak solution exist , such as integral solutions , viscosity solutions , variational solutions , and various types of regularized solutions , to name a few@xcite .
not all forms of weak solution are applicable to all pde s .
the types of weak solution which are most relevant for systems of conservation laws are _ regularized solutions _ and _ integral solutions_. in the former case , the pde is modified by the addition of a small term that serves to smooth out the defects ; for the hopf equation , such a modification looks like @xmath9 where @xmath10 controls the strength of the regularization . to recover a solution to the original system of conservation laws , we take a limit as @xmath11 .
note that , in principle , the form of the regularizing term should correspond to the microphysics of a given system
so for example , systems with viscosity have a second - order regularization , @xmath12 , whereas inviscid systems ( like plasmas or superfluids ) typically have dispersive ( third - order ) regularization , @xmath13 .
below , in eqs .
( [ modhopf1]-[modhopf3 ] ) , we will consider a different form of regularization altogether , which nonetheless bears a familial resemblance to eq .
( [ reghopf ] ) .
the main alternative kind of weak solution , the integral solution , proceeds instead by integrating the original pde to remove the problematic derivatives .
integral solutions are not of direct interest to us here , but we point out one tangential relationship to our current discussion : integral solutions are generally not _ unique _ , and to select a proper weak solution requires additional microphysical information . this information is usually specified as a relationship between the values of the solution @xmath5 on either side of a defect ; when the relationship takes the form of an inequality it is called an _ _ entropy condition__@xcite , and when it is an equality it is called a _ _ kinetic relation__@xcite .
the stationarity condition we propose below is similar to a kinetic relation , but with the generalization that the _ speed _ of the defect can also figure into the equality .
for our purposes , a delta shock ( also written `` @xmath14-shock '' ) may be defined informally as any defect which has a finite amount of mass concentrated at a point .
we note in passing that there are also meaningful notions of @xmath15-shocks ( and @xmath16-shocks , more generally ) corresponding to defects containing _
derivatives _ of delta functions@xcite . our numerical solutions ( section [ subsec : regularizations ]
, figs [ hopf ] and [ coshfig ] ) appear to contain an upward and downward - pointing singularity the sum being the strength of the delta shock and the difference a @xmath17 component . in this section
we shall derive conditions for the evolution of the strength of the delta shock component .
delta shocks are known to arise naturally in a number of systems of conservation laws for example , the system @xmath18 considered by tan , zheng , & zhang@xcite , and the system @xmath19 studied by keyfitz & kranzer@xcite . in the former case , the delta shocks have been shown to be vanishing viscosity limits of the regularized system @xmath20 the mechanisms which cause delta shocks to form in systems of conservation laws are not fully understood@xcite , though in some cases the necessity of delta shocks is obvious .
we highlight two such cases here : firstly , if a conserved quantity @xmath21 is a derivative of another variable @xmath5 , and the latter exhibits regular ( viscous ) shocks , then clearly @xmath21 must contain a delta shock . for example , if we take a derivative of the hopf equation ( [ hopf ] ) and set @xmath22 , we find that @xmath23 which is precisely the latter half of system ( [ tanseq ] ) .
thus ( provided the initial data for system ( [ tanseq ] ) satisfies @xmath24 ) we see that this system can be interpreted as the evolution of the conserved quantity of the hopf equation , along with that of its derivative .
this interpretation is not the only available , nor is it necessarily the best .
indeed , the closely related system@xcite @xmath25 where @xmath26 is an arbitrary smooth & convex function and g is an arbitrary smooth & increasing function , also displays delta shocks , though they can not in this case be interpreted as derivatives of regular shocks .
a more general interpretation for the appearance of delta shocks in such conservation laws is as a consequence of _ incompatibility of defect motion and the flow of the conserved quantity _ , which we explore in the next two subsections .
the rankine - hugoniot relation is a statement of conservation of mass across a shock . for the case of a regular shock in one spatial dimension which moves with speed @xmath27
, mass conservation requires that the net flux into the shock balance the mass of the `` hole '' left as the shock moves away . in symbols , this reads @xmath28 where @xmath29 and @xmath30 are the values of @xmath5 to the immediate left and right ( resp . ) of the shock ( see evans@xcite for a more formal derivation of this relation ) .
if we allow the possibility of delta shocks storing mass on the defect , it is now the combination of the hole behind the shock _ and _ the mass @xmath31 of the delta which must balance the incoming flux . the modified rankine - hugoniot relation is thus @xmath32 note that both eqs .
( [ rh ] ) and ( [ mrh ] ) are vector identities in general , and may be interpreted as specifying the shock speed @xmath27 and delta mass @xmath31 ( or the derivative thereof ) in terms of the local environment , given by @xmath29 and @xmath30 .
in particular , this implies that eq .
( [ rh ] ) is only satisfiable in general when @xmath5 is scalar , since vector @xmath5 leads to an overspecification in ( [ rh ] ) of the single unknown @xmath27 .
( [ mrh ] ) has no such problem , since @xmath31 has the same number of components as @xmath5 , so that there is always one more unknown than there are equations in ( [ mrh ] ) .
this shows that , except in the presence of high degeneracy so that eq .
( [ rh ] ) is soluble for @xmath27 , we expect multicomponent conservation laws to require delta shocks in order to conserve mass in the vicinity of a defect .
this provides our preferred explanation of the delta shocks in system ( [ tanseq ] ) : the classical rankine - hugoniot relation ( [ rh ] ) for this system reads : @xmath33 which does not hold unless @xmath34 ( i.e. the necessary degeneracy for ( [ rh ] ) to be satisfiable does not generally exist ) .
a similar statement shows why the more general system ( [ modtanseq ] ) also displays delta shocks .
the rankine - hugoniot relations illustrate how multicomponent systems can develop delta shocks when a defect s motion can not be simultaneously compatible with the flow of all components of the conserved quantity .
we can also produce a similar effect even in scalar conservation laws by insisting that a defect move according to specified kinematics which are not compatible with the flow of the scalar conserved quantity .
the simplest such example is to demand that a defect be _ stationary_. this might be an appropriate kinematic constraint in , for example , an electrical network containing fuses , where a blown fuse is modeled as a stationary defect
. a stationarity condition could also be physically appropriate for the continuum dislocation dynamics example we will describe below .
this stationarity condition is analogous to a kinetic relation@xcite , with the generalization that the speed of the defect can now figure in as well ( rather than just the value of the local conserved quantity ) . as an illustration
, we consider the riemann problem @xmath35 for the hopf equation ( [ hopf ] ) .
mass flows into the defect at @xmath36 with rate @xmath37 , and thus by the modified rankine - hugoniot relation ( [ mrh ] ) with @xmath38 we find that the mass of the delta at @xmath36 grows at a rate @xmath39 .
everywhere else the value of @xmath5 is uniquely specified by projection of characteristics .
hence a reasonable candidate solution for all times @xmath40 is @xmath41 we note that this candidate solution is not an integral solution in the usual sense , as the square of a delta function ( needed to make sense of an integral solution to eq .
( [ hopf ] ) ) is undefined by classical distribution theory .
( however , it is possible to overcome this technicality using generalized distributions@xcite . )
our _ ad hoc _ construction of a solution ( [ hopfsol ] ) to the hopf equation which satisfies the condition of stationarity does not lend itself well to numerical simulation .
we thus find it advantageous to try to construct a `` irregularization '' of the hopf equation which yields such a solution .
our attempted irregularization looks like : @xmath42 where the smooth & bounded function @xmath43 satisfies : @xmath44 + @xmath45 + @xmath46 .
for example , @xmath47 is such an irregularization .
the intuition behind this is that when a shock develops in the hopf equation , @xmath48 diverges to @xmath49 , sending the modified flux @xmath50 to zero .
the quenching of the flux will hopefully cause a pileup of mass behind the defect . on the other hand , for small values of @xmath51 and away from defects ,
condition 3 above implies that @xmath52 , and the original hopf equation is recovered ( approximately ) .
( this prescription for the irregularization is not as general as possible for example , one also could add to the denominator of eq .
( [ modhopf2 ] ) a non - vanishing polynomial in @xmath5 , which would reflect some sort of self - interaction effect . however , the `` minimal '' irregularization of eq .
( [ modhopf2 ] ) is sufficient to produce the defects we are interested in .
moreover , the effect is not peculiar to this particular form : we have also experimented with the irregularization @xmath53 and seen the same qualitative behavior . ) for numerical work , it is sometimes helpful to also add a small regularizing term to ensure solutions stay relatively smooth e.g .
@xmath54 where k is typically 2 or 4 . ) starting with sinusoidal initial conditions , for two values of @xmath51 at time @xmath55 ( simulated with a conservative upwind differencing ) . note that the singularity has both a positive and negative peak .
only the sum contributes to the delta shock ( as a weak limit ) ; the difference can be described as a derivative of a @xmath14-function . ]
does this succeed in making the defect stationary ?
we begin to answer this by examining numerical results .
( [ fig - hopf ] ) shows simulations of eq .
( [ modhopf3 ] ) for @xmath56 & @xmath57 on the unit interval with periodic boundary conditions , for parameters @xmath55 , @xmath58 , @xmath59 , and a grid spacing @xmath60 of .002 .
clearly the irregularization has substantially slowed the movement of the shock and thereby induced a delta - like spike to conserve mass . the extent to which this is a bona fide delta shock is a delicate question ( we will thus cautiously refer to this feature as a _ delta spike _ ) .
we mention three relevant considerations : 1 . to begin with , the effect is inherently non - smooth , as for smooth solutions of eq .
( [ modhopf2 ] ) , we can prove the following maximum principle : + a solution @xmath5 to eq .
( [ modhopf2 ] ) on a compact domain satisfies @xmath61 for all times @xmath62 while the solution remains smooth .
+ we motivate this as follows : if we look at a curve @xmath63 such that @xmath64 attains a maximum at @xmath63 , then we find @xmath65 where the last expression vanishes because @xmath66 at the maximum @xmath63 .
this shows that the maximum value of @xmath5 is unchanging in time .
( this argument can be made more rigorous , avoiding the assumption that the maximum can be parameterized by a differentiable curve @xmath63 , but we will not do so here . ) note that a similar result holds also in the case of eq .
( [ modhopf3 ] ) with a 2nd order ( viscous ) regularization , with the modification that the maximum of @xmath5 is then decreasing in time .
2 . examining
the delta spike closely reveals an internal structure . to provide some perspective , we note first that eq .
( [ modhopf2 ] ) admits two relevant exact solutions , one of the form @xmath67 ( i.e. a straight line , moving in time ) , and the other a stationary solution of the form @xmath68 + experimentally , we find that the shape of the spike is well approximated by a profile of the form ( [ coshprofile ] ) , as shown in fig .
( [ coshfig ] ) .
more precisely , the form of the delta spike consists of two partial coshes , one positive and one negative , separated by an abrupt discontinuity .
away from the defect , the solution to eq .
( [ modhopf2 ] ) approaches the aforementioned linear solution asymptotically in time .
this eventual shape is analogous to the asymptotic `` n - wave '' profile of solutions to the regular hopf equation . as discussed in the appendix , it is possible to derive analytic expressions for the parameters @xmath69 , @xmath70 , @xmath71 , @xmath72 describing this asymptotic form , and we find good agreement between simulation and these analytic values .
note that the cosh profile shown here is not a best fit , but rather an analytically derived approximation , as discussed in the appendix .
the delta spike comes from a simulation with parameters @xmath73 , and initial data @xmath74 .
again , the positive and negative spikes must be added together to give the net weight of the delta shock . ] + an important feature of the cosh profile ( [ coshprofile ] ) is that its width scales like @xmath75 . to conserve mass ,
the height of the shock must therefore scale like @xmath76 .
this suggests that , although the delta spike is not truly a delta _ shock _ for non - zero values of @xmath51 ( indeed , the cosh profile ( [ coshprofile ] ) certainly does not qualify as a delta shock ) , as @xmath77 the spike does indeed converge to a delta function .
experimentally , this is what we seem to find ( conditional upon some finessing of the numerical methods ) , as discussed in the next point .
3 . establishing numerical convergence of our simulations is tricky for two reasons : firstly , since we are looking for convergence to a delta shock , we can not hope to have convergence in e.g. the @xmath78 sense .
secondly , the equation ( [ modhopf2 ] ) is highly unstable ( not surprisingly , since the irregularization is designed specifically to produce a delta shock , which may be viewed as a sort of instability ) . after trying various finite - differencing schemes
, we found that the best results came from a type of conservative upwind differencing , wherein @xmath5 is differenced upwind but @xmath48 is not .
explicitly , the semi - discrete formulation of eq .
( [ modhopf1 ] ) is @xmath79 where @xmath80 is the discrete forward difference operator ( @xmath81 ) .
all figures presented herein were produced using this discretization .
( remark : for the regularized eq .
( [ modhopf3 ] ) we performed an operator splitting , evaluating the regularization term in fourier space for stability . ) despite being the most successful numerical method we found , the stability properties of this discretization are not entirely satisfactory .
in particular , we found that when the grid size became substantially smaller than the width of the delta spike ( which is on the order of @xmath75 ) , the delta spike would become unstable and break apart into multiple , smaller delta spikes . to overcome this difficulty
, we let the mesh size @xmath60 and the width of the delta @xmath75 go to zero together , keeping @xmath82 fixed ( recall that it was our intention to let @xmath83 anyhow , so as to recover a solution to the original hopf equation ) .
this technique appears to work , as seen in fig .
( [ deltacvrg ] ) : the delta spike gets narrower and higher , while remaining in essentially the same place , as @xmath84 in this manner . + , @xmath85 . ] although our simulations seem to establish convergence to a delta shock , it is a very peculiar convergence wherein the strength @xmath51 of the irregularization must vanish together with the discretization length @xmath60 .
the fact that the width of the delta spike must be kept close to @xmath60 suggests that there may be some sort of resonance between the irregularization of eq .
( [ modhopf2 ] ) and the grid which drives the formation of the spike ( especially in light of the maximum principle mentioned previously ) . nonetheless
, we can still hope to produce a meaningful weak solution for the hopf equation in this way .
we will not establish anything more rigorous at this time , though .
our interest in delta shocks is motivated by a tangible physical question .
why do dislocations in crystals form walls @xcite ? briefly , a crystal is a regular array of atoms .
a dislocation line is a flaw in that array , such as the edge dislocation formed by the boundary of an extra plane of atoms , or the screw dislocation forming the central line where planes form a ` spiral staircase ' @xcite .
these dislocations move to mediate plastic deformation when the crystal is bent , and form tangles that organize into wall - like structures called _ cell walls _ @xcite . in a continuum theory of dislocation dynamics ,
such walls must be described as delta shocks : their density scales as the inverse square of the lattice constant , which vanishes in the continuum limit .
one class of continuum dislocation dynamics theories @xcite do form such delta shocks . in these theories ,
the dislocations move with a common velocity @xmath86 .
if one describes an incipient wall in the @xmath87 plane with a dislocation density that depends only on @xmath88 , the dependence of this velocity on the dislocation density simplifies @xcite to the hopf equation : @xmath89 when the hopf equation forms a step - like shock , the entrained dislocations pile up into a delta shock , forming a wall .
the central physical question is how that wall should evolve after it forms .
if we regularize the hopf equation ( [ pkf ] ) into burger s equation ( i.e. viscous regularization ) , the dislocation wall moves along with the rankine - hugoniot velocity .
this produces a self - consistent , sensible model @xcite , but one that is unsatisfying in two regards .
first , the velocity of dislocation walls in crystals is determined by the microstructure of the walls and not by continuum properties .
secondly , generalizations of this continuum theory have differing velocities for the different components of the dislocation density smearing the resulting walls . on the other hand , _ physically _ at some junctions where two dislocations intersect , the point of intersection can be pinned in place ( so - called _ sessile _ dislocation junctions @xcite ) .
dislocation walls can also be pinned by impurities that segregate to the boundary .
this provides the motivation for our study in section ( [ sec : stationarity ] ) of the hopf equation ( [ pkf ] ) with stationary shocks .
we should note that the behavior of both our model in two and three dimensions and experimental dislocation systems under stress is more complex than simple wall singularities .
there they produce complex cellular structures , which in our models and some experiments form self - similar , fractal morphologies .
our theories in higher dimensions show clear analogies @xcite to the behavior of the euler equation ( the inviscid limit of navier - stokes ) .
nonetheless , these structures are complex , ramified wall - like entities , whose dynamics should be controlled by physics on the microscopic atomic scale , not by the continuum laws .
we have seen that delta shocks arise naturally in a variety of systems of conservation laws , and their presence can often be traced to an incompatibility of the flow of the conserved quantity and the motion of defects .
although most of the existing examples of delta shocks do not have direct physical interpretation , the case of cell walls in crystals does appear to furnish such an example .
a recurring theme in our presentation has been the importance of developing numerical techniques capable of handling defects with more exotic behavior than traditional , viscous shocks .
ideally , one would like to be able to specify whatever defect behavior is believed appropriate for a given system and have numerical techniques which respect that behavior . in lieu of such a very general approach to simulating conservation laws , we must instead focus on specific classes of defect behavior which we believe are important . to this end
, we propose further study of the stationarity condition introduced above . we have concocted an `` irregularization '' ( [ modhopf1]-[modhopf3 ] ) of the hopf equation which appears numerically to exhibit delta shocks , suggesting a possible avenue to development of the aforementioned numerical methods for delta shocks .
though we are far from showing that these apparent deltas are _ bona fide _ , we nonetheless think that the irregularization and the defects it yields are interesting in their own right and merit continued study .
we present here some further numerical and analytical observations ( mostly without details or proofs ) for the irregularization @xmath90 in particular , we sketch a systematic way to approximate the asymptotic form of solutions to this equation . three useful characteristics of the regular hopf equation can be shown to also hold for smooth solutions of this more general equation : 1 .
the maximum principle mentioned above . 2 .
zero - crossings of a solution to eq .
( [ rairreg ] ) remain fixed in time .
the total mass between any two zero - crossings is conserved .
the second and third items above allows us to predict ( analytically , approximately ) the asymptotic form of solutions to eq .
( [ rairreg ] ) .
we sketch this analysis as follows : shocks tend to form on downwards - going zero crossings , and away from shocks the solution becomes linear at large times .
the overall profile of the solution thus is linear everywhere except at the shocks , where it is approximated by the cosh profile mentioned above .
the stationarity of _ upwards_-going zero crossings allows for analytical evaluation of the slope there , which asymptotically must equal the slope of the entire linear region of the solution .
explicitly , the slope @xmath48 at a zero crossing @xmath91 satisfies the ode @xmath92 which is obtained by differentiating eq .
( [ rairreg ] ) w.r.t .
@xmath88 and noting that all terms with a factor of @xmath5 vanish .
this ode ( [ ode ] ) can be solved analytically , providing the asymptotic form of the linear part of the solution .
the third item above allows us to evaluate the mass of the delta , which is just the difference between the total initial mass between two zero - crossings and the eventual mass of the linear solution on the same region .
examining the cosh profile of the delta mentioned above , @xmath93 we see that @xmath71 must equal the height at which the delta begins , which can be determined from the linear solution .
thus @xmath71 is determined , and then @xmath72 can be determined from the known mass of the delta .
stringing together the above arguments allows a piecewise description of the solution across the entire domain .
this analytic , approximate solution is found to agree reasonably well with simulation for instance , the cosh profile shown in fig .
( [ coshfig ] ) uses parameters calculated in this way .
we thank randall leveque for helpful comments and guidance on hyperbolic conservation laws and numerical methods .
we are especially indebted to alexander vladimirsky for his exceptionally insightful suggestions for the production of this paper .
this work was supported by us doe / bes grant de - fg02 - 07er46393 .
b. l. keyfitz and h. c. kranzer _ a viscosity approximation to a system of conservation laws with no classical riemann solution _ `` nonlinear hyperbolic problems , '' lecture notes in mathematics , springer - verlag , 1989 r. meppelink , s. b. koller , j. m. vogels , p. van der straten , e. d. van ooijen , n. r. heckenberg , h. rubinsztein - dunlop , s. a. haine , and m. j. davis , _ observation of shock waves in a large bose - einstein condensate _ , physical review a , 2009 r. j. taylor , d. r. baker , and h. ikezi , _ observation of collisionless electrostatic shocks _ , physical review letters , 1970 , semidiscrete central - upwind schemes for hyperbolic conservation laws and hamilton jacobi equations , siam j. sci .
comput . , 2001 y. s. chen , w. choi , m. bierbaum , j. p. sethna , and s. papanikolaou , _ scaling theory of continuum dislocation dynamics in three dimensions : self - organized fractal pattern formation _
, international journal of plasticity , 2013 | we explore new ways of regulating defect behavior in systems of conservation laws .
contrary to usual regularization schemes ( such as a vanishing viscosity limit ) , which attempt to control defects by making them smoother , our schemes result in defects which are _ more singular _ , and we thus refer to such schemes as `` irregularizations '' . in particular , we seek to produce _ delta shock _ defects which satisfy a condition of _
stationarity_. we are motivated to pursue such exotic defects by a physical example arising from dislocation dynamics in materials physics , which we describe .
a remarkable feature of systems of conservation laws is the tendency of smooth initial data to give way to non - smooth defects after only finite time .
examples include shocks in hydrodynamics , cracks and dislocations in crystalline solids , and traffic jams in continuum models of traffic flow .
such defects are important in determining the qualitative behavior of a system , but they pose various theoretical and practical challenges , stemming from the fact that the time evolution of defects can not be determined solely from the evolution of the ambient continuum . the physical origin of this difficulty can be understood as follows : continuum equations of motion are derived from more fundamental , microscopic laws by discarding the `` high - order '' terms which are irrelevant at macroscopic scales .
thus , systems with different microphysics can yield the same continuum equations , and conversely the continuum equations alone do not specify the microphysics .
defects are inherently microscopic , and thus their behavior can not be inferred from the continuum laws .
the ambiguity of defect behavior in systems of conservation laws is reflected in the mathematical theory by _ non - uniqueness _ of solutions containing defects . to obtain uniqueness
, one must impose additional conditions upon the solution , presumably corresponding to the microphysics not accounted for by the continuum equations .
common conditions of this sort include _ entropy conditions _ and any of various types of _ regularization_. in the former case , inequalities are imposed in analogy with the thermodynamical principle that entropy must always increase , while in the latter case the conservation laws themselves are modified by the addition of small terms that serve to smooth out defects .
we will return in more detail to both of these cases in the next section .
for the most common kinds of conservation laws ( notably , those of hydrodynamics ) , the `` proper '' choice for solutions with defects turns out to be relatively tame , with defects consisting only of jump discontinuities and being described by regularization with a vanishingly small viscosity term .
some more exotic systems feature different classes of defects which require other types of regularization ; for example , superfluids are known@xcite to exhibit _ dispersive shock waves _ which are described by a very small dispersive term .
still another type of defect ( which is a primary subject of this paper ) is a _
delta shock _ , which consists of a delta function - like spike , possibly riding atop a regular ( viscous ) shock .
this latter type of defect is much more poorly understood than the former two and differs in some important regards : firstly , there is no particular regularization associated with delta shocks ( for some systems , delta shocks may even be contained in the vanishing viscosity solutions@xcite ) . secondly , there are few physically - realizable systems which are currently known to exhibit delta shocks .
the present paper was motivated by the discovery of delta shocks in a system arising from materials physics
continuum dislocation dynamics ( cdd)which to our knowledge provides the best physical interpretation of delta shocks currently available .
these delta shocks have presented substantial challenges to physicists interested in simulating cdd numerically , which has led to a desire for new numerical methods tailored to better handle such exotic defects .
we explore in this paper ( in a mostly informal way ) perspectives on delta shocks which may be of use to practitioners interested in simulating them numerically .
( we do not , however , explicitly construct any numerical schemes here . ) in particular , we describe a condition of _ stationarity _ which forces the formation of delta shocks , and we construct a new kind of `` irregularization '' which also produces delta - like defects ( though we do not rigorously prove these to be true delta shocks ) .
the outline of this paper is as follows : in the first section we provide some background on classical shocks . in the second section
we describe delta shocks and discuss various ways in which they can arise in systems of conservation laws , and we introduce the condition of stationarity and the irregularization mentioned above .
then in the third section we describe in more detail the system which initiated the preceding work
continuum dislocation dynamics . |
superallowed fermi @xmath7 transitions in nuclei , @xmath8 , provide an excellent laboratory for precise tests of the properties of the electroweak interaction , and have been the subject of intense study for several decades ( cf .
@xcite . according to the conserved - vector - current ( cvc )
hypothesis , for pure fermi transitions the product of the partial half - life , @xmath9 , and the statistical phase - space factor , @xmath10 , should be nucleus independent and given by @xmath11 where @xmath12 , @xmath13 is the vector coupling constant for nuclear @xmath7 decay , and @xmath14 is the fermi matrix element , @xmath15 . by comparing the decay rates for muon and nuclear fermi @xmath7 decay , the cabibbo - kobayashi - maskawa ( ckm ) mixing matrix element @xcite between @xmath16 and @xmath17 quarks ( @xmath18 ) can be determined and a precise test of the unitarity condition of the ckm matrix under the assumption of the three - generation standard model is possible @xcite . for tests of the standard model , two nucleus - dependent corrections must be applied to experimental @xmath19 values .
the first is a series of radiative corrections to the statistical phase - space factor embodied in the factors @xmath20 and @xmath21 , giving @xcite @xmath22 where @xmath20 is due to standard , electromagnetic ( `` inner '' ) radiative corrections ( cf .
p. 45 in ref .
@xcite ) and @xmath21 is what has been referred to as the `` outer '' radiative correction ( cf .
p. 47 of ref .
@xcite ) and includes axial - vector interference terms @xcite . the second correction , which is the subject of this work , arises because of the presence of isospin - nonconserving ( inc ) forces ( predominantly coulomb ) in nuclei that lead to a renormalization of the fermi matrix element .
this correction is denoted by @xmath23 @xcite and modifies the fermi matrix element by @xmath24 , where @xmath25^{1/2}$ ] is the value of the matrix element under the assumption of pure isospin symmetry . with the corrections @xmath20 , @xmath21 , and @xmath23 ,
a `` nucleus - independent '' @xmath26 can be defined by @xmath27 and the ckm matrix element @xmath18 is given by @xcite @xmath28 where the fermi coupling constant , @xmath29 , is obtained from muon @xmath7-decay , and includes radiative corrections .
currently , @xmath19 values for nine superallowed transitions have been measured with an experimental precision of 0.2% or better @xcite . with these precise measurements and reliable estimates for the corrections ,
the cvc hypothesis can be confirmed by checking the constancy of the @xmath26 values for each nucleus , while the unitarity condition of the ckm matrix is tested by comparing the average value of @xmath18 with the values determined for @xmath30 @xcite and @xmath31 ( 90% confidence level ) @xcite , i.e. , @xmath32 . in the past
, the nuclear structure correction @xmath23 has been computed within the framework of the nuclear shell model @xcite . in general , the isospin - nonconserving components of the nuclear hamiltonian are small , and can be treated perturbatively . due to computational limitations and uncertainties associated with determining an effective hamiltonian , almost all calculations for nuclei with @xmath33 have been performed within a single major oscillator shell , e.g. , for @xmath1c the model space spanned by the @xmath34 and @xmath35 oribitals ( _ p_-shell ) . within this context , two types of isospin mixing must be accounted for .
the first is due to the mixing between states that lie within the shell - model configuration space .
for example , for @xmath36 , there are 2 , 7 , and 1 _ p_-shell configurations leading to @xmath37 and @xmath38 , 1 , and 2 , respectively .
because of its two - body nature , the inc interaction is composed of isospin operators of rank zero ( isoscalar ) , one ( isovector ) , and two ( isotensor ) , and in the case of @xmath36 it is capable of mixing together all @xmath37 states .
traditionally , the configuration mixing correction is denoted as @xmath39 and in ref .
@xcite it was shown that the best estimates for @xmath39 are obtained using an inc interaction that correctly describes the coulomb energy splittings of the binding energies between members of the isospin multiplet , e.g. , the @xmath40 states in @xmath1c , @xmath1b , and @xmath1be .
of the two types of mixing , @xmath39 is the smallest with a magnitude of approximately 0.04 - 0.1% . in addition to
the mixing between states contained within the shell - model configuration space , mixing with states that lie outside the model space must also be accounted for . in particular , the coulomb interaction can strongly mix one particle - one hole ( @xmath41 ) @xmath42 excitations , e.g. , @xmath43 , into the ground state . in previous works ,
excitations of this type were accounted for by examining differences in the single - particle radial wave functions . indeed , for closed - shell configurations , mixing with @xmath41 states is properly accounted for at the level of hartree - fock .
hence , the second correction to the fermi matrix element , denoted by @xmath44 , was estimated by evaluating the mismatch in the radial overlap between the single - particle wave functions of the converted proton and the corresponding neutron .
the explicit details for the calculation of @xmath44 , which involve a sum over intermediate @xmath45 parent states that then determine the proton and neutron separation energy for the radial wave function , are given in refs .
@xcite . for the most part
, @xmath44 is found to be the larger of the two components ( with @xmath46 ) and has a magnitude of the order 0.1 - 0.8% . at present ,
two methods for evaluating @xmath44 are espoused .
the first ( thh ) @xcite uses woods - saxon ( ws ) radial wave functions , while in the second ( ob ) @xcite , hartree - fock ( hf ) wave functions are employed .
generally speaking , the two methods yield approximately the same dependence on nucleon number @xmath47 , but the hf values are systematically smaller by 0.1% for the magnitude of the correction .
the reason for the difference lies in the hf mean field .
the principal effect of the coulomb interaction is to push the proton wave functions out relative to the neutrons , hence , providing a mismatch in the radial overlap . in hartree - fock , however , the proton and neutron mean fields are coupled , and the coulomb interaction actually induces an attractive isovector mean field between the protons and neutrons . in effect , the coulomb interaction pushes the protons out , but because of the strong interaction , the protons pull the neutrons out with them , hence , reducing the magnitude of the radial overlap mismatch .
when all _ known _ corrections , i.e. , @xmath20 , @xmath21 , @xmath44 , and @xmath39 , are applied to the nine experimental data @xcite , it is found that the @xmath26 values are essentially constant within the limits of uncertainty but the unitarity limit is violated at the level of approximately 0.4(1)% or 0.3(1)% for the ob and thh corrections , respectively .
in addition , preliminary data from a new experiment for @xmath1c@xcite leads to an @xmath26 value that is significantly smaller than that of ref .
@xcite , and has been interpreted as possible evidence for an , as yet , unaccounted for correction that might lead to satisfying the unitarity condition of the ckm matrix .
in addition , it must be admitted that the present separation between the configuration mixing and radial overlap contributions to @xmath23 is somewhat unsatisfying .
a much better approach would be to perform a shell - model calculation that includes several @xmath48 excitations , so that both corrections would evaluated on the same footing and simultaneously .
because of recent improvements in computational capabilities and the ability to determine an effective model - space hamiltonian based on realistic nucleon - nucleon interactions , it is now possible to perform such a calculation for the lightest of the nine accurately measured transitions .
we report here the results of large - basis shell - model calculations that include excitations up to @xmath0 for @xmath36 nuclides , with an emphasis on evaluating the isospin - mixing corrections to the matrix element for the fermi decay of @xmath1c .
the organization for the paper is as follows .
first , in section [ sec2 ] we discuss the shell - model hamiltonian with a bound center - of - mass , the method used to derive the starting - energy - independent effective interaction , and the renormalization of the transfer operator .
results of the fermi matrix element calculations are presented in section [ sec3 ] and concluding remarks are given in section [ sec4 ] .
in our calculation we use the one- plus two - body hamiltonian for the a - nucleon system , i.e. , @xmath49 where @xmath50 is the nucleon mass and @xmath51 the nucleon - nucleon interaction , modified by adding the center - of - mass harmonic - oscillator potential @xmath52 , @xmath53 .
this potential does not influence intrinsic properties of the many - body system .
it provides , however , a mean field felt by each nucleon and allows us to work with a convenient harmonic - oscillator basis .
the modified hamiltonian , depending on the harmonic - oscillator frequency @xmath54 , may be cast into the form @xmath55 + \sum_{i
< j}^a \left [ v_{\rm n}(\vec{r}_i-\vec{r}_j ) -\frac{m\omega^2}{2a } ( \vec{r}_i-\vec{r}_j)^2 \right ] \ ; .\ ] ] the one - body term of the hamiltonian ( [ hamomega ] ) is then re - written as a sum of the center - of - mass term @xmath56 , @xmath57 , and a term depending on relative coordinates only .
shell - model calculations are carried out in a model space defined by a projector @xmath58 . in the present work
, we will always use a complete @xmath59 model space .
the complementary space to the model space is defined by the projector @xmath60 .
in addition , from among the eigenstates of the hamiltonian ( [ hamomega ] ) , it is necessary to choose only those corresponding to the same center - of - mass energy .
this can be achieved by projecting the center - of - mass eigenstates with energies greater than @xmath61 upwards in the energy spectrum .
the shell - model hamiltonian , used in the actual calculations , takes the form @xmath62p + \sum_{i < j}^a p\left [ v_{ij}-\frac{m\omega^2}{2a } ( \vec{r}_i-\vec{r}_j)^2 \right]_{\rm eff } p \nonumber \\ & + & \beta p(h^\omega_{\rm cm}-\frac{3}{2}\hbar\omega)p\ ; , \end{aligned}\ ] ] where @xmath7 is a sufficiently large positive parameter . the effective interaction introduced in eq .
( [ phamomegabeta ] ) should , in principle , exactly reproduce the full - space results in the model space for some subset of states . in practice , the effective interactions can never be calculated exactly as , in general , for an a - nucleon system an a - body effective interaction is required .
consequently , large model spaces are desirable when only an approximate effective interaction is used . in that case , the calculation should be less affected by any imprecision of the effective interaction .
the same is true for the evaluation of any observable characterized by an operator . in the model space ,
renormalized effective operators are required .
the larger the model space , the less renormalization is needed .
usually , the effective hamiltonian is approximated by a two - body effective interaction determined from a two - nucleon system . in this study
, we use the procedure as described in ref .
@xcite . to construct the effective interaction
we employ the lee - suzuki @xcite similarity transformation method , which gives an interaction in the form @xmath63 , with @xmath64 the transformation operator satisfying @xmath65 .
the projection operators @xmath66 project on the two - nucleon model and complementary space , respectively .
our calculations start with exact solutions of the hamiltonian @xmath67 which is the shell - model hamiltonian ( [ hamomega ] ) applied to a two - nucleon system .
we construct the effective interaction directly from these solutions .
let us denote the relative - coordinate two - nucleon harmonic - oscillator states , which form the model space , as @xmath68 , and those which belong to the q - space , as @xmath69 .
then the q - space components of the eigenvector @xmath70 of the hamiltonian ( [ hamomega2 ] ) can be expressed as a combination of the p - space components with the help of the operator @xmath64 @xmath71 if the dimension of the model space is @xmath72 , we may choose a set @xmath73 of @xmath72 eigenevectors , for which the relation ( [ eigomega ] ) will be satisfied . under the condition that the @xmath74 matrix @xmath75 for @xmath76 is invertible
, the operator @xmath64 can be determined from ( [ eigomega ] ) . in the present application we select the lowest states obtained in each channel .
once the operator @xmath64 is determined the effective hamiltonian can be constructed as follows @xmath77 \ ; .\ ] ] this hamiltonian , when diagonalized in a model - space basis , reproduces exactly the set @xmath73 of @xmath72 eigenvalues @xmath78 .
note that the effective hamiltonian is , in general , quasi - hermitian .
it can be hermitized by a similarity transformation determined from the metric operator @xmath79 .
the hermitian hamiltonian is then given by @xcite @xmath80^{1/2 } h_{\rm 2eff}\left[p_2(1+\omega^\dagger\omega ) p_2\right]^{-1/2 } \ ; .\ ] ] finally , the two - body effective interaction used in the present calculations is determined from the two - nucleon effective hamiltonian ( [ hermeffomega ] ) as @xmath81 .
note that we distinguish the two - nucleon system projection operators @xmath82 from the a - nucleon system operators @xmath83 .
to at least partially take into account the many - body effects neglected when using only a two - body effective interaction , we employ the recently introduced multi - valued effective interaction approach @xcite . as a consequence ,
different effective interactions are used for different @xmath48 excitations .
the effective interactions then carry an additional index indicating the sum of the oscillator quanta for the spectators , @xmath84 , defined by @xmath85 where @xmath86 and @xmath87 are the total oscillator quanta in the initial and final many - body states , respectively , and @xmath88 and @xmath89 are the total oscillator quanta in the initial and final two - nucleon states @xmath90 and @xmath91 , respectively .
@xmath92 is the minimal value of the spectator harmonic - oscillator quanta for a given system . here , for a=10 , @xmath93 .
different sets of the effective interaction are determined for different model spaces characterized by @xmath84 and defined by projection operators [ projop]@xmath94 in eqs .
( [ projop ] ) , @xmath95 characterizes the two - nucleon model space .
it is an input parameter chosen in relation to the size of the many - nucleon model space .
this multi - valued effective - interaction approach is superior to the traditional single - valued effective interaction , as confirmed also in a model calculation @xcite .
our goal in this study is to evaluate the fermi matrix element @xmath96 which is equal to @xmath97 for an isospin - invariant system .
note that for a system with isospin breaking , the isospin - lowering operator @xmath98 should be renormalized in a similar way , as the interaction used for calculation of the eigenstates appearing in eq .
( [ fermime ] ) .
in fact , we can apply the formalism described in refs .
@xcite to construct a two - body effective operator @xmath99 consistent with the two - body effective interaction derived above and exact for the two - nucleon system .
then we could use such an operator in the a - body calculation .
we studied such a possibility in a solvable - model calculation as described in refs .
@xcite . here
, we did two - nucleon calculations with the effective @xmath98 operator .
the observed renormalization of the bare operator for the model spaces of the size used in our calculations , was , however , insignificant compared to the other effects as described further .
therefore , in the a - body calculations we used the bare @xmath98 operator .
in order to evaluate the fermi matrix element [ eq .
( [ fermime ] ) ] , we apply the formalism outlined in section [ sec2 ] for a=10 nuclei . in the calculations we use the reid93 nucleon - nucleon potential @xcite and consider the following isospin - breaking contributions .
first , the reid93 potential differs in the @xmath3 channels for pn and pp , nn systems , respectively .
second , we add the coulomb potential to the pp reid93 potential . consequently , using the eqs .
( [ eigomega])-([hermeffomega ] ) , we derive different two - body effective interactions for the pn , pp , and nn systems .
no other mechanisms for isospin breaking are considered .
as we derive the effective interaction microscopically from the nucleon - nucleon interaction , the number of freely adjustable parameters in the calculation is limited .
first , we have the choice of the model - space size in the shell - model diagonalization .
that is , however , constrained by computer capabilities .
the largest model space we were able to use was the space allowing all @xmath0 excitations relative to the unperturbed ground state .
most of the calculations were done in the m - scheme using the many - fermion - dynamics code @xcite extended to allow the use of different pn , pp , nn interactions .
we also performed some calculations with the oxbash shell - model code @xcite . in the m - scheme ,
the dimensions associated with @xmath1b and @xmath1c are 581,740 and 430,137 , respectively . to study the dependence on the model - space size , we performed calculations in the @xmath42 space as well . in that space , the dimensions drop to 14,502 and 10,111 , respectively .
second , we have the choice of the two - nucleon model space used for the evaluation of the effective interaction .
this is related to the many - nucleon model - space size , and , in principle , is determined by that size .
traditionally , however , the @xmath100 space used to determine the g - matrix does not , necessarily , coincide with the many - particle model space @xcite . in our calculation , the two - nucleon model space is characterized by a restriction on the number harmonic - oscillator quanta @xmath101 , @xmath102 , @xmath103 . here , @xmath104 is the harmonic - oscillator quantum number for the nucleon @xmath105 .
this type of restriction guarantees an orthogonal transformation between the two - particle states and the relative- and center - of - mass - coordinate states . for the present @xmath0 calculation
, the choice of @xmath106 appears to be appropriate .
however , it has been observed in the past @xcite that when the lee - suzuki procedure combined with the g - matrix calculation according to ref .
@xcite ( which is equivalent to the procedure we are using ) is applied to calculate the two - body effective interaction , the resulting interaction may be too strong .
this is , in particular , true , when the multi - valued approach is used .
several possible adjustments were discussed to deal with this problem @xcite and amounted to introducing an extra parameter . in the present calculations , we do not introduce any new parameter , but rather we treat @xmath95 as a free parameter and use @xmath107 for the @xmath0 calculations and @xmath106 for the @xmath42 calculations , respectively . with this choice
we obtain quite reasonable binding energies for the studied nuclei .
we have also performed several @xmath42 test calculations with single - valued interactions that were derived following ref.@xcite , as opposed to the multi - valued interaction discussed in the previous section . to obtain reasonable binding energies with the single - valued interaction
we do not have to change the @xmath95 value from that corresponding to the many - nucleon space , e.g. , @xmath108 for the @xmath42 calculation .
this difference of treatment of the two types of interactions follows from the fact that the overall strength of the single - valued interaction is weaker .
third , our results depend on the harmonic - oscillator frequency @xmath54 .
we have studied this dependence by performing calculations for the values @xmath109 , and @xmath110 mev .
let us also mention one important feature of the present approach .
for both the multi - valued and the single - valued interactions our calculations do not break the separation of the center - of - mass and the internal relative motion . in particular , a variation of the parameter @xmath7 introduced in eq .
( [ phamomegabeta ] ) does not change the eigenenergies and other characteristic of the physical states .
this is so due to the choice of a complete @xmath59 many - nucleon space and the triangular two - nucleon model space for deriving the effective interaction as well as the due to the procedure used to derive the effective interaction . in figs .
[ fig : b1014 ] , [ fig : b10155 ] , and [ fig : b1017 ] we present the experimental and calculated spectra of @xmath1b for @xmath109 , and @xmath110 mev , respectively , for the @xmath42 and @xmath0 model spaces . in general
, we observe an overall improvement in the spectra with the enlargement of the model space in all three cases . also , the @xmath0 calculations exhibit more stability with regard to changes in the harmonic - oscillator frequency than do the @xmath42 results .
the agreement with experiment improves when going from @xmath111 mev to @xmath112 mev ; in particular for the ground state and the lowest states . in fact , from fig .
[ fig : b1017 ] we find that a very reasonable description of the spectra is obtained for @xmath112 mev . in table
[ tab : tab1 ] , the overall behavior with respect to @xmath48 is illustrated .
in general , we observe a reasonable reproduction of the binding energy , with a moderate decrease occuring for increasing @xmath48 .
using free - nucleon effective charges , we find that although the quadrupole moment for the @xmath113 state is underestimated considerably , the magnetic dipole moment is well reproduced . in addtion , the point - proton rms radius exhibits a fairly strong dependence , and increases with decreasing @xmath48 . for the rms radius , we find that the best agreement with experiment @xcite is achieved for @xmath111 mev
. from the point of view of the beta decay of @xmath1c , a good description of the @xmath3 states is important . from figs .
[ fig : b1014]-[fig : b1017 ] we can see that the calculated @xmath1b @xmath3 states have the right relative positions and are reasonably stable with variations of both the model - space size and @xmath48 . we have also performed @xmath0 calculations for @xmath1be to study the splitting of the isospin analog states in the whole isospin - multiplet @xmath1c - @xmath1b - @xmath1be . the experimental ground state splitting between @xmath1c and @xmath1be is 4.66 mev , while our calculated values are 4.68 , 4.83 , and 4.94 mev for @xmath109 , and @xmath110 mev , respectively .
the best agreement with experiment is achieved for @xmath111 mev , where the calculated rms point - proton radius is also in agreement with the experimental value . on the other hand ,
the splitting between the @xmath114 states of @xmath1c and @xmath1b , which is experimentally 2.69 mev , is overestimated in our calculations by @xmath115 , and @xmath116 for the @xmath1175 , and @xmath110 mev calculations , respectively .
since the correct @xmath1c - @xmath1be splitting is obtained for @xmath111 mev , the excess in the @xmath1c - @xmath1b splitting suggests that the isospin breaking due the strong @xmath3 force may be too large .
one possible explanation is that our approach for deriving the effective interaction tends to exaggerate the differences between the pn and nn , pp potentials .
such an artificial effect should decrease with increasing model - space size . on the other hand , it is also possible that the reid93 potential itself overestimates differences between the pn and pp , nn systems in the @xmath3 channel . for the most part , we find the best overall agreement for the rms point proton radius , binding energy , and coulomb energy splitting for @xmath111 mev .
given that the isospin mixing is largely driven by the coulomb interaction , which is then dependent on the size of the nucleus , we feel that the best value for the isospin - mixing correction to the fermi matrix element will be achieved for @xmath111 mev .
the most important results of our study are also summarized in table [ tab : tab1 ] in the last two lines .
the calculated isospin - mixing corrections @xmath118 , in % , are presented for all three choices of @xmath54 and for both @xmath0 and @xmath42 model spaces .
again , a correlation between the radius and the isospin - mixing correction is clearly observed , as @xmath23 decreases with increasing radius .
this is simply understood in terms of a larger radius implying weaker coulomb effects . on the other hand , with an increase in the model - space size , a significant increase in the isospin - mixing correction
is apparent .
this is due to the fact that in the larger model space , the excitation energies of the 1p-1h @xmath114 states decrease , hence leading to greater mixing .
for this reason , the more realistic multi - valued effective interaction is important .
we have also performed test calculations with the single - valued interaction in the @xmath42 space , and found @xmath6 to be smaller by approximately 30% .
our @xmath0 results suggest an isospin - mixing correction @xmath119 .
this is compatible with the previously published value of @xmath120 by ormand and brown @xcite . that value was a sum of two contributions .
first , about 0.04% came from the shell - model wave - function renormalization due to the isospin mixing and was obtained in a @xmath5 shell - model calculation using phenomenological effective interactions .
second , the amount 0.09% was due to the deviation from unity of the radial overlap between the converted proton and the corresponding neutron .
this effect was attributed to the influence of states lying outside the @xmath5 space .
the radial wave - functions were obtained in a hartree - fock calculation using skyrme - type interactions .
because we use a multi - configuration model space in the present calculation , we should have both effects included consistently at the same time .
another important factor in the calculation is the position of the @xmath42 states . as discussed before
, the position of the 1p-1h states influences the ground - state isospin mixing .
unfortunately , the excitation energy of these states is not known experimentally . however , in our calculations the multi - valued effective interaction is used and a more realistic description of these states should be obtained ; especially , in the @xmath0 model space . on the other hand , in an analogous calculation for @xmath121he , it was observed that an @xmath122 model space is needed to get the @xmath42 dominated @xmath123 state close to the experimental excitation energy @xcite .
there are states like @xmath124 at 5.18 mev in @xmath1b or @xmath114 at 6.18 mev in @xmath1be , which are believed to be 2p-2h , @xmath42 excitations .
we do not observe any such states below 7.5 and 12 mev , respectively , in our calculations .
the first excited @xmath114 state in @xmath1be obtained in the @xmath0 calculation with @xmath111 mev lies at 9.8 mev .
it is , however , predominately a @xmath5 state .
there can be two reasons why we do not get such states .
first , these states have not yet converged in the lanczos procedure .
second , and more likely , the @xmath0 model space is too small for the right description of the @xmath42 excitation states .
therefore , it would be desirable to extend the present calculations to a larger , e.g. , @xmath125 model space . unfortunately , due to the computational limitations , it is not possible at this time to perform a calculation of this magnitude .
however , from table [ tab : tab1 ] we observe an increase of @xmath6 by @xmath126 between the @xmath42 and the @xmath0 calculation .
therefore , we might expect an increase of similar magnitude for an increase of the model - space size beyond @xmath0 .
therefore , the more realistic value of the isospin - mixing correction from our calculation would be @xmath127 , where the uncertainty is estimated from the change in @xmath23 obtained when using an increased model space .
the effects of isospin mixing on the transition matrix element for the superallowed fermi @xmath7-decay of @xmath1c were estimated within the context of a large - basis , shell - model calculation .
the calculations were performed assuming no closed core and an effective interaction based on a realistic two - body nucleon - nucleon interaction , while including the coulomb interaction between protons .
contrary to previous estimates for the isospin corrections , this calculation was carried out within a model space that included many @xmath48 excitations . as a consequence , the conventional configuration mixing and radial mismatch contributions were evaluated within a unified framework , simultaneously and the usual separation was not necessary . with regard to parameters used within the calculation , we find a correlation between the isospin - mixing correction and the coulomb splitting between the isotopic multiplets , which , in turn , is governed by the nuclear size through the oscillator parameter . given that the isospin - mixing correction is primarily a coulomb effect , the best value for @xmath23 is taken to coincide with the oscillator parameter that correctly reproduces the coulomb splittings . with regard to the model - space size , a clear improvement ( or an indication towards convergence ) in most observables is evident when the size of the model space is increased from @xmath42 to @xmath0 , but @xmath23 is found to increase by only 0.03% ( in magnitude ) in this case . hence , our final estimate for @xmath23 is taken to be 0.12(3)% ( where the @xmath0 result has been increased by 0.03% to account for the possible effects of an increased model space ) .
this result also happens to be in excellent agreement with the previous estimates that relied on the conventional separation of the configuration and radial mismatch contributions . finally , we note that the magnitude of the isospin - mixing correction obtained in our calculation does not lead to a resolution to the deviation from unitarity for the cabibbo - kobayashi - maskawa matrix .
r. j. blin - stoyle , in _ isospin in nuclear physics _ , edited by d. h. wilkinson ( north - holland , amsterdam , 1969 ) , p. 115 ; d. h. wilkinson , phys
65b * , 9 ( 1976 ) ; i. s. towner , s. raman , t. a. walkiewicz , and h. behrens , at .
data nucl .
data tables * 16 * , 451 ( 1975 ) .
t. t. s. kuo , p. j. ellis , jifa hao , zibang li , k. suzuki , r. okamoto , and h. kumagai , nucl .
a560 * , 621 ( 1993 ) ; p.navrtil , h.b . geyer and t.t.s .
kuo , phys .
* b 315 * , 1 ( 1993 ) ; k. suzuki and r. okamoto , prog .
. phys . * 93 * , 905 ( 1995 ) .
.experimental and calculated binding energies , in mev , magnetic moments , in @xmath128 , and quadrupole moments , in @xmath129 @xmath130 , of @xmath1b .
also the experimental and calculated binding energies , in mev , and the point proton radius , in fm , of @xmath1c are presented .
the results correspond to the 4@xmath48 calculations .
in addition , the isospin - mixing correction @xmath6 , in % , is shown as obtained both in the 4@xmath48 calculations and the 2@xmath48 calculations .
results of three different calculations with the harmonic - oscillator parameter taken to be @xmath131 mev , respectively , are presented .
the effective interaction used was derived from the reid 93 nucleon - nucleon potential .
the experimental values are taken from refs .
[ cols="^,^,^,^,^",options="header " , ] | we use a @xmath0 shell - model calculation with a two - body effective interaction derived microscopically from the reid93 potential to calculate the isospin - mixing correction for the @xmath1c@xmath2b superallowed fermi transition .
the effective interaction takes into account the coulomb potential as well as the charge dependence of @xmath3 partial waves .
our results suggest the isospin - mixing correction @xmath4 , which is compatible with previous calculations .
the correction obtained in those calculations , performed in a @xmath5 space , was dominated by deviation from unity of the radial overlap between the converted proton and the corresponding neutron . in the present calculation
this effect is accommodated by the large model space .
the obtained @xmath6 correction is about a factor of four too small to obtain unitarity of the cabibbo - kobayashi - maskawa matrix with the present experimental data . |
the first evidence for an accelerated expansion of the universe at the present epoch was presented in the late 1990s by riess et al .
@xcite and perlmutter et al .
@xcite . fitting a cosmological model including a cosmological constant matched the supernovae type ia of the considered data sets significantly better than a model without such a constant . in the following years , a lot of effort was put into getting larger supernova ( sn ) sets of increasing quality @xcite . on the other hand , a variety of cosmological models
have been developed that could be fit to the newest data sets .
those models are characterized by certain parameters which are used to derive a redshift - luminosity relation that can be compared to the observed values .
the problem with this so - called dynamical approach is that it is impossible to test without assumptions on the matter and energy content of the universe if there really is a phase of acceleration in the expansion history of the universe .
some authors tried to avoid this problem by taking a kinematical approach , i.e. they only considered the scale factor @xmath1 and its derivatives , such as the deceleration parameter @xmath2 , without using any model specific density parameters or a dark energy equation of state .
the first kinematical analysis of sn data was done by turner and riess @xcite who considered averaged values @xmath3 for redshift @xmath4 and @xmath5 for @xmath6 concluding that a present acceleration and a past deceleration is favoured by the data .
other authors tested a variety of special parameterizations of @xmath2 @xcite or used @xmath7 @xcite or the hubble rate @xmath8 @xcite . instead of considering a special parameterization ,
a more general approach has been done by shapiro and turner @xcite who expanded the deceleration parameter @xmath9 into principal components .
rapetti et al .
@xcite expanded the jerk parameter @xmath10 into a series of orthonormal functions .
however , one has to be careful when doing a series expansion in @xmath11 as sne with a redshift larger than 1 are not within the radius of convergence .
this problem can be solved by reparameterizing redshift @xcite . in the present work we will neither make any assumptions about the content of the universe , nor about the parameterization of @xmath2 or another kinematical quantity .
moreover , we do not need to assume the validity of einstein s equations .
instead , we will only ask the question if the hypothesis holds that the universe never expanded accelerated between the time when the light was emitted from a sn and today .
our assumption for the test is that the universe is isotropic and homogeneous .
this is certainly not true for the real universe in a strict sense .
however , we follow the standard approach in assuming that the cosmic structure does not modify the observed sn magnitudes and redshifts apart from random peculiar motion .
the basic idea of this analysis has already been presented by visser @xcite and has been applied to sn data by different groups @xcite .
but while santos et al .
@xcite made some mistakes in their analysis ( that we will discuss later in section [ concl ] ) and gong et al .
@xcite only state that accelerated expansion is `` evident '' , we are able to give a quantitative value for this evidence
. additionally , we study the size of systematic effects
. for the fit one usually marginalizes over a function of the absolute magnitude @xmath12 and the hubble constant @xmath13 because these two values can not be determined independently by only considering sne .
marginalization is not suitable for our analysis because in order to do so a special cosmological model or at least a parametization of @xmath2 has to be inserted , which is what we want to avoid .
but the sne can be calibrated by cepheid meassurements and thus the values of @xmath12 and @xmath13 are determined . using this additional information ,
no marginalization is needed .
as there is still a controversy @xcite about the appropriate calibration method , we will take two quite different results of calibration into account @xcite . in order to test the robustness of our analysis ,
we consider two different sn data sets ( the 2007 gold sample @xcite and the essence set @xcite ) , where the data of the essence set are once obtained by the multicolour light - curve shape ( mlcs2k2 ) @xcite fitting method and once by the spectral adaptive light - curve template ( salt ) @xcite .
both calibration methods are applied to each set .
the analysis shows that in all cases the data indicate an accelerated expansion .
but the confidence level at which acceleration can be stated strongly depends on the data set , the fitting method and the calibration .
we begin our analysis by assuming a flat universe , but later also consider the cases of an open and a closed universe .
we want to keep our test as model - independent as possible , but still a few assumptions have to be made .
we consider inflationary cosmology to be correct .
this implies large scale homogeneity and isotropy of the universe as well as spatial flatness .
( yet , we will give up the assumption of a flat universe later in section [ openclosed ] . )
we also assume that cosmological averaging ( here along the line of sight ) does not modify the result obtained in a friedmann - lematre model .
in such a universe , the luminosity distance @xmath14 is given by @xmath15 as we are interested in the question whether the universe accelerates or decelerates at the present epoch , we need to examine the deceleration parameter @xmath2 : if @xmath9 is positive , the universe expands decelerated , for a negative @xmath9 it accelerates .
@xmath2 can be expressed in terms of the hubble parameter @xmath8 : @xmath16 where the prime denotes the derivative with respect to @xmath11 . integrating this equation yields @xmath17
our null hypothesis is that the universe never expanded accelerated , i.e. @xmath18 for all @xmath11 . under this assumption
, the above equation turns into the inequality @xmath19 or @xmath20 .
thus , for the luminosity distance we have @xmath21 in order to test our hypothesis , we consider different data sets of sne type ia .
if the observed luminosity distance is significantly larger than the luminosity distance of a universe with a constant @xmath22 , the hypothesis can be rejected at a high confidence level .
note that the rejection of the hypothesis does not mean that there was no deceleration between the time the light was emitted from a sn until now , it only gives evidence that there was at some time a phase of acceleration .
thus , we can not determine the transition redshift between deceleration and acceleration .
this restriction to our analysis comes from the integral over redshift in the calculation of @xmath14 . as a first step ,
the data in the considered sets have to be calibrated consistently .
the distance modulus @xmath23 is related to the luminosity distance by @xmath24 where @xmath25 is given in units of mpc .
the distance modulus of the sne ia in those sets is always given in arbitrary units because the absolute magnitude @xmath12 can not be measured independently of the hubble constant @xmath13 . only the apparent magnitude @xmath26 and thus the relative distance moduli are measured at high precision . in order to determine the absolute magnitude , the sne have to be calibrated by measuring the distance of cepheids in the host galaxies . then also the hubble constant can be determined .
but there is still disagreement between different groups about the correct calibration analysis . in this work
we will consider two results for @xmath12 and @xmath13 , namely that of riess et al .
@xcite and that of sandage et al .
@xcite . in the following
we will refer to those results as riess calibration and sandage calibration , respectively .
the difference in the calibration comes mainly from the different assumptions on the evolution of the cepheid period - luminosity relation with host metallicity @xcite .
so for preparing the data for our analysis , first the distance moduli @xmath27 have to be adjusted to the assumed absolute magnitude .
we define the magnitude @xmath28 as the observed distance modulus @xmath27 of the @xmath29-th sn minus the distance modulus in a universe with constant deceleration parameter @xmath22 at the same redshift @xmath30 : @xmath31 -25 \,.\ ] ] if the error in redshift and the peculiar velocities of the sne are already included in the error @xmath32 given in a certain data set , then the error @xmath33 of @xmath28 equals @xmath32 .
otherwise , the resulting error of @xmath28 is calculated by : @xmath34^{\frac{1}{2 } } \,.\ ] ] let @xmath35 be the theoretical distance modulus of a cosmological model , which describes the expansion of the universe correctly
. then @xmath36 would be the `` true '' value corresponding to the meassured sn value @xmath28 .
the null hypothesis for our test is that the universe never expanded accelerated , i.e. @xmath37 for each sn .
we reject that hypothesis if the measured value @xmath28 lies above a certain action limit @xmath38 , otherwise we accept it
. we want to keep the risk low that we conclude a late time acceleration of the universe , when there is indeed no acceleration at all .
therefore the action limit must be relatively high .
if we want the confidence level for concluding an accelerated expansion to be 99% , the action limit must be @xmath39 .
for a confidence level of 95% , the limit is @xmath40 .
here we assume that the measured values @xmath28 at a given redshift follow a normal distribution . on the other hand , we can test the hypothesis that the universe expanded accelerated all the time from light emission of a sn until today . this hypothesis can be rejected at a 99% cl if @xmath28 is below the action limit @xmath41 .
that would mean that there must have been a phase of deceleration in the late time universe , but it would not exclude a phase of acceleration .
the important parameters of the sne are obtained by fitting their light - curves .
the results depend on the fitter that is used .
the most common fitter is mlcs2k2 @xcite . here
we will also consider the salt fitting method @xcite .
the main difference between the two methods is in how the peak luminosity corrections are determined by the colour of the sne @xcite .
while mlcs2k2 assumes that the corrections that have to be made are only due to dust , salt takes an empirically approach to determine the relation between the colour and the luminosity correction . for the analysis we take the following sn ia
data sets : gold 2007 ( mlcs2k2 ) : the gold sample by riess et al .
( 2007 ) @xcite , which was obtained by using the mlcs2k2 fitting method essence ( mlcs2k2 ) : the set given by wood - vasey et al .
( 2007 ) @xcite , which includes data from essence , snls @xcite and nearby sne @xcite , fitted with mlcs2k2 essence ( salt ) : the same set fitted with salt as suggested by riess et al .
@xcite , we discarded all sne with a redshift of @xmath42 from the gold sample , which leaves us 182 sne with @xmath43 .
in essence ( mlcs2k2 ) and essence ( salt ) , the sne with bad light curve fits were rejected for each fitting method
this leaves 162 sne for essence ( mlcs2k2 ) and 178 for essence ( salt ) with @xmath44 .
153 of the sne are contained in both sets . due to the differences of the sne contained in essence ( mlcs2k2 ) and essence ( salt )
we will refer to them as different sets in the following .
nevertheless , you should keep in mind that they share a large number of sne and thus are not independent sets .
the riess calibration yields a value for the v - band magnitude of @xmath45mag , where @xmath46 is the time of the b - band maximum . for the gold 2007 set @xmath47
was considered to be @xmath48mag .
therefore , in order to get the appropriate distance modulus @xmath49 , we need to substract @xmath50mag from the value given in the gold sample . from the distance modulus values in the essence sets
@xmath51mag have to be substracted because in these sets a b - band magnitude of @xmath52mag is assumed and riess et al .
@xcite give the relation @xmath53 .
for this calibration one gets a hubble constant of @xmath54 . the results of sandage et al .
@xcite for the sne absolute magnitudes are @xmath55mag and @xmath56mag and for the hubble constant @xmath57(statistical ) @xmath58 5.0(systematic ) .
so considering the sandage calibration , we have to add 0.02mag to distance moduli of the gold sample and substract 0.01mag from the values given in the essence sets .
the relevant parameter for our analysis is @xmath28 given by equation .
its values are plotted in figure [ deltamufig ] for the sne of the three data sets and the two different calibration methods .
the curve for a flat @xmath59cdm model ( with @xmath60 ) is shown in the plots in order to give the reader a notion of how the redshift - dependency of @xmath61 could look like .
most of the data values are positive which indicates an accelerated expansion .
a similar diagram has been presented by gong et al .
@xcite for single sne as well as for binned sn data where they used a combined gold and essence data set .
they kept the arbitrary value of the absolute magnitude @xmath12 given in the set and then determined the hubble constant by using sne with @xmath62 to be @xmath63 . at this point
they stopped their analysis by concluding that due to the large number of sne that lie above the curve of a universe with @xmath22 , accelerated expansion is evident .
but the fact that there are also several data points below that curve and that the errors @xmath33 ( with a typical value of 0.23mag for mlcs2k2 and 0.18mag for salt ) are approximately of the same size as the values @xmath28 give rise to the question how certain acceleration really is .
thus , we will provide a more quantitative analysis in this work .
+ + we counted the sne of each set whose values of @xmath28 were above the action limit @xmath64 and those above @xmath39 ( which indicates acceleration at a 95% and a 99% cl , respectively ) and those with values below @xmath65 and below @xmath66 ( which indicates deceleration ) .
the results are shown in table [ acctab1 ] .
the clearest evidence for an accelerated expansion is given by the essence ( salt ) set with the riess calibration with 64 sne indicating acceleration at a 99% confidence level and none indicating deceleration . also in the other sets accelerated expansion is preferred with the exception of essence ( mlcs2k2 ) in the sandage calibration where none of the two expansion histories is preferred in the test using the action limit @xmath67 .
@lllllll & & & + & riess & sandage & riess & sandage & riess & sandage + acceleration ( 95% cl ) & 37 & 27 & 37 & 14 & 96 & 53 + acceleration ( 99% cl ) & 13 & 7 & 11 & & & 30 + deceleration ( 95% cl ) & 2 & 4 & 3 & 7 & 1 & 7 + deceleration ( 99% cl ) & 0 & 1 & 2 & & & 2 + number of sne & 182 & 182 & 162 & 162 & 178 & 178 + it is noticeable that the two light - curve fitting methods for the essence set yield very different results .
( for a discussion on these differences at low redshifts , see @xcite . )
this could be an indication that at least one of the two fitting methods does not give the correct result . in order to consider this possibility we take a look at the differences between @xmath28 obtained by salt and @xmath28 obtained by mlcs2k2 .
they are shown in figure [ mudiff ] for the 153 sne contained in both essence sets .
for some sne the difference is very large ( namely up to one magnitude ) . in order to see
if the two sets are consistent , we need to know how large the difference is in terms of the error @xmath68^{1/2}$ ] .
the result is shown in figure [ weightmudiff ] .
we want the systematical error due to the different fitting methods to be smaller than the statistical error of the observational data .
thus , we discard all sne with a difference in @xmath28 larger than @xmath69 and those that are only contained in one of the two essence sets , which leaves 129 sne in the sets .
again we counted the sne indicating acceleration or deceleration for each set and each calibration method .
the result is shown in table [ acctab2 ] .
the outcome of this test did not change qualitatively by discarding suspicious sne from the sets .
thus , we will use all the sne of the essence sets in the following .
@lllll & & + & riess & sandage & riess & sandage + acceleration ( 95% cl ) & 33 & 11 & 69 & 30 + acceleration ( 99% cl ) & 8 & & * 40 * & 14 + deceleration ( 95% cl ) & 1 & 2 & 0 & 6 + deceleration ( 99% cl ) & 0 & & & 1 + the error of the distance modulus contains the peculiar velocities @xmath70 of the sne . in the essence
sets @xmath71 is assumed to be 400km / s for all sne .
we wanted to know how sensitive our test is with respect to changes of the peculiar velocity .
table [ acctabvel ] shows that varying @xmath71 almost does not change the result with the exception of essence ( mlcs2k2 ) in the sandage calibration , where the number of sne indicating deceleration at a 95% cl is 7 for @xmath72km / s and 500km / s , but 11 for @xmath73km / s .
@llllll & & & + @xmath71[km / s ] & & riess & sandage & riess & sandage + 300 & acceleration ( 95% cl ) & 37 & 14 & 99 & 53 + & acceleration ( 99% cl ) & 11 & 3 & 64 & 30 + & deceleration ( 95% cl ) & 3 & 11 & 1 & 7 + & deceleration ( 99% cl ) & 2 & 3 & 0 & 2 + 400 & acceleration ( 95% cl ) & 37 & 14 & 96 & 53 + & acceleration ( 99% cl ) & 11 & 3 & 64 & 30 + & deceleration ( 95% cl ) & 3 & 7 & 1 & 7 + & deceleration ( 99% cl ) & 2 & 3 & 0 & 2 + 500 & acceleration ( 95% cl ) & 36 & 13 & 95 & 53 + & acceleration ( 99% cl ) & 10 & 3 & 63 & 30 + & deceleration ( 95% cl ) & 3 & 7 & 1 & 5 + & deceleration ( 99% cl ) & 2 & 3 & 0 & 2 + & 162 & 162 & 178 & 178 + until now , we have only made tests for single sne .
the problem in combining these data is that the quantity @xmath61 depends on the redshift @xmath11 .
thus the data from different redshifts do not have the same mean value .
nevertheless it is possible to average over @xmath28 .
then the result of course depends on how many sne from a certain redshift are used to calculate this mean value .
therefore , the average over all sne of a set does not characterize the function @xmath74 .
but still , when combined with its standard deviation , the mean value can be an evidence for acceleration or deceleration .
thus , we define @xmath75 where @xmath76 .
the mean value is calculated in such a way that data points with a small error are weighted more than those with large errors .
the standard deviation of the mean value is calculated by @xmath77^{\frac{1}{2 } } \;.\ ] ] we start with averaging @xmath28 over redshift bins of width 0.2 . for all data sets the averaged value of @xmath61 increases with redshift ( see figure [ zbinfig ] ) which could be expected for an accelerated expansion .
the curve for a flat @xmath59cdm with @xmath78 has its maximum at @xmath79 and then decreases again .
unfortunately , there are not enough sne at high redshifts to state a possible decrease of @xmath61 at some point .
the differences in the data points of the two essence sets again seem to be too large .
mlcs2k2 gives smaller values than the salt fitter for all redshift bins .
as @xmath74 should be close to zero at low redshifts in a homogeneous and isotropic universe , the mlcs2k2 data look more consistent with our assumptions in the riess calibration , whereas salt gives the better results in the sandage calibration .
the gold sample has sensible values in both calibrations .
@xmath80 devided by the error @xmath81 gives the evidence for acceleration in each redshift bin .
as can be seen in table [ zbin ] , the strongest evidence is given in redshifts between 0.4 and 0.8 .
@lllllll & & & + z & riess & sandage & riess & sandage & riess & sandage + flat universe & & & & & & + 0.0 0.2 & 2.1 & 0.4 & 1.6 & 2.8 & 5.7 & 0.8 + 0.2 0.4 & 4.2 & 2.9 & 3.9 & 0.7 & 7.3 & 4.3 + 0.4 0.6 & 9.8 & 7.6 & 7.1 & 3.1 & 12.1 & 7.7 + 0.6 0.8 & 5.4 & 4.0 & 7.7 & 3.7 & 8.5 & 4.8 + 0.8 1.0 & 4.8 & 3.5 & 6.8 & 4.1 & 6.0 & 4.2 + 1.0 1.2 & 2.8 & 2.2 & & & & + 1.2 1.4 & 2.3 & 1.8 & & & & + open universe & & & & & & + 0.0 0.2 & 2.1 & 0.4 & 1.5 & 2.8 & 5.7 & 0.7 + 0.2 0.4 & 3.5 & 2.2 & 3.2 & 0.0 & 6.6 & 3.6 + 0.4 0.6 & 7.4 & 5.2 & 5.5 & 1.4 & 10.2 & 5.8 + 0.6 0.8 & 2.8 & 1.4 & 4.9 & 0.9 & 5.9 & 2.2 + 0.8 1.0 & 1.4 & 0.2 & 3.9 & 1.2 & 4.0 & 2.2 + 1.0 1.2 & 0.9 & 0.4 & & & & + 1.2 1.4 & 0.2 & 0.7 + next we average over all sne of each data set with a redshift @xmath82 . we discard sne with a smaller redshift for the following reasons : ( a ) @xmath74 is expected to be relatively close to zero for small redshifts .
therefore , nearby sne do not contribute to the evidence for acceleration .
( b ) considering only the nearby universe , local effects can modify the results for the distance modulus as the cosmological principal is not valid on small scales .
( c ) another disadvantage of nearby sne is that they were observed with many different telescopes and thus the obtained systematic error is potentially higher .
table [ meantab2 ] shows the mean values @xmath80 and their standard deviations @xmath81 evaluated by using only sn data with @xmath82 .
@xmath80 is positive for all data sets .
@xmath80 divided by @xmath81 indicates the confidence level at which an accelerated expansion can be stated . the weakest evidence for acceleration is given for essence ( mlcs2k2 ) in the sandage calibration . here
the mean value lies @xmath83 above 0 , i.e. above the value for a universe that neither accelerates nor decelerates . in the other cases the confidence level is even larger , up to @xmath84 for essence ( salt ) in the riess calibration .
@lllllll & & & + & riess & sandage & riess & sandage & riess & sandage + @xmath85 & 0.63 & 0.63 & 0.54 & 0.54 & 0.51 & 0.51 + @xmath80 & 0.2196 & 0.1655 & 0.2398 & 0.1056 & 0.3457 & 0.2115
+ @xmath81 & 0.0167 & 0.0167 & 0.0201 & 0.0201 & 0.0203 & 0.0203 + @xmath86 & 13.1 & 9.9 & 11.9 & 5.2 & 17.0 & 10.4 +
although there are good reasons to believe that the universe is flat we give up this assumption in the following section . in an open universe the luminosity distance of a universe that neither accelerates nor decelerates is given by @xmath87 and thus depends on the density parameter of the scalar curvature @xmath88 .
@xmath25 increases with increasing @xmath88 and thus the evidence for acceleration becomes weaker .
as we are interested in the lower limit of this evidence we have to take the highest possible value for the scalar curvature , i.e. @xmath89 , which corresponds to an empty universe .
( here we allow ourself to make use of the einstein equation . )
then equation for @xmath28 changes to @xmath90 \right ] -25 \,.\ ] ] @lllllll & & & + & riess & sandage & riess & sandage & riess & sandage + open universe & 8.0 & 4.9 & 8.8 & 1.8 & 13.8 & 7.2 + flat universe & 13.1 & 9.9 & 11.9 & 5.2 & 17.0 & 10.4 + closed universe & 13.1 & 9.9 & 11.9 & 5.2 & 17.0 & 10.4 + the evidence for accelerated expansion is then calculated in the same way as for a flat universe .
the result obtained by averaging over redshift bins is shown in table [ zbin ] .
for the overall average we only used sne between redshift 0.2 and 1.2 because including higher redshifts would weaken the evidence .
this is due to the fact that the values of @xmath28 become negative when the phase of deceleration within the redshift range over which is integrated is large enough .
the result is shown in table [ opentab ] . as could be expected , the evidence is now much weaker than for a flat universe .
but we still find a hint of acceleration at 1.8@xmath0 .
for a closed universe we have a different situation : here @xmath25 decreases with increasing spatial curvature .
thus the lower limit of the evidence for acceleration is given in the case of the lowest possible curvature which corresponds to a flat universe .
we tested the cosmic expansion without specifying any density parameters or parameterizing kinematical quantities .
this was not possible without assuming a certain calibration of @xmath12 and @xmath13 .
therefore , we considered two very different calibrations .
we also considered two different data sets and two different light curve fitters and varied the peculiar velocity dispersion . using single sne for the test
has already given a clear indication of acceleration in a flat universe although a few sne strongly favour deceleration .
we find large systematic effects already at the level of individual sne .
similar analyses have already been done by other groups @xcite . but note that the work of santos et al . @xcite exhibits two major shortcomings .
first , they did not calibrate the sn data consistently .
they took a certain value for @xmath13 but kept the arbitrary value of the absolute magnitude @xmath12 given in the set which led to distance moduli that are too high .
thus , they concluded erroneously that there was a recent phase of super - acceleration .
the second problem is that they did not realize that due to the integration over redshift this method is not suitable to determine the transition redshift @xmath91 between deceleration and acceleration .
thus , their value of @xmath92 does not hold .
the shortcoming of the work of gong et al .
@xcite is that they have not applied any statistics at all . instead of only considering single sne
, we obtained a more significant result by averaging over the sn data of each set . here it is justified to discard nearby sne in order to decrease systematics from using different telescopes or from local effects .
we argue that we reject the null hypothesis of no acceleration for a spatially flat , homogeneous and isotropic universe at high confidence ( @xmath93 ) .
however , the results show large differences for each data set , calibration and light - curve fitter .
e.g. changing the calibration from sandage to riess calibration in a flat universe using essence(mlcs2k2 ) increases the evidence for acceleration from 5.2@xmath0 to 11.9@xmath0 .
for the two different light - curve fitters applied to the essence sne using the riess calibration , we get values of 11.9@xmath0 and 17.0@xmath0 . wood - vasey et al .
argue that the differences due to different light curve fitters are not important as they disappear when marginalization is applied @xcite .
this is true if the data are only used to determine the parameter values of a certain cosmological model as in this case no calibration of @xmath12 and @xmath13 is needed . but
nevertheless , the fitters should give the same result for a given absolute magnitude @xmath12 .
if this is not the case , a systematic error in the determination of the apparent magnitude @xmath26 is introduced by at least one of the fitters .
systematics due to different fitters have also been described in @xcite and those within the gold sample in @xcite .
the different results obtained by the two calibrations are very surprising .
our test only depends on the reduced distance modulus @xmath94 and not on the absolute magnitude or the hubble constant individually .
( note that this is the quantity which is marginalized when fitting cosmological parameters . )
starting from different calibrations @xmath95 and @xmath96 leads to different values of @xmath13 , determined by observation of sne , in such a way that @xmath97 should in principle be the same for both values of @xmath12 .
thus our test should not depend on the calibration .
however , the fact that the calibration changes our result significantly can be explained as follows : @xmath13 is determined by observing sne .
riess et al . and sandage et al .
use different sets of sne and different fitting methods for this determination .
the systematic errors due to different sets and fitters then of course also influence the result for the hubble constant leading to different values @xmath98 and @xmath99 . for a conservative conclusion , we need to take the set that gives the weakest evidence for acceleration , namely essence ( mlcs2k2 ) in the sandage calibration . using this set ,
accelerated expansion can be stated at 5@xmath0 if we assume a spatially flat or closed universe . in an open universe
the evidence is much weaker , namely 1.8@xmath0 .
these results only hold if the correct analysis of sne is somewhere in the range of the cases we considered here . but as we observe enormous systematic effects in our test , we can not be sure of this assumption .
thus , it is a major issue to better understand how sne have to be observed and analyzed .
remember that results of this work are only valid if the universe is homogeneous and isotropic .
the situation changes dramatically if we give up those assumptions .
there exist enormously large structures in the universe , an example is the sloan great wall with an extension of @xmath100400 mpc @xcite . besides , there are also big voids and superclusters on a 100 mpc scale .
this roughness of the universe could spoil the validity of equation ( 1 ) .
indeed it is possible to construct inhomogeneous models that can describe the observational data without the need of an average acceleration @xcite . some observational evidence for inhomogeneity and anisotropy in sn hubble diagrams has recently been presented by @xcite , probably due to large scale bulk motion and perhaps systematic effects . a non - trivial scale dependence of the hubble rate due to the so - called cosmological backreaction @xcite has been shown in @xcite .
this effect of averaging can also mimic curvature effects @xcite and thus strongly influence the reconstruction of the dark energy equation of state , as has been shown in @xcite .
thus we think a major task must be to establish the acceleration of the universe independently of the assumption of strict homogeneity and isotropy .
we thank dragan huterer , thanu padmanabhan , aleksandar raki , adam riess and bastian weinhorst for discussions , comments and references to the literature .
this work is supported by the dfg under grant grk 881 .
99 riess a _ et al . _ , _ observational evidence from supernovae for an accelerating universe and a cosmological constant _ , 1998 _ astron . j. _ * 116 * 1009 perlmutter s _
_ , _ measurements of @xmath101 and @xmath59 from 42 high - redshift supernovae _ , 1999
_ astrophys . j. _ * 517 * 565 jha s , riess a and kirshner r , _ improved distances to type ia supernovae with multicolor light curve shapes : mlcs2k2 _ , 2007 _ astrophys .
j. _ * 659 * 122 astier p _
et al . _ , _ the supernova legacy survey : measurement of @xmath102 , @xmath103 and @xmath104 from the first year data set _ , 2006 _ astron .
astrophys . _ * 447 * 31 riess a _
_ , _ new hubble space telescope discoveries of type ia supernovae at @xmath105 : narrowing constraints on the early behavior of dark energy _ , 2007 _ astrophys . j. _ * 659 * 98 wood - vasey w _
_ , _ observational constraints on the nature of the dark energy : first cosmological results from the essence supernova survey _ , 2007 turner m and riess a , _
do sne ia provide direct evidence for past deceleration of the universe ? _ , 2002 _ astrophys
. j. _ * 569 * 18 riess a _
_ , _ type ia supernova discoveries at @xmath106 from the hubble space telescope : evidence for past deceleration and constraints on dark energy evolution _ , 2004
_ astrophys .
j. _ * 607 * 665 elgary and multamki t , _ bayesian analysis of friedmannless cosmologies _
, 2006 _ j. cosmol .
_ jcap09(2006)002 wang y and tegmark m , _ uncorrelated measurements of the cosmic expansion history and dark energy from supernovae _ , 2005 _ phys .
d _ * 71 * 103513 john m , _ cosmography , decelerating past , and cosmological models : learning the bayesian way _
, 2005 _ astrophys .
j. _ * 630 * 667 shapiro c and turner m , _ what do we really know about cosmic acceleration ?
_ , 2006 _ astrophys
. j. _ * 649 * 563 rapetti d , allen s , amin m and blandford r , _ a kinematical approach to dark energy studies _
, 2007 _ mon . not .
. soc . _ * 375 * 1510 catton c and visser m , _ the hubble series : convergence properties and redshift variables _ , 2007 visser m , _ general relativistic energy conditions : the hubble expansion in the epoch of galaxy formation _
, 1997 _ phys .
d _ * 56 * 7578 santos j , alcaniz j , pires n and rebouas m , _ energy conditions and cosmic acceleration _ , 2007 _ phys .
d _ * 75 * 083523 gong y , wang a , wu q and zhang y - z , _ direct evidence of acceleration from distance modulus redshift graph _ , 2007 _ j. cosmol .
_ jcap08(2007)018 jackson n , _ the hubble constant _
, 2007 _ liv .
_ 10(2007)4 riess a _
_ , _ cepheid calibrations from the hubble space telescope of the luminosity of two recent type ia supernovae and a re - determination of the hubble constant _ , 2005 _ astrophys
. j. _ * 627 * 579 sandage a , tammann g , saha a , reindl b , macchetto f and panagia n , _ the hubble constant : a summary of the hst program for the luminosity calibration of type ia supernovae by means of cepheids _ , 2006 _ astrophys .
j. _ * 653 * 843 guy j , astier p , nobili s , regnault n and pain r , _ salt : a spectral adaptive light curve template for type ia supernovae _
, 2005 _ astron .
astrophys . _ * 443 * 781 conley a , carlberg r , guy j , howell d , jha s , riess a , sullivan m , _ is there evidence for a hubble bubble ?
the nature of type ia supernova colors and dust in external galaxies _ , 2007 _ astrophys . j. _ * 664 * l13 jassal h k , bagla j s and padmanabhan t , _ observational constraints on low redshift evolution of dark energy : how consistent are different observations ? _ , 2005 _ phys .
d _ * 72 * 103503 jassal h k , bagla j s and padmanabhan t , _ the vanishing phantom menace _
, 2006 nesseris s and perivolaropoulos l , _ tension and systematics in the gold06 snia dataset _ , 2007 _ j. cosmol .
_ jcap02(2007)025 gott j r i _ et al .
_ , _ a map of the universe _ , 2005 _ astrophys
. j. _ * 624 * 463 clrier m - n , _ do we really see a cosmological constant in the supernovae data ? _ , 2000 _ astron .
_ * 353 * 63 alnes h and amarzguioui m , _
the supernova hubble diagram for off - center observers in a spherically symmetric inhomogeneous universe _ , 2006
d _ * 75 * 023506 enqvist k and mattsson t , _ the effect of inhomogeneous expansion on the supernova observations _ , 2007 _ j. cosmol .
jcap02(2007)019 ishak m , richardson j , whittington d and garred d , _ dark energy or apparent acceleration due to a relativistic cosmological model more complex than flrw ? _ , 2007 mcclure m l and dyer c c , _ anisotropy in the hubble constant as observed in the hst extragalactic distance scale key project results _
, 2007 _ new astron . _ * 12 * 533 schwarz d j and weinhorst b , _
( an)isotropy of the hubble diagram : comparing hemispheres _ , 2007 _ astron .
astrophys . _ * 474 * 717 buchert t , _ on average properties of inhomogeneous fluids in general relativity .
i : dust cosmologies _ , 2000 _ gen .
_ * 32 * 105 schwarz d j , _ accelerated expansion without dark energy _ , 2002 _ on the nature of dark energy _
ed p brax , j martin and j - p uzan ( frontier group , paris ) rsnen s , _ dark energy from backreaction _ , 2004 _ j. cosmol .
_ jcap02(2004)003 kolb e w , matarrese s , notari a and riotto a , _ the effect of inhomogeneities on the expansion rate of the universe _ , 2005 _ phys .
d _ * 71 * 023524 vanderveld r a , flanagan e e and wasserman i , _ systematic corrections to the measured cosmological constant as a result of local inhomogeneity _ , 2007 _ phys .
d _ * 76 * 083504 leith b m , ng s c c and wiltshire d l , _ gravitational energy as dark energy : concordance of cosmological tests _ , 2007 li
n and schwarz d j , _ signatures of cosmological backreaction _ , 2007 li
n and schwarz d j , _ on the onset of cosmological backreaction _ , 2007 _ phys .
* 76 * 083011 bassett b a _ et al .
_ , _ is the dynamics of tracking dark energy detectable ? _ , 2007 | we test the present expansion of the universe using supernova type ia data without making any assumptions about the matter and energy content of the universe or about the parameterization of the deceleration parameter .
we assume the cosmological principle to apply in a strict sense .
the result strongly depends on the data set , the light - curve fitting method and the calibration of the absolute magnitude used for the test , indicating strong systematic errors .
nevertheless , in a spatially flat universe there is at least a 5@xmath0 evidence for acceleration which drops to 1.8@xmath0 in an open universe .
_ keywords _ : classical tests of cosmology , dark energy theory , supernova type ia |
in mathematical literature , there are important examples of models designed to describe the dynamical interaction between a set of players in a game - like context @xcite .
players are undistinguishable members of a large population , characterized by a phenotype which determines the fixed strategy they choose among the @xmath0 available , when playing with any other individual randomly selected in the population . + the payoff earned in the game by each player depends on the elements of a game - specific payoff matrix , while the system dynamics is described by an ordinary differential equation defined in the n - simplex , namely the replicator equation @xcite .
+ the solutions of the replicator equation may evolve towards evolutionary stable strategies , which are nash equilibria of the game and asymptotically stable stationary states @xcite .
evolutionary stable strategies are robust against invasion by competing strategies @xcite .
other kind of nash equilibria can also exist , such as equilibria corresponding to lyapunov stable stationary states of the replicator equation @xcite .
+ replicator equation has been used to describe several phenomena , such as biological evolution driven by replication and selection @xcite , and reaction - diffusion dynamics @xcite .
replicator dynamics including mutation have been studied in @xcite and may lead to more complex behaviors , characterized by hopf bifurcation and limit cycles @xcite . moreover ,
the replicator equation has been used for solving decision and consensus control problems @xcite and machine learning for optimization @xcite .
significant application have been also developed in the field of social science @xcite .
although widespread , the replicator equation is grounded on several strong assumptions on the system under investigation . 1 .
the population is very large . indeed , dominating strategies emerge due to higher rates of replication than the others , causing the frequency of less efficient strategies to become irrelevant in the total population .
any member of the population can play with any other member with the same probability .
participants of each game are chosen randomly and no social structures are present in the population .
the payoff earned by each player is defined by the payoff matrix , which is unique for all the population . in some cases , two or more subpopulations with different payoff matrices
are considered @xcite .
4 . players are constrained to behave according to a single fixed strategy at each round of the game they are playing .
for instance , if the phenotype of an individual is to be aggressive or generous , he will show the same level of aggressiveness or generosity in any situation at any time , without having the capacity of regulating the level of his natural impulses .
the above assumptions are the basis of the well known equivalence between the replicator equation and some ecological models , such as the predator - prey model introduced by lotka and volterra @xcite .
+ several modifications have been introduced in the replicator equation to overcome the limitations caused by the above assumptions .
for example , in @xcite is presented a generalization of the replicator equation to @xmath1 players . concerning networked populations ,
many efforts have been done to include the topology of connections between players in @xcite to deal with scenarios in which the connection between agents plays a fundamental role and may yield to the interpretation of inspected phenomena .
for example , in @xcite is proposed an algorithm that uses the connections among players and specific updating rules to induce cooperative behavior in an evolutionary prisoner dilemma game . on the other hand , a seminal paper by ohtsuki et al .
@xcite presented the replicator equation on infinite graphs , under the assumption that every vertex is connected to the same number of neighbours . in particular , he showed that the replicator equation on a graph with fixed degree is equivalent to a classical replicator equation when the degree goes to infinity .
+ in this paper , we derive a suitable mathematical model to describe static and dynamical behaviour of an @xmath1-players game interaction , where an agent is meant to engage his challenges only with a restricted set of players connected to him . to this aim , we make use of the standard non - cooperative game theory results and graph theory .
the results presented in the paper generalize many assumptions under the classical replicator equation and recent results on the replicator equation on graphs .
specifically , the equation derived in this paper meets the following : 1 . a finite ( even small ) or infinite population is considered .
the elements of the population are the vertices of a graph .
each element can be engaged in a game with another element only if they are connected in the graph .
no constraints on the topology of the graph are assumed .
moreover , the connections can be weighted to remark different perceived importance of each interaction .
3 . the payoff matrix can be player - specific , including the situations where the perception of the game is different for each individual .
4 . each player can behave according to a combination of strategies .
he is a sort of `` mixed player '' , thus incorporating composite and multiple personality traits .
his behavior can be driven contemporarily by heterogeneous impulses with different strengths , such as , for example , being cooperative and non cooperative , generous and selfish , at the same time .
the proposed approach makes players more realistic than in the classical framework and naturally extends the evolutionary game theory to a social context with human players .
the new framework presented in the paper generalizes the classical replicator equation , that can be obtained as a special case by assuming that any individual of the population possesses the same payoff matrix and starts playing from an identical initial condition .
+ the paper is structured as follows .
section [ sec : ncgames ] presents some preliminaries on noncooperative games on graphs .
then , the extended version of the replicator equation on graphs with generic topology is introduced in section [ sec : reg ] .
some properties of the new replicator equation are presented in section [ sec : mathreg ] , including the equivalence to the classical replicator equation when homogeneous initial conditions are used .
extended simulations are reported in section [ sec : sim ] , while some conclusions and future work are discussed in section [ sec : conclusions ] .
in real world situations , interactions between a finite number of rational players can be influenced by topological constraints ; in most cases , each player is only able to meet with a reduced number of opponents which are close with respect to a suitable topology , such as the distance between them . in this sense
, we talk about networks of players , where interconnections depend on the context .
an interesting case of interaction is represented by non - cooperative games , extensively studied in @xcite .
+ in this section , we extend the classical game theory by introducing a network , represented by a graph , which describes the connection among the involved players . typically , networks are described by means of graphs , and in a game context , each player is represented by a vertex .
an edge between two players indicates that they interact .
however , a player can consider that some interactions are more important than others .
moreover , two connected players can have different perception of the importance of their interaction .
these aspects can be accounted by assuming that the graph is weighted and directed ; an edge starting from a player and ending to another , is labeled with a positive weight to indicate the importance that the first player attributes to the game .
+ formally , let @xmath2 be a directed weighted graph of order @xmath3 , and let @xmath4 be the set of vertices ( players ) ( @xmath5 ) .
the graph @xmath2 is fully described by its adjacency matrix @xmath6 ; in particular , when player @xmath7 is meant to play with player @xmath8 , then there is an edge which starts from @xmath7 and ends to @xmath8 . in this case ,
@xmath9-entry of @xmath10 , @xmath11 , is the positive weight attributed by @xmath7 to the game against @xmath8 .
when @xmath12 and @xmath13 , there is an interaction between @xmath7 and @xmath8 , but only @xmath7 will get a payoff after the challenge . finally , if both @xmath14 and @xmath15 are equal to @xmath16 , then there is no interaction between these players . in general ,
we assume that @xmath2 has no self - edge , which means that no player has interaction with himself ( i.e. @xmath18 ) .
we indicate with @xmath19 the neighborhood of @xmath7 ( i.e. the set of vertices that interact with @xmath7 ) , and with @xmath20 the out - neighborhood of @xmath7 ( i.e. the set of vertices @xmath7 is connected to with and exiting edge ) .
the cardinalities of these set are indicated with @xmath21 and @xmath22 , and they represent the degree and out - degree of @xmath7 respectively .
note that , in general , @xmath23 , and also @xmath24 .
+ a player @xmath7 will play exactly @xmath21 two - players games with all its neighbours . in each one - to - one competition ,
the set of available strategies for both players is @xmath25 , while the outcome that player @xmath7 can obtain is defined by a payoff matrix @xmath26 ; when player @xmath7 uses strategy @xmath27 in a two - players game against a player which uses strategy @xmath28 , then he earns a payoff equal to the @xmath29-entry of @xmath30 , @xmath31 , where @xmath32 and @xmath33 are the @xmath34-th and @xmath35-th versors of @xmath36 , respectively .
+ each player decides to use the same strategy @xmath27 in all the games he is involved in .
he will play against all vertices in @xmath37 , but he will earn a payoff only when he plays with a player @xmath38 , since when @xmath39 and @xmath40 , there is an interaction which is meaningful only for player @xmath8 .
+ in an interconnected context , the effective payoff earned ( or the _ fitness _ of a strategy ) must be defined as an _ environmental measure _
depending on all the interactions between near players .
this measure must quantify how well a strategy behaves .
since each connection between two players has a positive weight , we pose that the effective payoff for a generic player @xmath7 is the weighted average of all obtained payoffs .
let s denote with @xmath41 the strategy of the generic player @xmath8 .
then , the effective payoff of player @xmath7 , @xmath42 is the following : @xmath43 where @xmath44 is the normalization factor .
this model of payoff based on weighted average will be denoted with wa . however , there are situations in which payoffs are cumulative and the weighted sum is used without the normalization factor @xmath45 .
in this case we have that : @xmath46 the payoff model based on weighted sum ( ws ) can be considered as wa , where each payoff matrix is substituted by @xmath47 .
for this reason , we will mainly work on wa model , unless differently specified .
+ the term @xmath48 that appears in both wa and ws models , is a vector where all components are non - negative numbers which sum up to @xmath49 . in a certain way ,
player @xmath7 fights against one _ virtual player _ which summarize all the strategies used by its opponents in the set @xmath50 ; in general , the strategy used by the virtual player is a mixed strategy which represents what player @xmath7 effectively _ sees _ around him .
this aspect will be deeply investigated later in this paper , because it plays a fundamental role to reach our aim .
+ notice that , for each @xmath7 , @xmath52 can be interpreted as a @xmath1-dimensional tensor , where the @xmath53-entry is @xmath42 . in this way , the game interaction between interconnected players on a finite graph is equivalent to a @xmath1-players game , where the set of pure strategies is @xmath54 , and the payoff of player @xmath7 is represented by the tensor @xmath52 . the structure of the graph is embedded in this definition , since the payoff tensor depends on the adjacency matrix @xmath10 . moreover , there are no assumptions made on the structure of the graph itself .
+ for example , consider the following matrices : @xmath55 , ~~\boldsymbol{b } = \left [ { } \begin{array}{cc } a & b \\ c & d \end{array } \right],\ ] ] where @xmath56 , and assume that @xmath57 for all @xmath7 . in this case ,
@xmath58 and @xmath59 .
table [ table : payoffs ] shows the payoff tensors @xmath52 of each player , which depend on the model parameters @xmath60 and @xmath61 .
both models wa and ws are considered .
+ .payoff tensor of the game on graphs defined by matrices in equation for wa and ws payoff models .
for each combination of strategies ( @xmath62 ) the payoffs @xmath63 and @xmath64 of players @xmath49 , @xmath65 , @xmath66 are reported . [ cols="^,^,^,^,^,^,^,^,^ " , ] it is evident that the presence of weights , the asymmetry of the matrix , and the use of a particular payoff model may lead to very different calculation of the payoff tensor , and hence , the structure of the game itself changes .
indeed , the effective payoff obtained by a player when he is engaged in a game is essentially evaluated by means of tensors , depending on the adjacency matrix of the graph .
these payoffs define the _ virtual player _ mentioned at the end of section [ sec : payoff_graphs ] , which embodies all the strategies used by the player s opponents . as a consequence ,
each player in the game is a sort of `` mixed player '' , thus incorporating composite and multiple personality traits , and behaving according to heterogeneous impulses with different strengths .
+ as a natural consequence , a @xmath1-players @xmath0-strategies game ( from now on , @xmath67-game ) can be extended over the set of mixed strategies @xmath68 : @xmath69^{t } \in \mathbb{r}^{m } : \sum_{i=1}^{m } z_{i } = 1 \wedge z_{i } \geq 0 ~~ \forall i \in \mathcal{s } \}.\ ] ] we indicate with @xmath70^{t } \in \delta_{m}$ ] the mixed strategy of player @xmath7 . recall that @xmath71 is the probability that player @xmath7 uses strategy @xmath34 , while he takes part in the games .
the formula of the expected effective payoff that player @xmath7 obtains , is similar to equation : @xmath72 where @xmath73 represents the expected outcome for player @xmath7 of the one - to - one game played by @xmath7 itself against @xmath8 .
from now on , we pose that @xmath74 , where @xmath75 indicates the group of all the vectors @xmath76 , with @xmath77 .
+ when a pure strategy @xmath34 is used by player @xmath7 , vector @xmath78 takes the form of the standard versor @xmath32 of @xmath36 .
if each @xmath76 is a versor , say @xmath79 , then equations and coincide .
furthermore , we can define the expected payoff @xmath80 of player @xmath7 when he is preprogrammed to use the strategy @xmath34 ( hence @xmath81 ) in all games played against its neighbours : @xmath82 equation easily leads to a more convenient definition of the expected payoff obtained by the player @xmath7 .
that is : @xmath83 the present work uses the same theoretical issues developed in the classical non - cooperative @xmath51-games theory ( @xcite ) .
the differences here introduced , consist with the possibility of embedding any topological structure in the game ; indeed , the payoff tensor used to describe the game depends on the adjacency matrix of the graph .
thanks to game theory , we are able to predict the strategies of opponents , assuming that all of them behave in a rational way during their decision - making tasks .
in fact , rational players choose pure strategies which may lead to a pure nash equilibrium , whenever it exists .
recall that nash theorem @xcite asserts that a game has always at least one nash equilibrium within the set of mixed strategies .
if the game is repeated over time , we can imagine that a mixed strategy describes , for each time , the probability that a player uses a certain pure strategy independently from the choices made until that moment .
payoff is computed as the average outcome that a player obtains when the game is reiterated for an infinite number of times , and nash equilibria are evaluated accordingly . although players behavior is randomized over time , there is a precise rational scheme that they follow .
hence , pure and mixed strategies games are quite similar , since in both cases players decide their behaviors through a rational decision - making task at the beginning of the game .
+ often , in real world situations , players do not have a full knowledge about the game , and the decision - making process suggested by game theory is not applicable .
however , players can learn from the context ; time after time , they are able to compare their payoffs with the outcomes of their opponents , and strategies are changed accordingly . for example , after some game iterations , one can understand that its opponent is preprogrammed to play always the same strategy @xmath35 , and then he decides to adopt the strategy @xmath34 which is the best reply to @xmath35 ( i.e. he obtains the maximum payoff knowing the strategy of his opponent ) .
also , a player can simply imitate the strongest opponent to reach a greater payoff .
+ evolutionary biology gives an interesting interpretation for mixed strategies .
let s consider a large population in which each individual is programmed to play a particular pure strategy .
population is divided according to a mixed strategy , which indicates the frequency of pure strategies in the population .
pairs of individuals are randomly drawn from the population to play games .
the average payoff obtained by all players playing a certain strategy is a measurement of the _ fitness _ of that strategy .
+ nature promotes fittest strategies ; time after time , when a strategy has a payoff greater than average , its frequency must increase , and consequently , frequencies relative to poorly fit strategies decrease .
this _ natural selection _
process may yield to interesting dynamical phenomena , since the fitness of each strategy changes over time according to the frequencies of the subpopulations .
natural selection is realized through reproduction .
generally , this happens asexually and the offspring produced is identical to parent .
for these reasons , a player in this context is also known as _ replicator_.
he is preprogrammed to use a certain strategy and there is no rational decision - making process .
a strategy is something included in the genes of a replicator , and hence , exhibited from birth .
a replicator uses a `` good '' strategy when has a fitness greater than the average .
nature favors players with a good strategy , by allowing them to survive and to produce offspring . +
the dynamics of replicator populations can be used to describe several non - biological situations : scientific ideas , life - styles , political orientations diffuse by means of imitation and education process that easily replace the concept of asexual reproduction .
in fact , this becomes clearer if we interpret fitness as a measurement quantifying how well a strategy behaves in a certain context , whatever it is . in this section
we develop a key idea which leads to the definition of a replicator equation based on a generic graph , where the number of players and the network structure are arbitrary . +
* main idea * - we imagine that each vertex of the network contains an infinite population of individuals .
we will refer to such elements as * atomic players * , and to vertices of the network as * vertex players*. the first are replicators , while the latter are the players introduced in paragraph [ sec : nmgames ] for @xmath51-games .
basically , each atomic player behaves like the corresponding vertex player : an atomic player of @xmath7 takes part to @xmath65-players games , described by payoff matrix @xmath84 , against exactly one atomic player randomly drawn from each connected vertex , and his effective payoff is the average of the payoffs obtained in all the one - to - one competitions . by the way , atomic players are different from vertex players .
indeed , all atomic players inside a vertex are indistinguishable , except for the fact that each of them is preprogrammed to use a certain pure strategy in @xmath54 during all the games he is involved in . on the contrary , vertex players can also adopt mixed strategies . atomic players reproduce themselves by replication , after their participation to games , inside their population .
furthermore , their capacity to produce offspring is related to the effective payoff obtained .
+ assume that one atomic player is randomly draw from each populations , and let @xmath85 be the strategy used by the one extracted from population @xmath8 .
then , the effective payoff earned by the atomic player of population @xmath7 is defined by equation .
now , @xmath71 can be interpreted as the share of atomic players preprogrammed to use the pure strategy @xmath34 inside the vertex @xmath7 .
this implies that @xmath80 is the expected effective payoff obtained by an atomic player of @xmath7 when he uses strategy @xmath34 , while @xmath86 represents the expected effective payoff for a generic atomic player randomly drawn from population @xmath7 ( see equations and ) .
a mixed strategy of a vertex player can be interpreted as the way in which its internal population of atomic players is distributed , according to the pure strategy they are preprogrammed to play .
suppose now that the games are iterated in time . we will refer to game session as the whole set of @xmath65-players games performed on the graph .
the probability for a replicator to survive and to reproduce himself between two games sessions , depends on the comparison between the effective payoff obtained and the average effective payoff of all other players .
let s assume that games sessions take place at discrete and equidistant times ( say , a session after each @xmath87 seconds ) .
let @xmath88 indicate the share of population inside vertex @xmath7 , which is preprogrammed to use the pure strategy @xmath34 , at time @xmath89 .
+ what happens to @xmath90 ?
first of all , suppose for a while that the population size in the generic node @xmath7 at time @xmath89 is @xmath91 . according to @xcite
, we can consider that @xmath92 represents a reproductive rate and therefore , @xmath93 is the number of offspring produced by one atomic player in @xmath7 that uses strategy @xmath34 between @xmath89 and @xmath94 .
hence , the population size after a time @xmath87 is equal to the previous size plus the produced offspring , since each atomic player reproduces himself within his population .
that is : @xmath95 where @xmath96 is the size of the subpopulation which uses strategy @xmath35 at time @xmath89 , and @xmath97 is the number of offspring produced by this subpopulation . by definition , @xmath90 is the ratio between the size of subpopulation @xmath34 and the total population . therefore : @xmath98 notice that equation does not depend on @xmath91 , and hence this relationship is valid for any starting size of the population . +
our aim is to develop a mathematical model that describes the evolutionary process on a graph when the time between replication events goes to @xmath16 , thus making atomic players able to reproduce themselves continuously in time .
let s consider the difference ratio of @xmath88 : @xmath99 letting @xmath100 , we obtain that : @xmath101 where the `` dot '' indicates the derivative with respect to time @xmath89 .
finally , we can write the following cauchy problem : @xmath102 where , for consistency , it is assumed that the distribution of strategies at the initial time @xmath103 is known for each vertex ( i.e. @xmath104^{t } \in \delta_{m}$ ] ) .
+ systems represents the replicator equation on a graph .
note that no assumptions on the structure of the graph is needed to derive the equation .
indeed , the adjacency matrix of the network is fully embedded in the payoff tensors .
+ it is straightforward to note that the equation has a structure similar to the classical replicator equation ; for example , dominant strategies are the fittest , and hence when the relative fitness @xmath80 is better than the average @xmath86 , the corresponding frequencies will grow over time . in the next section , the very strong correlation between the two equations will be rigorously shown .
furthermore , the relationship between nash equilibria of the underlying @xmath51-game and the rest points of the dynamical equation will also be discussed in section [ sec : reproperties ] .
let @xmath105 be the unique solution of problem , obtained by posing @xmath106 .
in addition , suppose that there exists a time instant @xmath107 where @xmath108 .
since all the components of the solution are continuous and non - negative at @xmath103 , then there must be a time @xmath109 such that @xmath110 .
following equation , we can state that @xmath111 , and hence , this component will be @xmath16 for all times after @xmath112 . for the unicity of the solution
, this implies that no time @xmath107 for which @xmath108 exists .
thus , for each @xmath113 we have that : @xmath114 for all strategies @xmath27 and for all times @xmath115 .
notice that the total variation of the strategies distribution in a vertex is null at time @xmath89 when @xmath116 .
in fact : @xmath117 this means that : @xmath118 imposing that @xmath106 , the last equation asserts that @xmath116 for all time @xmath115 .
joining the results provided by and , we conclude the following : @xmath119 in other words , all trajectories that start inside @xmath68 remain inside @xmath68 itself for all time @xmath115 . at any time
, @xmath105 can be always interpreted as a distribution of strategies .
recall that the best response function for the static @xmath51-game is : @xmath120 suppose that @xmath121 is a nash equilibrium .
then : @xmath122 for each vertex @xmath7 .
this means that : @xmath123 moreover , from we know that : @xmath124 and then : @xmath125 we can conclude that every nash equilibrium is also a rest point of the replicator equation on graph .
suppose that @xmath126 .
then @xmath127 and @xmath128 in addition , @xmath129 if @xmath130 , and again : @xmath131 for this reason : @xmath132 this implies that if each @xmath105 represents a pure strategy ( i.e. it is equal to a versor of @xmath36 ) , then we have a rest point of the replicator equation on graph .
suppose to fix a time lag @xmath133 , and assume that the mixed strategies are all the same for each vertex and for any time @xmath134 .
that is : @xmath135^{t } \in \delta_{m } ~~ \forall v \in \mathcal{v } , ~~\forall t_{0 } \in [ 0 , \tau).\ ] ] consider the payoff model wa and suppose that @xmath57 for all vertices @xmath7 .
following equations and , we obtain that @xmath136 and @xmath137 . in this case , we can rewrite the difference equation as follows : @xmath138 since previous equations do not depend on @xmath7 , we are able to impose that @xmath139 and @xmath140 , @xmath141 , and hence : @xmath142 it s straightforward to note that any other iteration of the previous map leads to quantities that are independent from @xmath7 .
for example , applying a second iteration we get that : @xmath143 and hence we can pose that @xmath144 . generalizing to any time lag , @xmath145 for any non negative integer @xmath146 .
similarly , @xmath147 and @xmath148 are also independent from @xmath7 . for these reasons , we pose that @xmath149 and @xmath150
then , the discrete map becomes the following : @xmath151 note that , for any @xmath152 there exist a non - negative integer @xmath146 and a real number @xmath134 , with @xmath133 fixed , such that @xmath153 .
then , equation becomes : @xmath154 considering the difference ratio @xmath155 , and letting @xmath100 , we obtain the following differential equation : @xmath156 which is the classical replicator equation . +
this result is quite straightforward if we imagine to divide a wide population of replicators into @xmath1 subpopulations , assuming that all of them are described by the same mixed strategy of the total one at initial time .
then , each subpopulation will behave exactly as the total one .
hence , the dynamics of a single subpopulation in a vertex can be described by the classical replicator equation applied to the single population , whatever is the graph used .
in this chapter , we present some simulations produced by equation . the @xmath157 payoff model is used . in particular , we set up experimental sessions by considering different @xmath65-strategies payoff matrices ( @xmath158 ) ; it is assumed that every vertex has the same payoff matrix .
each session has been developed over @xmath66 different graphs with @xmath159 vertices as reported in figure [ fig : graphs ] .
all edges represented in figure [ fig : graphs ] have the same weight , except for thicker ones in the asymmetric weighted graph .
note that we are using only undirected graphs ( i.e. @xmath160 ) .
our aim is to show the behavior of the replicator equation on graphs when initial players strategies are almost pure .
in fact , a vertex player with a pure strategy is in steady state ; for this reason , initial conditions used for vertex players are equals to slightly perturbed pure strategies ( i.e. @xmath161^{t}$ ] and @xmath162^{t}$ ] are used in place of pure strategies @xmath49 and @xmath65 , respectively ) .
replicator equation on graphs has been simulated until a steady state behavior is reached , starting from @xmath163 different distribution initial conditions on the graph .
+ the steady state situations are shown in figures [ fig : sym_bistable ] , [ fig : asym_bistable ] , [ fig : prisoner ] and [ fig : coexistence ] .
the first column of each figure gives a picture of the initial conditions used , while others report the solution of the simulations when steady state is reached for each of the considered graphs .
the color of each vertex indicates the value of @xmath164 , and hence it visually quantifies the inclination of player @xmath7 toward one of the @xmath65 feasible pure strategies ; yellow is used for player with strategy @xmath49 ( @xmath165 ) , red is for strategy @xmath65 ( @xmath166 ) .
mixed strategies ( @xmath167 ) are indicated by shaded colors , according to the color bar at the bottom of the figures .
moreover , figures [ fig : timecourse_pd ] and [ fig : timecourse_coex ] report the dynamical evolution obtained on the asymmetric weighted graph ; the same initial condition is used in both figures , while payoff matrices are different .
the following sections will discuss in detail the results of each simulation . in this first experimental session , we used the following payoff matrix : @xmath168,\ ] ] with @xmath169 .
the @xmath65-players game described by @xmath170 has @xmath65 strict pure nash equilibria ( i.e. both players use strategy @xmath49 or @xmath65 ) and a mixed nash equilibrium @xmath171^{t}.\ ] ] the classical replicator equation , based on matrix @xmath170 , has exactly @xmath66 rest points which coincide with the nash equilibria reported above .
moreover , mixed equilibrium is repulsive , while pure equilibria are attractive ; for this reason , we say that @xmath172 is a bistable payoff matrix .
+ figure [ fig : sym_bistable ] reports some results obtained when @xmath173 .
row ( a ) of the figure shows what happens when an homogeneous initial condition is used ; as said in section [ sec : classical ] , the dynamics is the same for each vertex player , and it is equivalent to the solution given by the classical replicator equation , whichever is the underlying graph structure . after a certain time
, all vertex players adopt pure strategy @xmath49 , since it represents an attractive rest point , and initial condition is in inside the relative basin of attraction . + in the row ( b ) of figure
[ fig : sym_bistable ] are reported the steady state situations obtained by using an homogeneous initial condition , where only one peripheral player uses the quasi - pure strategy @xmath65 . at the end of simulation ,
the pure strategy @xmath49 spreads all over the considered graphs .
let s consider the open graph situation : the vertex player @xmath49 , which is the unique neighbors of player @xmath65 , has no will to change his own strategy , since he is surrounded by @xmath163 yellow players .
similarly , on the closed and asymmetric weighted star , neighbors of player @xmath65 see an equivalent player which is almost yellow .
thus , none of them wants to change , and player @xmath65 must modify his strategy .
hence player @xmath65 must change his strategy to obtain a good payoff . in a certain way ,
the rebel peripheral player decides to adapt himself to the majority .
+ the dynamical behavior is slightly different when the central hub is the rebel . in the row ( c ) of the figure [ fig : sym_bistable ]
are shown the solutions of the replicator equation on graphs for this initial condition .
when the open star is used , player @xmath49 sees a yellow equivalent player , while all peripheral players have only him as neighbor .
player @xmath49 decides to change his own strategy to yellow , while all others do the exact opposite . after a certain time
, they meet half way , at the mixed equilibrium @xmath174^{t}$ ] .
the different position of the rebel player in the graph influences a lot the dynamics of the whole system ; the leader ( player @xmath49 ) understands that he must modify his own strategy according to his neighborhood , while all other players do the same , since their only opponent is player @xmath49 himself .
however , closed and asymmetric weighted graphs are more resistant to the influence of player @xmath49 , because the peripheral players have more than one neighbors ; in these situations , player @xmath49 does not play anymore as a leader able to change the whole dynamics .
+ the last row ( d ) of figure [ fig : sym_bistable ] reports the final solutions when both player @xmath49 and @xmath65 use the quasi - pure strategy @xmath65 .
while the closed star structure remains resistant to the influence of rebel players , the other graphs do not .
the open star becomes all red at final time .
this is because player @xmath65 sees only player @xmath49 : they are both red , so player @xmath65 does nt want to change strategy .
simultaneously , yellow neighbors of @xmath49 change their strategy to red , since they see only a red player . + changing the value of the parameter @xmath175 leads to different behaviors .
when @xmath176 , first strategy becomes stronger and it spreads all over the considered graphs as @xmath175 goes to @xmath16 . in figure
[ fig : asym_bistable ] are reported some results obtained with @xmath177 .
in particular , when player @xmath49 uses strategy @xmath65 at the beginning , then mixed equilibrium is not reached anymore on the open star graph ; all vertices adopt strategy @xmath65 , which is slightly better than strategy @xmath49 .
the strength of strategy @xmath65 is also visible on the asymmetric weighted star , when at the beginning both player @xmath49 and @xmath65 adopt strategy @xmath65 ; in figure [ fig : sym_bistable ] ( @xmath178 ) we have shown that on steady state , players @xmath65 and @xmath159 are the only ones red , while when @xmath179 , also players @xmath49 and @xmath180 do . in general , when @xmath181 , strategy @xmath65 becomes stronger and it spreads all over the considered graphs as @xmath175 grows up . in this section ,
we show the results obtained with the replicator equation on graphs , by using a modified version of the prisoners dilemma game , as proposed in @xcite .
the payoff matrix is the following : @xmath182,\ ] ] where @xmath181 .
+ _ cooperate _ and _ defect _ are the names typically used to indicate , respectively , the strategy @xmath49 and @xmath65 of this classic game .
the dilemma is that mutual cooperation produces a better outcome than mutual defection ; however , at the individual level , the choice to cooperate is not rational from a selfish point of view . in other words ,
the @xmath65-players game has only one nash equilibrium , reached when both players defect .
note that in this version of the prisoner s dilemma , the nash equilibrium is non strict .
moreover , classical replicator equation based upon payoff matrix reported in equation , has @xmath65 rest points , which correspond to the pure strategies @xmath183 and @xmath184 . in particular ,
the first one is repulsive , while the latter is attractive .
+ although mutual defection represents both a nash and a dynamical equilibrium , many works have shown that cooperation does not vanishes when games are played over graphs and @xmath175 is equal to suitable values ( see @xcite ) .
the resilience of cooperation is shown in [ fig : prisoner ] , where @xmath175 is set to @xmath185 .
steady states depend on the initial conditions and on the type of graph used , and behaviors can be very heterogeneous .
when an homogeneous initial condition is considered ( row ( a ) ) , all players on graphs become defectors ( again , this is the case when the classical and proposed replicator equations are the same ) . on the other side , when initial conditions are not homogeneous ( rows ( b ) , ( c ) and ( d ) ) , cooperation does not always completely vanish .
figure [ fig : timecourse_pd ] shows the time course of the variable @xmath164 for each vertex of the graph . in particular , the initial conditions with external outlayer and the asymmetric weighted graph have been used . in some @xmath65-players games
there are no pure nash equilibria .
nevertheless , nash theorem guarantees that at least a mixed equilibrium exists .
for example , this happens when payoff matrix is defined as follows : @xmath186.\ ] ] the unique mixed nash equilibrium is : @xmath187^{t}.\ ] ] classical replicator equation has @xmath66 rest points ; symmetric couples of pure strategies , @xmath183 and @xmath184 are repulsive , while the mixed equilibrium is attractive . in this case
, we speak about @xmath188 of both feasible strategies . + in figure [ fig : coexistence ] the steady state solutions when payoff matrix defined in equation is used , are reported .
again , when we have an homogeneous initial condition , everything works like a classical replicator equation , and hence , all players go to the mixed nash equilibrium .
when initial condition is not homogeneous , behaviors obtained through the replicator equation on graphs are strongly based on the topological structure of the underlying graph and on the initial conditions .
figure [ fig : timecourse_coex ] shows in details the behavior of the population when the asymmetric weighted graph and initial conditions with external outlayer are supposed .
in this work a new mathematical model for evolutionary games on graphs with generic topology has been developed .
we proposed a replicator equation on graphs , dealing with a finite population of players connected through an arbitrary topology .
a link between two players can be weighted by a positive real number to indicate the strength of the connection .
furthermore , the different perception that each player has about the game is modeled by allowing the presence of directed links and different payoff matrices for each member of the population .
a player obtains his outcome after @xmath65-players games are played with his neighbors ; payoffs of each game are averaged ( wa model ) or simply summed up ( ws model ) .
moreover , it has been shown that the proposed replicator equation on graphs extends the classical one , under the hypotheses that wa model for payoffs is used , homogeneous initial conditions over the vertices are considered , all vertex players have the same payoff matrix . in any case , no limitations are imposed to the underlying graph .
+ experimental results showed that the dynamics of evolutionary games are strongly influenced by the network topology . as expected , more complex behavior emerges with respect to the classical replicator equation .
for example , in the prisoner s dilemma game , cooperative and non - cooperative behaviors can coexist over the graph .
moreover , when a @xmath65-player game with strictly dominant strategies is considered , heterogeneous behavior is obtained , i.e. a part of the population chooses to play a dominant strategy , while others use different strategies .
then , players become mixed ( coexistence of strategies ) .
+ the very first step for extending this work is the study of dynamical and evolutionary stability of the rest points . by the way , we imagine that the concept of evolutionary stability must be revisited to deal with the proposed evolutionary multi - players game model based on graph , for which a theoretical effort is needed . indeed , in our opinion , the basic question `` is strategy @xmath34 resistant to invasion ? ''
must be reformulated to fit with the new model , where the population of players is finite and is organized according to a social structure .
+ the theory developed in this paper can also be extended to 3 or more strategies and can consider more complex topologies of the graph , such as small world , scale free , and random complex networks .
+ from an applicative point of view , the authors intend to use the replicator equation on graphs to deal with biological and physical processes , such as bacterial growth @xcite , model of brain dynamics @xcite and reaction - diffusion phenomena @xcite .
the developed model can be also profitably applied to solve networked socio - economics problems , such as decision making for the development of marketing strategies . | a new mathematical model for evolutionary games on graphs is proposed to extend the classical replicator equation to finite populations of players organized on a network with generic topology .
classical results from game theory , evolutionary game theory and graph theory are used .
more specifically , each player is placed in a vertex of the graph and he is seen as an infinite population of replicators which replicate within the vertex . at each time
instant , a game is played by two replicators belonging to different connected vertices , and the outcome of the game influences their ability of producing offspring .
then , the behavior of a vertex player is determined by the distribution of strategies used by the internal replicators . under suitable hypotheses ,
the proposed model is equivalent to the classical replicator equation .
extended simulations are performed to show the dynamical behavior of the solutions and the potentialities of the developed model .
* keywords : * replicator equation on graph , evolutionary game theory , finite populations , complex networks . |
doped eufe@xmath1as@xmath1 superconductors stand out among fe - based superconductors ( fescs ) , because the eu@xmath3 local spins ( @xmath4=7/2 ) may order _ ferromagnetically _ in the superconducting state@xcite .
the undoped eufe@xmath1as@xmath1 undergoes an @xmath5-type antiferromagnetic ( namely , in - plane ferromagnetic while inter - plane antiferromagnetic ) transition at 19 k in the eu sublattice@xcite , in addition to a spin - density wave ( sdw ) transition at 190 k in the fe sublattice . in the magnetically ordered state
, the eu spins align along the crystallographic [ 110 ] direction , collinear with the fe moments@xcite . upon partial p - for - as substitution , which effectively induces a chemical pressure ,
the fe - site sdw order is suppressed , and then superconductivity ( sc ) emerges at @xmath6 26 k@xcite .
simultaneously , the eu spins become ferromagnetically ordered at @xmath7 20 k , accompanying with a spin reorientation towards the [ 001 ] ( or @xmath8-axis ) direction@xcite .
similar coexistence of fe - based sc and eu - spin ferromagnetism ( fm ) was also observed and verified in several fe - site electron - doped eufe@xmath1as@xmath1 systems@xcite .
nevertheless , there have been some debates on the details of the eu - spin ordering . in refs .
@xcite , it was argued that sc coexists with eu - spin antiferromagnetism ( afm ) in the superconducting regime of eufe@xmath1(as@xmath9p@xmath10)@xmath1 .
later , zapf et al.@xcite proposed that sc coexists with a canted afm , such that it is virtually ferromagnetic along the @xmath8 axis .
these authors@xcite further revised the electronic phase diagram because of the discovery of a reentrant spin glass state .
recent x - ray resonant magnetic scattering@xcite and neutron scattering@xcite experiments however indicated long - range ferromagnetic orderings for eu spins in superconducting eufe@xmath1(as@xmath9p@xmath10)@xmath1 with @xmath11 = 0.19 and 0.15 , respectively .
it was demonstrated that the eu spins align exactly along the @xmath8 axis , in contradiction to the spin - canting scenario .
so far , this discrepancy remains unresolved .
note that the spin - tilting angle ( @xmath220@xmath12 from the @xmath8 axis , as detected by mssbauer measurements@xcite ) coincides with the direction that connects the interlayer next - nearest ( nn ) eu atoms because of the body - centered eu sublattice . to clarify whether the eu - sublattice type is relevant to eu spin orientations ,
it is desirable to study a related material system in which eu atoms form a primitive tetragonal lattice .
local - moment fm and spin - singlet sc are known to be mutually incompatible@xcite , which makes their coexistence ( hereafter abbreviated as fm+sc ) very rare@xcite .
the fm+sc phenomenon observed in fescs has been ascribed to the multi - orbital character as well as the robustness of superconductivity against magnetic fields@xcite .
on the one hand , the zero - temperature upper critical magnetic field , @xmath13 , of fescs is typically higher than 50 t@xcite , which is large enough to fight the internal exchange field that is comparable to the hyperfine field on the eu nucleus ( @xmath2 25 t)@xcite . on the other hand
, the eu - spin fm can be satisfied even in the presence of sc , because the fe-3@xmath14 multi - orbitals enable both superconducting pairing ( dominated by the @xmath15 and @xmath16 electrons@xcite ) and the ruderman - kittel - kasuya - yosida ( rkky ) exchange interaction between eu local moments .
the rkky interaction can be mediated by different fe-3@xmath14 orbitals such as @xmath17 and @xmath18@xcite .
therefore , both sc and local - moment fm can be favored in fescs@xcite .
the crucial factor that leads to eu local - moment fm should be the interlayer rkky interaction , since the in - plane rkky coupling remains ferromagnetic even in the parent compound eufe@xmath1as@xmath1 .
the interlayer exchange coupling ( @xmath19 ) is simplified to be proportional to @xmath20 for a large @xmath21 , where @xmath22 is the fermi wave vector and @xmath21 denotes the distance between local moments .
this means that @xmath19 can be changed from negative ( afm ) to positive ( fm ) by tuning the @xmath23 value .
the afm - to - fm transition in eufe@xmath1(as@xmath9p@xmath10)@xmath1@xcite and eu(fe@xmath9ni@xmath10)@xmath1as@xmath1@xcite is mainly due to the change in @xmath22 ( where the heavy three - dimensional hole pocket@xcite seems to be involved ) , simply because the interlayer eu interatomic distance ( @xmath24 ) varies only slightly .
indeed , based on a first - principles calculation@xcite , the effective interlayer nn magnetic coupling changes from antiferromagnetic in eufe@xmath1as@xmath1 to ferromagnetic in eufe@xmath1p@xmath1 , while the in - plane coupling remains ferromagnetic .
the above statement suggests an alternative approach to realize the afm - to - fm transition which may lead to an fm+sc state as well . by constructing a crystal structure in which @xmath24 changes significantly
, the sign of @xmath19 may be altered accordingly .
we previously designed a related structure , exemplified as klafe@xmath0as@xmath0 ( 1144)@xcite , which can be viewed as an intergrowth of kfe@xmath1as@xmath1 and lafe@xmath1as@xmath1 . in the eu analog , @xmath25eufe@xmath0as@xmath0 ( @xmath25 denotes an alkali metal ) , the eu atoms form a _ primitive _ tetragonal lattice , such that the lines that connect the interlayer - nn eu atoms are exactly parallel to the @xmath8 axis .
notably , the @xmath24 value is roughly doubled because every alternate eu atoms are replaced by @xmath25 along the @xmath8 axis .
very recently , we became aware of the report of @xmath26fe@xmath0as@xmath0 ( @xmath27=ca , sr , ba ) superconductors which possess the identical 1144-type structure@xcite .
this work inspires us to re - investigate our target material @xmath25eufe@xmath0as@xmath0 .
consequently , we succeeded in synthesizing a 1144 compound rbeufe@xmath0as@xmath0 . in this paper , we report the crystal structure and physical properties of this new material .
we indeed observe an eu - spin fm at 15 k which coexists with a bulk sc at an unexpectedly high @xmath28 of 36.5 k. remarkably , the evidence for fm+sc are very robust , compared to the previous fm+sc phenomena in doped eufe@xmath1as@xmath1 systems@xcite . apart from fm+sc , additional intriguing phenomena were also observed and discussed .
rbeufe@xmath0as@xmath0 polycrystalline samples were prepared via a solid - state reaction method .
first , feas , euas and rbas were prepared using source materials rb ( 99.75% ) , as ( 99.999% ) and eu ( 99.9% ) pieces and fe powder ( 99.998% ) .
the presynthesized arsenides were then ball milled separately for 10 minutes in a glove box filled with pure ar ( the water and oxygen content is below 1 ppm ) .
second , stoichiometric mixtures ( namely , rb : eu : fe : as=1:1:4:4 ) of fe , feas , euas and rbas were homogenized , pressed into pellets , and then loaded in an alumina tube - like crucible which was sealed in a ta tube .
the ta tube was protected in a quartz ampoule filled with ar gas ( @xmath2 0.6 bar ) .
third , the sample - loaded ampoule was heated rapidly to 1123 - 1173 k in a muffle furnace . after holding for 6 hours ,
the sample was quenched , similar to the approach in synthesizing @xmath26fe@xmath0as@xmath0@xcite .
the sample s quality could be improved by repeating the solid - state reaction .
powder x - ray diffraction ( xrd ) was carried out on a panalytical x - ray diffractometer with a monochromatic cu - k@xmath29 radiation .
the crystal structure was refined by a rietveld analysis using a rietan software@xcite .
the electrical transport and heat capacity measurements were conducted on a quantum design physical property measurement system ( ppms-9 ) . in the resistivity measurement , a four - electrode method and the ac transport option were employed .
the sample pellet was cut into a thin rectangular bar with a dimension of [email protected]@xmath300.5 mm@xmath31 , on which thin gold wires were attached with silver paint .
the excitation current was set to 5.18 ma .
the hall coefficient was measured by permutating the voltage and current electrodes@xcite , using a thin - square sample ( [email protected]@xmath300.12 mm@xmath31 ) with four symmetric electrodes attached .
the excitation current was 20 ma , and the applied magnetic field was 80 koe .
the heat capacity was measured by a thermal relaxation method using a square - shaped sample plate with a mass of 19.5 mg .
the dc magnetization for a regular shape sample ( in order to estimate the demagnetization factor ) was measured on a quantum design mpms-5 equipment .
different kind of protocols of zero - field cooling ( zfc ) and field - cooling ( fc ) were employed for probing the superconducting and magnetic transitions .
the as - prepared sample was characterized by powder xrd .
the result shows that most of the reflections can be indexed using a tetragonal lattice , whose unit - cell size ( @xmath32 3.89 and @xmath33 13.31 ) is close to other 1144-type compounds@xcite .
the intensity of the strongest reflection of the impurity phase , which is identified to be unreacted feas , is only 2.5% of that of the ( 103 ) diffraction peak of the main phase , indicating high quality of the sample .
the appearance of ( @xmath34 ) reflections with @xmath35=odd numbers confirms that the tetragonal lattice is primitive rather than body centered .
as@xmath0 whose crystal structure is displayed in the inset .
the relatively strong reflections in low diffraction angles are indexed.,title="fig:",width=302 ] [ fig1 ] figure 1 shows the rietveld refinement profile for the powder xrd data of rbeufe@xmath0as@xmath0 .
the refinement yields a weighted reliable factor @xmath36 of 3.45% and a
goodness - of - fit " parameter @xmath4 of 1.4 , which indicate high reliability for the crystal structure refined . as shown in the inset of fig .
1 , it is indeed an intergrowth of rbfe@xmath1as@xmath1 and eufe@xmath1as@xmath1 .
it can be also viewed as a modified eufe@xmath1as@xmath1 in which every alternate eu atoms are replaced by rb atoms along the @xmath8 axis .
it is the structural modification that changes the lattice from body - centered to primitive .
additionally , the local coordination environment for fe atoms turns out to be asymmetric . according to the refined structural parameters listed in table [ structure ] , the fe@xmath37as bondlengths are no longer equal .
consequently , as1 ( at rb side ) and as2 ( at eu side ) heights from the fe plane are remarkably different ( 1.249 and 1.467 , respectively ) . meanwhile , the bond angles as1@xmath37fe@xmath37as1 and as2@xmath37fe@xmath37as2 are 112.50@xmath12 and 109.04@xmath12 , respectively .
note that the as relative heights as well as the bond angles are opposite to the result in rbcafe@xmath0as@xmath0@xcite .
this suggests that the fe - coordination asymmetry does not interfere with the occurrence of sc , although it could influence the superconducting pairing symmetry .
.[structure ] crystallographic data of rbeufe@xmath0as@xmath0 at room temperature with @xmath38 = 3.8897(1 ) , @xmath8 = 13.3146(6 ) and space group @xmath39 ( no .
[ cols="^,^,^,^,^,^",options="header " , ] comparison of the lattice parameters of rbeufe@xmath0as@xmath0 with those of eufe@xmath1as@xmath1 and rbfe@xmath1as@xmath1 hints the interaction " between the two building blocks .
the @xmath38 axis is 0.009 ( 2 ) smaller than the average of those of eufe@xmath1as@xmath1@xcite and rbfe@xmath1as@xmath1@xcite , meanwhile , the @xmath8 axis is 0.020(5 ) smaller than half of the sum of those of eufe@xmath1as@xmath1 and rbfe@xmath1as@xmath1 .
this lattice shrinkage suggests stabilization of the hybrid 1144 phase .
the interaction of the two building units is also manifested by the shortening of the
rbfe@xmath1as@xmath1 " block ( from 7.267 to 6.914 ) and the stretching of the eufe@xmath1as@xmath1 " block ( from 6.068 to 6.401 ) .
this structural variation seems to be associated with the charge redistribution , since the fe formal valence in rbeufe@xmath0as@xmath0 has to be averaged to 2.25 + instead of being either 2 + in eufe@xmath1as@xmath1 or 2.5 + in rbfe@xmath1as@xmath1 .
the elongation of the eufe@xmath1as@xmath1 " block suggests weakening of the effective coupling between eu-4@xmath40 spins and fe-3@xmath14 itinerant electrons in rbeufe@xmath0as@xmath0 .
as@xmath0 polycrystalline sample .
the inset zooms in the superconducting transition under zero field .
( b ) the superconducting resistive transitions under increased external magnetic fields .
( c ) plot of the upper critical field as a function of temperature , derived from the data shown in ( b ) .
the dashed line is the linear fit.,title="fig:",width=302 ] [ fig2 ] figure 2 shows electrical transport measurement results for the rbeufe@xmath0as@xmath0 polycrystalline sample .
the resistivity @xmath41 exhibits a metallic behavior with a broad hump around 180 k , a common feature of hole - doped fescs@xcite . as expected , the conduction is dominated by a hole - carrier transport , which is verified by the positive hall coefficient ( @xmath42 ) in the whole temperature range of the normal state .
the @xmath42 values appear to be extremely low ( equivalent to a hole content of @xmath431.0 holes / fe using the simple formula @xmath44 ) for a single - band case ) , which indicates a multiband scenario ( the electron - hole compensation effect accounts for the low @xmath42)@xcite .
a sharp superconducting transition appears at 36.5 k , and zero resistance is achieved at 36 k ( note that the transition temperature in @xmath45 is decreased because it was measured under an 80-koe magnetic field ) . this @xmath28 value is the highest among all the doped eufe@xmath1as@xmath1 superconductors@xcite .
furthermore , no re - entrance superconductivity can be observed , in contrast to the case in the eufe@xmath1as@xmath1-related ferromagnetic superconductors where non - zero resistance often re - enters the superconducting state@xcite . under external magnetic fields ,
the superconducting transitions shift mildly to lower temperatures , as shown in fig .
2(b ) . from these data , the upper critical fields , @xmath46 , can be obtained by defining the transition temperature @xmath47 at which the resistivity drops to 90% of the linearly extrapolated one ( this criterion satisfies the result of heat - capacity measurement ) .
the resulting @xmath46 , shown in fig .
2(c ) , is almost linear , in contrast with the pronounced positive curvature observed in eufe@xmath1(as@xmath48p@xmath49)@xmath1@xcite and eu(fe@xmath50ir@xmath51)@xmath1as@xmath1@xcite .
the initial slope , @xmath52d@xmath53/d@xmath54 , is as large as @xmath375.6 t / k , over 4 times of that of eufe@xmath1(as@xmath48p@xmath49)@xmath1 .
here we note that the measurement was performed on the polycrystalline sample , the @xmath55 data obtained actually represent some kind of average of @xmath56 and @xmath57 .
figure 3 shows the dc magnetic susceptibility measured under low magnetic fields using fc and zfc protocols , respectively .
consistent with the resistivity measurement above , a superconducting onset transition occurs at @xmath58 = 36.5 k. the volume fraction of magnetic shielding , measured in a zfc process , is almost 100% at the lowest temperature after making a demagnetization correction .
the volume fraction of magnetic repulsion ( also called meissner volume fraction ) , scaled by the drop in @xmath59 in a fc process , is reduced to 16% due to flux pinning while cooling down under magnetic fields .
nevertheless , the meissner fraction is still over six - fold higher than any impurity content ( estimated to be less than 2.5% from the xrd ) in the sample .
therefore , bulk superconductivity is clearly demonstrated for rbeufe@xmath0as@xmath0 . in comparison
, the previous related superconductor eufe@xmath1(as@xmath48p@xmath49)@xmath1 shows no diamagnetism even in a zfc process for the polycrystalline sample@xcite .
this facts highlight the robustness of sc in the present rbeufe@xmath0as@xmath0 system .
as@xmath0 polycrystalline sample under low magnetic fields in field - cooling ( fc ) and zero - field - cooling ( zfc ) modes , respectively .
the superconducting transition at @xmath58 = 36.5 k and the magnetic transition at @xmath60 = 15 k are marked by arrows , respectively .
the inset magnifies low - temperature data from which a minor anomaly at @xmath61 = 5 k can be distinguished .
right axis : fc data measured in cooling ( fcc ) and heating ( fch ) processes ; left axis : zfc data.,title="fig:",width=302 ] [ fig3 ] interestingly , there is an anomaly at around 15 k in the superconducting state , featured by a sign change in @xmath62 at @xmath63=10 oe . at first sight
, this behavior resembles the ac magnetic susceptibility ( @xmath64 ) curve of the classical reentrant superconductor errh@xmath0b@xmath0 in which sc _ disappears _ when the er magnetic long - range order sets in@xcite . as a matter of fact , however , @xmath64 is equivalent to @xmath65 rather than @xmath62 .
the @xmath65 value of rbeufe@xmath0as@xmath0 remains diamagnetic ( in contrast with the positive @xmath64 in errh@xmath0b@xmath0 ) in the region where @xmath62 becomes positive , indicating sc survives at the low temperatures .
when the applied field is increased to 30 oe , the @xmath62 value is enhanced remarkably , suggesting that it is possibly associated with a ferromagnetism rather than a paramagnetism .
additionally , there is a minor kink at @xmath66 5 k in both fc and zfc data , which can be seen in the magnified plot shown in the inset of fig .
the bifurcation of fcc ( measured in cooling process ) and fch ( measured in heating process ) can also be seen at @xmath67 5 k. we will discuss the possible implications later on . in order to further identify the 15-k transition under the superconducting state , we performed isothermal magnetization [ @xmath68 measurements . as shown in fig .
4 , the @xmath69 data are essentially linear at @xmath70 , indicating a simple paramagnetic state .
when the temperature is decreased to 30 k , which is below @xmath71 but well above 15 k , a typical superconducting loop is superposed on the paramagnetic background . at @xmath72 = 20 k , which is close to the transition temperature , the paramagnetic component of the @xmath69 curve behaves in a shape of a brillouin function .
below 15 k , the overall @xmath69 loops look like a ferromagnet , but the magnetization curves do not merge together at fields higher than the saturation one ( about 1.5 koe ) .
this is due to the existence of sc which shows a flux - pinning effect as mentioned above .
note that the saturation magnetization at 2 k achieves 6.5 @xmath73/eu , basically consistent with the expected value of @xmath74=7.0 @xmath73/eu for the eu@xmath3-spin ferromagnetic alignment .
the above results undoubtfully indicate an eu - spin fm at @xmath60 = 15 k for rbeufe@xmath0as@xmath0 , although the spin orientation is not clear .
the robust ferromagnetic properties contrast with those of the previous ferromagnetic superconductor eufe@xmath1(as@xmath48p@xmath49)@xmath1 which shows a much lower coersive field ( 20 oe at 2 k)@xcite and a much higher saturation field ( @xmath2 7 koe)@xcite .
as@xmath0 at several representative temperatures .
the top - left inset zooms in the data in the low - field region for @xmath72 = 20 , 30 and 40 k.,title="fig:",width=302 ] [ fig4 ] figure 5 plots the magnetic susceptibility @xmath75 ( left axis ) and its reciprocal ( right axis ) of rbeufe@xmath0as@xmath0 as functions of temperature .
the @xmath75 data exhibit a linear dependence in the high-@xmath72 region , indicating dominant curie - weiss paramagnetism .
we thus fit the data from 50 to 300 k by an extended curie s law , @xmath76 , which yields @xmath77 = 0.00178 emu mol@xmath78 , @xmath79 = 7.91 emu k mol@xmath78 and @xmath80 = 23.6 k. if one assumes that @xmath77 is mostly contributed from pauli paramagnetism , a density of state at fermi level [ @xmath81 can be estimated to be 55 ev@xmath78 fu@xmath78 , which is unusually large . from the curie constant @xmath79 , an effective local moment of @xmath82 = 7.95 @xmath73 fu@xmath78 ( fu denotes formula unit )
is obtained , which is almost equal to the expected value of @xmath837.94 ( @xmath73 ) for an eu@xmath3 spin .
the paramagnetic curie temperature @xmath80 is positive , indicating dominant ferromagnetic interactions among the eu spins . indeed , in the low-@xmath72 data shown in the inset of fig . 5
, a canonical ferromagnetic transition can be seen .
the transition temperature can be estimated by the dip in d@xmath84/d@xmath72@xcite , which is about 2 k higher than the @xmath60 value at @xmath63 = 10 oe . ) of rbeufe@xmath0as@xmath0 measured under @xmath63 = 1 koe .
the data from 50 to 300 k are fitted by the extended curie s law .
the reciprocal of susceptibility ( @xmath85 ) is shown using right axis .
the inset shows temperature dependence of magnetization in the low-@xmath72 region , in which a canonical ferromagnetic transition can be seen .
the @xmath60 value can be estimated from the derivative of magnetization ( right axis).,title="fig:",width=302 ] [ fig5 ] to further characterize the superconducting and magnetic transitions , we conducted heat capacity measurement for rbeufe@xmath0as@xmath0 . from the raw @xmath86 data shown in fig .
6(a ) , two anomalies at @xmath60 = 15 k and @xmath71 = 36.5 k can be identified , respectively , verifying the ferromagnetic and superconducting transitions demonstrated above .
the inset shows the superconducting transition more clearly .
the thermodynamic transition temperature , determined by entropy - conserving construction , is 36.0 k , coincident with the zero - resistance temperature .
impressively , the specific - heat jump @xmath87 is as high as 7.5 j k@xmath78 mol@xmath78 , which further confirms bulk sc in rbeufe@xmath0as@xmath0 .
the @xmath88 value achieves 208 mj k@xmath89 mol@xmath78 .
thus the electronic specific - heat coefficient @xmath90 can be estimated to be 145 mj k@xmath89 mol@xmath78 by assuming @xmath91 = 1.43 in the bcs weak - coupling scenario
. a similarly large @xmath90 value of @xmath2 150 mj k@xmath89 mol@xmath78 can be _ independently _ estimated by the enhanced room - temperature specific heat that is 46 j k@xmath78 mol@xmath78 larger than the dulong - petit limit 3@xmath92 ( @xmath93 = 10 , being the number of atoms in a formula unit ) .
therefore , the real @xmath90 value should be around 150 mj k@xmath89 mol@xmath78 , equivalent to 38 mj k@xmath89 mol - fe@xmath78 . such an enhanced @xmath90 is often observed in hole - doped fescs@xcite .
notably , the estimated @xmath90 value corresponds to @xmath94 60 ev@xmath78 fu@xmath78 , consistent with the value derived from the magnetic measurement .
namely , the wilson ratio is about unity albeit of an enhanced @xmath95 .
[ fig6 ] figure 6(b ) shows the variations in @xmath86 under a magnetic field of 8 t. first , the superconducting transition temperature decreases slightly to 35.4 k , consistent with the result of magneto - resistivity measurement above .
second , the anomaly at @xmath96 is smeared out by the field . in addition , a minor anomaly at @xmath66 5 k , which can be seen more easily in fig . 6(c ) and ( d ) , tends to disappear as well .
third , the external field induces a @xmath86-weight transfer from lower to higher temperatures . at @xmath97 20 k ,
@xmath86 is suppressed by the field ; while in the temperature range of 22 k @xmath98 100 k , @xmath86 is enhanced substantially .
the field - induced change in @xmath86 has to be ascribed to the eu - spin magnetism , because electronic and phonon specific heat ( @xmath99 and @xmath100 ) generally do not change with magnetic fields .
the external fields force the eu spins along the field direction , which severely broadens the phase transition , making the @xmath86 weight shifts to higher temperatures consequently .
similar effect is also observed in eufe@xmath1(as@xmath48p@xmath49)@xmath1@xcite .
the @xmath86-weight transfer under different magnetic fields is explicitly shown in the @xmath101 plot shown in fig .
besides , an upward shift of @xmath60 can also be seen , further supporting the ferromagnetic transition . at @xmath66 5 k ,
the zero - field @xmath101 curve exhibits a tiny jump , which also changes with magnetic fields . under a high field
, the jump turns into a shoulder that moves to high temperatures with increasing field . this sensitive response to magnetic fields suggests that it should be an intrinsic phenomenon .
we will return to this topic in the discussion section .
shown in the inset of fig .
6(c ) is the @xmath101 vs @xmath102 plot for the low-@xmath72 data .
as is known , the expected @xmath101 value at @xmath103 0 k is zero for a fully - gaped superconductor , which increases linearly with magnetic fields .
however , the situation here is apparently abnormal : the zero - field @xmath101 value is as large as 573 mj k@xmath89 mol@xmath78 at 2 k , which is then suppressed to 86 mj k@xmath89 mol@xmath78 at 8 t. this result suggests that the eu - spin magnetic contribution ( @xmath104 ) remains dominant around 2 k. future measurements down to lower temperatures are expected to reveal the eu - spin wave excitations as well as the quasi - particle excitations from the superconducting state .
intriguingly , the specific - heat anomaly at @xmath60 is very different from those of most magnetic orderings ( such as those in eufe@xmath1as@xmath1@xcite and eufe@xmath1(as@xmath48p@xmath49)@xmath1@xcite ) which show a clear jump because of a second - order transition . in the present case for rbeufe@xmath0as@xmath0 , however , it simply shows a kink instead .
we thus plot the derivative of specific heat , as shown in fig .
. an obvious jump at @xmath60 can be seen in the zero - field d@xmath79/d@xmath72 data .
the result suggests that the magnetic transition is of a rare third order , according to the ehrenfest classification@xcite . to the best of our knowledge
, no third - order magnetic transition has been reported in a real material before .
as we know , most magnetic ordering transitions are described by landau s second - order phase transition theory .
therefore , the observation of third - order magnetic transition deserves further investigations .
rbfe@xmath1as@xmath1 is known to be a 2.6 k superconductor@xcite , while eufe@xmath1as@xmath1 belongs to a parent compound of fescs@xcite . here
we show that , interestingly , their hybrid material rbeufe@xmath0as@xmath0 turns out to be a ferromagnetic superconductor .
similar material hybridization effect in fescs is witnessed in ba@xmath1ti@xmath1fe@xmath1as@xmath0o@xcite which is actually an intergrowth of non - superconducting bati@xmath1as@xmath1o and bafe@xmath1as@xmath1 .
ba@xmath1ti@xmath1fe@xmath1as@xmath0o shows sc at 21 k owing to a charge transfer between different layers , in relation with an electron - correlation effect@xcite . here in rbeufe@xmath0as@xmath0 ,
the isolated
rbfe@xmath1as@xmath1 " block is heavily hole doped , in contrast with the undoped state in the eufe@xmath1as@xmath1 " block .
consequently , the structural hybridization leads to a charge homogenization , since there is only one equivalent fe site .
namely , rbeufe@xmath0as@xmath0 is naturally hole doped by 22.5% .
we hope that such a structural hybridization approach by design@xcite may be utilized for future material explorations .
notably , the @xmath71 value of rbeufe@xmath0as@xmath0 is unexpectedly high , which is nearly equal to the maximum @xmath71 of 36.8 k for cssrfe@xmath0as@xmath0 in the @xmath26fe@xmath0as@xmath0 series@xcite .
previous studies on doped eufe@xmath1as@xmath1 superconductors show that the @xmath71 is always significantly lower than the eu - free counterparts , which is ascribed to the eu4@xmath105fe@xmath106 interaction .
therefore , the unsuppressed @xmath71 in rbeufe@xmath0as@xmath0 suggests vanishingly small eu4@xmath105fe@xmath107 coupling .
the elongation of eu@xmath37fe interatomic distance from 3.604 in eufe@xmath1as@xmath1 to 3.685 in rbeufe@xmath0as@xmath0 supports this point of view . in addition
, the high value of @xmath52d@xmath53/d@xmath54 ( @xmath375.6 t / k ) as well as the large specific - heat jump ( 7.5 j k@xmath78 mol@xmath78 ) further indicates robustness of sc in rbeufe@xmath0as@xmath0 .
in addition to the robust sc , evidence of eu fm is also strong and sufficient .
the @xmath69 data show an obvious magnetic hysteresis ( with a coersive field of 360 oe at 2 k ) as well as a saturation magnetization ( 6.5 @xmath73/eu ) that corresponds to eu@xmath3-spin ferromagnetic alignment .
besides , both the positive paramagnetic curie temperature and the increase of @xmath96 with magnetic field support the ferromagnetic transition scanario . hopefully , future measurements of the anisotropic magnetic properties using single crystal samples will give more information on the magnetic state .
needless to say , investigations of the mssbauer spectra , x - ray resonant magnetic scattering and neutron scattering will be definitely helpful to clarity the eu - spin orientation and other related information .
the appearance of eu fm in rbeufe@xmath0as@xmath0 can be explained in terms of modification of @xmath19 , since @xmath19 [ @xmath108 _ oscillates _ and tends to decay with @xmath24 . in rbeufe@xmath0as@xmath0 ,
the interlayer eu interatomic distance @xmath24 becomes 13.31 , which is almost twice as the @xmath24 value in eufe@xmath1as@xmath1 ( 6.657 ) .
this may change the sign of @xmath19 , resulting in an eu fm . the lowered curie temperature [ about 4 k lower than that of eufe@xmath1(as@xmath48p@xmath49)@xmath1 ] seems to be associated with the decay of @xmath19 .
the weak interlayer magnetic coupling not only accounts for the relatively low curie temperature , but also justifies the robustness of sc .
additionally , it could be related to the third - order transition observed , because of the enhanced two dimensionality which allows strong magnetic fluctuations above @xmath96 .
finally , let us discuss how sc and fm compromise in rbeufe@xmath0as@xmath0 .
there were a few theoretical proposals available on the fm+sc issue decades ago@xcite .
the first solution is that fm is modified in the form of a
cryptoferromagnetic " state or a multidomain structure ( with domain size @xmath14 ) , hence the superconducting cooper pairs feel no net magnetization if the superconducting coherent length satisfies @xmath110@xcite .
the second idea considers that superconducting cooper pairs are magnetized " by the exchange fields , such that they possess non - zero momentum@xcite , or equivalently , sc is modulated to be inhomogeneous in real space@xcite .
this scenario is often called fulde - ferrell - larkin - ovchinnikov ( fflo ) state .
alternatively , the internal spontaneous magnetization field from fm penetrates the superconductor , which induces spontaneous vortices " even at zero external field , hence called spontaneous vortex ( sv ) state@xcite .
since non - superconducting defects tend to trap spontaneous vortices , the fflo state is believed to be preferably stabilized in a clean - limit superconductor . as for the rbeufe@xmath0as@xmath0 system , the first candidate seems to be unlikely , because the estimated coherence length at zero temperature is only 1.52 nm@xcite , which constrains the ferromagnetic domain size too much to present a long - range order .
note that the residual resistivity tends to be very small by a rough extrapolation ( see fig .
2 ) even for the polycrystalline sample , which means a quite long electron mean free path and , the flux - pinning effect is not severe from the fc diamagnetic signal ( see fig .
these facts suggest that the superconductor is probably in a clean limit at lower temperatures .
we thus speculate that the anomaly at @xmath66 5 k could be related to the change in the way of fm+sc .
there are indeed some signatures .
the bifurcation of fcc and fch curves above @xmath61 , shown in fig .
3 , implies the transformation from fflo to sv state .
the small specific - heat jump at @xmath61 [ fig .
6(c ) ] suggests a weak phase transition .
in addition , @xmath61 increases with field .
obviously , more investigations are needed to address this interesting issue .
to summarize , we have discovered an iron - based compound rbeufe@xmath0as@xmath0 in which an unprecedented coexistence of sc and fm is observed .
the new material crystallizes in an intergrowth structure of body - centered rbfe@xmath1as@xmath1 and eufe@xmath1as@xmath1 , which drastically changes the physical properties .
the feas layers are found to be hole doped by the material itself , presenting a large normal - state electronic specific - heat coefficient ( @xmath2 150 mj k@xmath89 mol@xmath78 ) and a pauli spin susceptibility ( @xmath2 0.00178 emu mol@xmath78 ) .
bulk superconductivity at @xmath28 = 36.5 k and eu - spin ferromagnetism at 15 k are unambiguously demonstrated .
the specific - heat jump at @xmath28 is as high as 7.5 j k@xmath78 mol@xmath78 .
the eu - spin ferromagnetism is manifested by a magnetic hysteresis with a 360-oe coersive field at 2 k , a saturation magnetization of 6.5 @xmath73/eu , and a saturation field of @xmath2 1.5 koe at low temperatures .
there are two additional novel phenomena observed in rbeufe@xmath0as@xmath0 .
first , the eu - spin ferromagnetic transition is of a rare third order , evidenced by the continuity in @xmath86 and discontinuity in d@xmath79/d@xmath72 at @xmath96 .
second , a weak anomaly at about 5 k possibly reflects the interplay between sc and fm . these intriguing observations together with the robust coexistence of sc and fm call for further investigations on the interesting title material .
50ifxundefined [ 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 `` , '' link:\doibase 10.1103/physrevlett.102.137002 [ * * , ( ) ] `` , '' http://stacks.iop.org/0953-8984/23/i=6/a=065701 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.89.014512 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.90.094407 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.78.052501 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.78.052502 [ * * , ( ) ] `` , '' http://stacks.iop.org/1367-2630/11/i=2/a=025007 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.80.134411 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.80.174424 [ * * , ( ) ] `` , '' http://stacks.iop.org/0953-8984/23/i=46/a=464204 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.83.054511 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.80.184514 [ * * , ( ) ] `` , '' http://stacks.iop.org/0295-5075/95/i=6/a=67007 [ * * , ( ) ] `` , '' http://stacks.iop.org/1367-2630/15/i=11/a=113002 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.88.214516 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.91.064506 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.86.220505 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.84.140503 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevlett.110.237002 [ * * , ( ) ]
`` , '' @noop * * , ( ) `` , '' link:\doibase 10.1080/00018738500101741 [ * * , ( ) ] `` , '' \doibase http://dx.doi.org/10.1016/j.physc.2015.02.050 [ * * , ( ) ] @noop `` , '' http://stacks.iop.org/1742-6596/391/i=1/a=012123 [ * * , ( ) ] `` , '' http://dx.doi.org/10.1080/00018732.2010.513480 [ * * , ( ) ] `` , '' link:\doibase 10.1103/revmodphys.83.1589 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.77.220503 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.79.094426 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevlett.100.237003 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.86.155119 [ * * , ( ) ] `` , '' http://stacks.iop.org/1674-1056/22/i=8/a=087410 [ * * , ( ) ] `` , '' link:\doibase 10.1021/jacs.5b12571 [ * * , ( ) ] `` , '' in <
go to isi>://wos:000249626200003 [ _ _ ] , , vol .
, ( ) pp .
`` , '' @noop * * ( ) `` , '' < go to isi>://wos : a1984aav2700033 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevlett.101.107006 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.84.184512 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.78.092406 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevlett.38.987 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevb.79.174501 [ * * , ( ) ] `` , '' link:\doibase 10.1007/s004070050021 [ * * , ( ) ] `` , '' \doibase http://dx.doi.org/10.1016/j.physc.2009.11.103 [ * * , ( ) ] `` , '' link:\doibase 10.1021/ja304315e [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevlett.113.266407 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrev.116.898 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrev.135.a550 [ * * , ( ) ] `` , '' < go to isi>://wos : a19656224900049 [ * * , ( ) ] `` , '' link:\doibase 10.1103/physrevlett.46.49 [ * * , ( ) ] `` , '' \doibase http://dx.doi.org/10.1016/0038-1098(80)90620-1 [ * * , ( ) ] @noop `` '' | we discover a robust coexistence of superconductivity and ferromagnetism in an iron arsenide rbeufe@xmath0as@xmath0 .
the new material crystallizes in an intergrowth structure of rbfe@xmath1as@xmath1 and eufe@xmath1as@xmath1 , such that the eu sublattice turns out to be primitive instead of being body - centered in eufe@xmath1as@xmath1 .
the feas layers , featured by asymmetric as coordinations , are hole doped due to charge homogenization .
our combined measurements of electrical transport , magnetization and heat capacity unambiguously and consistently indicate bulk superconductivity at 36.5 k in the feas layers and ferromagnetism at 15 k in the eu sublattice .
interestingly , the eu - spin ferromagnetic ordering belongs to a rare third - order transition , according to the ehrenfest classification of phase transition .
we also identify an additional anomaly at @xmath2 5 k , which is possibly associated with the interplay between superconductivity and ferromagnetism . |
it was the common expectation that , with the increase in the precision of rv measurements , the detected secondary masses of spectroscopic binaries will gradually decrease , first into the brown dwarf mass range ( @xmath5 ) and eventually into the planetary mass regime of roughly @xmath6 ( below the deuterium burning limit ) .
quite the opposite turned out to happen .
the pioneering long - term doppler survey of campbell , walker , & yang ( 1988 ) , and later also in the southern hemisphere by murdoch , hearnshaw , & clark ( 1993 ) , did not reveal any brown dwarf companions to nearby solar - type stars , despite their sufficient sensitivity .
there appeared to be an abrupt drop in the frequency of companions at the hydrogen burning limit of @xmath7 .
the first very low - mass companion ( @xmath8 ) detected by the rv technique is the companion to hd 114762 ( latham et al .
the discoverers classified the object as a brown dwarf companion , but according to cochran , hatzes , & hancock ( 1991 ) this companion might be in fact a late m dwarf due to a possible low value of the inclination angle @xmath9 . in the years after the discovery of the first unambiguous extrasolar planet orbiting a main - sequence star by mayor & queloz ( 1995 ) ,
a total of more than a @xmath10 extrasolar planets have been accumulated by several precise doppler searches all around the globe .
still , with a few exceptions ( hd 127506 : mayor , queloz , & udry 1998 ; hd 29587 & hd 140913 : mazeh , latham , & stefanik 1996 ; hd 168443 ( second companion ) : marcy & butler 2001 ; hd 184860 : vogt et al .
2002 ) , the part of the mass function for spectroscopic secondaries , in between the minimum hydrogen burning mass and the deuterium burning limit of @xmath11 , remained surprisingly sparsely populated . a notable exception is hd 10697 , where vogt et al .
( 2000 ) detected a companion with @xmath12 , while zucker & mazeh ( 2000 ) determined a true mass for the secondary of @xmath13 after combining the spectroscopic solution with @xmath14 astrometry .
udry et al . (
2002 ) found 4 companions with minimum masses close to the planet / brown dwarf border , which could be either `` superplanets '' or real brown dwarfs .
however , most of the candidate brown dwarf secondaries considered in the halbwachs et al .
( 2000 ) study turned out to be stellar mass companions , again after combination with @xmath14 astrometry .
this suggests the existence of two distinctive binary formation processes for certain mass - ratios .
the gap in the mass function between the stellar and planetary mass range for close binaries is generally dubbed the `` brown dwarf desert '' ( this desert does not appear to exist at wide separations , e.g. gizis et al .
2001 ; neuhuser & guenther 2004 ) .
the gap is particularily obvious as a sudden drop at the deuterium burning limit in the substellar mass function for close companions . while planets are relatively common as companions to solar - type stars , brown dwarfs are rare
this fact is especially surprising since brown dwarf companions are much easier to detect , even for lower precision surveys , due to the larger rv amplitudes .
moreover , these objects cover a more than 4 times larger mass range than planetary companions , still their detected numbers per mass bin is extremely low compared to lower and higher mass companions . in this work we present the discovery of another possible `` oasis '' in this desert , a brown dwarf orbiting the g - type star hd 137510 .
hd 137510 (= hr 5740 = hip 75535 ) is a bright ( @xmath15 ) g - type star at a distance of @xmath16 pc , according to the @xmath14 parallax of @xmath17 mas ( esa 1997 ) . in order to derive the stellar parameters for the primary we used the ( b - v)-@xmath18 calibration of alonso , arribas , & martnez - roger ( 1996 ) , compared @xmath19 and @xmath18 with stellar evolution calculations ( e.g. reddy et al .
2003 ) , and performed a model atmosphere analysis of a high resolution , high signal - to - noise spectrum of hd 137510 .
we find that the star has a mass of @xmath20 , an effective temperature @xmath21 k , a @xmath22 of @xmath23 , a radius @xmath24 of @xmath25 and an age of @xmath26 gyrs .
an iron abundance analysis yields @xmath27}= 0.16\pm0.07 $ ] dex .
the object is thus metal rich compared to the sun .
we also find substantial rotational broadening of the spectral lines , the value of @xmath28 can be constrained in the range @xmath29 to @xmath30 .
these parameters are consistent with ( or very close to ) those determined by lbre et al .
( 1999 ) and do nascimento et al .
it thus appears that hd 137510 was a metal rich late f - type star which has now started its evolution away from the main sequence .
interestingly , smith , cunha , & lazzaro ( 2001 ) have noted that hd 137510 possibly underwent chemically fractioned accretion and represents a good candidate for having a close - in giant planetary companion .
we carried out observations of hd 137510 as part of the ongoing doppler surveys for extrasolar planets using the harlan j. smith 2.7 m telescope at mcdonald observatory ( mcd ) in west texas and the 2.0 m alfred jensch telescope at thringer landessternwarte tautenburg ( tls ) in germany . in september 1998 a first spectrum of hd 137510
was taken at mcd , but rv monitoring was discontinued for almost three years . in may
2001 observations of hd 137510 were started at tls , while at the same time monitoring was resumed at mcd .
for both programs we use a temperature stabilized molecular iodine vapor ( i@xmath31 ) cell as the velocity metric . to obtain the high precision differential rv measurements we employ specialized data modeling techniques , which include the detailed reconstruction of the instrumental profile of the respective spectrometers at the time of observation ( valenti , butler , & marcy 1995 ; butler et al .
1996 ; endl , krster , & els 2000 ) .
[ orbit_time ] displays our rv data as a function of time and with the best - fit keplerian orbital solution superimposed .
we employ the program @xmath32 ( mcarthur , jefferys , mccartney 1994 ) , which uses a robust estimation method to find orbital solutions for each data set ( mcd and tls ) independently , as well as a combined solution , where the arbitrary velocity zero points ( @xmath33 velocities ) of both data sets are included as fit parameters .
table [ orbittab ] summarizes the parameters we determined for the orbit . the resulting mass function of this binary yields a minimum mass for the companion of @xmath34 .
table [ vels ] lists our doppler measurements for hd 137510 .
the residual scatter around this orbital fit is unusually high , mcd : @xmath35 and tls : @xmath36 , roughly a factor of @xmath37 worse than the routine rv precision for each telescope system .
it also exceeds the average internal measurement errors of both data sets ( mcd : @xmath38 , tls : @xmath39 ) .
stellar activity can introduce additional noise in rv data at that level of measurement precision ( e.g. saar & donahue 1997 ) .
for hd 137510 we can rule out this possibility , based on the general low ca ii h&k emission of this star .
the mcd spectra , for which we determined the rvs , also include the ca ii h&k lines , two well known activity indicators .
we find a mcd s - index @xmath40 ( see section 2.3 in paulson et al .
2002 for a detailed description of how we obtain the s - index for the ca ii k line core emission ) .
this is a comparable emission level as the well known inactive star @xmath41 ceti ( @xmath42 ) and shows that hd 137510 is not an active star .
furthermore , the star does not show up as an x - ray source in the rosat all - sky survey results ( hnsch , schmitt , & voges 1998 ) .
one possible explanation for the higher scatter is the substantial rotational broadening of the spectral lines in the case of hd 137510 ( @xmath43 ) .
the larger scatter for the tls data likely reflects the differences both in telescope aperture ( tls : @xmath44 m ; mcd : @xmath45 m ) as well as in the relative seeing conditions at these two sites .
another possible explanation is the presence of a second low - level keplerian signal .
we searched in the rv residuals for periodic signals using the lomb - scargle periodogram ( lomb 1976 , scargle 1982 ) .
while the two separate periodograms of the individual sets of residuals do not reveal a significant signal , we do find a strong signal with a period of @xmath46 days in the combined residuals .
this warrants further intensive monitoring of hd 137510 to confirm a possible planetary companion at a separation of @xmath47 au ( which would make this system very interesting from a dynamical standpoint ) .
due to the general sparseness of brown dwarf rv companions we have to look for additional clues in order to rule out a stellar mass companion producing the detected keplerian signal : simply based on the probabilty of a random distribution of the inclination angle @xmath9 , we already have a @xmath48 probability that @xmath9 is larger than @xmath49 , which translates into a true mass of @xmath50 , still below the hydrogen burning limit .
we measure a @xmath28 of @xmath29 to @xmath30 for hd 137510 , which is a typical value for a g0iv star .
the mean @xmath28 value of the f8 - g2 subgiants in the de medeiros & mayor ( 1999 ) catalogue is @xmath51 ( excluding one fast rotating f8iv star the mean value drops to @xmath52 ) .
this shows that the value for hd 137510 is not atypically low , which means that we probably do not have a near pole - on view on the primary star . assuming that the orbital plane is roughly perpendicular to the rotation axis of the star we , again
, do not seem to face a low @xmath9 scenario in this case .
mcalister et al . (
1987 ) obtained a negative result for this star in their speckle survey for duplicity among bright star catalogue stars using the cfht .
granted , a positive detection would have been a difficult one , close to the diffraction limit of the telescope ( @xmath53 arcsec ) .
hd 137510 has no flag in the multiple / binary annex of the @xmath14 catalogue , showing that no suspicious astrometric variation was discovered .
according to lindegren et al .
( 1997 ) a stochastic solution is chosen for objects where no acceptable single or double star solution can be found and additional noise , typically ranging from @xmath54 to @xmath55 mas , needs to be added until an acceptable fit is found .
if we regard the @xmath54 mas as the lower limit of an astrometric pertubation ( angular size of the semimajor axis ) which would have shown up in the analysis of the @xmath14 data of a bright target like hd 137510 , we can exclude inclination angles of @xmath56 and a true mass of the companion of @xmath57 , using the constraint of pourbaix & jorissen ( 2000 ) .
we have presented observational evidence for the substellar nature of the companion to hd 137510 . _
the question , why this star constitutes one of the very rare oases in the brown dwarf desert , remains unanswered .
_ however , due to the well established parameters of the primary and the orbit , this new system might help to understand binary formation processes for these mass - ratios at small orbital separations .
we are grateful to the mcdonald observatory time - allocation - committee for generous allocation of observing time .
this material is based upon work supported by the national aeronautics and space administration under grant nag5 - 9227 issued through the office of space science , and by national science foundation grant ast-9808980 .
we thank our referee adam burgasser for his helpful comments .
lindegren , l. , mignard , f. , sderhjelm , s. , badiali , m. , bernstein , h .- h . , lampens , p. , pannunzio , r. , arenou , f. , bernacca , p.l . , falin , j.l .
, froeschl , m. , kovalevsky , j. , martin , c. , perryman , m.a.c . , & wielen , r. 1997 , , 323 , l53 lrrr period [ days ] & @xmath58 & @xmath59 & @xmath60 + t@xmath61 [ jd ] & @xmath62 & @xmath63 & @xmath64 + k[@xmath65 & @xmath66 & @xmath67 & @xmath68 + @xmath69 & @xmath70 & @xmath71 & @xmath72 + @xmath73 & @xmath74 & @xmath75 & @xmath76 + @xmath77 $ ] & @xmath78 & @xmath79 & @xmath78 + @xmath80 [ au ] & @xmath81 & @xmath82 & @xmath81 + @xmath83 $ ] & @xmath84 & @xmath85 & mcd:@xmath86 tls:@xmath87 + rrrr 51067.6406 & 165.7 & 7.5 & mcd + 52037.7836 & -209.7 & 7.5 & mcd + 52040.8490 & -182.1 & 7.1 & mcd + 52114.7432 & -154.0 & 7.5 & mcd + 52141.6609 & -133.2 & 8.2 & mcd + 52142.6541 & -151.1 & 8.3 & mcd + 52144.7022 & -133.4 & 17.4 & mcd + 52145.6828 & -123.6 & 8.3 & mcd + 52331.9521 & 101.9 & 7.7 & mcd + 52353.9910 & 134.5 & 15.2 & mcd + 52357.9625 & 151.6 & 6.6 & mcd + 52388.9414 & 265.5 & 9.1 & mcd + 52451.7614 & 473.5 & 7.4 & mcd + 52472.7426 & 577.6 & 7.6 & mcd + 52492.6720 & 665.2 & 7.4 & mcd + 52576.5355 & 822.8 & 9.3 & mcd + 52689.0031 & 62.4 & 9.4 & mcd + 52689.0124 & 53.4 & 8.9 & mcd + 52741.8983 & -124.1 & 11.9 & mcd + 52741.9064 & -131.8 & 9.4 & mcd + 52743.9543 & -119.9 & 7.7 & mcd + 52804.8642 & -184.4 & 7.3 & mcd + 52804.8749 & -183.0 & 7.5 & mcd + 52807.6845 & -191.4 & 7.7 & mcd + 52807.6914 & -190.1 & 7.5 & mcd + 52839.7200 & -186.1 & 8.3 & mcd + 52895.5874 & -174.6 & 8.5 & mcd + 52895.6010 & -168.5 & 8.0 & mcd + 52896.6072 & -179.5 & 7.4 & mcd + 52931.5547 & -176.0 & 8.2 & mcd + 52931.5686 & -180.9 & 8.4 & mcd + 53018.9949 & -83.5 & 9.2 & mcd + 53018.0065 & -85.7 & 8.8 & mcd + 53038.9840 & -27.5 & 8.7 & mcd
+ & & & + 52039.5781 & -220.5 & 7.1 & tls + 52040.5039 & -233.4 & 6.4 & tls + 52042.4922 & -224.5 & 6.4 & tls + 52066.4219 & -239.6 & 7.6 & tls + 52090.4375 & -186.8 & 6.8 & tls + 52213.2188 & -81.1 & 18.7 & tls + 52280.6875 & -31.6 & 11.0 & tls + 52334.5703 & 77.4 & 9.5 & tls + 52365.5430 & 162.0 & 6.3 & tls + 52366.5469 & 173.5 & 8.6 & tls + 52394.5391 & 280.6 & 9.7 & tls + 52448.4219 & 445.9 & 5.7 & tls + 52452.3984 & 471.8 & 7.7 & tls + 52478.4805 & 529.5 & 19.8 & tls + 52483.4062 & 573.5 & 16.1 & tls + 52506.3320 & 656.7 & 15.7 & tls + 52516.3125 & 754.5 & 14.3 & tls + 52517.2969 & 722.5 & 28.6 & tls + 52519.3086 & 762.8 & 12.5 & tls + 52684.6445 & 15.5 & 14.0 & tls + 52745.5664 & -156.9 & 11.0 & tls + 52754.5312 & -174.1 & 8.5 & tls + 52804.4180 & -217.1 & 10.3 & tls + 52831.3984 & -187.3 & 7.6 & tls + 52833.3984 & -224.4 & 8.5 & tls + 52834.4102 & -193.4 & 10.5 & tls + 52835.3984 & -192.3 & 8.4 & tls + 52836.4766 & -195.8 & 10.2 & tls + 52838.4766 & -198.3 & 14.6 & tls + 52839.4141 & -187.7 & 9.9 & tls + 52840.4023 & -190.5 & 10.8 & tls + 52841.3516 & -225.3 & 12.3 & tls + 52861.3750 & -185.1 & 13.2 & tls + 52863.3516 & -237.1 & 17.1 & tls + 52872.3398 & -211.5 & 18.7 & tls + 52877.3672 & -193.1 & 10.3 & tls + 52878.3594 & -189.0 & 13.0 & tls + 52930.2266 & -195.6 & 9.8 & tls + 52931.2227 & -222.3 & 7.2 & tls + 52952.1992 & -135.6 & 11.2 & tls + 52956.1992 & -164.3 & 16.1 & tls + 52982.7031 & -144.2 & 12.5 & tls + | since the beginning of precise doppler surveys , which have had stunning success in detecting extrasolar planetary companions , one surprising enigma has emerged : the relative paucity of spectroscopic binaries where the secondary mass lies in between the stellar and planetary mass regime .
this gap in the mass function for close - in ( @xmath0 au ) companions to solar - type stars is generally referred to as the `` brown dwarf desert '' . here
we report the detection of a companion to hd 137510 ( g0iv ) , with a minimum mass of @xmath1 , moving in an eccentric orbit ( @xmath2 ) with a period of @xmath3 days and an orbital semimajor axis of @xmath4 au .
the detection is based on precise differential radial velocity ( rv ) data obtained by the mcdonald observatory and thringer landessternwarte tautenburg planet search programs . |
the hegra ( high energy gamma ray astronomy ) experiment is located at the observatorio del roque de los muchachos on the canary island of la palma , ( 2200 m a.s.l .
, 28.75@xmath2 n , 17.89@xmath2 w ) covering an area of about 200 m @xmath3 200 m .
the layout is shown in fig.[layout ] .
the experiment consists of 6 imaging air cherenkov telescopes and 3 arrays , the scintillator array , the wide angle cherenkov array airobicc @xcite and the geiger tower array@xcite .
one of the major aims of the hegra experiment is the search for galactic and extragalactic sources of tev @xmath0-radiation and a detailed study of their characteristics with the imaging atmospheric cherenkov technique at energies above 500 gev .
hereby the stereoscopic observation with the system of telescopes will be exploited .
the tev @xmath0-astronomy is continued towards higher energies ( @xmath4 20 tev ) with the three arrays . for the arrays
a further important aim is a measurement of the energy spectrum and elemental composition of primary cosmic rays in the energy region from @xmath5 500 tev to @xmath5 10 pev . in the following first the status and performance of the imaging atmospheric cherenkov telescopes and their recent results concerning mrk 501 and the microquasar grs1915
+ 105 are given , then the recent results from the arrays are presented concerning galactic and extragalactic @xmath0 sources , tev counterparts of grb s , cosmic ray anisotropies and the energy spectrum and elemental composition of the primary cosmic rays .
presently we are very close to complete a stereoscopic system consisting of 5 imaging telescopes ( reported by g.hermann and m.ulrich at this symposium ) which during the next several years will serve as one of the most powerful instruments in the field , operating at energies above 500 gev @xcite .
the multi mirror reflector of each telescope ( fig . [ telescop ] ) has an area @xmath6 .
the 60 cm diameter front aluminized and quartz coated spherical glass mirrors with a focal length of about 5 m are independently mounted on an almost spherical frame of an alt - azimuth mount .
the angular resolution of the reflector is better than 10 the mount is driven by two stepping motors , and controlled by shaft encoders which provide a tracking accuracy of @xmath7 .
each telescope is equipped with a 271-channel camera @xcite consisting of a close - packed cluster of circular pmts ( emi 9083 kafl ) faced with hollow hexagonal - to - round light guides ( fig [ camera ] ) .
the diameter of a pixel is about @xmath8 , which results in the telescope s field of view fov@xmath9 .
the photon to photoelectron conversion efficiency including all kinds of losses ( due to the mirror reflectivity , the quantum efficiency of pmts , losses in light guides , _ etc . _ ) , and averaged over the spectrum of the cherenkov light at the level of observations , is around @xmath10 . the entire electronics , including the trigger , readout , and the telescope control , is based on the vme standard . the digitization of the pmt pulses is performed with a flash adc system which allows to reconstruct the shape and time distribution of the pulses .
the trigger decision is taken on two levels @xcite .
the fast trigger requires a minimum number of pixels to be fired , and the second ( slower ) level restricts the topology of an image . near each telescope
there is a small hut which houses the local electronics .
each telescope is equipped with a local cpu which communicates via an ethernet line with the central computer .
although each telescope is built as stand alone detector , all telescopes can be integrated into the system by software command through the central computer ( pentium pc ) .
= 7.5 cm = 5.5 cm the first nominal ` system - telescope ' called ( for chronological reasons ) ct-3 , is in operation since august 1995 .
the data taken in the 95/96 season generally confirm the expected performance of a single telescope .
the next 3 telescopes ct-4 , ct-5 , and ct-6 have been installed , tested and calibrated during july - december 1996 , and are now operating together with ct-3 as a system in the 2-fold coincidence mode .
the preliminary results are very encouraging and already demonstrate the power of the stereoscopic approach ( see below ) .
currently , we use different fadc systems in different telescopes , but in the near future ( until march 1997 ) we plan to operate the above four telescopes with the 120 mhz fadc system . at approximately the same time
we plan to replace the present 61-channel ( pixel size @xmath11 , and fov@xmath12 ) camera of ct-2 by a standard 271-pixel high resolution camera .
thus we hope that after the test and calibration work , we will be able to run in the second half of 1997 the whole system as a very sensitive instrument for deep study of vhe @xmath0-rays in the energy region 0.5 - 10 tev .
in addition we plan to upgrade the first ( prototype ) hegra telescope ct-1 .
this equatorial mount telescope with @xmath13 reflector was installed and is taking data since 1992 @xcite . in 1994 the original 37-channel camera was replaced by a new high resolution 127-channel camera ( pixel size @xmath8 ) based on the same pixels used in the ` system ' telescopes , but designed in the camac standard @xcite . in 1997 the round glass mirrors will be replaced by light hexagonal aluminum mirrors with a total reflector area @xmath14 .
thus the energy threshold of the telescope will be reduced to 500 gev .
we plan to use this telescope for individual observational programs , especially in immediate reaction to potential vhe @xmath0-ray emitters showing flaring activity in other wavelengths . at the same time
this telescope being located very near to the central telescope ct-3 , can be effectively integrated , if necessary , with other iacts in order to set up 2 or 3 independent stereoscopic iact systems for simultaneous observations of different sources .
= 7.5 cm = 5.5 cm although @xmath0-ray showers of energy @xmath15 with core position within 100 m from the telescope produce @xmath16 photoelectrons in the camera , sufficient for proper image analysis , the energy threshold of a single telescope is somewhat larger , and is determined by the suppression efficiency of the night sky background ( n.s.b . ) at the trigger level .
the hardware trigger criterion for each telescope , optimized by monte carlo calculations , is organized such that the signals in at least 2 neighboring pixels from the inner 169 pixels should exceed @xmath17 photoelectrons ( @xmath18 ) .
the monte carlo calculations of the detection rates of @xmath0-rays from a point source , presented in fig . [ rates ] , show that the effective energy threshold of the telescope ct-3 is between 700 gev and 1 tev , depending on the slope of the primary photon spectrum .
the first observations with ct-3 generally confirm the calculated performance of a single telescope , namely , a detection rate of @xmath0-rays from the crab nebula above 700 gev of about @xmath19 on top of a cosmic ray ( cr ) detection rate @xmath20 , the significance of the signal per 1 hour of observations being @xmath21 .
= 6 cm = 6 cm = 6.5 cm = 7 .
cm the requirement of simultaneous detection of showers by locally triggered @xmath22 telescopes significantly reduces the effect of the n.s.b . as well as that of the another important background connected with local muons .
this allows a lower energy threshold of the system down to @xmath23 .
also , the ability of the system to analyze the cherenkov images in different projections where the images are only partially correlated , provides a superior rejection of hadronic showers . the preliminary results of observations of the crab nebula in october 96 by ct3/ct4/ct5
, operating in the 2-fold telescope coincidence mode , generally confirm this prediction . in fig .
[ system ] we present the on - source and off - source alpha distributions for the crab in the camera of , obtained after filtering the events through the ` standard ' image cuts of the showers detected by @xmath22 telescopes .
this preliminary result shows that the significance of detection of @xmath24 @xmath0-rays from the crab is achieved at a level of @xmath25 per 1 hour .
in fact , we expect further significant improvement of the signal using _ energy _ and _ distance _ dependent ` cuts ' to the detected images .
( note that the stereoscopic approach allows to determine the energy and the distance to the shower core with accuracies @xmath26 and @xmath27 , respectively ) .
furthermore , the obvious advantage of the joint operation of @xmath22 telescopes is the possibility to determine , without any _ a priory _ assumption , the arrival directions of _ individual _
@xmath0-ray primaries with accuracy @xmath28 .
for the case of point - like sources , this provides by itself , rejection of the cr background based only on the orientational characteristics of showers by a factor of at least several 100 .
the procedure obviously requires good understanding of the characteristics of all telescopes linked in the system .
although the work regarding the ct3/ct4/ct5 data set is still in progress , our results obtained earlier using the combination of telescopes ct1/ct2 @xcite , and telescopes ct2/ct3 allows us an optimistic view concerning the practical realization of this important aspect of the stereoscopic approach . indeed , despite the short time of simultaneous observations from the crab by ct-2 and ct-3 taken during the season 1995/1996 and a low coincidence rate ( due to the high energy threshold of ct-2 ) , the stereoscopic approach reveals a clear signal with 23 ` excess ' events against 4 ( ! ) ` background ' events within an angle @xmath29 around the direction to the crab ( see fig . [ stereo ] ) , the latter being in good agreement with the monte carlo predictions for the angular resolution of the ct2/ct3 system .
= 8 cm = 4 cm thus , our current understanding of the system based on the recent experimental results as well as on comprehensive monte carlo studies , allows us to make conclusive predictions for the performance of the system of the 5 hegra imaging telescopes , namely @xmath30 energy threshold around 500 gev ; @xmath30 angular and energy resolutions of about @xmath31 and @xmath26 , respectively ; @xmath30 minimum detectable fluxes of @xmath0-rays @xmath32 from point sources , and @xmath33 from moderately extended sources with an angular size @xmath34 . while the 5-iact system is nearing the final stage of its commission , we continue to use the available telescopes in different configurations not only for methodological purposes , but also for observations of a number of astronomical objects .
the first positive signal of @xmath0-rays was seen from the crab nebula in 1992 by ct-1 in its first version @xcite .
later we have detected tev @xmath0-rays from the crab with a significance of @xmath35 using ct-2 @xcite .
the current sensitivity of the system consisting of ct3/ct4/ct5 has achieved a level which allows to obtain a strong signal from the crab for just a few hours of observations under almost background - free conditions ( see fig . [ system ] ) .
thus , we can now use the high energy photon beam of this persistent source for calibration purposes and for an understanding of detector characteristics . over the last several years
ct-1 and ct-2 have been used for the search for tev @xmath0-rays from other objects , in particular cygnus x-3 , ae aquary , _ etc .
_ , reported earlier by different groups as possible episodic @xmath0-ray emitters ( see _ e.g. _ @xcite ) .
unfortunately , we did not find dc @xmath0-ray signals from any one of these sources .
however , the observations of markarian 421 in 1995 resulted in the independent detection of statistically significant positive signals by ct-1 and ct-2 @xcite , thus confirming the discovery by the whipple collaboration of tev emission from this bl lac object @xcite .
.[mrktab ] summary of mrk 501 and crab observations with ct-1 .
[ cols="<,<,<,<,<,<",options="header " , ] the first observations of the source in spring 1995 with the telescopes ct1 ( 13 h ) and ct2 ( 25 h ) in on / off mode showed a small evidence for a signal with statistical significance @xmath36 and @xmath37 , respectively .
however , the observations by the same telescopes after aug 15 with doubled exposition time did not reveal any indication of a signal .
since this might be a result of a highly sporadic behavior of grs 1915 + 105 , we decided to continue observations in 1996 . in 1996
grs 1915 + 105 has been observed during the period may - july 1996 by ct2 and ct3 .
ct2 with an energy threshold @xmath38 had been successfully used earlier for observations of the crab @xcite and mrk 421 @xcite .
the main characteristics of this telescope are studied by monte carlo simulations which reproduce quite well the distributions of different image parameters .
the rejection power of the telescope based on both _
shape _ and _ orientation _ parameters allows to suppress the cr background by a factor of 100 , the @xmath0-ray acceptance being at the level of @xmath39 .
the sensitivity of the telescope based on the observations of the crab nebula is estimated as @xmath40 detection rate of a signal from a point source at the flux level @xmath41 .
remarkably , this parameter has not been changed noticeably during the last 2 years which demonstrates the high stability of operation of the telescope .
since grs 1915 + 105 is a highly variable object , we observed the source in the tracking mode .
in addition , each month certain on / off pairs were taken in order to monitor the shape of the background `` @xmath42-distribution '' .
the statistics of observations at zenith angles @xmath43 and the results are summarized in table [ grstab ] .
the alpha - distribution of on - off source events shown in fig .
[ grs ] indicate the existence of a possible signal with significance @xmath44 .
the average rate of excess events is @xmath45 which corresponds to @xmath46 of the @xmath0-ray detection from the crab .
note that most of the `` excess '' events are due to the june period of observations which can be explained by @xmath47 of statistics contained in this period . at the same time ,
perhaps one can not exclude an intrinsic variability of the signal itself .
for example the statistical significance of the signal in june 19 alone is of about @xmath48 , with a rate exceeding the average rate by a factor of 3 .
this is seen in fig .
[ grsd ] where we present the daily rates of the `` excess '' events .
another feature of the time profile of the rate is its rather interesting coincidence with 3 prominent few - day - duration flares of grs 1915 + 105 detected in radio and x - rays ( see fig .
[ grsd ] ) .
this possible correlation in time of the rate of `` excess '' events may serve as an additional argument that strengthens the belief that the tev @xmath0-ray detection from this source is a real effect rather than a result of statistical fluctuation of the background .
if so , assuming that the radiation originates in the jet of grs 1915 + 105 approaching us , the intrinsic average tev luminosity of the source is estimated as @xmath49 .
= 8 cm = 4.3 cm however , we would like to emphasize that before going too far in the astrophysical interpretation of this statistically marginal signal , we obviously need independent confirmations of this result by other instruments and groups . as was mentioned before ,
the source was observed simultaneously also by the ct-3 telescope . at the perpignan symposium ,
a positive signal was also shown from this telescope .
further analysis of these data however has revealed evidence for hardware problems during this observation period in some of the camera pixels , resulting in large spurious signals , which may fake @xmath0-ray topologies .
work is in progress to correct these problems in software and to reprocess the data . at this point ,
no statements concerning grs 1915 + 105 signals in ct-3 data can be made .
in an analysis using monte carlo and experimental data it has been shown that the angular resolution of the airobicc array above a threshold of 18 tev for all events inside the array border is @xmath50 = 0.29 @xmath51 .
it decreases to 0.1 @xmath51 at higher energies .
the corresponding angular resolution of the scintillator array above 20 tev is @xmath50 = 0.9@xmath51 and reaches at higher energies 0.4 @xmath51 @xcite .
the numbers given refer to hadron - initiated showers of vertical incidence . for photon - initiated showers an angular resolution of @xmath50
= 0.22 @xmath51 is expected according to monte carlo calculations for the airobicc array .
a method to achieve a gamma - hadron separation has been presented by j. prahl .
the separation of gamma- and hadron - initiated showers is , besides a good angular resolution , crucial for a search of @xmath0-ray sources .
the method called pearl uses first the well know fact that electromagnetic showers decline after the shower maximum faster than hadronic showers .
therefore the ratio @xmath52/ of particles measured at ground @xmath53 and the number of cherenkov photons is a well known separation quantity @xcite .
the other newly introduced quantity in the method pearl is related to the larger transverse extent of hadronic showers as compared to electromagnetic showers due to the hadronic transverse momentum arising in hadronic interactions .
the quantity used for the transverse spread is defined by @xmath54 / @xmath55 whereby @xmath56 is the reconstructed core position , @xmath57 the position of the @xmath58 scintillator counter and @xmath59 the number of particles in the @xmath58 scintillator counter .
the separation quantity @xmath60 is related to the age parameter from a nkg fit . instead of using quantities obtained from a fit to the lateral distributions the method pearl uses the corresponding quantities derived from counting the amplitudes in the different counters i.e. the sum of the pulseheights in the scintillators n(e),the sum of the cherenkovlight measured in the airobicc array n(l ) , and the quantity @xmath61 sensitive to the lateral extent defined above .
the advantage of these quantities is that showers at the trigger threshold can be analysed without requiring a fit to converge .
( studies have shown that for showers for which the lateral fits converge the results are the same in the energy range of interest ) . in fig .
[ pearl ] the two separation quantities are plotted against each other for photon and hadron initiated showers .
a clear distinction can be seen and a straight line ( pearl cut ) can be used to separate photon- and hadron initiated shower in the 2-dim logarithmic plot . the q factor defined as / obtained for this method varies from q=1.5 at threshold of 18 tev to q=2.5 at 50 tev . with the gamma - hadron separation method pearl described above , based on the scintillator and airobicc arrays , a search for galactic sources has been performed using the 94/95 data . amongst 13 selected sources the most significant excess of 3.3 @xmath62 after gamma - hadron separation is observed for her x-1 .
no excess has been observed for the crab nebula in the 93/95 data whereas in the 1992/93 data a 3.6 @xmath62 effect has been reported ( @xmath63 icrc rom 1995 , unpublished ) . the results for the crab nebula are summarized in fig .
[ crab ] . up to now
two extragalactic tev @xmath0 sources mrk 421 and mrk 501 have been observed with the technique of imaging air cherenkov telescopes ( iact ) up to @xmath64 10 tev .
whereas with the iact only one source at a time can be observed , the arrays allow a simultaneous observation of several sources in the field of view of hegra , however for the prize of a higher energy threshold .
this may severely limit the chance of an observation due to the absorption of tev photons on the not well known infrared - optical background fields ( @xmath0-ray horizon ) . in order to perform a search a list of 11 source candidates with redshifts @xmath64
0.1 has been set up , belonging to the class of blazars visible in the field of view of hegra @xcite .
the sources have been looked at in the stacking mode and individually .
the search has been performed for datasets with and without gamma - hadron separation capability .
the gamma - hadron separation with the geiger towers using a neural net analysis ( q - factors 2 - 2.5 ) @xcite requires a threshold of 50 tev and an angular bin of 1 degree ( using the direction of the scintillator array ) therefore the results from other data sets are for comparison also evaluated for 50 tev and 1 degree .
stacking all source candidates the highest significances reported by n. magnussen are observed in the scintillator data from 6.89 -6.92 without gamma - hadron separation ( 3.5 @xmath62 ) and from 11.94 - 3.95 with gamma - hadron separation using the geiger towers the most prominent single source candidate is 0116 + 319 which contributes with in the scintillator data .
no excess has been observed for a sample of blazars with higher redshifts and for a sample of 49 galaxies characterized by steep radio spectra . as reported by j. prahl
the scintillator data from 1994/95 show an excess of 1.2 @xmath62 and the airobicc data from 1994/95 with @xmath0-hadron separation exploiting the much better angular resolution and lower energy thresholds possible for the airobicc data , the significances do not increase as expected but decrease . in order to compare the results of the different
data sets the excesses with the highest significance are converted tentatively into a flux and the other results in upper limits ( fig .
[ agn ] ) . a definite conclusion can not yet be drawn and more systematic checks are under way especially concerning the absolute pointing which is crucial in order to exploit the good angular resolution of the airobicc array . also further airobicc data from 1992/93 are investigated .
if a signal could be established this would imply , that the infrared background radiation has to be much less than presently assumed and the sources according to present day models would probably be proton accelerators . [ [ section ] ] gamma - ray - bursts ( grb s )
are known for more than 25 years yet their physical origin and distance scale are still unknown .
observations of tev counterparts would impose limits on the unknown distance scale because of the absorption at the infrared intergalactic background fields which limits the mean free path of tev photons to distances@xmath64 100 mpc .
to search for tev emission associated with gamma - ray bursts registered by the batse instrument all hegra detector components have been used .
especially searches with the large field of view of the air shower arrays have been performed for tev emission which may occur simultaneous , preceding or following a burst .
a delayed emission may be expected if grb s account for the acceleration of protons up to the highest energies which in turn generate a secondary intergalactic electromagnetic cascade due to interactions with the microwave background fields . no significant signal has been observed for the different searches as reported by b. funk and h. krawczynski .
the results from a search for coincident emission obtained for 31 batse bursts are shown in fig .
[ grb ] . to guide the eye an extrapolated grb spectrum
@xcite is displayed . in future
a lower energy threshold of about 2 tev will be possible due to higher data taking rate and an angle sensitive trigger .
a search for anisotropies has been reported by d. schmele using the data taken during 1994/95 with the scintillator array ( 60 mio events ) , with the airobicc array ( 50 mio ) and in the data with the airobicc array using the @xmath0-hadron separation method pearl . in order to search for anisotropies ,
all data ( for each data set separately ) have been used to determine a two dimensional shape of the background distribution in local coordinates @xmath65 and @xmath66 .
this shape of the background distribution is corrected for the air pressure by a multiplicative term depending on @xmath66 and air pressure . using this ( pressure corrected ) background shape in local coordinates a background map in equatorial coordinates
is generated according to the actual event rates recorded . in order to obtain the actual event rates in this procedure ,
the experimental event rates have first been corrected for the ( variable ) deadtime and air pressure . in this way the actual up time and performance of the array is taken into account .
the deviations of the data from the background ( for bin sizes corresponding to the angular resolution ) follow a normal distribution n(0,1 ) rather well . for binsizes from @xmath67 to @xmath68 an upper limit of @xmath69 for the smallest bin and @xmath70 for the largest bin
is obtained .
these values are essentially determined by the statistics of @xmath71 50mio events per event sample .
= 8.0 cm in order to look for an emission from the galactic disk , the maps are transformed into galactic coordinates and projected onto galactic latitudes and compared .
a disk of width @xmath72 has been assumed ( l = 30@xmath73 - 220@xmath73 ) .
the results for the upper limits for a flux from the galactic disk for the three data sets are shown in fig.[disk ] .
also shown are the limits of casa mia @xcite , the expected theoretical flux of @xmath74 for diffuse emission from interaction of cosmic rays with interstellar matter @xmath75 @xcite and a naive egret extrapolation with a constant slope of @xmath76 .
the hegra limits of @xmath77 can be interpreted as limits for an additional flux due to unresolved galactic sources in the energy range above 20 tev .
finally a search for a contribution from the region around the direction of the ehe events as seen by fly s eye and yakutsk has been performed . a contribution in the tev energy range
may be expected via the intergalactic cascade ( @xmath78 ) . on the scale from @xmath79
to @xmath80 radius around the fly s eye event position ( ra=85.2 , dec=48.0 ) @xcite , no excess has been found .
an upper limit for the contribution of an ehe source @xmath81 above 20 tev can be given .
this limit includes a possible contribution from the galactic disk , which is near to the estimated source position .
all searches are carried out with the three data samples described above and all three agree well within their different sensitivities and statistics . in the near future
the size of the data sample will will be increased substantially .
the measurement of the energy and composition of primary cosmic rays in the energy range of the knee is only accessible to the indirect methods of air showers and is still controversial and a matter of debate . here
a new method is presented to reconstruct the energy and masses of primary particles with energies above 300 tev from the electromagnetic component of an air shower as measured by a scintillator array and an array of cherenkov counters .
the main steps of the reconstruction procedure as presented by h. krawczynski , can be summarized as follows : \(1 ) the _ distance to the shower maximum @xmath82 _ is reconstructed from the lateral distribution of the cherenkov light as measured at detection level .
it can be shown by monte carlo simulation that @xmath83 can be determined independent of the primary mass ( and almost independent of energy ) .
this is due to the fact that the shower development behind the shower maximum , which dominates the determination of the distance to the maximum , is independent of the primary mass of nuclei .
\(2 ) _ electromagnetic energy @xmath84 . _ from the number of electrons @xmath53 as obtained from a nkg fit and the distance to the showermaximum @xmath85 the number of electrons in the maximum @xmath86 can be inferred .
this again is independent of the the particle mass .
the number of electrons in the maximum is in a good approximation proportional to the electromagnetic energy @xmath84 .
\(3 ) _ energy per nucleon e / a . _ from the distance to the showermaximum the penetration depth is obtained .
since the penetration depth depends ( logarithmically ) on e / a , the energy per nucleon e / a can be ( coarsely ) determined .
\(4 ) _ primary energy e. _ the fraction of the total energy which is transformed into electromagnetic energy increases in a known way ( i.e. from monte carlo simulations ) with the energy / nuleon : . with e / a from step ( 3 ) and e@xmath87 from step ( 2 )
the total energy e can be determined .
\(5 ) _ primary nucleon number a. _ with e and e / a known from step ( 3 ) and ( 4 ) also the nucleon number a is coarsely known .
more details of the method can be found in @xcite . in order to first test the new method ,
it is applied to the energy range from 500 tev to 1 pev where the composition is known or can be extrapolated from direct measurements taken with satellites and balloone - borne experiment .
the preliminary results shown in fig .
[ compos ] are in good agreement with the results from direct measurements . for the differential energy spectrum a spectral index of [email protected]@xmath880.1
is obtained in the energy range from 500 tev to 2 pev .
the abundance of light nuclei to all nuclei is obtained to be [email protected]@xmath880.10 in the energy range from 500 tev to 1 pev .
the distributions in fig.[compos ] reflect the fluctuations in the penetration depth for the different primaries .
it should be stressed that for the mass determination the mean values as well as width ( i.e. fluctuations ) of the curves contribute .
results with a slightly different method have been reported at @xmath89 icrc rom @xcite . in the near future
the new method is applied to the energy range above 1 pev ( knee ) .
the hegra collaboration thanks the instituto de astrofisica de canarias for the use of the hegra site at the roque de los muchachos and its facilities .
this work was supported by the bmbf , the dfg and the cicyt .
99 a.karle et al . , astroparticle physics 3 ( 1995 ) 321 w. rhode et al . , nim a378 ( 1996 ) 399 f.a.aharonian et al . , proc .
`` towards a major atmospheric cherenkov detector - ii '' , calgary ( 1993 ) 81 .
m.panter et al . , proc .
24th icrc , rome 1 ( 1995 ) 958 .
g.hermann , proc .
`` towards a major atmospheric cherenkov detector - iv '' , padova ( 1995 ) 81 .
r.mirzoyan et al . , nim a351 ( 1994 ) 51 .
g.rauterberg et al . , proc .
24th icrc , rome 3 ( 1995 ) 460 .
a.kohnle et al . , astroparticle physics 5 ( 1966 ) 119 .
a. konopelko et al . , astroparticle physics 4 ( 1996 ) 199 .
t.c.weekes , space sci .
59 ( 1992 ) 315 .
d.petry et al . , astron .
astrophys . 311
( 1996 ) l13 .
m.punch et al . ,
nature 358 ( 1992 ) 477 .
von montigny c. et al .
, ap.j . 440 ( 1996 ) 525 .
j.quinn et al . , ap.j .
456 ( 1996 ) l83 .
s.bradbury et al . , astron .
astrophys . , submitted .
i.f.mirabel and l.f.rodriguez , nature 371 ( 1994 ) 46 .
h. krawczynski et al . , nim a ( 1996 ) in press f. arqueros et al .
, 24th icrc rom k. mannheim et al . , astron . & astrophys . 315
( 1996 ) 77 s. westerhoff et al .
, astroparticle physics 4 ( 1995 ) 119 .
m. sommer et al . , astrophys .
j. 422 ( 1994 ) l63 a. borione et al . , 24th icrc rom 2 ( 1995 ) 463 f.a .
aharonian and a.m. atoyan , astron .
astrophys 917a ( 1996 ) 309 bird , d.j .
et al . , astrophys .
j. 441 ( 1995 ) 114 nagano , m. and takahara , f. ; astrophysical aspects of most energetic cosmic rays ; world scientific , singapore ; ( 1991 ) a. lindner astro - ph/9611175 publication in preparation r. plaga et al.,24th icrc rom 2 ( 1995 ) 693 | a report is given on the status and recent results from the hegra imaging atmospheric cherenkov telescopes ( iact ) concerning mrk 501 and grs 1915 + 105 and on recent results from the hegra arrays concerning galactic and extragalactic @xmath0-sources , counterparts of tev gamma - ray - bursts , cosmic ray anisotropies , energy spectrum and composition of primary cosmic rays . the report summarizes the presentations given at the 15@xmath1 european cosmic ray symposium 1996 . |
recent analysis of fermi gamma ray space telescope data has shown an excess of gamma ray from the galactic center ( gc ) with a feature similar to annihilating dark matter @xcite ( for a review of dark matter , please see @xcite ) .
previous studies @xcite also identified a similar excess of @xmath0 gev gamma rays from the region surrounding the gc . according to the analysis presented in @xcite , the fermi telescope signal of gamma ray excess in the gc can very well be fit by a @xmath1 gev dark matter particle annihilating into @xmath2 pairs with an annihilation cross section of @xmath10 , normalized to a local dark matter density of @xmath11 .
the required annihilation cross section is coincidentally very close to the annihilation cross section of typical weakly interacting massive particle ( wimp ) dark matter candidate in order to produce the correct dark matter relic abundance observed by the planck experiment @xcite @xmath12 where @xmath13 is the density parameter and @xmath14 is a parameter of order unity .
several interesting particle physics models have already been proposed @xcite which explain the gc excess of gamma rays . here
we study the possibility of providing such an explanation within the framework of an abelian extension of standard model , originally proposed by @xcite and later studied in the context of dark matter and ev scale sterile neutrino in @xcite and @xcite respectively .
the salient feature of the model is the way it relates dark matter with neutrino mass where neutrino masses arise at one loop level with dark matter particles running inside the loops : more popularly known as `` scotogenic '' model @xcite .
the additional abelian gauge symmetry @xmath4 and the corresponding gauge charges for the fields are chosen in such a way that it gives rise to a remnant @xmath7 symmetry so that the lightest @xmath7-odd particle is stable and hence can be a cold dark matter candidate . as studied in details in @xcite , this model has several dark matter candidates namely , fermion singlet , fermion triplet , scalar singlet and scalar doublet .
scalar dark matter phenomenology is similar to the higgs portal models discussed extensively in the literature . in these scenarios , the scalar dark matter annihilates into the standard model ( sm ) particles through the higgs boson .
co - annihilations through gauge bosons can also play a role if the cp even and cp odd components of the neutral higgs have a tiny mass difference as discussed recently in @xcite . in the context of gc gamma ray excess ,
several higgs portal models have already been studied and there exists at least one neutral higgs lighter than the sm higgs which acts as a mediator between scalar dark matter and the sm particles .
the mass of this light neutral higgs is approximately equal to twice the scalar dark matter mass in order to satisfy experimental bounds on relic density as well as direct detection experiments . instead of pursuing higgs portal like scalar dark matter scenarios in the model , we study the fermionic dark matter sector . since the neutral component of fermion triplet needs to be very heavy ( @xmath15 ) in order to reproduce correct dark matter relic density @xcite
, we confine our discussion to fermion singlet dark matter in this work .
that is , we explore the possibility of fermion singlet dark matter in this model with mass around 30 gev which can simultaneously give rise to gc gamma ray excess as well satisfy dark matter bounds on relic density as well as direct detection cross section .
such a light fermion singlet dark matter particle will self - annihilate through the abelian vector boson @xmath16 into sm particles .
we also incorporate the collider constraints on such additional vector boson and its gauge couplings .
we find that , although the relic density and direct detection constraints allow a significant region of the parameter space , the collider constraints reduce the parameter space into the s - wave resonance region where the gauge boson mass is approximately twice that of dark matter mass . finally , we check whether such a light fermion singlet dark matter is compatible with neutrino mass which arise at one loop level . this letter is organized as follows : in section [ model ] , we briefly discuss the model . in section [ sec : darkmatter ] , we discuss the singlet fermion dark matter as a source of gc gamma ray excess taking into account all necessary experimental constraints . in section [ neutrino ]
, we discuss the compatibility of light singlet fermion dark matter with neutrino mass and finally conclude in section [ conclude ] .
the model which we take as a starting point of our discussion was first proposed in @xcite .
the authors in that paper discussed various possible scenarios with different combinations of majorana singlet fermions @xmath5 and majorana triplet fermions @xmath17 . here
we discuss one of such models which we find the most interesting for our purposes .
this , so called model c by the authors in @xcite , has the the following particle content shown in table [ table1 ] .
.particle content of the model [ cols="^,^,^,^",options="header " , ] xenon100 experiment gives the lowest upper bound on spin dependent cross section as @xmath18 for a wimp mass of 45 gev at @xmath19 confidence level .
here we take this conservative upper bound for both 25 gev and 35 gev dark matter analysis and draw the exclusion line . as can be seen from figure [ fig1 ] , [ fig2 ] , [ fig3 ] and [ fig4 ]
, the black solid line corresponds to this direct detection bound such that the parameter space towards the left of this line is ruled out . to apply the collider bounds on @xmath20 and @xmath21 we follow the analysis of @xcite which studies the scenario of a new heavy abelian gauge boson coupling to dark matter as well as sm fermions in the light of collider and dark matter direct detection data . as discussed by the authors of @xcite , the large hadron collider ( lhc ) bounds on abelian vector boson
coupled to sm , which is approximately @xmath22 tev , can be relaxed if @xmath16 has non - negligible couplings to dark matter .
they showed that for @xmath16 decaying into sm particles with branching ratio @xmath19 and @xmath23 , the lowest allowed value of @xmath24 is approximately @xmath25 tev .
this limit gets pushed up to 4 tev and @xmath26 tev , if @xmath21 is increased to @xmath27 and weak gauge coupling @xmath28 respectively . to implement these bounds
we compute the branching ratio of @xmath16 decaying into dark matter particles and plot them as a function of @xmath24 in figure [ fig5 ] .
it can be seen that the maximum branching ratio is around @xmath29 .
we then apply three different lower limits on @xmath24 namely , 2 tev , @xmath30 tev and 3 tev and check how much of the parameter space remains .
these limits can be seen in figure [ fig1 ] , [ fig2 ] , [ fig3 ] and [ fig4 ] in three different colors such that , the parameter space towards the left of that region is ruled out . it can be seen that even if we take a conservative bound @xmath31 tev , then also we have available parameter space which satisfies all other dark matter constraints .
it should also be noted that near the resonance region in the @xmath32 as well as @xmath33 planes , the values of @xmath21 is much below @xmath34 for which the collider bound is @xmath35 tev as mentioned above .
thus the bound @xmath24 is supposed to get further relaxed as we go below @xmath36 resulting in more and more allowed parameter space .
we note that the parameter space shown in the @xmath32 plane has regions where mass of dark matter is larger than @xmath20 allowing the possibility of dark matter annihilation into two @xmath16 bosons .
this particular case however , corresponds to the region towards the left side of the plot where @xmath37 gev . in our model
, such annihilation of dark matter into two @xmath16 bosons can occur through t - channel exhange of the majorana fermion singlet dark matter itself .
since , the allowed parameter space after incorporating all the constraints correspond to the s - wave resonance region where @xmath38 , such annihilation of dark matter into two @xmath16 bosons will not alter the allowed parameter space and hence we have not included this process in our calculations .
as discussed in details in @xcite , the sm light neutrino mass can arise at one loop level in this model as can be seen in figure [ fig6 ] . at tree level only one of the neutrinos acquires non - zero mass from usual type i seesaw mechanism @xcite where the singlet fermion @xmath39 acts as the heavy right handed neutrino . writing the yukawa lagrangian for our model as @xmath40 @xmath41 the tree level light neutrino mass can be written as @xmath42 from figure [ fig2 ] and [ fig4 ]
, we see that the allowed region from dark matter as well as collider constraints suggest @xmath43 tev .
since @xmath44 gev , for neutrino mass to be of ev scale , the equation ( [ neutmass ] ) suggest that the yukawa couplings @xmath45 have to be fine tuned to @xmath46 which is approximately same as the electron yukawa coupling in the sm .
the other two sm neutrinos can acquire non - zero masses only when loop contributions in figure [ fig6 ] are taken into account . as discussed in @xcite ,
the one - loop contribution @xmath47 to neutrino mass is given by @xmath48 assuming all the scalar masses in the loop diagram to be almost degenerate and written as @xmath49 then @xmath50 , \label{scaldeg}\end{aligned}\ ] ] where @xmath51 . for fermion singlet
light dark matter , @xmath52 and hence the above expression can be approximated as @xmath53 the one - loop neutrino mass can be written as @xmath54 taking @xmath55 to be at few tev s , @xmath56 at electroweak scale and the singlet mass @xmath57 at few tens of gev ( for light fermion singlet dark matter ) , the above expression can give rise to ev scale neutrino mass if @xmath58 which can be achieved if each of the dimensionless couplings is tuned to be around @xmath59 .
thus , a light fermion singlet dark matter of mass 25 gev or 35 gev is consistent with the requirement of ev scale sm neutrino masses .
we have studied fermion singlet dark matter in the light of recently observed galactic center gamma ray excess within the framework of an abelian extension of standard model .
the model not only gives rise to a stable dark matter candidate , but also gives rise to tiny neutrino masses both at tree level as well as one - loop level .
we take two different dark matter masses @xmath60 and check whether they can give rise to the desired annihilation cross section in order to satisfy dark matter relic density constraint as well as annihilation into @xmath2 pairs to explain the galactic center gamma ray excess .
we also take into account the constraints from dark matter direct detection experiments on spin dependent scattering cross section of dark matter off nuclei .
since the annihilation and scattering of light fermion singlet dark matter is mediated by the abelian vector boson @xmath16 , these scenarios can also be constrained from lhc limits on additional gauge boson mass @xmath20 and its coupling @xmath21 . without performing a detailed calculation for collider signatures
, we use the results from @xcite where the authors found the lower bound on @xmath24 to be @xmath25 tev for @xmath61 and @xmath23 . in the present work ,
we find the maximum branching ratio of @xmath16 boson into sm particles to be approximately @xmath19 .
we find that , even after applying a conservative lower limit on @xmath24 as @xmath62 tev , we still have parameter space which can satisfy all dark matter constraints .
the allowed parameter space is essentially an s - wave resonance region where mass of the @xmath16 boson is twice that of dark matter mass . since the allowed region of parameter space is limited , these scenarios can be further constrained or even ruled out by future data from dark matter direct detection as well as collider experiments .
db would like to thank the organizers of the workshop `` frontiers of physics : colliders and beyond '' during 23 - 27 june , 2014 at ictp , trieste where this work was started .
ad likes to thank council of scientific and industrial research , govt .
of india for financial support through senior research fellowship ( emr no . 09/466(0125)/2010-emr - i ) .
t. daylan , d. p. finkbeiner , d. hooper , t. linden , s. k. n. portillo , n. l. rodd and t. r. slatyer , arxiv:1402.6703 .
, , , * * , ( ) , .
l. goodenough and d. hooper , arxiv:0910.2998 ; a. boyarsky , d. malyshev and o. ruchayskiy , phys . lett . * b705 * , 165 ( 2011 ) ; d. hooper and t. linden , phys . rev . *
d84 * , 123005 ( 2011 ) ; k. n. abazajian and m. kaplinghat , phys . rev . *
d86 * , 083511 ( 2012 ) ; c. gordon and o. macias , phys . rev . *
d88 * , 083521 ( 2013 ) ; d. hooper and t. r. slatyer , phys .
dark . univ .
* 2 * , 118 ( 2013 ) ; k. n. abazajian , n. canac , s. horiuchi and m. kaplinghat , phys .
rev . * d90 * , 023526 ( 2014 ) .
p. a. r. ade _ et al .
_ [ planck collaboration ] , arxiv:1303.5076 .
a. berlin , d. hooper and s. d. mcdermott , phys .
rev . * d89 * , 115022 ( 2014 ) .
j. d. ruiz - alvarez , c. a. de s. pires , f. s. queiroz , d. restrepo and p. s. r. da silva , phys . rev . *
d86 * , 075011 ( 2012 ) ; w .- c .
huang , a. urbano and w. xue , jcap * 1404 * , 020 ( 2014 ) ; k. p. modak , d. majumdar and s. rakshit , arxiv:1312.7488 ; n. okada and o. seto , phys . rev . * d89 * , 043525 ( 2014 ) ; k. hagiwara , s. mukhopadhyay and j. nakamura , phys . rev . * d89 * , 015023 ( 2014 ) ; c. boehm , m. j. dolan , c. mccabe , m. spannowsky and c. j. wallace , jcap * 1405 * , 009 ( 2014 ) ; d. k. ghosh , s. mondal and i. saha , arxiv:1405.0206 ; a. martin , j. shelton and j. unwin , arxiv:1405.0272 ; p. agrawal , b. batell , d. hooper and t. lin , arxiv:1404.1373 ; d. g. cerdeno , m. peiro and s. robles , jcap * 1408 * , 005 ( 2014 ) ; p. ko , w .- ii . park and y. tang , arxiv:1404.5257 ; s. ipek , d. mckeen and a. e. nelson , arxiv:1404.3716 ; k. kong and j .- c .
park , arxiv:1404.3741 ; b. kyae and j .- c . park , phys .
b732 * , 373 ( 2014 ) ; l. a. anchordoqui and b. j. vlcek , phys . rev . *
d88 * , 043513 ( 2013 ) ; e. izaguirre , g. krnjaic and b. shuve , arxiv:1404.2018 ; m. abdullah , a. difranzo , a. rajaraman , t. m. p. tait , p. tanedo and a. m. wijangco , phys . rev . *
d90 * , 035004 ( 2014 ) ; c. boehm , m. j. dolan and c. mccabe , phys . rev . * d90 * , 023531 ( 2014 ) ; a. alves , s. profumo , f. s. queiroz and w. shepherd , arxiv:1403.5027 ; d. hooper , c. kelso and f. s. queiroz , astropart . phys . * 46 * , 55 ( 2013 ) ; t. basak and tanmoy mondal , arxiv:1405.4877 ; a. berlin , p. gratia , d. hooper and s. d. mcdermott , phys . rev .
* d90 * , 015032 ( 2014 ) ; j. m. cline , g. dupuis , z. liu and w. xue , arxiv:1405.7691 ; l. wang , arxiv:1406.3598 ; c. arina , e. d. nobile and p. panci , arxiv:1406.5542 ; c. cheung , m. papucci , d. sanford , n. r. shah and k. m. zurek , arxiv:1406.6372 ; j. huang , t. liu , l .- t .
wang and f. yu , arxiv:1407.0038 ; c. balazs and t. li , arxiv:1407.0174 ; p. ko and y. tang , arxiv:1407.5492 ; s. baek , p. ko and w .- ii park , arxiv:1407.6588 ; n. okada and o. seto , arxiv:1408.2583 ; a. d. banik and d. majumdar , arxiv:1408.5795 ; e. hardy , r. lasenby and j. unwin , jhep * 1407 * , 049 ( 2014 ) ; j. beringer et al .
, phys . rev .
* d86 * , 010001 ( 2012 ) .
g. arcadi , y. mambrini , m. h. g. tytgat and b. zaldivar , jhep * 1403 * , 134 ( 2014 ) .
p. minkowski , phys . lett .
* b67 * , 421 ( 1977 ) ; m. gell - mann , p. ramond , and r. slansky ( 1980 ) , print-80 - 0576 ( cern ) ; t. yanagida ( 1979 ) , in proceedings of the workshop on the baryon number of the universe and unified theories , tsukuba , japan , 13 - 14 feb 1979 ; r. n. mohapatra and g. senjanovic , phys .
lett * 44 * , 912 ( 1980 ) ; j. schechter and j. w. f. valle , phys . rev . *
d22 * , 2227 ( 1980 ) . | the fermi gamma ray space telescope data have pointed towards an excess of gamma rays with a peak around @xmath0 gev in the region surrounding the galactic center .
this anomalous excess can be described well by a dark matter candidate having mass in the range @xmath1 gev annihilating into @xmath2 pairs with a cross section of @xmath3 . in this work
we explore the possibility of having such a dark matter candidate within the framework of a radiative neutrino mass model .
the model is a simple extension of the standard model by an additional @xmath4 gauge symmetry where the standard model neutrino masses arise both at tree level as well as radiatively by the anomaly free addition of one singlet fermion @xmath5 and two triplet fermions @xmath6 with suitable higgs scalars .
the spontaneous gauge symmetry breaking is achieved in such a way which results in a residual @xmath7 symmetry and hence providing a stable cold dark matter candidate .
we show that the singlet fermionic dark matter candidate in our model can give rise to the galactic center gamma ray excess .
the parameter space which simultaneously satisfy the constraints on relic density , direct detection scattering as well as collider bounds essentially corresponds to an s - wave resonance where the gauge boson mass @xmath8 is approximately twice that of dark matter mass @xmath9 .
we also discuss the compatibility of such a light fermion singlet dark matter with light neutrino mass . |
20 vysotskii s l , nikitov s a and filimonov yu a 2005 _ j. exp .
phys . _ * 101 * 636 chumak a v , serga a a , hillebrands b and kostylev m p 2008 _ appl .
lett . _ * 93 * 022508 chumak a v , serga a a , wolff s , hillebrands b and kostylev m p 2009 _ j. appl .
phys . _ * 105 * 083906 krawczyk m and puszkarski h 2008 _ phys . rev .
b _ * 77 * 054437 chumak a v , serga a a , wolff s , hillebrands b and kostylev m p 2009 _ appl .
lett . _ * 94 * 172511 ustinov a b , grigoreva n yu and kalinikos b a 2009 _ jetp lett .
_ textbf88 31 chumak a v , neumann t , serga a a , hillebrands b and kostylev m p 2009 _ j. phys .
phys . _ * 42 * 205005 gubbiotti g , tacchi s , carlotti g , singh n , goolaup s , adeyeye a o and kostylev m p 2007 _ appl .
* 90 * 092503 kostylev m p , schrader p , stamps r l , gubbiotti g , carlotti g , adeyeye a o , goolaup s and singh n 2008 _ appl .
_ * 92 * 132504 wang z k , zhang v l , lim h s , ng s c , kuok m h , jain s and adeyeye a o 2009 _ appl .
* 94 * 083112 chumak a v , tiberkevich v s , karenowska a d , serga a a , gregg j f , slavin a n and hillebrands b 2010 _ nat .
commun . _ * 94 * 141 dvornik m and kruglyak v v 2011 _ phys .
b _ * 84 * 140405 lee k s , han d s and kim s k 2009 _ phys . rev .
lett . _ * 102 * 127202 kim s k , lee k s and han d s 2009 _ appl .
lett . _ * 95 * 082507 kruglyak v v , hicken r j , kuchko a b and gorobets v y 2005 _ j. appl . phys . _ * 98 * 014304 chumak a v , pirro p , serga a a , kostylev m p , stamps r l , schultheiss h , vogt k , hermsdoerfer s j , laegel b , beck p a and hillebrands b 2009 _ appl . phys .
_ * 95 * 262508 chi k h , zhu y , mao r w , nikitov s a , gulyaev y v and tsai c s 2011 _ ieee trans .
magn . _ * 47 * 3708 gurevich a g and melkov g a 1996 _ magnetization oscillations and waves _
( new york : crc press ) serga a a , chumak a v and hillebrands b 2010 _ j. phys .
phys . _ * 43 * 264002 karenowska a d , gregg j f , tiberkevich v s , slavin a n , chumak a v , serga a a and hillebrands b 2012 _ phys
_ * 108 * 015505 demidov v e , demokritov s o , rott k , krzysteczko p and reiss g 2008 a__ppl .
_ _ * 92 * 232503 to note that the spin - wave intensity was recorded in the experiment after 8 structure periods while in the simulation after 4 periods .
hence , one may observe a difference in the depth and the width of the first rejection band between the experimental data and simulation , and also a slight shift in frequency . | the potential to control the number of the spin - wave band gaps of a magnonic crystal ( mc ) by variation of its geometry is investigated by numerical simulations .
the magnonic crystal is represented by a micro - sized planar ferromagnetic waveguide with periodically variable width . by choosing a step - like or sinusoidal variation of the width ,
the magnonic crystal reveals multiple or single band gaps , respectively .
this allows for additional degrees of freedom in the design of mc - based microwave filters and phase shifters with desired characteristics .
the mcs band gaps have been studied in the space and frequency domains exploring the spin - wave spectrum dependence on the probing position inside the magnonic crystal .
the linear and nonlinear magnetization dynamics in magnonic crystals ( mcs ) @xcite attracts increasing attention due to the potential application of these devices for signal processing in the microwave frequency range @xcite .
significant progress has recently been made in the development of magnonic crystals based on yttrium - iron - garnet ( yig ) ferrite films @xcite having extremely small spin - wave damping @xcite .
the attention has been focused on two types of yig based magnonic crystals : ( 1 ) crystals designed as spin - wave waveguides of periodically variable thickness @xcite and ( 2 ) crystals formed by placing of a plane spin - wave waveguide in a bias magnetic field spatially modulated by electrical currents @xcite .
grooved structures represent probably the simplest design of an one - dimensional magnonic crystal , while the current controlled mcs propose such unique features as gradual tuning and fast manipulation of crystal properties @xcite .
the transmission characteristics of these two types of mcs are markedly different : whereas a current - controlled mc has only one rejection - band ( i.e. the frequency region where spin waves are not allowed to propagate ) , a grooved mc possesses multiple rejection - bands in its transmission band .
this difference is clarified if one takes into account that the reflection coefficient of a wave with wavenumber @xmath0 is proportional to the @xmath1 fourier component of the magnetic periodic structure @xcite .
the spatial variations of the magnetic properties of the grooved crystals can be described by a _ step - like function _ expressing the change of the yig film thickness with a quasi - rectangular profile @xcite .
the current - controlled crystal is designed in such a way that variations of the bias magnetic field are practically _ sinusoidal _ @xcite .
since the spectrum of the latter crystal contains only one space component , the single rejection - band appears solely .
thus , by choosing a profile of a magnetic non - uniformity in the yig film one can control not only the width and depth of the rejection bands @xcite but also their number .
it allows , for example , creating of single- and multiple - band microwave filters .
however , yig - based structures do not satisfy the demands of modern microwave technology mainly due to their comparatively large sizes .
therefore , it opens up the necessity to design planar mcs of micron and sub - micron size with desired band gaps characteristics .
here we propose a method to control the transmission band gap characteristics of a microstructured magnonic crystal .
using micromagnetic simulations @xcite we analyze and compare two magnonic crystals designed as permalloy ( ni@xmath2fe@xmath3 , py ) spin - wave waveguides of periodically variable width .
the structures are planar and can easily be integrated on a chip within a cmos process .
the width of one waveguide is according to a step function ( that forms rectangular notches ) , while the width of another structure follows a sinusoidal function .
we show that by substituting the sharp notches with a harmonic edge profile one can change the mc operational transmission characteristics from a multiple rejection band state to a single band state .
the role of spin - wave channeling , spin - wave wavelength transformation , and geometrical scattering in the formation of the crystal characteristics is discussed .
( color online ) sketch of the simulated structures : mcs with rectangular a ) and sinusoidal notches b ) .
c ) and d ) distribution of the magnetic internal field ( @xmath4 component ) for the simulated structures .
e ) and f ) @xmath4 component of the internal field extracted along the magnetic stripes in the center ( y = 0 , white lines from ( a ) and ( b ) ) at half the distance between the center and the edge of the notches ( y = 375 nm , dashed line ) , and at y = 500 nm ( dotted blue line).,width=9 ] the two types of magnonic crystals under investigation are presented in fig . [ fig1]a and fig .
[ fig1]b .
they consist of 7.25 @xmath5m - long permalloy waveguides of 40 nm thickness .
the width of both mcs varies between @xmath6 m and @xmath7 m with a period @xmath8 m .
the geometry was chosen in such a way that the structure with rectangular notches is similar to the mc studied experimentally in ref .
@xcite .
the standard material parameters of py used to simulate the dynamics of the magnetization are : saturation magnetization @xmath9 t , exchange stiffness constant @xmath10 j / m , and the magnetocrystalline anisotropy is assumed to be zero . the damping constant in the landau - lifshitz - gilbert equation
is considered to be @xmath11 . in order to avoid spin - wave reflection at the ends of the structure ( @xmath12 and @xmath13 m )
the damping boundary conditions were used : in the boundary areas ( @xmath14 m on each side ) the parameter @xmath15 gradually increases more than fifty times resulting in a strong decay of the spin - wave intensity towards the boundaries . in order to excite spin waves within a wide frequency range
we apply a 70 ps - long video pulse to a 300 nm wide microstrip antenna ( see fig .
[ fig1 ] ) . in this case
the upper frequency limit given by the pulse duration is @xmath1613.5 ghz while the antenna can excite spin - waves with wavenumbers up to @xmath17 rad/@xmath5 m .
the static biasing magnetic field of 500 oe is applied in the film plane transverse to its long axis in order to satisfy the conditions for the excitation and propagation of the magnetostatic surface spin waves ( mssws ) .
these waves are known to have the highest values of the group velocity in magnetic microstructures and ensure spin - wave free paths of the order of several microns in py . for the understanding of the spin - wave excitation spectra in the magnonic crystals
one need to take into account the distribution of the internal field shown in fig .
[ fig1]c and fig .
[ fig1]d for both mcs configurations .
one can see that the magnetic field is non - uniform along both the long and short axis of the crystals .
the field variation along the waveguides is not of crucial importance since it changes mostly near the ends of the waveguides ( see the longitudinal field profiles along shown in fig .
[ fig1]e and fig .
[ fig1]f ) where the spin - wave damping is artificially increased and no spin - wave propagation occurs .
in contrast , the transverse field variation is of high importance : even a small deviation from the longitudinal axis of @xmath18 @xmath5 m results in a shift in the field of about 30 oe .
this means that the internal field creates a spin - wave propagation channel in the middle of the waveguide @xcite . as is visible in the figure
the magnetic field is practically not modulated along this channel and the spin wave propagates in a non - periodic potential .
nevertheless , one can see that the bias field oscillates closer to the lateral edges of the waveguide ( see dotted curves in fig .
[ fig1]e - f )
. this can be considered as a periodic variation of the effective width of the spin - wave channel . in fig .
[ fig2]a the solid line represents the spin - wave transmission characteristics extracted from the simulated data for the notched mc .
it was obtained by integrating the spin - wave intensity along the line drawn across the crystal at @xmath19 m away from the antenna ( see dotted line in fig .
[ fig1]a ) . as it is expected no spin - wave transmission
is allowed below the ferromagnetic resonance ( fmr ) frequency @xmath20 ghz as well as above the excitation cut - off frequency @xmath21 ghz .
the spin - wave band gap corresponding to the first bragg wavenumber @xmath22 rad/@xmath5 m is visible at @xmath23 ghz . the second band gap ( @xmath24 rad/@xmath5 m ) and the third one ( @xmath25 rad/@xmath5 m ) are not as pronounced but are also clearly visible around 10.2 ghz and 11.9 ghz , respectively . the experimentally measured transmission characteristics @xcite for the same magnonic crystal is shown in fig .
[ fig2]a with a dashed line .
one can see that the first band gap is pronounced and has practically the same frequency as in the simulated data @xcite .
the higher - order rejection bands are not visible in the experiment due to limitation of the wavenumbers that can be excited by the @xmath26m - wide antenna : @xmath27 rad / cm ( note that in the simulations we use a 300 nm - wide antenna ) .
nevertheless , the results of the simulations are in good agreement with the experimental ones .
thus , our approach is trustworthy and can be used for further studies of mcs properties and characteristics .
( color online ) comparison of spin - wave transmission characteristics extracted from : a ) experiment @xcite and simulation of mc with rectangular notches and b ) both simulated structures ( rectangular and sinusoidal mcs).,width=6 ] the comparison of the simulated transmission characteristics extracted from both notched and sinusoidal magnonic crystals is shown in fig .
[ fig2]b .
one sees that the first band gap is visible for both structures and has exactly the same frequency
. however , the second and the third band gaps do not exist for the sinusoidal magnonic crystal in contrast to the notched structure .
this brings us to an important conclusion : the mcs transmission characteristics can be changed qualitatively by designing the mcs geometry . in addition , one can see that the first band gap is not so pronounced for the sinusoidal structure .
this is in a good agreement with the results obtained in ref .
@xcite , where it has been shown that approximately half of the incident wave energy is reflected back from the notches due to the geometrical mismatch between the waveguide sections . in the case of a sinusoidal magnonic crystal
this reflection vanishes which results in a smaller reflection parameter and a weaker rejection efficiency , respectively . in general , several scattering mechanisms
can contribute to the formation of the band gaps in the in - plane shaped magnonic crystals : ( 1 ) the reflection caused by the variation of the transverse spin - wave wavenumber ( this wavenumber is inverse to the effective width of the spin - wave waveguide determined by the parameters of the spin - wave propagation channel ) @xcite , ( 2 ) the scattering connected with the transformation of the spin - wave wavevector in the non - uniform internal magnetic field @xcite , and ( 3 ) the spin - wave reflection from the notches due to the geometrical mismatch ( this mechanism is pronounced only for the notched mc ) .
since in our case there is a relatively smooth internal field distribution inside of the spin - wave channel , the first mechanism appears to be dominant .
we have estimated its role in the formation of the multiple rejection bands . in the approach of the uniform internal field
, we can express the variation of the width in terms of wavenumbers as follows : the total wavevector of the propagating wave is given by : @xmath28 , in which @xmath29 represents the longitudinal wavevector , @xmath30 and @xmath31 are the transverse quantized wavenumbers due to the finite width ( @xmath32 ) and finite thickness ( @xmath33 ) , respectively .
the integer parameters @xmath34 and @xmath35 represent the width and the thickness mode numbers . in our case
@xmath36 is due to the small py thickness and the uniform distribution of the dynamic magnetization on the film thickness .
moreover , @xmath37 becomes a function of the coordinate along the structure , and for sinusoidal and step function modulation we will have the corresponding relations : @xmath38\ ] ] @xmath39\ ] ] where @xmath40 is the modulation amplitude , @xmath41 is the spatial coordinate along the structure , @xmath42 is the period of the modulation , and @xmath43 is the number of space harmonics in the fourier series .
the term @xmath44 represents the higher space harmonics of the parameter @xmath45 with amplitudes orders smaller than the first term .
one can observe that both equations predict multiple rejection bands .
however , the amplitudes given by the higher order terms @xmath44 are negligible in comparison with the ones given by the first term .
this analytical prediction has been successfully proved by performing an additional simulation in which we considered a sinusoidally shaped waveguide with the maximal width variation of 0.75 @xmath5 m ( @xmath46 increase in the modulation depth in comparison with the previously shown data ) .
the simulation has shown a single band gap in the spin - wave transmission characteristics .
( color online ) a ) and b ) transmission characteristics extracted locally in the points @xmath47 ( in the center ) and @xmath48 ( at @xmath49 nm ) ( see fig . [ fig1]a and fig .
[ fig1]b ) inside the mcs .
transmission characteristics as a function of the probing position across the spin - waves propagation channel for the c ) step function and d ) sinusoidal structures at a distance @xmath50 m from the antenna .
e ) and f ) @xmath4 component of the magnetic internal field across the simulated structures.,width=8 ] a deeper analysis reveals the dependence of the frequency and the depth of the rejection bands on the distance from the excitation antenna . it has been found that the band gaps appear already after the spin wave has passed two structural periods of the magnonic crystal .
furthermore , the simulation data shows that the rejection band depth depends also on the probing position across the mcs width .
figures [ fig3]a and [ fig3]b present the transmission characteristics extracted in two points : @xmath47 in the center of the structure ( see fig
. [ fig1]a ) , and @xmath48 at half the distance between the center and the edge of the spin - wave propagation channel ( @xmath49 nm ) .
one may note that the depth of the rejection bands is increased for both simulated structures .
the dependence of the magnitude of the precessing magnetization component ( or spin - wave amplitude ) as a function of the position across the mc structures are shown in fig .
[ fig3]c and fig . [ fig3]d .
several features are clearly visible : ( 1 ) the band gaps are more pronounced in the regions close to the edges of the waveguide rather than in the center , ( 2 ) the band gap frequency is slightly varied with the position , and ( 3 ) the absence of the rejection bands just near the waveguide edges . to understand the first feature one should consider that besides the modulation of the transversal wavevector , there is still a variation of the internal field .
this variation is more pronounced closer to the notches than in the center of the magnonic crystal .
therefore , the spin wave which propagates close to the mcs edge is more sensitive to the modulation of the internal field which results in an increase of the rejection efficiency and a more pronounced band gap .
the second feature is related to the decrease of the internal field towards the notches ( see fig . [ fig3]e and [ fig3]f ) .
the change of the internal field produces a shift of the spin - wave dispersion characteristics triggering the shift of the band gap frequency .
moreover , the temporal evolution of the spatial distribution of the magnetization shows a complicated behavior at the edge of the structure .
standing waves formed inside the anti - notches are mixed with secondary higher - order width modes generated due to the scattering of the originally excited one at the edges @xcite . as a consequence
, the band gaps are vanishing at the edges of the spin - wave propagation channel ( the third feature ) . in conclusion , we performed a micromagnetic study of the spin - wave transmission in in - plane shaped microstructured magnonic crystals .
two kinds of spin - wave permalloy waveguides magnetized across their longitudinal axes were investigated : with notched and sinusoidal edge profiles . in both cases
spin waves propagate through a magnetically induced channel of varying effective width .
it has been found that the waveguide width variation leads to the modulation of the spin - wave wavevectors , and consequently to the formation of frequency rejection bands in the spin - wave transmission characteristics .
both the band gap frequency and its depth depend strongly on the probing position inside the magnonic crystal due the non - uniform distribution of the internal magnetic field .
the band gap properties including their number can be predefined by a proper choice of the magnonic crystal geometry . financial support by the dfg se-1771/1 - 2 is gratefully acknowledged . |
a suitable definition of the chiral symmetry has been a long standing problem in lattice field theories .
recently an ultimate solution to this problem seems to appear in the form of the ginsberg - wilson relation @xcite .
two explicit examples of the lattice fermion operators which satisfies the ginsberg - wilson relation have been found so far : one is the perfect lattice dirac operator constructed via the renormalization group transformation@xcite and the other is the overlap dirac operator@xcite derived from the overlap formalism@xcite or from the domain - wall fermion(dwf)@xcite in the limit of the infinite length of the 5th dimension . since the explicit form is simpler for the latter , a lot of numerical investigations@xcite as well as analytic considerations@xcite have been carried out for the domain - wall fermion or the overlap dirac fermion . in this report
we review the recent numerical investigations on the chiral properties of the dwf , which , however , bring puzzling results . to resolve these puzzles
, we derive a formula which connects a chiral symmetry breaking term with the 4-dimensional hermitian wilson - dirac operator . using this formula and the eigenvalue distribution of the wilson - dirac operator , we propose a consistent interpretation of puzzling results .
throughout this report we employ a shamir s variant of the original domain - wall fermion action@xcite , which is equivalent to the 5-dimensional massive wilson fermion with the free boundary condition in the 5-th direction , interacting with the 4-dimensional gauge field .
symbolically the action is given by @xmath0\psi { \nonumber}\\ & + & \bar\psi\left [ d_\mu(a)^2 + \partial_5 ^ 2\right]\psi + m \bar\psi\psi .\end{aligned}\ ] ] note that the sign of the mass term @xmath1 is opposite to the usual wilson fermion . since
gauge fields are 4-dimensional , the 5-th coordinate @xmath2 can be interpreted as `` flavors '' , so that the action can be rewritten in the 4-dimensional form : @xmath3 { \nonumber}\\ & + & \bar\psi_{n , s } \left [ { \cal m } p_r + { \cal m}^\dagger p_l \right]_{s , t}^{n , m } \psi_{m , t},\end{aligned}\ ] ] where the last term is considered as the flavor mixing `` mass '' term .
now we consider the fermion spectrum derived from this action for the free theory . in this case
the action in the momentum space becomes @xmath4\end{aligned}\ ] ] where @xmath5 @xmath6 with @xmath7 $ ] .
.relation between @xmath1 and number of poles [ cols="<,<,<",options="header " , ] by setting @xmath8 , one observe that the solution to the equation @xmath9 ( @xmath10 ) gives a right - handed ( left - handed ) zero mode .
the solutions to these equations are given by @xmath11 where @xmath12 is the number of sites in the 5-th dimension . in order to satisfy the boundary conditions @xmath13 for the above solutions ,
the @xmath14 limit is necessary , together with the condition that @xmath15 , which is equivalent to @xmath16 since we have the kinetic term @xmath17 for the fermion , the condition that @xmath18 or @xmath19 is required for the existence of the massless pole in the continuum limit .
( [ eq : condw ] ) gives a condition for @xmath1 and @xmath20 , which is summarized in table [ tab : pole ] .
it is clear that the value of @xmath1 ( wilson fermion mass ) controls the number of massless fermion poles .
for @xmath21 we have one physical massless dirac fermion .
the form of zero - modes obtained above is too complicated in the coordinate space even for the free theory to define the quark field as @xmath22 .
instead we define the quark field as @xmath23 this definition is reasonable since the right - handed zero mode has a peak at @xmath24 and the left - handed one at @xmath25 .
according to this definition , the quark mass term is given by @xmath26 , which leads to the non - zero pole mass of the fermion for @xmath27 : @xmath28.\end{aligned}\ ] ] the formula tells us that @xmath29 even at @xmath30 for @xmath31 , corresponding to the fact that no solution to zero modes exist for finite @xmath12 . on the other hand the fact that @xmath32 in the @xmath33 limit strongly suggests that the chiral symmetry is realized for the dwf .
the lattice qcd with the dwf is called the domain - wall qcd ( dwqcd ) .
= 6.0 cm as suggested by the behavior of the pole mass , the dwqcd satisfies the almost exact axial ward - takahashi identity , which is given by @xmath34 where @xmath35 is an axial - vector current , @xmath36 is an arbitrary operators which contain @xmath37 and @xmath38 only as the fermion variable , and @xmath39 , defined in the middle of the 5-th dimension , represents the explicit breaking term of the chiral symmetry . in perturbation
theory it is proven that @xmath40 as @xmath41 for the non - singlet sector(@xmath42 ) , while @xmath43 for the singlet sector ( @xmath44 ) .
the latter one corresponds to the u@xmath45(1 ) anomaly , well - known in the continuum qcd .
again the presence of the exact axial ward - takahashi in the dwqcd suggests the existence of the chiral symmetry in the @xmath46 limit even for interacting theories .
= 6.0 cm in order to investigate the chiral properties of the dwqcd , we define the anomalous quark mass in axial ward - takahashi identity as follows .
@xmath47 we expect that @xmath48 behaves as @xmath49 + \gamma m_f + o(m_f^2 ) .\end{aligned}\ ] ] the existence of the massless fermion requires @xmath50 . in fig .
[ fig : mq5mf ] , we have plotted @xmath48 as a function of @xmath51 with @xmath52 on a @xmath53 lattice at @xmath54 of the renormalization group ( rg ) improved gauge action @xcite .
the figure shows that @xmath51 dependence of @xmath48 is much milder than that of @xmath55 and is almost constant , so that the extrapolation of @xmath48 to @xmath56 becomes much easier .
in addition @xmath48 is very precise and sensitive to @xmath12 , so that the @xmath12 dependence can be extracted .
hereafter we exclusively employ @xmath48 to investigate the chiral properties of the dwqcd , because of these advantages of @xmath48 .
we have plotted @xmath48 as a function of @xmath12 , calculated on the cp - pacs for the plaquette gauge action in fig.[fig : mq5nsp ] and the rg improved gauge action(lower ) in fig.[fig : mq5nsr ] , at coarse(@xmath57 fm ) and fine(@xmath58 fm ) lattice spacings@xcite . at both lattice spacings @xmath48 is smaller for the rg improved action than for the plaquette action . at the coarse lattice
spacing(@xmath59 0.2 fm ) it seems that @xmath60 for both actions . on the other hand , @xmath61 for the rg action while @xmath60 for the plaquette action at the fine lattice spacing ( @xmath59 0.1 fm ) .
in this section we try to understand the dependences of @xmath48 on the lattice spacings and the gauge actions .
the existence of massless fermion of the dwqcd can be proven in the absence of zero eigenvalues of the hermitian wilson - dirac operator @xmath62 , where @xmath63 is the 4-dimensional wilson fermion operator with the fermion mass @xmath1 and @xmath64 is satisfied
. in the case of the 4-dimensional wilson fermion , zero eigenvalues of @xmath65 leads to the spontaneous breaking of parity - flavor symmetries , @xmath66 , as shown below@xcite .
@xmath67 { \nonumber}\\ & = & -i\lim_{h\rightarrow+0 } { \rm tr } \left[\frac{1}{h_w+i h } - \frac{1}{h_w - i h } \right ] { \nonumber}\\ & = & -i\lim_{h\rightarrow + 0}\sum_n { \left\langle \lambda_n\left| \left(\frac{1}{\lambda_n+i h } - \frac{1}{\lambda_n - i h } \right ) \right| \lambda_n\right\rangle}{\nonumber}\\ & = & -i\lim_{h\rightarrow+0}\int d\,\lambda \ , \rho_{h_w}(\lambda ) { \left\langle \lambda\left| \left(\frac{1}{\lambda+i h } - \frac{1}{\lambda - i h } \right ) \right| \lambda\right\rangle } { \nonumber}\\ & = & -i \int d\,\lambda \,\rho_{h_w}(\lambda ) ( -2\pi i)\delta(\lambda){\nonumber}\\ & = & -2\pi\,\rho_{h_w}(0 ) \end{aligned}\ ] ] where @xmath68 is the density of the eigenvalues of @xmath69 , which is defined by @xmath70 = 6.0 cm in fig.[fig : aoki_phase ] , we have drawn the expected phase structure of the lattice qcd with the wilson fermion in the @xmath71@xmath1 plane , where @xmath72 is the gauge coupling constant@xcite .
the small value of @xmath73 corresponds to the coarse lattice spacing while the large value to the fine lattice . at @xmath74 ( @xmath75 )
no zero eigenvalues exist for @xmath27 , which is the region where the domain - wall fermion has the zero modes .
if we increase the gauge coupling , the allowed region of @xmath1 for the massless dwf becomes narrow and is shifted to larger values of @xmath1 .
finally the allowed region disappears at @xmath76 .
this implies that massless fermions disappear at @xmath77 for the ( quenched ) dwqcd . according to this interpretation it seems that @xmath78 ( @xmath79 fm ) , 6.0(@xmath80 fm ) for the plaquette gauge action , while the numerical data indicate that 2.6 ( @xmath80 fm ) @xmath81 2.2 ( @xmath79 fm ) for the rg improved gauge action
in this section we directly investigate the distribution of small eigenvalues of @xmath69 .
= 7.5 cm in figs .
[ fig : evdistp ] and [ fig : evdistr ] we plot monte carlo time histories for the lowest six eigenvalues of @xmath83 for the plaquette and rg - improved actions@xcite .
the figures at @xmath84 fm are shown at the same lattice size with the previous work of @xmath85 . in each figure
the left panel shows results for @xmath57 fm and the right panel for @xmath84 fm .
open squares show the minimum eigenvalue @xmath86 and filled diamonds are the higher five eigenvalues .
there is a clear trend that the minimum eigenvalues become larger for smaller lattice spacings .
another interesting point is that the rg - improved action gives larger values @xmath86 than the plaquette action , which indicates that the rg - improved action has a better chiral behavior .
= 7.5 cm the spectral density of @xmath69 is defined by @xmath87 where the summation is over the eigenvalues of @xmath69 .
the density of zero - eigenvalues @xmath82 , related to the order parameter of the parity - flavor breaking , has been calculated by the accumulation method proposed in @xcite , which is based on the relation @xmath88 we note that , for the small-@xmath89 expansion in ( [ eq : ansatz ] ) , analyticity of @xmath90 at the origin is assumed .
the result of @xcite , where @xmath82 at several @xmath73 s have been plotted as a function of @xmath1 , is given in fig .
[ fig : rhom ] for the plaquette gauge action .
although @xmath82 becomes smaller for smaller @xmath91 (; larger @xmath73 ) , no region without zero eigenvalues seems to exist . this impression is indeed confirmed in fig .
[ fig : rho - a ] , where @xmath82 has been plotted as a function of @xmath92 at @xmath93@xcite .
= 6.0 cm it has been speculated in @xcite that the @xmath91 dependence of @xmath82 is well reproduced by the from that @xmath94 .\ ] ] if this form is correct for all @xmath1 , the region with @xmath95 does not exist as long as @xmath96 .
the similar result has been obtained for the rg improved action .
these results immediately lead to the conclusion that there is no gap of the parity - flavor breaking phase in the @xmath72@xmath1 plane for the wilson fermion , so that the domain - wall qcd fails to realize the chiral symmetry at all @xmath96 .
however this conclusion seems to contradict the numerical data for the rg improved gauge action at @xmath97 fm .
puzzles still remain .
in this section we try to understand the pathological behavior of @xmath48 as a function of @xmath12 , in terms of the eigenvalues of @xmath69 .
we first derive the analytic expression for @xmath48 as follows .
@xmath98 where @xmath99 represents site , spinor and color indices , @xmath100 for the derivation of the above formula , see @xcite . based on the above formula @xmath48
can be approximately estimated at @xmath56 as follows .
@xmath101 { \nonumber}\\ & & = \frac{1}{n_{all}}\left [ c_0\sum_n f(\tilde\lambda_n)^{-2 } + \sum_c f(\tilde\lambda_c)^{-2 } \right ] \\
\tilde\lambda & = & \log\frac{1-\lambda}{1+\lambda}\end{aligned}\ ] ] where @xmath102(@xmath103 ) is the continuous(discrete ) eigenvalue of @xmath104 , @xmath105 is the number of discrete modes , @xmath106 is the number of degrees of freedom , and @xmath107 is a weight factor for the discrete modes .
see @xcite for the detail of the estimation .
= 6.0 cm the important point is that there are two types of eigenvalues , the continuous eigenvalues and the discrete eigenvalues .
the eigenfunction for the former is the plain - wave like and spreads over whole space , while the one for the latter is exponentially localized . in the strong coupling region at @xmath77 , the continuous eigenvalues can become zero , so that @xmath108 . therefore the dwqcd does not work as expected .
( see @xcite for more details . ) in the weak coupling region , on the other hand , the continuous eigenvalues has a minimum value @xmath109 , such that @xmath110 for the continuous eigenvalues . instead of the continuous eigenvalues ,
discrete eigenvalues may become almost zero .
the expected distribution of eigenvalues @xmath89 in the weak coupling region is schematically drawn in fig .
[ fig : dist ] .
( 150,90 ) ( 0,10)(0,1)80 ( 0,10)(1,0)140 ( 60,10)(60,80)(140,80 ) ( 100,80)@xmath90 ( 1,10)(0,1)45 ( 5,10)(0,1)80 ( 10,10)(0,1)70 ( 15,10)(0,1)65 ( 17,10)(0,1)75 ( 20,10)(0,1)55 ( 25,10)(0,1)70 ( 29,10)(0,1)75 ( 34,10)(0,1)45 ( 36,10)(0,1)75 ( 41,10)(0,1)70 ( 47,10)(0,1)55 ( 55,10)(0,1)75 ( 0,5)0 ( 145,8)@xmath89 ( 10,0)descrete modes ( 90,0)continuous modes ( 60,3)@xmath109 it is likely that small ( localized ) discrete eigenvalues are caused by dislocations . in particular , @xmath111 , the value of the topological charge in some configuration should move to the different value @xmath112 during the simulation of the lattice qcd , so that the dislocation appears during the transition from @xmath111 to @xmath112 .
if the dislocation appears , the wilson - dirac operator has an almost zero eigenvalue at some value of @xmath1 .
therefore it is very difficult to avoid small ( localized ) eigenvalues of @xmath69 and @xmath113 . from the approximated estimate for @xmath48 , only near zero eigenvalues can contribute to it as @xmath14 .
hence only small discrete eigenvalues are dangerous for the dwqcd in the weak coupling region .
the analysis in @xcite leads to the followings . in the @xmath14 limit @xmath48
vanish at finite volume ( ; @xmath114 ) , since the probability of having exact zero eigenvalues is zero .
if the infinite volume ( ; @xmath115 ) limit is taken before the @xmath14 limit , there are two possibilities for the distribution of discrete eigenvalues . unless the number of near zero eigenvalues increases linearly in @xmath116 , they does not contribute to @xmath48 , hence the dwqcd works fine .
if it increases linearly in @xmath116 , @xmath82 and hence @xmath48 becomes non - zero , so that the dwqcd can not realize the chiral symmetry .
the numerical data obtained so far seem to prefer the latter possibility
. however more detailed investigations will be necessary for the definite conclusion .
the distribution of near zero eigenvalues of @xmath69 plays a crucial role for the chiral properties of the dwqcd .
we have pointed out that there are two types of eigenvalues , continuous and discrete . in the weak coupling region ,
only discrete eigenvalues become relevant for @xmath48 .
however the effect of discrete eigenvalues to @xmath48 is rather small even if the number of near zero discrete eigenvalues increases linearly in the volume .
we think that this has brought confusing numerical results for the @xmath12 dependence of @xmath48 .
our analysis suggests that the gap is closed in the phase structure of the wilson fermion due to the near zero discrete eigenvalues , if the quenched approximation is employed . in the dynamical qcd simulation ,
the near zero discrete eigenvalues are strongly suppressed by the fermion determinant , so that the gap opens and the expected phase structure is correct in the real lattice qcd .
since such a suppression of small discrete eigenvalues is not known for the dwqcd , it is unclear how the dynamical dwqcd change the distribution of the near zero discrete eigenvalues .
i would like to thank dr .
y. taniguchi , with whom i have completed the main part of this work .
i also thank the members of the cp - pacs collaboration for some results presented in this report , and prof .
h. neuberger for the useful discussion .
this work is supported in part by grants - in - aid of the ministry of education(no .
12640253 ) . | we investigate chiral properties of the domain - wall fermion ( dwf ) system . after a brief introduction for the dwf ,
we summarize the recent numerical results on the chiral properties of the domain - wall qcd ( dwqcd ) , which seem mutually inconsistent .
we next derive a formula which connects a chiral symmetry breaking term in the five dimensional dwf ward - takahashi identity with the four - dimensional hermitian wilson - dirac operator . based on this formula
, we propose a solution , which resolves the inconsistency among recent numerical data , and give a consistent picture of the chiral properties of the dwqcd . |
in the present work the orientational degrees of freedom of the particles are treated within the zwanzig model @xcite , hence a particular orientation can be identified with a number @xmath30 ( cf . tab .
[ tab1 ] ) . 1 & x & y & z + 2 & z & x & y + 3 & y & z & x + 4 & x & z & y + 5 & y & x & z + 6 &
z & y & x + according to density functional theory it is possible to express the free energy of a system as a functional of the single - particle density @xmath29 of particles with orientation @xmath99 ( @xmath30 ) belonging to species @xmath26 ( @xmath20 ) as @xcite @xmath31}{k_b t } = \int d\mathbf{r}\sum_{\alpha , i } \rho_i^\alpha(\mathbf{r } ) \bigl [ \ln(\rho_i^\alpha(\mathbf{r } ) \lambda_\alpha^3 ) -1 \bigr ] + \frac{\mathcal{f}^{ex}[\rho]}{k_b t } , \label{eq2}\ ] ] where for brevity @xmath100 the excess term @xmath101 $ ] has in general a non - trivial dependence on @xmath29 . for short - range potentials
it is always possible to express @xmath101 $ ] as a virial series in the single - particle density .
therefore , by truncating the series at second - virial order and thus disregarding higher - order contributions , one obtains @xmath35}{k_b t } = - \frac{1}{2 } \int d\mathbf{r } \ , d\mathbf{r } ' \sum_{\alpha,\alpha',i , i ' } f_{i i'}^{\alpha \alpha'}({\mathbf r } - { \mathbf r } ' ) \rho_i^\alpha(\mathbf{r } ) \rho_{i'}^{\alpha'}(\mathbf{r } ' ) , \label{eq2b}\ ] ] where the mayer function @xmath102 is defined in terms of the pairwise interaction potential @xmath37 as @xmath103 -1 .
\label{eq3}\ ] ] the single - particle density @xmath104 is related to the number of particles @xmath19 through the normalization condition @xmath105 for hard cuboids the interaction potential , which expresses the impenetrability of the particles , is @xmath106 according to the index notation defined in tab .
[ tab1 ] , the @xmath107-dimensional vectors @xmath108 , @xmath109 and @xmath110 of species @xmath26 are given in terms of the dimension of the particles by @xmath111 the main goal of this work is to study the stability of spatially homogeneous phases ( i.e. isotropic and nematic ) . in order to simplify the problem we therefore neglect spatial modulations in the single - particle density by imposing @xmath38 .
consequently , eq . ( [ eq2 ] ) becomes @xmath112 + \frac{1}{2 } \sum_{\alpha,\alpha',i , i ' } e^{\alpha \alpha'}_{i i ' } \rho_i^\alpha \rho_{i'}^{\alpha ' } , \label{eq7}\ ] ] which is the restricted orientation version of the onsager free energy @xcite .
the matrix @xmath113 in eq .
( [ eq7 ] ) is the excluded volume between two particles belonging to species @xmath26 and @xmath114 with orientations @xmath99 and @xmath115 interacting through the potential eq .
( [ eq5 ] ) @xmath116 in the homogeneous case the normalization condition eq .
( [ eq4 ] ) becomes @xmath117 the single - particle density at equilibrium is the one which minimizes eq .
( [ eq7 ] ) with the constraints of eq .
( [ eq9 ] ) for all @xmath20 , hence it is found by solving the euler - lagrange equation @xmath118 which is achieved by standard numerical ( iterative ) techniques .
while studying the homogeneous equilibrium phases of the system , we are also interested in estimating their upper bound in the phase diagram , where spatially inhomogeneous phases tend to be thermodynamically favored .
bifurcation theory @xcite provides a way to investigate the limit of stability of a particular phase .
the condition of thermodynamic stability of a phase described by the single - particle density @xmath119 requires that the system corresponds to a minimum of the free energy @xmath120 , i.e. a stationary point that satisfies @xmath121 for any arbitrary perturbation @xmath122 . by inserting the functional expression eq .
( [ eq2 ] ) into eq .
( [ eq10 ] ) , one finds that the reference phase ( described by @xmath119 ) ceases to be stable at the smallest density @xmath123 at which a perturbation @xmath122 exists such that @xmath124 here we are interested in calculating the limit of stability of the ( uniaxial or biaxial ) nematic phase with respect to smectic fluctuations . with this in mind , in eq .
( [ eq11 ] ) we neglect spatial modulations in the reference phase , i.e. @xmath125 , and a positional dependence of the fluctuations only along the @xmath40 direction , i.e. @xmath126 . after some rearranging eq .
( [ eq11 ] ) becomes @xmath127 where @xmath128 and @xmath129 a symmetric ( hermitean ) kernel . by inserting the explicit form of the inter - particle potential ( cf
( [ eq3 ] ) and ( [ eq5 ] ) ) into eq .
( [ eq25 ] ) , we obtain @xmath130 eq .
( [ eq24 ] ) can be more conveniently solved in fourier space , where it reads @xmath131 with @xmath132 and @xmath133 . in conclusion ,
the limit of stability of the nematic phase with respect to smectic fluctuations can be numerically found as the minimum packing fraction @xmath134 at which there exists a wave vector @xmath135 such that the @xmath136 matrix with entries @xmath137 has a unit eigenvalue .
the periodicity of the corresponding bifurcating smectic phase is given by @xmath138 .
when dealing with mixtures , the phase diagram is conveniently expressed in terms of the osmotic pressure @xmath140 vs. the mole fraction @xmath141 of @xmath142 components . in this way it is possible to visualize the coexistence of phases characterized by a different composition with respect to the parent distribution .
this phenomenon , called demixing or fractionation , is a consequence of the first - order character of the transition . ( a ) phase diagram of a binary mixture of hard cuboids in terms of the reduced osmotic pressure @xmath143 vs. mole fraction of the first species @xmath43 .
( b)-(d ) @xmath12 ( red solid line ) , @xmath3 ( blue dotted line ) and @xmath4 ( green dashed line ) branches of the gibbs free energy per particle @xmath144 at ( b ) @xmath145 , ( c ) @xmath146 and ( d ) @xmath147 .
a straight line with slope @xmath148 ( with @xmath149 and @xmath150 the compositions of the coexisting isotropic and nematic phases ) was subtracted in each case to enhance the visualization of the common tangent construction . ] here we analyze demixing in a binary mixture of cuboids parameterized as in eq .
( 3 ) with @xmath50 , @xmath51 and @xmath151 . in fig .
[ fig3](a ) we report the phase diagram for such a system as a function of the mole fraction @xmath43 of the larger species .
the expected first - order character of the @xmath152 transitions is not detectable at this scale ( see below ) , whereas the @xmath153 transitions appear to be second order . at three different values of the reduced pressure @xmath154 we calculated the isotropic and uniaxial nematic branches of the gibbs free energy per particle @xmath155 .
the coexistence between the two phases is given by a common tangent construction , which allows to evaluate the difference in composition @xmath156 of the coexisting phases .
the results are reported in fig . [ fig3](b)-(d ) for @xmath145 , @xmath157 and @xmath158 , respectively . in the three cases ,
two of which describe a @xmath159 and one a @xmath160 transition , @xmath161 and can therefore be neglected .
the situation does not change when one considers different values of the bidispersity parameter @xmath66 .
although landau - de gennes theory predicts the @xmath152 transition to be first order @xcite , we have just shown that its discontinuous character can be safely neglected for the binary mixture of boardlike particles we consider in this work . in our opinion , this fact is tightly related to the shape of the particles close to the @xmath13 value .
in fact , when considering a monodisperse system , the closer @xmath6 is to zero the weaker is the first - order character of the @xmath152 transition ( see also sec .
[ mono ] ) .
this fact allows us to assume that for an arbitrary number of components of volume - polydisperse cuboids close to @xmath13 the @xmath152 transition can be approximated as continuous . as a consequence
, we can neglect demixing in the phase behavior analysis reported in fig .
4 , thus reducing enormously the complexity of the problem .
the main goal of the present work is to investigate how polydispersity affects the phase behavior , and in particular the stability , of the @xmath0 phase in a system of hard cuboids .
for this reason , it is instructive to study what the theoretical framework described in sec .
[ dft ] predicts in the monodisperse case @xmath162 .
in particular , we will focus here on the role of the particles dimensions on the phase behavior of the system .
phase diagram of a monodisperse system of hard cuboids as a function of the shape parameter @xmath70 ( with @xmath50 fixed and @xmath71 variable ) .
the solid lines indicate phase boundaries as calculated by minimizing the onsager - zwanzig functional , the dashed line indicates the limit of stability of the nematic with respect to the smectic phase and the open circle the landau tetracritical point .
the inset highlights the first order character of the @xmath152 transition and how this tends to become continuous by approaching @xmath13 . ] in fig .
[ fig1 ] we report the phase diagram of a monodisperse system of hard cuboids as a function of the aspect ratio @xmath71 at fixed @xmath50 .
consequently , by varying @xmath71 one varies the shape parameter @xmath70 , in such a way that by crossing the point @xmath13 one expects a transition from plate- to rod - like behavior .
this is precisely what fig .
[ fig1 ] shows , where the phase separation lines are calculated by minimizing the onsager - zwanzig functional eq .
( [ eq7 ] ) with the constraint of eq .
( [ eq8 ] ) for each value of the packing fraction @xmath53 . moreover , bifurcation theory ( cf . sec .
[ bif ] ) provides a way to estimate the upper limit of stability of homogeneous phases with respect to the smectic ( dashed line in fig .
[ fig1 ] ) .
[ fig1 ] shows that to observe a stable @xmath0 phase , the shape of the particles should be designed with extremely high precision in a small @xmath6-regime about @xmath13 .
in fact , for @xmath163 the @xmath0 phase disappears unless @xmath164 .
this is due both to the tight cusp - like shape of the @xmath153 transition line and to the preempting character of inhomogeneous phases .
analogous results can be obtained by varying the shape parameter through @xmath72 , while keeping @xmath71 fixed ( not shown ) . finally , in the inset of fig .
[ fig1 ] ( note the different scale ) we show the first order character of the @xmath152 transition , which tends to become second - order by approaching the critical point at @xmath13 .
orientation distribution function of a monodisperse system of hard cuboids as a function of the packing fraction @xmath53 obtained by minimization of the onsager - zwanzig functional eq .
( [ eq7 ] ) for @xmath162 .
the cuboids have dimensions @xmath50 and ( a ) @xmath165 , ( b ) @xmath166 , ( c ) @xmath167 , ( d ) @xmath51 .
the different lines indicate the probability of a particular orientation @xmath30 ( cf . tab .
[ tab1 ] ) .
the dashed vertical line shows the limit of stability of the nematic phases with respect to the smectic . ] for the sake of completeness , in fig .
[ fig2 ] we report the orientation distribution function @xmath168 , which is the probability of a given orientation @xmath30 as a function of the packing fraction @xmath53 for different values of the shape parameter @xmath6 . in the monodisperse case this function coincides with the single - particle density divided by the number density : @xmath169 .
the values of the orientation distribution function characterize the symmetry of the corresponding phase .
in fact , at a given packing fraction @xmath53 in fig . [ fig2 ] one can have one of the following possibilities : * the probabilities @xmath168 are all the same , i.e. @xmath170 ( isotropic @xmath12 phase ) ; * the probabilities @xmath168 are coupled two - by - two , demonstrating the presence of a symmetry axis ( uniaxial nematic @xmath1 phase ) ; * the probabilities @xmath168 are different between each others ( biaxial nematic @xmath0 phase ) .
this classification is easily generalized to the multi - component case . with this in mind , one can observe the difference in the orientation distribution function when @xmath172 ( @xmath165 , fig . [
fig2](a ) ) , @xmath13 ( @xmath166 , fig .
[ fig2](b ) ) and @xmath173 ( @xmath167 , fig .
[ fig2](c ) ) .
the vertical dashed line indicates the limit of stability with respect to smectic fluctuations as given by bifurcation theory .
finally , fig .
[ fig2](d ) shows the predicted orientation distribution function when the experimental value @xmath51 is considered @xcite , and highlights how according to the model the @xmath0 phase is expected to be preempted by inhomogeneous phases . | inspired by the observations of a remarkably stable biaxial nematic phase [ e.v.d .
pol _
et al
. _ , phys . rev . lett . * 103 * , 258301 ( 2009 ) ] , we investigate the effect of size polydispersity on the phase behavior of a suspension of boardlike particles . by means of onsager
theory within the restricted orientation ( zwanzig ) model we show that polydispersity induces a novel topology in the phase diagram , with _ two _ landau tetracritical points in between which _ oblate _ uniaxial nematic order is favored over the expected _ prolate _ order .
additionally , this phenomenon causes the opening of a huge stable biaxiality regime in between uniaxial nematic and smectic states .
pacs numbers : 82.70.dd , 61.30.cz , 61.30.st , 64.70.m- since its first prediction back in the early 1970s @xcite , the biaxial nematic ( @xmath0 ) phase has strongly attracted the interest of the liquid crystal ( lc ) community @xcite .
in contrast to the more common uniaxial nematic ( @xmath1 ) phase , where cylindrical symmetry with respect to the nematic director determines optical uniaxiality , the @xmath0 phase is characterized by an orientational order along three directors and consequently by the existence of two distinct optical axes .
the prospect of inducing orientational ordering along three directions , while maintaining a nematic fluid - like mechanical behavior @xcite , renders biaxial nematics preeminent candidates for next generation lc - based displays @xcite .
although experimental evidences of stable @xmath0 phases were reported already 30 years ago in lyotropic lcs @xcite , in thermotropics this result was achieved in systems of bent - core molecules only a few years ago @xcite .
actually , when trying to experimentally reproduce an @xmath0 phase , one often encounters practical problems related to its unambiguous identification @xcite and to the presence of competing thermodynamic structures @xcite .
stabilizing @xmath0 states is therefore an open , challenging scientific problem with huge potential applications .
motivated by the exciting results of a recent experiment on a colloidal suspension @xcite , we use here a mean - field theory to investigate the role played by size polydispersity on the stability of biaxial nematics in systems of boardlike particles .
we show that a difference in the particle volume of a binary mixture can favor _ oblate _ uniaxial orientational ordering over _ prolate _ , in sharp contrast with the behavior of the pure systems .
this phenomenon gives rise to a new phase diagram topology due to the appearance of _ two _ landau tetracritical points , leading to a wider region of @xmath0 stability .
this feature is shown to hold also for a larger number of components , thus offering an explanation to the results of ref .
@xcite . finally , we argue that our findings could furnish a new way to look for biaxiality in thermotropic lcs .
( a ) cuboidal particle with dimensions @xmath2 .
( b ) schematic representation of a system of freely rotating cuboids in the biaxial nematic phase @xmath0 , ( c ) the uniaxial nematic prolate @xmath3 and ( d ) the uniaxial nematic oblate @xmath4 . in this work
the rotational degrees of freedom are discretized according to the zwanzig model @xcite . ] at low density in lyotropics , and at high temperature in thermotropics , the @xmath0 phase appears as a crossover regime in between `` rod - like '' and `` plate - like '' behavior @xcite .
in fact , one can distinguish between the @xmath1 phase developed by rods , in which particles align the longest axis along a common direction ( uniaxial nematic _ prolate _ , @xmath3 ) , and that developed by plates , in which particles align the shortest axis ( uniaxial nematic _ oblate _ , @xmath4 ) . a natural candidate system for developing
an @xmath0 phase is a binary mixture of rods and plates @xcite ; however , in most cases a demixing transition into two uniaxial nematic phases , i.e. @xmath3 and @xmath4 , prevents its stabilization @xcite .
alternatively , a stable @xmath0 state is expected in a system of particles with cuboid ( i.e. rectangular parallelepiped ) shape defined by the lengths of the principal axes @xmath5 , as depicted in fig .
[ f1](a ) @xcite . in this case , it is convenient to introduce a _ shape parameter _ @xmath6 , defined by @xmath7 . by increasing the packing fraction and disregarding the possible stability of inhomogeneous phases ,
a system of cuboids undergoes an @xmath8 sequence of phases if @xmath9 , whereas an @xmath10 sequence is found if @xmath11 ( @xmath12 stands for the isotropic phase ) @xcite . a schematic representation of these nematic phases is given in fig .
[ f1](b)-(d ) .
the case @xmath13 describes the optimal `` brick '' shape exactly in between `` rod - like '' and `` plate - like '' . in this case
the @xmath1 phase is suppressed and substituted by a second - order @xmath14 transition @xcite .
the first experimental realization of the hard - cuboid model was found only recently in a colloidal suspension of boardlike mineral goethite particles @xcite . by producing particles with shape parameter @xmath15 close to zero ( @xmath16 and size polydispersity of @xmath17 )
, the authors were able to produce an @xmath0 phase stable over a pressure range surprisingly much wider than to be expected from theory @xcite and simulations @xcite for particles whose shape parameter deviates even slightly from zero .
even more interestingly , the authors affirm that _ no _ @xmath1 phase was observed , contrasting ref .
@xcite .
they suggest that a possible reason for this disagreement should be found in ingredients whose effects have never been studied so far because of their complexity , i.e. fractionation , sedimentation and polydispersity .
these unexpected results motivate our interest in analyzing the effect of the above mentioned ingredients , in particular polydispersity , on the stability of the @xmath0 phase in a fluid of hard cuboids .
we consider an @xmath18-component suspension of @xmath19 colloidal cuboidal particles of species @xmath20 with dimensions @xmath21 ( @xmath22 ) in a volume @xmath23 at temperature @xmath24 .
the total number density of colloids is @xmath25 , the mole fraction of species @xmath26 is @xmath27 and the packing fraction is @xmath28 .
the theoretical framework used in this letter consists of onsager theory of lcs @xcite , which is a density functional theory truncated at second - virial order . in order to facilitate the calculations we follow zwanzig by restricting the orientations of the particles to the six in which their principal axes are aligned along a fixed cartesian frame @xcite .
although quantitative agreement with real systems is not expected because of the simplifications introduced in the model , the same model was shown to successfully predict non - trivial phenomena like demixing in rod - plate mixtures @xcite , orientational wetting due to confinement and capillary nematization @xcite .
moreover , we expect that transitions between different nematic phases and smectic phases are better described by this model than transitions from isotropic to nematics . in density
functional theory the free energy of the system is expressed as a functional of the local density @xmath29 of particles of species @xmath20 with orientation @xmath30 as @xcite @xmath31}{k_b t } = \int d\mathbf{r}\sum_{\alpha , i } \rho_i^\alpha(\mathbf{r } ) \bigl [ \ln(\rho_i^\alpha(\mathbf{r } ) \lambda_\alpha^3 ) -1 \bigr ] + \frac{\mathcal{f}^{ex}[\rho]}{k_b t } , \label{e2}\ ] ] where @xmath32 is the boltzmann constant and @xmath33 the thermal volume of species @xmath26 . at second - virial order
the excess free energy @xmath34 reads @xmath35}{k_b t } = - \frac{1}{2 } \int d\mathbf{r } \ , d\mathbf{r } '
\sum_{\alpha,\alpha',i , i ' } f_{i i'}^{\alpha \alpha'}({\mathbf r } - { \mathbf r } ' ) \rho_i^\alpha(\mathbf{r } ) \rho_{i'}^{\alpha'}(\mathbf{r } ' ) , \label{e2b}\ ] ] where @xmath36 - 1 $ ] is the mayer function , defined in terms of the pair - wise potential @xmath37 . by neglecting spatial modulations , i.e. by imposing @xmath38 , the free energy eq .
( [ e2 ] ) reduces to an onsager - type functional whose minimization ( under the constraints that @xmath39 for all @xmath20 ) allows to identify the spatially homogeneous equilibrium phase ( see appendix a ) . since at sufficiently high density one expects spatially inhomogeneous phases to be thermodynamically favored , we apply bifurcation theory @xcite to determine the limit of stability of the homogeneous equilibrium phases with respect to smectic fluctuations . by considering spatial density modulations only along the @xmath40 axis , i.e @xmath41 in eq .
( [ e2 ] ) , the smectic bifurcation density is the minimum density at which the hessian second - derivative matrix of the free energy has an eigenvalue equal to zero ( see appendix b ) .
our analysis starts by considering the simplest case of polydispersity , i.e. a mixture of @xmath42 components with mole fractions @xmath43 and @xmath44 , respectively . among the different ways one can parameterize polydispersity , our preliminary analysis suggests to consider _ volume polydispersity _
( i.e. same particle shape but different volume ) .
therefore , we study the phase behavior of a binary mixture of hard cuboids whose dimensions are @xmath45 where the parameter @xmath46 describes the degree of bidispersity .
notice that eq .
( [ e4 ] ) implies the same aspect ratios for both species @xmath47 and @xmath48 ( hence @xmath49 ) . here
we set @xmath50 and @xmath51 ( @xmath52 ) in order to reproduce the experimental system of ref .
@xcite , thereby neglecting the small effect of the ionic double layer used by the authors to interpret the experimental data .
( color online ) phase diagram of a binary mixture of hard cuboids in terms of packing fraction @xmath53 vs. mole fraction of the larger component @xmath43 showing isotropic ( @xmath12 ) , uniaxial ( @xmath3 and @xmath4 ) and biaxial ( @xmath0 ) nematic and smectic ( @xmath54 ) phases . the size of the particles is defined by eq . ( [ e4 ] ) with @xmath50 , @xmath51 ( @xmath52 ) and bidispersities ( a ) @xmath55 , ( b ) @xmath56 , ( c ) @xmath57 , ( d ) @xmath58 .
the solid lines separate different homogeneous phases , the dashed lines indicate the limit of stability of the homogeneous phases with respect to smectic fluctuations , whereas the open circles represent the landau tetracritical points . ]
fig .
[ f2 ] shows density - composition phase diagrams of binary mixtures ( @xmath42 ) of boardlike particles with the experimental shape parameter @xmath52 for various bidispersity parameters ( a ) @xmath55 , ( b ) @xmath59 , ( c ) @xmath60 and ( d ) @xmath61 , featuring isotropic ( @xmath12 ) , uniaxial nematic ( @xmath3 _ and _ @xmath4 ) , biaxial nematic ( @xmath0 ) and smectic ( @xmath54 ) phases .
due to the near - perfect `` biaxial '' shape of the particles , fractionation is extremely weak and invisible on the scale of fig .
[ f2 ] ( see appendix c ) . at the extreme mole fractions @xmath62 and @xmath63 ( pure systems )
all phase diagrams feature the phase sequence @xmath64 that is well known and expected for board - shaped particles with @xmath9 , with the @xmath0 phase metastable with respect to the @xmath54 phase @xcite ( see also appendix d ) .
however , for all @xmath65 there is an intermediate composition regime in which the @xmath0 phase is found to be stable , the more so for increasing @xmath66 . whereas the opening - up of a stable @xmath0 regime is only quantitative for @xmath55 ,
there is a qualitative change of the phase diagram topology beyond @xmath56 , where _ two _ landau tetracritical points appear ( open circles in figs . [ f2](b)-(d ) ) . in between these critical points a region of stable @xmath4 phase , which is _ not
_ expected for the rod - shaped particles ( @xmath9 ) of interest , opens up .
clearly , figs . [ f2](c ) and ( d ) show that this unexpected @xmath4 regime enlarges with bidispersity , accompanying a further increased @xmath0 stability .
in other words , excluded - volume interactions in mixtures of board - shaped rods with the same shape and different volume tend to favor @xmath0 stability as a consequence of an unexpected @xmath67 competition . at higher packing fractions
the increased @xmath0 stability with respect to the @xmath54 phase is not a surprise , given that regular packing into layers is hindered by size differences between particles @xcite .
it is interesting to understand how the remarkable features of the binary mixture described in fig .
[ f2 ] change with the shape of the particles .
here we are mainly interested in two properties of the phase diagram : ( i ) the minimum threshold bidispersity @xmath68 at which the landau tetracritical points appear and ( ii ) the tetracritical mole fractions @xmath69 in terms of the bidispersity @xmath66 .
we change the particle shape ( @xmath70 ) by fixing in eq .
( [ e4 ] ) one aspect ratio ( @xmath71 ) and varying the remaining one ( @xmath72 ) .
fig .
[ f3](a ) shows for @xmath73 , @xmath74 , @xmath75 and @xmath76 a similar trend : the minimum threshold bidispersity @xmath68 increases the more the shape deviates from the optimal `` brick '' one . at the same time , the fact that at fixed @xmath6 the threshold bidispersity decreases with @xmath71 , indicates that the appearance of the landau tetracritical points is favored by an increasing aspect ratio of the particles , in qualitative agreement with ref . @xcite . moreover , by fixing the aspect ratio @xmath51 , we can observe the tetracritical mole fraction as a function of the bidispersity for different values of @xmath77 , @xmath78 and @xmath79 in fig . [
f3](b ) .
the closer the shape is to the optimal `` brick '' , the wider is the difference in value of the two tetracritical mole fractions @xmath69 and , consequently , the stability regime of the @xmath4 phase .
finally , we note that no critical composition is observed if the particles are closer to the `` plate - like '' shape , i.e. if @xmath80 one finds the @xmath4 in between the @xmath12 and @xmath0 phases for every value of @xmath66 and @xmath43 ( not shown ) ; the @xmath3 phase does _ not _ occur in this case .
( color online ) ( a ) threshold bidispersity @xmath68 for the appearance of a tetracritical point as a function of the shape parameter @xmath6 for different fixed values of @xmath71 .
( b ) critical mole fraction @xmath69 as a function of the bidispersity parameter @xmath66 for a binary mixture of hard cuboids for different shape of the particles ( cf .
eq .
( [ e4 ] ) ) . ] in order to analyze proper polydispersity , and thus more realistically model the experimental system of ref .
@xcite , we extend our phase - diagram calculations to a system with @xmath81 components of cuboids . inspired by our analysis of the binary mixture and by the experiments @xcite
, we fix the aspect ratios of all species to @xmath82 and @xmath83 , such that ( i ) all species have the same shape @xmath84 and ( ii ) the size of each species is completely determined by @xmath85 .
we consider @xmath85 to be distributed according to a discretized gaussian function with average @xmath86 and standard deviation @xmath87 , where @xmath88 is the size polydispersity .
in general the calculation of a ( high - dimensional ) phase diagram of a multi - component system is a daunting task @xcite . in this case , however , it is justified to ignore fractionation ( see appendix c ) , which reduces the problem to minimizing the functional with respect to @xmath89 at fixed @xmath90 .
the resulting phase diagram in the density - polydispersity representation is shown in fig .
[ f4](a ) , featuring again @xmath12 , @xmath3 , @xmath4 , @xmath0 and @xmath54 equilibrium states and a tetracritical point at @xmath91 , which is surprisingly close to the size polydispersity in the experiments @xcite .
the strikingly large stability regime of the @xmath0 is caused by the reduced stability of @xmath54 and @xmath3 ( cf .
fig .
[ f4](b ) ) , not unlike in the binary case .
however , a direct @xmath92 transition similar to that observed in ref .
@xcite is not expected in this model due to the reentrant character of the @xmath93 phase transition ( cf .
fig .
[ f4](c ) ) .
( color online ) ( a ) phase diagram of @xmath81 components of hard cuboids ( packing fraction @xmath53 ) with aspect ratios @xmath94 and @xmath95 ( @xmath20 ) and gaussian distributed dimensions with polydispersity @xmath88 ( see text ) .
the dashed line indicates the limit of stability of the homogeneous phases with respect to smectic fluctuations . the dotted rectangles highlight ( b ) the absence of the @xmath0 phase at polydispersity @xmath96 due to the direct @xmath97 phase transition and ( c ) the reentrant character of the @xmath98 transition close to the tetracritical point ( open circle ) . ] in conclusion , by means of a mean - field theoretical approach with discrete orientations we have shown that size polydispersity strongly affects the phase behavior of boardlike particles , driving the emergence of a novel topology of the phase diagram .
this topology change is due to the appearance of landau tetracritical points , which in turn is related to a competition between the prolate `` rod - like '' ordering typical of the pure components and the oblate `` plate - like '' purely induced by the mixing . in combination with the destabilization of the @xmath54 phase , we can conclude that polydispersity dramatically increases the stability regime of the @xmath0 phase .
the usual stability limitations of @xmath0 phases , such as @xmath67 demixing of rod - plate mixtures and ordering into smectics , are therefore overcome in the present system . although this work focuses on a particular value of the particles dimensions , its predictions hold for a more general choice of the relevant parameters , as reported in fig .
[ f3 ] .
moreover , we do not expect the homogeneous phase behavior to be crucially dependent on the form of the interaction ( cuboidal ) , on the contrary it should be qualitatively similar to other excluded - volume interactions with the _ same _ symmetry ( e.g. spheroid , spheroplatelet ) . finally , it is tempting to consider this work in the perspective of stabilizing @xmath0 thermotropic liquid crystals . in this case , the soft - core character of the inter - molecular interactions does not allow for a univocal definition of `` shape '' , and van der waals forces can significantly influence the phase diagram .
nonetheless , it is widely accepted that hard - core models contain the essential physical ingredients for a first - approximation description of the structure of a molecular or colloidal fluid @xcite . following this interpretation scheme ,
it is intriguing to wonder whether it is possible to enhance the @xmath0 stability by considering two- or multi - component mixtures of molecules with biaxial symmetry and different size .
we hope our findings will stimulate further research in this direction .
this work is financed by a nwo - vici grant and is part of the research program of fom , which is financially supported by nwo . |
one of most interesting topics in nonlinear dynamics is the issue of leaking or escaping orbits ( e.g. , @xcite ) . over the years
many studies have been devoted on hamiltonian systems with escapes ( e.g. , @xcite ) .
however the topic of escaping orbits in hamiltonian systems is by far less explored than the closely related problem of chaotic scattering , where a test particle coming from infinity approaches and then scatters off a complex potential ( e.g. , @xcite ) .
undoubtedly , one of the most important topics in dynamical astronomy is the classical restricted three - body problem ( rtbp ) .
the rtbp describes the motion of a test body with an infinitesimal mass moving under the gravitational effects of a two primary bodies with finite masses which move in circular orbits around their common center of mass @xcite .
it is assumed that the third test body does not influence the motion of the two primary bodies .
this problem has many applications in celestial mechanics , stellar systems , artificial satellites , galactic dynamics , chaos theory and molecular physics and therefore is still a very active and stimulating field of research .
generally , the shapes of the two primary bodies in the classical version of the rtbp are assumed to be spherically symmetric .
however , we found in nature that several celestial bodies , such as saturn and jupiter are sufficiently oblate @xcite or even triaxial
. furthermore , irregular shapes are also possible , especially in the case of minor planets ( e.g. , ceres ) and meteoroids @xcite . the oblateness or triaxiality of a celestial body can produce perturbation deviations from the two - body motion .
the study of oblateness coefficient includes the series of works of @xcite by considering the more massive primary as an oblate spheroid with its equatorial plane co - incident with the plane of motion of the primaries .
all planets of the solar system are in fact not spherically symmetric but oblate spheroidals .
consequently the exact shape of the planets and their oblateness must be taken into account when someone wants to model a space flight mission . in the work of @xcite
it was proved that the theoretical data from the celestial systems saturn - tethys - satellite and saturn - dione - satellite are much more accurate with respect with the corresponding observational data when the oblateness of saturn is taken into consideration .
similarly , the oblateness of the neptune was included in the work of @xcite regarding the dynamics of a spacecraft in the neptune - triton system . in the present paper
we continue the work initiated in @xcite following the same numerical techniques .
our aim is to numerically investigate how the oblateness coefficient influences the nature of orbits in the restricted three - body problem .
the paper is organized as follows : in section [ mod ] we describe the properties of the considered dynamical model along with some necessary theoretical details .
all the computational methods we used in order to determine the character of orbits are explained in section [ cometh ] . in the following section , we conduct a thorough numerical investigation revealing the overall orbital structure ( bounded regions and basins of escape / collision ) of the system and how it is affected by the value of the oblateness coefficient with respect to the total orbital energy .
our paper ends with section [ conc ] , where the main conclusions of this work are given .
it would be very illuminating to briefly recall the basic properties and some aspects of the planar circular restricted three - body problem @xcite .
the two primaries move on circular orbits with the same kepler frequency around their common center of gravity , which is assumed to be fixed at the origin of the coordinates .
the third body ( test particle with mass much smaller than the masses of the primaries ) moves in the same plane under the gravitational field of the two primaries .
the non - dimensional masses of the two primaries are @xmath2 and @xmath3 , where @xmath4 , with @xmath5 is the mass ratio .
we consider an intermediate value of the mass ratio , that is @xmath6 .
this value remains constant throughout the paper .
we choose as a reference frame a rotating coordinate system where the origin is at ( 0,0 ) , while the centers @xmath7 and @xmath8 of the two primary bodies are located at @xmath9 and @xmath10 , respectively .
the total time - independent effective potential according to @xcite is @xmath11 where @xmath12 @xmath13 @xmath14 are the distances to the respective primaries and the angular velocity @xmath15 , while @xmath16 is the oblateness coefficient which is defined as @xmath17 where @xmath18 and @xmath19 are the equatorial and polar radius , respectively of the oblate primary , while @xmath20 is the distance between the centers of the two primaries .
we consider values of the oblateness coefficient is in the interval @xmath21 $ ] ( see e.g. , @xcite ) , while we study the effect of oblateness up to the linear coefficient @xmath22 only @xcite .
the human species stands on the edge of a new frontier , the transition from a planet - bound to a space - faring civilization .
the expansion into the solar system , in terms of dynamics of artificial satellites , requires the formulation of new models that include the effects of some of the perturbing forces on the satellite .
therefore , the problem is not only of mathematical interest but has astrophysical applications .
the most striking example of perturbations arising from the oblateness in the solar system is the orbit of the fifth satellite of jupiter , amalthea .
this planet is so oblate and the satellite s orbit is so small that its line of apsides advances about @xmath23 in a year ( e.g. , @xcite ) .
it should be pointed out that in general terms the values of the oblateness in the solar system are relatively low ( see e.g. , @xcite for updated values of the oblateness coefficient ) .
there are however close - in extrasolar planetary systems with hot jupiters in which the oblateness coefficient varies in the interval [ 0.001 , 0.01 ] . in this work
we consider even higher values of the oblateness coefficient @xmath24 because in the near future it is possible new highly oblate extrasolar planets to be discovered .
the scaled equations of motion describing the motion of the test body in the corotating frame read @xcite @xmath25 @xmath26 the dynamical system ( [ eqmot ] ) admits the well know jacobi integral @xmath27 where @xmath28 and @xmath29 are the velocities , while @xmath30 is the numerical value of the orbital energy which is conserved and defines a three - dimensional invariant manifold in the total four - dimensional phase space .
thus , an orbit with a given value of it s energy integral is restricted in its motion to regions in which @xmath31 , while all other regions are forbidden to the test body .
it is widely believed that @xmath32 is the only independent integral of motion for the rtbp system @xcite .
the value of the energy @xmath30 is related with the jacobi constant by @xmath33 .
the dynamical system has five equilibria known as lagrangian points @xcite at which @xmath34 three of them , @xmath35 , @xmath36 , and @xmath37 , are collinear points located in the @xmath38-axis .
the central stationary point @xmath35 is a local minimum of the potential @xmath39 .
the stationary points @xmath36 and @xmath37 are saddle points .
let @xmath36 located at @xmath40 , while @xmath37 be at @xmath41 .
the points @xmath42 and @xmath43 on the other hand , are local maxima of the gravitational potential , enclosed by the banana - shaped isolines .
the lagrangian points are very important especially for astronautical applications .
this can be seen in the sun - jupiter system where several thousand asteroids ( collectively referred to as trojan asteroids ) , are in orbits of equilibrium points . in practice
these lagrangian points have proven to be very useful indeed since a spacecraft can be made to execute a small orbit about one of these equilibrium points with a very small expenditure of energy @xcite .
the projection of the four - dimensional phase space onto the configuration ( or position ) space @xmath0 is called the hill s regions and is divided into three domains : ( i ) the interior region for @xmath44 ; ( ii ) the exterior region for @xmath45 and @xmath46 ; ( iii ) the forbidden regions .
the boundaries of these hill s regions are called zero velocity curves ( zvcs ) because they are the locus in the configuration @xmath0 space where the kinetic energy vanishes . [ cols="<,^,^,^,^",options="header " , ] the values of the jacobi integral at the five lagrangian points @xmath47 , @xmath48 are critical energy levels and they are denoted as @xmath49 ( note that @xmath50 ) .
the structure of the equipotential surfaces strongly depends on the value of the energy . in particular , there are five distinct cases * @xmath51 : all necks are closed , so there are only bounded and collisional basins .
* @xmath52 : only the neck around @xmath35 is open thus allowing orbits to move around both primaries . * @xmath53 : the neck around @xmath36 is open , so orbits can enter the exterior region and escape form the system .
* @xmath54 : the necks around both @xmath36 and @xmath37 are open , so orbits can escape through two different escape channels .
* @xmath55 : the banana - shaped forbidden regions disappear and therefore , motion over the entire configuration @xmath0 space is possible . in fig .
[ isos](a - d ) we present for @xmath56 the structure of the first four possible hill s region configurations .
we observe in fig .
[ isos]d the two openings ( exit channels ) at the lagrangian points @xmath36 and @xmath37 through which the test body can enter the exterior region and then leak out .
in fact , we may say that these two exits act as hoses connecting the interior region of the system where @xmath44 with the outside world " of the exterior region .
the position of the lagrangian points as well as the critical values of the energy are functions of the oblateness coefficient @xmath16 ( e.g. , @xcite ) . in table
[ table1 ] we provide the location of the lagrangian points and the critical values of the total orbital energy when @xmath57 .
the motion of the test third body is restricted to a three - dimensional surface @xmath58 , due to the existence of the jacobi integral . with polar
coordinates @xmath59 in the center of the mass system of the corotating frame the condition @xmath60 defines a two - dimensional surface of section , with two disjoint parts @xmath61 and @xmath62 .
each of these two parts has a unique projection onto the configuration @xmath0 space . in order to explore the orbital structure of the system we need to define samples of initial conditions of orbits
whose properties will be identified .
fore this purpose , we define for several values of the total orbital energy @xmath30 , as well as for the oblateness coefficient @xmath16 dense uniform grids of @xmath63 initial conditions regularly distributed on the configuration @xmath0 space inside the area allowed by the value of the total orbital energy . following a typical approach , the orbits are launched with initial conditions inside a certain region , called scattering region , which in our case is a square grid with @xmath64 . inside the system
s disk with radius @xmath65 , while the motion is unbounded and the numerical integration stops when the test body crosses the system s disk with velocity pointing outwards . collision or ( crash ) with one of the primaries occurs when the test body crosses the disk of radii @xmath66 and @xmath67 of the primaries . ] in the rtbp system the configuration space extends to infinity thus making the identification of the type of motion of the test body for specific initial conditions a rather demanding task .
there are three possible types of motion for the test body : ( i ) bounded motion around one of the primaries , or even around both ; ( ii ) escape to infinity ; ( iii ) collision into one of the two primaries .
now we need to define appropriate numerical criteria for distinguishing between these three types of motion .
the motion is considered as bounded if the test body stays confined for integration time @xmath68 inside the system s disk with radius @xmath69 and center coinciding with the center of mass origin at @xmath70 .
obviously , the higher the values of @xmath68 and @xmath69 the more plausible becomes the definition of bounded motion and in the limit @xmath71 the definition is the precise description of bounded motion in a finite disk of radius @xmath69 .
consequently , the higher these two values , the longer the numerical integration of initial conditions of orbits lasts . in our calculations
we choose @xmath72 and @xmath65 ( see fig . [ crit ] ) as in @xcite and @xcite .
we decided to include a relatively high disk radius @xmath73 in order to be sure that the orbits will certainly escape from the system and not return back to the interior region .
furthermore , it should be emphasized that for low values of @xmath68 the fractal boundaries of stability islands corresponding to bounded motion become more smooth .
moreover , an orbit is identified as escaping and the numerical integration stops if the test body body intersects the system s disk with velocity pointing outwards at a time @xmath74 .
finally , a collision with one of the primaries occurs if the test body , assuming it is a point mass , crosses the disk with radius @xmath75 around the primary . for the larger oblate primary we choose @xmath76
generally , it is assumed that the radius of a celestial body ( e.g. , a planet ) is directly proportional to the cubic root of its mass .
therefore , for the sake of simplicity of the numerical calculations we adopt the simple relation between the radii of the primaries @xmath77 which was introduced in @xcite . in @xcite it was shown that the radii of the primaries influence the area of collision and escape basins .
as it was stated earlier , in our computations , we set @xmath78 time units as a maximum time of numerical integration .
the vast majority of escaping orbits ( regular and chaotic ) however , need considerable less time to escape from the system ( obviously , the numerical integration is effectively ended when an orbit moves outside the system s disk and escapes ) .
nevertheless , we decided to use such a vast integration time just to be sure that all orbits have enough time in order to escape .
remember , that there are the so called sticky orbits " which behave as regular ones during long periods of time . here
we should clarify , that orbits which do not escape after a numerical integration of @xmath78 time units are considered as non - escaping or trapped .
the equations of motion ( [ eqmot ] ) for the initial conditions of all orbits are forwarded integrated using a double precision bulirsch - stoer ` fortran 77 ` algorithm ( e.g. , @xcite ) with a small time step of order of @xmath79 , which is sufficient enough for the desired accuracy of our computations .
here we should emphasize , that our previous numerical experience suggests that the bulirsch - stoer integrator is both faster and more accurate than a double precision runge - kutta - fehlberg algorithm of order 7 with cash - karp coefficients . throughout all our computations , the jacobian energy integral ( eq .
( [ ham ] ) ) was conserved better than one part in @xmath80 , although for most orbits it was better than one part in @xmath81 . for collisional orbits where the test body moves inside a region of radius @xmath79 around one of the primaries
the lemaitre s global regularization method is applied .
the main numerical task is to classify initial conditions of orbits in the @xmath61 part instead of the @xmath62 part simply because in @xcite we seen that it contains more interesting orbital content .
] of the surface of section @xmath60 into three categories : ( i ) bounded orbits ; ( ii ) escaping orbits and ( iii ) collisional orbits .
moreover , two additional properties of the orbits will be examined : ( i ) the time - scale of collision and ( ii ) the time - scale of the escapes ( we shall also use the terms escape period or escape rates ) . in this work
we shall explore these dynamical quantities for various values of the total orbital energy , as well as for the oblateness coefficient @xmath16 .
in particular , we choose three energy levels which correspond to the last three hill s regions configurations .
the first two hill s regions configurations contain only bounded and collisional orbits , so the orbital content is not so interesting . in the following color - coded grids ( or orbit type diagrams - otds )
each pixel is assigned a color according to the orbit type .
thus the initial conditions of orbits are classified into bounded orbits , unbounded or escaping orbits and collisional orbits . in this special type of poincar surface of section
the phase space emerges as a close and compact mix of escape basins , collisional basins and stability islands .
our numerical calculations indicate that apart from the escaping and collisional orbits there is also a considerable amount of non - escaping orbits . in general terms
, the majority of non - escaping regions corresponds to initial conditions of regular orbits , where an adelphic integral of motion is present , restricting their accessible phase space and therefore hinders their escape .
our numerical exploration begins in the energy region iii @xmath82 and we choose the energy level @xmath83 . in fig .
[ e1 ] the otd decompositions of the @xmath61 part of the surface of section @xmath60 reveal the orbital structure of the configuration @xmath0 space for four values of the oblateness coefficient .
the black solid lines denote the zvc , while the inaccessible forbidden regions are marked in gray .
the color of a point represents the orbit type of a test body which has been launched with pericenter position at @xmath0 . when the oblateness coefficient is zero @xmath84 we observe in fig .
[ e1]a the followings : ( i ) around the centers of the two primaries there are stability islands , while in the exterior region another ring - shaped stability island is present .
the stability islands located in the interior region correspond to regular orbits around one of the primary bodies , while the stability island in the exterior region is formed by initial conditions of regular orbits that circulate around both primaries ; ( ii ) the region in the boundaries of the stability islands of the interior region is mainly occupied by orbits which lead to collision ; ( iii ) near the center of primary 1 we can identify a small hole which contains a mixture of collisional and escaping orbits ; ( iv ) outside the forbidden regions there is a well - formed circular basin of escaping orbits .
it is evident in fig .
[ e1]b that even a very low values of the oblateness coefficient @xmath85 is able to influence the orbital content . in particular , there are two main differences with respect to the case of zero oblateness shown in fig .
first the area of the stability region around the center of the oblate primary is smaller , while the central hole contains now only initial conditions of collisional orbits .
moreover , the collisional basin around primary 1 is larger in size .
the orbital structure changes even further when @xmath56 as we can see in fig .
almost all the area around the oblate primary 1 is occupied by initial conditions of collisional orbits , while the corresponding stability island splits into two small parts .
it seems that so far the region around smaller primary 2 as well as the exterior region remain almost unperturbed by the shift on the value of the oblateness coefficient .
things however change drastically when the larger primary is highly oblate with @xmath86 .
it is seen in fig .
[ e1]d that the entire interior region around primary 1 is dominated by a large collisional basin .
the exterior region on the other hand , is highly fractal , while the stability island of regular orbits circulating around both primaries is now absent . at least at this energy level
our computations suggest that the increase in the value of the oblateness coefficient does not affect the stability island around primary 2 , however at the highest studied value the collisional basin around this stability island is weakened .
it should be noted that when @xmath86 several spiral collisional basins are inside the fractal exterior region . the following fig . [
e1 t ] shows how the escape and collisional times of orbits are distributed on the configuration @xmath0 space for the four values of the oblateness coefficient discussed in fig .
light reddish colors correspond to fast escaping / collional orbits , dark blue / purple colors indicate large escape / collional times , while white color denote stability islands of regular motion .
note that the scale on the color bar is logarithmic . inspecting the spatial distribution of various different ranges of escape time ,
we are able to associate medium escape time with the stable manifold of a non - attracting chaotic invariant set , which is spread out throughout this region of the chaotic sea , while the largest escape time values on the other hand , are linked with sticky motion around the stability islands of the two primary bodies . as for the collisional time we see that orbits with initial conditions very close to the vicinity of the center of the oblate primary 1 collide with it almost immediately , within the first time step of the numerical integration , while this phenomenon is not observed for the case of primary body 2 which is not oblate ( see also @xcite ) . looking more carefully fig .
[ e1t]d we observe that when @xmath86 the area of the stability region around primary 2 ( the only stability region that survives ) is smaller with respect to the tree previous cases shown in figs .
[ e1t](a - c ) .
thus we may say that high enough values of the oblateness coefficient influence also the stability region corresponding to smaller spherically symmetric primary .
we continue our investigation in the energy region iv @xmath87 .
it is evident from the critical energy values given in table [ table1 ] that there is not a single energy level which we can use for all four cases regarding the value of the oblateness coefficient .
therefore for the first three cases , that is @xmath88 , we choose the value @xmath89 , while for the last case @xmath90 we use the energy level @xmath91 .
the orbital structure of the configuration @xmath0 space is unveiled in fig .
[ e2 ] through the otd decompositions of the @xmath61 part of the surface of section @xmath60 .
we observe in fig .
[ e2]a where @xmath92 that in this case the stability island in the exterior region is no longer present .
the vast majority of the exterior region is occupied by initial conditions of escaping orbits , while inside the extended escape basin we identify delocalized initial conditions of orbits that collide to the oblate primary . on the other hand , the initial conditions of orbits which collide to primary 2 form well - defined spiral basins .
the interior region is almost the same with that discussed earlier in fig .
[ e1]a , however there is a minor difference ; the collisional basin around the stability island of primary 2 is absent now . with the introduction of the oblateness coefficient it is seen in figs .
[ e2](b - d ) that the interior region displays exactly the same behaviour as in the previous case . in the exterior region two phenomena take place as the value of the oblateness coefficient increases : ( i ) the amount of the escaping orbits decreases while at the same time the portion of the collisional orbits to oblate primary 1 increases and ( ii ) the fractality of the exterior region gradually decreases with increasing @xmath16 and several escape and collisional basins emerge . the distribution of the escape and collisional times of orbits on the configuration space is shown in fig .
one may observe that the results are very similar to those presented earlier in fig .
[ e1 t ] , where we found that orbits with initial conditions inside the escape and collisional basins have the smallest escape / collision rates , while on the other hand , the longest escape / collisional rates correspond to orbits with initial conditions in the fractal regions of the otds .
our calculations reveal , and this can be seen better in figs .
[ e2t](a - d ) , that in this energy region the oblateness coefficient does not affect the size of the stability island of motion around primary 2 .
the last case under consideration involves the results in the energy region v where the entire configuration space is available for motion since the forbidden regions completely disappear .
once more , all the different aspects of the numerical approach remain exactly the same as in the two previously studied cases .
[ e3 ] presents the orbital structure of the configuration space through the otd decompositions of the @xmath61 part of the surface of section @xmath60 .
when both primaries are spherically symmetric , that is when @xmath92 , we see in fig .
[ e3]a that most of the configuration @xmath0 plane is highly fractal , while a basin of collisional orbits is located around the stability island of primary 1 . around primary 2
there is another stability island but now the corresponding collisional basin identified in energy region iii is not present now .
in addition , the central hole inside the stability island of primary 1 still survives . at the lowest value of the oblateness coefficient @xmath85 the observed changes in fig .
[ e3]b are the same with those reported earlier in fig .
[ e2]b . the same applies in the case for @xmath56 where the results are shown in fig .
[ e3]c . finally in fig .
[ e3]d where we have the scenario with the highest possible value of the oblateness coefficient @xmath90 it is seen that the vast majority of the @xmath0 plane is occupied by well - formed escape basins and collisional basins , while the fractal areas are confined to the boundaries between the several basins .
it should be pointed out that in this case the area of the stability region around primary 2 is considerably reduced .
thus one may reasonably conclude that in the energy region v and for high enough values of @xmath16 the stability region around spherically primary 2 is highly affected by the oblateness coefficient . in fig .
[ e3 t ] we depict the distribution of the escape and collisional times of orbits on the configuration space .
one can see similar outcomes with that presented in the two previous subsections . at this point
, we would like to emphasize that the basins of escape can be easily distinguished in fig .
[ e3 t ] , being the regions with intermediate colors indicating fast escaping orbits . indeed
, our numerical computations suggest that orbits with initial conditions inside these basins need no more than 10 time units in order to escape from the system .
furthermore , the collisional basins are shown with reddish colors where the corresponding collisional time is less than one time unit of numerical integration . the otds shown in figs .
[ e1 ] , [ e2 ] and [ e3 ] have both fractal and non - fractal ( smooth ) boundary regions which separate the escape basins from the collisional basins .
such fractal basin boundaries is a common phenomenon in leaking hamiltonian systems ( e.g. , @xcite ) . in the rtbp system
the leakages are defined by both escape and collision conditions thus resulting in three exit modes . however , due to the high complexity of the basin boundaries , it is very difficult , or even impossible , to predict in these regions whether the test body ( e.g. , a satellite , asteroid , planet etc ) collides with one of the primary bodies or escapes from the dynamical system .
the color - coded otds in the configuration @xmath0 space provide sufficient information on the phase space mixing however , for only a fixed value of the energy integral and also for orbits that traverse the surface of section retrogradely .
hnon @xcite , introduced a new type of plane which can provide information not only about stability and chaotic regions but also about areas of bounded and unbounded motion using the section @xmath93 , @xmath94 ( see also @xcite ) .
in other words , all the initial conditions of the orbits of the test particles are launched from the @xmath38-axis with @xmath95 , parallel to the @xmath96-axis @xmath97 . consequently ,
in contrast to the previously discussed types of planes , only orbits with pericenters on the @xmath38-axis are included and therefore , the value of the energy @xmath30 can now be used as an ordinate . in this way , we can monitor how the energy influences the overall orbital structure of our dynamical system using a continuous spectrum of energy values rather than few discrete energy levels . in fig .
[ xe ] we present the orbital structure of the @xmath1 plane for four values of the oblateness coefficient when @xmath98 $ ] , while in fig .
[ xet ] the distribution of the corresponding escape and collision times of the orbits is depicted .
we can observe the presence of several types of regular orbits around the two primary bodies .
being more precise , on both sides of the primaries we identify stability islands corresponding to both direct ( counterclockwise ) and retrograde ( clockwise ) quasi - periodic orbits .
it is seen that a large portion of the exterior region , that is for @xmath45 and @xmath46 , a large portion of the @xmath1 plane is covered by initial conditions of escaping orbits however , at the left - hand side of the same plane two stability islands of regular orbits that circulate around both primaries are observed .
additional numerical calculations reveal that for much lower values of @xmath38 @xmath99 these two stability islands are joined and form a crescent - like shape .
we also see that collisional basins to oblate primary 1 leak outside the interior region , mainly outside @xmath36 , and create complicated spiral shapes in the exterior region . on the other hand ,
the thin bands represent initial conditions of orbits that collide with primary body 2 are much more confined .
it should be pointed out that in the blow - ups of the diagram several additional very small islands of stability have been identified . as the value of the oblateness coefficient increases the structure of the @xmath1 planes exhibits the following changes : ( i ) the collisional basins to oblate primary 1 significantly grow and for @xmath86 they dominate ; ( ii ) the area of the stability islands around primary 1 reduces and at the highest studied value of @xmath16 they completely disappear . according to broucke s classification @xcite
the periodic orbits around the primaries belong to the families @xmath100 ( at the left side of the primary ) and @xmath101 ( at the right side of the primary ) , while @xcite proved for the planar hill s problem that the stability regions of the @xmath100 family are more stable than those of the @xmath101 family .
our computations show that both families disappear for large values of the oblateness coefficient ; ( iii ) the stability islands around primary body 2 are almost unperturbed , at least in the interval @xmath102 $ ] , by the shifting on the value of the oblateness .
the phenomenon that stability islands can appear and disappear as a dynamical parameter is changed has also been reported in earlier paper ( e.g. , @xcite ) ; ( iv ) inspecting figs .
[ xet](a - d ) we see that orbits with initial conditions very close to the center of the oblate primary 1 collide almost immediately with it , while their portion ( or in other words the thickness of the vertical basin ) increases for larger values of the oblateness coefficient @xmath16 ; ( v ) another interesting phenomenon is the fact that as the primary body 1 becomes more and more oblate the fractality of the @xmath1 plane reduces and the boundaries between escaping and collisional basins appear to become smoother .
it should be emphasized that the fractality of the structures was not measured by computing the corresponding fractal dimension . when we state that an area is fractal we mean that it has a fractal - like geometry .
it would be very informative to monitor the evolution of the percentages of all the different types of orbits as a function of the total orbital energy @xmath30 for the @xmath1 planes shown in figs .
[ xe](a - d ) .
our results are presented in figs .
[ pxe](a - d ) .
we see that in all four cases the rate of the escaping orbits starts at about 80% for @xmath103 and gradually reduces until about 20% for @xmath104 . for @xmath105
it suddenly increases but this increase is controlled by the oblateness coefficient . in particular , the higher the value of @xmath16 the lower the increase of the escaping percentage in the energy interval @xmath106 $ ] . in the same interval however , according to fig .
[ xet]a escaping orbits dominate the @xmath1 plane when @xmath92 .
the evolution of the percentage of regular orbits displays similar patterns in the energy interval @xmath107 $ ] however , the corresponding rates are reduced with increasing @xmath16 . for larger values of the energy
the percentage of bounded orbits fluctuate and in general terms we may say that again the oblateness coefficient has an influence .
the percentage of collisional orbits to oblate primary 1 has a monotone behaviour in the energy interval @xmath108 $ ] , while for @xmath105 it increases .
this increase becomes stronger and stronger as the value of the oblateness coefficient increases . for large enough values of the energy the rate of collisional orbits to primary 1 decreases in all four cases , while at the same time escaping orbits is the most populated type of orbits .
the percentage of collisional orbits to primary 2 on the other hand , is extremely low , while it peaks only in the energy interval @xmath109 $ ] .
once more , the peaks in this energy range are influenced by the value of @xmath16 ; the larger the oblateness coefficient the lower the peaks . another interesting aspect would be to reveal how the oblateness coefficient influences the escape as well as the collision time of orbits .
the evolution of the average value of the escape time @xmath110 of orbits as a function of the total orbital energy is given in fig .
[ stats]a .
it is evident , especially in the energy interval @xmath106 $ ] , that as the value of the oblateness coefficient increases the escape rates of orbits are reduced . in the same vein in fig .
[ stats]b we present the evolution of the average collision time @xmath111 of orbits that collide to oblate primary 1 .
once more we observe that the collision time of orbits decreases with increasing value of @xmath16 , especially in the energy interval @xmath112 $ ] .
additional numerical calculations , not shown here , indicate that the oblateness coefficient also influences the collision rates of orbits which collide to spherical primary 2 .
in particular , there are energy ranges in which the collision rates to primary 2 are reduced with increasing @xmath16 , however the diagrams are not so clear as in fig .
[ stats]b , so we decided not to include them . before closing this section
we would like to add that the particular value of the mass ratio @xmath3 does not really change the qualitative nature of the numerical outcomes presented in this section . indeed after conducting some additional calculations with larger and lower values of @xmath3 we concluded to the same results .
the parameters which mostly influence the orbital dynamics are the total energy and of course the oblateness coefficient .
the aim of this work was to reveal how the oblateness coefficient influences the character of orbits in the classical planar circular restricted three - body problem . after conducting an extensive and thorough numerical investigation we managed to distinguish between bounded , escaping and collisional orbits and we also located the basins of escape and collision , finding also correlations with the corresponding escape and collision times .
our numerical results strongly suggest that the oblateness coefficient plays a very important role in the nature of the test s body motion under the gravitational field of two primaries . to our knowledge , this is the first detailed and systematic numerical analysis on the influence of the oblateness coefficient on the character of orbits and this is exactly the novelty and the contribution of the current work . for several values of the oblateness coefficient in the last three hill s regions configurations we defined dense uniform grids of @xmath63 initial conditions regularly distributed on the @xmath61 part of the configuration @xmath0 plane inside the area allowed by the value of the total orbital energy .
all orbits were launched with initial conditions inside the scattering region , which in our case was a square grid with @xmath64 . for the numerical integration of the orbits in each type of grid
, we needed about between 11 hours and 5 days of cpu time on a pentium dual - core 2.2 ghz pc , depending on the escape and collisional rates of orbits in each case .
for each initial condition , the maximum time of the numerical integration was set to be equal to @xmath78 time units however , when a particle escaped or collided with one of the two primaries the numerical integration was effectively ended and proceeded to the next available initial condition . in this article
we provide quantitative information regarding the escape and collisional dynamics in the restricted three - body problem with oblateness .
the main numerical results of our research can be summarized as follows : 1 .
it was observed that as the value of the oblateness coefficient increases the size of the stability islands around primary 1 is reduced and at relatively high values of the oblateness coefficient there is no indication of regular motion around the oblate primary , while at the same time the stability region around spherical primary 2 is also reduced .
the collisional basins to primary 1 were found to significantly increase in size with increasing value of the oblateness coefficient , while the basins formed by initial conditions of orbits which collide to the primary 2 are practically unaffected by the shifting on the value of the oblateness coefficient .
it was detected that in the vicinity of the center of the oblate primary 1 a portion of orbits collide to the same primary almost immediately .
furthermore , the amount of these type of orbits increases with increasing oblateness . on the other hand this behaviour do not apply to the case of the spherical primary 2 .
our calculations reveal that as the primary body 1 becomes more and more oblate the fractality of the planes is reduced and the boundaries between the escaping and collisional basins appear to become smoother .
we presented evidence that the oblateness coefficient also influences the escape as well as the collision time of the orbits .
in particular , both types of times are reduced as the value of the oblateness coefficient increases .
judging by the detailed and novel outcomes we may say that our task has been successfully completed . we hope that the present numerical analysis and the corresponding results to be useful in the field of escape dynamics in the restricted three - body problem with oblateness .
the outcomes as well as the conclusions of the present research are considered , as an initial effort and also as a promising step in the task of understanding the escape mechanism of orbits in this interesting version of the classical three - body problem .
taking into account that our results are encouraging , it is in our future plans to properly modify our dynamical model in order to expand our investigation into three dimensions and explore the entire six - dimensional phase thus revealing the influence of the oblateness coefficient on the orbital structure .
i would like to express my warmest thanks to the anonymous referee for the careful reading of the manuscript and for all the apt suggestions and comments which allowed us to improve both the quality and the clarity of the paper .
kalantonis , v.s . ,
markellos , v.v . , perdios , e.a .
: computing periodic orbits of the three - body problem : effective convergence of newton s method on the surface of section .
space sci . * 298 * , 441 - 451 ( 2005 ) kalantonis , v.s . ,
douskos , c.n . ,
perdios , e.a . : numerical determination of homoclinic and heteroclinic orbits as collinear equilibria in the restricted three - body problem with oblateness .
astron . * 94 * , 135 - 153 ( 2006 ) kalantonis , v.s . , perdios , e.a . ,
perdiou , a.e . : the sitnikov family and the associated families of 3d periodic orbits in the photogravitational rtbp with oblateness .
space sci . * 315 * , 323 - 334 ( 2008 ) markellos , v.v . ,
papadakis , k.e . , perdios , e.a . :
non - linear stability zones around triangular equilibria in the plane circular restricted three - body problem with oblateness .
space sci .
* 245 * , 157164 ( 1996 ) sharma , r.k .
: the linear stability of libration points of the photogravitational restricted three - body problem when the smaller primary is an oblate spheroid .
space sci . * 135 * , 271281 ( 1987 ) sharma , r.k .
, subba rao , p.v .
: stationary solutions and their characteristic exponents in the restricted three - body - problem when the more massive primary is an oblate spheroid .
* 13 * , 137 - 149 ( 1976 ) sharma , r.k .
, subba rao , p.v . : on finite periodic orbits around the equilibrium solutions of the planar restricted three - body problem . in : bhatnagar ,
space dynamics and celestial mechanics , pp . 7185 .
d. reidel publishing company , dordrecht ( 1986 ) singh , j. , leke , o. : effect of oblateness , perturbations , radiation and varying masses on the stability of equilibrium points in the restricted three - body problem .
space sci . * 344 * , 51 - 61 ( 2013 ) | we numerically investigate the case of the planar circular restricted three - body problem where the more massive primary is an oblate spheroid .
a thorough numerical analysis takes place in the configuration @xmath0 and the @xmath1 space in which we classify initial conditions of orbits into three categories : ( i ) bounded , ( ii ) escaping and ( iii ) collisional .
our results reveal that the oblateness coefficient has a huge impact on the character of orbits . interpreting the collisional motion as leaking in the phase space we related our results to both chaotic scattering and the theory of leaking hamiltonian systems .
we successfully located the escape as well as the collisional basins and we managed to correlate them with the corresponding escape and collision times .
we hope our contribution to be useful for a further understanding of the escape and collision properties of motion in this interesting version of the restricted three - body problem .
= 1 |
underdoped cuprate superconductors exhibit a `` pseudogap '' phase that becomes apparent below a temperature @xmath3 , where @xmath3 decreases with increased doping , approaching the superconducting transition temperature @xmath4 for hole concentrations beyond optimal ( where optimal doping corresponds to maximum @xmath4 ) @xcite .
it has become common to discuss `` competing '' orders that may be responsible for the pseudogap and which may compete with superconductivity .
there are certainly many good examples of systems where the transition temperature for one type of order , such as charge density wave ( cdw ) or spin density wave ( sdw ) order , drops to zero at the same point where the superconducting @xmath4 reaches its maximum .
for example , pressure causes the cdw transitions in tas@xmath5 to decrease from 550 k towards zero , while the superconducting @xmath4 rises to 5 k , and continues in the absence of cdw order at higher pressures @xcite . at this meeting
, nez - regueiro reviewed a variety of related systems in which an order such as cdw or sdw competes for fermi surface area with superconductivity @xcite .
a rather different sort of picture has emerged from studies of @xcite . here , sdw and cdw orders appear to be intertwined with a spatially - modulated superconductivity , referred to as a pair - density - wave ( pdw ) superconductor @xcite .
antiferromagnetic and pairing correlations seem to develop in a cooperative fashion , forming a spatially self - organized pattern .
such behavior has motivated the perspective that intertwined orders are a common feature of the cuprate superconductors @xcite .
recent observations of cdw order in @xcite have led to renewed discussions of competing order . while many questions remain
, it appears that the concept of intertwined order is relevant to the interpretation of these results @xcite .
the nominal fermi surface of an underdoped cuo@xmath5 plane forms a hole pocket about @xmath6 . in the superconducting state
, the @xmath1-wave pairing gap has a node near @xmath7 , with the gap maxima occurring in the vicinity of @xmath8 and @xmath9 ( known as the antinodal regions ) . in the temperature range above @xmath4 but below @xmath3 ,
the superconducting gap closes along a finite arc about the nodal point , but a gap remains in the antinodal region @xcite .
the energy scale of the pseudogap is quite similar to that of the superconducting gap , just as the magnitude of @xmath3 is similar to @xmath4 .
is the pseudogap associated with some type of competing order ?
there have certainly been proposals of various specific orders intended to explain the pseudogap @xcite . in a conventional picture , a well - defined fermi
liquid would be present at @xmath10 , and the pseudogap would appear and grow as the competing order develops .
a problem with such a scenario is that there is no fermi liquid state at @xmath10 .
optical conductivity studies demonstrate the absence of a drude peak , and hence the absence of coherent quasiparticles @xcite .
the appearance of coherent states , associated with the nodal arc states , occurs only on cooling below @xmath3 .
this happens in parallel with the development of antiferromagnetic spin correlations , as indicated by the bulk spin susceptibility @xcite .
thus , charge coherence and spin correlations appear to develop at the same time , suggesting cooperation rather that competition .
the case of provides an example of how spins and charges can organize themselves in an intertwined fashion that enables strong pairing of holes . after the initial discovery of high - temperature superconductivity @xcite , exploration of the phase diagram revealed a sharp minimum of @xmath4 at the hole concentration @xmath11 @xcite .
such a strong dip in @xmath4 is not observed in chemically - similar @xcite .
it was quickly demonstrated that the unusual behavior in is associated with a phase transition to the low - temperature tetragonal ( ltt ) phase at a temperature of @xmath12 k @xcite .
when single crystals became available , neutron diffraction measurements on the related system la@xmath13nd@xmath14sr@xmath15cuo@xmath16 , which also exhibits the ltt phase , found the presence of charge and spin stripe order @xcite .
the occurrence of stripe order was eventually confirmed in @xcite and also in @xcite . despite the absence of bulk ltt order ,
very recent x - ray scattering measurements have demonstrated the presence of charge stripe order in la@xmath17sr@xmath15cuo@xmath16 @xcite , where spin stripe order had previously been detected @xcite . the occurrence of charge and spin stripe order was predicted long before it was detected @xcite .
these analyses properly captured the fact that the charge stripes act as antiphase domain walls for the spin stripes , so that wave vectors @xmath18 for charge order and @xmath19 for spin order ( with the latter measured relative to @xmath20 ) are related by @xmath21 . where they disagree with experiment is in predicting that the charge stripes should be insulating , with one hole per cu site along a stripe , and with a period @xmath22 ; experiment finds @xmath23 @xcite , corresponding to about half of a hole per cu site along a stripe .
after stripes were detected , the role of stripe fluctuations was considered in order to account for metallic behavior in cuprates @xcite .
[ the existence of charge - stripe fluctuations in the analog system ( with insulating charge stripes ) has now been confirmed by neutron scattering experiments @xcite ( and the system becomes a bad metal when the stripes are fluctuating @xcite ) . ]
an alternative theoretical approach is the concept of coulomb - frustrated phase separation @xcite .
an early analysis of the @xmath24@xmath25 model indicated that a low density of holes in an antiferromagnet would tend to phase separate @xcite ; the phase separation is possible because the model does not take account of extended coulomb interactions .
accounting for such interactions with an effective hamiltonian led to predictions of a variety of charge - ordered states , including stripe and checkerboard orders @xcite .
after stripes were experimentally observed , the possibility of superconductivity arising from pairing correlations within the hole - rich stripes was explored @xcite . at the time , only in - phase coupling of pairing between stripes was considered , and it was concluded that superconducting order would require fluctuating stripes in order to avoid competition from cdw order @xcite . the formation of stripes can be viewed as a competition between the kinetic energy of the holes and the exchange energy between the spins . this effect has been efficiently captured by density matrix renormalization group ( dmrg ) calculations of the @xmath24@xmath25 model on finite systems @xcite .
these calculations yield a charge density of @xmath26 hole / site , consistent with experiment .
furthermore , strong pairing correlations within the stripes are obtained @xcite .
experimentally , stripe order in with @xmath27 strongly depresses bulk superconducting order ; however , careful studies of the anisotropic resistivity and magnetic susceptibility in single crystals revealed evidence for the onset of two - dimensional ( 2d ) superconducting correlations at @xmath28 k , simultaneously with spin stripe order @xcite .
( similar observations of 2d superconductivity have been reported for stripe - ordered @xcite . ) while superconductivity in the cuprates is always driven by the electronic correlations within the cuo@xmath5 planes , there is inevitably a josephson coupling between the layers that induces 3d superconducting order when the correlation length for 2d superconducting phase order becomes sufficiently large .
the suppression of 3d order in @xmath27 suggests that the interlayer josephson coupling must be frustrated .
this frustration effect was first observed in a @xmath29-axis optical conductivity study of la@xmath30nd@xmath31sr@xmath32cuo@xmath16 , where the increasing nd concentration leads to a transition to the ltt phase and stripe order , together with a collapse of the josephson plasma resonance , despite the survival of bulk superconductivity @xcite .
these observations motivated the concept of the pdw @xcite . because of the rotation of the charge - stripe direction by 90@xmath33 between layers in the ltt phase @xcite , josephson coupling between the pdw order in neighboring layers is zero .
it is interesting to note that the pdw state has also been proposed from a different starting point , that of `` amperean pairing '' @xcite . while the pdw state provides a good explanation of the 1/8 anomaly ,
is it an energetically competitive state ?
recent variational tensor network studies of the @xmath24@xmath25 model suggest that it is @xcite .
using reasonable parameters , the stripe state with antiphase @xmath1-wave superconducting order is very close to that of stripes with in - phase @xmath1-wave order ; both are found to be slightly lower in energy than the uniform @xmath1-wave state .
there has also been an interesting mean - field analysis of the pdw state @xcite .
is the pdw state relevant at hole concentrations away from @xmath27
? optical conductivity studies of with @xmath34 provided evidence that superconducting layers become decoupled in a @xmath29-axis magnetic field @xcite . this led us to investigate the behavior of with @xmath2 , for which the zero - field @xmath4 is 32 k. measurements of resistivity parallel and perpendicular to the planes showed that the in - plane resistivity remains effectively zero while the @xmath29-axis resistivity can become finite in quite modest @xmath29-axis fields ( @xmath35 t ) @xcite .
such measurements have been extended to fields as high as 35 t , at which point
the in - plane resistivity is negligible at 13 k while the @xmath29-axis resistivity is compatible with strongly insulating behavior @xcite . while the stability of this novel high - field state is yet to understood , these results suggest that the pdw state may be induced by a @xmath29-axis magnetic field , leading to the frustration of the interlayer josephson coupling
. can the modulated superconducting state be relevant to good superconductors at zero field ?
we have recently investigated the spin fluctuations in the @xmath2 sample by inelastic neutron scattering @xcite .
this sample exhibits weak spin and charge stripe order in the superconducting state .
measurements of the incommensurate spin fluctuations at @xmath36 reveal gapless excitations with a strength comparable to that of spin waves in antiferromagnetic la@xmath5cuo@xmath16 .
similarly , there is no sign of a `` spin resonance '' peak .
the absence of a spin gap and resonance peak violates the current paradigm for antiferromagnetic superconductors @xcite .
empirically , superconductivity does not like to coexist with static or quasi - static local antiferromagnetic order .
near optimum doping , the spectral weight of low - energy spin fluctuations is weak in the normal state , and below @xmath4 the system may organize itself so as to gap the spin fluctuations below the energy of the superconducting gap . in the case of with @xmath2 ,
we have bulk superconductivity coexisting with spatially - modulated quasi - static antiferromagnetic spin correlations .
these observations seem to imply that the superconducting pair wave function must be spatially modulated , to minimize overlap with the low - energy spin correlations @xcite .
the modulated state in zero field is probably not a pure pdw , since the pdw is quite sensitive to disorder and would tend to frustrate the interlayer josephson coupling ; nevertheless , some degree of modulation appears unavoidable .
soon after quantum oscillations were observed in underdoped at high magnetic fields and low temperature @xcite , it was proposed that the corresponding fermi - surface pockets might result from a reconstruction of the fermi surface due to spin stripes @xcite or charge stripes @xcite .
similarities were demonstrated in the temperature dependence of transport properties , such as the thermoelectric power , between ( at finite magnetic field ) and stripe - ordered `` 214 '' cuprates @xcite .
nuclear magnetic resonance ( nmr ) studies revealed a magnetic - field - induced splitting of nmr lines , with an onset at a minimum field of 10 t , consistent with some type of charge order that appears at high field @xcite .
a sound - velocity study also provides evidence for a transition to a distinct high - field state at low temperature ( although the transition field seems to be different from that found by nmr ) @xcite .
thermal conductivity measurements provide further support for such a transition , and torque magnetometry provides evidence that diamagnetism is present above that transition ( although this is above the vortex melting transition ) @xcite .
it has been proposed that the transition between states with finite diamagnetism may correspond to a switch from uniform @xmath1-wave to pdw superconductivity @xcite . in the mean time
, distinct short - range cdw order has been detected in by both resonant soft - x - ray scattering @xcite and high - energy x - ray diffraction @xcite .
results for a range of dopings have recently been summarized @xcite . the most intense cdw scattering and
the highest onset temperatures occur near a hole concentration of 0.12 , the same point at which the upper critical field @xmath37 ( as measured by the onset of finite resistivity ) shows a dip @xcite , and remarkably similar to the 1/8 anomaly in .
the maximum cdw onset temperature is @xmath38 k , which is quite high compared to charge ordering in ; however , to the extent that this ordering may depend on anisotropic cu - o bonds , such an anisotropy is present in orthorhombic up to the cu - o chain ordering temperature , above 300 k. on cooling , the cdw intensity reaches a peak at the superconducting @xmath4 , and falls below it @xcite ; application of a @xmath29-axis magnetic field strong enough to depress the superconductivity causes an enhancement of the cdw intensity @xcite .
( similar behavior is seen in for @xmath39 far enough from 1/8 that the charge order is not close to saturation @xcite . )
diffraction studies have found similar short - range cdw correlations in bi@xmath5sr@xmath40la@xmath41cuo@xmath42 @xcite , @xcite , and hgba@xmath5cuo@xmath43 @xcite .
( cdw order has even been seen in the electron - doped superconductor nd@xmath40ce@xmath41cuo@xmath16 @xcite . ) for the former two materials , it has been shown @xcite that the cdw correlations correspond to modulations previously detected by scanning tunneling spectroscopy studies , especially at biases comparable to the pseudogap energy @xcite .
a significant observation from many of these studies is that the cdw wave vector tends to decrease slowly with hole doping , in contrast to the relationship @xmath44 already noted for and closely related compounds .
the characteristic _ spin _ wave vectors for low - energy magnetic excitations in bi@xmath45sr@xmath40cuo@xmath46 and _ increase _ with hole doping in a fashion quantitatively similar to and @xcite ; it follows that the cdw modulations in and these other compounds has no direct connection to the spin correlations .
it is notable that that inducing static incommensurate spin order in through the partial substitution of zn causes the cdw signal present without zn to weaken @xcite ; this result suggests that the high - temperature cdw correlations are distinct from the charge and spin stripes found in .
the connection between the cdw orders seen at low and high fields in has yet to be resolved .
nmr studies show that the modulation in the low - field cdw is relatively weak , as it only results in a small , temperature - dependent line broadening @xcite .
the same appears to be true in hgba@xmath5cuo@xmath43 , where , despite narrow @xmath47o nmr lines , an early study found no sign of anything unusual @xcite .
it seems likely that the high - field cdw is related to the modulations detected by scanning tunneling spectroscopy within halos about magnetic vortices in @xcite ; these modulations have been detected via resonances at an energy smaller than the superconducting gap .
further studies of this kind may provide a better understanding . for , a plot of @xmath4 measured in a @xmath29-axis field of 15 or 30 t shows a doping dependence with a strong dip at the hole concentration of 0.12 @xcite .
this two - hump behavior is quite similar to that seen in @xcite .
while the discussion so far has focused on the charge correlations that are optimized at 0.12 , it is also interesting to consider the character of the robust superconducting state in the low - doping dome . for , this corresponds to @xmath2 , which was argued in the previous section to have a spatially - modulated superconducting wave function related to the pdw state . for
, the lower dome has a maximum at the hole concentration 0.08 , where neutron scattering studies have found quasi - static , incommensurate spin correlations @xcite , with the elastic response enhanced by an applied magnetic field @xcite .
this behavior is quite similar to , and suggests the relevance of pdw - like superconductivity .
the intertwined orders of spin , charge , and superconductivity found in at @xmath27 , where bulk superconductivity is strongly depressed , appear to be relevant also to understanding the robust superconductivity found at @xmath2 .
while there may be differences with the type of cdw correlations found in other cuprate families , the fact that charge order is optimized near the hole concentration of 1/8 suggests a common `` ineluctable complexity '' @xcite .
quasi - static incommensurate spin correlations are likely intertwined with superconductivity in with a hole concentration of 0.08 , suggesting that such behavior may be common in the cuprates .
it will be of interest to test the generality of such behavior in future experiments , as well as to perform direct tests for pdw superconducting order @xcite .
i am grateful to many experimental collaborators , to s. a. kivelson and e. fradkin for collaborations on interpretation of the experiments , and to a. v. chubukov for useful discussions .
work at brookhaven is supported by the office of basic energy sciences , division of materials science and engineering , u.s .
department of energy under contract no .
de - ac02 - 98ch10886 .
this paper was written in part while the author was in residence at the kavli institute for theoretical physics , which is supported in part by the national science foundation under grant no .
nsf phy11 - 25915 .
g. ghiringhelli , m. le tacon , m. minola , s. blanco - canosa , c. mazzoli , n. b. brookes , g. m. de luca , a. frano , d. g. hawthorn , f. he , t. loew , m. m. sala , d. c. peets , m. salluzzo , e. schierle , r. sutarto , g. a. sawatzky , e. weschke , b. keimer , l. braicovich , science 337 ( 2012 ) 821825 .
j. chang , e. blackburn , a. t. holmes , n. b. christensen , j. larsen , j. mesot , r. liang , d. a. bonn , w. n. hardy , a. watenphul , m. v. zimmermann , e. m. forgan , s. m. hayden , nat .
( 2012 ) 871876 .
i. m. vishik , m. hashimoto , r .- h .
he , w .- s .
lee , f. schmitt , d. lu , r. g. moore , c. zhang , w. meevasana , t. sasagawa , s. uchida , k. fujita , s. ishida , m. ishikado , y. yoshida , h. eisaki , z. hussain , t. p. devereaux , z .- x .
shen , proc .
109 ( 2012 ) 1833218337 .
n. b. christensen , j. chang , j. larsen , m. fujita , m. oda , m. ido , n. momono , e. m. forgan , a. t. holmes , j. mesot , m. hcker , m. v. zimmermann , bulk charge stripe order competing with superconductivity in la@xmath40sr@xmath41cuo@xmath16 ( @xmath48 ) , arxiv:1404.3192 ( 2014 ) .
s. a. kivelson , v. j. emery , in : k. s. bedell , z. wang , d. e. meltzer , a. v. balatsky , e. abrahams ( eds . ) , strongly correlated electronic materials : the los alamos symposium 1993 , addison - wesley , reading , ma , 1994 , pp . 619656 .
d. leboeuf , n. doiron - leyraud , j. levallois , r. daou , j .- b .
bonnemaison , n. e. hussey , l. balicas , b. j. ramshaw , r. liang , d. a. bonn , w. n. hardy , s. adachi , c. proust , l. taillefer , nature 450 ( 2007 ) 533536 .
j. chang , r. daou , c. proust , d. leboeuf , n. doiron - leyraud , f. lalibert , b. pingault , b. j. ramshaw , r. liang , d. a. bonn , w. n. hardy , h. takagi , a. b. antunes , i. sheikin , k. behnia , l. taillefer , phys .
104 ( 2010 ) 057005 . f. yu , m. hirschberger , t. loew , g. li , b. j. lawson , t. asaba , j. b. kemper , t. liang , j. porras , g. s. boebinger , j. singleton , b. keimer , l. li , n. p. ong , diamagnetic response in under - doped yba@xmath5cu@xmath49o@xmath50 in high magnetic fields , arxiv:1402.7371 ( 2014 ) . a. j. achkar , r. sutarto , x. mao , f. he , a. frano , s. blanco - canosa , m. le tacon , g. ghiringhelli , l. braicovich , m. minola , m. moretti sala , c. mazzoli , r. liang , d. a. bonn , w. n. hardy , b. keimer , g. a. sawatzky , d. g. hawthorn , phys .
109 ( 2012 ) 167001 .
e. blackburn , j. chang , m. hcker , a. t. holmes , n. b. christensen , r. liang , d. a. bonn , w. n. hardy , u. rtt , o. gutowski , m. v. zimmermann , e. m. forgan , s. m. hayden , phys .
lett . 110 ( 2013 ) 137004 .
g. grissonnanche , o. cyr - choinire , f. lalibert , s. ren de cotret , a. juneau - fecteau , s. dufour - beausjour , m. .
delage , d. leboeuf , j. chang , b. j. ramshaw , d. a. bonn , w. n. hardy , r. liang , s. adachi , n. e. hussey , b. vignolle , c. proust , m. sutherland , s. krmer , j. h. park , d. graf , n. doiron - leyraud , l. taillefer , direct measurement of the upper critical field in cuprate superconductors , nat .
commun . 5 ( 2014 ) 3280 .
r. comin , a. frano , m. m. yee , y. yoshida , h. eisaki , e. schierle , e. weschke , r. sutarto , f. he , a. soumyanarayanan , y. he , m. le tacon , i. s. elfimov , j. e. hoffman , g. a. sawatzky , b. keimer , a. damascelli , science 343 ( 2014 ) 390392 .
w. tabis , y. li , m. l. tacon , l. braicovich , a. kreyssig , m. minola , g. dellea , e. weschke , m. j. veit , m. ramazanoglu , a. i. goldman , t. schmitt , g. ghiringhelli , n. barii , m. k. chan , c. j. dorow , g. yu , x. zhao , b. keimer , m. greven , connection between charge - density - wave order and charge transport in the cuprate superconductors , arxiv:1404.7658 ( 2014 ) .
e. h. da silva neto , r. comin , f. he , r. sutarto , y. jiang , r. l. greene , g. a. sawatzky , a. damascelli , charge ordering in the electron - doped superconductor nd@xmath40ce@xmath41cuo@xmath16 , arxiv:1410.2253 ( 2014 ) .
t. wu , h. mayaffre , s. krmer , m. horvati , c. berthier , w. n. hardy , r. liang , d. a. bonn , m .- h .
julien , short - range charge order reveals the role of disorder in the pseudogap state of high-@xmath4 superconductors , arxiv:1404.1617 ( 2014 ) . | the concept of intertwined orders has been introduced to describe the cooperative relationship between antiferromagnetic spin correlations and electron ( or hole ) pair correlations that develop in copper - oxide superconductors .
this contrasts with systems in which , for example , charge - density - wave ( cdw ) order competes for fermi surface area with superconductivity . with @xmath0
provides an example in which the ordering of spin stripes coincides with the onset of two - dimensional superconducting correlations .
the apparent frustration of the interlayer josephson coupling has motivated the concept of the pair - density - wave superconductor , a state that theoretical calculations show to be energetically competitive with the uniform @xmath1-wave superconductor .
even at @xmath2 , where there is robust superconductivity below 32 k in zero field , the coexistence of strong , low - energy , incommensurate spin excitations implies a spatially modulated and intertwined pair wave function .
recent observations of cdw order in and other cuprate families have raised interesting questions regarding the general role of charge modulations and the relation to superconductivity .
while there are differences in the doping dependence of the modulation wave vectors in and , the maximum ordering strength is peaked at the hole concentration of 1/8 in both cases .
there are also possible connections with the quantum oscillations that have been detected about the same hole concentration but at high magnetic fields .
resolving these relationships remains a research challenge .
high - temperature superconductors , copper oxides , stripes , pair density wave 74.72.-h , 75.25.dk , 74.81.-g |
in this supplement we present the calculations that were quoted in the main text .
we wish to evaluate @xmath102 the integral over the brillouin zone is dominated by those values of @xmath103 where both green functions correspond to states near a saddle point .
expanding the energy about the saddle points , we rewrite the integral as @xmath104 where the integral is understood to have a uv cutoff for @xmath103 of order @xmath105 .
we now define @xmath106 and @xmath107 , and rewrite the above integral as @xmath108 we now define @xmath109 and rewrite the integral as @xmath110 we now assume @xmath111 ( which should certainly be the case for weak / moderate coupling ) . in this limit , we can perform the integral over @xmath112 approximately , using the cauchy integral formula , to get @xmath113 the imaginary part of the above integral is odd in @xmath114 and hence vanishes upon performing the matsubara sum to leave an integral that is purely real @xmath115 with logarithmic accuracy . performing the integral over @xmath116 ( again with logarithmic accuracy )
gives @xmath117 where we take @xmath118 , @xmath119 and perform the discrete sum on mathematica .
this time we want to evaluate @xmath120 again , we anticipate this integral will be dominated by regions of the brillouin zone where all three green functions correspond to states near saddle points .
expanding the dispersion about the saddle points , we obtain @xmath121 making the same coordinate substitutions as in the preceding section , we recast this as @xmath122 after scaling out @xmath123 , we can rewrite it as @xmath124 where the rescaled integral is fully convergent , and can be done numerically on mathematica .
the sum over matsubara frequencies can also be done on mathematica , and yields the answer @xmath125 comparing with the previous expression for @xmath43 , we see that @xmath126 .
thus , @xmath127 provided the log is large .
( if the log is not large then the evaluation of @xmath43 with logarithmic accuracy does not suffice , and sub - logarithmic contributions to @xmath43 must also be taken into account . ) we want to evaluate @xmath128 this time it is not possible to place all the green functions at the saddle points .
in fact , we can not even place all the green functions at the fermi surface - the best that can be done is to place three of the green functions near a saddle point , but the fourth has to be off fermi surface .
thus , we obtain , @xmath129 making the usual substitutions , and assuming @xmath130 , we obtain @xmath131 the integral is convergent . as usual , the imaginary part is odd in @xmath114 and vanishes and we care only about the real part . performing the integral on mathematica and taking the real part , we obtain @xmath132 which is parametrically smaller than @xmath43 and @xmath55 by @xmath52 .
we now calculate the coefficient of the sixth order chirality sensitive term in the free energy , @xmath133 . after some analysis of diagrams we find that , @xmath134 we can now place all the green functions on the near a saddle point . making the usual substitutions , we obtain @xmath135 where in the second line we have assumed that the matsubara sum is controlled by the first few matsubara frequencies .
the integral can be done numerically , and it is negative for small @xmath52 .
it then follows that at weak / moderate coupling @xmath136 , so that the free energy at sixth order also disfavors chirality .
the above calculations were performed at weak / moderate coupling , assuming @xmath97 .
however , the integrals can be evaluated at arbitrary @xmath52 . to this end
, it is useful to define the scaling functions @xmath137 .
these scaling functions are evaluated numerically and shown in fig.[fig : scaling functions ] .
\a ) b ) c )
at strong coupling , when @xmath138 and @xmath100 , our analysis based on ginzburg - landau expansion is less accurate because fermions can no longer be approximated as free particles ( the self - energy corrections to fermionic lines and vertex corrections to square and hexagonal diagrams are generally of order one ) . nevertheless , if we apply our analysis to @xmath100 , we find that @xmath44 , @xmath46 and @xmath139 all change signs at some @xmath52 ( see fig . [ fig : scaling functions ] ) .
the first sign change occurs in the sixth order chirality sensitive term @xmath46 , which becomes positive for @xmath140 .
when @xmath46 is positive , the chiral sdw state @xcite is energetically favored , provided we are sufficiently far below @xmath0 for the sixth order term to dominate over the quartic term @xmath44 .
thus , at large @xmath52 , the uniaxial sdw phase has a much narrower region of stability , and the transition into the chiral sdw phase happens quite close to @xmath0 .
the term @xmath141 is next to change sign , becoming negative for @xmath142 .
once this term becomes negative , the system prefers instead a @xmath143 collinear state , of the form discussed in @xcite , wherein sdw order develops only at a single nesting vector .
the subsequent sign change of @xmath44 at @xmath144 has no physical consequences .
the @xmath143 collinear sdw state that forms at @xmath145 is a ( full ) metal because the entire fs is not gapped out .
the competition between a metallic collinear state and non - coplanar insulating state has been detected numerically in the mean - field analysis at strong coupling @xcite , and our results for the strong coupling case are in line with this earlier study .
our strong coupling results are also consistent with the studies that found a non - coplanar , chiral sdw order in the models of spins of the same fixed length at every lattice site @xcite . however , in the weak / moderate coupling limit , our results indicate that the preferred state is a uniaxial @xmath59 state of a sort not considered before , which can only be realized in a model starting from itinerant fermions .
we also analyzed the evolution of @xmath99 with @xmath52 for fermions on a triangular lattice .
we found similar trends , e.g. , sign change of @xmath46 .
however , for a triangular lattice , the first sign change ( in @xmath46 ) occurs at a much larger @xmath146 , when the itinerant approach is very questionable . | we consider electrons on a honeycomb or triangular lattice doped to the saddle point of the bandstructure .
we assume system parameters are such that spin density wave ( sdw ) order emerges below a temperature @xmath0 and investigate the nature of the sdw phase .
we argue that at @xmath1 the system develops a uniaxial sdw phase whose ordering pattern breaks @xmath2 symmetry and corresponds to an eight site unit cell with non - uniform spin moments on different sites .
this state is a half - metal it preserves full original fermi surface , but has gapless charged excitations in one spin branch only .
it allows for electrical control of spin currents and is desirable for nano - science .
the electronic properties of single layer graphene have been the subject of considerable experimental and theoretical interest @xcite . near half - filling
, a description in terms of non - interacting dirac electrons captures the essential physics , since interactions effects are suppressed by the low density of states ( dos ) .
a sharply different behavior arises when graphene is strongly doped to @xmath3 or @xmath4 filling @xcite . at this filling ,
a divergent density of states and nested fermi surface ( fs ) conspire to produce weak coupling instabilities to an extensive buffet of ordered states , including spin density waves ( sdw ) @xcite , pomeranchuk metals @xcite , and @xmath5 wave superconductors ( sc ) @xcite .
a similar situation arises on a triangular lattice at @xmath6 filling @xcite .
it has recently been established using renormalization group ( rg ) methods @xcite that the two most relevant instabilities at weak coupling are towards sdw and a @xmath5-wave sc .
the sdw vertex is the largest at intermediate rg scales , but superconducting vertex eventually overshoots it both at perfect nesting and away from perfect nesting , making @xmath5-wave superconductivity the leading weak coupling instability .
the sc state has a @xmath7 gap structure and breaks time - reversal symmetry @xcite . in this paper
we assume that superconductivity is destroyed by an applied magnetic field , or alternatively that system parameters are such that corrections to the rg flow become relevant before sc vertex overshoots sdw vertex . in both cases ,
the sdw instability becomes dominant , and an sdw order emerges .
previous work argued that the sdw state is non - coplanar and has a non - zero spin chirality @xcite .
such a state gaps out the entire fermi surface ( fs ) , i.e. , is an insulator .
the chiral sdw state has also been found in the strong coupling analysis for classical spins of fixed length @xcite .
we argue that the situation is more complex than originally thought , and the chiral sdw state is present only at the lowest temperatures . over a wide intermediate range of temperatures , a different sdw state emerges in which sdw order develops simultaneously at three inequivalent wavevectors @xmath8 , but the three vector order parameters are all aligned along the same axis .
this state has a eight site unit cell with non - uniform spin moments and zero net magnetization ( fig .
[ fig : lattice]b ) .
such a state can not be accessed starting from a spin hamiltonian for local moments with a fixed length , and can only be accessed starting from a model of itinerant fermions .
we show that in this state , unlike in any other known sdw state , the chemical potential shifts proportionally to the sdw order parameter preserving the original fermi surface for one spin branch and gapping out the other spin branch .
the uniaxial sdw state is therefore a ` half - metal ' that allows for electrical control of spin currents .
such a state is highly desirable for nano - science applications .
\a ) b ) _ the model : _ for definiteness we focus on doped graphene at @xmath3 filling .
our point of departure is the tight binding model @xcite , with the nearest - neighbor dispersion @xmath9 where @xmath10 at @xmath3 filling .
the fs then forms a perfect hexagon inscribed within a hexagonal bz ( fig .
[ fig : lattice]a ) .
the perfect nesting of the fs in doped graphene is quite robust
it is broken only by third and higher neighbor hoppings , which are generally quite small .
the fermi velocity vanishes near the hexagon corners @xmath11 , which are saddle points of the dispersion : @xmath12 where each time @xmath13 denotes the deviation from a saddle point .
saddle points give rise to a logarithmic singularity in the dos and control the sdw instability at weak coupling .
there are three in - equivalent nesting vectors connecting in - equivalent pairs of saddle points ( see fig . [
fig : lattice]a ) : @xmath14 each @xmath15 is equivalent to @xmath16 modulo a reciprocal lattice vector .
there are two electron - electron interactions that contribute to the sdw channel .
one is a forward scattering interaction @xmath17 , while the other is an umklapp interaction , @xmath18 .
we label these interactions @xmath19 and @xmath20 respectively , for consistency with the notation introduced in @xcite . the partition function for @xmath21 model
can then be written as @xmath22 \exp(- s[\psi^{\dag } , \psi])$ ] , where @xmath23 and @xmath24 where the action is written in terms of electron operators , @xmath25 are patch labels , and @xmath26 and @xmath27 are spin components .
each nesting vector @xmath28 has associated with it an sdw order parameter @xmath29 .
the condition for the emergence of each @xmath30 is the same : @xmath31 @xcite , leaving a large number of sdw states as potential candidates .
we study the selection of the sdw order within ginzburg - landau theory and by comparing different sdw solutions in the mean - field approximation for eq .
( [ eq : sdw action ] ) at arbitrary @xmath32 . _
ginzburg - landau theory : _ to construct the ginzburg - landau theory
, we decouple the quartic interaction terms by restricting the interaction to the spin channel and performing a hubbard stratonovich transformation to introduce the order parameters @xmath33 .
we integrate out the fermions in the matsubara frequency representation and obtain an action in terms of the order parameter fields @xmath30 , which takes the form @xmath34 .
\label{eq : grand potential}\end{aligned}\ ] ] for @xmath35 , we can expand ( [ eq : grand potential ] ) in small @xmath36 .
it is useful to define the expansion coefficients @xmath37 where the integrands @xmath38 are expressed in terms of fermionic green functions @xmath39 , @xmath40 , and @xmath41 as @xmath42
diagrammatically , @xmath43 @xmath44 are given by ` square ' diagrams with four fermionic propagators and @xmath45 in the vertices , and @xmath46 is given by a ` hegagonal ' diagram with six fermionic propagators , ( see fig .
[ fig : diagrams ] ) . the free energy evaluated at @xmath47 can be expressed in terms of these coefficients as @xmath48 where @xmath26 is an inessential positive constant .
the quadratic term and the first quartic term in ( [ eq : free energy ] ) set the overall magnitude of @xmath49 , but do not differentiate between different sdw states . the second quartic term in ( [ eq : free energy ] )
determines whether sdw order develops only at one nesting vector , or at all three ( depending on the sign of @xmath50 ) . finally , the third quartic term and sixth order term control the relative orientation of the vector order parameters , if sdw order develops at multiple wavevectors .
close to @xmath0 the expansion to order @xmath51 is generally sufficient , but we include the sixth order term because @xmath44 is suppressed by an extra factor of @xmath52 , which is exponentially small in the weak coupling limit .
the relative smallness of @xmath44 arises because in the integrals for @xmath53 , and @xmath46 , all fermions can be simultaneously brought to the saddle points , whereas in the integral for @xmath44 , three fermions can be brought simultaneously to saddle points , but the remaining fermion stays far away from not only the saddle point but also the fs .
we evaluate the coefficients @xmath43 @xmath46 to leading order in small @xmath52 and obtain @xcite @xmath54 the positivity of @xmath43 guarantees a second order phase transition , with the type of sdw order depending on the signs and relative magnitudes of @xmath55 , @xmath44 , and @xmath46 . since @xmath44 is smaller by @xmath52 than @xmath56 , and @xmath55 is smaller by @xmath57 than @xmath43 , it follows that @xmath58 , so the system forms sdw order simultaneously at all three nesting vectors ( the @xmath59 state ) .
meanwhile , the relative orientation of the three sdw order parameters is controlled by the sign of @xmath44 at the smallest @xmath60 , and by the sign of @xmath46 at somewhat larger @xmath60 .
both @xmath44 and @xmath46 are negative and favor the non - chiral sdw order with the three @xmath60 all aligned along the same axis .
an order parameter of the form @xmath61 leads to spin moments on the lattice of the form shown in fig .
[ fig : lattice ] .
a quarter of lattice sites have spin moment @xmath62 , the other three quarters have moment @xmath63 .
such an order can not be obtained from any spin hamiltonian for local moments of constant magnitude on every site .
our result differs from earlier mean - field analysis @xcite which found non - coplanar insulating sdw order at weak coupling .
we note , however , the @xmath59 state that we found , with non - equal spin length on different sites , was not considered in that work and other earlier considerations of the type of sdw order .
we found analogous results for fermions on a triangular lattice at van hove filling .
this system is identical to graphene , except that the nesting is less robust and is spoiled already by second neighbor hopping .
_ properties of a uniaxial sdw : _ is the uniaxial sdw state a metal or an insulator ? to address this issue we need to compute the fermionic spectrum . without loss of generality , we take the sdw to be uniaxial along the @xmath64 axis , so that @xmath65 is a good quantum number , and spin - up and spin - down fermions decouple .
consider the state with @xmath66 .
the up spins near the three van hove points are described by a simple @xmath67 hamiltonian @xmath68 where @xmath69 are the dispersions near the van hove points , eq .
( [ n_1 ] ) , and @xmath70 is the sdw - induced shift of the chemical potential .
the @xmath67 hamiltonian describing the spin down branch is obtained by taking @xmath71 . at @xmath72 ( i.e. , at van hove points )
the energies of spin - up excitations @xmath73 are @xmath74 , and @xmath75 , and the energies of spin - down excitations are @xmath76 , and @xmath77 . in conventional sdw states
( e.g. , sdw on a 2d square lattice ) @xmath78 is negligibly small and can be safely neglected .
we find , however , that in our case @xmath79 , so that gapless excitations arise in the spin - down spectrum . to see the unexpected shift of the chemical potential , we diagonalize eq .
( [ m_1 ] ) and the corresponding equation for down spins and inspect six branches of excitations .
we find that fixing @xmath79 ensures that both in the paramagnetic and in the @xmath59 uniaxial sdw state there are four bands with @xmath80 and two bands with @xmath81 for all momenta in the reduced bz ( see fig .
[ fig : spectrum ] ) .
since the chemical potential is fixed by the constraint that the total number of electrons ( equal to the number of states below the chemical potential ) must not change between @xmath82 and @xmath83 @xcite , it follows that we must set @xmath84 . for verification
, we computed the thermodynamic potential @xmath85 from ( [ eq : grand potential ] ) , numerically solved the simultaneous equations @xmath86 and @xmath87 , and confirmed that @xmath79 to a high accuracy . having determined that @xmath79 , we find from ( [ m_1 ] ) that gapless excitations emerge when @xmath88 , which has solutions along three lines passing through each van hove point .
two of then coincide with the original fs , the third is directed towards the center of the bz .
the @xmath59 uniaxial sdw state is then obviously a metal .
we emphasize , however , that gapless states exist only for the electrons with spin projection opposite to @xmath89 .
the electrons with spin projection along @xmath89 are fully gapped .
since a fermi surface exists for one spin projection only , we dub this state a ` half metal . '
we found an analogous ` half - metal ' spectrum for the @xmath59 uniaxial sdw phase on the triangular lattice .
the half - metallic nature of the sdw should manifest itself in numerous experiments .
for example , in tunneling experiments conducted with electrons spin polarized along the @xmath64 axis , a hard gap will be seen for down spins , but a fermi surface will be seen for up spins . furthermore , since the low energy charged excitations involve up spins only , any charge currents will necessarily also be spin currents .
thus , the half metal state allows for electrical control of spin currents , which may be beneficial for nanoscience applications .
_ order parameter manifold : _ the uniaxial sdw order obviously breaks @xmath90 spin - rotational symmetry .
it also breaks @xmath46 discrete symmetry associated with either parallel or antiparallel ordering of @xmath60 , i.e. , in addition to the @xmath91 state which we considered above , there are also @xmath92 , @xmath93 , and @xmath94 states .
these states have an identical structure of fermionic excitations , and correspond to the four in - equivalent ways to choose which two of the eight sites of the sdw unit cell carry large spins ( see fig .
[ fig : lattice](b ) ) .
equivalently , the three other states from the @xmath46 manifold are obtained from the @xmath91 state by shifting the origin of coordinates to the center of one of three neighboring hexagons .
an interesting possibility , which deserves further study , is that @xmath46 symmetry can be broken before @xmath90 symmetry , leading to a nematic - like state @xcite . _
the phase diagram : _ thus far we have constructed the ginzburg - landau expansion in small @xmath95 .
this expansion becomes less justified as we move towards zero temperature . to investigate the behavior at arbitrary @xmath96
we calculate numerically the full free energies of the various sdw states from ( [ eq : grand potential ] ) . upon doing this ,
we find that the @xmath59 uniaxial state has the lowest free energy over a wide range of intermediate temperatures , but undergoes a first order transition at a lower temperature to the insulating chiral sdw state discussed in earlier works @xcite .
we show the free energy profile in fig .
[ fig : spectrum]b .
we found this behavior both for graphene and for fermions on a triangular lattice .
intuitively , the chiral sdw state wins at the lowest @xmath96 because it has spin - degenerate excitations and opens a full spectral gap , unlike the half - metal state . the free energy profile in fig . [ fig : spectrum]b is for weak / moderate coupling , when @xmath97 . at @xmath98 ,
the phase diagram is more complex and less universal . for completeness
, we discuss the forms of @xmath99 and the phase diagram at @xmath100 in the supplementary material @xcite .
_ conclusion : _ we considered in this work the sdw instability on the honeycomb and triangular lattices , when doped to the saddle points of the dispersion .
the sdw instability is subleading to a @xmath101wave superconducting instability at weak coupling , but becomes the leading instability if superconductivity is suppressed .
we found that if the sdw ordering temperature @xmath0 is much smaller than the fermionic bandwidth , then a uniaxial sdw order develops simultaneously at three inequivalent nesting vectors .
this has an order parameter manifold @xmath2 and corresponds to the ordering pattern shown in fig.[fig : lattice ] .
such a state can only be obtained from an model of itinerant electrons with interactions , and not from a spin model of local moments .
we found that such sdw state is a half - metal in which gapless excitations exist in one spin branch only .
such a state may be beneficial for nanoscience applications particularly because charge currents will necessarily also be spin currents , which allows for electrical control of the latter .
we thank l. levitov for numerous discussions concerning the interplay between superconductivity and sdw order .
we are also thankful to c. batista , r. fernandes , i. martin , and fa wang for useful conversations .
g.w.c .
is supported by icam and nsf - dmr-0844115 , and a.v.c .
is supported by nsf - dmr-0906953 . |
even after several decades of research the microscopic foundations of itinerant ferromagnetism are not sufficiently understood .
indeed , in contrast to other collective electronic phenomena such as antiferromagnetism or conventional superconductivity there exists a remarkable gap between theory and experiment in this field .
this has mainly to do with the fact that itinerant ferromagnetism is a quantum - mechanical _ strong - coupling _ phenomenon whose explanation requires the application of _ non - perturbative _ techniques .
thus it belongs into the class of the most difficult many - body problems in condensed - matter physics .
significant progress was made in this field in the last few years due to the development and application of several new analytic and numerical approaches .
it is the purpose of this paper to present and discuss some of these new , exciting results . in sec . 2 we recapitulate the derivation of a general lattice model for correlated electrons starting from a continuum model , and discuss the various truncation steps which eventually lead to the hubbard model .
in particular , the implications of the lieb - mattis theorem on the impossibility of ferromagnetism in these truncated models in @xmath1 dimension are analyzed . in sec .
3 several microscopic mechanisms favoring ferromagnetism are discussed .
new and very recent results concerning ferromagnetism in hubbard - type models obtained by various non - perturbative methods are presented and put into perspective .
metamagnetic phase transitions and the need for the application of non - perturbative techniques for these investigations are the subject of sec .
finally , a conclusion is presented in sec .
within the occupation number formalism the hamiltonian for electrons with spin @xmath2 interacting via a _ spin - independent _ interaction @xmath3 in the presence of an ionic lattice potential @xmath4 has the form @xcite @xmath5 , where @xmath6 \hat{\psi}_\sigma ( { \bf r } ) \label{h0}\\ \hat{h}_{\mbox{\scriptsize\em int } } & = & \frac{1}{2 } \sum_{\sigma \sigma ' } \int d^{3}r \int d^{3}r\ , ' v^{ee } ( { \bf r } - { \bf r\ , ' } ) \hat{n}_{\sigma } ( { \bf r } ) \hat{n}_{\sigma ' } ( { \bf r } \ , ' ) \ , . \label{hint}\end{aligned}\ ] ] here @xmath7 are the usual field operators and @xmath8 is the local density .
we note that the interaction term is diagonal in the space variables @xmath9 i.e. it depends only on the ( operator - valued ) densities of the electrons at site @xmath10 which interact via @xmath11 the lattice potential entering the non - interacting part ( [ h0 ] ) leads to a splitting of the parabolic dispersion into infinitely many bands which we enumerate by the index @xmath12 the non - interacting problem is then characterized by the bloch wave functions @xmath13 and the band energies @xmath14 we may introduce wannier functions localized at site @xmath15 by @xmath16 where @xmath17 is the number of lattice sites , and thus construct creation and annihilation operators @xmath18 for electrons with spin @xmath19 in the band @xmath20 at site @xmath21 as @xmath22 thereby the hamiltonian may be written in lattice representation as @xcite @xmath23 where the matrix elements are given by @xmath24 \chi_{\alpha i } ( { \bf r } ) \label{matr1}\\ { \cal v}^{\alpha\beta\gamma\delta}_{i j m n } & = & \int d^{3}r \int d^{3}r\ , ' v^{ee } ( { \bf r } - { \bf r}\ , ' ) \nonumber \\ & & \chi^{*}_{\alpha i } ( { \bf r } ) \ \chi^{*}_{\beta j } ( { \bf r}\ , ' ) \
\chi_{\delta n } ( { \bf r}\ , ' ) \ \chi_{\gamma m } ( { \bf r } ) \ , . \label{matr2}\end{aligned}\ ] ] we note that in contrast to the field - operator representation defined in the continuum , the wannier representation does _ not _ lead to a site - diagonal form of the electron - electron interaction , i.e. the interaction does not only depend on the _ densities _ @xmath25 but contains explicit off - diagonal contributions which will be discussed later .
the hamiltonian ( [ ham ] ) is too general to be tractable in dimensions @xmath26 .
hence it has to be simplified using physically motivated truncations @xcite .
in particular , if the fermi surface lies within a single conduction band , and if this band is well separated from the other bands and the interaction is not too strong , it may be justified to restrict the discussion to a _
single _ band ( @xmath27 ) . in this case
( [ ham ] ) reduces to @xmath28 for most purposes this single - band hamiltonian is still too complicated . taking into account the weak overlap between neighboring orbitals in a tight - binding description one may expect that the overlap between nearest - neighbors is most important . hence the site - indices in ( [ 1band ] ) are restricted to nearest - neighbor positions . in the interaction
this leaves us with a purely local contribution @xmath29 , the hubbard term , and the four nearest - neighbor contributions @xmath30 , @xmath31 , @xmath32 , and @xmath33 , which are _
off_-diagonal in the site indices . the remaining one - band , nearest - neighbor hamiltonian has the form @xcite @xmath34 where @xmath35 is the hubbard model and @xmath36 is the contribution of nearest - neighbor interactions . here
@xmath37 and @xmath38 , where @xmath39 denotes the vector of pauli matrices . in ( [ vnn ] )
the @xmath40-term describes a density - density interaction , the @xmath41-term is a bond - charge site - charge interaction ( `` density - dependent hopping '' ) , the @xmath42-term describes the hopping of local pairs consisting of an up and a down electron and , finally , the @xmath43-term corresponds to the _ direct _ heisenberg exchange which is generically ferromagnetic in nature .
the occupation number formalism illustrates particularly clearly that a spin - independent interaction plus the pauli principle is able to lead to a mutual orientation of spins . of all interactions in ( [ nn ] )
the hubbard interaction @xmath0 is certainly the strongest .
hence , in a final truncation step one may try to neglect even the nearest - neighbor interactions and retain only the on - site interaction @xmath0 .
this leaves us with the hubbard model ( [ hub ] ) , the simplest correlation model for lattice electrons @xcite .
the hubbard model was originally introduced in an attempt to understand itinerant ferromagnetism in @xmath44-transition metals @xcite .
the expectation was that in this model ferromagnetism would arise naturally since in a polarized state the electrons do not interact at all .
however , it soon became clear that in a ferromagnetic state the _ kinetic _ energy is also reduced .
this makes the stability of ferromagnetism in the hubbard model a particularly delicate problem .
indeed , the kinetic energy with nearest - neighbor hopping usually favors _ anti_ferromagnetism . at half - filling @xmath45 and on bipartite lattices antiferromagnetism
is a generic effect since it appears both at weak coupling ( hartree - fock or slater mean - field theory ) and strong coupling ( anderson s `` superexchange '' mechanism ) .
hence it arises naturally in any perturbational approach and , in particular , is tractable by renormalization group methods @xcite . by contrast , ferromagnetism is a non - trivial strong - coupling phenomenon which can not be investigated by any standard perturbation theory .
the above discussion shows that , to understand the microscopic origin of itinerant ferromagnetism , non - perturbative techniques are required . unfortunately , there are not many approaches of this type available ; rigorous mathematical methods ( for recent reviews see ref .
@xcite ) , large - scale numerical methods @xcite , and variational approaches @xcite are the ones most frequently used .
as to rigorous results about ferromagnetism in the hubbard model the lieb - mattis theorem @xcite of 1962 is one of the most famous .
it proves that for spin- and velocity - independent forces between electrons ferromagnetism can not occur in one spatial dimension .
this theorem applies to the general hamiltonian ( [ ham ] ) with infinitely many bands . on the other hand the general _
single-_band model ( [ 1band ] ) is _ not _ covered by the theorem unless ( a ) the interaction matrix element @xmath46 is site - diagonal such that the interaction depends only on densities @xmath47 , and ( b ) the hopping and interaction does not extend beyond nearest neighbors .
hence the theorem applies to the model ( [ nn ] ) with @xmath48 but @xmath49 , and thereby also to the hubbard model @xmath50 .
we note that any critique of the single - band model ( [ nn ] ) in view of the fact that it can lead to ferromagnetism in @xmath51 in contrast to the lieb - mattis theorem would apply even more so to the hubbard model since the latter is a particularly special single - band model .
in other words : the fact that the lieb - mattis theorem applies to the hubbard model but _ not _ to the _ general _ single - band model ( [ nn ] ) does not make the hubbard model a `` more physical '' model than ( [ nn ] ) ; after all it is only a special case of ( [ nn ] ) . _
_ _ _ another well - known theorem , that by nagaoka @xcite of 1966 , provides explicit , albeit highly idealized , conditions under which ferromagnetism _ is _ stable in the hubbard model with nearest - neighbor hopping .
it proves that for @xmath52 the microscopic degeneracy of the ground state at half - filling ( number of electrons @xmath53 = number of lattice sites @xmath17 ) is lifted by a single hole , i.e. when one electron is removed @xmath54 . in this case
a saturated ferromagnetic ground state is stable for any value of the hopping @xmath55 on simple cubic and bcc lattices , and for @xmath56 on fcc and hcp lattices . for the nagaoka mechanism to work the lattice
needs to contain loops along which the holes can move .
once the hole moves , the maximal overlap between the initial and final state clearly occurs in a _ ferromagnetic _ configuration .
the problem is that nagoaka s proof does not even extend to two holes , that a single hole is thermodynamically irrelevant , and that the limit of @xmath57 is highly unrealistic .
the hubbard interaction is the result of an extreme truncation of the interaction in the general hamiltonian ( [ ham ] ) .
all interactions beyond the purely local part ( e.g. nearest - neighbor density - density interactions , direct exchange , band degeneracy and the associated hund s rule couplings ) are totally neglected .
the hubbard interaction is therefore very unspecific it does not depend on the lattice at all and hence not on the spatial dimension .
the lattice structure only enters via the kinetic energy .
therefore the stability of ferromagnetism in the hubbard model can be expected to depend in a sensitive way on the precise form of the kinetic energy @xcite .
strategies to find the essential `` kick '' for ferromagnetism in the hubbard model and more general models should then proceed in different directions : one may ( i ) keep interactions beyond the hubbard-@xmath0 ( in particular the direct exchange term @xmath43 in ( [ nn ] ) ) , ( ii ) keep band degeneracy and hund s rule couplings , or ( iii ) find an appropriate kinetic energy and lattice structure .
we will now discuss several recent results obtained along these lines .
the heisenberg exchange interaction , caused by direct quantum - mechanical exchange of electrons at nearest - neighbor positions ( the @xmath43-term in ( [ vnn ] ) with @xmath58 ) , favors the alignment of the electronic spins and hence supports ferromagnetism in a straightforward way @xcite .
however , since this interaction is rather weak ( hubbard @xcite estimated @xmath59 ev for @xmath44-metals , such that @xmath60 ) it can not be the _ sole _ origin of itinerant ferromagnetism in systems like fe , co , ni . nevertheless it may be qualitatively important , since it may well give a correlated system with more or less strong ferromagnetic tendencies the ultimate push and trigger ferromagnetism in spite of its smallness .
it is therefore unjustified to neglect the exchange interaction for merely _ quantitative _ reasons .
this becomes particularly clear in the limit of large @xmath0 ( with @xmath61 , @xmath62 , @xmath63 , @xmath43 , @xmath42 ) close to @xmath64 , when the one - band model ( [ nn ] ) can be transformed into an effective @xmath55-@xmath65-model @xcite @xmath66 where @xmath67 projects onto the subspace without doubly occupied sites .
the effective exchange coupling @xmath68\ ] ] has an antiferromagnetic part , due to anderson s superexchange but modified by the @xmath41-term , and a ferromagnetic part , due to the direct heisenberg exchange .
hence for large enough heisenberg exchange @xmath43 and/or hubbard repulsion @xmath0 the exchange becomes effectively ferromagnetic .
this effect is completely neglected in the hubbard model where even in the nagaoka - limit ( @xmath57 ) the dimensionless parameter @xmath69 is kept zero ! if the heisenberg exchange coupling is taken into account , it is possible to generalize nagaoka s theorem to @xmath70 @xcite .
we start with the hamiltonian @xmath71 , ( [ nn ] ) .
this hamiltonian , like the hubbard hamiltonian , commutes with the total spin @xmath72 .
the eigenvalues of @xmath73 are denoted by @xmath74 .
we are concerned with saturated ferromagnetic states with one hole below half - filling , i.e. with largest possible eigenvalue @xmath75 .
there are @xmath76 such states with the same energy eigenvalue .
it can be shown that the ground states of @xmath77 with one hole ( i. e. @xmath78 ) have maximum total spin @xmath79 and are non - degenerate ( apart from the above - mentioned @xmath80-fold spin degeneracy ) in the following cases @xcite : case 1 : on any lattice , if @xmath58 , @xmath81 and ( a ) @xmath82 and @xmath83 , or ( b ) @xmath84 and @xmath85 . case 2 : on lattices with loops , if @xmath86 , @xmath87 , and @xmath88 . in both cases @xmath89 is allowed if the lattice is bipartite .
these results are summarized in table [ crtab ] .
the constants @xmath90 and @xmath91 are given by @xmath92 where @xmath93 is the number of nearest neighbors .
hence , if @xmath58 ferromagnetic ground states are stable on any lattice for @xmath0 larger than a _ finite _ critical value .
for @xmath94 we have @xmath95 , thus yielding nagaoka s condition for the pure hubbard model .
this shows once more that the heisenberg interaction @xmath43 , which is neglected in the hubbard model , provides an obvious mechanism for stabilizing ferromagnetic ground states at finite @xmath0 .
note that since @xmath41 and @xmath55 are expected to be of the same order of magnitude , the sensitive dependence on @xmath43 , due to the term @xmath96 , may cancel from @xmath90 .
the dependence of @xmath97 on @xmath98 is depicted in fig .
the case @xmath84 is special , since in this case the stability of ferromagnetism can be achieved either by @xmath58 , or by @xmath99 and @xmath100 if the lattice has loops .
note that the case @xmath58 is _ not connected _ to the case @xmath87 by a limiting procedure , since only in the latter case the lattice is required to have loops .
the critical couplings @xmath90 and @xmath91 are sums of terms , each of which corresponds to a typical energy scale .
this means that the on - site interaction @xmath0 has to be larger than the energy describing the paramagnetic state ( bandwidth @xmath101 ) , as well as the threshold energies for the onset of a charge - density wave or phase separation ( @xmath102 ) , @xmath103-pairing superconductivity @xcite ( @xmath104 ) , and a spin - density wave ( @xmath105 ) .
however , these terms do not enter separately , but appear in combinations , i. e. the effects interfere , as should be expected .
the above conditions are _ sufficient _ conditions .
the occurrence of ground states with maximum spin outside the above parameter region is not ruled out . as in nagaoka s theorem for the pure hubbard model
, the ferromagnetic ground state is an itinerant state with non - zero kinetic energy , but the proof of its stability can not yet be extended to doping beyond a single hole . even for the simplest electronic correlation model , the hubbard model , exact solutions are not available in @xmath106 dimensions , and numerical methods whether exact diagonalizations or quantum monte carlo ( qmc ) techniques are limited by the smallness of the systems that can be studied .
hence one would like to construct , at least , a thermodynamically consistent mean - field theory which is valid also at strong coupling .
such a ( non - perturbative ) approximation is provided by the exact solution of a model in @xmath107 .
it is now known that in the limit @xmath108 @xcite one obtains a _
dynamical _ mean - field theory for hubbard - type models where the spatial dependencies become local , but all quantum fluctuations of the @xmath109-dimensional model are included @xcite .
in fact the problem is equivalent to an anderson impurity model complemented by a self - consistency condition @xcite , leading to dynamical mean - field equations which can be solved numerically , e.g. within a finite - temperature qmc scheme @xcite ; for reviews see refs .
we employed this numerical approach ( for details see ref .
@xcite ) to investigate the influence of the direct exchange interaction @xmath43 on the stabilization of ferromagnetism in the hubbard model , using the one - band hamiltonian ( [ nn])-([vnn ] ) with @xmath110 @xcite .
to take the limit @xmath108 the couplings in ( [ nn])-([vnn ] ) have to be scaled appropriately @xcite , i.e. @xmath111 in the following we set @xmath112 . in the limit
@xmath108 the @xmath113-term @xcite and @xmath114-term reduce to their hartree - contributions .
hence their influence is that of a generalized , i.e. spin- and site - dependent , chemical potential . in the homogeneous phase the spin- and site - dependent terms vanish
this implies that the nearest - neighbor interactions @xmath114 and @xmath113 become important only in the symmetry-_broken _ phase .
consequently the susceptibilities of the model ( [ nn])-([vnn ] ) with @xmath110 may be calculated from the pure hubbard model @xcite ; this simplifies the matter considerably .
the phase boundaries between the paramagnetic , antiferromagnetic and ferromagnetic phases may then be calculated in principle as follows @xcite : ( i ) qmc simulations are performed in the homogeneous phase of the pure hubbard model for given @xmath0 , temperature @xmath115 , filling @xmath116 , and number of matsubara frequencies @xmath117 ; ( ii ) for arbitrary values of @xmath114 , @xmath113 the appropriate susceptibilities are calculated ; ( iii ) the inverse susceptibilities are extrapolated to @xmath118 , if they are negative the homogeneous phase is found to be unstable ; ( iv ) to obtain ground state properties the calculated quantities have to be extrapolated to @xmath119 .
the results of these calculations for @xmath64 , @xmath120 and a semi - elliptic dos are collected in fig .
[ phaset0 ] , where the exchange coupling @xmath114 is plotted versus the hubbard interaction @xmath0 .
we neglect the density - density terms in ( [ vnn ] ) , since they become important only for @xmath121 @xcite .
the solid line marks the phase boundary to the ferromagnetic phase .
as expected the critical value of @xmath114 , @xmath122 , decreases with @xmath0 : it depends on @xmath0 as @xmath123 for small @xmath0 ( hartree - fock limit ) and as @xmath124 for large @xmath0 ( heisenberg limit ) . at @xmath125 the value of @xmath114 necessary to induce ferromagnetism
is seen to be as small as @xmath126 .
this shows how important even a weak exchange coupling @xmath43 is for the stability of ferromagnetism .
for @xmath127 there may well be a direct transition between the antiferromagnetic and the ferromagnetic phases .
the shape of the dos and the band filling @xmath116 are very important factors concerning the stability of ferromagnetism , as will be discussed below ( sec .
3.3 ) . indeed ,
a _ symmetric _ dos and filling @xmath64 , as used in the above calculation , are especially disadvantageous for ferromagnetism , partly because antiferromagnetism will be the generic magnetic order in this case .
another important route to ferromagnetism may be taken by considering more than one energy band , namely by starting from @xmath128 wannier ( or tight - binding ) orbitals .
it is known from atomic magnetism that there are ferromagnetic couplings between electrons on the same atom leading to hund s rules .
these on - site `` hund s rule couplings '' express the fact that by putting electrons in a maximum spin state an atomic exchange energy may be gained by the following mechanism . a spin wave function with maximum
spin is always symmetric .
therefore , for the total wave function to be antisymmetric , the coordinate wave function must be antisymmetric .
this reduces the probability for electrons to come close to each other which in turn lowers the coulomb energy between them .
this consideration establishes that in general there will be ferromagnetic on - site interactions _ even in a bulk system .
_ whether the presence of these terms is sufficient for ferromagnetism to appear in the ground state is , however , strongly dependent on their relative strength compared with the kinetic energy , on the lattice structure , electron density , etc .
let us therefore take a closer look at the terms in the hamiltonian ( [ ham ] ) in the case of @xmath129 relevant bands .
in this case we have to retain the band index @xmath130 .
now there exist important on - site interactions even beyond the hubbard interaction @xmath131 , namely the following couplings that are off - diagonal in the _ band _ indices , and are hence only present for @xmath128 bands : density - density interaction @xmath132 , direct exchange interaction @xmath133 , and hopping of double occupancies @xmath134 . for simplicity
we assume the orbitals to be equivalent , i.e. @xmath0 is the same for all orbitals @xmath20 , and @xmath135 , @xmath136 , @xmath137 each have a fixed value for all pairs of orbitals @xmath20 , @xmath138 .
furthermore it should be noted that for equivalent orbitals these parameters are not independent , but are related by @xmath139 , @xmath140 @xcite .
in addition to these hund s rule couplings there are still the inter - site terms , namely the hopping @xmath141 which takes place only between like orbitals ( this follows from the general derivation above ) , and the next - neighbor interactions @xmath142 , @xmath143 , @xmath144 , @xmath145 . for simplicity
, these next - neighbor couplings are assumed to be diagonal in the band indices , i.e. they act only between like orbitals on neighboring sites . finally , since we are dealing with equivalent orbitals , we assume that the next - neighbor parameters @xmath55 , @xmath146 , @xmath147 , @xmath148 each have the same value for all bands @xmath20 .
the resulting multi - band hamiltonian then reads @xmath149 where @xmath150 \end{aligned}\ ] ] is a straightforward generalization of the one - band hamiltonian ( [ nn ] ) to more than one band , and @xmath151\ , .
\end{aligned}\ ] ] several of the processes contained in the hamiltonian ( [ mbham ] ) are illustrated in fig .
[ orb ] .
the physical picture of bulk ferromagnetism , originally put forward by slater , is the following .
if the on - site hund s rule couplings are strong enough , they lead to an independent ferromagnetic alignment of spins on each atom . in this situation
the kinetic energy plays a decisive role since it can serve to communicate the spin alignment across the solid . indeed ,
if the alignment on neighboring atoms were different , the hopping of electrons from one atom to the next would generate on - site hund s rule interactions and thus increase the energy .
these interactions can only be avoided if the spin alignment on neighboring atoms is the same , implying _ global _ ferromagnetism .
this mechanism for the stabilization of ferromagnetic order works all the better the larger the number @xmath129 of orbitals , i. e. bands , is onto which an electron can hop ; it does not work for a single band ( @xmath152 ) .
therefore one should expect band degeneracy to favor ferromagnetism .
this very qualitative picture is indeed found in stoner mean - field theory @xcite : since the non - interacting dos is proportional to the number of degenerate bands , the critical interaction decreases as @xmath153 . on the other hand ,
few _ exact _ results are known in the case of degenerate bands .
for example , for a one - dimensional chain with two orbitals ( @xmath154 ) and infinite on - site coulomb interactions it has been shown that the ground state is ferromagnetic for @xmath155 electrons @xcite .
this statement has been extended to @xmath156 , i.e. up to one electron less than half - filling @xcite .
the important point is that on a one - dimensional chain with only one orbital and hopping between nearest neighbors , nagaoka s theorem is not applicable , since the lattice does not have loops . in that case
the ground state is degenerate with respect to the total spin @xmath157 . on the other hand , if there are two orbitals the loop property is restored , and the ferromagnetic states with maximum spin become the only ground states .
furthermore , the following rigorous result can be established for the hamiltonian ( [ mbham ] ) at half - filling @xcite . for @xmath158 electrons ,
the ground states of @xmath159 have maximum spin @xmath160 if @xmath161 where @xmath162 are the critical values for @xmath0 in the _ single _ band system , as given in ( [ bounds1])-([bounds2 ] ) .
the meaning of these bounds is the following . the requirement that @xmath163 and @xmath136 be greater than a certain threshold leads to an alignment of the electronic spins on an isolated atom .
on the other hand , ferromagnetism within each band is brought about by the next - neighbor exchange @xmath164 and the hubbard interaction @xmath0 larger than a threshold related to @xmath162 .
the combination of these two effects ( using the fact that @xmath139 for equivalent bands ) leads to a critical value for the hund s rule coupling @xmath136 , which indeed becomes lower as the number of bands increases .
while this result contains some ideas of slater s picture , it does not explain the itinerant aspects of multi - band ferromagnetism , since at half - filling the ferromagnetic ground states are insulating .
so far , this result can only be modified to apply also to nagaoka s case ( one hole , @xmath165 ) @xcite .
the stability of ferromagnetism is intimately linked with the structure of the underlying lattice and the kinetic energy ( i.e. the hopping ) of the electrons @xcite .
this is supported by several facts : ( a ) nagaoka s proof of ferromagnetism in the hubbard model for a single hole at @xmath57 @xcite depends on the existence of closed loops along which the hole can move , and ( b ) on bipartite lattices @xmath166ferromagnetism is the generic magnetic state making it hard for ferromagnetism to become stable .
hence @xmath167-bipartite lattices with loops ( or with a kinetic energy involving hopping between nearest @xmath168 next - nearest neighbors sites effectively leading to a motion on loops ) should be expected to support ferromagnetism because the competing antiferromagnetic tendencies are severely weakened , and because the corresponding dos of non - interacting electrons is asymmetric and thus has a peak at a non - symmetric position . indeed ,
a peak at one of the band edges as in the case of the fcc lattice is favorable for ferromagnetism @xcite .
this is supported by the observation @xcite that co and ni , having non - bipartite hcp and fcc lattice structure , respectively , show a full magnetization while bcc - fe has only a partial magnetization . to gain insight into why a dos with a peak ( or more precisely with a high spectral density ) at the band edge , e.g. at the lower band edge for @xmath169 , may be favorable for ferromagnetism
, we study non - interacting electrons with the following model dos : @xmath170 where @xmath171 is the width of the band and @xmath172 parameterizes the asymmetry .
the lower band edge is @xmath173 such that the first moment of @xmath174 vanishes . for @xmath175
the dos has a triangular shape with the peak at the lower ( upper ) edge , while for @xmath176 it is flat .
we wish to calculate the energy difference @xmath177 between the fully polarized and the paramagnetic state @xmath178 as a function of @xmath172 and the band filling .
it is easy to confirm that for all @xmath116 @xmath177 is lowest for @xmath179 .
the reason is this : in the paramagnetic state @xmath53 non - interacting electrons ( @xmath180 electrons with spin up and down each ) fill the lowest @xmath180 @xmath181-states up to an energy @xmath182 , while in the ferromagnetic state the @xmath53 singly occupied @xmath181-states below @xmath183 are filled , i.e. the @xmath180 @xmath181-states with energy above @xmath182 are also occupied .
the higher the dos is at the lower edge the less the additional @xmath180 @xmath181-states are forced into high - energy states , i.e. the lower @xmath184 will be .
then @xmath177 is kept at a minimum . as explained in the previous section , a lattice structure which gives a high dos at low energies
is expected to be favorable for ferromagnetism ( at @xmath169 ) because this situation reduces the loss in kinetic energy .
the question remains : is a strongly peaked dos _ sufficient _ to induce ferromagnetic order in the single band hubbard model without additional interactions ?
_ _ in this section we discuss the results of a qmc - investigation of the stability of ferromagnetism in the single band hubbard model in the limit of infinite dimensions . based on the considerations in sec . 3.3.1
, the fcc lattice is expected to be a good candidate for ferromagnetism because of its high ( divergent ) dos at the lower band edge .
the fcc - lattice can be generalized to higher dimensions in different ways @xcite .
here we use the definition of an fcc lattice as the set of all points with integer cubic coordinates summing up to an even integer @xcite .
it is a non - bipartite bravais lattice for any dimension @xmath185 .
for @xmath186 it is identical to the square lattice .
nearest neighbors are connected by _ two different _ unit vectors on a simple hypercubic ( hc ) lattice .
the coordination number is hence @xmath187 . with the proper scaling of the hopping term , ( [ scale ] )
, the non - interacting dos of the generalized fcc lattice can be calculated in @xmath107 @xcite as : @xmath188 which has a strong square - root divergency at the lower band edge , @xmath189 , and no upper band edge .
since we choose a _ positive _ hopping integral @xmath190 , @xmath191 might be regarded as the dos of holes rather than electrons .
a full polarization of holes would hence correspond to a _ maximal _ ( though not full ) polarization in a more than half - filled band . _ _ _ _ _ _ while the three dimensional fcc lattice has no square - root but only a logarithmic divergency at the band edge it is worth mentioning that a square - root divergency arises on the fcc lattice in any dimension if there is an additional next nearest neighbor hopping of the size @xmath192 between sites that are linked by two unit vectors in the _ same _ direction on the hc lattice .
the energy dispersion and the dos of this so - called `` half - hypercubic '' ( hh ) lattice @xcite can easily be obtained from the hc lattice as : @xmath193 in the limit @xmath108 , the hh and fcc lattices become equivalent . _ _ the fact that the fcc lattice provides a good ` environment ' for ferromagnetism has also been supported by variational studies of the stability of the nagaoka state @xcite .
variational calculations provide limits for the critical density @xmath194 and the critical interaction @xmath97 where saturated ferromagnetism becomes unstable : while for the hc lattice in @xmath107 the stability regime shrinks to the point @xmath195 , @xmath196 @xcite , for the fcc dos ( [ glferro1 ] ) there is a critical line @xmath197 with @xmath198 and @xmath199 @xcite .
the nagaoka state is always unstable in the case of electron doping @xmath200 .
recently , these variational boundaries were qualitatively confirmed by uhrig s calculation of the exact single spin - flip energy of the nagaoka state in @xmath107 @xcite .
while on the hh lattice @xmath97 vanishes at low densities , @xmath97 remains finite for all densities in the case of the `` laminated '' lattice which is a different generalization of the fcc lattice without a divergent dos at the band edge .
his results , too , emphasize the subtle dependence of the stability of ferromagnetism on the lattice structure .
antiferromagnetism is not expected on the fcc - lattice in high dimensions because the difference of the numbers of not frustrated bonds and frustrated bonds is only of the order of @xmath109 resulting in an effective field of the order of @xmath201 @xcite . even in @xmath202 antiferromagnetism
is frustrated and is expected only very close to half - filling . to detect a ferromagnetic instability we calculated the temperature dependence of the uniform static susceptibility , @xmath203 , from the two - particle correlation functions @xcite . at an intermediate interaction strength of @xmath204 we found the ferromagnetic response to be strongest around quarter filling ( @xmath205 ) .
@xmath203 obeys a curie - weiss law ( fig .
[ figferro1 ] ) and the curie temperature @xmath206 can safely be extrapolated from the zero of @xmath207 to a value of @xmath208 at @xmath209 .
below @xmath206 the magnetization @xmath210 grows rapidly , reaching more than 80% of the fully polarized value ( @xmath211 ) at the lowest temperature which is only 30% below @xmath206 .
the three data points @xmath212 ( fig .
[ figferro1 ] ) are consistent with a brillouin function with the same critical temperature of @xmath213 and an extrapolated full polarization at @xmath120 .
a saturated ground state magnetization is also consistent with the single spin - flip energy of the fully polarized state which is positive at the present parameter values @xcite
. translated to the three dimensional fcc lattice with @xmath214 nearest neighbors and a bandwidth of @xmath215 , the critical temperature becomes @xmath216 @xcite .
thus , despite the oversimplifications of the single band hubbard model , the resulting curie temperature has a realistic order of magnitude of 500 - 800k for typical values of @xmath217 around 5 ev . in order to answer the question if the system is still metallic we also calculated the single particle spectrum @xcite .
we find that the system is metallic since both spectra have a finite value at the fermi level .
we conclude that within the dynamical mean - field theory a dos with sufficiently large spectral weight at low energies is able to induce itinerant ferromagnetism in the single band hubbard model .
detailed investigations of the dependence of ferromagnetism on the electronic density , the interaction strength , and in particular the extension to more realistic densities of states ( e.g. fcc lattice in @xmath202 ) is under progress and will be presented elsewhere @xcite . in the @xmath107
limit the dynamics of the correlated system is fully taken into account , while spatial fluctuations are suppressed @xcite .
one might therefore suspect that the stability of ferromagnetism is somehow @xmath218estimated in this approach ( in particular since the rivaling antiferromagnetism is completely absent on the fcc lattice ) .
however , most recently similar results were reported for the hubbard model with nearest and next - nearest neighbor hopping @xmath55 and @xmath219 , respectively , in dimensions @xmath1 @xcite and @xmath186 @xcite , which are consistent with the results in @xmath107 .
hence the existence of itinerant ferromagnetism in the pure hubbard model with a suitable kinetic energy seems to be established at last .
an issue related to the stability of ( itinerant ) ferromagnetism in the hubbard model is the question concerning metamagnetism @xcite in this and other models . here
it is an external magnetic field @xmath220 which helps to suppress the ( not necessary long - range ) antiferromagnetic correlations in the system and thereby induces a pronounced transition from a state with low magnetization to one with high magnetization . at the metamagnetic transition the magnetization curve @xmath221 shows an up - turn such that the susceptibility @xmath222 has some kind of maximum
this feature serves as a convenient general definition for `` metamagnetism '' .
metamagnetic transitions were first observed in strongly anisotropic antiferromagnets of which fecl@xmath223 and dy@xmath224al@xmath225o@xmath226 ( dag ) are well - studied prototypes @xcite .
these materials are insulators where the valence electrons are localized at the fe and dy ions , respectively
. the arising local moments order antiferromagnetically and are strongly anisotropic in the sense that they are constrained to lie along an easy axis @xmath227 . in this case
a spin - flop transition in an external magnetic field @xmath228 can not occur .
apart from the above materials there are also conducting systems that most probably belong to this class , e.g. the conductors ua@xmath229b@xmath230 ( where @xmath231 , @xmath232 ; @xmath233 , @xmath234 ) @xcite , smmn@xmath223ge@xmath223 @xcite and tbrh@xmath235ir@xmath230si@xmath223 @xcite .
hitherto completely different theories are employed to describe metamagnetic phase transitions in these different , strongly anisotropic antiferromagnets .
investigations of localized systems are usually based on the ising model , where more than one interaction has to be introduced to describe the experimentally observed first order phase transitions @xcite
. with antiferromagnetic coupling @xmath65 between the @xmath93 nearest - neighbor ( nn ) spins and a ferromagnetic coupling @xmath236 between the @xmath237 next - nearest - neighbors ( nnn ) one obtains @xmath238 in weiss mean - field theory this model shows two different types of phase diagrams depending on the parameter @xmath239 @xcite .
for @xmath240 the first and second order phase transition line join smoothly at the same point , producing a tricritical point ( tcp ) , while for @xmath241 there is no common endpoint ( see fig . [ scheme ] ) .
however the scenario of fig .
[ scheme]b was not found when evaluating ( [ isi ] ) beyond mean - field theory @xcite . for _ itinerant electron metamagnetism _
( iem ) moriya and usami @xcite proposed a landau theory , where the parameters have to be deduced from microscopic models .
the idea is to calculate first the independent electron band structure and to introduce the coulomb interaction within the random phase approximation .
it is our purpose to investigate the origin of metamagnetism in strongly anisotropic antiferromagnets from a microscopic , quantum - mechanical point of view , and to describe different kinds of metamagnets ( i.e. metallic _ and _ insulating , band - like _ and _ localized systems ) qualitatively within a single model . to this end
we study the hubbard model ( [ hub ] ) with the additional constraint that the antiferromagnetic magnetization @xmath242 lies _
parallel _ to the external magnetic field @xmath243 @xcite . in this way
the existence of an easy axis @xmath227 along which @xmath243 is directed , such that @xmath244 , is incorporated in a natural way . by this approach ,
both kinetic energy and coulomb interaction are captured microscopically , whereas the relativistic corrections ( responsible for the easy axis ) are not .
this procedure is justified since the relativistic corrections are of @xmath245 ev ) and are thus small compared to the kinetic and coulomb energy which are of @xmath246ev ) .
therefore the existence of an anisotropy axis @xmath227 and the correlations described by the hubbard model are quite unrelated .
note , that the existing ising and iem theories do not treat the kinetic energy and coulomb interaction microscopically , but within an effective model .
a perturbative treatment of the hubbard model with easy axis in the weak and strong coupling limit shows that the appearance of a tri- or multicritical point is a delicate matter , since neither of these two limits is able to describe a change of the transition from first to second order @xcite
. apparently the entire transition scenario depends sensitively on the value of the electronic on - site interaction @xmath0 . to study this point in greater detail
we have to go to intermediate coupling . in this non - perturbative regime
we employ again qmc simulations to calculate the magnetization @xmath221 and the staggered magnetization @xmath247 of the hubbard model ( [ hub ] ) in @xmath107 @xcite .
as the results do not much depend on the precise form of the density of states we choose @xmath248^{1/2}/(2 \pi t^{*2})$ ] , setting @xmath249 in the following .
all calculations are performed at half - filling .
the results for @xmath221 and @xmath247 are used to construct the @xmath250 phase diagram at @xmath204 ( fig .
[ hvst ] ) .
it displays all the features of fig .
[ scheme]b .
in particular , the first order line continues _ into _ the ordered phase , separating two different af phases : af@xmath251 ( where @xmath252 ) and af@xmath253 ( where @xmath254 ) .
the position of its endpoint can not , at present , be determined accurately ( dotted line ) .
this phase diagram is surprisingly similar to the experimental phase diagram of febr@xmath223 @xcite
. a change in @xmath0 and the filling @xmath116 will affect the phase diagram quantitatively and qualitatively .
these results will be reported elsewhere @xcite .
in the last few years , and especially most recently , considerable progress was made in our understanding of the microscopic origin of itinerant ferromagnetism .
these results were obtained on the basis of well - defined lattice models of correlated electrons , of which the one - band hubbard model is a particularly important ingredient , by applying new , non - perturbative techniques , ranging from rigorous to large - scale numerical methods .
there exists convincing evidence now that on appropriate , non - artificial lattices , or for an appropriate kinetic energy , itinerant ferromagnetism is stable even in the pure hubbard model , for electronic densities not too close to half - filling and large enough @xmath0 .
important ingredients are : + ( i ) lattices with loops ( or a kinetic energy allowing for motion on loops , e.g. with @xmath255 hopping ) such that the nagaoka mechanism works and antiferromagnetism is suppressed , + ( ii ) a large spectral weight near the band edge .
( we note that this condition goes far beyond the mean - field stoner criterion for ferromagnetism , @xmath256 where @xmath257 is the dos at the fermi energy . ) the direct exchange interaction , as well as band degeneracy , will strongly reduce the critical value of @xmath0 above which ferromagnetism becomes stable .
furthermore , using the dynamical mean - field theory to solve the hubbard model with easy - axis in the intermediate coupling regime , it appears to be possible to describe various , different forms of metamagnetism within a single microscopic theory . the development and application of controlled , non - perturbative techniques will continue to be of particular importance for the investigation of correlated electron systems .
one of us ( dv ) thanks elliott lieb for very useful discussions . in its early stages
this work was supported in part by the deutsche forschungsgemeinschaft under sfb 341 .
m. m. p. de azevedo et al .
, j. magn .
magn . mater .
* 140 - 144 * , 1557 ( 1995 ) ; j. pelloth et al . ,
b * 52 * , 15372 ( 1995 ) ; h. a. katori , k. katsumata , and m. katori , phys .
b * 54 * , r 9620 ( 1996 ) ; k. katsumata , h. a. katori , s. m. shapiro , and g. shirane , preprint . | the microscopic basis for the stability of itinerant ferromagnetism in correlated electron systems is examined . to this end several routes to ferromagnetism
are explored , using both rigorous methods valid in arbitrary spatial dimensions , as well as quantum monte carlo investigations in the limit of infinite dimensions ( dynamical mean - field theory ) . in particular
we discuss the qualitative and quantitative importance of ( i ) the direct heisenberg exchange coupling , ( ii ) band degeneracy plus hund s rule coupling , and ( iii ) a high spectral density near the band edges caused by an appropriate lattice structure and/or kinetic energy of the electrons .
we furnish evidence of the stability of itinerant ferromagnetism in the pure hubbard model for appropriate lattices at electronic densities not too close to half - filling and large enough @xmath0 .
already a weak direct exchange interaction , as well as band degeneracy , is found to reduce the critical value of @xmath0 above which ferromagnetism becomes stable considerably . using similar numerical techniques the hubbard model with an easy axis
is studied to explain metamagnetism in strongly anisotropic antiferromagnets from a unifying microscopic point of view .
+ + 71.27.+a,75.10.lp |
polarized interacting electrons in the lowest landau level in the tao - thouless limit , which corresponds to a thin torus or cylinder , has been a subject of recent study , see _ eg _ ref . .
in this limit the problem becomes equivalent to a one - dimensional lattice model with a purely electrostatic repulsive interaction between the electrons .
the hamiltonian can be diagonalized for any filling factor @xmath0 and the energy eigenstates are the states where the electrons have fixed positions on the lattice@xcite . in the ground state , the tao - thouless state ,
the electrons are as far separated as possible ; the state has a unit cell with @xmath3 electrons on @xmath4 sites and is @xmath4-fold degenerate .
there is a gap to all excitations and the low energy charged excitations are domain walls between the degenerate ground states ; these domain walls have charge @xmath5 .
the tao - thouless states , and their excitations , are limits of hierarchy wave functions describing abelian quantum hall states and there is strong evidence that they develop adiabatically into such states as the circumference of the torus increases in cases where such states are observed .
the hierarchy construction of quantum hall states is manifest in the tt - limit : each state is obtained by condensing quasiparticles , _ ie _ domain walls , in a parent state@xcite . in our opinion ,
the tt - states are abelian quantum hall states . in this article
we consider the low - energy excitations in the tt - limit in general . it was previously noted that the unit cell @xmath6 for the tt qround state can be written as @xmath7 where @xmath8 is the quasielectron and @xmath9 is the quasihole excitation @xcite .
thus the ground state is a string @xmath10 .
interchanging a nearest neighbor particle and hole gives @xmath11 .
this state has two nearby domain walls , @xmath12 and @xmath13these are the quasiparticles with charges @xmath5 . here
we show that , given a certain condition on the electron - electron interaction , the low - energy sector of states consists of all reorderings of the @xmath8 s and @xmath9 s that make up the ground state .
we study how the low - energy states develop away from the tt - limit using exact diagonalisation at and around @xmath1 and @xmath2 .
we find that the low energy sector for any circumference consists of states that develop from the low energy states in the tt - limit .
the exclusion statistics @xcite was determined for @xmath1 in the tt - limit in ref . .
we here generalize this to arbitrary filling factor @xmath0 and find a simple result for the exclusion statistics , which , however , in general differs from the values established for the excitations in the physical regime@xcite .
in particular it disagrees in certain cases with the values obtained by assuming the quasiparticles are anyons and counting anyon wave functions .
for example , at filling factors 1/3 and 2/5 the diagonal statistics in the tt - limit agrees with that in the physical regime for the holes but differs for the particles .
we study numerically for these filling factors how the exclusion statistics for the quasiparticles change as the circumference increases and the values that are consistent with the particles being anyons are recovered .
a recent discussion of exclusion statistics in connection to entanglement spectra can be found in ref .
, see also ref . .
the quasiparticles in the tt - limit are domain walls and as a consequence the exchange statistics can not be calculated in this limit@xcite .
however , the exchange statistics can be considered by making use of the tt - limit and invoking the adiabatic continuity to the bulk quasiparticles@xcite . in sec .
[ low_section ] we determine the low - energy states in the tt - limit and in sec . [ numerics ] we investigate to what extent these results hold away from the tt - limit by numerically investigating small systems .
the exclusion statistics is discussed in sec .
[ exclusion ] and our findings are discussed in sec .
[ discussion ] .
derivations and details of our analysis are given in three appendices .
in this section we identify the space of low - energy states in the tao - thouless limit for arbitrary filling factor @xmath14 .
mathematical details are given in the appendices and details about the tt - limit can be found in ref . .
we consider a single landau level of spin polarized electrons on a torus with lengths @xmath15 .
using landau gauge , the natural one - particle eigenstates @xmath16 , @xmath17 , are gaussians concentrated along circles around the torus ( in the @xmath18-direction say , @xmath19 is the momentum in this direction ) .
this maps the landau level onto a one - dimensional lattice model with @xmath20 sites numbered by @xmath19 , where each site is either empty or occupied by an electron .
the lattice constant is @xmath21 in units of the magnetic length @xmath22 .
a basis of many - particle states is given by @xmath23 , where @xmath24 , and @xmath25 is the number of electrons ; the filling factor is @xmath26 . since the position @xmath19 gives the conserved momentum
, it follows that a general translationally invariant two - body hamiltonian consists of all terms where two electrons hop while preserving the position of their center of mass .
let @xmath27 be the amplitude for this process when the separations between the electrons are @xmath28 .
when @xmath29 , the lattice constant goes to infinity and the overlap between the single - particle states with different @xmath19 vanishes . as a consequence ,
only the electrostatic terms @xmath30 remain .
this is the thin torus , or tao - thouless , limit .
it should be noted that one can think of this limit as simply being a change of the hamiltonian , _ ie _ the @xmath27 , while keeping the configuration space unchanged .
thus the tt - limit does describe a two - dimensional qh system , albeit with a peculiar anisotropic interaction .
in the tt - limit the interacting electron problem can be solved exactly since the interaction is purely electrostatic : the energy eigenstates are simply the states @xmath23 . for an interaction that obeys the concavity condition @xmath31 , the ground state and the low energy excitations
have the same qualitative properties as the standard abelian qh states .
the @xmath4-fold degenerate ground state is crystal - like with a unit cell of length @xmath4 containing @xmath3 electrons .
there is a gap to excitations and the lowest energy charged excitations have charge @xmath5 and are the domain walls between the degenerate ground states .
the unit cell @xmath6 for the ground state is obtained by considering a circle with @xmath4 equidistant sites and @xmath3 equidistant electrons .
letting each electron relax to the closest site gives the unit cell@xcite .
is then placed at site @xmath33 , where @xmath34 \ .\ ] ] here , @xmath35 denotes equality modulo @xmath4 and @xmath36 $ ] is the rounding function : @xmath37=n$ ] if @xmath38 ; the parameter @xmath39 corresponds to the freedom to rotate all the initial ( unrelaxed ) electron positions by @xmath39 , or , equivalently , to rotate the sites by @xmath40 .
there is strong evidence that the tt - ground state , given by at @xmath0 , @xmath4 odd , is adiabatically connected to the standard ground state that describes an abelian qh state for a realistic interaction .
we claim that the tt - states ( [ relaxation ] ) are abelian qauntum hall states .
the hierarchy construction of qh states is proven in the tt - limit : each tt - ground state is a condensate of quasiparticles in another parent tt - ground state .
the unit cell @xmath41 can be represented in @xmath4 different ways by translating the cell ( imposing periodic boundary conditions ) .
transposing the cell gives a cell that is a translation of the original one .
it is convenient to introduce a standard representation @xmath42 ( @xmath43 ) which is the cell where the electrons are as far to the right ( left ) as possibile , _
ie _ the cell with the maximal ( minimal ) value of @xmath44 .
the ground state unit cell at @xmath0 can be written as @xmath45 , see [ app : ph - properties ] , where @xmath46 and @xmath47 are the quasihole and quasielectron with charges @xmath5 . when inserting @xmath47 or @xmath46 between two unit cells in the ground state they produce domain walls with charges @xmath48 .
the number of sites , @xmath49 , and the number of electrons , @xmath50 , in the quasiparticles are given by @xmath51 where @xmath52 ; throughout the article we use @xmath53 and @xmath8 interchangeably to denote the quasielectrons ( and @xmath54 and @xmath9 to denote the quasiholes ) .
these relations are derived in [ app : ph - properties ] . here , @xmath55 is euler s totient function .
( @xmath55 is the number of positive integers less than or equal to @xmath56 that are coprime to @xmath56 , _ ie _ have no common positive factor with @xmath56 other than 1 . ) @xmath57 and @xmath58 are the lengths of ( _ ie _ the number of sites in ) the quasielectron and quasihole in the tt - limit ; @xmath59 .
for example , for @xmath1 , 2/5 , 4/11 and 5/13 one finds @xmath60 , @xmath61 , @xmath62 and @xmath63 respectively .
moreover , due to the hierarchical structure , @xmath46 and @xmath47 are themselves the unit cells at @xmath64 .
this means that they are given by the relaxation procedure above .
let us consider the two examples @xmath65 and @xmath66 . for @xmath65 , choosing @xmath67 in ( [ relaxation ] ) , we find the positions of the electrons in the unit cell to be : @xmath68 , corresponding to @xmath69 ( since @xmath44 is maximal this is in the @xmath70-representation ) . from ( [ ln ] )
we find that the quasielectron and quasihole have lengths @xmath71 , @xmath72 and contain @xmath73 electrons respectively , hence @xmath74 and @xmath75 .
note that 001 is the unit cell for the ground state at @xmath1 as well as the quasihole at @xmath65 ( the corresponding result holds for 01 and 1/2 , although the latter does not continue to a qh state since @xmath76 is even ) .
the cells are written in the @xmath70-representation and we see that @xmath77 . inspecting the string @xmath78 we see that by moving the first electron one step to the left , or the last electron one step to the right ( assuming periodic boundary conditions ) , we obtain a translation of @xmath79 . when these moves are performed in the tt - ground state , _ ie _ in a string of @xmath77
, one finds that they interchange a nearest neighbor @xmath8 and @xmath9 ; the first gives @xmath80 whereas the second leads to @xmath81 . repeating the analysis for @xmath82
, we find from ( [ relaxation ] ) ( with @xmath67 ) the positions of the electrons in the unit cell to be : @xmath83 , corresponding to @xmath84 .
translating @xmath6 to maximize @xmath44 we find the unit cell @xmath85 . from ( [ ln ] ) we find that the quasielectron and quasihole have lengths @xmath86 , @xmath87 and contain @xmath88 , @xmath89 electrons respectively .
applying the relaxation procedure ( [ relaxation ] ) , then gives @xmath90 and @xmath91 .
note that 00101 is the unit cell for the ground state at @xmath65 as well as the quasielectron at @xmath66 ( similarly for 00100101 and 3/8 ) .
we have chosen to write the cells in @xmath70-representation and we see that @xmath77 . inspecting the string @xmath92 we see that by moving the second electron one step to the left , or the last electron one step to the right ( assuming periodic boundary conditions ) , we obtain a translation of @xmath79 . when these moves are performed in the tt - ground state , one finds , just as for 2/5 , that they interchange a nearest neighbor @xmath8 and @xmath9 ; the first gives @xmath80 whereas the second leads to @xmath81 .
these results obtained for @xmath93 and @xmath94 generalize to arbitrary filling factor and will be crucial in identifying the low energy excitations .
we show in [ app : ph - properties ] that in the unit cell @xmath77 that gives the ground state at @xmath95 ( _ ie _ the unit cell that is given by ( [ relaxation ] ) ) there are precisely two electrons with the property that moving one of them one step ( to the left for the one to the left and to the right for the one to the right ) interchanges a nearest neighbor @xmath8 and @xmath9 in the string of @xmath79 that make up the ground state ; moving the left electron gives @xmath80 whereas moving the right leads to @xmath80 .
( the first acts within a unit cell @xmath96 , whereas the second acts within a translated unit cell . )
furthermore , the energy of these excitations is of the order of @xmath97 ; we denote this by @xmath98 , and it is the minimum energy excitation of the system ( except for the zero energy excitations that translate the whole system giving one of the degenerate ground states ) .
all other excitations are of order @xmath99 , @xmath100 .
( we assume @xmath101 , in the tt - limit , see below . )
we assume that @xmath8 , @xmath9 ( just as @xmath79 ) are given in the @xmath70-representation ; this makes the identification of a certain state given in the @xmath102 notation as a @xmath103 state unique .
note that interchanging a nearest neighbor @xmath9 and @xmath8 in the ground state , which is a string of @xmath104 , corresponds to creating two domain walls , @xmath105 and @xmath13 with charges @xmath5 , next to each other this is a minimally separated particle - hole pair , which is a natural minimum energy excitation .
for @xmath106 the above still holds with the exception that there is now one electron per unit cell that can be moved both to the right and to the left leading to an interchange of @xmath8 and @xmath9 .
any interchange @xmath107 , wherever it takes place in an arbitrary string of @xmath9 and @xmath8 , is obtained by a minimum energy hop @xmath108 with energy of order @xmath109 as described above .
this follows since @xmath110 differ in energy by @xmath97 ; whereas @xmath111 and @xmath112 differ in energy by @xmath113 .
the last statement can be understood by noting that the two states are degenerate if the surrounding elements are symmetric around @xmath114 .
the energy difference is brought about by any antisymmetry @xmath19 sites away . for a proof of these results , see [ app : ph - proofs ] .
if @xmath115 , for all @xmath19 , then we can identify the low - energy sector as consisting of states that can be described using only @xmath9 and @xmath8 .
however , several comments are in order . first , taking the limit of a standard interaction this condition may not be satisfied .
it holds for a screened coulomb interaction , @xmath116 ( periodized over the torus ) but not for the unscreened case , @xmath117 , see fig .
[ fig:5_17_yuk ] below .
( because of the periodization in the @xmath118direction , the interaction in the tt - limit , @xmath119 , becomes the electrostatic interaction between two homogeneously charged parallel lines , giving @xmath120 for the unscreened coulomb case and @xmath121 for the screened case , where @xmath122 is some non - exponential function . )
what one finds quite generally though is that @xmath123 .
second , while the specific form of the interaction in the physical regime is well - defined , this is not the case in the tt - limit
one may simply modify the interaction in this limit so that the condition is fulfilled .
third , even if @xmath115 it may be that a sufficiently large number of @xmath8 , @xmath9 excitations has a higher energy than the lowest energy states that are not of this type . bearing the provisos above in mind , we identify the low - energy sector as consisting of states that can be described using only @xmath9 and @xmath8 . one further argument for this is of course that it is in accord with the expectation in the physical regime . as shown in [ app : ph - proofs ] these states and no others differ in energy compared to the ground state by @xmath124 . as an illustrative aside we consider the hierarchy construction of 5/13 . since @xmath125 it follows that 5/13 is obtained by starting from the laughlin state at 1/3 and condensing quasielectrons to obtain 2/5 ; condensing quasiholes in 2/5 then gives 5/13 . in the tt - limit
, the 1/3 ground state is 001 and @xmath126 , @xmath127 .
inserting @xmath74 in the 001 ground state once per unit cell gives a state with unit cell 00101this is the 2/5 ground state .
( note that if one had inserted a lower homogeneous density of @xmath74 one would have obtained the states @xmath128 ; these are the @xmath129 ground states . )
one finds from ( [ ln ] ) that the quasiparticle excitations at 2/5 are @xmath74 , @xmath75 . inserting @xmath75 once per every two unit cells gives @xmath130this
is the 5/13 ground state .
this illustrates how the hierarchy construction , where states are obtained by successive condensation of quasiparticles , is manifest in the tt - limit .
in this section we consider how the low - energy sector identified above in the tt - limit develops as @xmath131 increases .
we do this by performing exact diagonalization of small systems at and around filling factors @xmath132 .
the results are presented in fig . 1 - 5 .
in each figure the spectra are shown for @xmath133 and an unscreened coulomb interaction , @xmath117 , that is periodized , in the @xmath131 and @xmath18 directions , over the torus .
the black and red lines are continuations of the low - energy states identified above in the tt - limit , whereas grey lines correspond to higher energy states in this limit .
red lines mark the very lowest states in the tt - limit , where they are identified in terms of particle and/or hole excitations of the ground state .
we use the convention that when two lines with different colors but with the same quantum numbers approach each other and an avoided crossing occurs then the lower line is chosen as red or black ( in this order of priority ) after the crossing . in each figure
@xmath97 is indicated by a star ; this is , as discussed above , a characteristic energy for the low energy excitations in the tt - limit .
energies are given relative to the energy of the state that has lowest energy in the tt - limit and scaled by @xmath131 to cancel the @xmath134 behaviour of @xmath27 as @xmath119 .
the main conclusion from these figures is that for each filling factor the lowest energy states at any @xmath131 develop from states in the low - energy sector in the tt - limit ( black or red lines ) , without mixing of high - energy tt - states ( grey lines ) . there is no obvious simple rule for precisely how many of the lowest states this is true for , but it is always true for some set for all the @xmath0 that we have studied and we propose that it is true in general . how strong this restriction on the number of states involved in determining the low energy sector is depends on the filling factor : at @xmath135 , since @xmath136 all states are in the low - energy sector in the tt - limit , whereas the fraction of low - energy states decreases rapidly with increasing @xmath4 .
[ fig:4_10_ph_5_15_ph ] shows the spectra for @xmath1 and @xmath2 .
the ground states and the one particle - hole excitations ( identified in the tt - state ) are marked in red .
as noted before @xcite the ground states develop continuously , without the gap closing .
the one particle - hole excitations remain low energy excitations for all @xmath131 , but there are additional ( black ) states that mix with these for large @xmath131
. some of these are presumably states that approach the particle - hole tt - states in the limit @xmath137 ( for fixed @xmath138 ) . with six unit cells @xmath139 ( upper panel ) and @xmath65 with three cells @xmath140 ( lower panel ) .
energies , scaled by @xmath131 , are shown for @xmath141 and are given relative to the energy of the state that has lowest energy in the tt - limit .
red lines denote the ground states and the one particle - hole excitations ( identified in the tt - limit ) .
black lines denote the remaining low energy states . the characteristic variation in the energy of the low - energy excitations in the tt - limit , @xmath97 ,
is indicated by a star .
( each state is , at least , 3 and 5-fold degenerate respectively . )
[ fig:4_10_ph_5_15_ph],title="fig : " ] with six unit cells @xmath139 ( upper panel ) and @xmath65 with three cells @xmath140 ( lower panel ) .
energies , scaled by @xmath131 , are shown for @xmath141 and are given relative to the energy of the state that has lowest energy in the tt - limit .
red lines denote the ground states and the one particle - hole excitations ( identified in the tt - limit ) .
black lines denote the remaining low energy states .
the characteristic variation in the energy of the low - energy excitations in the tt - limit , @xmath97 , is indicated by a star .
( each state is , at least , 3 and 5-fold degenerate respectively . )
[ fig:4_10_ph_5_15_ph],title="fig : " ] fig .
[ fig:5_15_h_hh ] shows the energy spectra for one and two holes in the 1/3 ground state , whereas fig .
[ fig:5_15_p_pp ] shows the data for one and two particles .
the corresponding spectra for 2/5 are shown in fig .
[ fig:4_10_h_hh ] and [ fig:4_10_p_pp ] .
we see that the one hole , as well as the one particle , states are the lowest energy states for all @xmath131 at both 1/3 and 2/5 .
the same holds true for the two hole states , however at 1/3 the gap at small @xmath131 is small .
thus for all these cases the lowest energy states identified in the tt - limit remain the low - energy states for all @xmath131 .
for the two - quasiparticle excitations the situation is different . at 1/3
one of the low energy states identified in the tt - limit disappears to higher energies as @xmath131 increases .
similarly , at 2/5 the low energy states also change .
we conclude that the few particle and hole excitations identified in the tt - limit show a large , but not complete , correspondence to the states at large @xmath131 . in the next section
we analyze these excitations , and the relation between the tt - limit and the bulk , in terms of exclusion statistics . in all the figures something happens in the spectrum at @xmath142 ,
this is where the hopping matrix elements @xmath143 become important . in fig .
[ fig:4_10_h_hh ] , for @xmath144 , there is a change of ground state .
note that this describes the ground state at 3/8 as well as two holes at 2/5 .
the change in ground state as @xmath131 increases is typical for even denominator filling factors @xcite .
the finite size dependence of the crossover caused by the onset of hopping , has been carefully studied at @xmath135 where it was found that a transition occurs at @xmath145 virtually independent of @xmath146 @xcite .
this agrees with the very weak dependence of size we see here and we conclude that the crossover remains at finite @xmath131 as @xmath146 increases .
we have verified that the qualitative features obtained above are unchanged if the unscreened coulomb potential is replaced by a screened one or by a short - range , pseudo - potential , interaction . and two ( 5,17 ) holes at @xmath1 with five unit cells these excitations , identified in the tt - limit , are shown as red lines . [ fig:5_15_h_hh],title="fig : " ] and two ( 5,17 )
holes at @xmath1 with five unit cells these excitations , identified in the tt - limit , are shown as red lines .
[ fig:5_15_h_hh],title="fig : " ] and two ( 7,19 ) particles at @xmath1 with five unit cells these excitations , identified in the tt - limit , are shown as red lines .
[ fig:5_15_p_pp],title="fig : " ] and two ( 7,19 ) particles at @xmath1 with five unit cells
these excitations , identified in the tt - limit , are shown as red lines .
[ fig:5_15_p_pp],title="fig : " ] and two ( 6,16 ) holes at @xmath65 with two unit cells these excitations , identified in the tt - limit , are shown as red lines .
[ fig:4_10_h_hh],title="fig : " ] and two ( 6,16 ) holes at @xmath65 with two unit cells these excitations , identified in the tt - limit , are shown as red lines .
[ fig:4_10_h_hh],title="fig : " ] and two ( 6,14 ) particles at @xmath65 with two unit cells
these excitations , identified in the tt - limit , are shown as red lines .
[ fig:4_10_p_pp],title="fig : " ] and two ( 6,14 ) particles at @xmath65 with two unit cells these excitations , identified in the tt - limit ,
are shown as red lines .
[ fig:4_10_p_pp],title="fig : " ]
the low energy quasiparticle states can be analyzed in terms of the exclusion statistics introduced by haldane @xcite .
the exclusion statistics parameters @xmath147 are defined as @xmath148 where @xmath149 is the change in the number of states , @xmath150 , available to a particle of species @xmath151 when the number , @xmath152 , of particles of species @xmath153 is changed by @xmath154 ( keeping the size of the system unchanged ) . by exact diagonalization of small systems on the sphere and by using the composite fermion construction , the following values have been established @xmath155 for the jain series @xmath156@xcite . ( except that ref . claims that @xmath157 . )
( the diagonal parameters were determined for any level of hierarchy states in ref . . ) since we identified in sec .
[ low_section ] the low energy states in the tt - limit as particle and hole excitations it is straightforward to determine the exclusion statistics in this limit for any @xmath0 , see [ app : exclusion ] .
this gives @xmath158 , @xmath159 , @xmath160 and @xmath161 , where the lengths @xmath162 of the quasiparticles are given by ( [ ln ] ) .
this differs from the established values , in particular we note that the tt - limit gives @xmath163 rather than @xmath164 , which , at @xmath106 , is obtained by assuming the quasiparticles are anyons and counting the number of anyon wave functions @xcite . at 1/3
the tt - limit gives @xmath165 , @xmath166 as compared to @xmath165 , @xmath167 , whereas at 2/5 the values are @xmath168 , @xmath169 and @xmath168 , @xmath170 respectively .
we see that the values for the quasiholes agree , whereas those for the quasielectrons are shifted by one .
this is just a rephrasing in terms of exclusion statistics of the results we found in sec .
[ numerics ] . in the figures above
, the statistics parameters in the tt - limit can be read off from the red states , which are the one and two particle / hole states in this limit , see [ app : exclusion ] .
the exclusion statistics is unchanged as @xmath131 increases if and only if the red states remain the low energy states .
we see that this is the case for two holes at 1/3 and 2/5 , _ cf _ fig .
[ fig:5_15_h_hh ] , [ fig:4_10_h_hh ] , whereas red states disappear from the low energy spectrum as @xmath131 increases for two particles , _ cf _ fig .
[ fig:5_15_p_pp ] , [ fig:4_10_p_pp ] .
we conclude that the exclusion statistics may change when @xmath131 increases from the tt - limit to the physical regime .
the non - diagonal exclusion parameters , @xmath171 , can be inferred from fig .
[ fig:4_10_ph_5_15_ph ] , where the red lines give these parameters . when @xmath131 increases these states mix with others , indicating that mutual statistics of particles and holes changes .
however , there is no obvious separation of a low energy sector at @xmath172 and a value for the off - diagonal components can not be extracted from these small systems .
let the ground state contain @xmath364 unit cells @xmath6 : @xmath365 .
let @xmath366 , @xmath367 denote the number of @xmath8 , @xmath9 in a general state ; for the ground state @xmath368 .
a quasielectron @xmath8 ( quasihole @xmath9 ) can then be inserted once per quasihole @xmath9 ( quasielectron @xmath8 ) to produce different quasielectron ( quasihole ) states .
thus the number of one - quasielectron states is equal to the number of quasiholes @xmath367 present , @xmath369 , and the number of one - quasihole states is equal to the number of quasielectrons @xmath366 present , @xmath370 . to determine how this changes in the presence of quasiparticles , we insert @xmath4 quasielectrons @xmath8 in the ground state and remove @xmath371 unit cells @xmath6 in order for the size of the system to remain unchanged .
there are now @xmath372 quasielectrons @xmath8 and @xmath373 quasiholes @xmath9 present in the state .
a quasielectron @xmath8 can still be inserted once per quasihole @xmath9 to produce different one - quasielectron states .
thus there are now @xmath373 one - quasielectron states and @xmath374 one - quasihole states .
this gives @xmath375 and @xmath376 because we actually inserted @xmath4 quasielectrons @xmath8 .
inserting instead @xmath4 quasiholes in the ground state leads to @xmath160 and @xmath161 .
the lenghts of the quasiparticles are given by ( [ ln ] ) .
we note that while the diagonal parameters are positive as they should be in order for the number of states of a particular species not to increase with the addition of particles of this species , the off - diagonal parameters are always negative . in the tt - limit
we have that @xmath163 , @xmath377 and @xmath378 , which differs from the results @xmath164 and @xmath379 , that hold away from this limit @xcite .
this discrepancy is further discussed in sec .
[ numerics ] . when calculating the statistics parameters above the size of the system was maintained .
however , when comparing to exact diagonalisation of small systems this constraint needs to be relaxed .
we do this by first considering how the corresponding constraint can be relaxed when calculating the fractional charge of the quasiparticles .
normally , one adds several quasiparticles and removes an integral number of unit cells so that the size of the system is unchanged .
we can however reformulate the calculation and insert just one @xmath151 particle .
this particle contains @xmath380 electrons and extends over @xmath381 sites . to obtain the charge of the quasiparticle
we must subtract the charge of the background , which has charge @xmath3 per unit cell of length @xmath4 .
thus we find that the charge of the @xmath151 quasiparticle is @xmath382 , where we have used ( [ eq : main - equation ] ) wich @xmath383 .
a similar reasoning can be applied to the calculation of the exclusion statistics .
consider the tt - ground state with @xmath364 unit cells at filling factor @xmath0 .
we now insert one @xmath151 particle ( @xmath384 or @xmath385 ) .
let @xmath386 be the number of low - energy states thus obtained .
( if there is one such state per unit cell , which is typically the case , then these @xmath386 states are all degenerate . )
we now , in another tt - ground state of @xmath364 unit cells , insert instead a @xmath153-particle .
( @xmath151 and @xmath153 could be the same kind of particle . ) here we will get @xmath152 low energy states , just like for @xmath151 above .
we fix the position of the first inserted @xmath153-particle and insert an @xmath151-particle .
let @xmath387 be the number of low - energy states thus obtained .
( these low energy states will not be degenerate since the first insertion breaks the translational invariance . )
the difference @xmath388 in the number of low energy states before and after the insertion of the @xmath153 particle tells us how much the state space of @xmath151 changes by the insertion of a @xmath153 .
we obtain the statistics parameter @xmath147 from @xmath389 by compensating for the change in size of the system the insertion of the @xmath153-particle gave rise to . we know that a unit cell contains exactly one particle and one hole , therefore the number of particles and holes per site is @xmath390 .
we added a @xmath153-particle containing @xmath391 sites and by that very reason the number of low energy states for the particle and hole increased by @xmath392 respectively . deducting this modification from the number of low - energy states we find the exclusion statistics @xmath393 .
applying the above procedure to the tt - limit gives just another way of deriving the @xmath147 given above .
here we apply it to obtain the exclusion statistics for small systems analyzed using exact diagonalization .
we start with @xmath1 .
[ fig:5_15_h_hh ] shows the insertion of one ( 5/16 ) and two ( 5/17 ) holes at @xmath1 with @xmath394 unit cells .
the tt - ground state at @xmath395 is @xmath396 . adding one hole one obtains 5/16 , where the lowest energy state in fig .
[ fig:5_15_h_hh ] is @xmath397 in the tt - limit .
there are obviously 16 such states related by translation , but choosing one particular ground state initially , the additional @xmath9 can be inserted in five different places ; thus , @xmath398 . at 5/17 , the three lowest energy states in fig .
[ fig:5_15_h_hh ] are found to be @xmath399 , @xmath400 and @xmath401 ( each of them 17-fold degenerate ) in the tt - limit , in order of increasing energy . to determine @xmath402
we must keep the first hole fixed and count the number of states for the second hole @xmath403the three low - energy states ( and translations of them ) then give the states : @xmath404 , @xmath405 , @xmath406 , @xmath407 , and @xmath408 ; thus @xmath409 . ( @xmath9 and @xmath410 are of course identical so the order of them does not matter . ) of the five states , there are two pairs that are related by translation and hence degenerate .
since the numbers of low - energy quasihole states , @xmath367 and @xmath402 , are equal , we conclude that @xmath411 .
the tt - limit is particle - hole symmetric , thus the analysis for insertion of particles , _ cf . _
[ fig:5_15_p_pp ] , is obtained by simply interchanging @xmath9 and @xmath8 , consequently @xmath412 . since @xmath413 and @xmath414 we get @xmath415 and @xmath416 .
[ fig:4_10_h_hh ] shows the insertion of one ( 5/13 ) and two ( 6/16 ) holes at @xmath65 with @xmath417 unit cells .
the tt - ground state at @xmath418 is @xmath419 .
adding one hole one obtains 5/13 , where the lowest state is @xmath420 .
there are obviously 13 such states related by translation , but choosing one particular ground state initially , the additional @xmath9 can be inserted in two different places ; thus , @xmath421 . at @xmath422 ,
the two lowest energy states are found to be @xmath423 and @xmath424 , in the tt - limit , in order of increasing energy .
the first of these is 8-fold degenerate whereas the second is 16-fold .
( the additional degeneracy disappears when @xmath131 increases and linear combinations of states with different positions of the holes are formed . ) to determine @xmath402 we keep the first hole fixed and count the number of states for the second hole @xmath403these states are : @xmath425 and @xmath426 ; thus @xmath427 .
since the numbers of low - energy hole states , @xmath367 and @xmath402 , are equal , we conclude that @xmath411 .
the tt - limit is particle - hole symmetric , thus the analysis for insertion of particles , _ cf . _
[ fig:4_10_p_pp ] , is obtained by simply interchanging @xmath9 and @xmath8 , consequently @xmath412 . since @xmath413 and @xmath428 we get @xmath429 and @xmath430 .
the off - diagonal elements of the statistics parameter are obtained from fig .
[ fig:4_10_ph_5_15_ph ] , which shows the spectra for @xmath1 with six unit cells and @xmath65 with three cells . at 6/18
the lowest energy states are identified as @xmath431 , @xmath432 , @xmath433 , @xmath434 , @xmath435 and @xmath436 in the tt - limit ( in order of increasing energy ) .
( four of these states are pairwise degenerate . )
the first of these is of course the 1/3 ground state , whereas the others are particle - hole excitations thereof .
all these states can be obtained by inserting a hole and a particle in the 5/15 ground state @xmath396 .
adding one hole to @xmath396 one obtains the five low energy states @xmath437 according to the analysis above , thus @xmath398 .
adding a particle , keeping the added hole fixed , then gives precisely the six low - energy states found at 6/18 .
thus there are @xmath438 particle states .
thus , when we move from @xmath439 to @xmath440 the number of states increases by one from 5 to 6 , @xmath441 , ( the process @xmath442 to @xmath443 of course gives the same counting ) .
this gives the off - diagonal elements @xmath444 and @xmath445 .
for @xmath446 at 6/15 , _ cf _ fig .
[ fig:4_10_ph_5_15_ph ] , we identify the low - energy states as @xmath447 , @xmath448 , and @xmath449 in the tt - limit ( the last two being degenerate ) . by a similar analysis as for 1/3
this gives again @xmath450 , and hence @xmath451 and @xmath452 .
earlier studies have shown that the ground state in the tao - thouless limit , at any filling fraction @xmath14 , is adiabatically connected to an abelian quantum hall state in the two - dimensional bulk system and the fractionally charged quasiparticles , @xmath173 , are domain walls in the tt - limit . in this article
we have studied the full low - energy sector of states in the tt - limit .
using the fact that the unit cell , @xmath6 , that gives the ground state is @xmath174 we found that , assuming a certain condition for the interaction ( @xmath115 ) , the low - energy sector consists of the reorderings of the @xmath8 and @xmath9 in the string of @xmath6 s that make up the ground state ; these reorderings correspond to ( multiple ) particle - hole excitations . from the low - energy sector
we determined the exclusion statistics in the tt - limit .
we studied numerically , at and around @xmath1 and 2/5 , how the low - energy states develop as one moves away from the tt - limit .
we found that the low energy states in the physical regime always develop from low - energy tt - states .
this is a non - trivial result which supports the tt - approach .
however , there is not a perfect match between the lowest energy states in the tt - limit and those in the physical regime and this lack of matching is reflected in different exclusion statistics in the two regimes .
the states that give the tt - exclusion statistics at @xmath1 and 2/5 are the states marked in red in the figures above .
consider for example two holes and two particles at @xmath1 , fig .
[ fig:5_15_h_hh ] and [ fig:5_15_p_pp ] .
the two - hole states are low - energy states for all @xmath131 , whereas one of the two - particle states disappears to higher energies as @xmath131 increases . however , although the red states are the lowest energy states in the tt - limit , @xmath119 , the gap to higher energy states is not big , and is certainly not diverging as @xmath119 .
this is because the coulomb interaction obeys only @xmath175 , with a small difference between successive terms . from just considering the energies in the tt - limit
, one may question if the red states are really the low - energy states .
if instead a screened coulomb interaction is used , see fig .
[ fig:5_17_yuk ] , then the red states are clearly separated as the low - energy states .
this indicates that the precise identification of the very lowest energy states in the tt - limit depends on the details of the interaction .
note that there is no unique way of taking the limit of , say , the coulomb interaction as @xmath131 decreases ; since this is an unphysical process one may allow the interaction to change .
the tt - ground state is insensitive to details of the interaction ( as long as @xmath176 ) ; it is not surprising that this is not the case for the gaps between the low - energy excitations .
it is conceivable that one could modify the interaction , for small @xmath131 , so that the exclusion statistics in the tt - limit agrees with the one in the physical regime .
however , we find this unnatural since the counting of the low - energy states would then no longer be the counting of the domain walls in the tt - limit .
thus we conclude that the exclusion statistics does change as @xmath131 increases .
still , one should keep in mind that there is a large correspondence between the lowest energy states for large and small @xmath131 ; all low energy states in the physical regime emanate from low energy states in the tt - limit .
this supports the validity of the tt - limit for the abelian quantum hall states .
quasiholes and two ( 7,19 ) quasiparticles at @xmath1 .
the low - energy states are clearly separated in the tt - limit which is not the case for the unscreened coulomb interaction in fig .
[ fig:5_15_h_hh ] and [ fig:5_15_p_pp ] .
[ fig:5_17_yuk],title="fig : " ] quasiholes and two ( 7,19 ) quasiparticles at @xmath1 .
the low - energy states are clearly separated in the tt - limit which is not the case for the unscreened coulomb interaction in fig .
[ fig:5_15_h_hh ] and [ fig:5_15_p_pp ] .
[ fig:5_17_yuk],title="fig : " ]
the relaxation scheme states that the positions of the electrons on the sites are given by @xmath177 \ , \end{aligned}\ ] ] where @xmath178 , and @xmath36 $ ] is the rounding function @xmath37=n$ ] if @xmath38 .
we immediately see that @xmath179 thus we need only consider @xmath180 and @xmath181 .
+ * definition : * the ordered set @xmath182 is the unit cell ; the transposed set is defined as @xmath183 .
the sets @xmath184 , for different @xmath39 , differ by a translation , @xmath185 , at most .
choosing different @xmath39 , exactly @xmath4 different unit cells are obtained ; these are related by rigid translations and correspond to the @xmath4-fold degeneracy on the torus . the integers @xmath186 change in discrete steps when @xmath39 changes .
specifically , the change to the set @xmath184 under @xmath187 is that one @xmath33 changes by one .
this hopping of one specific electron one site is thus equivalent to a translation of the unit cell .
transposition of a set , @xmath188 , is equivalent to a translation .
one way to see this is to consider the transposition of @xmath189 , see below .
for this set , transposition is equivalent to hopping a specific electron which is equivalent to a translation . noticing that any transposition can be divided into a translation to @xmath79 , a transposition and a translation back again leads to the conclusion that all transpositions are translations .
since all sets @xmath184 are related by rigid translations , it follows that when @xmath39 changes by @xmath190 there exists a translation @xmath191 and a corresponding relabeling @xmath192 that has the same effect : @xmath193 or , equivalently , @xmath194=\left[\frac{j+n_a}{p}q+\delta - l_a\right ] \ . \end{aligned}\ ] ] this relation is trivially true when @xmath195 we see that if @xmath196 and @xmath197 satisfy then so do @xmath198 and @xmath199 .
this gives the general solution to .
since all the variables in are integers the equation is a modulo equation of type @xmath200 we solve these equations explicitly using euler s theorem : if @xmath201 , then @xmath202 , where @xmath203 is the number of integers smaller than or equal to @xmath3 that are relatively prime to @xmath3 .
this gives @xmath204 from and it is trivial to see that @xmath205 and @xmath206 . thus moving one particular electron one step ( @xmath207 ) corresponds to translating the unit cell @xmath208 steps .
this defines a partition of the unit cell where one part consists of @xmath209 sites with @xmath210 electrons and the other part consists of @xmath211 sites with @xmath212 electrons , @xmath213 .
these parts have charges @xmath214 relative to the ground state .
we denote the partitions with charges @xmath215 and @xmath216 by @xmath8 and @xmath9 respectively .
these are the charges of the elementary quasiparticles and since the parts obey the relaxation condition it follows that they are the quasiparticles . moving the electron one step interchanges @xmath8 and @xmath9 in the unit cell .
we verify that the partition corresponding to the quasiparticle obeys the relaxation condition by checking that @xmath217 given by for @xmath218 are unchanged if @xmath219 are replaced by @xmath220 .
consider , to be definite , the relaxation procedure @xmath221_\mathrm{f}$ ] , where @xmath36_{\mathrm f}$ ] is rounding downwards ( @xmath222_{\mathrm f}=\left[x-\frac{1}{2}\right]$ ] ) and @xmath223 .
this corresponds to the @xmath224 representation introduced below .
the partitions are then given by @xmath225_\mathrm{f}$ ] , where @xmath226 .
if @xmath227 for all @xmath226 then the relaxation procedure works for the partitions as well . we have checked this condition for all @xmath3 , @xmath4 with @xmath228 and @xmath229 but lack a general proof . , @xmath230 , and ( b ) @xmath224 , @xmath231 at @xmath232 .
blue @xmath53 are positions of the sites .
black @xmath233 are unrelaxed electrons at @xmath67 .
red @xmath234 and @xmath235 are relaxed and unrelaxed electrons for @xmath79 and @xmath224 respectively .
[ fig : relax ] ] we now consider @xmath66 as an explicit example .
this illustrates the results in the lemmas below , which follow by simple generalizations of the example .
[ fig : relax ] shows two constructions of unit cells by relaxation for different @xmath39 . in ( @xmath236 ) , @xmath237 and the unit cell is @xmath85 ; whereas in ( @xmath238 ) , @xmath239 and @xmath240 .
( @xmath39 is defined modulo @xmath241 because of the periodic boundary conditions . )
@xmath79 ( @xmath224 ) has the property that the electrons are as far to the right ( left ) as possible .
the particle and hole at 5/13 are , as shown above , @xmath242 and @xmath243 respectively ( in the @xmath70-representation ) .
we see that @xmath79 ( @xmath224 ) starts with a 0 ( 1 ) and ends with a 1 ( 0 ) while the interior is symmetric and the same for @xmath79 and @xmath224 : @xmath244
. consequently , @xmath79 and @xmath224 are mirror images of each other @xmath245 . here , ( [ mirror ] ) and ( [ hopping ] ) follow directly from the definition ; ( [ hopping ] ) implies ( [ third ] ) because the electron that hops when @xmath256 is the electron at the edge ; ( [ third ] ) then implies ( [ symmetry ] ) because of ( [ mirror ] ) .
* definition : * let @xmath257 be the set of all unit cells ( including all translations for all @xmath0 ) and let @xmath258 , where @xmath259 , be the subsets of left and right unit cells , @xmath260 , @xmath261 .
the interpretation of the above equations is as follows .
if one considers the @xmath4 unit cells of @xmath271 and takes the mean density of the @xmath196 leftmost ( or rightmost ) sites one gets the mean density of @xmath0 , see the example in table [ tab : cells-2_5 ] .
denote this segment of @xmath6 by @xmath272 and its complement by @xmath70 so that @xmath273 .
we see that in @xmath274 cases the density of @xmath272 is @xmath275 , whereas in the @xmath276 remaining @xmath272 the density is @xmath277 . for @xmath278 ( @xmath279 )
@xmath272 has the correct size and density to be the quasiparticle @xmath47 ( quasihole @xmath46 ) discussed earlier . because of ( [ eq : main - equation ] ) , @xmath70 has to be the quasihole @xmath46 ( quasiparticle @xmath47 ) . the unit cell where @xmath280 ( @xmath281 ) has to be @xmath79 ( @xmath224 ) because in that cell the electrons are as far to the right ( left ) as possible .
property [ enu : transposition ] above has as a consequence that all of the expansions ( [ eq : repetetive_terms ] - [ eq : re_indexed_terms ] ) have transposed variants .
the same holds true for the lemmas in the following section .
_ proof : _ since @xmath289 is a unit cell , @xmath290 is also a unit cell , @xmath291 . for any unit cell @xmath292 , @xmath255 ( property [ symmetry ] above ) , then @xmath293 which is ( [ eq :
trans : hp^m = hp^m ] ) . equation ( [ eq : trans : h^mp = h^mp ] ) follows analogously .
assume a representation @xmath184 with the corresponding particle - hole decomposition @xmath303 exists .
assume that a state , @xmath304 , can be expressed as a sequence of only @xmath47 and @xmath46 , @xmath305 , where @xmath306 .
then any permutation @xmath307 , where @xmath306 , differs from @xmath304 in energy with terms of order @xmath4 or higher , _ ie _ @xmath308 any other state @xmath70 , _ ie _ a state that can not be obtained by reordering the @xmath8 and @xmath9 in the ground state has energy @xmath309 _ proof : _ we start by proving , by induction , the first part of the theorem . any permutation @xmath310 can be written as a series of ( pairwise ) transpositions @xmath311 . if all transpositions fulfill the requirement above then so does any permutation . we prove this using the @xmath251-representation , but the result is valid for all representations since they are equivalent .
( equivalent in the sense that the same sequence of @xmath8 and @xmath9 , in different representations , gives ( under periodic boundary conditions ) the same sequence of 0 and 1 , upp to a translation of @xmath312 sites ) . in this representation @xmath313
is affected using @xmath314 as is seen from ( the transpose of ) .
we show only the proof for @xmath320 ; the case @xmath321 is analogous and @xmath322 is trivial . here
@xmath323 and @xmath324 are shorthands for @xmath325 and @xmath326 respectively .
also @xmath327 and @xmath328 , which are the minimal expansions of @xmath329 and @xmath324 , do appear .
consider the first case : @xmath330 where @xmath331 according to ( [ eq : trans : hp^m = hp^m ] ) .
@xmath301 denotes an arbitrary string of @xmath332 s in the @xmath70-representation it necessarily ends with a 1 ; this string is included since it gives the leading contribution to the energy .
@xmath333 denotes a string that is @xmath56 sites long such that @xmath334 is symmetric around @xmath335 .
because of this reflection symmetry , @xmath336 and @xmath337 have the same energy .
it then follows that the leading contribution to the energy is given by the interaction with the next electron on both sides and is @xmath338 , thus @xmath339 . the third case , using ( [ eq : trans : hp^m = hp^m ] ) , @xmath345 here the nearest pair of electrons that contributes is more than @xmath342 sites to the left and more than @xmath4 sites to the right , thus @xmath346 .
this finishes the proof of the first part of the theorem .
we now turn to the second part which states that all states that can not be obtained by reordering the @xmath8 and @xmath9 that make up the ground state differ in energy from the ground state by @xmath349 .
assume that we have a state @xmath350 that can be described by @xmath351 .
we say that @xmath351 covers @xmath350 if and only if we can describe @xmath350 using @xmath351 .
@xmath350 is our reference state ; it could be the ground state .
let @xmath304 be a state that is not covered by @xmath351 .
assuming @xmath352 , we expand @xmath353 . note that @xmath354 trivially cover @xmath350 since @xmath351 do .
if @xmath354 cover @xmath304 then the sequence @xmath355 must occur somewhere in @xmath304 otherwise @xmath356 would also cover @xmath304 .
this means that when @xmath357 there has been a move of the type @xmath358 where the last step costs @xmath359 as seen in the proof the first part of the theorem .
since @xmath360 , this proves the theorem .
if @xmath354 do not cover @xmath304 , then expand the longer of @xmath361 and @xmath8 and repeat the argument .
eventually a particle and a hole is found that covers @xmath304 and the theorem follows . | we consider spin - polarized abelian quantum hall states in the tao - thouless limit , _ ie _ on a thin torus . for any filling factor @xmath0 a well - defined sector of low - energy states
is identified and the exclusion statistics of the excitations is determined .
we study numerically , at and near @xmath1 and @xmath2 , how the low energy states develop as one moves away from the tt - limit towards the physical regime .
we find that the lowest energy states in the physical regime develop from states in the low energy sector but that the exclusion statistics is modified . |
doubly - magic nuclei are key benchmarks to our understanding of nuclear shell structure . until experiments around @xmath2sn become feasible , @xmath1ni is the heaviest experimentally accessible doubly - magic nucleus in which protons and neutrons occupy the same orbitals .
an early neutron - pickup experiment populating states in @xmath0ni suggested that the low - lying excited states may not be purely single - particle in nature , but may include contributions from @xmath1ni excited states @xcite .
the @xmath3-decay of @xmath0cu indicated @xcite that the gamov - teller matrix element of its decay to the ground state of @xmath0ni is significantly reduced compared to the single - particle value .
measurements of unusually high collectivity of @xmath1ni @xcite reinforced the questions about the single - particle character of the ground and lowest excited states of @xmath0ni .
the spectroscopic factors for the first three states of @xmath0ni were measured in a @xmath4 transfer reaction at low energy @xcite .
in contrast to the indications of collectivity from the measured @xmath1ni @xmath5 excitation strength , the transfer experiment indicated that the first three states of @xmath0ni are almost pure single - particle states .
pickup and stripping reactions and extraction of spectroscopic factors can be used to resolve the nature of @xmath1ni and the single - particle character of @xmath0cu and @xmath0ni .
the accuracy of spectroscopic factors measured in this mass region via low - energy transfer reactions is limited by the lack of good optical model potential parameters . in experiments performed at high energies , analyses based on the sudden approximation and eikonal theory @xcite
are applicable and thus the model dependency is reduced .
the corresponding experimental technique is one - nucleon knockout in inverse kinematics , a method which has been used to measure single - particle configurations over a range of nuclei since its development as a spectroscopic tool @xcite . in this method ,
single - particle spectroscopy on the nucleus of interest is performed by extracting the following quantities : partial cross sections to ground and excited states of the knockout residues ; the spectroscopic factors for the removal of a nucleon from a specific single - particle orbit of the projectile ; and the orbital angular momentum of the removed nucleon .
measurements are performed in inverse kinematics at beam energies of greater than 50 mev / nucleon . in previous one - nucleon knockout experiments on well - bound nuclei , a reduction in measured spectroscopic strength with respect to shell - model predictions has been observed . the resulting reduction factor @xmath6 ,
the ratio of experimental and theoretical spectroscopic strength , is 0.5 - 0.7 for one - proton and one - neutron knockout from stable @xmath7c and @xmath8o @xcite and for one - neutron removal from well - bound @xmath9 isotones @xcite , but closer to unity for one - proton knockout from the lighter weakly proton - bound nuclei @xmath10b , @xmath11c @xcite and for the one - neutron knockout from the weakly neutron - bound @xmath12c to the ground state of @xmath13c @xcite .
recently , a pronounced reduction of 0.24(4 ) has been observed in the one - neutron removal from @xmath14ar @xcite , the most deeply - bound neutron system studied so far .
in the technique of one - nucleon knockout , the measured cross sections are used to derive the one - nucleon removal spectroscopic factors via an extension of the eikonal model @xcite . the cross section @xmath15 to a specific final state of the knockout residue ( core ) is related to the spectroscopic factor via @xmath16 where the sum is taken over all non - vanishing nucleon - core configurations .
the quantity @xmath17 is the spectroscopic factor for the removal of a nucleon with given single - particle quantum numbers @xmath18 .
the factor @xmath19 is a center - of - mass correction that has to be applied if spectroscopic factors are taken from shell model to obtain a theoretical cross section . in the center - of - mass correction , @xmath20 is the mass number of the initial nucleus and @xmath21 is the main harmonic oscillator quantum number associated with the relevant major shell @xcite .
the spectroscopic factor depends on the structure of the nucleus , while the single - particle cross section @xmath22 comes from reaction theory , assuming unit single - particle strength .
the single - particle cross sections are highly dependent on the nucleon separation energy @xmath23 and the orbital angular momentum @xmath24 of the removed nucleon .
these were calculated in the eikonal model and include contributions from both the stripping and diffraction processes @xcite .
the removed - neutron wave functions in the @xmath0ni ground state were calculated in woods - saxon potential wells with a diffuseness 0.7 fm .
their radius parameters were chosen so that the _ rms _ radius of each neutron orbital was consistent with those from a skyrme skx hartree - fock ( hf ) @xcite calculation for @xmath0ni when the depth of the woods - saxon potential was adjusted to reproduce the effective separation energy .
the measured separation energy of @xmath25 mev @xcite was used for the @xmath26 state and an effective separation energy of 14 mev was estimated from the centroid of the @xmath27 multiplet for the @xmath28 level .
the orbital _ rms _ radii were assumed to be 4.14 and 4.17 fm for the @xmath26 and @xmath28 levels .
the @xmath1ni neutron and proton densities were also taken from hf , resulting in a _
matter radius of 3.64 fm for @xmath1ni , consistent with @xmath1fe .
a gaussian matter distribution was assumed for the @xmath29be target with an _ rms _ matter radius of 2.36 fm .
excited levels of @xmath1ni and associated spectroscopic factors were calculated in the many - body shell model @xcite using the fpd6 effective interaction @xcite . for @xmath0ni ,
the @xmath30 and @xmath31 closed shells plus one neutron in the @xmath26 orbit were assumed . for @xmath1ni , calculations were performed with the @xmath30 and @xmath31 closed shells for the ground state , and for the excited states @xmath32 and @xmath33 with the @xmath26 , @xmath34 , and @xmath35 neutron shells active . in this approach , the spectroscopic factor for the removal of a neutron from the @xmath26 orbit leading to the ground state of @xmath1ni is @xmath36 and the spectroscopic factors for removing a neutron from @xmath28 summed over all final excited states of @xmath1ni ( @xmath37 and @xmath38 ) is @xmath39 . considering the calculated energy levels of @xmath1ni
, we predict 7.36 of this to be below the proton separation energy .
the shape of the measured parallel momentum distribution of the @xmath1ni residues is used to assign the orbital angular momentum for the removed nucleon .
the theoretical parallel momentum distributions were calculated in a black - disk model @xcite , core - target collisions at small impact parameters ( @xmath407.2 fm ) being assumed to result in fragmentation of the core .
interaction radii were chosen to reproduce the reaction cross sections of the free constituents .
the experiment was performed at the national superconducting cyclotron laboratory ( nscl ) .
acceleration of a beam of stable @xmath41ni to an energy of 105 mev / nucleon and an intensity of 2 pna was carried out in the k1200 cyclotron of the nscl .
the @xmath41ni beam was incident on a 249 mg/@xmath42 @xmath11be production target .
the a1200 fragment separator @xcite was used to select a secondary beam of @xmath0ni with an energy of 73 mev / nucleon , an intensity of approximately 30,000 particles per second and a momentum spread of 0.5% .
a 56.1 mg/@xmath42 @xmath11be target was located at the center of an array of 38 position - sensitive nai(tl ) detectors @xcite at the entrance to the s800 spectrograph @xcite .
gamma rays were measured in coincidence with the knockout residues detected by the s800 focal - plane detector system @xcite .
the identification of the @xmath1ni fragments was performed with energy loss and position information measured with the s800 spectrograph focal - plane detectors and using the time of flight taken between plastic scintillators at the exit of the a1200 fragment separator and the s800 focal plane ( see fig.[pid ] ) . ni beam . @xmath1ni and @xmath0ni isotopes can be clearly separated .
the x - position relates to the parallel momentum .
[ pid],width=302 ] ) projectile - frame @xmath43-ray spectrum in coincidence with @xmath1ni fragments detected in the spectrograph focal plane .
the solid black line is the fit to the experimental spectrum containing the sum of the simulated response functions for seven @xmath43-ray peaks ( solid grey lines ) and a double - exponential coincident background ( dashed line ) , assumed to arise from excitations of the target .
the inset shows a partial decay scheme for @xmath1ni with a new proposed level at 5661(72 ) kev .
[ gamspec],width=324 ] after the @xmath1ni residues from one - neutron knockout were unambiguously identified , the coincident @xmath43-ray spectrum was analyzed .
only the inner ring of 11 nai(tl ) detectors @xcite was used in this experiment .
geant @xcite simulations successfully modeled the @xmath43-ray spectrum and the detector response .
simulations and experimental efficiencies measured with standard calibration sources agreed within 7.5% .
the doppler - reconstructed ( @xmath44 ) @xmath43-ray spectrum in coincidence with @xmath1ni fragments is shown in fig .
[ gamspec ] .
the analytical curves , which were fitted to the simulated spectra and subsequently scaled to fit the experimental spectrum , can be seen as solid grey lines .
the dashed line indicates the prompt , coincident background , described in this experiment by a double exponential curve .
a similar background has been observed in previous knockout experiments @xcite where it has been attributed to neutrons , @xmath43 rays and charged particles interacting with the experimental apparatus and scintillators .
the background determined for this experiment was 67% to 85% higher than that quoted for knockout on si and s @xcite .
there may be evidence for an increase in coincident background with mass number or binding energy of the knocked - out nucleons .
a simplified level scheme including all @xmath43-ray transitions observed in this experiment is shown in the inset of fig .
[ gamspec ] .
the placement of the 1726(10 ) and 3027(71 ) kev @xmath43 rays in coincidence with the 1224.5(7 ) and 2700.6(3 ) kev @xmath43 rays resulted in a level at 5661(72 ) kev .
a 5668 kev level and associated 1744 kev de - excitation @xmath43 ray were observed in 1985 @xcite and reported without uncertainty . while an angular momentum assignment of was proposed for the 5668 kev level , only are expected to be populated via the one - neutron knockout mechanism due to angular momentum considerations .
no placement in the proposed level scheme was possible for the @xmath43 ray observed at 1379(10 ) kev .
the proton separation energy for @xmath1ni is significantly higher than the energy of the highest excited state observed in this experiment .
while excited states up to an energy of 7900 kev were predicted in a shell - model calculation , no higher - lying excited states could be identified experimentally above 5661 kev .
low statistics at high @xmath43-ray energies contributed to the difficulty in resolving any transitions above 3100 kev .
therefore the branching ratios to the individual excited states of @xmath1ni were not calculated due to the possibility of indirect feeding from higher - lying , unobserved excited states .
ni detected at the s800 focal plane .
the distribution associated with @xmath1ni fragments in coincidence with all @xmath43 rays between 250 and 7000 kev is shown in panel ( a ) , and that associated with all remaining @xmath1ni fragments in panel ( b ) . the distribution associated with knockout to all excited states of @xmath1ni
is shown in panel ( d ) and that to the ground state in panel ( c ) .
the distributions in panels ( c ) and ( d ) were constructed assuming an average efficiency for the nai(tl ) array of and a background probability of ( see text for details ) .
[ bothmom],width=324 ] the parallel momentum distributions of the @xmath1ni knockout residues were determined using the reconstructed scattering angle and fractional kinetic energy after the target , the momentum ( 20.26 gev/@xmath45 ) and beam velocity @xmath44 at the point of the @xmath43-ray emission , and the mass of the fragments .
due to the complexity of the @xmath43-ray spectrum of @xmath1ni , it was not possible to isolate momentum distributions associated with individual final excited states . instead , the momentum distributions associated with the knockout to all excited states and the ground state of @xmath1ni were reconstructed from the measured momenta in coincidence and anti - coincidence with detected @xmath43 rays .
the coincident ( @xmath46 ) and anti - coincident ( @xmath20 ) spectra shown in fig .
[ bothmom](a ) and ( b ) , respectively , can be written in terms of the momentum distributions to the ground state ( @xmath47 ) and all excited states ( @xmath48 ) of @xmath1ni as @xmath49 the average efficiency @xmath50 for the nai(tl ) array is expected to be approximately 50% for an average @xmath43-ray cascade of two , but was varied in the analysis to determine the best separation between @xmath47 and @xmath48 .
the probability that a @xmath43 ray not originating from the de - excitation of a knockout residue would be detected in coincidence with that fragment in the s800 focal plane is @xmath51 . for this experiment ,
a value of was chosen , in agreement with previous experiments which determined a @xmath43-ray background of 10% per fragment .
the background appears to be higher in this experiment than the 10% assumed in experiments on lighter nuclei .
however , with this choice of @xmath51 , a consistent separation of the ground state and excited state events was obtained with a choice of . thus the higher background used in the fit to the @xmath43-ray spectrum in this experiment may be a result of multiple @xmath43 rays that could not be resolved due to the level density and low statistics at high energies .
a 3.5% systematic error on the number of counts was included in the calculation of experimental cross sections due to the choice of the two parameters @xmath50 and @xmath51 .
the resulting momentum distributions associated with the ground state and all excited states of @xmath1ni are shown in fig .
[ bothmom](c ) and ( d ) .
the errors shown include both , statistical uncertainties and the systematic error from the choice of @xmath50 and @xmath51 .
a similar analysis has been performed for the one - nucleon removal reactions on @xmath52 shell nuclei @xcite .
the shell model predicts that one - neutron removal to the ground state would be associated with an distribution , due to the removal of the @xmath53 valence neutron .
similarly , one - neutron removal to the excited states of @xmath1ni would correspond to an momentum distribution due to the orbital angular momentum carried by a neutron removed from the @xmath54 shell .
the ground state distribution shown in fig .
[ bothmom](c ) is well described by an theoretical curve centered at 20.42 gev/@xmath45 .
the distribution associated with knockout to the excited states of the @xmath1ni fragments , shown in fig .
[ bothmom](d ) is consistent with the theoretical curve for with a low - momentum excess extending to approximately 20 gev/@xmath45 .
this low - momentum tail as seen around 20.25 gev/@xmath45 in the excited - state momentum distribution ( d ) , has been reported in several previous one - nucleon knockout experiments @xcite .
tostevin @xcite has shown that the asymmetric shape as observed for the knockout reactions on the halo nuclei @xmath55be and @xmath12c can be reproduced by using a proper dynamical treatment of continuum - coupling effects in the diffractive channel of the knockout process . however , the same explanation is unlikely to hold for the well - bound @xmath0ni .
the present observation is very similar to the pronounced tail at low momentum reported in the one - neutron knockout from @xmath56ar to the @xmath57 ground state of @xmath58ar @xcite , where this asymmetry is discussed in the framework of deviations from eikonal theory . .
experimental cross sections @xmath59 in mb and spectroscopic factors ( c@xmath60s@xmath61/@xmath62 for the various final states of @xmath1ni populated in the @xmath11be(@xmath0ni,@xmath1ni+@xmath63x reaction at 73 mev / nucleon .
theoretical single - particle cross sections ( @xmath64 were calculated in the eikonal model @xcite for the stripping @xmath65 and diffractive @xmath66 processes .
spectroscopic factors from shell model ( c@xmath60s@xmath67 ) and the theoretical cross section @xmath68 , which combines the spectroscopic factors from the shell model and the single - particle cross section from the reaction theory following eq .
( 2 ) , are compared to the experimental values in terms of the ratio @xmath69/@xmath70 ( see text).[xsec ] [ cols="<,<,<,<",options="header " , ] the inclusive cross section @xmath71 was calculated from the number of @xmath1ni fragments detected in the spectrograph focal plane relative to the number of incident @xmath0ni projectiles , normalized to the number of target nuclei .
the inclusive cross section was calculated run by run , and the average for the three runs closest to the unreacted @xmath0ni normalization run was used to determine @xmath72 mb .
the partial cross sections to all excited states and the ground state of @xmath1ni were determined using the same method as for the parallel momentum distributions .
this resulted in branches of 81.3(35)% to all excited states and 18.7(35)% to the ground state , corresponding to a partial cross section to the ground state of 7.7(15 ) mb and to all excited states of 33.7(17 ) mb .
the calculated stripping and diffraction components for the single - particle cross sections to the ground state and excited states of the fragments are listed in table [ xsec ] , along with the total theoretical cross sections to both final states .
the deduced spectroscopic factors were for the ground state and for all excited states . versus incident beam energy for one - nucleon knockout from the well - bound nuclei @xmath7c , @xmath8o and @xmath0ni .
the circles ( squares ) indicate results from one - neutron ( proton ) knockout .
@xmath73 values for @xmath7c and @xmath8o were calculated in @xcite from measurements performed at the lawrence berkeley national laboratory @xcite .
[ quenchfig],width=324 ] the shell model is expected to most accurately predict the properties of nuclei at or near closed shells . as expected for @xmath0ni , a semi - magic nucleus with only one neutron outside the doubly - magic @xmath74 core
, the shell model correctly predicts the angular momentum of the knocked - out neutron for the ground - state configuration , shown by the good fit of the shape to the momentum distribution associated with the @xmath1ni ground state .
although the orbital angular momentum of knocked - out neutrons populating excited states of @xmath1ni could not be well determined , a value of at least should be assigned .
the measured partial and inclusive cross sections , however , do not show good agreement with theory .
the measured cross section to the ground state is 55(11)% of the theoretical ground - state cross section .
the measured cross section to all excited states amounts to 47(2)% of the shell - model sum - rule theoretical cross section , and the inclusive cross section is 48(2)% of the theoretical inclusive cross section .
the reduction in cross section translates into a reduction in measured spectroscopic strength relative to the shell model .
several studies have been performed recently to explore this reduction in spectroscopic strength observed in one - nucleon removal experiments @xcite .
proton removal from inelastic electron scattering has been considered the standard for the determination of absolute spectroscopic factors .
it was shown that deduced spectroscopic factors from ( @xmath75 ) experiments for nuclei from mass 12 to 208 exhaust only about 60% of the single - particle shell model prediction @xcite .
electron scattering data is only available for stable nuclei , and it is not possible to probe neutron occupancies within this approach .
in contrast , the technique of one - nucleon knockout can be used to probe both proton and neutron occupancies in radioactive nuclei as well as stable species .
it then becomes of interest to compare the spectroscopic factors deduced via one - proton knockout reactions to those from electron scattering .
et al . _
@xcite compared theoretical cross sections calculated in the eikonal model with experimental cross sections from a series of one - nucleon knockout experiments . in order to facilitate a comparison between experiments ,
the reduction factor @xmath73 was defined as the ratio between the experimental and theoretical cross sections .
one - proton knockout on the stable nuclei @xmath7c and @xmath8o resulted in @xmath73 values of 0.53(2 ) and 0.68(4 ) , in agreement with the reduction factors deduced from the electron scattering experiments .
one - neutron knockout from the same two stable nuclei yielded @xmath73 values of 0.49(2 ) and 0.56(3 ) , suggesting that one - neutron knockout experiments measure the same quantity as one - proton knockout , and thus can also be used to measure absolute spectroscopic factors .
the reduction in single - particle strength observed in the case of the radioactive nucleus @xmath0ni is comparable to the reductions obtained for the single - neutron knockout of the two stable nuclei @xmath7c and @xmath8o @xcite ( see fig .
[ quenchfig ] ) , consistent with the reduction observed in the one - neutron knockout on well - bound @xmath9 isotones @xcite and agrees with the quenching of single - particle strengths observed for stable nuclei in electron scattering experiments @xcite .
it is suggested @xcite that correlation effects ( short - range , long - range and tensor nucleon - nucleon interactions ) which are absent from or only approximately treated in effective - interaction theory may explain some of the observed reduction in spectroscopic strength
. those correlations result in physical nucleon occupancies being reduced and spectroscopic strength shifted to higher energies ; see @xcite . for example , the repulsive short - range part of the interaction becomes important at distances of less than 0.4 fm , and leads to high - momentum components in the nucleon wave function @xcite .
the importance of the coupling to surface phonons and giant resonances as contribution to the reduction has been discussed in @xcite . in summary ,
spectroscopic factors for one - neutron removal from @xmath0ni to the ground and all observed excited states of @xmath1ni have been measured at 73 mev / nucleon .
the @xmath76 character of the @xmath0ni ground state and the @xmath77 character of the first few excited states have been confirmed .
the level scheme of @xmath1ni was extended to include a level at 5661(72 ) kev , with two de - excitation @xmath43 rays .
finally , the phenomenon of a reduction in the measured spectroscopic strengths , as compared to shell - model predictions , has been confirmed in the exotic , doubly - magic @xmath1ni the heaviest nucleus yet measured via one - nucleon knockout .
the observed inclusive cross section for knockout to @xmath1ni was 48(2)% of the predicted shell - model value , in agreement with the reductions typically observed in stable , well - bound nuclei .
this work was supported by the national science foundation under grants phy-0110253 and phy-0244453 , and by the united kingdom engineering and physical sciences research council ( epsrc ) under grant ep / d003628 . . | the single - particle structure of @xmath0ni and level structure of @xmath1ni were investigated with the reaction at 73 mev / nucleon . an inclusive cross section of 41.4(12 )
mb was obtained for the reaction , compared to a theoretical prediction of 85.4 mb , hence only 48(2)% of the theoretical cross section is exhausted .
this reduction in the observed spectroscopic strength is consistent with that found for lighter well - bound nuclei .
one - neutron removal spectroscopic factors of 0.58(11 ) to the ground state and 3.7(2 ) to all excited states of @xmath1ni were deduced . |
any cellular automaton ( ca ) with two states and local transition rule can be used for definition of a reversible second - order ca with new rule @xmath1 on a pair @xcite @xmath2 + c ' \mod 2\ , , c ) .
\label{2nd}\ ] ] an inverse rule is @xmath3 + c ' \mod 2 ) .
\label{inv2nd}\ ] ] and also may be rewritten @xmath4 where @xmath5 is exchange of states @xmath6 the ca acting on pairs of binary states @xmath7 can be considered as four - state ca due to simple correspondence @xmath8 .
let us denote @xmath9 state of a cell with notations @xmath10 let us consider a few different two - dimensional ca 1 .
@xmath11 with local rule : @xmath12 2 .
@xmath13 with local rule : @xmath14 : 3 .
@xmath15 with local rule : @xmath16 + @xmath17 with local rule : @xmath18 and second - order reversible ca @xmath19 derived from them using eq .
( [ 2nd ] ) .
if to start with a single cell with value one and all others zero , then total number of cells with nonzero values at @xmath0-th stage is some sequence @xmath20 .
it is also possible to consider sequences @xmath21 , @xmath22 for number of cells with value @xmath23 .
the sequence was initially introduced due to consideration of `` noise '' in computationally universal ca @xmath24 , but it is shown below , that for other three ca the sequences are the same and @xmath25 . due to definition of second - order ca eq .
( [ 2nd ] ) a simple property is true @xmath26 and so @xmath27 initial terms of the sequences are represented in the table below : @xmath28
few recursive equations are proved in this paper : @xmath29 @xmath30 the negative value of @xmath0 can be used because ca are reversible . due to eq .
( [ r1r2 ] ) last formula is equivalent with @xmath31 both eq . ( [ recr1 ] ) and eq . ( [ recr ] )
are simply derived from the equation eq .
( [ recr2 ] ) : @xmath32 @xmath33 an _ alternative form of recursive equations _ is also valid for @xmath34 and @xmath35 : @xmath36 @xmath37 these equations are equivalent due to eq .
( [ r1r2 ] ) and together with eq .
( [ rsum ] ) imply a simple relation between the sequences @xmath38 equations eq .
( [ rec2r1 ] ) and eq .
( [ rec2r2 ] ) can be proved _ by induction _ using eq .
( [ recr1 ] ) and eq .
( [ recr2 ] ) respectively . due to
( [ r1r2 ] ) , it is enough to consider only one of them . the eq .
( [ recr2 ] ) holds for @xmath39 .
assume eq .
( [ recr2 ] ) holds for any @xmath40 , @xmath41 with @xmath42 , @xmath43 .
( [ recr2 ] ) allows us to express @xmath35 as a linear combination with terms smaller than @xmath23 and to show that the equation holds also for @xmath44 : @xmath45 @xmath46 where @xmath47 , @xmath48 .
@xmath49 it remains to prove eq .
( [ recr2 ] ) .
the recursion is proved below for simpler case with ca @xmath50 and @xmath51 with straightforward demonstration of equivalence for ca @xmath52 and @xmath24 .
let us start with consideration of @xmath11 and @xmath13 .
these ca are linear ( additive ) @xcite , _ i.e. _ for any two configurations @xmath53 and @xmath54 local rule defines global map @xmath55 with property @xmath56 where @xmath57 is symmetric difference configurations @xmath53 and @xmath54 considered as sets ( regions ) of cells with unit values . a configuration of 2d ca can be described with ( characteristic ) polynomial @xmath58 \equiv p_{x , y}[c ] = \sum_{i , j=-\infty}^\infty c_{i , j } x^i y^j \label{poly}\ ] ] and eq .
( [ lin2 ] ) corresponds to @xmath59 = p[f(a ) ] \oplus p[f(b ) ] \equiv p[f(a ) ] + p[f(b ) ] \mod 2 .
\label{polin2}\ ] ] it is convenient further for ca with two states to treat eq .
( [ poly ] ) as a polynomial over @xmath60 .
let us consider evolution of pattern @xmath61 with single nonzero cell @xmath62 for ca @xmath11 .
it can be described using equation for global transition rule @xmath63 & \mapsto & ( x^{-1}y^{-1 } + x y^{-1 } + x^{-1}y + xy)\ , p_{x , y}[c ] \notag\\ & = & ( x^{-1}+x ) ( y^{-1 } + y)\,p_{x , y}[c ] .
\label{c1poly } \end{aligned}\ ] ] here treatment of @xmath64 $ ] as a polynomial over @xmath60 is especially useful and after @xmath0 steps due to eq .
( [ c1poly ] ) @xmath65 = ( x^{-1}+x)^n ( y^{-1 } + y)^n\,p_{x , y}[c ] .
\label{c1poln}\ ] ] the polynomial of pattern @xmath61 is @xmath64=p[\delta_{0,0}]=1 $ ] and the eq .
( [ c1poln ] ) corresponds to decomposition @xmath66 on two characteristic polynomials @xmath67 of 1d cellular automata with local rule @xcite @xmath68 also known as `` rule 90 '' @xcite and initial pattern @xmath69 with single nonzero cell @xmath70 .
the number of cells on @xmath23-th step may be described by equation @xmath71 there @xmath72 is number of units in binary decomposition of @xmath23 @xcite .
the polynomial is over @xmath60 and a property used further @xmath73 is simply derived using recursion on @xmath23 : @xmath74 eq .
( [ binom2k ] ) can be used for inductive proof of eq .
( [ n90 ] ) . for @xmath75
( [ n90 ] ) holds : @xmath76 .
assume @xmath77 for @xmath78 .
for @xmath79 characteristic polynomial is @xmath80 and because @xmath81 and @xmath82 are not `` overlapped , '' @xmath83 . due to @xmath84 for @xmath85 : @xmath86 .
so , eq . ( [ n90 ] ) holds for @xmath87 .
@xmath49 the decomposition eq .
( [ c1poly ] ) produces some simplification with comparison to @xmath13 @xmath88 \mapsto ( x^{-1 } + x + y^{-1 } + y)\ , p_{x , y}[c ] .
\label{c2poly}\ ] ] on the other hand , @xmath11 ( @xmath50 ) may be considered as two independent copies of @xmath13 ( @xmath51 ) on two _ `` diagonal '' sublattices _ corresponding @xmath9 with even and odd @xmath89 respectively : @xmath90 visually , they correspond to cells with black and white colors on checkerboard pattern after @xmath91 rotation of the board . because @xmath92 belongs to even sublattice @xmath93 , configuration of @xmath11 after any @xmath0 steps always belongs to @xmath93 and it is equivalent with @xmath13 acting on the diagonal sublattice . due to
( [ binom2k ] ) and eq .
( [ c1poln ] ) application of @xmath94 steps of @xmath11 to arbitrary configuration @xmath95 may be expressed as @xmath96 \mapsto ( x^{-2^k}y^{-2^k}+x^{2^k}y^{-2^k } + x^{-2^k}y^{2^k}+x^{2^k}y^{2^k})\,p_{x , y}[c ] \label{c1expmov}\ ] ] and analogue property can be proved for @xmath13 @xmath97 \mapsto ( x^{-2^k}+x^{2^k}+y^{-2^k } + y^{2^k})\,p_{x , y}[c ] .
\label{c2expmov}\ ] ] so patterns bounded by @xmath98 are replicated into four copies after @xmath94 steps both for @xmath11 and @xmath13 . for @xmath11 coordinates of four copies
are shifted due to eq .
( [ c1expmov ] ) as @xmath99 , @xmath100 , @xmath101 , @xmath102 and for @xmath13 due to eq .
( [ c2expmov ] ) the shifts are @xmath103 , @xmath104 , @xmath105 , @xmath106 .
such ca with replicating property was initially considered by e. fredkin in 1970s @xcite .
for @xmath11 and @xmath13 an analogue of eq .
( [ n90 ] ) is true @xmath107 configuration of @xmath11 is represented as product eq .
( [ c1poln ] ) @xmath108 of two `` rule 90 '' ca and eq .
( [ n90sqr ] ) can be derived directly from eq .
( [ n90 ] ) . in more general case for such products of two 1d configurations @xmath53 and @xmath54 an equation @xmath109 can be used , where @xmath110 is 2d configuration with values of cells @xmath111 .
a direct proof by induction for @xmath13 or @xmath11 is also useful due to similarity with further approach to second - order ca . for @xmath75
( [ n90sqr ] ) holds : .
assume @xmath112 for @xmath78 .
for @xmath79 characteristic polynomial for @xmath13 satisfies eq .
( [ c2expmov ] ) @xmath113 and describes four shifted nonoverlapping copies of region @xmath114 .
so , @xmath115 and eq . ( [ n90sqr ] ) holds for @xmath87 .
@xmath49 similar proof by induction for @xmath11 uses eq .
( [ c1expmov ] ) .
a second - order ca corresponds to pair of polynomials @xmath116 . for second - order ca derived from ca with two states described by polynomials over gf(2 ) local rule eq .
( [ 2nd ] ) can be simply rewritten as a global one @xmath117 + p_2(x , y)\ , , p_2(x , y)\bigr ) .
\label{plin2nd}\ ] ] for @xmath11 , @xmath13 due to eq .
( [ c1poly ] ) and eq .
( [ c2poly ] ) @xmath118 = t(x , y ) p(x , y ) \mod 2 \label{tp}\ ] ] with @xmath119 @xmath120 let us prove that for @xmath11 , @xmath13 with initial configuration @xmath121 with single nonempty cell @xmath122 after @xmath23 steps the configuration is described by polynomial @xmath123 = \bigl(f_{k+1}(t),f_k(t)\bigr ) , \label{pft}\ ] ] where @xmath124 are polynomials over gf(2 ) defined using recursive equation @xmath125 and @xmath126 is application of the polynomial to @xmath127 eq .
( [ tp ] ) also considered over gf(2 ) . for @xmath128
( [ pft ] ) holds @xmath129 = ( 1,0)$ ] assume eq .
( [ pft ] ) holds for @xmath130 , for @xmath131 due to eq .
( [ plin2nd ] ) @xmath132 the eq .
( [ fibpol ] ) defines _
fibonacci polynomials_. the _ lucas polynomials _ ( also used below ) are defined by the same recursive equation with other initial conditions @xcite @xmath133 @xmath134 with simpler correspondence over gf(2 ) @xmath135 some relations with lucas and fibonacci polynomials @xcite are useful further @xmath136 @xmath137 for gf(2 ) multiplier @xmath138 can be omitted and due to relation eq .
( [ fibluc2 ] ) from eq .
( [ fiblucrec ] ) for polynomials @xmath126 over gf(2 ) follows @xmath139 for @xmath140 eq .
( [ tftrec ] ) gives @xmath141 and eq .
( [ fib2rec ] ) gives for @xmath140 @xmath142 it again may be modified for polynomials over gf(2 ) @xmath143 let us show for polynomials over gf(2 ) @xmath144 it holds for @xmath75 and for @xmath131 due to eq .
( [ tft2n ] ) @xmath145 let us consider @xmath146 with @xmath147 .
due to eq .
( [ tftrec ] ) and eq .
( [ fpow2 ] ) @xmath148
a state of cell in the second - order ca for pair @xmath149 was encoded as @xmath150 .
the values one and two correspond to pairs @xmath151 and @xmath152 respectively .
show that these properties are valid after next step .
let us denote @xmath155 and @xmath156 configurations corresponding to set of cells with nonzero first and second elements of pair @xmath149 respectively the properties above claim that configurations @xmath155 and @xmath156 belong to _ diagonal sublattices _ with opposite colors .
the sublattices are represented by polynomials with odd and even degrees , so configurations with properties above correspond to either ( even , odd ) or ( odd , even ) pairs of polynomials .
the operator @xmath157 eq .
( [ tc2 ] ) changes degree of monomial on unit and so eq . ( [ tp ] )
exchanges odd and even polynomials and eq .
( [ plin2nd ] ) maps configuration ( odd , even ) into ( even , odd ) and vise versa .
initial configuration also has desired properties and so equation @xmath158 is proved by induction . @xmath49 it is more convenient sometimes to use @xmath11 instead of @xmath13 and it is possible to introduce analogues of structures discussed below .
it was already mentioned that @xmath13 itself corresponds to _ diagonal sublattice _ of @xmath11 and so notion of cells with the _ `` same color '' _ needs for some clarification .
relation between `` coloring '' for @xmath51 ( @xmath13 ) and @xmath50 ( @xmath11 ) is shown on fig .
[ fig : check2x ] . for @xmath51 ( @xmath13
) coloring of cell @xmath159 used for illustration properties above is corresponding to value @xmath160 .
next , all the @xmath13 board is mapped into sublattice of @xmath11 producing new type of coloring with _
`` light '' _ and _ `` dark '' _ cells illustrated on fig . [ fig : check2x ] .
let us use for @xmath50 notation already introduced for @xmath51 with @xmath155 and @xmath156 configurations corresponding to set of cells with nonzero first and second elements in the pair representing a state of second - order ca .
it was shown that for configurations derived from a single cell with unit state such patterns have opposite color .
for @xmath13 it corresponds to different diagonal sublattices and in each pattern nonempty cells can not have adjoint sides , but may have common corners . for @xmath13 with new scheme of coloring the corners of cells are also separated .
let us first prove such expressions as eq .
( [ rec2r1 ] ) and eq .
( [ rec2r2 ] ) .
they already were derived above from eq .
( [ recr1 ] ) and eq .
( [ recr2 ] ) , but direct proof provided below illustrates some useful relations . the equation eq .
( [ rec2r2 ] ) may be derived from eq .
( [ tft2n ] ) and eq .
( [ fft2n1 ] ) .
let us recollect that for any polynomial @xmath163 over gf(2 ) @xmath164 and so for representations of two - states pattern via polynomials used earlier the square corresponds to rescaling of the pattern @xmath165 .
( [ tft2n ] ) corresponds to multiplication of @xmath127 on the rescaling pattern . for @xmath11 @xmath127
is described by eq .
( [ tc1 ] ) .
it was already shown , that for @xmath11 any cells with same value are separated , so after the scaling distances between nonzero cells are enough to put four new cells generated by @xmath127 without overlap .
[ fig : compos2n ] illustrates that for @xmath168 so , eq . ( [ tft2n ] ) proves first part of eq .
( [ rec2r2 ] ) , @xmath169 .
next , due to eq .
( [ 2nd ] ) two polynomials @xmath170 , @xmath171 in eq .
( [ fft2n1 ] ) describes @xmath172 on a step @xmath0 and it was already shown that the pattern are not intersecting for chosen initial conditions .
square of the sum only rescales the union without changing number of nonzero cells . fig .
[ fig : compos2n ] illustrates that for @xmath173 so , eq . ( [ fft2n1 ] ) proves second part of eq .
( [ rec2r2 ] ) , @xmath174 .
@xmath49 recursive polynomial equation eq .
( [ f2kj ] ) can be simply adopted for proof of eq .
( [ recr2 ] ) for number of cells in @xmath50 and @xmath51 and it is enough to demonstrate both eq .
( [ recr ] ) and eq .
( [ recr1 ] ) .
let us prove eq .
( [ recr2 ] ) for number of cells with state 2 in @xmath50 using eq .
( [ f2kj ] ) . the fact , that all cells with state 2 on each step @xmath23 are contained within a square region represented as direct product of two _ open _
intervals @xmath175 is also used and proved .
assume that equations hold for all patterns @xmath178 and consider @xmath179 .
due to eq .
( [ f2kj ] ) and eq .
( [ c1expmov ] ) the polynomial representation is @xmath180 the multiplier before @xmath181 produces four copies moved in directions @xmath182 , @xmath183 , @xmath184 , @xmath185 and @xmath186 corresponds to pattern in the center , fig .
[ fig : compos2 ] . the five patterns are not overlapped : central one with @xmath187 nonempty cells
is contained within @xmath188 and other four others with @xmath189 nonempty cells are distributed within a `` four - fold '' disjointed region described by product @xmath190 total number of nonempty cells is @xmath191 so the equation for number of cells eq .
( [ recr2 ] ) holds for @xmath192 .
the union of the five regions belongs to square @xmath193 .
@xmath49 the five patterns have a unit gap between them ( fig . [
fig : compos2 ] ) and only after consideration of all cells with nonzero values corresponding to union of both `` checkerboard sublattices '' the final patterns ( fig .
[ fig : compos ] ) belong to square regions described by product of _ closed _
intervals @xmath197 \times [ -k , k]$ ] and recursive equation eq .
( [ recr ] ) corresponds to union of five disjoint regions without gaps , fig .
[ fig : compos ] .
let us check recursive equation for pair of polynomials eq .
( [ pft ] ) representing all states of second - order ca and used for calculation of @xmath20 @xmath201 & = & \bigl(f_{2^k+j+1}(t),f_{2^k+j}(t)\bigr ) \notag \\ & = & \bigl(t^{2^k}f_{j+1}(t)+f_{2^k - j-1}(t),t^{2^k}f_j(t)+f_{2^k - j}(t)\bigr ) \notag \\ & = & t^{2^k}\bigl(f_{j+1}(t),f_j(t)\bigr)+ \bigl(f_{2^k - j-1}(t),f_{2^k - j}(t)\bigr ) \notag \\ & = & t^{2^k } p[c_j ] + p[{{\mathsf x}}c_{2^k - j-1 } ] , \label{c2kj } \end{aligned}\ ] ] where @xmath5 operation eq .
( [ swap ] ) swaps values @xmath202 .
( [ c2kj ] ) illustrates dynamics of pattern growth , fig .
[ fig : compos ] . due to
( [ invswap ] ) application of transition rule @xmath1 to pattern @xmath203 for any index @xmath204 satisfies property @xmath205 so , application of @xmath1 to eq .
( [ c2kj ] ) corresponds to increase of four patterns @xmath206 and decrease of central region @xmath207 until @xmath208 . for @xmath209 four outer configurations
reach maximal size and may not grow more , so on next step they are joined into single central configurations @xmath210 and four cells @xmath211 appear near corners as centers for future growth .
proofs of eqs .
( [ recr][rec2r2 ] ) for @xmath51 directly follow from consideration of @xmath50 , because ( similarly with relation between @xmath11 and @xmath13 discussed earlier ) @xmath51 is equivalent with @xmath50 acting on _ a diagonal sublattice_. in such representation patterns for @xmath51 may look more closely packed fig .
[ fig : composr2 ] , but it does not change recursive equations due to above mentioned equivalence . let us now consider @xmath52 and @xmath24 .
local rule for both @xmath51 and @xmath52 uses only four closest cells with common sides in so - called _ von neumann neighborhood_. due to eq .
( [ 2nd ] ) it is enough to consider actions of local rules for @xmath13 and @xmath15 on the first element of pair to describe differences between rules .
if the rules act in the same way for any configuration under consideration , then actions of @xmath51 and @xmath52 for patterns derived from @xmath61 are also the same .
comparison of definition @xmath13 and @xmath15 shows that local rules differ only for _ three nonempty cells _ in von neumann neighborhood . on fig .
[ fig : compos1r2 ] for simplicity are shown only cells with nonzero first components in the pair for configurations used earlier , fig .
[ fig : composr2 ] .
all such pattern have 0,1,2,4 nonempty cells in von neumann neighborhood and so @xmath51 and @xmath52 act in the same way for such pattern .
let us proof the property by induction .
any new configuration is composition of five previous patterns and it is enough to consider new configurations near contiguities of they boundaries . due to consideration below for @xmath213
there are four contacts of central pattern with outer configurations .
four cells with two neighbors corresponds them .
the cases @xmath214 correspond to contacts of four outer patterns and due to symmetry number of neighbors there are always even .
in fact , it may be simply shown that all such configuration ( of cells with state 1 ) are simple diamond - like checkerboard patterns with @xmath215 cells , fig .
[ fig : compos1r2 ] .
let us now consider @xmath216 .
the only difference between @xmath217 and @xmath15 is additional requirement about cells with common corners .
the limitation always holds due to `` coloring '' properties already discussed earlier on page .
indeed , each new generation of cells with state 1 for @xmath51 may appear only on checkerboard sublattice with opposite colors , _
i.e. _ all cells with common corner for an empty cell going to be switched into the state 1 are empty .
so , evolution of @xmath216 starting with configuration @xmath61 is also the same as for @xmath52 and @xmath51 . @xmath49 9 t. toffoli and n. margolus , `` invertible cellular automata : a review , '' _ physica d * 45 * _ , 229253 ( 1990 ) . s. wolfram , _ cellular automata and
complexity : collected papers _ , ( addison - wesley , reading ma 1994 ) . l. le bruyn and m. van den bergh , `` algebraic properties of linear cellular automata , '' _ lin .
157 * _ , 217234 ( 1991 ) .
s. wolfram , `` statistical mechanics of cellular automata , '' _ rev .
phys . * 55 * _ , 601644 ( 1983 ) ; _ reprinted _ in @xcite , 370 .
o. martin , a. m. odlyzko , and s. wolfram , `` algebraic properties of cellular automata , '' _ comm .
phys . * 93 * _ , 219258 ( 1984 ) ; _ reprinted _ in @xcite , 71113 .
b. chopard and m. droz , _ cellular automata modeling of physical systems _ , ( cambridge university press , cambridge 1998 ) .
t. koshy , _ fibonacci and lucas numbers with applications _ , ( john wiley & sons , new york , 2001 ) . | recursive equations for the number of cells with nonzero values at @xmath0-th step for some two - dimensional reversible second - order cellular automata are proved in this work .
initial configuration is a single cell with the value one and all others zero . |
in 1985 , hoste - ocneanu - millett - freyd - lickorish - yetter @xcite had discovered the homfly polynomial which is a 2-variable oriented link polynomial @xmath2 motivated by the jones polynomial .
also , prztycki and traczyk @xcite independently had done some work related to the homfly polynomial .
the calculation of the homfly polynomial is based on the homfly skein relations as follows . + 1 .
@xmath3 is an isotopy invariant .
2 . @xmath4(unknot)=1 .
3 . @xmath5 .
lickorish and millett @xcite and kanenobu and sumi @xcite gave a formula to calculate the homfly polynomials of 2-bridge knots by using a representation of the continued fraction of a rational knot into a group of @xmath6 matrices .
( see proposition 14 of @xcite . )
+ in this paper , i use the homfly blacket polynomial and the plat presentation of knots to calculate the homfly polonomials of rational knots by using a representation of the braid group @xmath0 into a group of @xmath1 matrices .
also , we can extend the method to evaluate the homfly polynomials of @xmath7-plat presentations of knots .
+ now , we define the plat presentation of a knot and a rational tangle .
let @xmath8 be a sphere smoothly embedded in @xmath9 and let @xmath10 be a link transverse to @xmath8 .
the complement in @xmath9 of @xmath8 consists of two open balls , @xmath11 and @xmath12 .
we assume that @xmath8 is @xmath13-plane @xmath14 .
now , consider the projection of @xmath10 onto the flat @xmath15-plane .
then , the projection onto the @xmath15-plane of @xmath8 is the @xmath16-axis and @xmath11 projects to the upper half plane and @xmath12 projects to the lower half plane .
the projection gives us a _ link diagram _
, where we make note of over and undercrossings .
the diagram of the link @xmath10 is called a @xmath17 , denoted by @xmath18 , if it is the union of a @xmath7-braid @xmath19 and @xmath7 unlinked and unknotted arcs which connect pairs of consecutive strings of the braid at the top and at the bottom endpoints and @xmath8 meets the top of the @xmath7-braid .
( see the first and second diagrams of figure [ p2 ] . )
any link @xmath10 in @xmath9 admits a plat presentation , that is @xmath10 is ambient isotopic to a plat ( @xcite , theorem 5.1 ) .
the bridge ( plat ) number @xmath20 of @xmath10 is the smallest possible number @xmath21 such that there exists a plat presentation of @xmath10 on @xmath7 strings .
+ we know that the braid group @xmath22 is generated by @xmath23 which are twisting of two adjacent strings .
for example , @xmath24 is the word for the 4 braid of the first diagram of figure [ p2 ] .
+ then we say that a plat presentation is @xmath25 if the @xmath26-braid @xmath19 of @xmath27 involves only @xmath28 .
+ a @xmath29 is the disjoint union of @xmath30 properly embedded arcs in a 3-ball @xmath31 .
then we say that a @xmath30-@xmath32 @xmath33 is @xmath34 if there exists a homeomorphism of pairs @xmath35 , where @xmath36 are distinct points in @xmath37 and @xmath38 $ ] .
we have the two @xmath30-tangles @xmath39 and @xmath40 .
we note that @xmath41 and @xmath42 are rational if @xmath10 has a plat presentation .
let @xmath43 be the rational @xmath30-tangle in @xmath12 if @xmath10 has a plat presentation .
+ now , we define a @xmath44 for rational @xmath30-tangles @xmath45 ( refer to @xcite . ) denoted by @xmath46 if it is the union of a @xmath26-braid @xmath19 and @xmath30 unlinked and unknotted arcs which connect pairs of strings of the braid at the bottom endpoints with the same pattern as in a plat presentation for a knot and @xmath47 meets the top of the @xmath26-braid .
+ we note that @xmath46 is a rational @xmath30-tangle in @xmath12 .
( see @xcite . ) + we say that @xmath48 = @xmath27 is the @xmath49 of @xmath46 . + the tangle diagrams with the circles in figure [ p3 ] give the diagrams of trivial rational @xmath30-tangles as in @xcite , @xcite , @xcite , @xcite .
we note that @xmath50 is alternating if and only if @xmath51 is alternating .
+ a tangle @xmath52 is @xmath53 alternating if @xmath52 is alternating and @xmath52 does not have a self - crossing which can be removed by a type i reidemeister move . [ t1 ] if @xmath10 is a 2-bridge knot , then there exists a word @xmath19 in @xmath0 so that the plat presentation @xmath54 is reduced alternating , standard and represents a knot isotopic to @xmath10 . by theorem [ t1 ]
, any 2-bridge knot @xmath10 can be represented by a word @xmath19 as a plat which involves only @xmath55 and @xmath56 ( or @xmath57 and @xmath58 ) .
i.e. , @xmath59 for some positive ( negative ) integers @xmath60 for @xmath61 .
we notice that if @xmath62 for some positive ( negative ) integer @xmath63 then it is not a reduced alternating form .
i.e. , we can twist the right unlinked and unknotted bottom arc to reduce some crossings to have fewer crossings for @xmath10 .
so , in order to have a reduced alternating form , @xmath19 needs to start from @xmath64 and end at @xmath64 .
+ in section 2 , we introduce the homfly polynomial of rational 2-tangles and give a formula to calculate the homfly polynomial of 4-plat presentations of knots . + in section 3
, we give a method to find the orientation of each crossing of a knot from a given orientation of the knot . +
then , we give some examples of knots for which we calclulate the homfly polynomials and especially give examples of a 2-bridge knot and a 3-bridge knot that have the same jones polynomial , but different homfly polynomials in section 4 .
let @xmath10 be a 2-bridge knot
. by theorem [ t1 ] there exists a plat presentation @xmath54 which is reduced alternating , standard and represents a knot isotopic to @xmath10 .
+ for a given orientation of @xmath10 , we will give an induced orientation to the rational @xmath30-tangle @xmath65 such that @xmath43 has the same orientation with the oriented knot @xmath66 in @xmath12 .
let @xmath67 be the plat presentation of @xmath43 with the induced orientation .
+ now , we define the homfly polynomial of an oriented plat presentation of rational @xmath30-tangle @xmath67 in @xmath12 as @xmath68 , where the coefficients @xmath69 and @xmath70 are polynomials in @xmath71 and @xmath72 that are obtained by starting with @xmath73 and using the skein relations repeatedly until only the three tangles in the expression given for @xmath43 .
+ let @xmath21 be the number of crossings of @xmath19 .
+ we note that if we switch one of the alternative crossings of @xmath19 @xmath74 from positive ( negative ) to negative ( positive ) to have @xmath75 , then we have the plat presentation @xmath76 for @xmath75 so that @xmath77 is a reduced alternating , standard and @xmath77 has lower crossings than @xmath19 .
( refer to @xcite . )
so , by the skein relations , we can reduce the number of crossings of @xmath19 .
however , if we have an oriented rational @xmath30-tangle with @xmath78 then we can not reduce the number of crossing by the skein relations .
this is the reason that we need @xmath79 .
+ we remark that polynomials @xmath69 and @xmath70 are invariant under isotopy of @xmath67 . + also , we note that even if we apply the skein relationship to one of the crossings of @xmath80 , the orientations of the rest of crossings will be preserved .
i.e. , all the rest of crossings keep the directions for the given orientation while we are applying the skein relationship to one of the crossings of @xmath67 to calculate the homfly polynomial of @xmath80 .
+ let @xmath81 , @xmath82 and @xmath83 .
+ recall that @xmath10 is alternating and standard .
+ since @xmath54 is standard , we consider @xmath84 instead of @xmath0 .
then let @xmath85 and @xmath55 be the two generators of @xmath84 .
i want to emphasize here that we are changing from @xmath55 and @xmath58 to @xmath85 and @xmath55 .
+ suppose that @xmath86 for positive integers @xmath87 ( @xmath61 ) .
then , we will give an orientation to @xmath19 which is induced by @xmath66 .
so , @xmath85 and @xmath57 have four different cases @xmath88 and @xmath89 for @xmath90 as in figure [ p5 ] . actually , there are two possible directions for the orientation of @xmath66 .
however , the skein relation does not depend on the choice of the direction .
now , we consider the sub - directions which is induced by the orientation of @xmath66 as in figure [ p5 ] .
we note that there are the other corresponding eight cases which are obtained from the given cases by taking the opposite arrows .
however , we will not distinguish the corresponding cases since they play the same role as the corrsponding cases when we construct @xmath1 matrices for calculation of the homfly polynomial .
+ so , by considering the orientation , we can describe @xmath19 as @xmath91 instead of @xmath86 , where @xmath92 .
+ let @xmath93 $ ] , @xmath94,\\$ ] and , @xmath95 $ ] , @xmath96.\\$ ] let @xmath97 + then we have the main theorem to calculate the homfly polynomial of @xmath10 as follows .
+ [ t2 ] suppose that @xmath50 is a plat presentation of a rational 2-tangle @xmath43 which is alternating and standard so that @xmath98 for some positive integers @xmath60 ( @xmath61 )
. then @xmath99 , where @xmath100 and @xmath70 are obtained from @xmath101=[0~1~0]m^t$ ] .
i.e. , the second column of @xmath102 . moreover , @xmath103 . by theorem [ t1 ] , for a two bridge knot @xmath10 , there exists a word @xmath19 in @xmath0 so that the plat presentation @xmath54 is alternating , standard and represents a link isotopic to @xmath10 .
+ then , by the argument above , we have @xmath99 for some 2-variable polynomials @xmath100 and @xmath70 .
+ since @xmath19 is standard , we consider two generators @xmath85 and @xmath55 for @xmath84 as mentioned before .
+ let @xmath104 and @xmath105 be the rational two tangles which are obtained from @xmath43 by adding @xmath106 or @xmath64 respectively to cancel the first @xmath107 in @xmath19 .
so , we have a new word @xmath108 of smaller length than @xmath19 so that @xmath109 for the rational 2-tangles @xmath104 and @xmath105 . without loss of generality , we will consider the cases that @xmath110 or @xmath111 .
+ first , we consider the case that @xmath110 .
then , we have that @xmath112 for some 2-variable polynomials @xmath113 and @xmath114
. + also , by figure [ p6 ] , we know that @xmath115 , where @xmath116 is the tangle which has a plat presentation @xmath50 with @xmath117 . + by figure [ p6 ] , we also know that @xmath118 or @xmath119 .
+ therefore , @xmath120 or , + @xmath121 + therefore , the following gives the operations .
@xmath122\left[\begin{array}{c } f(a , m)\\ g(a , m)\\ h(a , m)\\ \end{array } \right]=\left[\begin{array}{c } f'(a , m)\\ g'(a , m)\\ h'(a , m)\\ \end{array } \right ] , \left [ \begin{array}{ccc } 1 & -a^{-1}m & 0 \\ 0 & 0 & 1 \\ 0 & -a^{-2 } & 0\\ \end{array } \right]\left[\begin{array}{c } f(a , m)\\ g(a , m)\\ h(a , m)\\ \end{array } \right]=\left[\begin{array}{c } f'(a , m)\\ g'(a , m)\\ h'(a , m)\\ \end{array } \right].$ ] now , we consider the case that @xmath111 .
+ then we have @xmath123 .
+ by figure [ p7 ] , we also know that @xmath119 or @xmath118 .
+ therefore , @xmath124 or , + @xmath125 + therefore , the following gives the operations .
+ @xmath126\left[\begin{array}{c } f(a , m)\\ g(a , m)\\ h(a , m)\\ \end{array } \right]=\left[\begin{array}{c } f''(a , m)\\ g''(a , m)\\ h''(a , m)\\ \end{array } \right ] , \left [ \begin{array}{ccc } 0 & 0 & 1 \\ -am & 1 & 0 \\ -a^{2 } & 0 & 0\\ \end{array } \right]\left[\begin{array}{c } f(a , m)\\ g(a , m)\\ h(a , m)\\ \end{array } \right]=\left[\begin{array}{c } f''(a , m)\\ g''(a , m)\\ h''(a , m)\\ \end{array } \right].$ ] + now , let @xmath127,$ ] @xmath128,\\$ ] and , @xmath129 $ ] , @xmath130.\\$ ] now , recall that @xmath93,$ ] @xmath131,\\$ ] and , @xmath95 $ ] , @xmath96.\\$ ] we note that each @xmath132 is invertible and @xmath133 is actually the inverse of @xmath134 .
+ also , we note that each @xmath135 is invertible and @xmath136 is actually the inverse of @xmath137 .
+ therefore , @xmath138=\left[\begin{array}{c } 0\\ 1\\ 0\\ \end{array } \right]$ ] since @xmath139 + this implies that @xmath140 .
+ we remark that @xmath85 and @xmath57 are corresponding to @xmath141 and @xmath142 depending on the given orientation of @xmath19 .
+ now , by attaching the three unlinked and unknotted arcs in @xmath11 , we can calculate @xmath143 by figure [ p8 ] .
+ to see this , we need the fact that the homfly polynomial of a link @xmath144 that is a split union of two links @xmath145 and @xmath146 is given by @xmath147 .
+ so , we have @xmath148 for the disjoint union of two unknots .
+ finally , we have @xmath149 + we note that if we have a word @xmath19 which involves only @xmath150 and @xmath55 then we can get a @xmath1 matrix @xmath151 which is a composing sequences of @xmath152 and @xmath153 which are corresponding to @xmath150 and @xmath55 depending on a given orientation of @xmath19 .
+ we have the equality about the mirror image of @xmath10 as follows
. + @xmath154 .
first , assume that the projection onto the @xmath15-plane of a 2-bridge knot @xmath10 has a standard plat presentaion @xmath54 with @xmath98 for some positive ( negative ) integers @xmath60 ( @xmath61 ) .
+ then we have the plat presentation @xmath50 of the tangle @xmath157 so that @xmath158 .
+ let @xmath159 be the @xmath1 matrix which is obtained by interchanging the @xmath160 and @xmath161 rows of @xmath162 .
+ then @xmath163 extends to a homomorphism from @xmath84 to @xmath164 .
+ for an element @xmath19 of @xmath84 , let 1,2,3 be the upper endpoints of the three strings for @xmath84 from the left .
also , let @xmath165 be the upper endpoint of the left most string for @xmath0 .
let @xmath166 $ ] .
then we assign the same number to the other endpoint of the three strings .
then we say that the new ordered sequence of numbers @xmath167 is the @xmath168 .
[ t3 ] suppose that @xmath19 is an element of @xmath84 so that @xmath98 for some positive ( negative ) integers @xmath60 ( @xmath61 ) .
+ then @xmath169\mathcal{p}(w)$ ] is the permutation which is induced by @xmath19 .
this is proven by induction on @xmath170 .
now , let @xmath171=[1,2,3]\mathcal{p}(w)$ ] .
+ without loss of generality , give the orientation ( clockwise ) to the trivial arc @xmath172 in @xmath11 with @xmath173 from @xmath174 to @xmath165 along @xmath172 .
so , the initial point of @xmath172 is 1 and the terminal point of @xmath172 is 0 for the given orientation .
then , we can give the orientation to the other trivial arcs @xmath175 in @xmath11 as follows , where @xmath176 .
+ now , we consider @xmath177
. then we note that @xmath178 . if not
, then @xmath10 is a link , not a knot .
[ t4 ] if @xmath179 then the trivial arc @xmath175 has the same direction ( clockwise ) as @xmath172 for the orientation .
if @xmath180 then the trivial arc @xmath175 has the opposite direction ( counter clockwise ) as @xmath172 for the orientation . if @xmath179 then @xmath181 .
so , the direction of the orientation at @xmath182 is upward .
so , the @xmath175 has the same direction with @xmath172 .
+ if @xmath180 then @xmath183 .
then the direction of the orientation at @xmath30 is upward .
so , the @xmath175 has the opposite direction with @xmath172 .
recall the ordered sequence of numbers @xmath166 $ ] .
now , we will define a new sequence of numbers @xmath184 $ ] . for the given orientation , we replace the original number for the initial point of @xmath175 by 1 as follow .
+ @xmath184 $ ] so that @xmath185 and for @xmath186 , @xmath187 if @xmath188 and @xmath189 if @xmath190 .
+ for the three strings of the braids @xmath19 , we assign the number @xmath191 to each string with the upper endpoint @xmath192 for @xmath193 .
+ now , let @xmath194 .
+ let @xmath195= r\mathcal{p}(\sigma_{1}^{\epsilon_1}\sigma_{2}^{-\epsilon_2}\sigma_1^{\epsilon_3}\cdot\cdot\cdot \sigma_{\delta}^{(-1)^{i-1 } \epsilon_{i}})$ ] for @xmath196 , where @xmath197 if @xmath160 is odd and @xmath198 if @xmath160 is even .
+ let @xmath199if @xmath200if @xmath201 .
+ [ t5 ] suppose that the projection of a knot @xmath10 onto the @xmath15-plane has a plat presentation @xmath54 with @xmath202 for some positive ( negative ) integers @xmath60 ( @xmath61 ) .
+ then a given orientation for @xmath10 , we have @xmath203 for @xmath155 which is defined above . without loss of generality ,
we give the orientation ( clockwise ) to @xmath172 from @xmath174 to @xmath30 along @xmath172 .
then the direction of the orientation at 1 is up and the direction at 2 is down .
then we know that the orientation at @xmath160 is up if @xmath187 and is down if @xmath204 .
+ fix a value @xmath160 .
+ case 1 : suppose that @xmath205 .
+ then the two strings for the ( @xmath206-th crossing have the same direction of the orientation since @xmath205 so , the directions of the orientation are upward .
+ then @xmath207 .
this is consistant with the @xmath155 that is defined above .
+ case 2 : suppose that @xmath208 .
+ then the two strings for the ( @xmath206-th crossing have different directions for the orientation since @xmath208 .
+ then @xmath209 . this is consistant with the @xmath155 that is defined above .
first , we will calculate the homfly polynomials of @xmath210 ( trefoil knot ) , @xmath211 and @xmath212 .
+ 1 .
@xmath210 is represented by @xmath213 .
then we have @xmath214 .
+ so , @xmath101=[-a^2+a^2m^2~a^3m~0]$ ] .
+ therefore , @xmath215 .
@xmath211 is represented by @xmath216 .
then we have @xmath217 .
+ so , @xmath101=[a^4 - 3a^4m^2+a^4m^4~a^5m(-2+m^2)~0]$ ] .
+ therefore , @xmath218 .
@xmath212 is represented by @xmath219 .
then we have @xmath220 .
+ so , @xmath101=[a^3m - a^3m^3-a^5m+a^5m^3~ a^4-a^4m^2+a^6m^2~0]$ ] . + therefore , @xmath221 .
+ now , i will give a set of knots @xmath222 and @xmath223 so that they have the same jones polynomial but different homfly polynomials .
+ consider the two knots @xmath224 and @xmath225 .
+ first , we can check that the kauffman ( jones ) polynomial of @xmath222 and @xmath223 are the same as follows .
+ @xmath226 .
+ we note that @xmath227 is represented by @xmath228 .
+ so , @xmath101=[0~(-m^7a+4am^5 - 5am^3+a^{-1}m^5 - 2a^{-1}m^3+a^{-1}m+a^4a^{-1}m^5 - 3a^3m^3 + 2a^3m+am)~(-m^6a^2 + 3a^2m^4+m^4 - 3a^2m^2-m^2+a^4m^4 - 2a^4m^2+a^4)].$ ] + therefore , @xmath229 .
+ however , @xmath230 is prepresented by @xmath231 .
+ so , @xmath101=[0~(-am^3+a^{-1}m+a^3m)~(-a^2m^2+a^4)]$ ] .
+ therefore , @xmath232 . +
this implies that @xmath233 .
+ we can check that @xmath234 .
+ now , consider the two knots @xmath235 and @xmath236 as in figure [ p11 ] .
+ we can check that they have the same kauffman ( jones ) polynomial as follows .
+ @xmath237 .
+ then @xmath235 is represented by @xmath238 .
+ so , @xmath101=[(a^{10}-2a^8m^2 + 2a^6m^2+m^4a^6 - 2a^4m^4+a^2m^4+a^2m^2-m^2)/a^2,-m(a^2 - 1)(a^6-a^4m^2+a^2m^2 - 1)/a^3,0]$ ] .
+ therefore , @xmath239 .
+ @xmath240 is represented by @xmath241 .
+ so , @xmath101=[(a^6-a^4m^2 + 2a^2m^2-m^2)/a^2,m(a^2 - 1)/a^3,0]$ ] . + therefore , @xmath242 .
+ now , we know that @xmath243 .
+ this implies that @xmath244 . | in this paper , i give a method to calculate the homfly polynomials of two bridge knots by using a representation of the braid group @xmath0 into a group of @xmath1 matrices .
also , i will give examples of a 2-bridge knot and a 3-bridge knot that have the same jones polynomial , but different homfly polynomials . |
in the standard model of elementary particle physics , the @xmath4 symmetry is spontaneously broken to a residual @xmath5 , generating mass for the @xmath6 and @xmath7 gauge bosons and the matter fields . a possible cause for the symmetry breaking is the presence of an additional scalar field , the higgs field .
although there is as yet no experimental evidence for the expected higgs boson , we can still explore the implications of this symmetry breaking mechanism using radiative corrections to standard - model processes .
for example , the condition that perturbative calculations be reliable provides a theoretical upper bound on the mass of a weakly interacting higgs boson @xcite . in a recent paper @xcite , nierste and
riesselmann analyzed one- and two - scale processes involving the higgs field with a particular emphasis on the running of the quartic higgs coupling @xmath3 .
they assessed the reliability of perturbation theory using two criteria : the relative difference of physical quantities calculated in different renormalization schemes ; and the dependence of @xmath1 on the renormalization scale @xmath8 .
if perturbation theory is to be reliable , the choices of the renormalization scheme and scale should not be important for physical quantities . to determine @xmath3 in their analysis , nierste and riesselmann integrated the renormalization group equation using the three - loop @xmath2 function , and solved the resulting equation for @xmath1 iteratively using four different approximation schemes .
the solutions differed significantly for large values of the coupling or mass scale , and determined one constraint on in a perturbative theory .
this uncertainty in @xmath3 carries over to physical quantities such as scattering amplitudes and again affects the ranges of and @xmath8 over which perturbative calculations are reliable .
we show here that is is possible to explore the regime of large coupling without the ambiguities that arise from the direct iterative solution for the coupling .
we approach the problem by emphasizing the @xmath2 function @xmath0 , and show that it can be determined reliably to rather large values of @xmath1 by using pad approximates @xcite to sum the perturbation series for @xmath0 @xcite .
integration of the renormalization group equation then gives an implicit equation for @xmath3 that can be inverted numerically .
the results can be used to study the the validity of perturbation theory for scattering amplitudes in the region of large higgs - boson masses and high energies where the running coupling @xmath3 is the natural renormalization group expansion parameter .
we will not pursue those applications here as a number of authors @xcite have considered them in detail .
we first investigate the pad approach in sec .
[ sec : pade ] using the the results for @xmath0 in the minimal subtraction ( ms ) renormalization scheme for which the perturbation series for @xmath2 is known to five loops @xcite .
after establishing the effectiveness of the padapproach , we apply it in sec .
[ sec : applications ] to the more physical on - mass - shell ( oms ) renormalization scheme where @xmath2 is only known to three loops @xcite .
we find that pad summation of the series apparently gives a reliable result for @xmath0 for quite large values of the coupling , @xmath9 , and conclude , after inversion of the renormalization group expression , that @xmath3 is known reliably in the oms scheme for @xmath10 tev for @xmath11 gev .
the region in which @xmath3 is known well extends to very large mass scales if is sufficiently small , for example , to @xmath12 gev for @xmath13 gev .
in the following , we deal with the quartic higgs - boson coupling @xmath1 defined at tree level in terms of and the electroweak vacuum expectation value @xmath14 gev or the fermi coupling by @xmath15 .
we will work in the interesting limit of large higgs - boson masses , corresponding to the limit of large quartic couplings , and neglect the effects of couplings with fermions .
the running coupling @xmath3 is defined as the solution of the renormalization group equation @xmath16 at the energy scale @xmath8 .
the function @xmath0 is given in perturbation theory as a power series in @xmath1 , @xmath17 the coefficients @xmath18 are renormalization - scheme dependent beyond two loops .
they are known through three loops in the on - mass - shell renormalization scheme @xcite , @xmath19 and to five loops in the minimal subtraction scheme @xcite , @xmath20 & \beta_3=-690759 , \quad \beta_4=4.91261\times 10 ^ 7 .
\label{beta_n_ms}\end{aligned}\ ] ] alternatively , the coefficients @xmath21 are given by @xmath22 @xmath23 & b_3=-0.0073090 , \quad b_4=0.0032917 . \label{b_n_ms}\end{aligned}\ ] ] to determine the running coupling , one must integrate the renormalization group equation , eq .
( [ rgeqn ] ) , and solve the implicit equation @xmath24 this equation determines @xmath3 in terms of the initial and final mass scales @xmath25 and @xmath8 and the initial value of the coupling at the scale @xmath25 , defined as @xmath26 .
different , typically iterative , methods of solution lead to different results for @xmath3 , with the differences increasing for large vales of or @xmath27 and for @xmath28 @xcite .
since the @xmath2 function is only known to finite order , the only constraint on this standard approach is that the different solutions satisfy the renormalization group equation , eq .
( [ beta : int ] ) , to that order . however , the resulting ambiguities for large values of can compromise tests of the reliability of perturbation theory , and the determination of limits on in a weakly interacting theory .
it is therefore useful to approach the problem differently , and concentrate on the @xmath2 function itself . if @xmath0 is known accurately for some range of @xmath1 , the integral in eq .
( [ beta : int ] ) will also be accurately determined , and the equation can be inverted numerically to find @xmath3 in that region .
pad approximates @xcite give a very useful way of summing or extrapolating series for which only a finite number of terms are known .
the @xmath29 $ ] pad approximate for a function @xmath30 defined by a truncated power series @xmath31 is a ratio of two polynomials , @xmath32(z)\equiv\frac{\textstyle{\sum_{n=0}^{n}\,a_nz^n } } { \textstyle{\sum_{n=0}^{m}\,b_nz^n}},\quad b_0=1,\quad n+m = m.\label{pade_define}\ ] ] the coefficients @xmath33 are determined uniquely by the requirement that the series expansion of @xmath34(z)$ ] agree term - by - term with the series for @xmath30 through terms of order @xmath35 .
the sequence of pad approximates @xmath34 $ ] is known to converge to @xmath30 as @xmath36 with @xmath37 fixed for large classes of functions @xcite , but the approximates can also give useful and rapidly convergent asymptotic approximations for finite @xmath38 and @xmath39 even if the sequence and the original series for @xmath30 do not converge @xcite . in the present case ,
the function in question is @xmath0 , known perturbatively to orders @xmath40 and @xmath41 , that is , to three and five loops , in the oms and ms renormalization schemes , respectively .
the perturbation series for @xmath2 is not expected to converge , but a pad summation of the series may still be useful for @xmath1 not too large . because the perturbative expansion of @xmath0 starts at order @xmath42 , we will extract the leading power explicitly , redefine the pad coefficients , and define the @xmath29 $ ] approximate for the n - loop @xmath2 function as was considered by yang and ni @xcite , but without applications to the present problem . those authors did not extract the overall factor @xmath42 , so use a different labeling of the approximates , and miss the diagonal approximates used here . ]
@xmath43=\beta_0\frac{\lambda^2}{16\pi^2}\,\frac{1+a_1\lambda+ a_2\lambda^2+\cdots+a_n\lambda^n } { 1+b_1\lambda+b_2\lambda^2+\cdots+b_m\lambda^m},\quad n+m = n-1 .
\label{eq : p_for_beta}\ ] ] note that the approximates are just the perturbation series for @xmath2 carried to n loops . the series for @xmath0 defined by eq .
( [ b_series ] ) are alternating series in which the ratios of coefficients @xmath44 change only slowly in either oms or ms renormalization in the range in which the @xmath45 s are known .
this suggests that the diagonal approximates with @xmath46 or the subdiagonal approximates with @xmath47 may be particularly effective in estimating the series . in the case of oms renormalization , the @xmath2 s
are known only to three loops , so @xmath48 .
the possible choices are then or if we use all the three - loop information , or if the perturbation series is truncated at two loops . and are just the three- and two - loop perturbation series . in the case of ms renormalization
, @xmath2 is known to five loops , @xmath49 , and we will consider the approximates at the four - loop level , and at five loops , keeping @xmath46 or @xmath47 . the additional five - loop approximates , , and are members of sequences two or more steps off the diagonal .
these are not expected to converge as rapidly as the sequences we consider .
the coefficients @xmath50 , @xmath51 for these approximates are given in appendix a. the fact that the perturbation series for @xmath2 is known to five loops gives us the opportunity to test the pad summation procedure using known results .
having established its reliability , we will the apply the method in sec .
[ sec : applications ] to the more physical oms renormalization scheme in which the connection between @xmath1 and is known . based upon the general convergence properties of padapproximates and the alternating character of the series at hand
, we expect the sequence , and to converge as we progress from three to five loops .
we plot these approximates in fig .
[ fig : ms5 ] to demonstrate that convergence .
the convergence of the padsequence is , in fact , relatively fast . for low values of @xmath1
there is excellent agreement . even for @xmath52 , and differ by @xmath53 10 % with the diagonal five - loop approximate lying roughly halfway between the other two .
we interpret the agreement and the pattern of convergence as strong evidence for the effectiveness of the sequence in summing the series for @xmath0 , and conclude that it is unlikely that @xmath2 would be found to differ significantly from in the region shown if higher - loop contributions were calculated . in fig . [
fig : n0ms ] , we look at the problem from the point of view of the purely perturbative approach , and show the sequence of the n - loop perturbation series for @xmath2 . this is not a sequence in which @xmath38 and @xmath39 increase together with the difference @xmath37 fixed , so the standard results on pad convergence do not apply .
the convergence of the sequence is very slow as shown in the figure , with large differences between successive terms already present for @xmath54 . for comparison
, we also show the three- and five - loop diagonal approximates and .
these forms interpolate the perturbative sequence very well , eliminating the dominance of the last term in the series for @xmath1 large
. since , , and the four - loop approximate differ from each other by less than 5% for @xmath55 , all are effective in extrapolating the perturbation series .
we conclude , in particular , that the three - loop approximate already gives a reliable extrapolation for @xmath0 , with uncertainties of only a few percent , out to @xmath56 , far beyond the range in which the five - loop perturbation series is reliable .
pad approximates often converge to the limit function faster than the power series used to construct them . in that case , the terms in the expansion of a pad approximate beyond the matched order may give reasonable estimates for the unknown higher - order coefficients in the power series .
as a simple test of this expectation in the present case , we can expand the three- and four - loop approximates and to one order higher in @xmath1 than the finite power series used to construct them , and compare the new coefficient with the known four- and five - loop results .
thus , the expansion @xmath57 $ } } } = \frac{\lambda^2}{16\pi^2}\,\beta_0\,\left [ 1+b_1\lambda + b_2\lambda^2 + ( b_2 ^ 2/b_1)\lambda^3 + ( b_2 ^ 3/b_1 ^ 2)\lambda^4 + \cdots \right]\ ] ] gives the estimates @xmath58 for the four- and five - loop coefficients @xmath59 and @xmath60 , results equivalent to @xmath61 the actual four- and five - loop results are @xmath62 the estimates of @xmath63 and @xmath64 from the three - loop coefficients are therefore about 0.67 and 0.36 of the actual coefficients . in the case of , we can estimate only @xmath64 , with the result @xmath65 this estimate gives @xmath66 , and a ratio @xmath67 .
the estimates for the first missing terms in the perturbation series are too small in both of the cases considered
. we can understand this result qualitatively as resulting from the averaging of an alternating series by the approximates , with the corresponding tendency to avoid large higher coefficients in the expansion .
we will use this observation below .
the effects of incorrect estimates of @xmath59 on the approximate are shown in fig .
[ fig : b3 ] . in these calculations , we have taken @xmath59 as five- and ten times the estimated value , and calculated using the new value as input .
the result is a @xmath68% change in @xmath2 for @xmath55 despite the very large values of the new coefficient .
the effect of the uncertainty in @xmath0 on the running of @xmath3 can be studied by integrating the renormalization group equation , eq .
( [ rgeqn ] ) , and solving numerically for @xmath1 as a function of its initial value @xmath27 and the ratio of energy scales @xmath69 .
is connected only indirectly to the physical pole mass of the higgs boson , so we can not state the results in terms of @xmath8 and without using a separate calculation of the self - energy function .
] we have done this calculation using the approximates and , choosing initial values @xmath70 .
the results are shown in fig .
[ fig : msrunning ] .
the result for the optimum five - loop approximate , , lies near the center of the shaded regions in that figure , as would be expected from the comparison of the approximates in fig .
[ fig : ms5 ] .
we believe the estimated range of uncertainty is quite generous given the rapid convergence of the sequence shown there toward .
the range of uncertainty in @xmath3 at fixed @xmath69 is quite small for @xmath71 over the entire range shown , @xmath72 .
the uncertainty is larger for @xmath73 , roughly 16% , at @xmath74 , but even then the boundary curves differ from the curve for by @xmath75% .
the rather small effect of uncertainties in @xmath2 on @xmath3 can be understood rather simply .
the renormalization group equation involves @xmath76 rather than @xmath2 .
the prefactor @xmath42 in the padexpression in eq .
( [ eq : p_for_beta ] ) leads to a rapid decrease in the integrand , and the value of the integral is determined mainly by the region near @xmath27 , the lower endpoint of the integration . for @xmath27 small , @xmath2 is well determined in the most important region , and the uncertainty in the integral is small .
the uncertainty in the integral , hence the uncertainty in @xmath3 , becomes large only for renormalization group evolution away from a large starting value for @xmath27 .
having tested the use of pad approximates in the ms scheme , we consider the implications of pad summation for the oms scheme .
the most significant difference is the limited order , three loops , to which the perturbation series for @xmath2 is known .
we are therefore restricted to two approximates that use the full information available , the diagonal approximate and the subdiagonal approximate .
we can also use at the two - loop level . based upon the convergence of the pad sequence demonstrated for the ms scheme , and the apparent reduction in the size of the coefficients in the oms scheme in the oms scheme
is smaller than that in the ms scheme by roughly a factor of three @xcite .
nierste and riesselmann @xcite have found similar reductions in the coefficients in the expansion of physical amplitudes .
we assume that the reductions in the size of the coefficients persist at higher orders . ]
, we will assume that these approximates again provide an accurate estimate for the @xmath2 function , with the diagonal approximate probably the most reliable . to determine the range of @xmath1 for which the @xmath2 function is reliable
, we first considered the differences among the three - loop functions and , and the two - loop function .
these approximates can barely be distinguished over the range of @xmath1 shown in fig .
[ fig : p11p02 ] with the scale used there , so only is shown .
this agreement is the result of the nearly geometric growth of the first coefficients in the perturbation series .
the three - loop approximates and continue to agree well to much larger values of @xmath1 . while one is tempted on this basis to conclude that the oms @xmath2 function is reliably known for @xmath9 , the range of current interest , the geometric character of the low - order perturbation series may well be accidental . we have therefore attempted to estimate a wider range of uncertainty in the @xmath2 functionin a different way by supposing , in agreement with the results of the ms analysis , that the coefficient @xmath78 estimated by expanding is too small , and constructing a new `` four - loop '' approximate using a greatly increased value of @xmath59 .
the result obtained using @xmath79 is shown in fig .
[ fig : p11p02 ] .
the change in the extrapolation of the perturbation series is quite small , with a difference of less than 2% between and for @xmath55 .
we also show the perturbation series for the @xmath2 function , , in fig .
[ fig : p11p02 ] for comparison . in the oms renormalization scheme , the parameter @xmath1 is defined by the relation @xmath80 to all orders in perturbation theory @xcite .
we will choose the starting value @xmath27 of the running coupling @xmath3 to have this value .
what remains to be decided is the energy scale @xmath25 at which this relation should be taken to hold .
the natural energy scale would appear to be @xmath81 .
however , other choices have been made .
thus , in an early investigation , sirlin and zucchini @xcite calculated the one - loop corrections to the four - point higgs - boson scattering amplitude and defined the parameters in the theory so that large electromagnetic effects appear only in such standard relations as that between and the muon decay rate . with this definition ,
the high mass limit of the four - point function gives @xcite @xmath82 .
\label{eq : sirlin}\ ] ] the logarithm in the expression above is just that which appears in the expansion of the one - loop expression for @xmath3 , @xmath83 for @xmath84 and @xmath81 .
the ambiguity in the choice of @xmath25 is in the treatment of the remaining constants in eq .
( [ eq : sirlin ] ) .
these have been incorporated in the running coupling by some authors@xcite by redefining @xmath25 as @xmath85 $ ] .
however , the constants do not appear naturally in the expression for the four - point function at two loops @xcite .
it is probably most reasonable , therefore , to treat them as separate `` radiative corrections '' and write @xmath86 to one loop as @xmath87 $ ] , with @xmath3 the one - loop running coupling defined above , and @xmath88 incorporating the remaining scale - independent corrections .
this question has been studied in more detail by nierste and riesselmann @xcite , who showed that the convergence of the perturbation series was improved for several physical amplitudes by adopting the natural scale @xmath81 instead of the choice noted above .
they note , furthermore , that in order to cancel large logarithmic terms in the perturbative result when one considers two - scale physical processes such as scattering , the scale @xmath8 must be related to the energy scale of the interaction by @xmath89 @xcite .
we will follow nierste and riesselmann and make the definite , physically motivated choices @xmath81 and @xmath89 in the following analysis .
this specification amounts as already noted to a definite specification of the `` radiative corrections '' in perturbatively calculated amplitudes once the couplings are expressed in terms of @xmath3 . with @xmath27 and @xmath25 specified , and the range of reliability of the @xmath2 function established
, it is straightforward to integrate the renormalization group equation and invert the result numerically to obtain @xmath3 .
the uncertainty in @xmath3 can be specified in terms of that in @xmath2 . with this procedure ,
no further uncertainties such as those illustrated in @xcite are introduced .
it is not necessary , for example , to obtain the solution of the renormalization group equation as a series in @xmath27 , in an iterative approximation .
we note in this connection that the `` naive '' and `` consistent '' forms for @xmath3 given in @xcite correspond respectively to the approximates , the perturbation series for @xmath2 , and , the series obtained by expanding @xmath90 .
neither sequence is expected to converge well with increasing @xmath38 . our results for @xmath3
are shown in fig .
[ fig : lmaxvmu ] for @xmath91 and 800 gev and @xmath92 tev .
we find for @xmath91 gev that all padapproximates , including the perturbation series , agree very for @xmath93 tev , a region in which @xmath94 . the residual uncertainty in @xmath3 is small enough not to affect perturbative results for physical processes .
different pad approximates also give very similar extrapolations for @xmath3 for @xmath95 gev , even when the predicted value of @xmath63 is changed by a large factor .
the only significant deviation involves the perturbation series which we do not believe is reliable on the basis of our earlier investigation . even if we restrict the range of @xmath1 in which we take @xmath2 as reliable to @xmath55 as in fig .
[ fig : p11p02 ] , the result for @xmath3 remains reliable for @xmath96 tev , a value well into the energy region of interest for experiments at the large hadron collider at cern .
the rapid growth of @xmath3 for the perturbative approximate in fig .
[ fig : lmaxvmu ] is the result of an landau pole at @xmath97 gev .
a pole can appear in @xmath3 if the integral of @xmath76 converges for @xmath98 , with the position of the pole in @xmath8 determined by the condition @xmath99 no pole can actually appear when the integration is restricted to the finite range of @xmath1 in which @xmath2 is known reliably , but the likely presence of a pole would be indicated by very rapid growth of @xmath3 with increasing @xmath8 in that region . in the present case , there is no reason to expect the perturbation series to be accurate for @xmath1 large . the results in fig .
[ fig : p11p02 ] indicate , in fact , that the perturbative approximation begins to fail badly for @xmath100 , while the starting point for the evolution of @xmath3 shown in fig .
[ fig : lmaxvmu ] is at @xmath101 for @xmath95 gev .
the remaining approximates do not lead to poles in the region shown .
we have shown that pad summation of the @xmath2 function improves the reliability of @xmath2 and the running quartic higgs coupling .
the method gives a best estimate for @xmath2 , and removes much of the uncertainty associated with different determinations of @xmath3 at the three - loop level @xcite .
we have tested the pad method using the @xmath2 function in the ms renormalization scheme , where @xmath2 is known to five loops in the perturbation expansion .
the test results suggest rapid convergence of the diagonal and subdiagonal pad sequences .
our applications are to the more physical oms renormalization scheme , where the first scheme - dependent coefficient in the oms expansion is significantly smaller than in the ms expansion .
this more rapid apparent convergence is reflected in the excellent agreement among the leading pad approximates for @xmath102 even for rather large values of the first unknown coefficient , @xmath63 .
application of the pad method to the three - loop results for the oms @xmath2 function leads to a running coupling that appears to be quite reliable in the region studied , namely a higgs - boson mass @xmath11 gev and a mass scale @xmath10 tev .
we conclude that uncertainties in @xmath3 are not an important source of uncertainty in perturbative results for physical scattering and decay amplitudes in this interesting region .
this work was supported in part by the u.s .
department of energy under grant no .
de - fg02 - 95er40896 .
one of the authors ( ld ) would like to thank the aspen center for physics for its hospitality while parts of this work were done .
the coefficients of the pad forms used in our analysis are given below .
we will state the results in terms of the coefficients @xmath103 . at two loops , @xmath104 and
we have only the truncated perturbation series and the approximate with @xmath105$}}}:\qquad b_1=-b_1\ ] ] at three loops , @xmath106 and we have the new approximants and .
the coefficients are given by : @xmath107$}}}:\qquad a_1 & = & ( b_1 ^ 2-b_2)/b_1,\label{eq : p11}\\ b_1 & = & -b_2/b_1 , \nonumber \\ & & \nonumber\\ { { \mbox{$\beta\,[{0},\,{2}]$}}}:\qquad b_1 & = & -b_1,\label{eq : p02}\\ b_2 & = & b_1 ^ 2-b_2.\nonumber \end{aligned}\ ] ] at five loops , @xmath108 and we will consider the new approximants , , and .
the coefficients are given by : @xmath109$}}}:\qquad a_1 & = & ( b_1b_3-b_4)/b_3 , \label{eq : p31 } \\ a_2 & = & ( b_2b_3-b_1b_4)/b_3 , \nonumber \\ a_3 & = & ( b_3 ^ 2-b_2b_4)/b_3 , \nonumber \\ b_1 & = & -b_4/b_3 , \nonumber \\ & & \nonumber \\ { { \mbox{$\beta\,[{2},\,{2}]$}}}:\qquad a_1 & = & ( b_1b_2 ^ 2-b_1 ^ 2b_3+b_1b_4-b_2b_3)/a_{22 } , \label{eq : p22}\\ a_2 & = & ( b_2 ^ 3 - 2b_1b_2b_3+b_1 ^ 2b_4+b_3 ^ 2-b_2b_4)/a_{22 } , \nonumber\\ b_1 & = & ( b_1b_4-b_2b_3)/a_{22 } , \nonumber
\\ b_2 & = & ( b_3 ^ 2-b_2b_4)/a_{22 } , \nonumber\\ a_{22 } & = & b_2 ^ 2-b_1b_3 , \nonumber \\ & & \nonumber \\ { { \mbox{$\beta\,[{1},\,{3}]$}}}:\qquad a_1 & = & ( b_1 ^ 4 - 3b_1 ^ 2b_2 + 2b_1b_3+b_2 ^ 2-b_4)/a_{13 } , \label{eq : p13 } \\
b_1 & = & ( -b_1 ^ 2b_2+b_2 ^ 2+b_1b_3-b_4)/a_{13 } , \nonumber \\ b_2 & = & ( b_1b_2 ^ 2-b_2b_3-b_1 ^ 2b_3+b_1b_4)/a_{13 } , \nonumber \\
b_3 & = & ( 2b_1b_2b_3-b_2 ^ 3-b_3 ^ 2+b_2b_4-b_1 ^ 2b_4)/a_{13 } , \nonumber \\ a_{13 } & = & b_1 ^ 3 - 2b_1b_2+b_3 .
\nonumber \end{aligned}\ ] ] we will also consider the approximates , the subdiagonal approximate for the four - loop expansion .
the coefficients in this case are : @xmath110$}}}:\qquad a_1 & = & ( b_1 ^ 3 - 2b_1b_2+b_3)/(b_1 ^ 2-b_2 ) , \label{eq : p12 } \\ b_1 & = & ( b_3-b_1b_2)/(b_1 ^ 2-b_2 ) , \nonumber \\
b_2 & = & ( b_2 ^ 2-b_1b_3)/(b_1 ^ 2-b_2 ) .
\nonumber\end{aligned}\ ] ]
the pad approximates we have used are all integrable analytically .
we will give only the results needed in our investigation of the oms renormalization scheme : @xmath111$}}}:\quad \frac{\beta_0}{16\pi^2}\ , \int^\lambda \frac{d\lambda}{{{\mbox{$\beta\,[{0},\,{1}]$}}}}&= & -\frac{1}{\lambda}-b_1\ln{\lambda } , \\ & & \nonumber \\ { { \mbox{$\beta\,[{2},\,{0}]$}}}:\quad \frac{\beta_0}{16\pi^2}\ , \int^\lambda \frac{d\lambda}{{{\mbox{$\beta\,[{2},\,{0}]$ } } } } & = & -\frac{1}{\lambda } -b_1\ln{\lambda } + \frac{1}{2}b_1 \ln\left(1+b_1\lambda+b_2\lambda^2\right ) \nonumber \\ & & + \frac{b_1 ^ 2 - 2 b_2}{\sqrt{4b_2-b_1 ^ 2 } } \arctan{\frac{b_1 + 2b_2\lambda}{\sqrt{4b_2-b_1 ^ 2}}},\\ & & \nonumber \\ { { \mbox{$\beta\,[{1},\,{1}]$}}}:\quad \frac{\beta_0}{16\pi^2}\,\int^\lambda \frac{d\lambda}{{{\mbox{$\beta\,[{1},\,{1}]$ } } } } & = & -\frac{1}{\lambda}-b_1\ln{\lambda}\nonumber\\ & & + b_1\ln{\left(1 + \frac{b_1 ^ 2-b_2}{b_1}\lambda\right)},\\ & & \nonumber \\ { { \mbox{$\beta\,[{0},\,{2}]$}}}:\quad \frac{\beta_0}{16\pi^2}\,\int^\lambda \frac{d\lambda}{{{\mbox{$\beta\,[{0},\,{2}]$ } } } } & = & -\frac{1}{\lambda } -b_1\ln{\lambda}+\left(b_1 ^ 2-b_2\right)\lambda.\end{aligned}\ ] ] | we apply pad summation to the @xmath0 function for the quartic higgs coupling @xmath1 in the standard electroweak model .
we use the @xmath2 function calculated to five loops in the minimal subtraction scheme to demonstrate the improvement resulting from the summation , and then apply the method to the more physical on - mass - shell renormalization scheme where @xmath2 is known to three loops .
we conclude that the oms @xmath2 function and the running coupling @xmath3 are reliably known over the range of energies and higgs - boson masses of current interest . |
in spite of being one of the simplest possible impurities , a detailed understanding of the physical properties of isolated hydrogen in group - iv materials requires the combination of advanced experimental and theoretical methods.@xcite experimental investigations on atomic , interstitial hydrogen have been so far scarce , since h has turned out to be difficult to observe as an isolated impurity in these materials . in silicon , electron paramagnetic resonance experiments@xcite indicated that hydrogen is located at a bond - center ( bc ) site , midway between two nearest - neighbor host atoms , in agreement with several theoretical calculations.@xcite however , for isolated hydrogen in diamond there is little detailed experimental information available in the literature . in this context
, muonium may be considered as a light pseudoisotope of hydrogen , due to the small mass of the muon @xmath0 ( about 1/9 that of the proton ) .
this means that in many respects muonium is expected to behave similarly to hydrogen , with the appropriate corrections due to zero - point motion and related effects .
thus , muon implantation experiments in diamond as well as several theoretical approaches have shown that this impurity is metastable at a tetrahedral interstitial site ( the so - called ` normal muonium ' ) , and has its lowest energy at or around the bond center site ( ` anomalous muonium').@xcite apart from its basic interest as an isolated impurity , an important property of hydrogen in solids is its ability to form complexes and passivate defects .
this has been extensively studied in the last twenty years for silicon and germanium.@xcite more recently , it has been noted that hydrogen can also passivate impurities in diamond.@xcite over the last years , there has been an intense theoretical activity on hydrogen - related defects in this material.@xcite for isolated hydrogen in the neutral charge state , most of the calculations found the bc site as the lowest - energy position for this impurity in diamond .
some deviations from the bc site were found by saada _
et al._@xcite from a tight - binding ( tb ) calculation .
also , for hydrogen in the positive charge state , density - functional theory calculations have found this impurity to be stable off - axis in a buckled bond - centered configuration.@xcite in standard electronic - structure calculations in condensed matter , despite of their quantum mechanical character , atomic nuclei are treated as classical particles , and therefore typical quantum effects like zero - point vibrations are not directly accessible .
these quantum effects can be important for vibrational and electronic properties of light impurities like hydrogen , especially at low temperatures .
in fact , the importance of taking into account the quantum character of proton and muon to study hydrogen - related point defects in diamond has been emphasized by kerridge _
et al._@xcite in particular , quantum tunneling is relevant to understand the properties of vacancy - hydrogen complexes in this material.@xcite finite - temperature properties of hydrogen - related defects in solids have been studied by _
ab initio _ and tight - binding molecular dynamics ( md ) simulations . in many previous applications of these methods ,
the atomic nuclei were treated as classical particles.@xcite in order to consider the quantum nature of the nuclei , the path - integral molecular dynamics ( or monte carlo ) approach has proved to be very useful .
a remarkable advantage of this method is that all nuclear degrees of freedom can be quantized in an efficient manner , thus permitting the inclusion of both quantum and thermal fluctuations in many - body systems at finite temperatures . in this way ,
monte carlo or molecular dynamics sampling applied to evaluate finite - temperature path integrals allows one to carry out quantitative and nonperturbative studies of highly - anharmonic effects in solids.@xcite in this paper , the path - integral molecular dynamics ( pimd ) method is used to investigate the role of the impurity mass on the properties of hydrogenic point defects .
we consider isolated hydrogen , deuterium ( d ) and muonium ( mu ) in diamond , in their neutral charge state .
special attention has been paid to the vibrational properties of these impurities , by considering anharmonic effects on their quantum dynamics .
the results of the present calculations show that such anharmonic effects lead to a significant deviation of the vibrational frequencies of the impurities , as compared to a harmonic approximation .
also , zero - point motion causes an appreciable shift of the defect levels in the electronic gap , which is a manifestation of the electron - phonon interaction induced by the hydrogenic impurities .
path - integral methods analogous to that employed in this work have been applied earlier to study hydrogen in metals@xcite and semiconductors.@xcite the paper is organized as follows . in sec.ii
, we describe the computational method and the models employed in our calculations .
our results are presented in sec.iii , dealing with the energy of the defects , vibrational frequencies , and defect levels in the electronic gap of diamond .
sec.iv includes a discussion of the results and a summary .
in the path - integral formulation of statistical mechanics , the partition function is evaluated through a discretization of the density matrix along cyclic paths , composed of a finite number @xmath1 ( trotter number ) of `` imaginary - time '' steps.@xcite in the numerical simulations , this discretization gives rise to the appearance of @xmath1 `` beads '' for each quantum particle .
thus , this method exploits the fact that the partition function of a quantum system is formally equivalent to that of a classical one , obtained by replacing each quantum particle by a ring polymer consisting of @xmath1 particles , connected by harmonic springs.@xcite in many - body problems , the configuration space is usually sampled by monte carlo or molecular dynamics techniques .
here , we have employed the pimd method , which has been found to require less computer time resources in our problem .
effective algorithms to perform pimd simulations in the canonical @xmath2 ensemble have been described in detail by martyna _
et al._@xcite and tuckerman.@xcite all calculations presented here were carried out in the canonical ensemble , using an originally developed md software , which enables efficient pimd simulations on parallel supercomputers .
the calculations have been performed within the adiabatic ( born - oppenheimer ) approximation , which allows us to define a potential energy surface for the nuclear coordinates . as in classical molecular dynamics simulations ,
an important issue of the pimd method is the proper description of interatomic interactions , which should be as realistic as possible . to overcome the limitations of effective classical potentials to reproduce the many - body energy surface ,
one has to resort to self - consistent quantum - mechanical methods .
nevertheless , true density functional or hartree - fock based self - consistent potentials require computer resources that would restrict enormously the size of our simulation cell . we have found a reasonable compromise by deriving the born - oppenheimer surface for the nuclear dynamics from an efficient tight - binding effective hamiltonian , based on density functional ( df ) calculations.@xcite the capability of tight - binding methods to simulate different properties of solids and molecules has been reviewed by goringe _
et al._@xcite we have checked the ability of this df - tb potential to predict frequencies of c h vibrations . in particular , for a methane molecule it predicts in a harmonic approximation frequencies of 3100 and 3242 @xmath3 for c
h modes with symmetry @xmath4 and @xmath5 , respectively , to be compared with experimental frequencies of 2917 and 3019 @xmath3.@xcite taking into account the typical anharmonic shift associated to these modes ( towards lower frequencies ) ,
the agreement is satisfactory .
a detailed analysis of vibrational frequencies of hydrocarbon molecules derived with the present df - tb potential , including anharmonicities , can be found elsewhere.@xcite simulations were carried out on a @xmath6 supercell of the diamond face - centered cubic cell with periodic boundary conditions , containing 64 c atoms and one impurity . for comparison , we also carried out simulations of diamond without impurities , using the same supercell size . in our calculations , four symmetry independent * @xmath7
* points in the brillouin zone of the simulation supercell were employed , distributed by following the monkhorst - pack generation scheme.@xcite we have checked the convergence of the internal energy for some selected atom configurations , by considering up to 32 * @xmath7 * points .
in particular , the internal energy of hydrogen on the tetrahedral t site and on the bc configuration changes by about 1 mev , respect to our calculation employing 4 * @xmath7 * points .
sampling of the configuration space has been carried out at temperatures between 100 and 800 k. the simulation - cell parameter employed in our calculations was taken from experimental data@xcite and thus changed from 7.1336 at 100 k to 7.1424 at 800 k. for a given temperature , a typical run consisted of @xmath8 md steps for system equilibration , followed by @xmath9 steps for the calculation of ensemble average properties . to have a nearly constant precision in the path integral results at different temperatures , we have taken a trotter number that scales as the inverse temperature .
at 300 k we took @xmath1 = 20 for h and d , and @xmath1 = 70 for mu .
for comparison with the results of our pimd simulations , we have carried out some classical md simulations with the same interatomic interaction ( setting @xmath1 = 1 ) .
note that for equilibrium properties in the canonical ensemble ( i.e. , mean energy , spatial distribution ) , the classical limit is equivalent in our context to the infinite - mass limit ( @xmath10 ) .
the quantum simulations were performed using a staging transformation for the bead coordinates .
chains of four nos - hoover thermostats with mass @xmath11 were coupled to each degree of freedom to generate the canonical ensemble.@xcite to integrate the equations of motion , we have used the reversible reference system propagator algorithm ( respa ) , which allows one to define different time steps for the integration of the fast and slow degrees of freedom.@xcite the time step @xmath12 associated to the calculation of df - tb forces was taken in the range between 0.2 and 0.5 fs , which was found to be appropriate for the interactions , atomic masses , and temperatures considered here . for the evolution of the fast dynamical variables , that include the thermostats and harmonic bead interactions , we used a smaller time step @xmath13 .
we note that for muonium at 200 k ( the lowest temperature considered for this impurity ) , a simulation run of @xmath9 md steps requires the calculation of forces and energy with the tight - binding code for about @xmath14 configurations , which needs the use of large parallel computers .
important characteristics of impurities in solids are their vibrational frequencies , which are known to be dependent on the particular site occupied by a given impurity in the crystal and on its interactions with the nearby hosts atoms . in our context , the question arises whether the oscillator frequencies associated to an impurity located at a given site can be extracted by assuming the host c atoms fixed in the relaxed geometry .
this is in fact a method frequently employed to calculate vibrational frequencies of impurities in solids . on the other side ,
when all c atoms are allowed to relax by following the impurity motion , the potential energy surface is much flatter ( smaller energy changes ) than when the host atoms are fixed . to obtain an approach for the actual vibrational frequencies of the impurities
, one can calculate the eigenvalues of the dynamical matrix of the whole simulation cell , and obtain the frequencies in the harmonic approximation ( ha ) .
however , for light impurities and atomic configurations displaying large strains ( important relaxations ) , the anharmonicity can be appreciable , making the harmonic frequencies meaningless .
to calculate these anharmonic frequencies we will employ a method based on the linear response ( lr ) of the system to vanishingly small forces applied on the atomic nuclei . with this purpose , we consider a lr function , the static isothermal susceptibility @xmath15 , that is readily derived from pimd simulations of the equilibrium solid , without dealing explicitly with any external forces in the simulation .
this approach represents a significant improvement as compared to a standard harmonic approximation.@xcite a sketch of the method is given in the following .
let us call @xmath16 the set of @xmath17 cartesian coordinates of the beads forming the ring polymers in the simulation cell ( @xmath18 ) .
we consider the set @xmath19 of centroid coordinates , with @xmath20 defined as the mean value of coordinate @xmath21 over the corresponding polymer : @xmath22 then , the linear response of the quantum system to small external forces on the atomic nuclei is given by the susceptibility tensor @xmath23 , that can be defined in terms of centroid coordinates as@xcite @xmath24 where @xmath25 = @xmath26 , @xmath27 is the mass of the nucleus associated to coordinate @xmath21 , @xmath28 is the covariance of the centroid coordinates @xmath20 and @xmath29 , and @xmath30 indicates an ensemble average along a md run .
the tensor @xmath23 allows us to derive a lr approximation to the low - lying excitation energies of the vibrational system , that is applicable even to highly anharmonic situations .
the lr approximation for the vibrational frequencies reads @xmath31 where @xmath32 ( @xmath33 ) are eigenvalues of @xmath34 , and the lr approximation to the low - lying excitation energy of vibrational mode @xmath35 is given by @xmath36 .
more details on the method and illustrations of its ability for predicting vibrational frequencies of solids and molecules can be found elsewhere.@xcite in connection with the vibrational modes that actually appear in our calculations , we note that the application of periodic boundary conditions is physically equivalent to the only consideration of lattice vibrations at the center ( * @xmath7 * = * @xmath37 * ) of the brillouin zone of the simulation cell.@xcite
we first discuss the stable sites for the hydrogenic impurities in diamond , as derived from classical calculations at @xmath38 , i.e. , point nuclei without spatial delocalization . in this respect ,
et al._@xcite noticed the importance of lattice relaxation for finding the most stable site for the impurities , and obtained that interstitial bond - centered hydrogen or muonium is stable as a result of unusually large lattice relaxation .
this is in fact the case for the df - tb model employed here .
after relaxation of the host atoms , the energy surface has minima only at two non - equivalent positions .
the lowest - energy position is the bc site , as in several earlier investigations.@xcite in the relaxed configuration , we find a distance c h of 1.17 , which means a backward relaxation of the adjacent c atoms of 0.40 .
this indicates that the formation of the c h c bridge requires a large dilation of the c c bond ( a 52% of the original one ) .
these values agree with those found earlier from tight - binding calculations of hydrogen in diamond.@xcite another local minimum of the energy surface is found for the impurity at the tetrahedral t site , located 1.44 ev above the absolute minimum . in this configuration ,
the relaxation of the nearest lattice atoms is much smaller than for the bond - center configuration ( 0.08 ) .
there appears in the literature a rather large dispersion of values for the relative energy of h on the t site .
these values range typically from 0.5 to 2.7 ev.@xcite then , our result lies in the intermediate region of these earlier values , and in particular it is close to that reported by kaukonen _
et al._@xcite from a df - tb calculation ( 1.6 ev ) . for the energy barrier from a bc to a t site
we find 1.7 ev , a value between those obtained from local - density functional theory@xcite ( 1.6 ev ) and earlier df - tb calculations@xcite ( @xmath39 ev ) .
for the migration barrier from t to bc sites we have found 0.3 ev , to be compared with @xmath40 ev obtained in ref . .
we note that the so - called anti - bonding site , between an atom and a t site along a @xmath41 direction , has turned out to be from our tb calculations a saddle point on the energy surface , contrary to ref . , where a local minimum was reported .
we now turn to our simulations at finite temperatures .
the internal energy , @xmath42 , at volume @xmath43 and temperature @xmath44 can be written as : @xmath45 where @xmath46 is the potential energy for the classical solid at @xmath38 , and @xmath47 is the vibrational energy . at finite temperatures , @xmath43 changes with @xmath44 due to thermal expansion .
the vibrational energy , @xmath47 , depends explicitly on both @xmath43 and @xmath44 , and can be obtained by subtracting the energy @xmath46 from the internal energy . in this way ,
path - integral molecular dynamics simulations allow us to obtain @xmath47 for a given volume and temperature .
diamond supercell with one impurity , as derived from pimd simulations .
results are shown for deuterium ( circles ) , hydrogen ( squares ) , and muonium ( triangles ) . for comparison , we also present results for a pure diamond supercell ( diamonds ) .
dashed lines are guides to the eye .
error bars are less than the symbol size . , width=321 ] shown in fig.[f1 ] is the temperature dependence of the vibrational energy @xmath48 per simulation cell for hydrogenic impurities at the bc site .
symbols indicate results of pimd simulations : deuterium ( circles ) , hydrogen ( squares ) , and muonium ( triangles ) . for comparison
we also show @xmath48 for pure diamond ( triangles ) . at 300 k , the vibrational energy of diamond amounts to 12.4 ev per simulation cell , i.e. , 0.19 ev / atom .
as expected , @xmath48 increases as temperature is raised , and eventually converges to the classical limit @xmath49 at high @xmath44 .
when we consider the hydrogenic impurities at the bc site , the vibrational energy increases as compared to pure diamond .
this increase is larger the lower the impurity mass , as a consequence of zero - point motion . at this point ,
an interesting characteristic of the different hydrogenic defects is their associated vibrational energy . at a given temperature ,
this energy is defined as the difference @xmath50 = @xmath48(64c + imp )
@xmath48(64c ) , where ` imp ' stands for h , d , or muonium . note that @xmath50 so defined is not just the vibrational energy of a given impurity in the crystal , but it also includes changes in the vibrations of the host atoms . in fig.[f2 ] , the energy @xmath50 is plotted vs temperature for the considered impurities at a bc site . within the numerical precision of our md simulations , @xmath50 turns out to be constant for the three hydrogenic impurities .
this means that at temperatures lower than 800 k we find for these defects , @xmath50 values basically the same as their corresponding zero - temperature limit .
a slight increase of this energy should be expected in the temperature region shown in fig.[f2 ] , especially for h and d ( with vibrational frequencies lower than for mu ) , due to thermal excitation of higher vibrational modes .
however , it seems that this increase is compensated for by a decrease due to lattice expansion , with its associated softening of the vibrational modes .
in fact , we have carried out a pimd simulation of h at a bc site with the lattice parameter corresponding to 200 k ( @xmath51 = 3.5668 ) , and found @xmath50 = 0.40 ev , about 50 mev higher than the value obtained using the actual lattice parameter at 800 k ( @xmath51 = 3.5712 ) .
for comparison with our results of pimd , we also show in fig.[f2 ] results for @xmath50 derived from classical md simulations ( diamonds ) .
these data points follow closely the trend expected for a classical three - dimensional harmonic oscillator : @xmath52 . [
cols="^,^,^,^,^ " , ] in a one - particle harmonic approximation , the low - temperature vibrational energies corresponding to deuterium , hydrogen , and muonium scale as @xmath53 , i.e. , proportional to @xmath54 ( @xmath55 , impurity mass ) . in table
i we give the vibrational energies of the hydrogenic defects and ratios between them . at 300 k , @xmath50 values derived from our pimd simulations scale as @xmath56 , with ratios somewhat different from those for a harmonic approximation . in fact , we have estimated error bars of @xmath57 and @xmath58 for the ratios @xmath59 and @xmath60 , respectively . this is an indication of the anharmonicity present in the defect vibrations , which will be discussed below . by using the same procedure ,
we have calculated the vibrational energy @xmath50 for impurities at the t site .
the results at 300 k are given in table i. the resulting energies are larger than the corresponding ones for the bc configuration . at first sight , this result may seem surprising , taking into account the constrained geometry of the bc configuration , which should give a larger frequency for the stretching vibration along the c h c axis .
however , vibrations transverse to this axis are expected to have much lower frequency , making reasonable a larger @xmath50 for the t defect with threefold degenerate vibrational modes ( see below ) . for the energy ratios at the t configuration we find @xmath61 , which are compatible within error bars with a harmonic approach .
a detailed analysis of the temperature dependence of properties of the impurities at the t site is not feasible due to the metastability of these defects .
the metastability of hydrogen at the t site is illustrated in fig.[f3 ] , where we show the energy of the simulation cell along a pimd run at 600 k. this simulation run started with h at a t site , and after about 15000 steps we observed the passage of the impurity to a bc site .
this is reflected in the figure by an energy jump of about 1.6 ev , which corresponds to the difference between energy minima plus the difference between vibrational energies .
similar results were obtained in other simulations at the same temperature , with hydrogen typically jumping from t to bc positions after a number of md steps in the range from @xmath8 to @xmath9 . however , at 300 k hydrogen remained around the metastable t site during several simulation runs of @xmath9 steps , and the same happened for deuterium and muonium .
we remember that the time scale of a simulation run ( i.e. , @xmath12 times the number of md steps ) does not represent a real physical time , since the dynamics of the beads in the ring polymers does not correspond to the actual dynamics of the quantum particles .
the fictitious bead dynamics is merely used as an efficient way to calculate equilibrium properties of the system in the canonical ensemble .
thus , the plot in fig.[f3 ] illustrates the ability of pimd to explore the configuration space , and to drive the system to the most stable region , once the temperature is high enough . for comparison
, we note that the metastability of muonium at the t site has been observed by muon spin rotation.@xcite in fact , ` isotropic muonium ' ( at t ) converts to ` anisotropic muonium ' ( at bc ) in the range 500700 k. as indicated above , the df - tb potential gives a good description of vibrational modes in hydrocarbon molecules . to check the reliability of this potential to give vibrational frequencies of hydrogenic impurities in diamond , and to assess their anharmonicity , we have carried out calculations for a simple atomic cluster ( see fig.[f4 ] ) . in this cluster ,
the impurity occupies a position similar to a bc site in the diamond lattice , midway between two carbon atoms with a distance @xmath62(c , h ) = 1.17 , as that found in diamond with the df - tb potential . in this atomic cluster
, we have calculated total energies for hydrogen displacements along the c c axis , with both our tb hamiltonian and the b3lyp density - functional theory ( with the cc - pdvz basis set as implemented in the gaussian 98 package ) .
@xcite results obtained by both methods are shown in fig.[f5 ] , where the zero of energy corresponds to the bc position in either case .
the continuous and dashed lines correspond to df - tb and b3lyp models , respectively .
the curvature of the solid line at the minimum is smaller than that of the dashed line , and thus the harmonic approach gives for the df - tb method a vibrational frequency ( 1991 @xmath3 for h ) lower , but of the same order , than that of the b3lyp potential ( 2293 @xmath3 for h ) .
we emphasize that these energy curves do not correspond to hydrogenic impurities in bulk diamond , but they help to analyze the reliability of the df - tb potential and to illustrate qualitatively the vibrational potential of these impurities in the solid . in fig.[f5 ]
we also present the two lowest energy levels for hydrogen , obtained by numerically solving the one - dimensional schrdinger equation with each interaction model .
now the ground state for the b3lyp potential is higher than that of the df - tb model .
the first excited states yielded by both procedures are very close to one another , and in fact are indistinguishable on the scale of fig.[f5 ] .
we then find that the excitation energy @xmath63 corresponds to 2633 and 2577 @xmath3 for df - tb and b3lyp , respectively . in both cases we find excitation energies larger than in the harmonic approximation ,
i.e. , anharmonicity hardens the stretching vibration , as expected for an impurity in a highly - confined geometry .
the excitation energies obtained by both methods are similar , with a relative difference between them smaller than 3% , which gives us confidence in the reliability of the df - tb method to analyze vibrational properties of hydrogenic point defects in diamond . .
the minimum potential energy corresponds to the impurity at the bond - center site .
horizontal lines represent energy levels for hydrogen , obtained by numerically solving the one - dimensional schrdinger equation with those potentials .
results are shown for the tight - binding potential employed here ( solid lines ) and for the b3lyp model ( dashed lines ) .
the given frequencies correspond to first excitation energies in both cases .
, width=377 ] we now go back to the study of hydrogenic impurities in bulk diamond . before analyzing the anharmonic modes given by the linear response procedure described in sec .
ii.b , we will present the vibrational frequencies of the impurities derived from a harmonic approximation
. this will allow us to assess the importance of anharmonicity for the impurity vibrations . by calculating the dynamical matrix ,
we find the vibrational density of states ( vdos ) corresponding to the whole simulation cell ( 64 c atoms plus one impurity ) .
we then use the amplitude of the displacement of the impurity in each vibrational mode , to obtain its partial vdos .
this is shown in fig.[f6 ] for ( a ) deuterium , ( b ) hydrogen , and ( c ) muonium at the bc site .
the results can be most clearly understood for the case of muonium , where one observes basically two peaks at 2221 and 4775 @xmath3 , well separated from the region of crystal vibrations ( up to about 1500 @xmath3 ) . due to the low mass of mu ,
these vibrations are almost totally decoupled from the host - atom vibrations , and are associated to motion perpendicular ( bending , twofold degenerate ) and parallel ( stretching ) to the c c axis , respectively .
this is in fact the situation expected for an impurity at a bc site , assuming fixed host atoms .
the partial vdos is , however , more complicated for h and d , which participate in modes that are not totally localized at the impurity atom or at the impurity plus its nearest c neighbors . for h
, we find only one mode with frequency larger than the lattice modes , corresponding to the stretching vibration along the c h c bond ( @xmath64 1738 @xmath3 ) .
we obtain two other peaks at 448 and 719 @xmath3 with large participation of hydrogen motion .
they are assigned to transverse vibrations , perpendicular to the c h c axis , and include important participation of nearby c atoms . for deuterium ,
the most prominent feature is a large peak at 367 @xmath3 , corresponding to bending vibrations . in this case , the stretching vibration is highly coupled to crystal modes , and does not appear as a prominent peak in the vdos .
next , we apply the linear - response procedure to obtain the ( anharmonic ) vibrational frequencies of the impurities .
again , we obtain the vdos of the whole simulation cell and project from it the contribution of the considered impurity . this partial vdos is shown in fig.[f7 ] for the hydrogenic impurities studied here at 300 k. as before , we first comment on the results for muonium in panel ( c ) .
we observe two distinct peaks , as for the ha , but now at much higher frequencies .
this resembles the effect of anharmonicity observed for the atomic cluster considered above .
for muonium at a bc site in diamond , both vibrations parallel and perpendicular to the c mu c axis harden strongly due to anharmonicity .
something similar happens for hydrogen and deuterium , as noticed when one compares the corresponding panels in figs.[f6 ] and [ f7 ] . for hydrogen
, an interesting feature is the appearance of a second peak at 1559 @xmath3 , above but close to the largest frequency in bulk diamond , corresponding to a bending mode [ fig.[f7](b ) ] . for deuterium
, we obtain a well - resolved stretching mode at 1837 @xmath3 , far in frequency from the lattice vibrations . for the frequencies derived from the lr method
, we estimate error bars of 20 @xmath3 , due to the statistical noise present in the susceptibility tensor @xmath15 derived from the pimd simulations . for the different impurities , the highest - frequency mode @xmath65 scales as @xmath66 , to be compared with
the ratio @xmath53 expected for one - particle harmonic oscillators .
the stretching frequency derived for h from our lr calculations ( @xmath67 2544 @xmath3 ) is lower than that obtained for this impurity by goss _
et al._@xcite from df calculations in the harmonic approximation ( 2919 @xmath3 ) .
this difference can be explained , at least in part , by the difference in the c h distance obtained by both methods : 1.17 for our tb potential vs 1.13 for the df calculations@xcite ( the larger the distance , the lower is expected to be this frequency ) .
we also note that df calculations predict a frequency of 2456 @xmath3 for positively charged hydrogen in a buckled bond - centered configuration.@xcite = 300 k , as derived from a linear - response analysis of pimd results .
each panel corresponds to an impurity : ( a ) deuterium , ( b ) hydrogen , ( c ) muonium .
labels indicate wave numbers in @xmath3 . for clarity ,
discrete modes are presented with a gaussian profile with a standard deviation of 10 @xmath3 .
note the different horizontal scale in panel ( c ) .
, width=377 ] we have repeated the same lr procedure to calculate the vibrational frequencies of hydrogenic impurities at the t site . in this case , we find for each impurity a threefold degenerate mode , in agreement with the tetrahedral point symmetry of this site . coupling with lattice vibrational modes is negligible .
we find the frequencies 2088 , 2818 , and 7284 @xmath3 for d , h , and mu , respectively , which translates to a frequency ratio 0.74:1:2.58 . for impurities at the t site
, anharmonicity causes a softening of the modes . for example , for hydrogen we find in the ha a frequency of 3099 @xmath3 , more than 200 @xmath3 larger than that found with the lr method .
this softening is still more important for muonium , due to its larger zero - point vibration , and is the origin of the small ratio @xmath68 ( to be compared with a ratio of 2.97 expected for harmonic oscillators ) .
we note that anharmonicity affects the mode frequencies in opposite ways at the bc and t sites .
while the vibrational modes associated to the impurities at the bc site suffer a strong shift towards higher frequencies , for the impurities at the tetrahedral site the shift is towards lower ones .
this no doubt is related to the different geometries in both configurations : at the bc site the impurity is in a much more confined environment , with a strongly directional character ( along the bond ) , while at the t site the impurity is less constrained .
the df - tb method employed here allows us to study the influence of zero - point motion on the one - electron levels . for a classical hydrogen - like impurity at a bc site
, we find a defect level located in the electronic gap 2.61 ev above the top of the valence band .
this result is close to the donor level found by goss _
et al . _ from local - density - functional calculations.@xcite also , saada _ et al . _ found a defect level at @xmath69 0.5 ev above the middle of the energy gap for the minimum - energy configuration , which in their case corresponded to the impurity on off - bc sites.@xcite defect levels associated to the presence of the impurity in the crystal are expected to change as a consequence of electron - phonon interaction .
in fact , effects of this kind of interactions on the electronic structure have been experimentally investigated by measuring the temperature dependence of the optical spectra of solids.@xcite these effects appear even at low @xmath44 , due to zero - point motion , and can be large in some cases .
for example , in diamond the direct electronic gap ( at the @xmath70 point ) is reduced by about 0.6 ev with respect to that calculated when zero - point vibrations are neglected.@xcite such studies have been complemented in some cases by observing the dependence of the spectra on isotopic mass , whenever different stable isotopes are available.@xcite similar effects of quantum atomic vibrations should also appear in the defect levels induced by the hydrogenic impurities studied here . shown in fig.[f8 ] is the temperature dependence of the defect level energy for h , d , and mu at a bc site in diamond , as derived from our pimd simulations . for comparison , we also present the energy level @xmath71 derived from classical md simulations . in this figure
, the zero of energy corresponds to the classical limit at @xmath44 = 0 .
the most remarkable fact that one observes in this plot is an appreciable shift of the energy level as the impurity mass changes . in particular , for the impurity at the bc site , the level goes down as the mass is lowered . at 300 k
we find a decrease in @xmath71 of 13 mev when passing from d to h , and a further decrease of 70 mev when comparing muonium with deuterium .
we note also that the level in the case of deuterium is 20 mev lower than that corresponding to a classical impurity ( limit of infinite mass ) at 300 k. in all cases , the statistical error due to finite sampling of the canonical ensemble is of about 2 mev . for a given impurity ,
the defect level goes down as the temperature is raised .
in fact , for h and d we find a decrease in the level energy of about 30 mev when @xmath44 is increased from 100 to 800 k. for muonium , this temperature - induced change amounts to about 20 mev .
for the classical impurity this level shift is more important , resulting to be 63 mev .
the position of the defect level for the different impurities at 300 k is given in table ii , where we also present the energy of the level for impurities at the tetrahedral t site .
in this case , the classical result at 300 k is found to be 0.58 ev above the defect level for the impurity at a bc site . for hydrogenic atoms at site t ,
the level goes up as the impurity mass is lowered , contrary to the case of the bond - center configuration . also , the level for the t configuration is more sensitive to the impurity mass , since it changes by 0.47 ev from a classical impurity to muonium , vs a decrease of 0.10 ev for the bc configuration .
we note that , while the absolute position of the defect level can depend on the model employed , its relative change induced by temperature and isotopic mass is expected to be quite reliable , because in our analysis both electronic and nuclear dynamics are treated at the level of quantum mechanics , which permits a more realistic description of the electron - phonon interaction in this system .
iii we have presented results of our pimd simulations for h , d , and mu in diamond . the main advantage of this kind of simulations is the possibility of calculating defect energies at finite temperatures , with the inclusion of a full quantization of host - atom motions , which are not easy to be accounted for in fixed - lattice calculations .
isotope effects can be readily explored , since the impurity mass appears as an input parameter in the calculations .
this includes the consideration of zero - point motion , which together with anharmonicity causes appreciable non - trivial effects .
our results indicate that hydrogenic impurities at the bond - center site can not be accurately described as a particle moving in a harmonic potential .
even if anharmonicities of the interatomic potential are taken into account , a single - particle approximation does not give reliable results for the impurity complex at finite temperatures .
it is then necessary to treat the defect as a many - body problem with anharmonic interactions .
the most prominent feature in the vdos of the hydrogenic impurities at a bc site is a peak corresponding to the stretching vibration , with a frequency @xmath72 higher than the lattice vibrations .
this stretching mode is infrared - active , but to our knowledge has not been detected yet . on the theoretical side ,
there are not many calculations in the literature predicting the frequency of this mode .
et al._@xcite employed an ha from local - density - functional calculations and found for hydrogen at a bc site ( in the neutral charge state ) a stretching frequency @xmath73 = 2919 @xmath3 , somewhat higher than our result using lr calculations ( @xmath65 = 2544 @xmath3 ) . from pimd simulations
we have calculated in sec . iii.a
the defect energy @xmath50 for the different hydrogenic impurities . also , by using the lr procedure we have obtained an approach to the vibrational frequencies of the impurities , which is more realistic than a pure harmonic approximation .
these frequencies can now be used to estimate a zero - point energy for the impurities in diamond . in the case of muonium ,
the modes are almost totally localized on the impurity , and the zero - point vibrational energy of the defect is well approximated by @xmath74 .
this gives @xmath75 = 0.96 ev , close to the ground state energy of the defect presented in fig.[f1 ] ( @xmath76 ev ) .
using the frequencies @xmath77 and @xmath78 derived from the ha , one finds @xmath79 = 0.57 ev , much lower than that obtained directly from pimd simulations .
note that we have used here an expression for @xmath80 which is only rigorously valid for harmonic oscillators , but it nevertheless gives a good approximation to the zero - point energy of the defect when one introduces the frequencies derived from the lr method . in this sense , these lr frequencies may be considered as renormalized harmonic frequencies , which incorporate in a nonperturbative way anharmonicities of the interatomic interactions , as seen by the atoms in their quantum motion .
the zero - point energy of the defect requires a comment in the sense that , in the case of muonium , it could help to cross the adiabatic barrier for impurity diffusion .
nevertheless , we find that muonium is confined in the bc region , especially at low temperatures .
impurity migration to other regions of the crystal requires important relaxations of the neighboring c atoms , which are not probable at low temperatures .
this picture is similar to that described in the literature as `` opening of a door '' , @xcite which favors impurity diffusion .
an analysis of hydrogen diffusion in diamond is out of the scope of this paper . as noted above
, actual diffusion coefficients are not directly accessible with the kind of md simulations employed here , since the time scale in the simulations is not readily connected to the real one . in connection with this
, pimd simulations can be applied to study quantum diffusion ( including tunneling ) of hydrogen in pure and doped diamond , by calculating free - energy barriers in a way similar to that employed earlier to study h diffusion in metals@xcite and semiconductors.@xcite even though the rate of h tunneling in diamond is not expected to be high at low @xmath44 , due to the large lattice relaxation associated to the impurity on a bc site , thermally activated tunneling may be possible , as observed for hydrogen in silicon.@xcite this point will require further research .
theoretical techniques to calculate the electronic band structure of solids have been improving their precision for many years . for various purposes ,
the accuracy currently achieved by these methods is excellent , when comparing their predictions with experimental data .
however , zero - point motion is a significant factor limiting the accuracy of state - of - the - art techniques to predict energy bands .
the same happens for defect levels caused by impurities in solids , since their energy may change appreciably as the impurity mass is varied .
this effect has been illustrated here by shifts in the energy of the levels due to hydrogenic impurities in bc or t sites in diamond .
for muonium at the t site , we have found a level shift of 0.47 ev ( upwards ) , on the order of the renormalization of the diamond gap due to zero - point motion ( @xmath81 ev ) .
for the bc site , the corresponding shift for mu was 0.10 ev and for h we found 0.03 ev ( both downwards ) .
thus the magnitude and direction of these shifts depend markedly on the mass and position of the impurity in the solid .
this behavior can be rationalized by considering the results of perturbation theory for the electron - phonon interaction in the limiting case of a harmonic oscillator.@xcite the shift of an electronic level is expected to be proportional to the square of the mode amplitude , i.e. , proportional to @xmath54 in the low - temperature limit and to @xmath44 at high temperatures .
the extrapolation to zero temperature of the defect levels for d , h , and mu at a bc site ( see fig.[f8 ] ) gives shifts scaling as @xmath82 , to be compared to the ratio @xmath53 expected for a harmonic impurity .
the energy shift for mu deviates appreciably from the harmonic expectation , as a consequence of the larger anharmonicity of this impurity center . in summary
, the pimd method has turned out to be well - suited to study finite - temperature equilibrium properties of light impurities in diamond .
this has allowed us to notice the importance of anharmonicity in order to give a realistic description of the properties of these point defects .
this anharmonicity shows up in the vibrational modes of the impurities , causing important shifts respect to the harmonic expectancy .
also , zero - point motion introduces an appreciable shift in the defect levels , which depends on the impurity mass .
this type of analysis offers a promising way for studying other challenging effects of light impurities in diamond , such as quantum diffusion .
the calculations presented here were performed at the barcelona supercomputing center ( bsc - cns ) .
this work was supported by cicyt ( spain ) through grant no .
bfm2003 - 03372-c03 - 03 .
erh thanks dursi ( regional government of catalonia ) for funding through project 2005sgr683 . | isolated hydrogen , deuterium , and muonium in diamond have been studied by path - integral molecular dynamics simulations in the canonical ensemble .
finite - temperature properties of these point defects were analyzed in the range from 100 to 800 k. interatomic interactions were modeled by a tight - binding potential fitted to density - functional calculations . the most stable position for these hydrogenic impurities
is found at the c c bond center .
vibrational frequencies have been obtained from a linear - response approach , based on correlations of atom displacements at finite temperatures .
the results show a large anharmonic effect in impurity vibrations at the bond center site , which hardens the vibrational modes with respect to a harmonic approximation .
zero - point motion causes an appreciable shift of the defect level in the electronic gap , as a consequence of electron - phonon interaction .
this defect level goes down by 70 mev when replacing hydrogen by muonium . |
string theory predicts that our world has more dimensions than the 4 we see around us every day . the usual way out of this apparent discrepancy is to say that the additional dimensions are compactified and small . in the simplest case
the extra dimensions are just circles .
string theory is also a theory of gravity .
it is therefore a natural and interesting question to ask what is the behavior of the classical solutions of general relativity in spacetimes where one or more dimensions are compactified .
this question is however surprisingly hard to answer .
the periodic analogue of a schwarzschild black hole in a space of topology @xmath0 has been st udied more or less independently by majumdar @xcite , papapetrou @xcite , myers @xcite , korotkin and nicolai @xcite as well as by frolov and frolov @xcite . in higher dimensions
the full solution is known only implicitly @xcite .
the problem becomes even more interesting when one realizes , following gregory and laflamme @xcite , that there is a competition between different gravitational configurations with the same mass and charges but with different symmetries and even of different horizon topology .
which configuration is stable depends on the particular value of the mass and the charges .
for instance , a long and thin black string becomes unstable and will `` decay '' into a new configuration with higher entropy . this new configuration could for instance be a nontranslationally invariant black string or an infinite array of black holes .
the phase diagram of black objects in higher dimensional circle space times has a very rich structure @xcite .
only parts of it is accessible to analytical methods and the most interesting pices has been calculated only numerically @xcite or using perturbation theory @xcite .
it would therefore clearly be of great interest to have a better analytical understanding of the various solutions that could appear . in this paper
we perform a more modest task .
we generalize the solution given in @xcite to arbitrary electric and nut charge . in doing
so we find a new parameter of the solution which has not been recognized before .
we analyze the physical meaning of this parameter .
we also notice that adding nut charge to the solution makes it more well behaved .
maybe this is a feature which would persist in higher dimensions ?
our method heavily relies on weyl coordinates .
however , as shown in @xcite , because of symmetry reasons , weyl coordinates are expected to be useful at most in 4 and 5 dimensions .
this is something which would have to be overcome if one would like to generalize our results to higher dimensions .
this paper is organized as follows .
section 2 is a general discussion of properties of the four dimensional einstein - maxwells equations important for our problem . in section 3
we discuss the ( nonperiodic ) reissner - nordstrm - nut solution . in section 4
we find its periodic generalization and discuss the metric in various limits . in section 5
we give our conclusions .
we write a general stationary metric with timelike killing vector @xmath1 as @xmath2 where @xmath3 which implies that @xmath4 and that @xmath5 , @xmath6 and @xmath7 are stationary fields , lie dragged by @xmath1 .
we now decompose the electromagnetic field strength @xmath8 by defining the electric and magnetic fields @xmath9 and @xmath10 with respect to @xmath1 : @xmath11 where @xmath12 is the dual field strength @xmath13 it then follows that @xmath14 } - \epsilon_{ab}{}^{cd}b_c\mu_d \\ * f_{ab } & = 2b_{[a}\mu_{b ] } + \epsilon_{ab}{}^{cd}e_c\mu_d.\end{aligned}\ ] ] assuming that @xmath8 is stationary as well as source - free so that @xmath12 is closed , then both @xmath9 and @xmath10 are gradients of stationary scalars .
hence we introduce electric and magnetic potentials @xmath15 and @xmath16 according to @xmath17 moreover , with the electromagnetic field being the only matter source , a twist scalar @xmath18 , likewise stationary , can be introduced by where @xmath19 . ]
@xmath20 the tensor @xmath7 can be viewed as a metric on the three - dimensional manifold @xmath21 of killing orbits and we shall use that the full 4d einstein - maxwell equations are now equivalent to the 3d equations obtained from the action @xmath22,\ ] ] where @xmath23 , @xmath24 and @xmath25 are the ricci scalar , inverse and levi - civita connection of @xmath7 .
this action describes three - dimensional euclidean gravity coupled to a certain sigma model .
more precisely , the effective matter part of the action is that of a wave map from @xmath21 to a four - dimensional target space with metric @xmath26 ^ 2},\ ] ] where @xmath27 and @xmath28 are the complex ernst potentials @xmath29 this is an einstein metric , whose ricci tensor obeys @xmath30 if we use as target space coordinates @xmath31 the real variables @xmath32 , where @xmath33 and @xmath34 are the amplitude and phase of @xmath28 , @xmath35 the target metric takes the neat form @xmath36 which makes it evident that it has signature @xmath37 .
now , varying the action with respect to @xmath24 gives @xmath38 while variation with respect to the target space coordinates @xmath31 leads to the wave map equation @xmath39 with @xmath40 being the christoffel symbols of @xmath41 . for our purposes
it is important to note that if the @xmath31 all depend solely on one free function @xmath42 , then eqs . and become @xmath43 now , eq . tells us that the curve @xmath44 is a geodesic , since ,
as @xmath7 is supposed to be a positive - definite metric , it can not happen that @xmath45 vanishes identically unless @xmath42 is constant .
using the freedom to reparametrize the geodesic , we can assume that it is affinely parametrized and arrive at the equations @xmath46 in this paper we focus exclusively on this class of solutions , which we shall refer to as geodesic solutions .
let us now assume that the stationary spacetime is also axisymmetric with an axisymmetry generator @xmath47 that commutes with the timelike killing vector @xmath1 .
we can then introduce coordinates @xmath48 such that @xmath49 , @xmath50 while the three - metric @xmath7 and one - form @xmath6 take the forms @xmath51 the metric functions @xmath5 , @xmath52 , @xmath53 and @xmath54 obviously depend on @xmath55 and @xmath56 only .
if we assume that @xmath42 also depends solely on @xmath55 and @xmath56 ( although it could have a linear dependence on @xmath57 with constant coefficient , a possibility which we do not consider here ) , the @xmath58-component of eq . becomes @xmath59 making it possible to choose @xmath55 and @xmath56 such that @xmath60 , a choice well - known as weyl ( canonical ) coordinates .
the great advantage of being able to use these coordinates is that the unknown metric function @xmath52 does not enter the laplace equation ( nor the more general wave map equation ) , which will thus be identical to the laplace equation in flat space , expressed in cylindrical coordinates , for an unknown function that is independent of @xmath57 ; @xmath61 with the use of this equation , the remaining components of eq .
become @xmath62 \\ & k_{,z } = 2\epsilon\rho\,\omega_{,\rho}\,\omega_{,z } , \end{split}\ ] ] while eq .
gives @xmath63 with @xmath64 being a constant equal to the conserved geodesic momentum associated with the cyclic coordinate @xmath18 of the target space metric .
explicitly , @xmath65 a recipe for finding a solution of the type considered here is thus to choose an appropriate geodesic of the target space and an appropriate solution to the flat space laplace equation .
the geodesic then gives the four target space functions as functions of @xmath42 , including the metric component @xmath5 and the potentials @xmath15 and @xmath16 which completely determine the electromagnetic field via eqs . and .
what then remains is to integrate the two pairs of equations and to determine the remaining spacetime metric components @xmath52 and @xmath54 .
note that eq .
is the integrability condition for both pairs of equations ( when @xmath66 ) .
using schwarzschild type coordinates , the reissner - nordstrm - nut ( rnn ) solution with mass @xmath67 , electric charge @xmath68 and nut ( gravitational monopole ) charge @xmath69 is @xmath70 where @xmath71 here we have defined @xmath72 while @xmath73 is a constant which should be set to @xmath74 ( @xmath75 ) to make the half - axis @xmath76 ( @xmath77 ) explicitly regular , leaving the other half - axis the misner string singular , as @xmath78 is not a well - behaved one - form at @xmath79 .
however , it is well - known that both half - axes can be made regular , since changing @xmath73 from @xmath74 to @xmath75 can be mimicked by changing the time coordinate from @xmath80 to @xmath81 . the price
to pay is closed timelike curves since it requires that @xmath80 and @xmath82 should both be periodic with period @xmath83 .
the three - metric @xmath7 for this solution is seen to be @xmath84 and the three remaining target space scalars are given by @xmath85 since all the @xmath31 depend on @xmath86 only , this is a geodesic solution . in the non - extremal case @xmath87 ,
the geodesic is spacelike ( @xmath88 ) and explicitly given by @xmath89},\end{aligned}\ ] ] where @xmath90 is the arclength parameter for the geodesic , and here the function that satisfies the laplace equation , namely @xmath91 however , we shall instead think of the geodesic as parametrized by the rescaled affine parameter @xmath92 which implies that the norm of the geodesic tangent vector has norm @xmath93 .
the reason for this is that we will then straightforwardly be able to treat the extremal case which corresponds to a null geodesic ( @xmath94 ) as the @xmath95 limit of the non - extremal case . indeed , taking this limit for the above formulae , we find that the spacelike geodesic goes over into a lightlike one ; @xmath96},\end{aligned}\ ] ] while @xmath42 goes over into @xmath97 this limiting procedure would not have worked , had we used @xmath90 as the affine parameter .
moreover , @xmath42 has the large @xmath86 asymptotic behaviour @xmath98 which means that it is @xmath42 , rather than @xmath90 , that corresponds to a newtonian gravitational potential .
however , although we will think of the geodesic as parametrized by @xmath42 , we will often work with @xmath90 in subsequent calculations , since the factor @xmath99 would otherwise often appear merely as an annoying appendage . for future reference , we here finally calculate the conserved geodesic momentum @xmath64 according to the formula , to find the simple relation @xmath100 a special and somewhat surprising feature of the rnn solution is that although no sources are included for the electromagnetic field , the result of calculating the electric charge @xmath101 by integrating @xmath12 over a two - sphere of constant @xmath80 and @xmath86 and dividing by @xmath102 is not what one would naively expect from stoke s theorem , the result is not independent of the choice of two - sphere but depends on the radius ; @xmath103 we also calculate the magnetic charge @xmath104 by integrating @xmath8 over the same sphere to find @xmath105 obviously we have @xmath106 , @xmath107 asymptotically as @xmath108 , so at infinity the solution is purely electric with charge @xmath68 .
it is worth noting that the relation @xmath109 holds as an identity , for all @xmath86 .
the reason why it is possible that electric and magnetic charges depend on @xmath86 is that the gravitomagnetic vector potential @xmath110 enters not only the metric but also the electromagnetic field strength .
thus the latter is singular on the misner string , on @xmath111 or @xmath112 ( or both ) depending on how @xmath73 is chosen .
as the misner string goes through every two - sphere that we integrate over , we can not expect stoke s theorem to imply that the charges should be independent of the choice of sphere .
interestingly , if we refrain from removing the singularity by introducing periodic time , the misner string must be thought of as a concrete physical object as it carries electric and magnetic charge density . for the rnn solution
, the transformation from schwarzschild type coordinates to weyl coordinates reads @xmath113 with inverse transformation @xmath114 where @xmath115 the functions @xmath42 , @xmath52 and @xmath54 are now given by @xmath116 in the non - extremal case @xmath87 , the function @xmath42 is the potential of an infinitely thin rod located at @xmath117 , @xmath118 and having a line density @xmath119 per unit length , which means that the line density can be interpreted as the mass line density . note that in the extremal case @xmath120 , the function @xmath52 vanishes identically leaving the three - metric @xmath7 flat .
moreover @xmath121 reduces to @xmath122 and @xmath42 to the standard monopole solution @xmath123 .
we will now construct compactified versions of the rnn solution , which generalize the compactified schwarzschild solution discussed in @xcite .
the approach we take can be summerized as follows : * for every member of the rnn family of solutions , compactify the function @xmath42 by using canonical weyl coordinates and taking the @xmath56-periodic analogue of the original solution to the laplace equation .
this is straightforward due to the linearity of the latter . * insert @xmath42 into
the _ same _ geodesic that defines the original rnn solution . *
integrate eqs . and to find the remaining metric functions @xmath52 and @xmath54 . before we start , some general remarks about the first of these steps are in order . as proved by korotkin and nicolai @xcite ,
if @xmath124 is a solution to the laplace equation with the asymptotic behaviour @xmath125 where @xmath126 and @xmath67 is some constant ( in our case the mass ) , then the series @xmath127 , \qquad a_n = \left\ { \begin{array}{ll } \ds\frac{m}{l\abs{n } } & \textrm{if $ n\neq 0 $ } \\ 0 & \textrm{if $ n=0 $ } \end{array}\right . , \ ] ] is convergent for any @xmath128 such that @xmath129 does not coincide with a singular point of @xmath124 for any integer @xmath130 .
the resulting function @xmath42 then obviously defines a @xmath56-periodic solution to the laplace equation with period @xmath131 .
the constants @xmath132 are essential as they make the series converge , but one can not say that they are uniquely determined as one could add to @xmath132 any other @xmath130-dependent constant @xmath133 that falls off faster than @xmath134 as @xmath135 so that @xmath136 is a finite constant . in other words , the @xmath56-periodic analogue of the function @xmath124
is only defined up to an additive constant @xmath137 .
korotkin and nicolai fix this constant by taking eq . as it stands ( @xmath138 ) to define @xmath139 , but this choice has not been made by the other workers that have studied the compactified schwarzschild black hole ; while myers@xcite makes an explicitly different choice of @xmath132 leading to a certain non - zero @xmath137 , frolov and frolov@xcite instead use a green s function method to compactify which in effect corresponds to a third choice of @xmath137 .
now , since the function @xmath139 is periodic in @xmath56 , it does not make sense to study its behaviour for large @xmath140 , but for large @xmath55 it has the asymptotic behaviour @xmath141 if @xmath42 would tend to a constant in this limit , it would be natural to make that constant vanish by a suitable choice of @xmath137 .
however , since @xmath42 instead diverges logarithmically ( unless @xmath142 ) , we see no physical motivation for fixing @xmath137 at any particular value and hence we will keep it as a free parameter .
we will now implement the compactification scheme as outlined above .
we thus take @xmath143 to be the function given by eq . .
according to the above , its periodic analogue is @xmath144 + b,\ ] ] where @xmath145 , \quad z_n = z - nl \\ & a_n = \left\ { \begin{array}{ll } \ds\frac{m}{l\abs{n } } , & \textrm{if $ n\neq 0 $ } \\ 0 & \textrm{if $ n=0$}. \\
\end{array } \right .
\end{aligned}\ ] ] we take @xmath56 to have the range @xmath146 $ ] with the end points of the interval identified .
the horizon , where @xmath42 diverges to minus infinity , will thus be located at @xmath117 , @xmath118 , just as in the non - compactified case .
since we shall not consider the case when the horizon overlaps itself , we require that @xmath147 .
it will prove useful to introduce the dimensionless variable @xmath148 having the range @xmath149 $ ] . for convenience
, we will , for the time being , work with @xmath90 and use the dimensionless coordinates @xmath150 clearly @xmath90 is then given by @xmath151 + \bar{b},\ ] ] where @xmath152 , \quad y_n = y-2n\beta^{-1 } \\ & \bar{a}_n = \frac{\delta}{m}a_n = \left\ { \begin{array}{ll } \ds\frac{\beta}{2\abs{n } } , & \textrm{if $ n\neq 0 $ } \\ 0 & \textrm{if $ n=0 $ } , \\
\end{array } \right .
\\ & \bar{b } = \frac{\delta}{m}\,b.\end{aligned}\ ] ] we shall now explicitly evaluate @xmath90 close to the symmetry axis @xmath153 . as the behaviour on the horizon section @xmath154 of the axis is different from the off - horizon section @xmath155
, these two sections will have to be treated differently , but in both cases we shall express the result in terms of the function @xmath156 having the properties @xmath157 moreover , for reasons that will become clear , we shall replace the free constant @xmath158 by a free constant @xmath159 according to @xmath160},\ ] ] where @xmath161 is the euler constant .
now , for small @xmath162 and @xmath154 , we find that @xmath90 is given by @xmath163 where @xmath164 is the function @xmath165 for @xmath155 , on the other hand , one finds that @xmath166 this is all that is needed to generate the whole solution . indeed , for both axis sectors , @xmath90 is a solution to the laplace equation which can be expanded in terms of a constant @xmath167 and a function @xmath168 as @xmath169 inserting this expansion into the right hand sides of eqs .
and ( using eq .
for the latter ) , we find that @xmath52 and @xmath54 are determined up to additive constants @xmath170 and @xmath73 ; @xmath171x^2 + o(x^4 ) \\ \lbeq{omegaexpx } & \omega = \omega ' + 2l\left[cy + \sum_{k=1}^\infty\frac{2k\,(-1)^k}{(2^kk!)^2}\frac{d^{2k-1}a(y)}{dy^{2k-1}}\,x^{2k}\right],\end{aligned}\ ] ] where we refrain from giving the whole series for @xmath52 since it depends nonlinearly on @xmath90 , although it would not be difficult to write down at least a few more terms .
the value of @xmath170 has to be determined from the requirement that the axis be regular . starting with the off - horizon section ,
regularity means absence of a conical singularity which is in turn means that @xmath52 should vanish as @xmath172 , at least if we assume that @xmath54 vanishes in that limit , which can always be achieved as we shall see below .
thus , since the constant @xmath167 is zero in this case , we directly obtain @xmath173 .
for the horizon section we instead have @xmath174 .
in this case we may determine @xmath170 by inspecting the horizon metric , which reads @xmath175,\ ] ] while @xmath176 is the constant @xmath177 now , since @xmath178 it follows that for the horizon s polar points @xmath179 to be regular , we must set @xmath180 when we are on the horizon @xmath154 , the constant @xmath73 in eq
. precisely corresponds to the constant @xmath73 in eqs . and and
should be set to @xmath74 ( @xmath75 ) to make @xmath54 vanish at the horizon pole @xmath181 ( @xmath182 ) .
continuity at @xmath179 then requires us to set @xmath183 for @xmath184 ( @xmath185 ) and @xmath186 ( @xmath187 ) for @xmath185 ( @xmath184 ) .
again , these two choices do not correspond to different physics , but are related by a change of time @xmath188 which results in time periodicity with the period @xmath83 .
however , note that unlike the non - periodic case , both choices imply that @xmath54 has a jump discontinuity where @xmath189 is periodically identified with @xmath190 .
this discontinuity is of course not of a physical nature either , since we can shift it to any other position on the @xmath191-axis .
for instance , we may put the discontinuity somewhere on the horizon and obtain a completely regular off - horizon section on which @xmath54 is everywhere zero . in principle , the full rnn solution is given as the @xmath153 expansions - , but since these series converge very slowly for large @xmath162 and are quite useless when it comes to determining the asymptotic behaviour as @xmath192 , we shall now follow frolov and frolov and represent the solution in terms of fourier series , thus making explicit use of the periodicity of the coordiante @xmath191 . still using the rescaled
coordinates @xmath193 , we assume fourier expansions for the functions @xmath90 and @xmath52 of the forms @xmath194 since @xmath90 is a solution to the laplace equation , it follows that @xmath195 disqualifying solutions that diverges exponentially as @xmath192 , we obtain that @xmath196 is a linear function of @xmath197 while , for @xmath198 , @xmath199 where @xmath200 is the modified bessel function of the second kind . for small @xmath162
, we have @xmath201 we can now use that we already know @xmath90 for small @xmath162 to find that @xmath202 } + \beta\,\ln{\!\parfrac{\beta\,x}4 } \\ & \omegaal_q(x ) = -\frac2{\pi q}\,\sin{\!(\pi q\beta)}\,k_0{(\pi q\beta x ) } \quad \mbox{for $ q>0$}. \end{aligned}\ ] ] we use this fourier expansion for @xmath90 to plot the metric function @xmath5 for two choices of the parameters @xmath67 , @xmath68 and @xmath69 ( figure [ fig : gravpot ] ) . due to the nonlinearity of eqs . which governs @xmath52
, we shall not attempt to give its full fourier series . however , since all @xmath203 with @xmath198 decay exponentially as @xmath192 , it will suffice to determine @xmath204 to obtain the asymptotic behaviour in that limit .
now , the zeroth order fourier term of the first of eqs
. reads @xmath205_0 = x\left\ { \omegaal_0'(x)^2 + \frac12\sum_{q=1}^\infty \left[\omegaal_q'(x)^2-(\pi q\beta)^2\omegaal_q(x)^2\right]\right\}\ ] ] inserting the determined expressions for the @xmath206 gives @xmath207\right\},\ ] ] which can be integrated to yield @xmath208 \right\},\ ] ] where @xmath209 is some constant yet to be determined .
using now that @xmath210 = \frac2{(\pi k\beta)^2}\left[1+\gamma+\ln{\!\parfrac{\pi k\beta x}2}\right ] + o(x^2),\ ] ] we can compare @xmath204 as given by eq . to the zeroth order fourier term of the exact expression for @xmath52 we have near @xmath153 to find , after some manipulations , that @xmath209 is the @xmath211-dependent constant @xmath212 - \beta\ln{\![\varphi_\beta(2 ) ] } + \int_0^\beta\!\ln{\![\varphi_{\beta'}(2)]}d\beta'.\ ] ] finally , from eqs .
and we obtain that @xmath54 is given by the following `` quasi - fourier '' series : @xmath213 where @xmath73 is the same constant that appears in eq . for the case @xmath154 .
, plotted for @xmath214 , @xmath215 as well as for @xmath216 , with @xmath217 in both cases . as the two top plots show , the behaviour of @xmath5 for the two cases is very similar close to the axis @xmath218 . on
the other hand @xmath5 behaves qualitatively different when a larger range of @xmath55 is included , as indicated by the plots below ; in the first case @xmath5 increases monotonically with @xmath55 towards infinity , while in the second case @xmath5 reaches a maximum and subsequently decreases towards zero .
the latter behaviour is generic as long as @xmath219 , regardless of @xmath68 , but with @xmath220 , the maximum is replaced by a divergence , as discussed in subsection [ sec : large].,scaledwidth=95.0% ] as noted above , the ( outer ) event horizon has the geometry @xmath221 , \quad \mathcal{r } = e^{-\bar{u}}\sqrt{r_+^2+l^2}.\ ] ] it directly follows that the horizon area and gaussian curvature ( half the ricci scalar ) are given by the simple formulae @xmath222 it is worth noting that @xmath164 behaves for small @xmath211 as @xmath223 while for @xmath211 close to unity it behaves as @xmath224(1-\beta)^2 + o((1-\beta)^3),\ ] ] where @xmath225 is the digamma function ( the logarithmic derivative of the gamma function ) .
clearly , for @xmath226 , the metric is an exact two - sphere of radius @xmath227 .
note that @xmath228 corresponds to two independent and physically different limits , namely the extremal limit @xmath229 as well as the limit of infinite coordinate diameter @xmath230 .
this is natural since both limits , makes the _ proper _ distance between the poles of the black hole infinite , keeping the black hole from being distorted `` by itself '' , by the gravitational field from its periodic copies . as @xmath211 is increased , the sphere becomes deformed in a prolate manner . for small @xmath211
the deformation is very small , however , as the first correction term in eq . enters at order @xmath231 . as @xmath211
is further increased the deformed sphere gets stretched out to a long cigar , and for @xmath211 close to unity the geometry is almost everywhere a flat cylinder of length @xmath232 and circumference @xmath233 , except at the poles @xmath179 where the curvature is very high ( note that the mean gaussian curvature is always @xmath234 ) . up to a constant conformal factor
the geometry only depends on the parameter @xmath211 and hence we can refer the reader to frolov and frolov@xcite for a further discussion of its properties , as well as embedding diagrams .
now , the horizon area @xmath235 is clearly modified by a factor @xmath236 compared to the non - compactified case .
this is in agreement with the general treatment of distorted black holes with ( electric ) charge as presented by fairhurst and krishnan@xcite , based on geroch and hartle s treatment of the case without charge@xcite .
although it may not have been obvious from the way we introduced it , the parameter @xmath159 ( denoted the same way in @xcite but with an unbarred @xmath16 in @xcite and @xcite ) can be defined as @xmath237 @xmath159 is the value of @xmath238 evaluated at either of the poles of the event horizon , with @xmath238 being the difference between the function @xmath90 for the deformed black hole ( whether the deformation is due to compactification as in our case , or to external matter ) and the function @xmath239 for the undeformed rnn solution with the same @xmath67 , @xmath68 and @xmath69 . since as @xmath240
, we have @xmath241 it follows that @xmath159 is closely related to the change in horizon pole redshift factor that the deformation produces . in the schwarzschild case @xmath215 ,
the relation is more direct since @xmath242 in that case .
the term `` horizon redshift factor '' must not be taken too literally , however , since it normally refers to an asymptotically flat situation when the value of @xmath243 gives the redshift factor between the point of evaluation and infinity where @xmath244 .
anyhow , we shall refer to @xmath159 as the _ redshift parameter_. in table [ table : ubar ] we display how @xmath159 has in effect been fixed by previous workers that have studied the compactified schwarzschild solution .
as we see no reason why any one of these choices should be better than the other two , we prefer to keep @xmath159 as a free parameter .
other quantities that are interesting to calculate on the horizon are the surface gravity @xmath245 and komar mass @xmath246 with respect to the killing vector @xmath1 , as well as the electric and magnetic charges and potentials .
we find that these quantities are given by @xmath247 we note that the surface gravity is constant over the event horizon and thus the zeroth law of thermodynamics holds .
moreover the electric and magnetic potentials are constant as well .
in fact the electromagnetic charges and potentials take the exact same values as in the noncompactified case .
for a discussion of further thermodynamic properties , we refer the reader to the general framwork for distorted charged black holes@xcite .
unfortunately , to our knowledge no such framework exists in the case of nonzero nut charge , but providing it here would be beyond the scope of this paper .
.the redshift parameter @xmath159 as used in previous studies of the compactified schwarzschild solution .
[ cols="^,^",options="header " , ] a quantity that characterizes the off - horizon section of the symmetry axis is the proper spatial separation between the poles of the black hole .
it can be calculated as @xmath248 with @xmath5 evaluated at @xmath153 . to get a general idea of how this quantity depends on the choice of charges @xmath67 , @xmath68 and @xmath69
, we have held the quotients @xmath249 and @xmath250 fixed at four different values and , with the redshift parameter @xmath159 set to zero , plotted @xmath251 as a function of @xmath211 ( figure [ fig : l_sep ] ) .
since @xmath252 depends nontrivially on how one chooses @xmath159 to depend on @xmath211 , our plot for the case @xmath215 looks different than the one given by frolov and frolov @xcite .
in particular , in @xcite @xmath252 tends to a finite value rather than zero as @xmath253 , which is counter - intuitive since in that limit the polar points reach each other as the event horizon fills the whole axis .
again , this is directly related to the behaviour of @xmath159 in the same limit . however , any choice of @xmath159 which stays _ finite _ results in a vanishing @xmath252 as @xmath253 , just like for @xmath254 . to show the effect of different choices of @xmath159 , we have also plotted @xmath252 as a function of @xmath211 for the different choices of @xmath159 collected in table [ table : ubar ] , with the choice @xmath254
again included for comparison ( figure [ fig : l_sep2 ] ) .
, for @xmath255 and four particular choices of the quotients @xmath249 and @xmath250.,scaledwidth=95.0% ] , for @xmath215 and four different choices of the redshift parameter @xmath159 , three of which being taken from the previous studies of the compactified schwarzschild solution.,scaledwidth=95.0% ] in the limit @xmath256 , our compactified black hole solution approaches an exact solution corresponding to the field of a homogeneous line mass with electric as well as nut line charge .
this solution is obtained simply by truncating all terms that fall off exponentially with @xmath55 , which is easy to do given the fourier expansions derived in subsection [ sec : fourier ] .
explicitly , @xmath257}+\beta\ln{\!\parfrac{\rho}{2l } } \\ & k = k _ * + \beta^2\ln{\!\parfrac{\rho}{\delta } } \\ \lbeq{omegacyl } & \omega = \frac{4lz}{l}+\omega',\end{aligned}\ ] ] so the spacetime metric for our compactified reisser - nordstrm - nut solution , in the nonextremal case @xmath87 , rapidly approaches @xmath258,\ ] ] where @xmath259 ^ 2 + l^2\left[(\rho/\rho_0)^\beta-(\rho/\rho_0)^{-\beta}\right]^2 } \\\ ] ] as before , @xmath260 with @xmath261 , but we have here also introduced the constants @xmath262^{\frac1{2\beta}}2l \\ & \tilde{c } = \left[\frac{\varphi_\beta(2)\,e^{2\bar{u}}}2\right]^{-\frac{\beta}2}e^{\beta^2+\int_0^\beta\ln[\varphi_{\beta'}(2)]d\beta'}.\end{aligned}\ ] ] if we set both the electric charge @xmath68 and nut charge @xmath69 to zero , the function @xmath263 simplifies to @xmath264 which shows that this is a special case of levi - civita s static cylindrically symmetric vacuum solution .
as discussed for instance by bick @xcite , the levi - civita metric contains two essential parameters @xmath265 and @xmath266 and can be written in the standard form @xmath267 where the coordinates @xmath268 , @xmath269 and @xmath270 as well as the _ conicity parameter _ @xmath266 in general carry dimension in a nonstandard way .
the conicity parameter can obviously be set to unity by rescaling the coordinate @xmath57 as @xmath271 , but since we insist that @xmath57 be a periodic coordinate with the standard range @xmath272 , the value of @xmath266 can not be viewed as a coordinate choice .
in fact , since we are dealing with a compactified spacetime for which @xmath56 is also a periodic coordinate with a fixed range , the general static cylindrically symmetric solution what one might call the _ compactified levi - civita solution _
has instead three essential parameters , which for instance can be taken to be our @xmath211 , @xmath273 and @xmath274 , with @xmath211 directly corresponding to @xmath265 . setting the electric charge @xmath68 to zero while having a non - vanishing nut charge @xmath69 makes it possible to reexpress @xmath263 as @xmath275}},\ ] ] where ( note that @xmath276 in this case ) @xmath277 as should be expected , this vacuum solution is the `` cylindrical analogue of nut space '' that was constructed and studied by nouri - zonoz@xcite . according to nouri - zonoz ,
the `` gravitomagnetic charge per unit length '' is ( up to sign ) equal to the coefficient of @xmath56 in the expression for @xmath54 , in our case @xmath278 .
however , what we have constructed here is the asymptotic field of a black hole with nut charge @xmath69 in a space with a compactified dimension of ( @xmath56-coordinate ) length @xmath131 , so we expect the nut charge per unit @xmath56-length to be @xmath279 .
we believe that the discrepancy of a factor of @xmath280 is only a matter of definitions ( nouri - zonoz `` gravitomagnetic charge '' is four times nut charge , when the latter is definied so that its value for the rnn metric is @xmath69 .
setting @xmath281 but keeping a non - zero electric charge @xmath68 , the general asymptotic solution instead becomes the cylindrically symmetric einstein - maxwell solution with a purely electric maxwell field .
this is one of the solutions constructed and studied by richterek @xcite as well as miguelote @xcite .
the function @xmath263 can in this case be written as @xmath282 } } , \qquad \rho_2 = \parfrac{r_+}{r_-}^{\!\!\frac1{2\beta}}\rho_0.\ ] ] note that @xmath263 diverges at a finite cylinder radius @xmath283 , indicating that there should be a curvature singularity there .
this is indeed the case , as noticed in @xcite , but let us see this explicitly by calculating the energy density of the electromagnetic field with respect to the unit timelike vector field @xmath284 where @xmath285 , as before .
the result , in the general case with all three charges @xmath67 , @xmath68 and @xmath69 arbitrary , is @xmath286 clearly @xmath287 diverges at finite @xmath55 exactly when @xmath263 does .
however , the curvature singularity at @xmath283 disappears as soon as a nut charge @xmath69 is turned on , since the denominator of the right hand side of eq .
is then the sum of two squares which can not vanish simultaneously .
of course , one can make the curvature grow arbitrarily large close to @xmath283 by choosing the quotient @xmath288 sufficiently small .
this feature is not only present in the asymptotic solution discussed here , but also in the full compactified rnn solution with the @xmath56-dependent fourier terms turned back on .
the reason for this is that the range of the function @xmath42 changes from @xmath289 to @xmath290 when we compactify the spacetime by making the coordinate @xmath56 periodic ; in both cases @xmath291 occurs at the event horizon , at @xmath117 , @xmath118 , but @xmath256 sends @xmath42 to @xmath292 rather than zero in the compactified case . in particular , for @xmath293 , the compactified @xmath90 will take on the positive value @xmath294 for which @xmath263 diverges , as can be seen from eq . .
again , this gives a curvature singularity which however is regularized by turning on a nut charge @xmath69 , however small . the difference , compared to the asymptotic solution , is that the curvature singularity now occurs on some torus @xmath295 which is not a simple flat torus @xmath296 .
the asymptotic solution also has a curvature singularity on the axis @xmath117 for @xmath297 , but this is in contrast to the full compactified rnn solution which by construction is regular on that axis .
another interesting feature related to the nut parameter @xmath69 is that although the spacetime metric is cylindrically symmetric in the sense that no physically measurable quantities depend on the @xmath56 , this coordiate still appears in the expression for @xmath54 .
what is interesting is that if we demand that the metric be explicitly periodic in @xmath56 with period @xmath131 , then the metric must be unchanged under @xmath298 for any integer @xmath130 .
this is true , granted that @xmath80 is periodic with period @xmath83 ( just like for the non - compactified rnn solution we started out with ) , since we can mimick the shift in @xmath73 by letting @xmath299 .
turning finally to the extremal case @xmath300 , we must use @xmath301 instead of @xmath90 before taking the limit @xmath229 .
the limit then gives @xmath302 , \qquad u ' = \lim_{\beta\rarr 0}\beta^{-1}\bar{u } \\ & k = 0 \\ & \omega = \frac{4lz}{l } + \omega',\end{aligned}\ ] ] so the asymptotic spacetime metric is in this case @xmath303 with , using eq .
, @xmath304 ^ 2+\left[\dsfrac{2l}{l}\,\ln{\!(\rho/\rho_3)}\right]^2 } , \qquad \rho_3 = 2le^{-u'-\gamma}.\ ] ] just like in the nonextremal case , this asymptotic solution fails to be asymptotically flat , albeit in a weaker manner .
also , the discussion concerning the curvature singularity at finite @xmath55 applies here as well . in particular , from eq .
it follows that if @xmath281 , the asymptotic solution has a curvature singularity at @xmath305 , but the singularity is removed when the nut charge @xmath69 is turned back on .
having constructed and analyzed what we consider to be the natural periodic analogue of the reissner - nordstrm - nut solution , a question that naturally arises is whether or not any given asymptotically flat black hole has a periodic analogue and , if so , whether or not the periodic analogue is unique , given the period @xmath131 of the compactified coordinate . by identifying the target space geodesic that the non - compactified solution defines and assuming that the very same geodesic should be used for the compactified version
, we have found that the problem essentially reduces to finding the appropriate axisymmetric solution to the laplace equation in flat space , just as in the schwarzschild case .
there is thus a simple superposition principle at play , which makes compactification tractable when weyl coordinates are used .
however , the low dimension requires a regularization procedure which introduces a one - parameter ambiguity as there is no obvious way to fix the freedom to add an arbitrary constant to the solution to the laplace equation . in all previous works on the periodic schwarzschild solution , this constant the redshift parameter
was always fixed at an early stage and without any physical motivation for the specific choice made .
as we have seen there are three different choices in the existing papers on the subject , which we take as support for instead keeping the parameter free , or at least fixing it at a later stage from physical considerations .
one way of fixing it would be to require that a set of relations that hold for horizon quantities ( for instance the relation between the horizon area and the black hole charges ) should remain the same after compactification .
if one could motivate why this should be the case , our method would give a neat one - to - one correspondance between standard and periodic reissner - nordstrm - nut black holes .
of course , we have not given any proof that our method is correct , in some suitably defined sense .
one could perhaps argue that it is not , based on the fact that our solution family has a curvature singularity surrounding the horizon in the case of electric but no nut charge .
however , we see no way of avoiding such a singularity .
indeed , the singularity in question is actually a general feature not just for our full solution with a nontrivial dependence on the periodic coordinate , but also for the cylindrically symmetric solution that is approached for large cylinder radii .
since there simply is no other cylindrically symmetric solution available that could work as an asymptotic solution for a compactified reissner - nordstrm black hole , it seems at least likely that the exponentially decaying corrections to the asymptotic solution can not smooth out the singularity .
one can think of several ways to extend our work while staying in four dimensions .
one way would be to include a dilaton field with arbitrary coupling parameter .
this would be straightforward to do , as our method would still be applicable with the dilaton simply entering as an extra target space coordinate
. it would be interesting to see whether the inclusion of a dilaton could prevent the above - mentioned singularity from occuring .
a general framework for distorted charged dilaton black holes has been provided by yazadjiev@xcite .
it would be even more interesting if the periodic analogue of the kerr solution could be constructed .
some indications of how this could be done were given in @xcite , but no explicit formulae were given there , except for a few first steps .
since the noncompactifed kerr solution traces out a target space two - surface rather than a geodesic , it would no longer be possible to linearize the problem and thus more sophisticated methods , such as bcklund transformations , are needed .
it would of course be even more interesting if our work could give some insights into how things work in higher dimensions .
consider the einstein - maxwell equations for a static @xmath306-dimensional spacetime .
dimensionally reducing with respect to the static killing vector , one arrives at einstein gravity in @xmath307 euclidean dimensions , coupled to a two - dimensional ( three - dimensional if a dilaton is included ) sigma model .
one question one could ask is whether or not the target space geodesic that the @xmath306-dimensional reissner - nordstrm solution corresponds to would also be traced out when the solution is compactified on a circle .
the work of harmark and obers actually suggests that this could be the case , as their ansatz is such that the target space coordinates are functionally dependent , thus corresponding to geodesic solutions .
this does not mean that it would be easy to make analytical progress for @xmath308 , but nevertheless it could give a coordinate invariant way of thinking about the ansatz of these authors which in turn could lead to a deeper understanding of the general problem .
it would be very interesting to investigate this matter further .
the work of rvu was supported by the ministry of education of the czech republic under the project msm 0021622409 .
the work of mk was supported by a post doctoral fellowship of the faculty of science of masaryk university .
r. c. myers , `` higher dimensional black holes in compactified space - times , '' phys .
d * 35 * ( 1987 ) 455 .
d. korotkin and h. nicolai , `` a periodic analog of the schwarzschild solution , '' arxiv : gr - qc/9403029 .
d. korotkin and h. nicolai , `` the ernst equation on a riemann surface , '' nucl .
b * 429 * ( 1994 ) 229 [ arxiv : gr - qc/9405032 ] .
a. v. frolov and v. p. frolov , `` black holes in a compactified spacetime , '' phys .
d * 67 * ( 2003 ) 124025 [ arxiv : hep - th/0302085 ] .
r. gregory and r. laflamme , `` black strings and p - branes are unstable , '' phys .
* 70 * ( 1993 ) 2837 [ arxiv : hep - th/9301052 ] .
r. gregory and r. laflamme , `` the instability of charged black strings and p - branes , '' nucl .
b * 428 * ( 1994 ) 399 [ arxiv : hep - th/9404071 ] .
g. t. horowitz and k. maeda , `` fate of the black string instability , '' phys .
. lett .
* 87 * ( 2001 ) 131301 [ arxiv : hep - th/0105111 ] .
t. harmark and n. a. obers , `` black holes on cylinders , '' jhep * 0205 * ( 2002 ) 032 [ arxiv : hep - th/0204047 ] . t. harmark and n. a. obers , `` new phase diagram for black holes and strings on cylinders , '' class . quant . grav .
* 21 * ( 2004 ) 1709 [ arxiv : hep - th/0309116 ] .
t. harmark and n. a. obers , `` phase structure of black holes and strings on cylinders , '' nucl .
b * 684 * ( 2004 ) 183 [ arxiv : hep - th/0309230 ] .
t. harmark `` small black holes on cylinders , '' phys .
d * 69 * ( 2004 ) 104015 [ arxiv : hep - th/0310259 ] .
h. elvang , t. harmark and n. a. obers , `` sequences of bubbles and holes : new phases of kaluza - klein black holes , '' jhep * 0501 * ( 2005 ) 003 [ arxiv : hep - th/0407050 ] . s. s. gubser , `` on non - uniform black branes , '' class .
. grav .
* 19 * ( 2002 ) 4825 [ arxiv : hep - th/0110193 ] .
b. kol , `` topology change in general relativity and the black - hole black - string transition , '' arxiv : hep - th/0206220 .
b. kol and t. wiseman , `` evidence that highly non - uniform black strings have a conical waist , '' class .
grav . * 20 * ( 2003 ) 3493 [ arxiv : hep - th/0304070 ] .
b. kol , e. sorkin and t. piran , `` caged black holes : black holes in compactified spacetimes .
i : theory , '' phys .
d * 69 * ( 2004 ) 064031 [ arxiv : hep - th/0309190 ] .
e. sorkin , b. kol and t. piran , `` caged black holes : black holes in compactified spacetimes .
ii : 5d numerical implementation , '' phys .
d * 69 * ( 2004 ) 064032 [ arxiv : hep - th/0310096 ] .
d. gorbonos and b. kol , `` a dialogue of multipoles : matched asymptotic expansion for caged black holes , '' jhep * 0406 * ( 2004 ) 053 [ arxiv : hep - th/0406002 ] .
d. gorbonos and b. kol , `` matched asymptotic expansion for caged black holes : regularization of the post - newtonian order , '' arxiv : hep - th/0505009 .
t. wiseman , `` static axisymmetric vacuum solutions and non - uniform black strings , '' class .
* 20 * ( 2003 ) 1137 [ arxiv : hep - th/0209051 ] .
t. wiseman , `` from black strings to black holes , '' class .
* 20 * ( 2003 ) 1177 [ arxiv : hep - th/0211028 ] .
h. kudoh and t. wiseman , `` properties of kaluza - klein black holes , '' prog .
* 111 * ( 2004 ) 475 [ arxiv : hep - th/0310104 ] .
h. kudoh and t. wiseman , `` connecting black holes and black strings , '' phys .
* 94 * ( 2005 ) 161102 [ arxiv : hep - th/0409111 ] .
r. emparan and h. s. reall , `` generalized weyl solutions , '' phys . rev .
d * 65 * ( 2002 ) 084025 [ arxiv : hep - th/0110258 ] .
s. fairhurst and b. krishnan , `` distorted black holes with charge , '' int .
j. mod .
d * 10 * ( 2001 ) 691 [ arxiv : gr - qc/0010088 ] .
r. geroch and j. b. hartle , `` distorted black holes , '' jour .
* 23 * ( 1981 ) 680 j. bicak , t. ledvinka , b. g. schmidt and m. zofka , `` static fluid cylinders and their fields : global solutions , '' class .
* 21 * ( 2004 ) 1583 [ arxiv : gr - qc/0403012 ] .
l. richterek , j. novotn and j. horsk , `` new einstein - maxwell fields of levi - civita s type , '' czech .
j. phys .
* 50 * ( 2000 ) 925 [ arxiv : gr - qc/0003004 ] .
a. y. miguelote , m. f. a. da silva , a. wang and n. o. santos , `` levi - civita solutions coupled with electromagnetic fields , '' class .
* 18 * ( 2001 ) 4569 [ arxiv : gr - qc/0104018 ] . m. nouri - zonoz , `` cylindrical analogue of nut space : spacetime of a line gravomagnetic monopole , '' class .
quant . grav .
* 14 * ( 1997 ) 3123 [ arxiv : gr - qc/9706015 ] . s. s. yazadjiev , `` distorted charged dilaton black holes , '' class .
* 18 * ( 2001 ) 2105 [ arxiv : gr - qc/0012009 ] . | we construct and investigate a compactified version of the four - dimensional reissner - nordstrm - nut solution , generalizing the compactified schwarzschild black hole that has been previously studied by several workers .
our approach to compactification is based on dimensional reduction with respect to the stationary killing vector , resulting in three - dimensional gravity coupled to a nonlinear sigma model .
using that the original non - compactified solution corresponds to a target space geodesic , the problem can be linearized much in the same way as in the case of no electric nor nut charge . an interesting feature of the solution family is that for nonzero electric charge but vanishing nut charge , the solution has a curvature singularity on a torus that surrounds the event horizon , but this singularity is removed when the nut charge is switched on .
we also treat the schwarzschild case in a more complete way than has been done previously . in particular
, the asymptotic solution ( the levi - civita solution with the height coordinate made periodic ) has to our knowledge only been calculated up to a determination of the mass parameter .
the periodic levi - civita solution contains three essential parameters , however , and the remaining two are explicitly calculated here . |
the study of planetary nebulae ( pne ) has evolved enormously over the last half century , largely through the developing of new observational techniques .
however , some fundamental subjects within the field have changed only slightly .
outstanding among these latter subjects are distances and the three dimensional ( 3-d ) structure of individual pne .
unfortunately , only a few pne are close enough for accurate determination of trigonometric parallaxes @xcite and the determination of a 3-d model from high velocity slit spectroscopy has largely been limited to a narrow range of relatively high surface brightness but well resolved pne , c.f .
sabbadin et al .
( 2006 ) and references within . in the case of the closest classical pn ( ngc 7293 , the helix nebula )
there is an accurate trigonometric distance and in a series of papers utilizing imaging and spectroscopy a quadrapolar model was derived @xcite .
another classical nebula , ngc 6720 , the ring nebula , was the subject of 3-d modeling @xcite through extensive spectroscopic mapping , but in this case the trigonometric parallax distance of 700@xmath0 pc @xcite is quite uncertain , as are some properties of the central star .
fortunately , if a good kinematic 3-d model of a pn is available , measurements of tangential motions can be used to determine distances .
this has been done for a large number of pne @xcite using vla observations of the radio continuum with spatial resolutions of about 1 , but the necessary supporting kinematic models are usually not available and a defacto spherical model was assumed in deriving most of the distances .
there is also the uncertainty resulting from the radio continuum observations sampling a different portion of the nebula than that studied in the optical emission lines that were used for deriving expansion velocities and the problems associated with the fact that radial velocities measure bulk motion of the gas whereas apparent expansion also includes the effect of the changing position of the ionization front .
odell et al . ( 2007b ) determined the 3-d structure of ngc 6720 from optical line spectroscopy within the assumption of homologous ( resembling ballistic ) expansion , a general characteristic of classical pne known for more than half a century @xcite even though such a general expansion characteristic is difficult to theoretically understand @xcite .
a recent study @xcite demonstrates the errors that can arise from an incorrect assumption of homologous expansion , but unfortunately the probable errors in our current study do not allow questioning that assumption in any detail . in the odell et al .
( 2007 ) study it was determined that the nebula was a triaxial ellipsoid seen nearly pole - on , with most of the mass confined to an elliptical ring lying almost in the plane of the sky .
the basic correctness of the model was confirmed from application of a very different method of 3-d modeling in a study by steffen et al .
( 2007 ) , which also added modeling of the outer halos of the ring nebula . however , the conclusions drawn from the radial velocity study were not as secure as one would hope because radial velocities of a pole - on ellipsoid with most of the material nearly in the plane of the sky are not sensitive ways of mapping where the material is located and of the material s spatial velocity .
this is the underlying reason for formulating an observing program that will add information about the tangential velocities . in this article
we report on a study of the ring nebula that employed imaging observations of an order of magnitude higher spatial resolution than the vla studies of other pne , but comparable or longer time baselines .
moreover , we utilized the same optical lines for both spectroscopy and imaging .
the imaging observations and their analysis are described in 2.1 and 3 .
new spectrophotometric observations are described in 2.2 and their use for determining corrections to the apparent tangential expansion velocities is described in
4 . these combined data , together with the results of our earlier spectroscopic study @xcite , are the basis of the discussion presented in
we obtained new hubble space telescope ( hst ) wfpc2 images as parts of programs go 11231 and go 11232 .
the go 11231 observations were in the f469n ( he ii ) , f547 m ( continuum ) , and f673n ( [ s ii ] ) filters and executed on 2008 april 23 .
triple exposures of 60 , 200 , and 260 seconds respectively were combined to provide high signal to noise ratio images free from cosmic ray events .
the go 11232 observations were made in the f658n ( [ n ii ] ) and f502n ( [ o iii ] ) filters on 2008 may 6 . again
triple exposures of 260 seconds were made in each filter and similarly combined using iraf tasks .
identical pointing was employed with the central star of the nebula centered in ccd3 of the wfpc2 , duplicating as close as possible the conditions of earlier observations in f502n and f658n in program go 7632 , except that the current two - gyro pointing restrictions of the hst caused us to adopt a nearly reciprocal position angle ( p.a . ) of 213.2 , while the go 7632 observations were made with p.a.=33.8 .
the interval between the go 7632 and go 11232 observations in the f502n and f658n filters used for astrometry is 9.557 years .
the go 11232 images were aligned with the geomap and geotran iraf tasks using four stars , including the central star .
the alignment of the central star was corrected with the imshift task to an accuracy of better than 0.02 , leaving the possibility of a poorer alignment accuracy in rotation for objects further out in the nebula , but eliminating any effects of the other field stars not moving in the same manner as the central star .
the ring nebula has been the target of many spectroscopic programs at various spectral resolutions , the latest and most comprehensive high resolution study being that of odell et al .
( 2007b ) , while recent moderate resolution studies are those of guerrero , et al .
( 1997 ) , garnett & dinerstein ( 2001 ) , and liu et al ( 2004 ) .
moderate resolution spectroscopy offers the opportunity to measure the most useful diagnostic emission lines , but previous studies have not combined study of the critical position angles ( the major axis points towards p.a.=60 and the minor axis points towards p.a.=150 ) and full wavelength coverage .
we have closed this information - gap by making new observations with the boller and chivens spectrograph mounted on the 2.12 m telescope at the san pedro martir observatory .
it was expected that these observations would also provide ground based telescope reference data for calibrating the f469n and f673n filters of the hst s wfpc2 , which meant that we used an unusually wide slit ( in order to diminish the effects of comparing high and low spatial resolution data ) .
we employed a 300 @xmath1 m wide entrance slit ( 3.93 ) .
the scale of the site ccd was 1.05 pixel along the slit and 3.05 / pixel along the dispersion direction .
the full width at half maximum signal ( fwhm ) for the final combined spectra along the direction of the dispersion was 9.4 .
observations were made at p.a.=150 on the night of 2008 july 25 and at p.a.=60 on the night of 2008 july 26 under photometric conditions .
the flux standard bd+25 3941 in the iraf database was used for both nights .
various exposure times were used in order to avoid the problems of the bright lines saturating on the longer exposures . for p.a.=150
we obtained seven exposures of 180 seconds , five of 600 seconds , and four of 1200 seconds for a total exposure time of 9060 seconds .
for p.a.=60 we obtained three exposures of 180 seconds , three of 600 seconds and two of 1200 seconds for a total exposure time of 4740 seconds .
the seeing fwhm was about 2.3 on both nights .
twilight sky exposures were made on the evening of the p.a.=150 observations and were used for characterizing the small flat field corrections .
the data were reduced using iraf tasks , which included flat field correction , distortion correction , wavelength calibration , and rendering into cgs system surface brightness units ( ergs s@xmath2 @xmath3 @xmath2 steradian@xmath2 ) .
the spectra were combined to form a single spectrum for each position angle , with only the shorter exposure time observations used for the brightest emission lines .
digital versions of these spectra are available in the electronic form of this paper .
lllllllll 4026 & hei , heii & 0.29 & 2.33 & 2.21 & 1.59 & 2.25 & 2.86 & 2.66 + 4068 & [ s ii ] * & 0.28 & 4.42 & 5.58 & 3.96 & 4.71 & 5.58 & 2.40 + 4102 & h@xmath4 & 0.27 & 26.2 & 26.1 & 23.8 & 25.4 & 27.5 & 23.1 + 4267 & o ii & 0.19 & 0.73 & 0.71 & 1.79 & 0.74 & 0.82 & 1.39 + 4340 & h@xmath5 & 0.16 & 43.0 & 43.2 & 44.2 & 43.1 & 44.9 & 43.9 + 4363 & [ o iii ] & 0.16 & 7.93 & 7.50 & 9.64 & 6.93 & 6.32 & 9.59 + 4471 & hei & 0.12 & 4.18 & 4.29 & 2.86 & 4.62 & 4.73 & 2.82 + 4540 & heii & 0.11 & 0.63 & 0.56 & 1.69 & 0.35 & 0.42 & 1.65 + 4686 & heii & 0.04 & 16.5 & 15.2 & 49.3 & 11.6 & 11.6 & 45.5 + 4711 & [ ar iv ] * & 0.02 & 1.62 & 1.31 & 4.52 & 1.04 & 1.04 & 4.15 + 4739 & [ ar iv ] * & 0.02 & 0.95 & 0.82 & 2.98 & 0.52 & 0.48 & 2.79 + 4861 & h@xmath6 & 0.00 & 100.0 & 100.0 & 100.0 & 100.0 & 100.0 & 100.0 + 4922 & hei & -0.02 & 2.58 & 1.62 & 0.84 & 1.35 & 1.32 & 1.00 + 4959 & [ o iii ] & -0.02 & 405.5 & 260.6 & 361.4 & 376.7 & 380.2 & 397.2 + 5199 & [ n i ] & -0.10 & 6.90 & 7.50 & 0.71 & 8.81 & 8.87 & 0.80 + 5411 & heii & -0.14 & 1.85 & 1.60 & 5.08 & 1.21 & 1.22 & 4.55 + 5517 & [ cl iii ] & -0.17 & 1.03 & 1.07 & 0.62 & 0.87 & 0.91 & 0.57 + 5537 & [ cl iii ] & -0.17 & 0.82 & 0.86 & 0.47 & 0.70 & 0.74 & 0.43 + 5577 & [ o i ] & -0.19 & 0.35 & 0.35 & - & 1.21 & 0.13 & 0.57 + 5593 & o iii & -0.19 & - & - & 0.022 & - & - & 0.021 + 5755 & [ n ii ] & -0.21 & 5.86 & 6.30 & 0.90 & 6.35 & 6.30 & 1.09 + 5876 & hei & -0.23 & 12.4 & 12.6 & 8.07 & 13.1 & 13.1 & 7.80 + 6300 & [ o i ] & -0.31 & 29.58 & 31.4 & 2.00 & 36.8 & 37.2 & 2.70 + 6363 & [ o i ] & -0.31 & 9.86 & 10.6 & 0.63 & 12.3 & 12.4 & 0.86 + 6548 & [ n ii ] & -0.35 & 136.8 & 151.0 & 21.1 & 156.0 & 156.0 & 23.4 + 6563 & h@xmath7 & -0.35 & 292.4 & 294.4 & 288.4 & 283.8 & 284.4 & 269.2 + 6583 & [ n ii ] & -0.35 & 421.7 & 458.1 & 64.3 & 469.9 & 469.9 & 69.0 + 6678 & hei & -0.37 & 4.00 & 4.13 & 2.96 & 5.06 & 5.04 & 3.29 + 6717 & [ s ii ] & -0.38 & 26.3 & 29.7 & 5.48 & 24.3 & 24.4 & 5.12 + 6731 & [ s ii ] & -0.38 & 29.1 & 32.9 & 6.07 & 26.9 & 27.0 & 5.66 + lcllccll pa150top & 11 31 & 1.01 & 1.01 & 0.81 & 0.81 & 600 & 860 + pa150bot & -14 -36 & 1.15 & 1.00 & 0.88 & 0.80 & 850 & 860 + pa150mid & 3 10 , -3 -13 & 1.12 & 1.01 & - & - & 460 & 860 + pa60top & 19 46 & 0.99 & 0.99 & 0.86 & 1.19 & 850 & 860 + pa60bot & -20 -46 & 0.96 & 0.99 & 1.02 & 0.61 & 370 & 860 + pa60mid & 3 18 , -3 -19 & 1.08 & 1.06 & - & - & 410 & 860 + night sky lines were present as were the high ionization lines arising from the outer halo of the ring nebula .
the regions beyond the bright ring were used to determine how much should be subtracted from the ring spectra .
a final pair of background subtracted spectra of 107 pixels length centered on the bright ring was created .
both of these spectra were sampled in four regions called top ( 25 pixels ) , bottom ( bot , 25 pixels ) , and middle ( mid , 13 pixels ) with the results presented in table 1 .
an illustration of the quality of the spectra is shown in figure 1 .
the range of pixels measured from the central star , with positive numbers away from the direction of the p.a . , are given in table 2 .
the pairs of middle samples were adopted because of the much lower surface brightness near the central star .
the sample region spectra were measured with iraf task `` splot '' , with deconvolution of line blends where necessary .
not all features were measured , in particular those blended lines that would not be useful in a quantitative analysis .
the interstellar extinction correction was determined from the f(h@xmath7)/f(h@xmath6 ) and f(h@xmath4)/f(h@xmath6 ) ratios using the general extinction curve presented in osterbrock & ferland ( 2006 ) and their theoretical line ratios of 2.87 and 0.256 respectively .
the weighted average extinction was c@[email protected] ( in close agreement with the value of 0.14 derived by guerrero , et al .
( 1997 ) but about half that of barker ( 1987 ) ) and the same extinction correction was applied to each sample .
the extinction curve utilized and the results of this analysis of the samples are presented in table 1 .
we also derived c@xmath8 pixel by pixel along both slits .
this indicated that there is a sharp rise just outside of the peak in h@xmath6 brightness in the directions p.a.=150 , 240 , and 330 , but not towards 60 .
these rises are like those shown in garnett & dinerstein ( 2001 ) towards p.a.=271 , although they questioned the reality of the rise because it might be due to misalignment of their spectra , which is not an issue with our observations .
since the peaks in brightness occur just inside the projection of the ionization front and the front is curved , the rise in extinction just outside the peaks is probably due to dust in dense neutral gas projected onto ionized material inside the ionization front .
our spectra are comparable in resolution to the 7 resolution spectra of guerrero , et al .
( 1997 ) at similar p.a .
values and with a slit width of 1.2 , however , their exposure time along the major axis was 1800 sec ( ours was 4740 sec ) and the sum of the exposure times of their spectra nearly along the minor axis was 4500 sec ( ours was 9060 sec ) , although it should be noted that they covered a wider wavelength range ( 3727 through 7330 ) .
our resolution was poorer than the 2.1 - 2.3 reported by garnett & dinerstein ( 2001 ) , who had a total exposure time of 3600 sec at p.a.=91 with a 2.5 wide slit , but covered a wider wavelength range than their 4150 through 5000 sample .
the study of liu et al .
( 2004 ) was made with a p.a.=90 slit at a resolution of 1.5 in their blue samples and 3.5 in the red , both with a 1 " slit and a maximum total exposure time in their reddest sample of 2560 sec .
they reduced the long - slit spectrum to a single spectrum , which we show in 4.1 to blend regions of quite different conditions .
their total wavelength coverage was 3600 through 8010 . because of the wider entrance slit and longer exposure times ,
our data are of much higher spectrophotometric quality than these earlier studies , within the limitations of the spectral resolution .
a measure of the large signal in this program is that the 4922 line of hei in the south east peak of the main ring gave an individual 1200 sec exposure signal of 850 counts in one pixel along the slit , even though this line is only about 1% the flux of the h@xmath6 line and this observation represents only 13 % of the entire signal for the 4922 line .
since the central star has a surprisingly large spread in published v magnitude brightness ( 15.016.2 ) @xcite , we carefully measured this star on our spectra at 5550 (near the center of the v bandpass ) and determined that the correct value is [email protected] , in good agreement with the value of 15.75 determined by harris , et al .
the small difference in our new results and the value adopted in our earlier study of the nebula @xcite mean that the stellar properties adopted in that paper do not need to be changed .
in an earlier study of the ring nebula @xcite pairs of hst wfpc2 images in the f658n filter were compared by magnifying the first epoch images until the difference of the first and second images was a minimum .
the time interval between the images employed was 1.944 years and the magnification necessary was [email protected] .
this corresponds an annual expansion rate ( motion in mas yr@xmath2 divided by the distance @xmath10 from the central star in arcseconds ) of [email protected] mas yr@xmath2/. however , a later paper @xcite argued that these results were possibly incorrect since the central intersection of the wfpc2 centered on the central star , which meant that the image of the nebula fell into four different ccd s and the known drift of the ccd s with respect to one another @xcite meant that an internally self consistent but flawed answer could be obtained .
it is for this reason that we formulated the go 11232 observing program , where the images fell into a single ccd , whose results we present here .
our new data were processed in two ways , first by magnification and comparison , as done before , and by measurement of individual features .
our method of comparison was to magnify the first epoch images , align these with the second epoch image in the same filter , normalize them so that the total signal was the same , then take their ratio .
we noted that the ratio images began as being highly structured , indicating a mismatch , then became nearly grey in a small scaling range , then again became structured at larger scaling factors .
samples of the ratio images used in the analysis are shown in figure 2 .
this rather qualitative method gave a best scaling factor of [email protected] for the f658n images and [email protected] for the f502n images , yielding annual expansion rates over the 9.557 year interval of [email protected] mas yr@xmath2/ and [email protected] mas yr@xmath2/ respectively , a weighted average ( f502n being given one half weight because of less structure in that filter ) of [email protected] mas yr@xmath2/ , only a fraction of the annual expansion rates previously derived from segmented images and arguing that those rates are in error because of the use of images divided onto four ccd s .
the probable errors of our present study are very qualitative as they depend strictly on judgement about when the fit is best .
non - homologous expansion would have manifested itself by different parts of the ratio image becoming featureless at different scale factors .
qualitatively , this did not seem to be the case except in the bright northern quadrant , where some non - homogeneity could be seen in the inner and outer parts of the brightest section ( which was homologous ) .
ccccc 54.7 & 29.3 & 10.20 & 60.1 & 0.348 + 55.4 & 35.9 & 4.65 & 52.4 & 0.130 + 61.2 & 23.5 & 6.69 & 84.1 & 0.285 + 66.0 & 25.1 & 7.37 & 80.6 & 0.294 + 69.4 & 33.2 & 8.98 & 102.2 & 0.270 + 70.9 & 41.7 & 13.86 & 75.8 & 0.332 + 130.6 & 30.8 & 7.90 & 127.7 & 0.256 + 139.6 & 30.1 & 6.12 & 131.1 & 0.203 + 164.2 & 26.6 & 7.42 & 135.0 & 0.279 + 177.1 & 26.9 & 8.00 & 156.6 & 0.297 + 189.1 & 31.3 & 5.70 & 175.4 & 0.182 + 233.6 & 31.0 & 9.68 & 228.1 & 0.312 + 248.7 & 21.9 & 1.46 & 202.7 & 0.067 + 253.0 & 35.0 & 5.78 & 224.1 & 0.165 + 262.5 & 31.8 & 8.38 & 241.9 & 0.429 + 266.8 & 32.2 & 8.25 & 243.1 & 0.256 + 322.1 & 28.1 & 5.00 & 1.4 & 0.178 + 327.9 & 27.7 & 2.08 & 352.5 & 0.075 + 337.5 & 26.3 & 4.42 & 357.5 & 0.168 + 345.2 & 25.4 & 5.13 & 29.6 & 0.202 + ccccc 3.5 & 14.9 & 7.34 & 6.8 & 0.493 + 35.5 & 21.8 & 7.41 & 25.1 & 0.340 + 62.2 & 23.8 & 5.98 & 53.3 & 0.251 + 301.8 & 21.3 & 5.01 & 334.9 & 0.235 + the more quantitative method employed compared the position of isolated individual features ( bright in the case of f658n and dark in the case of f502n ) on aligned , normalized images .
the method is one of shifting small sections of an image and identifying when the differences are least , a method previously developed @xcite for measurement of herbig - haro features .
the results are shown in figure 3 and are presented in table 3 and table 4 . in the same manner as the comparison of images method , we can express our results as the motion divided by the distance to the central star , which would be a constant for homologous expansion .
the errors in the position for individual features expressed in relative motion were typically 0.02 mas yr@xmath2/. for f658n this method gave for 20 objects a global average for the expansion rate of [email protected] mas yr@xmath2/ and for four f502n objects [email protected] mas yr@xmath2/ , for a weighted average of [email protected] mas yr@xmath2/. these values are slightly larger than the results from simply comparing the scaled images ; however , these are to be preferred because of their more quantitative derivation . in this case
individual features are measured and an estimate of their accuracy is determined from the scatter of the results . in the scaling method described in
3.1 there is only the qualitative judgement of when the scaling is correct , with the uncertainty estimated by when the scaling would become unreasonable .
there were no statistically significant differences in the expansion properties in the four quadrants sampled .
we determined by comparing the p.a . of the measured motions with the p.a . from the central star
that possible slight misalignment of the images in rotation did not strongly affect our results .
we found that on the average the motion vectors were 5 smaller
. this could be the results of a misalignment of the first and second epoch images of only 0.0116 and this would affect the derived expansion motions by only about 0.05% .
ccccc 25 & f658n & 23.0 & 3.96 & 0.17 + - & - & 31.8 & 6.98 & 0.22 + 70 & - & 26.3 & 8.96 & 0.34 + 115 & - & 23.5 & 6.04 & 0.26 + 160 & - & 41.7 & 10.53 & 0.25 + - & - & 33.5 & 0.63 & 0.02 + - & - & 25.6 & 3.65 & 0.14 + - & - & 22.4 & 5.21 & 0.23 + 205 & - & 30.9 & 4.17 & 0.13 + - & - & 22.9 & 5.63 & 0.25 + 250 & - & 20.9 & 6.46 & 0.31 + - & - & 26.7 & 3.44 & 0.13 + - & - & 31.6 & 2.71 & 0.08 + 295 & - & 29.6 & 7.40 & 0.25 + 340 & - & 29.7 & 6.57 & 0.22 + 25 & f502n & 21.5 & 12.82 & 0.60 + 70 & - & 24.9 & 4.48 & 0.18 + 115 & - & 22.3 & 9.17 & 0.41 + 160 & - & 26.8 & 5.00 & 0.19 + - & - & 18.2 & 6.98 & 0.38 + 250 & - & 24.9 & 5.52 & 0.22 + 340 & - & 28.3 & 3.44 & 0.12 + we also measured features within the nebula employing a different approach .
the method was to create samples 4.0 wide along p.a .
values selected to lie along the major and minor axes of the nebula and intermediate angles displaced by 45 .
averages were taken perpendicular to the p.a .
direction and profiles were created .
well defined peaks not associated with knots were then measured in the first and second epoch images and in both filters .
fewer [ o iii ] features were measured because of the more amorphous appearance of the nebula in the f502n filter .
the results are shown in figure 3 and table 5 .
the expansion rate was [email protected] mas yr@xmath2/ for f502n , and [email protected] mas yr@xmath2/ for f658n .
because the f502n features are broader and more difficult to measure , they were assigned half the weight of the f658n features in determining the average expansion rate for the nebular features of [email protected] mas yr@xmath2/.
the ionization stratification of the ring nebula in the plane of the sky has been well defined by a number of studies including both imaging in various ions @xcite and long - slit spectroscopy like our own @xcite .
these studies indicate that in the main equator of material , which lies almost in the plane of the sky , that the nebula is optically thick to lyman continuum radiation ( e @xmath11 13.6 ev ) .
closer to the central star the radiation field is determined by the optical depth in helium , either singly ionized ( e @xmath11 24.6 ev ) or doubly ionized ( e @xmath11 54.4 ev ) .
as summarized earlier @xcite , these various zones of ionization have convenient optical emission - line tracers .
[ o i ] and [ s ii ] come from the outermost region where hydrogen is transitioning from being highly ionized to being neutral ( which we ll call the main ionization front or mif ) .
[ n ii ] comes from the next closest region to the central star , where h@xmath12 and he@xmath13 coexist ( which we ll call the he@xmath13 zone ) .
the next zone is populated by h@xmath12 and he@xmath12 ( which we ll call the he@xmath12 zone ) and is traced by [ o iii ] emission .
the innermost zone finds h@xmath12 and he@xmath14 and is traced by the heii recombination lines like 4686 ( this region will be called the he@xmath14 zone ) .
our spectra cover a sufficiently wide range of line intensities and ionic species , that we can independently determine densities and temperatures for most of the regions .
we have done this using the iraf / stdas / nebula task `` temden '' , which calculates the temperature ( for an assumed density ) or density ( for an assumed temperature ) from extinction corrected line ratios for a wide range of ions .
it must be kept in mind that in the best 3-d model for the ring nebula @xcite that the equatorial high density region is elliptical , with its plane lying about 7 out of the plane of the sky , c. f. 5.2 and odell , et al .
( 2007 ) , and there is a lower density polar region of about 1.5 larger in size that extends perpendicular to the equatorial region @xcite .
this means that samples in taken the bright ring ( our pa150top , pa150bot , pa60top , and pa60bot samples ) are measures of what is happening in the equatorial region , while the samples from close to the central star ( our pa150mid and pa60mid samples ) are seeing radiation from over a wide range of ionization states , with any low ionization emission lines coming from material physically distant from the central star .
the electron densities can in principle be determined from doublet line ratios with little dependence on the assumed electron temperature .
the [ s ii ] 6717 and 6731 doublet ratio gives densities at the mif , while the [ cl iii ] 5517 and 5537 doublet largely arises in the he@xmath12 zone , with the results shown in table 2 .
garnet and dinerstein ( 2001 ) used their higher resolution spectra to determine densities in the he@xmath12 zone from the [ ar iv ] 4711 and 4739 ratio , after correcting the 4711 for the unresolved hei 4713 line by scaling from the hei 4471 line , finding densities of about 700 @xmath15 , which are in basic agreement with our [ cl iii ] results for this same region .
the electron temperatures can be obtained from ratios of lines having very different excitation levels .
the most widely used is the f ( 4363)/[f(5007)+f(4959 ) ] of [ o iii ] , which arises in the he@xmath12 zone .
since the 5007 line was saturated in the bright ring portion of even our shortest exposure time spectra , we determined the f(5007)/f(4959 ) from the unsaturated central section of the spectra and found this to be 2.94 .
the value predicted using the transition probabilities presented in osterbrock and ferland ( 2006 ) is 2.9 and the observed value could be used to refine the determination of the transition probabilities .
the results are also given in table 2 and are in good agreement with those of earlier studies @xcite .
the [ n ii ] f(5755)/[f(6548)+f(6583 ) ] gives temperatures in the he@xmath13 zone which are similar to those in the he@xmath12 zone , which is not surprising since most of the heating arises from lyman continuum photons and that radiation field has suffered little modification in these regions .
in even finer sampling of our spectra we see the same rise in the [ o iii ] and [ n ii ] temperatures towards the central star that was found by garnett & dinerstein ( 2001 ) in [ o iii ] , except that the local peak in temperature at about 10 on both sides of the central star is better defined .
this position corresponds to the inner boundary of a peak region of heii 4686 emission , which suggests that there is a central lower density but higher temperature region along the polar axis of the nebula .
temperatures are determined in the mif by the [ f(4068)+f(4076)]/[f(6731)+f(6731 ) ] ratio of [ s ii ] and the f(5577)/[f(6300)+f(6363 ) ] lines of [ o i ] .
the intrinsically weak 5577 line is absent in our middle region samples , thus precluding derivation of a temperature there and the f(4068)+f(4076 ) lines are contaminated by line of sight higher ionization lines ( in particular o ii recombination lines ) . in our earlier paper @xcite that produced a 3-d model of the nebula
, we presented the results of calculations with the cloudy @xcite code using the appropriate stellar and nebular conditions .
excellent agreement was found between the observed and calculated properties for both the axes lying nearly in the plane of the sky .
less good agreement was found for the more poorly defined structure nearly along the line of sight , where the predicted nebular density was too small .
this discrepancy is exacerbated by our spectroscopic determination that the densities in the outer parts of these polar regions ( as determined from the [ s ii ] doublet in the middle samples ) are no lower than in the outer parts of the planar region .
this leaves a quandary about the structure in the polar region , as both the arguments of odell et al .
( 2007 ) and steffen et al .
( 2007 ) for the ionization of the outer halo call for this region being optically thin .
it is possible to combine the results of our earlier study of the radial velocities @xcite together with our new results from spectroscopy and expansion in the plane of the sky to improve our knowledge of the properties of the ring nebula . in this section
we present the velocity dependence upon the distance from the central star ( 5.1 ) , derive the tilt of the nebula with respect to the plane of the sky ( 5.2 ) , derive the nebula s dynamic distance ( 5.3 ) , comment on corrections that may be necessary due to the relative motion of the mif with respect to the nebular gas ( 5.4 ) , demonstrate that knots appear to be moving with the expanding nebular gas ( 5.5 ) , then comment on how these results alter the previous picture of the age of the nebula and stage of evolution of the central star ( 5.6 ) .
arguments for the existence of homologous expansion in the ring nebula appears in both the radial velocity data @xcite and in our tangential velocity data .
as noted earlier , this type of motion defies explanation by hydrodynamic models , but it can not be ignored because it has been well established for radial velocities for about half a century @xcite . in the case of the present study of tangential velocities ,
the average of the results from measuring motions of the knots and the nebula give a tangential expansion relation of v@[email protected] x @xmath10 mas yr@xmath2 , where @xmath10 is the distance from the central star in the plane of the sky in seconds of arc .
the relation of this result to the radial velocity expansion relation is determined by the distance to the nebula ( d ) . by comparing the ratio of radial velocities with respect to the systemic velocity ( v@xmath17 ) at various angular distances from the central star , odell et al .
( 2007b ) determined that v@xmath17/@xmath10=0.98 @xmath18/arcsec , without giving an uncertainty .
the relation of v@xmath17 to the spatial expansion velocity ( v@xmath19 ) will be determined by the geometry of the spheroid best describing the nebula .
odell et al . (
2007b ) argue that the structure perpendicular to the main ring of the nebula is a spheroid with a vertical length ( h ) 1.5 times the major axis of the elliptical main ring . in this case
the spatial expansion law would be v@xmath19=0.65x@xmath20 @xmath18 , where @xmath20 is the angle corresponding to the distance between a point in the nebula and the central star .
we can use our new results to more accurately determine the tilt of the plane of the elliptical main ring of the nebula with respect to the plane of the sky ( @xmath21 ) .
one knows from the non - tilted form of the velocity ellipses of the high resolution spectra made nearly across the minor axis ( p.a.=160 ) that this axis lies in the plane of the sky , whereas those nearly along the major axis ( p.a.=70 ) are tilted with a difference of velocity of 6@xmath92 @xmath18 , with the east - northeast end approaching the observer . assuming that the full major axis is 67 and using the expansion rate of [email protected] mas yr@xmath2/ means that the tangential velocity is 25.6 @xmath18 .
the tangent of the tilt angle will be the ratio of the radial velocity difference ( 3 @xmath18 ) divided by the tangential velocity ( 25.6 @xmath18 ) .
this yields a tilt angle of 6.7@xmath92 and decreases about linearly if the true distance is greater than the adopted distance of 700 pc .
this is in excellent agreement with the derivation from only the radial velocities of 6.5@xmath9 2 , the result presented in odell et al .
( 2007b ) .
this is small enough to simplify the calculation of the distance .
there are two complicating factors that need to be considered carefully when deriving a kinematic distance to the nebula .
the first is that the radial velocities derived from spectroscopy are most sensitive to the expansion of those regions of the nebula where the expansion direction is closest to the line of sight . in the case of the ring nebula ,
these are the polar regions of the nebula
. the proper motions , on the other hand , will be more representative of the equatorial regions of the nebula , where the expansion direction is closest to the plane of the sky . in order to deal with this complication
, it is necessary to adopt some model for the internal kinematics of the nebula . in this paper , we follow @xcite in assuming a `` hubble flow '' ( all gas velocities strictly radial and linearly proportional to distance from the central star ) , with a nebular radius along the polar axis of 1.5 times that along the major equatorial axis at @xmath22 .
although this simple kinematic model seems to be broadly consistent with the observations , it is important to test it by searching for deviations from a pure hubble flow , as emphasized by steffen et al .
( 2008 ) .
the second complicating factor is that the proper motions measure a pattern speed , whereas the spectroscopy measures the physical speed of the gas .
these are not necessarily the same thing , as elucidated in detail by mellema ( 2004 ) . in appendix
a , we derive the correction factor , @xmath23 ( ratio of pattern speed to physical speed ) as a function of two parameters : @xmath24 , which is the degree of concentration of the ionized gas at its outer edge , and @xmath25 , which is the rate of change of the stellar ionizing luminosity divided by the nebular expansion rate : @xmath26
. the density profile found from fitting cloudy models to nebular imaging and spectroscopy @xcite shows a roughly linear increase with radius , implying @xmath27 .
the dynamic timescale @xcite and the stellar evolution timescale @xcite are roughly equal at 45000 years , implying @xmath28 .
hence , we find @xmath29
. we can derive the distance to the ring nebula by assuming that the component of the spatial expansion ( v@xmath19 ) in the plane of the sky is equal to the corrected tangential component , the former being determined from the radial velocity study @xcite and the latter using the tangential angular motions determined in the present study . equating these two components one has v@xmath19 x cos @xmath21=4.74 @xmath1 x d x @xmath23@xmath2 , where units of arcsec / year , @xmath18 , and parsecs are used . within the assumption of homologous expansion ,
the distance from the central star drops out . employing the velocity expansion relation
( v@xmath19=0.65x@xmath20 @xmath18 ) determined from the radial velocities , the expansion rate of [email protected] yr@xmath2 determined from the observed tangential angular velocity , and the value @xmath23=1.34 from 5.3 , one finds d=790@xmath30 pc .
this is larger , but within the uncertaintanties of the parallax determined value of 700@xmath0 pc @xcite . since both methods determine the distance from the reciprocal of the measured quantity , it is illuminating to compare the measured parallax [email protected] mas with our predicted parallax [email protected] mas .
taking the average of these parallaxes ( [email protected] ) , one determines a best recommended distance of 740@xmath30 pc .
this is only a small change from the parallax determined distance used previously , but , it is determined from two independent methods and is the value to now be preferred .
the origin of the knots in the ring , helix , and other nearby pne @xcite remains highly uncertain and an important diagnostic observation is whether the knots share the general expansion characteristics of the host pn or if they trail behind . in a study of the radial velocities of the similarly nearly pole - on pn ngc 7293 ( the helix nebula )
@xcite argued that the knots were expanding , but at about one half the rate of expansion of the nebular material .
this interpretation of the radial velocity data was disputed @xcite when the knot velocities were compared with several sources of nebular velocities , the conclusion being that the expansion rates of the knots and the nebula are indistinguishable .
that also appears to be the case for the ring nebula , since the plane of the sky expansion rate for the knots ( [email protected] mas yr@xmath2/ ) is not distinguishably different from that for the nebular features ( [email protected] mas yr@xmath2/ ) .
this means that either the knots in the ring nebula move in a very different way than those in the helix nebula or that the helix nebula results based on radial velocities @xcite are incorrect .
the images of the ring nebula made in the h@xmath31 2.12@xmath1 line ( speck et al .
2003 ) indicate that this emission arises from the same ionization bounded knots that produce the bright cusps seen in [ n ii ] and their dark central knots seen in [ o iii ] .
the h@xmath31 emission is probably excited by stellar radiation more energetic than 13.6 ev in the same manner as that in the helix nebula @xcite .
a lower spatial resolution fabry perot study @xcite at a resolution of 24 @xmath18 also noted that the 2.12@xmath1 emission arose in these features .
hiriart ( 2004 ) used his data to derive a model in which the 2.12@xmath1 emission comes from a cylindrical shell seen almost along its rotational axis , which is fully consistent with the triaxial spheroid model ( odell et al .
2007b ) that has most of the material lying in an equatorial disk lying nearly in the plane of the sky . using the h@xmath31 velocities and distances from hiriart s study
, one finds a ratio of expansion velocity to angular separation of 1.1 @xmath18/arcsec , which is within the determination errors of the value of 1.0 @xmath18/arcsec derived from the optical line study @xcite ( c. f. 5.1 ) , which had a velocity resolution of 7 @xmath18 .
this agreement in radial velocity adds credence to the interpretation that the knots are moving at the same velocity as the nebular material . on a small scale , many planetary nebulae show imhomogeneities in the form of dense molecular knots @xcite . in the case where the knot is well inside the global ionization front ,
there is expected to be a locally divergent flow of ionized gas away from the head of the knot , e.g. , bertoldi & mckee ( 1990 ) .
this contradicts the assumption that all gas flows are radially away from the central star , so the above analysis can not be applied directly to this case .
however , the ionization front conditions on the knot should be approximately d - critical ( @xmath32 , where @xmath33 is the isothermal sound speed in the ionized gas ) .
therefore , one finds @xmath34 , where @xmath35 is the mach number of the ionized gas at the surface of the knot on a radial line from the central star , as measured in the rest frame of the star . in order to apply this formula ,
it is necessary that both the proper motion observations ( which measure @xmath36 ) and the spectroscopic observations ( which measure @xmath37 ) are detecting the knots , rather than the diffuse nebular emission .
this requires a very high angular resolution , which is not typically available from ground - based spectroscopy . in
2.2 we established that the visual brightness ( v=15.65 ) of the central star was 0.1 magnitudes brighter than assumed previously @xcite , but that the extinction was less ( c@xmath8=0.13 , a@xmath38 ) than had been previously assumed ( c@xmath8=0.20 , a@xmath38=0.43 ) , so that this would not significantly change the previous result that the absolute luminosity of the central star is about 200 l@xmath39 , ( these new numbers argue for its being 5% fainter ) . the largerer recommended distance ( 740 pc rather than 700 pc )
argues that the star is an additional 11% brighter .
this means that at present there are no good grounds to question our previous conclusions about the evolutionary state of the central star , that is , that it is a star of about 0.61 - 0.62 m@xmath39 about 7000 years beyond the end of the asymptotic giant branch wind .
we have obtained long time base images of ngc 6720 , the ring nebula and from these have derived its expansion characteristics in the plane of the sky .
supplemental ground based spectra along the major and minor axis were also obtained .
these data have allowed us to draw these conclusions .
scaled ratio images and the motion of individual features argues with considerable uncertainty for homologous expansion , that is , the rate of expansion varies linearly with the distance from the central star .
the dark knots seen against the bright nebular emission appear to be expanding at the same rate as the surrounding gas .
the spectra have allowed determination of the electron temperature in all of the ionization zones present in this high ionization pn and of the electron density in the lowest ( mif ) and medium ( he@xmath12 ) ionization zones .
\4 . combining the tangential and radial velocity information , we have derived a new distance for the ring nebula of d=790@xmath40 pc , slightly larger than the 700 @xmath0 pc determined by parallax , although subject to revision because of correction for the difference in velocity of the mif and the surrounding nebular gas .
the best distance is now 740 @xmath41 pc .
we are grateful to yilen gomez maqueo chew , who provided excellent assistance in obtaining the spectra at the san pedro martir observatory .
we thank garrelt mellema for useful comments on a draft of this paper .
part of cro s support for work on this project came from grants go 11231 and go 11232 from the space telescope science institute .
wjh acknowledges financial support from dgapa - unam , mexico , project in110108 .
_ facilities : _ , spm
we derive here the relationship between the apparent propagation speed , @xmath36 , of an ionization front and the gas speed , @xmath37 , of ionized material immediately inside the front .
this calculation is complementary to that of mellema ( 2004 ) , who employed a strictly local procedure , based on the ionization front jump conditions .
mellema s results are exact , but depend on a free parameter that must be supplied from the outside .
we follow a global procedure , considering the nebula as a whole , allowing us to find a unique solution . on the other hand ,
we are forced to make some simplifying assumptions , which are nevertheless justified in most cases .
we consider a spherically symmetric , ionization - bounded , pure - hydrogen nebula , surrounding a point source of ionizing radiation with an ionizing photon luminosity that varies with time as @xmath42 .
the ionization front radius is @xmath43 and the total mass of ionized gas is @xmath44 .
the number density of hydrogen ions just inside the ionization front is @xmath45 , whereas the mean density and mean square density within the ionized volume are given by @xmath46 and @xmath47 , respectively .
we define a clumping factor , @xmath48 and a shell concentration factor , @xmath49 , which are both assumed to be constant with time .
the proton mass is @xmath50 and the case b recombination coefficient is @xmath51 .
we define the ionization front dynamic timescale as @xmath52 and the stellar evolution timescale as @xmath53 .
_ note that @xmath54 is negative when the ionizing luminosity falls with time .
_ given these assumptions , the mass of ionized gas is given by @xmath55 and the dynamic ionization balance in the nebula is given by @xmath56 where the first term on the right hand side represents the rate of recombinations in the nebula , while the second term represents the rate at which new particles are _ advected _ through the ionization front .
it is straightforward to show that the relative size of the second term is of order @xmath57 , where @xmath58 is the recombination time .
this ratio is rather small for typical nebular conditions , so we neglect the advective term in what follows , which is equivalent to assuming that the nebula passes through a sequence of static photoionization equilibrium states . with this approximation , we can take the time derivative of equation ( [ eq : ionbalance ] ) , eliminating the density by use of equation ( [ eq : mass ] ) , to obtain @xmath59 which yields the rate of change of ionized mass as @xmath60 on the other hand , we can also express the rate of change of ionized mass in terms of the relative velocity , @xmath61 , with which newly ionized gas leaves the ionization front : @xmath62 equations ( [ eq : mdot ] ) and ( [ eq : massflow ] ) can be combined to find the correction factor , @xmath63 , that relates the real and apparent expansion speeds : @xmath64 where @xmath65 is the ratio of the rate of evolution of the ionizing luminosity to the rate of nebular expansion .
the ionizing luminosity of the central stars of planetary nebulae first rises with time as the post - agb atmosphere contracts towards higher effective temperatures , and then falls swiftly when envelope nuclear burning ceases , followed by a much slower decline as the white dwarf interior cools @xcite .
the overall timescale for this sequence is set primarily by the core mass at the end of the agb phase ( higher mass cores tend to evolve more rapidly ) , but with a secondary dependence on the agb mass - loss history .
the expansion timescale of the nebula generally increases monotonically with the nebular age , so that the three phases in the luminosity evolution ( rise , swift decline , slow decline ) produce three phases in the nebular evolution : ionization ( @xmath66 ) , recombination ( @xmath67 ) , re - ionization ( @xmath68 ) .
these phases can be seen , for example , in the evolution of the mass of ionized gas in the models of natta & hollenbach ( 1998 ) .
the detailed evolution of the nebula can be rather intricate , even when only one - dimensional models are considered @xcite , especially during the high - luminosity phase when the fast stellar wind from the central star is expected to be dynamically important in the inner regions of the nebula .
nonetheless , so long as the nebula is optically thick to ionizing radiation , the simple equation given above can be reliably applied in order to correct expansion parallaxes measured from the main nebula shell for cases where detailed tailored models , such as those given in schnberner et al .
( 2005 ) , are not available . the [ @xmath69 line is to be preferred over [ @xmath70 for both spectroscopic and proper motion measurements , since it traces gas close to the ionization front , giving the most reliable estimates of @xmath37 and @xmath36 .
note that the model given here does _ not _ apply to density - bounded nebulae , where the ionizing luminosity is high enough to ionize all the circumstellar material .
the lack of an ionization front means that other features , such as shocks @xcite , must be used to determine expansion parallaxes in such cases . if the ionizing luminosity , @xmath71 , is constant , then @xmath72 and we have @xmath73 . for a constant density nebula , @xmath74 and we obtain @xmath75 , which is the well - known result for the evolution of a classical strmgren sphere during the subsonic , d - type phase the maximum ionized gas velocity is one half the expansion speed of the ionization front . on the other hand ,
for a thin , homogeneous ionized shell , of thickness @xmath76 , one finds @xmath77 and @xmath78 , so that the correction factor is much reduced , tending towards unity as the shell thickness tends to zero . if @xmath79 , then @xmath25 is also positive , which will always lead to larger values of @xmath23 than in the constant luminosity case .
the greatest increase is seen for a constant density nebula , whereas larger values of @xmath24 tend to dampen the effect .
as @xmath80 , then equation ( [ eq : correct ] ) suggests @xmath81 .
however , the equation is not valid in this limit , which corresponds to a highly supersonic weak - r type ionization front , since for such fronts the advective term would need to be retained in equation ( [ eq : ionbalance ] ) .
if @xmath82 , then @xmath25 is negative , which causes a reduction in the correction factor , @xmath23 , with respect to the constant luminosity case .
irrespective of the value of the density factor @xmath24 , if @xmath67 , then @xmath83 , so that the gas velocity is now greater than the expansion pattern speed . in this case
, the nebula is bounded by a _ recombination front _ , instead of an ionization front , and the mass of ionized gas falls with time .
anderson , j. , & king , i. r. 2003 , pasp , 115 , 113 barker , t. 1987 , apj , 322 , 922 bertoldi , f. & mckee , c. f. 1990 , , 354 , 529 blcker , t. 1995 , a&a , 299 , 755 ferland , g. j. , korista , k. t. , verner , d. a. , ferguson , j. w. , kingdon , j. b. , & verner , e. m. 1998 , pasp , 110 , 761 garnett , d. r. , & dinerstein , h. l. 2001 , apj , 558 , 145 guerrero , m. a. , manchado , a. , & chu , y .- h .
1997 , apj , 487 , 328 harris , h. c. , dahn , c. c. , canzian , b. , guetter , h. h. , leggett , s. k. , levine , s. e. , luginbuhl , c. b. , monet , a. k. b. , & six others , 2007 , aj , 133 , 631 hartigan , p. , morse , j. a. , reipurth , b. , heathcote , s. , & bally , j. 2001 , , 559 , l157 henney , w. j. , williams , r. j. r. , ferland , g. j. , shaw , g. , & odell , c. r. 2007 , , 671 , l137 hiriart , d. 2004 , pasp , 116 , 1135 lame , n. j. , & pogge , r. w. 1994 , aj , 108 , 1860 liu , x .- w .
, & barlow , m. j. 1996 , mnras , 279 , 511 liu , y. , liu , x .- w . , luo , s .-
, & barlow , m. j. 2004 , mnras , 353 , 1231 meaburn , j. , clayton , c. a. , bryce , m. , walsh , j. r. , holloway , a. j. , & steffen , w. 1998 , , 294 , 201 meaburn , j. , lpez , j. a. , & richer , m. g. 2008 , , 384 , 497 mellema , g. 2004 , , 416 , 623 natta , a. , & hollenbach , d. 1998 , a&ap , 337 , 517 odell , c. r. 1998 , aj , 116 , 1346 odell , c. r. , balick , b. , hajian , a. r. , & henney , w. j. 2002 , aj , 123 , 3329 odell , c. r. , henney , w. j. , & ferland , g. j. 2007a , aj , 133 , 2343 odell , c. r. , mccullough , p. r. , & meixner , m. 2004 , aj , 128 , 2339 odell , c. r. , sabbadin , f. , & henney , w. g. 2007b , , 134 , 1679 osterbrock , d. e. , & ferland , g. j. 2006 , astrophysics of gaseous nebulae and active galactic nuclei ( second edition ) , university science books ( mill valley ) perinotto , m. , schnberner , d. , steffen , m. , & calonaci , c. 2004 , a&a , 414 , 993 sabbadin , f. , turatto , m. , benetti , s. , ragazzoni , r. , & cappellaro , e. 2008 , , in press sabbadin , f. , turatto , m. , ragazzoni , r. , cappellaro , e. , & benetti , s. 2006 , a&a , 451 , 937 schnberner , d. , jacob , r. , & steffen 2005 , 441 , 573 speck , a. k. , meixner , m. , jacoby , g. h. , & knezek , p. m. 2003 , pasp , 115 , 170 steffen , w. , lpez , j. a. , koning , n. , kwok , s. , riesgo , h. , micher , m. g. , & morisset , c. 2007 , in asymmetrical planetary nebulae iv , in press steffen , w. , garca - segura , g. , & koning , n. 2008 , , in press terzian , y. 1997 , in iau symposium 180 , planetary nebulae , ed . h. j. habing & h. j. g. l. m. lamers ( dordrecht , kluwer academic publishers ) , 29 weedman , d. w. 1968 , , 153 , 49 wilson , o. c. 1950 , apj , 111 , 279 | we have combined recent hubble space telescope wfpc2 images in the [ o iii ] 5007 and [ n ii ] 6583 lines with similar images made 9.557 years earlier to determine the motion of the ring nebula within the plane of the sky .
scaled ratio images argue for homologous expansion , that is , larger velocities scale with increasing distance from the central star .
the rather noisy pattern of motion of individual features argues for the same conclusion and that the silhouetted knots move at the same rate as the surrounding gas .
these tangential velocities are combined with information from a recent high resolution radial velocity study to determine a dynamic distance , which is in basic agreement with the distance determined from the parallax of the central star .
we have also obtained very high signal to noise ratio moderate resolution spectra ( 9.4 ) along the major and minor axes of the nebula and from this determined the electron temperatures and density in the multiple ionization zones present .
these results confirm the status of the ring nebula as one of the older planetary nebulae , with a central star transitioning to the white dwarf cooling curve .
= 1 |
since their introduction @xcite , topological field theories have been object of continuous and renewed investigations and have led to many interesting applications @xcite . their original motivation was to provide a natural framework for a field theory characterization and computation of several topological invariants .
for instance , the chern - simons and the bf theories in three dimensions play a relevant role in knots theory , their correlation functions being respectively related to the jones @xcite and the alexander - conway @xcite polynomials . in particular , a generalization of this framework to higher dimensions allows to describe the linking number of surfaces in terms of antisymmetric tensor fields , whose action reads @xcite @xmath1 where @xmath2 and @xmath3 are forms of degree @xmath4 and @xmath5 respectively .
indeed , it is possible to show @xcite that the correlation function @xmath6 gives the linking number between two smooth , closed nonintersecting surfaces @xmath7 and @xmath8 .
it is worth underlining also that , more recently , much attention has been given to the fact that topological terms of the kind of ( [ s - top ] ) appear frequently as parts of more general effective actions describing the low - energy dynamics of several field theory models .
for instance , the effective action corresponding to the bosonization @xcite of relativistic three - dimensional massive fermionic systems at @xmath9 can be written as the sum of a chern - simons term @xmath10 and of an infinite series of higher order terms in the curvature @xmath11 and its derivatives .
the same procedure can be applied when fermions are coupled to a dynamical vector field @xmath3 . in this case
the resulting bosonized action contains a mixed term of the type of ( [ s - top ] ) .
these effective actions are well suited to describe several three - dimensional phenomena such as the fermi - bose transmutation @xcite , the quantum hall effect @xcite , the incompressible chiral liquids @xcite .
further interesting examples are provided by the low energy effective supergravities studied in the context of the @xmath12 correspondence @xcite .
indeed , as discussed by @xcite , the dual supergravity action which describes the @xmath13sym theory with electric and magnetic fluxes is @xmath14 where @xmath15are the ramond - ramond and neveu - schwarz neveu - schwarz two forms , transforming as a doublet under the @xmath16 s - duality symmetry , @xmath17 is a negatively curved five - dimensional einstein manifold , and @xmath18 collects all the higher order terms in the curvatures @xmath19 . in all these models ,
the topological terms appear together with higher derivatives terms in the generalized curvatures , which depend on the metric .
therefore we are naturally led to analyse the possible contributions coming from these terms , as recently done for the three - dimensional abelian chern - simons theory @xcite . in this case
it has been possible to prove that the correlator @xmath20 describing the linking number between two smooth closed nonintersecting curves @xmath21 , is not affected by the presence of higher derivative interaction terms of the type @xmath22 .
the aim of the present work is that of generalizing this result to higher dimensions . in particular we shall see that , in analogy with the three - dimensional chern - simons , the correlator @xmath23 is not affected by the introduction of local perturbation terms in the curvatures of the type @xmath24 and gives as final result the linking number of the two surfaces @xmath25
this result means that the correlator ( [ 1 ] ) is stable against the introduction of perturbing local terms in the curvatures , regardless of their power - counting nonrenormalizability .
it is worth underlining that we will limit here ourselves only to effective actions which are abelian , and we shall consider only terms dependent on the curvatures @xmath26 which can be treated as true perturbations .
namely , we shall avoid in the effective action @xmath27 the inclusion of a term of the maxwell type @xmath28 where @xmath29 @xmath30 are mass parameters .
the presence of this term would completely modify the original properties of the model .
in fact , being expression ( [ max ] ) quadratic in the gauge fields , it can not be considered as a true perturbation term , as it will be responsible for the presence of massive excitations in the spectrum of the theory . rather , the presence of the maxwell term in the effective action @xmath31 will give rise to the existence of two distinct regimes corresponding to the long and short distance behaviors , respectively .
for distances larger than the inverse of the mass parameters @xmath29 @xmath32 ( _ i.e. _ , the low energy regime ) , the topological term will prevail , _ _
_ _ while the maxwell term will become the relevant one at short distances ( _ i.e. , _ the high - energy regime ) . in general
, the computation of the correlator ( [ corr - top ] ) for arbitrary surfaces in the presence of the maxwell term is a very difficult task , mainly due to the metric dependence of the propagator .
indeed , in the abelian three - dimensional maxwell - chern - simons case no closed analytic expression for the correlator ( [ c - s - l ] ) has been found so far , even in the simpler case in which the curves @xmath21 are two circles .
the inclusion of the maxwell term is beyond the aim of the present paper , being presently under investigation .
we would like to remark that , in the low energy regime , a useful local gauge invariant field redefinition has been recently applied @xcite for the maxwell - chern - simons theory to compute the corrections to the correlator ( [ c - s - l ] ) for two coinciding curves @xmath33 as power series in the inverse of the mass .
this field redefinition can be also generalized to the present case and is expected to be suitable for the characterization of the maxwell contributions in the low energy regime @xcite .
in order to analyse the aforementioned nonrenormalization properties of the linking number , let us proceed by computing the correlator @xmath34 in the case in which the total action @xmath35 is chosen to be @xmath36 where @xmath37 is an arbitrary parameter with negative mass dimension , and @xmath38 , @xmath39 are respectively the curvatures of the gauge fields @xmath2 and @xmath3 .
notice that the normal ordering prescription has been adopted for the quartic interaction , allowing us to rule out tadpoles diagrams and @xmath2 are each gauge invariant . ] .
the action @xmath35 is invariant under the gauge transformations @xmath40 remark that the transformations @xmath41 are reducible and therefore the quantization procedure will require the introduction of ghosts for ghosts . in order to gauge fix the action @xmath42 we choose the linear landau tansverse condition @xmath43 which implies that the whole set of ghosts for ghosts completely decouples , due to the abelian character of the theory .
thus , for the propagator we get @xmath44 where @xmath45 is the green function of the laplacian @xmath46 since the two surfaces @xmath47 must be homologically trivial @xcite @xmath48 we can suppose that they are contained in a region of @xmath49 diffeomorphic to the flat euclidean space @xcite , but we do not consider these particular cases here . ]
@xmath50 . in this case , the green function is given by @xmath51 } \frac{^1}{\left| x - y\right| ^{n-2}}\,,\ ] ] where @xmath52 denotes the euclidean distance between @xmath53 and @xmath54 .
let us now give the elementary wick contractions which shall be needed for the evaluation of the feynman diagrams .
from @xmath55 we obtain @xmath56 } ^{\nu _ 1 ... \nu _ p}\delta ( x - y)+\right .
\nonumber \\ & & \left .
+ \,\,\,\,\delta _ { [ \mu _ 1, .. ,\mu _ { i-1},\mu _ { i+1}, .. ,\mu _ p}^{\nu _ 1, .. ,\nu _ { i-1},\nu _ { i+1}, .. ,\nu _ p}\partial _ { \mu _ i]}^x\partial _ y^{\nu _ i}g(x - y)\right ) , \nonumber \\ \label{bhcprop}\end{aligned}\ ] ] @xmath57 } ^{\nu _ 1 ...
\nu _ { n - p-1}}\delta ( x - y)+\right .
\nonumber \\&&\left .
+ \,\,\,\,\delta _ { [ \mu _ 1, .. ,\mu _ { i-1},\mu _ { i+1}, .. ,\mu _ { n - p-1}}^{\nu _ 1, .. ,\nu _ { i-1},\nu _ { i+1}, .. ,\nu _ { n - p-1}}\partial _ { \mu _ i]}^x\partial _ y^{\nu _ i}g(x - y)\right ) , \nonumber
\\ \label{chbprop}\end{aligned}\ ] ] where the index @xmath58 runs from one to @xmath4 , and @xmath59 notice that the propagator @xmath60 has the same structure of @xmath61 in @xmath62 . in the above equations @xmath63 and @xmath64 are respectively the hodge duals of the curvatures @xmath65 and @xmath66 and the symbol @xmath67 $ ] means complete antisymmetrization in the indices . concerning now the perturbative expansion , it is easy to see that the first feynman diagram which contributes to the correlation function @xmath68 is of two - loop order .
for the three - dimensional case , in which @xmath69 @xmath70 are two smooth oriented nonintersecting curves , this diagram can be drawn as follows in the above figure , the dashed , the wavy and the double lines refer respectively to the wick contractions @xmath71 and @xmath72 .
apart from an irrelevant global symmetry coefficient , the feynman integral corresponding to the diagram of @xmath73 is given by @xmath74 in the analysis of the above expression it is worth to observe that the two fields @xmath75 and @xmath76 , lying on the hypersurfaces @xmath77 and @xmath78 , are respectively wick - contracted with @xmath79 and @xmath80 . from the explicit form of the propagator @xmath61 , we see that the term in the second line of @xmath62 yields a total derivative on the closed hypersurface @xmath77 ; implying that this term gives a vanishing contribution . the same result also holds for the contribution coming from the propagator @xmath60 of @xmath81 .
it is worth remarking that this argument relies only on the fact that the two surfaces @xmath77 and @xmath70 are closed .
therefore , it applies to the higher - order contributions as well . as a consequence
, all the wick contractions entering the feynman diagrams will lead to a product of delta functions .
in particular , for the 2-loop diagram of @xmath73 , the integral @xmath82 reduces to @xmath83 } ^{\gamma _ 1 .... \gamma _ { n - p-1}}\delta \left ( y - z_2\right ) \delta _ { \left [ \mu _ 1 .... \mu _ p\right ] } ^{\sigma _ 1 .... \sigma _ p}\delta \left ( x - z_1\right ) \nonumber \\ & & \varepsilon ^{\sigma _ 1 .... \sigma _ { p+1}\gamma _ 1 .... \gamma _ { n - p-1}}\partial
_ { \sigma _ { p+1}}\delta \left ( z_1-z_2\right ) \varepsilon ^{\alpha _ 1 .... \alpha _ { p+1}\lambda _ 1 .... \lambda _ { n - p-1}\,}\partial _ { \alpha _ { p+1}}\delta \left ( z_1-z_2\right ) \nonumber \\ & & \varepsilon _ { \alpha _ 1 .... \alpha _ p\widehat{\alpha } _ { p+1}\lambda _ 1 .... \lambda _ { n - p-1}}^{\,}\partial ^{\widehat{\alpha } _ { p+1}}\delta \left ( z_1-z_2\right ) .
\end{aligned}\ ] ] in spite of the presence of products of delta functions with the same argument , the above expression is easily seen to vanish .
we observe first of all that it is always possible to take an order of integration over the delta functions such that we end up with products of @xmath84 and not of @xmath85 . in the present case
, this would amount to integrate out first the two delta functions with arguments @xmath86 and @xmath87 , leading to @xmath88 } ^{\gamma _ 1 .... \gamma _ { n - p-1}}\delta _ { \left [ \mu _ 1 .... \mu _ p\right ] } ^{\sigma _ 1 .... \sigma _ p } \nonumber \\ & & \varepsilon ^{\sigma _ 1 .... \sigma _ { p+1}\gamma _ 1 .... \gamma _ { n - p-1}}\varepsilon ^{\alpha _ 1 .... \alpha _ { p+1}\lambda _ 1 .... \lambda } \varepsilon _ { \alpha _ 1 .... \alpha _ p\widehat{\alpha }
_ { p+1}\lambda _ 1 .... \lambda _ { n - p-1}}^{\ , } \nonumber \\ & & \partial
_ { \sigma _ { p+1}}\delta \left ( x - y\right ) \partial _ { \alpha _ { p+1}}\delta \left ( x - y\right ) \partial ^{\widehat{\alpha } _ { p+1}}\delta \left ( x - y\right ) , \end{aligned}\ ] ] which is zero owing to the fact that the two surfaces @xmath77 and @xmath78 never intersect , so that @xmath89 never vanishes .
moreover , as done in @xmath90 , we can adopt a more rigorous treatment by regularizing the @xmath0-dimensional delta functions with coinciding arguments through the point splitting procedure already used by polyakov @xmath91@xmath92 more precisely , whenever a product of @xmath93 delta functions with coinciding arguments occurs , it will be understood as @xmath94 ^m=\left [ \delta _ \epsilon \left ( z_1-z_2\right ) \right ] ^{m-1}\delta \left ( z_1-z_2\right ) , \ ] ] where the limit @xmath95 is meant to be taken at the end of all calculations . in this regularization scheme
we then have @xmath96 } ^{\gamma _ 1 .... \gamma _ { n - p-1}}\delta ( y - z_2)\delta _ { \left [ \mu _ 1 .... \mu _ p\right ] } ^{\sigma _ 1 .... \sigma _ p}\delta \left ( x - z_1\right ) \nonumber \\ & & \varepsilon ^{\sigma _ 1 .... \sigma _ { p+1}\gamma _ 1 .... \gamma _ { n - p-1}}\partial _ { \sigma _ { p+1}}\delta _ \epsilon \left ( z_1-z_2\right ) \varepsilon ^{\alpha _ 1 .... \alpha _ { p+1}\lambda _ 1 .... \lambda _ { n - p-1}\,}\partial
_ { \alpha _ { p+1}}\delta _ \epsilon \left ( z_1-z_2\right ) \nonumber \\ & & \varepsilon
_ { \alpha _ 1 .... \alpha _ p\widehat{\alpha }
_ { p+1}\lambda _ 1 .... \lambda _ { n - p-1}}^{\,}\partial ^{\widehat{\alpha } _ { p+1}}\delta \left ( z_1-z_2\right ) , \label{i22reg } \end{aligned}\ ] ] which is vanishing no matter the order of integration we take . indeed , before taking the limit @xmath97 we always obtain an expression containing @xmath98 , with @xmath99 .
we can thus conclude that the two - loop diagram in @xmath73 does not contribute to the correlator @xmath68 .
similar arguments can be used to show that also the higher - order contributions vanish .
let us consider , for instance , the topologically distinct diagrams contributing to the 3 and 4-loop order , shown respectively in @xmath100 and @xmath101 .
concerning the 3-loop contribution of @xmath100 , a typical contraction is @xmath102 } ^{\varphi _ 1 .... \varphi _ { n - p-1}}\delta ( y - z_2)\delta _ { \left [ \mu _ 1 .... \mu _ p\right ] } ^{\beta _ 1 .... \beta _ p}\delta \left ( x - z_1\right ) \varepsilon _ { \rho _ 1 ...
\rho _ { p+1}\alpha _ 1 ... \alpha _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\rho _ { p+1}}\delta \left ( z_1-z_3\right )
\nonumber \\ & & \;\;\;\;\;\varepsilon _ { \sigma _ 1 ...
\sigma _ { p+1}\alpha _ 1 ... \alpha _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\sigma _ { p+1}}\delta \left ( z_1-z_2\right ) \varepsilon _ { \beta _ 1 .... \beta _ { p+1}\theta _ 1 ... \theta _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\beta _
{ p+1}}\delta \left ( z_1-z_3\right ) \nonumber \\ & & \;\;\;\;\;\varepsilon ^{\sigma _ 1 ... \sigma _ p\widehat{\sigma } _ { p+1}\theta _ 1 ... \theta _ { n - p-1}}\partial _ { \widehat{\sigma } _
{ p+1}}\delta \left ( z_3-z_2\right ) \varepsilon _ { \,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\varphi _ 1 .... \varphi _ { n - p-1}}^{\rho _ 1 ... \rho _ p\widehat{\rho } _ { p+1}\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial _ { \widehat{\sigma } _ { p+1}}\delta \left ( z_2-z_3\right ) \;. \nonumber
\\ \label{13}\end{aligned}\ ] ] analogously , a typical term of the four - loop contribution of @xmath103 turns out to be @xmath104 } ^{\sigma _ 1 .... \sigma _ p}\delta \left ( x - z_1\right ) \delta _ { \left [ \nu _ 1 .... \nu _ { n - p-1}\right ] } ^{\beta _ 1 .... \beta _ { n - p-1}}\delta ( y - z_2)\varepsilon ^{\theta _ 1 ... \theta _ { p+1}\alpha _ 1 ... \alpha _ { n - p-1}}\partial _ { \theta _ { p+1}}\delta \left ( z_1-z_2\right ) \nonumber \\ & & \varepsilon _ { \theta _ 1 ... \theta _ p\widehat{\theta } _ { p+1}\alpha _ 1 ... \alpha _ { n - p-1}}\partial ^{\widehat{\theta } _ { p+1}}\delta \left ( z_1-z_2\right ) \varepsilon _ { \sigma _ 1 ...
\sigma _ { p+1}\varphi _ 1 ...
\varphi _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\sigma _ { p+1}}\delta \left ( z_1-z_3\right ) \nonumber \\ & & \varepsilon _ { \xi _ 1 ...
\xi _ { p+1}\beta _ 1 ... \beta _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\xi _
{ p+1}}\delta \left ( z_2-z_4\right ) \varepsilon _ { \xi _ 1 ... \xi _ p\widehat{\xi }
_ { p+1}\varphi _ 1 ...
\varphi _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\widehat{\xi } _
{ p+1}}\delta \left ( z_3-z_4\right ) \nonumber \\ & & \varepsilon ^{\eta _ 1 ... \eta _ { p+1}\gamma _ 1 ...
\gamma _ { n - p-1}}\partial
_ { \eta _ { p+1}}\delta \left ( z_3-z_4\right ) \varepsilon _ { \eta _ 1 ... \eta _ p\widehat{\eta } _ { p+1}\gamma _ 1 ...
\gamma _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\widehat{\eta } _ { p+1}}\delta \left ( z_3-z_4\right ) \;.
\label{14 } \end{aligned}\ ] ] integrating over @xmath105 and @xmath106 , and after regularizing the product of delta functions with coinciding arguments , we obtain @xmath107 } ^{\varphi _ 1 .... \varphi _ { n - p-1}}\delta _ { \left [ \mu _ 1 .... \mu _ p\right ] } ^{\beta _ 1 .... \beta _ p}\varepsilon _ { \rho _ 1 ...
\rho _ { p+1}\alpha _ 1 ... \alpha _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\rho
_ { p+1}}\delta \left ( x - z_3\right ) \nonumber \\ & & \;\;\;\;\;\varepsilon _ { \sigma _ 1 ...
\sigma _ { p+1}\alpha _ 1 ... \alpha _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\sigma
_ { p+1}}\delta \left ( x - y\right ) \varepsilon _ { \beta _ 1 .... \beta _ { p+1}\theta _ 1 ...
\theta _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\beta
_ { p+1}}\delta \left ( x - z_3\right ) \nonumber \\ & & \;\;\;\;\;\varepsilon ^{\sigma _ 1
... \sigma _ p\widehat{\sigma } _ { p+1}\theta _ 1 ... \theta _ { n - p-1}}\partial _
{ \widehat{\sigma } _
{ p+1}}\delta \left ( z_3-y\right ) \varepsilon _ { \,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\varphi _ 1 .... \varphi _ { n - p-1}}^{\rho _ 1 ... \rho _ p\widehat{\rho } _
{ p+1}\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial _ { \widehat{\sigma } _
{ p+1}}\delta _ \epsilon \left ( y - z_3\right ) \ ; , \nonumber \\
\label{13-reg}\end{aligned}\ ] ] and @xmath108 } ^{\sigma _ 1 .... \sigma _ p}\delta _ { \left [ \nu _ 1 .... \nu _ { n - p-1}\right ] } ^{\beta _ 1 .... \beta _ { n - p-1}}\varepsilon ^{\theta _ 1 ... \theta _ { p+1}\alpha _ 1 ... \alpha _ { n - p-1}}\partial _ { \theta _ { p+1}}\delta \left ( x - y\right ) \nonumber \\ & & \varepsilon
_ { \theta _ 1 ... \theta _ p\widehat{\theta } _ { p+1}\alpha _ 1 ... \alpha _ { n - p-1}}\partial ^{\widehat{\theta }
_ { p+1}}\delta \left ( x - y\right ) \varepsilon _ { \sigma _ 1 ...
\sigma _ { p+1}\varphi _ 1 ...
\varphi _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\sigma _ { p+1}}\delta \left ( x - z_3\right )
\nonumber \\ & & \varepsilon _ { \xi _ 1 ...
\xi _ { p+1}\beta _ 1 ... \beta _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\xi _ { p+1}}\delta \left ( y - z_4\right ) \varepsilon _ { \xi _ 1 ...
\xi _ p\widehat{\xi } _ { p+1}\varphi _ 1 ...
\varphi _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\widehat{\xi } _ { p+1}}\delta \left ( z_3-z_4\right ) \nonumber \\ & & \varepsilon ^{\eta _ 1 ... \eta _ { p+1}\gamma _ 1 ...
\gamma _ { n - p-1}}\partial _ { \eta _ { p+1}}\delta _ \epsilon \left ( z_3-z_4\right ) \varepsilon _ { \eta _ 1 ... \eta _ p\widehat{\eta } _ { p+1}\gamma _ 1 ...
\gamma _ { n - p-1}}^{\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,}\partial ^{\widehat{\eta } _ { p+1}}\delta _ \epsilon \left ( z_3-z_4\right ) \;.
\nonumber \\ \label{14-reg}\end{aligned}\ ] ] all terms in all possible diagrams may then be seen to be proportional to the @xmath0-dimensional delta function @xmath98 or its derivatives .
one may easily convince oneself that this mechanism also applies to any order in perturbation theory .
since we always have @xmath109 , these diagrams all amount to a null correction to the basic diagram , so that the correlation function @xmath68 gives the generalized linking number of @xmath77 and @xmath70 @xmath110
we have been able to prove that the correlator @xmath68 is unaffected by radiative corrections to all orders , provided that @xmath77 and @xmath78 are two smooth closed nonintersecting surfaces .
although we have carried out explicit computations for the interaction term @xmath111 , the same result may be achieved straightforwardly for any local interaction containing arbitrary powers of the curvatures @xmath112 .
this result , which generalizes to arbitrary dimensions that obtained in the case of two nonintersecting curves @xmath90 , may be interpreted as a kind of nonrenormalization property of the linking number in higher dimensions .
we are indebted to the departamento de campos e partculas ( dcp ) of the centro brasileiro de pesquisas fsicas ( cbpf ) for kind hospitality .
the conselho nacional de pesquisa e desenvolvimento ( cnpq / brazil ) , the fundao de amparo pesquisa do estado do rio de janeiro ( faperj ) and the sr2-uerj are gratefully acknowledged for financial support . | the topological antisymmetric tensor field theory in @xmath0-dimensions is perturbed by the introduction of local metric dependent interaction terms in the curvatures .
the correlator describing the linking number between two surfaces in @xmath0-dimensions is shown to be not affected by the quantum corrections . |
entanglement is at the heart of many paradoxes of quantum mechanics and is also a useful ressource in quantum information processing . in particular ,
entangled states play a key role in quantum error correction and in quantum computational speedup .
the characterization of entanglement of a general multipartite state remains a mathematical challenge and there exist several inequivalent measures .
a comprehensive account of the most important mathematical concepts used for quantifying and manipulating entangled states may be found in @xcite .
some properties of entangled states may also be approached from differential geometry ( two- and three - qubit hilbert space geometry corresponds to hopf fibrations , that are entanglement sensitive @xcite ) , from topological spaces ( for instance the @xmath15-qubit ghz state corresponds to a borromean ring @xcite ) , from finite geometries ( by using the hyperplanes of the relevant generalized quadrangles @xcite or in the black - hole analogy @xcite .
besides , group theoretical concepts have been repetitively put forward for describing classical @xcite and quantum error correction @xcite . in this paper , we exploit the unnoticed but organic link between real entangling gates of quantum computation and the group of automorphisms of some highly symmetric even euclidean lattices .
let us define an @xmath11-dimensional euclidean lattice @xmath8 as a discrete additive subgroup of the real vector space @xmath0 , endowed with the standard euclidean product , and spanned by a generator matrix @xmath16 with rows in @xmath0 .
the automorphism group @xmath7 is the set of orthogonal matrices @xmath9 such that under the conjugation action by the generating matrix @xmath16 , one gets a unimodular matrix @xmath17 , that is ( i ) @xmath18 and ( ii ) @xmath19 is an integer matrix ( see @xcite , p. 90 ) .
when the degree @xmath11 is equal to the number of basis elements of @xmath8 , then @xmath12 also acts on basis vectors and is generated with matrices @xmath9 such that the sum of squared entries in a row is one , i.e. @xmath9 may be seen as a quantum gate . it will be shown that for a suitable choice of even euclidean lattices @xmath8 such as @xmath20 lattices , the root lattices @xmath1 and @xmath2 , the barnes - wall lattice @xmath21 , the lattice @xmath4 and the leech lattice @xmath5 , one observes that the orthogonal group @xmath7 is a group of entangling _
quantum gates_. it can be considered as acting on a two- , three- , four- , a two - qubit / qutrit and a three - qubit / qutrit system , respectively .
in retrospect , such lattices bear out our recent work relating @xmath1 and @xmath2 weyl groups to entangled states arising from the joint basis of specific mutually commuting set of generalized pauli observables @xcite .
to illustrate our topic , let us start with the @xmath20 lattice whose lattice points are the integers and whose basis @xmath16 is the identity matrix .
the elementary cell is the @xmath11-dimensional hypercube and the automorphism group is a wreath product @xcite @xmath22 of order @xmath23 , where the symbol @xmath24 denotes the semidirect product and the wreath product @xmath25 corresponds to a permutation action of the symmetric group @xmath26 on the @xmath11 copies of the two - letter group @xmath27 .
it is straightforward to check that the automorphism group @xmath28 of the four - dimensional lattice contains the gate cnot=@xmath29 , that is well known to create two - qubit entanglement .
more precisely , taking the input qubit states as @xmath30 and @xmath31 , the action cnot@xmath32 ends up into the non - separable bell state @xmath33 .
as it is well known , the cnot gate is a primitive of universal quantum computing .
it has been found @xcite that the entanglement arising from the two - qubit pauli group is encapsulated in the following mermin s square @xmath34 where @xmath35 , @xmath36 and @xmath37 are the ordinary pauli spin matrices .
every row and every colum in the above square consists of mutually commuting operators sharing a basis of entangled states .
the mermin s square may be used for a proof of the kochen - specker theorem in a four - dimensional space by observing that the algebraic relations for the eigenvalues @xmath10 of the observables contradicts the one for the eigenstates @xcite .
let us denote @xmath38 and @xmath39 the first row and the r.h.s .
column in ( [ msquare ] ) , the only ones that possess real entries .
one may use the common eigenstate basis @xmath40 of the set @xmath41 for generating a lattice isometric to @xmath42 . the natural action of the automorphism group on the basis @xmath43 reads . ) ; l:=latticewithbasis(mr ) ; aut:=automorphismgroup(l : naturalaction ) ; l , aut ; " ] @xmath44 it is noticeable that the rows of the generators @xmath45 and @xmath46 of @xmath7 encode the eigenstates of the triples of observables @xmath47 and @xmath48 , respectively ( a partial list of the eigenstates , associated to the maximal sets of mutually commuting operators in the two - qubit pauli group , may be found in @xcite ) . then
, following a similar reasoning , we construct the @xmath49 lattice by means of a generating basis built from the eigenstates of the following triple of three - qubit observables @xmath50 that follows from @xmath47 by applying the left tensor product @xmath37 .
the basis we select is the matrix @xmath51 first introduced in @xcite ( eq . ( 9 ) ] . as for the two - qubit case above
, the natural action of the automorphism group on the basis is obtained as @xmath52 the matrix @xmath53 is not made explicit but it has a similar form and properties as @xmath54 .
let us now investigate the type of entanglement contained in the @xmath15-qubit generators / gates @xmath54 and @xmath53 .
we single out the state encoded by the fourth row of @xmath54 @xmath55 a quantitative measure of tripartite entanglement is the @xmath15-tangle @xcite @xmath56 which contains the four determinants + @xmath57 , @xmath58 , + @xmath59 and @xmath60 . to quantify bipartite entanglement the strategy is as follows @xcite .
take the density matrix @xmath61 of the @xmath15-qubit system and trace out over the bipartite subsystems to obtain a reduced density matrix @xmath62 .
a measure of two - qubit entanglement is the tangle @xmath63 , with the concurrence @xmath64 , where the @xmath65 are non - negative eigenvalues of the product @xmath66 arranged in decreasing order , @xmath67 is the spin - flipped density matrix and @xmath68 denotes the complex conjugate .
two important families are the ghz family with @xmath69 and all @xmath70-tangles @xmath71 vanishing , and the @xmath72 family with @xmath73 and all @xmath70-tangles equal to 1 .
both families are unequivalent under slocc transformations @xcite .
the @xmath15-tangle of the state ( [ state ] ) is @xmath74 and the reduced density matrices for subsystems @xmath75 , @xmath76 and @xmath77 are @xmath78 , @xmath79 and @xmath80 .
the set of eigenvalues of @xmath66 associated to subsystems @xmath75 and @xmath76 are @xmath81 leading to the tangles @xmath82 and the set of eigenvalues of @xmath66 associated to the subsystem @xmath77 is @xmath83 leading to a vanishing tangle @xmath84 .
all states encoded by rows of @xmath54 and @xmath53 behaves as state ( [ state ] ) .
we call a tripartite entanglement characterized by the tangles @xmath85 and @xmath86 an _ incomplete balanced entanglement_. fully balanced tripartite entanglement is found in @xcite and in the next section below devoted to the @xmath2 lattice and other dense lattices .
the family of barnes - wall lattices @xmath87 , of even dimension @xmath88 , is related to self - dual codes and the real clifford group .
the first four members of the family , for @xmath89 to @xmath90 , are isomorphic to the densest known lattices of the corresponding dimension , i.e. the root lattices @xmath91 , @xmath1 , @xmath2 and the barnes - wall lattice @xmath3 .
the automorphism group of @xmath87 ( in dimension @xmath88 , @xmath92 ) corresponds to the real clifford group @xmath93 ( defined as the normalizer in the general orthogonal group @xmath94 of the generalized pauli group on @xmath11 real qubits : the notation comes from @xcite ) . for @xmath95 qubits , @xmath96 is @xmath2 and @xmath97 is the second largest maximal subgroup of @xmath98 , the derived subgroup of the weyl group @xmath6 . in the sequel , we concentrate on the entanglement arising from the lattices embodying two , three and four qubits , respectively . among the several representations of the lattice @xmath1 , we select the one with basis @xmath99 , that has the automorphism group @xmath100 going back to ( [ autd4 ] ) , it is clear that the rows of the two generators of @xmath101 encode the ( entangled ) eigenstates of the triple @xmath47 ( as was the case for the generator @xmath45 ) .
thus , two - qubit maximal entanglement is at the heart of the symmetries of this representation on the lattice @xmath1 . the group @xmath101 , of order @xmath102 , is ( up to isomorphism ) the same object than the two - qubit real clifford group @xmath103 and the weyl group @xmath104 of the @xmath105-cell . among the several representations of the lattice @xmath1
, we select the one with basis @xmath106 , @xmath107 that has the automorphism group @xmath108 the group @xmath109 is of order @xmath110 .
let us consider the state encoded by the first row of @xmath53 @xmath111 the tangles attached to the state ( [ stateb ] ) are calculated , as we did for the state ( [ state ] ) of the previous section .
they are such that @xmath112 corresponding to fully _ balanced tripartite entanglement_. the same results holds for all states encoded by the rows of @xmath54 and @xmath53 .
one can conclude that this type of entanglement , first featured in @xcite , is a specific property of the lattice @xmath2 .
as mentioned previously , the @xmath15-qubit real clifford group @xmath97 of order @xmath113 is a maximal subgroup of @xmath114 ( the second largest one ) ; the largest maximal subgroup is the automorphism group @xmath115 of the lattice @xmath116 . as a second representation of the lattice @xmath117 , we select the one obtained from the construction a " on the extended hamming code @xmath118 $ ] @xcite .
@xmath119 here we get two types of generators . in the generator @xmath54 ,
the rows @xmath90 and @xmath120 encode non entangled states while the other rows encode a state such as @xmath121 with @xmath69 and vanishing @xmath70-tangles .
such a state is of the ghz type . the second generator @xmath53 encodes states of the type ( [ state ] ) corresponding to incomplete balanced entanglement .
for four real qubits , the clifford group is @xmath122 , of order @xmath123 , associated to the barnes - wall lattice @xmath21 @xcite .
we use construction b " associated to the @xmath124 $ ] reed - muller code ( see @xcite , p. 141 ) .
one gets two @xmath125 gates generating @xmath126 with rows encoding entangled states such as the factorizable state @xmath127 it is easy to quantify the residual tripartite entanglement of state ( [ state3 ] ) . using ( [ three ] )
, one gets true tripartite entanglement @xmath69 corresponding to the ghz family .
all states encoded by the rows in the generators of @xmath126 display vanishing @xmath90-tangle ( the four - tangle is calculated from the expressions given in @xcite ) . presumably , all residual three - tangles equal unity as for ghz - type states ) .
the unique indecomposable @xmath128-dimensional unimodular lattice is @xmath4 .
it can be generated from the construction a " applied to the @xmath129 $ ] extended golay code over @xmath130 .
the automorphism group is of order @xmath131 . as for the previous cases , the rows of the generators encode entangled states . if we look at the system as a bipartite @xmath132 system , the schmidt rank with entries in an euclidean ring ( a field @xmath133 , the ring @xmath134 of integers , the ring @xmath135 $ ] of polynomials with coefficients over a field @xmath133 ... ) . given an @xmath136 matrix @xmath137 over an euclidean ring @xmath138 , one can find invertible square matrices @xmath139 and @xmath140 such that @xmath141 , where @xmath45 is a diagonal matrix : the smith normal form of @xmath137 .
this is equivalent to the singular value decomposition @xmath142 .
the non - zero diagonal entries of the smith form of @xmath137 are unique up to multiplication by a unit of @xmath138 and are also called elementary divisors . thus , the schmidt rank of @xmath137 equals the rank of @xmath45 .
] of the generators is found to be three .
there are several types of states .
we restrict our account to one of them , that we consider as the state of a @xmath143 system , i.e. a qutrit coupled to two qubits .
@xmath144 the reduced density matrix of the two - qubit bipartite system is found to be @xmath145 the set of eigenvalues of @xmath66 associated the two - qubit subsystem @xmath77 is obtained as @xmath146 , corresponding the two - tangle @xmath147[multiblock footnote omitted ] .
then , one can compute the reduced density matrix @xmath148 and @xmath149 ; they correspond to a qutrit / qubit system . according to the peres criterion @xcite of a bipartite system a - b with dimension @xmath150 is separable iff @xmath151 has non negative eigenvalues . for @xmath152 ,
if the partial transpose matrix @xmath153 has negative eigenvalues , then the bipartite system @xmath75 is entangled .
but they are states , known as bound entangled states , that satisfy the peres positivity criterion , but still are entanged . ] , a necessary and sufficient condition for the existence of entanglement between parties @xmath137 and
@xmath9 is that the partial transpose matrix @xmath153 contains at least one negative eigenvalue .
a similar criterion holds for the entanglement between parties @xmath137 and @xmath154 .
the sets of eigenvalues of matrices @xmath155 are @xmath156 and @xmath157 , respectively .
these calculations ensure that bipartite entanglement is spread over all parties .
the densest known lattice in dimension @xmath105 is the leech lattice @xmath5 with kissing number @xmath158 .
it is the unique euclidean lattice that is even , unimodular and with non - zero vectors at least two .
the automorphism group is called @xmath159 , where the conway group @xmath160 is sporadic , with order @xmath161 one elegant construction of @xmath5 makes use of the mog ( miracle octad generator ) coordinates , that are obtained from the codewords of the hexacode ( @xcite , p. 132 ) . in this representation , @xmath162 is entangled .
one can see the states encoded by the rows of the two generators as coming grom a @xmath163 quantum system and for them one finds that the schmidt rank is four .
the amount of entanglement that we find with this approach is quite inhomogeneous .
we restrict to a single example , the two - qubit / sextit state @xmath164 with reduced density matrix @xmath165 and eigenvalues of @xmath66 as @xmath166 .
it follows that the two - tangle for this states is @xmath167 .
it can be shown that the partial traces @xmath153 and @xmath168 violate the peres positivity criterion so that there also is entanglement between the parties @xmath75 and @xmath76 .
it would be desirable to analyze the entanglement in @xmath5 as arising from a three - qubit / qutrit system but the methods to achieve this goal seem not yet to be available .
the symmetry group acting on the generating basis of even euclidean lattices has been described as a group of entangling quantum gates .
barnes - wall lattices are known to be related to the clifford group of quantum error correction @xcite and to discretised hilbert spaces @xcite but the occurrence of leech lattice is novel in this context .
it may be that real world systems , like quark systems , or codons , or even black holes @xcite , exploit dimension @xmath169 as three - qubits and qutrits at the same time .
leech lattice may also be constructed from leech roots in lorentzian space @xmath170 ( see @xcite , chaps .
26 - 27 ) and dimension @xmath171 corresponds to the original version of string theory through the no - ghost theorem .
these new vistas ( see also @xcite ) , and their connection to tripartite entanglement , will be explored in future work .
the author acknowledges bernd souvignier for his reading of an earlier version of the manuscript and r. mosseri for his feedback about the geometry of the discretised hilbert space . the work started in december 2009 during a visit at de brn center for computational algebra in galway under the invitation of michael mc gettrick .
it was pesented at international conference on quantum optics and quantum computing ( icqoqc-11 ) in noida and at the 18th central european workshop on quantum optics ( cewqo 2011 ) in madrid .
10 mosseri r and dandoloff r 2001 geometry of entangled states , bloch spheres and hopf fibrations .
_ j. phys .
a : math . gen . _ * 34 * 10243 ; + mosseri r and ribeirop 2007 entanglement and hilbert space geometry for systems with a few qubits .
struct . in comp
. science _ * 17 * 1117 .
aravind p k , in : r.s .
cohen et al ( eds ) 1997 _ potentiality , entanglement and passion - at - a - distance _
( kluwer academic publishers ) p. 53 ; + sugita a 2007 borromean entanglement revisited .
preprint 0704.1712 ( quant - ph ) .
calderbank a r , rains e m , schor p w and sloane n j a 1998 quantum error correction via codes over gf(4 ) _ ieee trans .
th . _ * 44 * 1369 ; + planat
m and sol p 2009 clifford groups of quantum gates , bn - pairs and smooth cubic surfaces .
_ j. phys .
a. : math . theor _ * 42 * 042003 .
bosma w , cannon j and playoust c 1997 the magma algebra system i. the user language _ j. symbolic comp .
_ * 24 * 235 ; + plesken w and souvignier b 1997 computing isometries of lattices .
_ j. symbolic comp . _ * 24 * 327 . | the group of automorphisms of euclidean ( embedded in @xmath0 ) dense lattices such as the root lattices @xmath1 and @xmath2 , the barnes - wall lattice @xmath3 , the unimodular lattice @xmath4 and the leech lattice @xmath5 may be generated by entangled quantum gates of the corresponding dimension .
these ( real ) gates / lattices are useful for quantum error correction : for instance , the two and four - qubit real clifford groups are the automorphism groups of the lattices @xmath1 and @xmath3 , respectively , and the three - qubit real clifford group is maximal in the weyl group @xmath6 . technically , the automorphism group @xmath7 of the lattice @xmath8 is the set of orthogonal matrices @xmath9 such that , following the conjugation action by the generating matrix of the lattice , the output matrix is unimodular ( of determinant @xmath10 , with integer entries ) . when the degree @xmath11 is equal to the number of basis elements of @xmath8 , then @xmath12 also acts on basis vectors and is generated with matrices @xmath9 such that the sum of squared entries in a row is one , i.e. @xmath9
may be seen as a quantum gate . for the dense lattices listed above
, maximal multipartite entanglement arises .
in particular , one finds a balanced tripartite entanglement in @xmath2 ( the two- and three- tangles have equal magnitude @xmath13 ) and a ghz type entanglement in bw@xmath14 . in this paper
, we also investigate the entangled gates from @xmath4 and @xmath5 , by seeing them as systems coupling a qutrit to two- and three - qubits , respectively . apart from quantum computing
, the work may be related to particle physics in the spirit of @xcite . |
the majority of the known stellar mass black hole candidates ( bhcs ) are transient low - mass x - ray binaries .
these systems remain in quiescence for periods of months to tens of years and occasionally show outbursts ( x - ray novae ) which are the result of a sudden increase in the mass accretion rate that generally lasts for a few months . during an outburst various spectral and variability properties
are observed which seem to a great extent , but not uniquely , to be determined by the mass accretion rate ( see e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
although the appearance of black hole x - ray binaries is often categorized into 4 to 5 canonical states , they in fact appear to go through a continuous range of states whose properties are determined by the interplay between the two dominating spectral components : a soft thermal disk component and a hard non - thermal power law component attributed to a comptonizing medium @xcite whose origin remains unclear but is often thought to be an outflow and/or corona surrounding the inner accretion disk .
the variability properties of bhcs are strongly related to the relative contribution of the disk and the non - thermal components .
when the latter dominates ( hard state ) , the power spectra of bhcs show a strong band limited noise component on which often are superposed one or more quasi - periodic oscillations ( qpos ) .
when the contribution of the non - thermal component to the spectrum decreases ( intermediate state ) , the noise becomes weaker and the qpos become more prominent , while their frequencies increase from typically 0.1 to 10 hz . when the thermal component dominates the spectrum ( soft state ) variability is very weak .
in addition to low frequency qpos ( lf qpos ; @xmath00.110 hz ) five bhc systems have also shown high frequency qpos ( hf qpos ) with frequencies of 40 hz and higher ( for a recent overview see * ? ? ?
they are mainly observed when neither the thermal nor the non - thermal component dominates the spectrum . unlike the khz qpos that are observed in the neutron star low - mass x - ray binaries @xcite and whose frequencies are variable and correlated to those of qpos at lower frequency , our view on hf qpos in bhcs is less complete and has evolved considerably during the past six years .
currently , three of the five sources have shown pairs hf qpos , sometimes simultaneously detected : gro 165540 shows a pair at 300 and 450 hz ( * ? ? ?
* ; see also * ? ? ?
* ) , grs 1915 + 105 at 41 and 67 hz ( * ? ? ?
* ; see also * ? ? ? * ) and another at 164 and 328 hz ( * ? ? ?
* not detected simultaneously with the other pair ) , and xte j1550564 at 184 and 272 hz @xcite .
note that in the case of xte j1550564 there is also evidence for a peak at 92 hz @xcite and that single peaks at intermediate frequencies are also found @xcite .
the frequencies of the 102284 hz peaks in xte j1550564 reported by @xcite seemed to correlate with the frequency of the lf qpos , although @xcite dispute this . in xte j1859 + 226
@xcite and 4u 163047 @xcite only single peaks are detected that are consistent ( at a @xmath2 level ) with having a constant frequency . although several detailed models had already been proposed to explain the frequencies of the hf qpos ( some of which will be discussed in section [ sec : disc ] ) , recently much interest has been sparked by the apparent harmonic ratios of the frequencies in the hf qpo pairs ( 3:5 and 1:2 in grs 1915 + 105 , 2:3 in gro j165040 , and 1:2:3 xte j1550564 ) , which , according to @xcite , might be suggestive of resonances between orbital and epicyclic motion in the kerr metric . in any case , the frequencies of the hf qpos strongly suggest that they find their origin in the inner parts of the accretion flow , close to the innermost stable orbit ( isco ) .
when eventually the mechanism behind these oscillations is understood , their properties can be used to study the effects of general relativity and constrain the mass and spin of the black hole .
already now such constraints can be made ; for example , assuming that the highest observed frequencies are caused by keplerian motion at the isco , for at least two of the above sources , gro j165540 and xte j1550564 , the derived values of the masses of the black holes ( 5.57.9 @xmath1 and 9.711.6 , * ? ? ?
* ; * ? ? ?
* ) and the highest observed frequencies imply that the black holes _ must _ be spinning .
the models proposed thus far to explain the frequencies of the hf qpos have all in common that their interpretation of the highest observed frequency requires even larger spin values than the ones based on the above simple assumption . in this paper
we report our search for high frequency qpos in the x - ray transient xte j1650500 .
this source was first detected with the rossi x - ray timing explorer ( rxte ) on september 5 2001 @xcite .
follow up observations @xcite revealed a hard x - ray spectrum , strong variability and qpos around a few hz , marking the source as a black hole candidate .
optical and radio counterparts were identified by @xcite and @xcite .
optical follow - up observations performed when the source was close to quiescence , in october 2002 , revealed an orbital period of 0.212 days and a mass function which requires an inclination @xmath3 for the black hole mass to be larger than 3 @xmath1 @xcite .
xmm - newton observations performed near the peak of the outburst revealed the presence of a broad , skewed fe k@xmath4 emission line that suggests the compact object in this system may be a kerr black hole @xcite .
@xcite studied the variability properties during a @xmath020 day interval centered around a transition from the soft to the hard state that occurred on 2001 november 18 . in the power spectra they found lf qpos between 4 and 9 hz , but also broad bumps around 25 and 80 hz .
no hf qpos were found , however . during the later stages of the outburst
a 14-day period was found in the asm light curves , which is probably the disk precession period @xcite . in this paper
we report the detection of high frequency variability in this source and in particular the detection of a 250 hz qpo . a more detailed spectral and variability study of the outburst
will be presented elsewhere .
the data used in this paper were taken with the proportional counter array ( pca , * ? ? ?
* ) onboard rxte @xcite .
the observations of xte j1650500 were performed between 2001 september 6 and 2001 november 22 .
a total of 85 pointed observations were made , amounting to @xmath0185 ks .
the data used in our analysis were in the following modes : standard 2 , which has a time resolution of 16 s and covers the 260 kev pca effective energy range with 129 energy channels , and e_125us_64m_0_1s / e_64us_64m_0_1s ( event modes ) , which provide a time resolution of @xmath5 s or @xmath6 s in 64 energy channels covering the same energy range as the standard 2 mode .
the standard 2 data were used to create a light curve , a hardness - intensity diagram ( hid ) and a color - color diagram ( cd ) .
only data from proportional counter units ( pcus ) 0 and 2 were used since these were the only two ( of the five ) that were always operational during our observations .
the observations were background subtracted , but no dead time corrections were applied ( @xmath73.6% ) .
two colors were defined ; a soft color , which is the ratio of counts in the 5.810.0 kev and 2.55.8 kev bands ( standard 2 [ i.e. 0128 ] channels 1019 and 29 ) , and a hard color , which is the ratio of counts in the 13.818.4 kev and 2.55.8 kev bands ( channels 2939 and 29 ) .
the intensity used for the hid was the count rate in the 2.520.5 kev band ( channels 244 ) .
fast fourier transforms ( ffts ) of the event mode data ( from all the active pcus ) were performed to create power spectra with a frequency range of 0.06254096 hz in the 7.522.2 kev band ( ` absolute ' [ i.e.0254 ] channels 15 - 49 ) .
this energy range was chosen since high frequency variability in black hole lmxbs is in general most significantly detected in a similar energy range - an additional advantage of this energy range is that it does not introduce the instrumental broad high frequency features found by klein - wolt , homan & van der klis ( in prep . ) in the lowest energy channels .
the data were not background subtracted and no dead time corrections were applied prior to the fft ; the effects of dead time were accounted for by our power spectral fit function ( see below ) .
power spectra were averaged based on time , colors or count rate , and normalized according to the recipe described in @xcite .
prior to fitting the power spectra the dead time modified poisson level was subtracted using the method of klein - wolt , homan & van der klis ( in prep . ) , which uses the analytical function from @xcite and @xcite as a basis .
they found that this function often fails to describe the high frequency part of the power spectrum where no contribution from the source is expected and , instead of improving the fit at high frequencies by varying the parameters of this function ( which leads to unacceptable values for e.g. the instrumental dead time and vle window ) , the zhang function is scaled to provide the best fit to the power spectrum in the 10004000 hz range . following a current trend initiated by @xcite @xcite we fit the resulting power spectra with a sum of lorentzians ,
each given by @xmath8^{-1}$ ] , where @xmath9 is the centroid frequency , @xmath10 the half - width - at - half - maximum , and @xmath11 the integrated fractional rms ( from @xmath12 to @xmath13 ) .
instead of @xmath9 and @xmath10 we will quote the frequency at which the lorentzian attains its maximum in @xmath14 , @xmath15 , and the quality factor , q , where @xmath16 and @xmath17 @xcite .
the fractional rms amplitudes quoted in this paper are the integrated power between 0 and @xmath13 .
errors on fit parameters were determined using @xmath18 .
upper limits on the strength of the lorentzians were determined by fixing the q and/or @xmath15 , but not the rms amplitude , to values similar to those obtained in another power spectrum and using @xmath19 ( 95% confidence ) .
, width=453 ]
figure [ fig : lc - hid ] shows the 2.520.5 kev light curve and hid of the outburst of xte j1650500 .
the first few observations were done during the final stages of the rise , when the source was very hard . after reaching a peak on september 10 ( day 3 ) a more or less exponential decline followed , lasting for @xmath08 days , during which the spectral hardness changed considerably .
the period between september 25 ( day 18 ) and october 7 ( day 30 ) , which clearly shows up as a dent in the light curve , marks a transition from the hard to the soft state - during this period the day to day variations in count rate and hardness were more erratic than before and after , while both show an overall decrease .
after this transition the changes were smoother , and around october 17 ( day 40 ) the count rate started to decline exponentially again .
a minimum in the hardness was reached around november 1 ( day 55 ) .
after november 16 ( day 70 ) the source showed a remarkable hardening accompanied by a strong increase in count rate .
part of the decline ( @xmath020 days centered on our last observation ) is discussed in @xcite .
our initial search for high frequency variability was conducted by selecting a few groups of power spectra based on their position in the hid with the aim to get an idea of how the variability changed along the track in the hid and also to increase the sensitivity to high frequency features .
six selections were made whose location in the light curve and hid is indicated by roman numerals i vi . the horizontal branch ( hereafter hard branch ) made up by groups i and ii was split in two parts since
power spectral properties are known to correlate well with spectral hardness and we wanted to avoid too much smearing of frequencies .
group iii represents the transitional zone discussed above , which stands out clearly in both the light curve and hid .
group iv encompasses all the observations in the vertical soft branch and groups v and vi represent the last two observations .
we note that , apart from the one between groups i and ii , all the group boundaries in the hid coincide with considerable changes in the shape of the low frequency part of the power spectrum .
figure [ fig : pds ] shows the power spectra of groups i iii and vi in a @xmath14 representation - the power spectra of groups iv and v are not shown since at high frequencies they were dominated by statistical errors .
a summary of the fit parameters can be found in table [ tab : fit ] .
the first three selections all show power at high frequencies , with the most eye - catching feature being the narrow qpo in selection iii .
the lorentzian fit to this peak gives a @xmath15of 250@xmath205 hz , a q - value of [email protected] , and an rms amplitude of 5.0@xmath21% ( 6.0@xmath22 detection ) , making xte j1650500 the sixth black hole candidate to show such coherent high frequency variations . in selections
i , ii , vi high frequency variability is found at lower frequencies ( @xmath15=53139 hz ) with a lower coherence ( q=0.251.4 ) , whereas only upper limits could be determined in selections iv and v. note that for the power spectrum of selection ii , two lorentzians were needed to fit the high frequency part , one with properties similar to that of selection i , and one that was much narrower and at a considerably higher frequency .
the properties of the @xmath050 hz and 250 hz features do not change significantly when using the non - scaled zhang function for the subtraction of the poisson level .
the thick solid line shows the best fit to the data .
the number of lorentzians used for each group was : 6 ( i ) , 8 ( ii ) , 6 ( iii ) , and 4 ( vi ) .
the individual components are shown as dotted lines ( noise component ) , dashed lines ( low frequency qpos ) , and thin solid lines ( high frequency qpos ) . in all cases the poisson level was subtracted . note that for display purposes additional rebinning was applied to the high frequency parts of the power spectra in order to reduce the statistical errors.[fig : pds],height=680 ] to determine whether the frequency of the high frequency variability changed within our selections , we made additional selections on hard and/or soft color . dividing group iii in two parts , based on position in the cd (
not shown ) , we found qpos with frequencies of 245@xmath205 hz ( [email protected] ) and 270@xmath2010 hz ( [email protected] ) , for the spectrally hard and soft part , respectively .
this frequency difference has a significance of 2.2@xmath22 . dividing the hard branch , consisting of groups i and ii , in four parts
, reveals that the broad feature around 50 hz has a frequency consistent ( @xmath23 ) with being constant .
the @xmath0140 hz feature in group ii seems to move from 109@xmath209 hz , in the hard part of that group , to 168@xmath2015 hz in the soft part .
both these peaks are significantly detected ( @xmath24 ) , have @xmath25 , and have frequencies that are more than @xmath2 apart .
although this might suggest a continuous change in frequency from the hard part of group ii to the soft part of group iii , we note that the four frequencies detected in the subselections of groups ii and iii are consistent ( within @xmath2 ) with frequencies that are 1/3 , 1/2 and 1 times the frequency of the qpo detected in group iii , i.e. 250 hz . unfortunately , dividing our data into smaller selections , to test the presence of intermediate frequencies did not result in significant detections .
the 250 hz qpo is only significantly ( @xmath24 ) detected in a few of the 33 observations of group iii .
combining those individual observations leads to a detection with a significance that is lower than that of the detection in the whole group , suggesting the qpo is also present in ( some of ) the other observations . the strength of the 250 hz qpo seems to increase with energy
; rms amplitudes of @xmath70.85% , 4.5% , and @xmath712.1% were found in , respectively , the 26.2 kev , 6.215.0 kev , and 15.060 kev bands .
the low frequency part of the power spectra also shows considerable changes between the different groups . except for the first observations , all power spectra of groups i and ii show low frequency qpos , whose frequency smoothly increases with hardness from @xmath01 hz to @xmath08 hz
the low frequency part of group iii generally show a qpo around 6 and/or 9 hz , although in a few observations no qpo is detected at all . in groups
iv and v no low frequency qpos are detected - they reappear in group vi with a frequency of @xmath05.5 hz . in general the frequencies and q values of the lorentzians increase when the spectrum softens and decrease when the spectrum hardens , although there is clearly no one - on - one relation between hard color and frequencies ( see figures [ fig : lc - hid]a and [ fig : pds ] ) .
although there is a trend for both the high and low frequency qpos to increase with hardness , the low number of simultaneous detections does not allow us to tell whether they show a tight correlation .
our detection of a 250 hz qpo makes xte j1650500 the sixth bhc to show such high frequency variations .
the properties like q and rms amplitude fall in the range observed for other sources . as was already mentioned in section [ sec : intro ] , hf qpos in black hole systems are predominantly observed when neither the non - thermal nor the thermal component completely dominates the energy spectrum - they have yet to be detected in the soft or hard state .
xte j1650500 seems to be no exception to this ; a preliminary analysis of the energy spectra shows that the 250 hz qpo was detected when the contribution of the thermal component in the 2200 kev range varied between @xmath050% and @xmath070% ( @xmath01070% when including the marginal detections at @xmath0110 and @xmath0170 hz ) .
interestingly , this group of observations also bridges the gap between two major branches in our hid , a hard branch ( groups i and ii ) and a soft branch ( group iv ) .
the erratic changes in count rate and hardness in this transitional zone , as opposed to the smooth changes in the soft and hard branches , indicate that part of the accretion flow had become unstable .
although it is not clear from our analysis what part of the accretion flow this would be , our results suggest that the hf qpos in xte j1650500 are enhanced when these instabilities in the accretion flow ( with a time scale of apparently hours to days ) occur .
in addition to the 250 hz qpo we have also found a broad feature around 50 hz . this feature is present in the hard branch where its frequency was consistent with being constant . at the moment
that the 250 hz qpo was detected this 50 hz disappeared , suggesting that at least to some extent the mechanisms behind them are connected . although the above might suggest that the broad 50 hz peak had evolved into the qpo at 250 hz , the two subselections of group ii suggest otherwise ; they show peaks at @xmath0110 and @xmath0170 , which are present simultaneously with the 50 hz peak .
the constant frequency of this broad peak is remarkable though , especially since the lf qpos ( and possibly also the hf qpos , see below ) show clear changes in their frequency .
four of the eight power spectra of xte j1650500 shown in figure 2 of @xcite also show broad peaks , three around 80 hz and one around 25 hz , and broad peaks at similar frequencies are also found in gx 3394 and cyg x-1 @xcite .
it is not clear whether these broad peaks are the same as the 50 hz peaks we found .
the frequency behavior of the 250 hz qpo itself is not clear either .
there are indications that the frequency increased from @xmath0110 hz to @xmath0270 hz with decreasing spectral hardness , but the four independent frequencies that were measured ( @xmath0110 hz , @xmath0170 hz , @xmath0245 hz , and @xmath0270 hz ) are also consistent with being 1:2:3 harmonics of each other . in a sense
this behavior is similar to what was found in xte j1550564 ; also in that source the frequency of the hf qpo seemed to increase with decreasing spectral hardness @xcite , while it was found later by @xcite that , when combining power spectra according to the type of lf qpo , the majority of the detected hf qpos appeared to have frequencies that were consistent with being 1:2:3 harmonics of each other ( with the relative strength of the harmonics being related to spectra hardness ) .
however , in xte j1550564 hf qpos were also found at frequencies that did not obey this harmonic relation .
we also note that in the case of xte j1650500 the relation between hf qpos and the type of lf qpo is not as strong as in xte j1550564 , since the power spectra of group iii showed considerable changes in the coherence and relative strength of the various low frequency components , while the frequency and harmonic content of the hf qpo did not change significantly .
most models proposed to explain hf qpos in black holes @xcite have in common that the oscillations are produced in the accretion disc . some of them predict stable frequencies @xcite , whereas others interpret them as the black hole counterparts of the khz qpos in neutron star sources , and try to predict and explain their frequencies with respect to those of the lf qpos .
unfortunately , the limited number of detections of hf qpos in this source does not allow us the make strong statements about any of the models .
stable frequencies are still possible , if one interprets the different frequencies as harmonics of each other as suggested by @xcite for the pairs of peaks in grs 1915 + 105 and gro j165540 . on the other hand , the different frequencies might also reflect real changes in the frequency .
as noted by other authors it might well be possible that more than one of the above mechanism are at work at once .
the broad and possibly stable feature around 50 hz has not been specifically addressed by the current models . if we simply interpret the maximum observed frequency of 270 hz qpo as being due to keplerian motion at the isco of a schwarzschild black hole , we derive a mass of 8.2 @xmath1 . for a black hole with a lower mass no spin is required ( and the radius of the keplerian orbit has to be larger than that of the isco ) , whereas for more massive black holes it is . since the properties of the broad , skewed fe k@xmath26 line detected with xmm - newton @xcite suggest that the black hole has a near maximum angular momentum
, the mass of the black hole might well be much larger than 8.2 @xmath1 . assuming a mass ratio for the companion star and the black hole of 0.1 , the derived mass function for the system @xcite
would then imply an inclination lower than 27@xmath27 , which is considerably lower than the values derived for grs 1915 + 105 ( @xmath28 * ? ? ?
* ) , gro j165540 ( @xmath28 * ?
* and references therein ) , and xte j1550564 ( @xmath29 * ? ? ?
if we assume that the non - modulated x - ray emission is isotropic , this suggests that the mechanism responsible for the hf qpos in black hole x - ray binaries does not produce highly beamed modulated radiation .
jh acknowledges support from cofin-2000 grant mm02c71842 .
jmm is grateful for the support of the nsf .
mk acknowledges support from the netherlands organization for scientific research ( nwo ) .
whgl is grateful for support from nasa . ,
r. a. , muno , m. p. , mcclintock , j. e. , & orosz , j. a. 2002 , in procs . of the 4th microquasar workshop , 2002
durouchoux , fuchs , & rodriguez , to be published by the center for space physics : kolkata , astro - ph/0208402 , y. & lewin , w. h. g. 1995 , in x - ray binaries ( cambridge astrophysics series , cambridge , ma : cambridge university press , |c1995 , edited by lewin , walter h.g . ; van paradijs , jan ; van den heuvel , edward p.j . ) , p. 126
, m. 1995 , in proceedings of the nato advanced study institute on the lives of the neutron stars , held in kemer , turkey , august 29-september 12 , 1993 .
editors , m.a .
alpar , u. kiziloglu , and j. van paradijs ; publisher , kluwer academic , dordrecht , the netherlands , boston , massachusetts , p. 301 | we report the detection of high frequency variability in the black hole x - ray transient xte j1650500 . a quasi - periodic oscillation ( qpo )
was found at 250 hz during a transition from the hard to the soft state .
we also detected less coherent variability around 50 hz , that disappeared when the 250 hz qpo showed up .
there are indications that when the energy spectrum hardened the qpo frequency increased from @xmath0110 hz to @xmath0270 hz , although the observed frequencies are also consistent with being 1:2:3 harmonics of each other . interpreting the 250 hz as the orbital frequency at the innermost stable orbit around a schwarzschild black hole leads to a mass estimate of 8.2 @xmath1 .
the spectral results by miller et al .
( 2002 , apj , 570 , l69 ) , which suggest considerable black hole spin , would imply a higher mass . |
the matrix element method is unique among the analysis methods used in experimental particle physics because of the direct link it establishes between theory and event reconstruction .
originally developed to minimize the statistical uncertainty in measurements of @xmath1events at the tevatron experiments d0and cdf @xcite , it has since been applied with great success to measurements of the top quark mass @xmath2 @xcite and also in the discovery of electroweak production of single top quarks @xcite .
the method can in principle be used for any measurement , with the largest gain compared to cut - based analysis techniques expected for processes involving intermediate resonances and leading to many - particle final states . in general , the matrix element method can be used to determine several unknown parameters ( theoretical parameters describing the physics processes measured as well as experimental parameters describing the detector response ) at the same time in one measurement , thus also allowing for a reduction of systematic uncertainties .
this paper presents the analysis method in general and also gives an example of how the determination of such additional parameters can be implemented .
recent measurements in @xmath1events containing one leptonic and one hadronic @xmath3decay ( `` lepton+jetsevents '' ) already exploit the known @xmath3mass to constrain the energy scale for light - quark jets and significantly reduce the main systematic uncertainty of early measurements of the top quark mass . among the largest remaining systematic uncertainties
is the uncertainty on potential differences between the energy scales @xmath4and @xmath5for @xmath0- and light - quark jetsand @xmath5denote scale factors relative to the nominal detector calibration . in the definition of these factors
, it is assumed that all experimental corrections to measured jet energies have been applied to the events that enter the analysis , such that @xmath4and @xmath5do not depend on quantities like the jet direction or energy .
uncertainties on such dependencies give rise to additional systematic uncertainties of the measurement e. g. of the top quark mass . ] . without further improvements
, it will soon become a limiting uncertainty for those measurements that dominate the world - average value @xcite . in this paper
we show how together with the top quark mass , a simultaneous additional measurement of the @xmath0-quark jet energy scale , which was first proposed in @xcite for the lepton+jetschannel , can be incorporated naturally in the matrix element technique not only for measurements in the lepton+jetschannel , but also in @xmath1events with two leptonic @xmath3decays ( `` dilepton events '' ) where the quantities to be measured can not be reconstructed based on the kinematical information of a single event alone .
this study considers the case of the tevatron ( proton - antiproton collisions at a center - of - mass energy of @xmath6 ) as a concrete example but is applicable to the two lhc experiments atlas and cms as well .
the paper is structured as follows : in section [ method.sec ] , an overview of the matrix element method is given .
section [ samples.sec ] discusses the generation and selection of @xmath1events used for the studies described in the further sections .
this is followed by a discussion of the implementation of the likelihood calculation for signal and background processes for @xmath1measurements in the lepton+jetsand dilepton channels in section [ application.sec ] .
we then describe studies of the performance of the new measurement technique , separately for lepton+jets(section [ enstestljets.sec ] ) and dilepton ( section [ enstestdilepton.sec ] ) @xmath1events . in section [ systuncs.sec ] , systematic uncertainties on @xmath2are addressed , with an emphasis on the effect of events with significant initial- and final - state radiation .
section [ conclusions.sec ] summarizes the findings and gives an outlook .
the matrix element method is based on the likelihood @xmath7to observe a sample of selected events in the detector .
the likelihood is obtained directly from the theory prediction for the differential cross - sections of the relevant processes and the detector resolution and is calculated as a function of the assumed values for each of the parameters to be measured .
the minimization of @xmath8 yields the measurement of the parameters , where the likelihood @xmath7for the entire event sample is computed as the product of likelihoods to observe each individual event .
this is in contrast to most analysis methods used in experimental particle physics , where distributions from observed events in the detector are compared with corresponding distributions obtained from simulated events that have been generated according to theory and then passed through a detector simulation and the same event reconstruction software .
this paper concentrates on the case of the measurement of the top quark mass in lepton+jetsand dilepton events at a hadron collider , where the parameters to be measured are the top quark mass @xmath2 , factors @xmath4and @xmath5describing the energy scales for @xmath0- and light - quark jets relative to the default energy scaleevents , only @xmath0-quark jets occur and thus only @xmath4is determined .
only one overall scale factor @xmath4is determined for all @xmath0jets ; thus for example no distinction is made between @xmath0jets with and without a reconstructed muon inside the jet .
] , and the fraction @xmath9of signal events in the channel under consideration .
a comparison of the matrix element method with other methods to measure the top quark mass can be found in @xcite .
the sample likelihood @xmath7for @xmath10 measured events to have measured properties @xmath11 can be written as @xmath12 where the symbol @xmath13 denotes assumed values of the physics parameters to be measured , @xmath14 stands for parameters describing the detector response that are to be determined , and @xmath15 is defined below .
the likelihood @xmath16 to observe event @xmath17 under the assumption of parameter values @xmath13 , @xmath14 , and @xmath15 is given as the linear combination @xmath18 where the sum is over all individual processes @xmath19 that could have led to the observed event @xmath17 , @xmath20 is the likelihood to observe this event under the assumption that it was produced via process @xmath19 , and @xmath21 denotes the fraction of events from process @xmath19 in the entire event sample , with @xmath22 . in total , the physics parameters @xmath13 , the detector response described by @xmath14 , and the event fractions @xmath15 are to be determined simultaneously from the minimization of @xmath8 .
the likelihood @xmath23 in turn is given by the theoretical description of the process and the resolution of the concrete experiment , and is computed as the convolution of the differential partonic cross section with the parton distribution functions of the colliding hadrons and with the detector response . to make the likelihood calculation manageable , simplifying assumptions
are introduced .
this concerns the description of both the detector response and the physics processes @xmath19 , where only the dominant ones are considered explicitly , and where the effects of parton shower and hadronization are accounted for with a simple parametrization . because of these simplifications , the technique has to be calibrated using fully simulated events before applying it in an actual measurement on data . in this paper ,
a conceptual study is presented .
we show how the method can be validated with events that have been generated under the same assumptions as made in the likelihood calculation , which allows to demonstrate that the measurement method as such is unbiased . in equation ( [ pevt.eqn ] ) ,
not all likelihoods @xmath24 necessarily depend on all parameter values ; e.g. for the measurement of the top quark mass , the likelihood @xmath25 depends on the assumed top quark mass , while the likelihoods for an event to be produced via a background process ( which by definition does not involve top quark production or decay ) do not .
even if a likelihood does depend on a certain parameter , this dependency does not necessarily have to be taken into account explicitly ; for example , for the top quark mass measurement it will be shown in sections [ enstestljets.sec ] and [ enstestdilepton.sec ] that the dependency of the likelihoods for background processes on the jet energy scales can be neglected without introducing large biases on the top quark mass and energy scale measurements .
the individual contributions to the likelihood for an observed event @xmath26 to be produced via a given process @xmath19 are described in this section .
they are visualized schematically in figure [ psgn_schematic.fig ] for the example of a lepton+jets@xmath1event at the tevatron .
the observed event @xmath26 , shown at the right , is fixed while integrating over all possible momentum configurations @xmath27 of final - state particles .
the differential cross section for the process is convoluted with the probability for the final - state partons to yield the observed event ( transfer function ) , and with the probability to find initial - state partons of given flavor and momenta inside the colliding proton and antiproton ( parton distribution function ) .
all possible assignments of final - state particles to measured objects in the detector are considered by the transfer function . for each partonic final state under consideration , the initial - state parton momenta are determined by energy and momentum conservation .
schematic representation of the calculation of the likelihood to obtain a given observed lepton+jetsevent at a proton - antiproton collider ( similar figures apply to dilepton events and to other processes ) . ]
the likelihood for a final state with @xmath28 partons and given four - momenta @xmath27 to be produced in the hard - scattering process is proportional to the differential cross section @xmath29 of the corresponding process , given by @xmath30 where @xmath31 and @xmath27 stand for the kinematic variables of the partonic initial and final states , respectively .
the symbol @xmath32 denotes the matrix element for this process , @xmath33 is the center - of - mass energy squared of the collider , @xmath34 and @xmath35 are the momentum fractions of the colliding partons @xmath36 and @xmath37 ( which are assumed to be massless ) within the colliding proton and antiprotoncollider as a concrete example but is equally valid for the lhc when the antiproton is replaced with a proton and the appropriate pdf is used . ] , and @xmath38 is an element of @xmath28-body phase space . to obtain the differential cross section @xmath39 in @xmath40collisions , the differential cross section from equation ( [ dsigmahs.eqn ] ) is convoluted with the parton density functions ( pdf ) and summed over all possible flavor compositions of the colliding partons , @xmath41 where @xmath42 and @xmath43 denote the probability densities to find a parton of given flavor @xmath44 and momentum fraction @xmath45 in the proton and one of flavor @xmath46 and momentum fraction @xmath47 in the antiproton , respectively .
this equation reflects qcd factorization @xcite .
the finite detector resolution is taken into account via a convolution with a transfer function @xmath48 that describes the probability to reconstruct a partonic final state @xmath27 as @xmath26 in the detector , given the values @xmath14 of the parameters describing the detector response . the differential cross section to observe a given reconstructed event
@xmath26 then becomes @xmath49 only events that are inside the detector acceptance and that pass the trigger conditions and offline event selection are used in the measurement . to obtain a properly normalized likelihood ,
the overall cross section of events observable in the detector , @xmath50 is used , where @xmath51 for selected events and @xmath52 otherwise .
one then obtains @xmath53 as the ( differential ) likelihood that an event produced via process @xmath19 has measured properties @xmath26 ( and not other properties that would still lead to an event passing the event selection criteria ) .
this section describes a parametrization of the transfer function @xmath48 appropriate for measurements of high-@xmath54objects at a hadron collider .
the matrix element method is based on a fast parametrization that reproduces the basic properties of the detector .
any biases introduced can be determined ( and then corrected for ) when ensemble tests as described in section [ method.enstest.sec ] are performed with events generated with a full detector simulation , typically based on the geant @xcite package .
this section is written with measurements in @xmath1events in mind but is applicable to other processes as well .
the transfer function @xmath48 describes the probability density @xmath55 to reconstruct an assumed partonic final state @xmath27 as a measurement @xmath26 in the detector : @xmath56 because the final - state partons are assumed to result in some measured event @xmath26 , the normalization condition @xmath57 holds , where the integral is over all possible events @xmath26 .
effects due to selection cuts or finite detector acceptance are discussed in section [ method.norml.sec ] .
the transfer function is assumed to factorize into contributions from each measured final - state particle .
aspects to be considered in the transfer function are in principle the measurement of the momentum of a particle ( both of its energy and of its direction ) as well as its identification .
thus @xmath0-tagging information for the jets can be included , which can help to distinguish signal from background events . in many applications like the description of @xmath1events ,
a number of assumptions can be made @xcite about how final - state particles are measured in the detector , such that the dimensionality of the integration over the final - state particle phase space described in section [ method.pprc.sec ] is reduced .
individual particles can be described in the transfer function as follows : @xmath58 * isolated energetic electrons : * electrons are assumed to be unambiguously identified ( i.e. an electron is not reconstructed as a muon or a jet ) .
the electron direction and energy are both assumed to be well - measured , i.e. during integration , the final - state electron is assumed to be identical to the measured particle .
this is justified for @xmath1events since the resolution for electrons is far better than that for jets , and the jet energy resolution will dominate all effects due to the finite detector resolution .
* isolated energetic muons : * as for electrons , muons are assumed to be unambiguously identified , and their directions to be precisely measured .
however , instead of the energy the detector typically yields a measurement of @xmath59 , the muon charge divided by the transverse momentum .
consequently , the muon energy resolution can be poor for high-@xmath54 muons , and thus a transfer function @xmath60 allowing for a finite resolution is introduced . in the studies presented in this paper ,
the function @xmath61 is used to describe the likelihood that a muon with charge and momentum @xmath62(described by the _ mat_rix element ) is _ rec_onstructed with @xmath63 .
the resolution @xmath64 depends on the pseudorapidity @xmath65 to account for muon tracks at large @xmath66 that do not reach the full radius of the tracking detector .
the parameter values are taken from @xcite .
* energetic @xmath67 leptons : * events with energetic @xmath67 lepton decays are typically selected if the visible decay products pass a minimum energy cut . in this case , the directions of the visible decay products are close to that of the original @xmath67 lepton , but only a fraction of the @xmath67 energy can be measured in the detector . in this paper , only leptonic decays @xmath68 are considered , where the symbol @xmath69 denotes an electron or muon . consequently , a transfer function @xmath70 is introduced to describe the likelihood to obtain a charged lepton with a given energy fraction @xmath71 of the decaying @xmath67 lepton . for the study presented here ,
it is parametrized as a @xmath72-orderpolynomial as shown in figure [ fig : tautransferfunction ] .
the @xmath67 direction is taken to be well approximated by the direction of the reconstructed charged lepton .
likelihood for an electron or muon to carry a given energy fraction of the initial @xmath67 lepton energy.,scaledwidth=55.0% ] for muonic @xmath67 decays , the muon transfer function introduced above in principle has to be taken into account as well to describe the transition from the assumed to the reconstructed muon transverse momentum .
however , muons from @xmath67 decays typically have low enough transverse momenta so that the muon @xmath54 can be assumed to be well - measured in most applications . in the following
, the muon transfer function is omitted for muonic @xmath67 decays .
also , in this study we consider the reconstruction efficiency to be independent of the lepton energy .
* energetic quarks and gluons : * the directions of final - state quarks and gluons are assumed to be well - measured by the jet directions , and transfer functions are introduced for the jet energy measurement . the probability density for a jet energy measurement @xmath73 in the detector if the true quark energy is @xmath74 ( depending on the overall jet energy scale @xmath4or @xmath5 ) is given by the jet energy transfer function @xmath75 . in principle , different transfer functions apply to gluon jets and jets from different quark flavors @xmath76 . for @xmath77 ,
the jet transfer function is computed as @xmath78 where the factor @xmath79 in the denominator ensures the correct normalization in the absence of selection cuts . in this paper , the same jet energy transfer function is used to describe light - quark ( @xmath80 , @xmath81 , @xmath33 , and @xmath82 ) and gluon jetsselection cuts , gluon jets arise in background processes whose description is anyway only approximate ; therefore no separate transfer function for gluon jets is introduced . ] ; an independent transfer function is used for @xmath0-quark jets .
the parametrization of the transfer function follows that of the d0experiment given in @xcite , with parameters depending on the jet energy and pseudorapidity . in a fraction of those @xmath0jets that contain a semimuonic @xmath0-hadron decay ,
the muon is identified , and these jets could in principle be described with a separate transfer function @xcite ( while the jets with unidentified semileptonic decays would still have to be described with one function together with all other @xmath0jets ) . in this paper , only one class of @xmath0jets is considered , because the focus is to show how an energy scale for @xmath0jets can be determined at all , and only one overall energy scale factor @xmath4is determined for @xmath0jets .
once this is achieved , it will be possible in principle to determine two independent energy scales for the different classes of reconstructed @xmath0jets .
the ability of the detector to distinguish quarks from gluons and to identify the quark flavor is limited .
nevertheless , identification of @xmath0-quarks ( @xmath0-tagging ) can be useful to distinguish signal and background events , or to identify the correct assignment of final - state quarks to measured jets in final states like lepton+jets@xmath1events that contain both light and @xmath0-quark jets . in this paper ,
we follow the approach introduced in @xcite to include a term @xmath83 in the transfer function which describes the likelihood for parton @xmath84 with assumed flavor @xmath76 to be reconstructed with @xmath0-tagging information @xmath85 . if @xmath0 tagging is used as a binary decision , then one simply has @xmath86 where @xmath87 is the @xmath0-tagging efficiency for a jet from a parton of flavor @xmath88 .
typically , @xmath0-tagging algorithms yield a continuous output ( for example , the decay length significance of a secondary vertex within a jet , or the output of an artificial neural network ) . instead of a binary decision
, the quantity @xmath83 can be parametrized as a function of this continuous value .
such an approach naturally makes optimal use of the information . as a compromise
, it is possible to use several ( rather than only two ) bins in the output value .
this is the concept used in this publication .
figure [ btagging.fig ] shows the values @xmath83 for jets in lepton+jets@xmath1events as used in the study presented here , which corresponds to the @xmath0-tagging performance of the d0 experiment @xcite . for the study in this paper ,
the @xmath83 functions are assumed not to depend on the jet transverse momentum or pseudorapidity , but this will be a straightforward extension of the method for future measurements .
the function @xmath83 used to parametrize the @xmath0-tagging performance . the likelihood @xmath83 for a jet to be reconstructed as @xmath0-tagged with given @xmath0-tagging output @xmath89 is shown for light , @xmath82-quark , and @xmath0-quark jets . in this paper
, the output @xmath90 of an artificial neural network used in the d0experiment @xcite is taken as a concrete example .
the first bin contains jets that fail the @xmath0-tagging preselection .
the structure in the histogram is due to the non - equidistant binning.,scaledwidth=80.0% ] * energetic neutrinos : * neutrinos are not measured in the detector , but still an integration has to be performed over assumed values for all momentum components of all final - state neutrinos in an event . information on neutrino momenta can be partly inferred from mass constraints ( e.g. , @xmath91or @xmath2 in @xmath1events ) .
the additional assumption is made in this paper that events are balanced in the transverse plane , i.e. that the @xmath1system has zero transverse momentum .
this assumption is dropped in section [ systuncs.ttj.sec ] , which means that an integration over two additional variables has to be carried out .
the presence of neutrinos in an event is typically inferred from an imbalance in the transverse plane ( non - zero missing transverse momentum @xmath92 ) .
it is not straightforward to parametrize the resolutions of the two @xmath92components since they depend on the resolutions of all other reconstructed objects in the event .
instead , the vector sum of transverse momenta of all reconstructed objects that are not assigned to the final state in question could be considered . in the case of @xmath1events at the tevatron , this would be calorimeter measurements outside of the jets assigned to the @xmath1 final state . in this paper , however , as in @xcite , no corresponding transfer function factor is introduced .
in addition to the detector resolution , one has to take into account the fact that the particles measured in the detector can not be assigned unambiguously to specific final - state particles .
consequently , all possibilities must be considered , and their contributions to the transfer function summed .
the total transfer function can be written as @xmath93 where the four lines represent the contributions from electrons , muons , tau leptons , and jets , respectively .
it is understood that a term only appears if the corresponding particle is present in the final state under consideration .
the number of possible assignments of reconstructed ( `` rec '' ) particles to final - state particles in the process described by the matrix element ( `` @xmath94 '' ) is denoted by @xmath95 , and @xmath96 stands for one specific permutation .
the symbols @xmath97 , @xmath98 , @xmath99 , and @xmath100 stand for the numbers of electrons , muons , tau leptons , and quarks or gluons in the final state . a reconstructed particle is denoted by @xmath101 , @xmath102 , @xmath103 , or @xmath84 .
the symbols @xmath104 , @xmath105 , @xmath106 , and @xmath107 stand for the corresponding final - state particle assumed in the matrix element integration , which is given by the index @xmath96 of the permutation and the index of the reconstructed particle : @xmath108 ( and accordingly for muons , tau leptons , and jets ) .
the flavor @xmath109 of final - state parton @xmath107 assigned to jet @xmath84 is given by the permutation @xmath96 .
the jet energy scale appropriate for jet @xmath84 ( @xmath4or @xmath5 ) is denoted by @xmath110 and selected according to the assumed flavour @xmath109 . the symbol @xmath85 stands for any output from a @xmath0-tagging algorithm . because of the assumption that the transfer function factorizes into independent contributions from the final - state particles , it may in principle also depend on @xmath13 .
for example , in a top quark mass measurement , smaller top quark masses correspond to a smaller mean angular separation of jets , which may lead to a broadening of the jet energy resolution .
such effects are however typically small and are therefore neglected in the simplified description of the detector response with transfer functions ; they would implicitly be taken into account in a calibration of the measurement with fully simulated events .
the normalization condition for the likelihood @xmath111for each process is given by @xmath112 where the inclusion of the factor @xmath113 is equivalent to integrating only over those configurations @xmath26 of observed events that pass the event selection criteria .
the normalization condition is fulfilled according to the definition of the observable cross section @xmath114 in equation ( [ sigmaobs.eqn ] ) .
the calculation of @xmath114is intimately related with the normalization of the transfer function .
both aspects are discussed in this section .
the normalization condition of the jet energy transfer function used in previous implementations of the matrix element method @xcite is given by @xmath115 we call this a _ process - based _ normalization scheme as it reflects the concept that a final - state quark or gluon gives rise to a jet of any energy ( or is not reconstructed as a jet if the energy is below the jet reconstruction threshold of the experiment ) ; thus this normalization scheme does not depend on the event selection cuts .
a modified , _ selection - based _ normalization scheme which simplifies the computation of the normalization integral given in equation ( [ sigmaobs.eqn ] ) but leaves the likelihood @xmath111unchanged is introduced in this section .
we assume that the event selection requires a reconstructed object for every charged lepton and every quark or gluon in the final state ( this means for example that the presence of four jets is required for lepton+jets@xmath1events ) , and that the jet selection cuts are identical for all jets .
the selection - based normalization of the transfer function is based on the concept that events only enter the analysis if all these reconstructed objects pass the corresponding selection criteria .
this means that every possible partonic final state in the integral in equation ( [ dsigmapp.eqn ] ) is assumed to yield an observed event that passed the event selection .
thus , a modified jet energy transfer function @xmath116 is introduced in the top quark mass measurement , which satisfies the condition @xmath117 i.e. the parton under consideration is assumed to have led to a jet that passed the selection cut @xmath118 , where the energy cut normally depends on the polar angle of the jet since a transverse energy cut is used in the event selection .
equation ( [ tfnormjt.eqn ] ) ensures that in equation ( [ sigmaobs.eqn ] ) , @xmath119 it is shown in section [ method.norml.sigmaobs.sec ] that the modified denominator @xmath120 which is then needed in equation ( [ normpevtp.eqn ] ) to compute the likelihood @xmath111becomes independent of the parameters @xmath14 that describe the detector response .
the effect of this selection - based normalization scheme on the jet energy transfer function @xmath116 is shown in figure [ tfnormalization.fig](a ) for the double - gaussian function used in this study : if the parton energy is assumed to be very small , then a small reconstructed jet energy just above the cut value is most likely . however , the function is still normalized to unit area as it is assumed that the parton must have given rise to a jet that passed the selection cut ( in this example set at @xmath121 corresponding to @xmath122 for the event selection criteria of section [ samples.sec ] ) .
the dependence of the jet energy transfer function on the parameters @xmath14 that describe the detector response must also be accounted for as explained in figure [ tfnormalization.fig](b ) : for every @xmath5hypothesis the same event sample is considered in the measurement , and therefore the event selection ( in this example the minimum jet transverse energy cut ) can not depend on the assumed @xmath5value . for different assumed values of @xmath5 , the @xmath116 curve varies , and the normalization of the curve must be adjusted to ensure that the normalization condition in equation ( [ tfnormjt.eqn ] ) is satisfied . jet energy transfer function @xmath116 in the modified normalization scheme .
plot ( a ) shows the transfer functions for light - quark jets at @xmath122 for an assumed value of @xmath123 and assumed parton energies of @xmath124 ( red ) , @xmath125 ( blue ) , and @xmath126 ( green line ) .
plot ( b ) shows the same transfer function for different assumed @xmath5values ( violet : @xmath127 , blue : @xmath123 , and cyan : @xmath128 ) and an assumed parton energy of @xmath125.,title="fig:",scaledwidth=48.0% ] jet energy transfer function @xmath116 in the modified normalization scheme .
plot ( a ) shows the transfer functions for light - quark jets at @xmath122 for an assumed value of @xmath123 and assumed parton energies of @xmath124 ( red ) , @xmath125 ( blue ) , and @xmath126 ( green line ) .
plot ( b ) shows the same transfer function for different assumed @xmath5values ( violet : @xmath127 , blue : @xmath123 , and cyan : @xmath128 ) and an assumed parton energy of @xmath125.,title="fig:",scaledwidth=48.0% ] it is assumed that every @xmath67 lepton decays to an electron or muon that passed the event selection .
the @xmath67 energy has to be larger than the energy of the reconstructed lepton .
consequently , the selection - based normalization condition is @xmath129 because of the non - zero lower integration bound the transfer function @xmath130 has to be scaled with an appropriate overall factor that depends on the reconstructed lepton energy and pseudorapidity to arrive at the modified function @xmath131 . in comparison with the jet energy resolution ,
the muon transverse momentum resolution is good for muons close to the minimum transverse momentum cut , and only a negligible fraction of muons is affected by this cut .
in addition , it is assumed that final - state muons passing the selection cuts are always reconstructed as muons .
thus , a calculation of the normalization corresponding to equation ( [ tfnormtau.eqn ] ) can be omitted for muons . to derive the denominator @xmath132with which to normalize the likelihood @xmath111for a given process @xmath19 , it follows from equations ( [ normcond.eq ] ) , ( [ normpevtp.eqn ] ) , and ( [ dsigmapp.eqn ] ) that @xmath133 where the normalization condition for the modified transfer function @xmath134 ( equation ( [ tfnorm_wprime.eqn ] ) ) has been used in the last step .
the quantity @xmath132is thus only a function of the physics parameters @xmath13 , but not of the detector performance parameters @xmath14 . in the definition of @xmath132 ,
the integral over the observed events @xmath26 is not over the full phase space , but instead only over that part of the phase space that passes the kinematic event selection .
typically , regions of small jet transverse energy @xmath135or large @xmath66 will be excluded from the integration region . because the normalization of the jet energy transfer function @xmath116 described in section [ method.norml.tfnormjt.sec ] accounts for the lower jet @xmath135cut , any jet energy scale dependence of @xmath132is eliminated .
in contrast , the jet angular resolution is approximated with a @xmath136 function to save integration time , and this means that the integration over @xmath27 must exclude those angular regions that do not pass the event selection . through the angular acceptance cuts ( and through the matrix element
@xmath137 itself , of course ) , @xmath132still depends on the physics parameters @xmath13 .
a similar argument holds for angular acceptance cuts in the charged lepton selection .
the above argument is only valid if the normalization condition of equation ( [ tfnorm_wprime.eqn ] ) is fulfilled for the modified transfer function @xmath134 . in practice , this is difficult to implement for event selection cuts based on quantities that depend on more than one reconstructed particle . for example , the @xmath138cut in section [ samples.sec ] does not fulfill this criterion since it depends on all measured final - state particles .
therefore , an @xmath4dependence of @xmath139is taken into account in the analysis of dilepton @xmath1events described in section [ enstestdilepton.sec ] . for the measurement with lepton+jetsevents ,
it is shown in section [ enstestljets.sec ] that the @xmath4and @xmath5dependence of @xmath139can be neglected for the less stringent @xmath138cut applied in the event selection .
it is possible to choose a process - based transfer function normalization according to equation ( [ tfnormjt_alt.eqn ] ) . in this case , the dependency of the transfer function normalization on the parameters @xmath14 describing the detector resolution is not taken into account .
this means that in the last step of the derivation in equation ( [ sigmaobs_independent_of_beta.eqn ] ) , a dependency on @xmath14 remains and @xmath114has to vary as a function of both @xmath13 and @xmath14 .
this process - based scheme has the advantage that the transfer function @xmath3 accomodates the possibility of jets not passing the selection cuts . for analyses like @xmath2measurements in @xmath1events
as described in this paper , the number of jets required in the event selection ensures that in any event where this is the case an additional hard parton would have to be produced which yields a jet passing the cuts . in principle
, such events have to be described by the signal process . at the tevatron their contribution to the event sample is so small that they do not have to be accounted for explicitly in the method until the final calibration step .
consequently , the selection - based transfer function normalization described in sections [ method.norml.tfnormjt.sec ] and [ method.norml.tfnormmutau.sec ] is chosen for the studies described in this paper , as it eliminates the dependency of @xmath139 on the parameters @xmath14 and thus facilitates the simultaneous measurement of several parameters
. this picture may change for measurements in @xmath1events at the lhc , where initial - state radiation becomes much more relevant .
the normalization of the likelihoods can in principle be determined in the same way for all processes considered . alternatively , if the normalization of the likelihood @xmath140for one specific process @xmath141(for example , the signal process ) has been determined as described above and if the fraction @xmath142of events from that process in the selected sample is left as a free parameter in the analysis , it is possible to relate the absolute normalization of the likelihoods for all other processes to that of process @xmath141 .
one can then make use of the fact that the fit described in section [ method.mefit.sec ] will yield a signal fraction @xmath142of the sample that is too small if the background likelihood is too large and vice versa , and one can adjust the relative normalization in the validation of the matrix element method until the signal fraction is determined correctly .
this concept can only be applied if the cross - section for the process @xmath141is well - known ( like for example for @xmath1production ) .
it is then helpful in particular if the likelihoods for background processes as implemented in the analysis do not depend on any of the parameters @xmath13 and @xmath14 : in such a case , only one normalization constant needs to be determined for each background process . for a given sample of selected events ,
the parameters to be measured are determined as those values that maximize the likelihood @xmath7 .
one wants to determine @xmath143 physics parameters , @xmath144 parameters describing the detector resolution , and @xmath145fractions of events from different processes @xmath19 . for every measured event ,
the likelihoods for each process are calculated for an @xmath146-dimensional grid of assumed parameter values . given these grids of likelihood values for each process
, the sample likelihood @xmath147 defined in equation ( [ lsample.eqn ] ) is available for an @xmath148-dimensional grid of assumed parameter values . the measurement value of a given parameter @xmath149 and the corresponding uncertainty
are then determined from a one - dimensional likelihood @xmath150 .
the value of @xmath150 is obtained by marginalization of all other parameters ; in practice , this is done by keeping the value of @xmath149 constant , varying the assumed values of all @xmath151 other parameters , and taking the maximum @xmath7value . the one - dimensional function @xmath152 is fitted with a parabola .
the parameter value that minimizes the parabola is taken to be the measurement value , and the measurement uncertainty is given by the parameter values where the fitted parabola rises by @xmath153 above the minimum . by construction , this procedure takes correlations between the parameters into account . to validate the measurement technique ,
tests are performed with simulated events generated under the assumptions used in the matrix element method , i. e. using the same pdf set , matrix element , and transfer function .
a pseudo - experiment emulates a measurement performed on data and consists of events randomly drawn from monte carloevent pools for signal and background processes .
the numbers of events taken from the different pools are chosen to reflect the fractions observed in the data .
an ensemble of several pseudo - experiments is performed for each of a number of sets of assumed input parameter values .
the range of assumed values is chosen according to previous determinations and the expected precision of the measurement .
taking the results from all ensembles , the following information is obtained : @xmath58 the relation between the expected ( mean ) measurement values and the corresponding true input values .
it is expected that the method yields unbiased results if the matrix element method reflects the properties of the events . the distribution of measurement uncertainties as a function of input parameter values . the width @xmath154 of the pull distribution . to test that the fitted uncertainties describe the actual measurement uncertainty ,
the deviation of the measurement value from the true value is divided by the fitted measurement uncertainty in each pseudo - experiment .
the width of this distribution of deviations normalized by the measurement uncertainties is referred to as _
pull width_. it is expected that @xmath155 if all features of the events are accommodated in the method . in a similar way , ensemble tests based on fully simulated events can be used to determine any correction of measurement values and fit uncertainties needed when applying the method ( which is based on a simplified detector model ) to real data .
because the computation of likelihoods is time - consuming , the size of the pools of simulated events is usually limited and individual events are allowed to be redrawn , i.e. to appear several times even in the same pseudo - experiment . this technique maximizes the information about the expected uncertainties and pull widths , but it has to be taken into account when evaluating the uncertainties of the ensemble test results @xcite . to summarize , the validation tests described in sections [ enstestljets.sec ] and [ enstestdilepton.sec ] each comprise the three following steps : 1 .
likelihood fit : build one pseudo - experiment and determine @xmath13 , @xmath14 and @xmath15 ( cf .
section [ method.mefit.sec ] ) ; 2 .
ensemble test : repeat step 1 with 1000 pseudo - experiments and obtain mean results , expected uncertainties , and pull widths ; and 3 .
validation : repeat step 2 for several input parameter values to obtain calibration curves .
as an example for a concrete implementation of the matrix element method described previously , the measurement of the top quark mass in lepton+jetsand dilepton @xmath1events is described in this and the following sections .
the discussion of dilepton events is restricted to events containing one electronic and one muonic @xmath3decay , which yield the most precise top quark mass measurement in dilepton events .
this section summarizes the generation of smeared events to study the top quark mass measurement and introduces the event selection criteria .
the characteristics of lepton+jets@xmath1events at a hadron collider are the presence of one energetic isolated charged lepton , at least four energetic jets ( two of which are @xmath0-quark jets ) , and missing transverse momentum due to the unreconstructed neutrino .
the main background is from events where a leptonically decaying @xmath3is produced in association with four or more jets .
multijet background where one jet mimicks an isolated charged lepton can also enter the event sample .
dilepton @xmath1events are characterized by two oppositely charged energetic isolated leptons ( in the case considered here , one electron and one muon ) , two energetic @xmath0-quark jets , and missing transverse momentum due to the two neutrinos from the @xmath3decays .
the largest physics background in the @xmath156channel is from events with @xmath157 decays where the @xmath158boson is produced in association with two or more jets ; another background channel is the production of two leptonically decaying @xmath3bosons together with two jets .
instrumental background arises from events where a leptonically decaying @xmath3is produced in association with three or more jets , one of which is misidentified as the second isolated charged lepton . in principle
, events with leptonically decaying @xmath67 leptons from @xmath3decay contribute to both the lepton+jetsand dilepton event samples .
however , because of lower transverse momentum or transverse energy cuts on the charged lepton(s ) in the event selection ( see below ) , these contributions are typically small .
thus , @xmath1events including leptonic @xmath67 decays are not simulated for the study presented here ( whereas in a real measurement , the effect of such decays has to be accounted for ) . for the study presented here
, events containing a @xmath159 reaction in a @xmath40collision at @xmath6 center - of - mass energy are simulated with the madgraph @xcite generator .
events are generated for each of the different top quark masses , varied between @xmath160 and @xmath161 in steps of @xmath162 .
the alpgen @xcite program is used to generate events containing a leptonic @xmath3or @xmath158decay in association with additional light partons ; events with @xmath0quarks are simulated by smearing the light partons with the transfer function for @xmath0jets . to simulate the decay of a @xmath67 lepton to an electron or muon in @xmath163events , the @xmath67 transfer function shown in figure [ fig : tautransferfunction ]
is applied , while the direction of the lepton is left unchanged . for the modeling of the parton distribution functions , the leading - order pdf cteq5l @xcite is chosen .
multijet background without leptonic @xmath3or @xmath158decay is not simulated because it was shown in @xcite that its effect on the measurement in the lepton+jetschannel is similar to that of additional @xmath3+jetsbackground+jetsand multijet background by the @xmath3+jetsprocess to calculate the likelihood @xmath7 , and to account for any differences between @xmath3+jetsand multijet background when calibrating the measurement using full simulation . ] .
all simulated events are passed through the parametrized detector simulation discussed in section [ method.tf.sec ] , which describes the response of the d0detector @xcite .
no simulation of the parton shower and hadronization is performed since the transfer functions account for their effects in addition to the detector resolution .
samples with different true values of @xmath5are obtained by scaling the smeared light - quark jet energies ; values of @xmath5between @xmath164 and @xmath165 in steps of @xmath166 are used .
similarly , @xmath0-quark jets are scaled by @xmath4 , with @xmath4varied between @xmath167 and @xmath168 in steps of @xmath169 , to obtain samples for different true @xmath0-jet energy scales . as the @xmath4constraint is weaker than the @xmath5one a wider range of generated values
was studied for this observable .
the association of final - state partons to jets is not assumed to be known in the subsequent analysis .
the reconstructed missing transverse momentum @xmath92is taken to be the negative vector sum of all other reconstructed transverse momenta ( i.e. , after the smearing and scaling described above ) ; this means that before smearing and scaling the @xmath1system has zero @xmath54 .
typical event selection criteria as used by the tevatron experiments are then applied to the smeared events .
candidate lepton+jetsevents are required to contain @xmath58 one charged lepton within a pseudorapidity range of @xmath170 ( electrons ) or @xmath171 ( muons ) and with a transverse energy or momentum of at least @xmath172 , four jets within @xmath173 and with ( scaled ) transverse energies of @xmath174 , and missing transverse momentum with magnitude @xmath175 .
the angular separation between the charged lepton and any jet is required to be @xmath176 , and similarly , any jet - jet pair has to be separated by @xmath177 .
no @xmath0-tagging requirements for the jets are included , but @xmath0-tagging information is used later in the analysis .
similarly , dilepton events must contain @xmath58 one electron and one muon of opposite charges within pseudorapidity ranges of @xmath170 or @xmath178 ( electrons ) . ] or @xmath171 ( muons ) and with a transverse energy or momentum of at least @xmath124 , two jets within @xmath173 and with ( scaled ) transverse energies of @xmath174 , and missing transverse momentum with magnitude @xmath179 . the same @xmath180separation cuts as above
are applied , and in addition the two charged leptons are required to be separated by @xmath181 .
the event samples described here are used for validating the measurement technique as discussed in sections [ enstestljets.sec ] and [ enstestdilepton.sec ] . while the exact event selection criteria are not critical to the method , it is mandatory to adjust the likelihood calculation accordingly .
this section describes the calculation of the signal and background likelihoods for lepton+jetsand dilepton @xmath1events .
when the likelihood for a certain process has to be evaluated for many hypotheses , a dedicated implementation of the matrix element optimized for speed is beneficial , and it is helpful to limit the number of evaluations of the transfer function .
section [ application.psgn.sec ] discusses the evaluation of the signal @xmath1likelihoods for a top quark mass measurement ( in the lepton+jetsor dilepton channel ) as an example for such a case .
in contrast , when the number of hypotheses is smaller and/or there are many individual diagrams contributing to a process , interfacing to routines from existing monte carlo generators is a powerful solution .
such a case is the evaluation of the background likelihoods for an @xmath2measurement , which is described in section [ application.pbkg.sec ] .
an overview of the event likelihood calculation for the different decay channels and processes in a top quark mass measurement is given in table [ pevt.table ] .
cccc channel & & & + lepton+jets & .[pevt.table ] overview of the @xmath182calculation in the lepton+jetsand dilepton channels .
the column entitled `` processes '' lists the signal and background processes taken into account in the calculation of the event likelihood @xmath182 .
the symbol `` @xmath84 '' refers to any light parton , i.e. a @xmath80 , @xmath81 , @xmath33 , or @xmath82quark ( or antiquark ) or a gluon .
the rightmost column shows the parameters on which the likelihoods @xmath111for each individual process depend .
in principle the background likelihoods depend on @xmath5 , but as shown later in the paper it is possible to omit this dependence without introducing a significant bias on the @xmath2measurement . [
cols="^ " , ] this test shows that it is possible to adequatly describe @xmath183events in the method . to reduce the systematic error on the top quark mass that arises from the uncertainty on the fraction of @xmath183events in the data sample
, the method should be extended in the future by introducing the fraction of such events as an additional unknown parameter to be measured from the data , similar to the parameters @xmath4and @xmath5 .
the matrix element method is a powerful analysis tool that has been applied with great success in measurements of the top quark mass , the discovery of electroweak single top quark production , and searches for the higgs boson . in this paper , a detailed introduction into the method
is given with the aim of facilitating its application to further measurements .
the principle of the method is introduced , and details concerning the description of the detector response are given .
it has been proposed previously to overcome the current limitation in top quark mass measurements arising from experimental systematic uncertainties by a simultaneous determination of the top quark mass as well as the absolute energy scales for both @xmath0-quark and light - quark jets .
the paper discusses how this strategy can be implemented naturally in the matrix element method for measurements in both lepton+jetsand dilepton events at hadron colliders .
it is shown that the limiting systematic uncertainty in current measurements ( arising from the absolute energy scale for @xmath0-quark jets ) can be overcome . in the future
, it should be possible to render the method stable also against systematic uncertainties related to the fraction of events with significant initial- or final - state radiation . in conclusion
, we have given a general introduction to the matrix element method , and we have shown how future measurements of the top quark mass can be performed with the matrix element method in order to reduce the experimental systematic error .
the authors would like to thank gaston gutierrez and juan estrada for their fundamental contributions to the development of the matrix element method , many of which are part of the foundation for the work presented here .
also , the authors would like to thank raimund strhmer for his careful reading of the manuscript and his very valuable comments , and all their colleagues at the tevatron experiments d0and cdf for many helpful discussions .
all authors have previously been employed at munich university ( lmu ) , where a substantial part of the work towards this paper has been performed , and would like to thank dorothee schaile , otmar biebel , and all members of the lmu experimental particle physics group .
99 v. m. abazov _ et al .
_ , nature * 429 * ( 2004 ) 638 ; + v. m. abazov _ et al .
_ , phys .
b * 617 * ( 2005 ) 1 ; + k. kondo , j. phys .
* 60 * ( 1991 ) 836 ; + r. h. dalitz and g. r. goldstein , phys .
d * 45 * ( 1992 ) 1531 . | the most precise measurements of the top quark mass are based on the matrix element method .
we present a detailed description of this analysis method , taking the measurements of the top quark mass in final states with one and two charged leptons as concrete examples . in addition , we show how the matrix element method is suitable to reduce the dominant systematic uncertainties related to detector effects , by treating the absolute energy scales for @xmath0-quark and light - quark jets independently as free parameters in a simultaneous fit together with the top quark mass . while the determination of the light - quark jet energy scale has already been applied in several recent measurements , the separate determination of the absolute @xmath0-quark jet energy scale is a novel technique with the prospect of reducing the overall uncertainty on the top quark mass in the final measurements at the tevatron and in analyses at the lhc experiments .
the procedure is tested on monte carlo generated events with a realistic detector resolution . |
recently k2k collaboration has reported experimental results on the inclusive single pion production induced by charged current by neutrinos in the energy region of 0.4 - 3 gev@xcite .
similar results were earlier reported by miniboone collaboration in the energy region of 1gev @xcite .
the study of the energy dependence of the pion production cross sections in neutrino reactions for nuclei in this energy is important in modeling the neutrino nucleus cross sections for various monte carlo neutrino event generators used in analyzing the present neutrino oscillation experiment at miniboone , k2k and future experiments to be done by t2k@xcite and no@xmath3a @xcite collaborations . the present experimental result by k2k collaboration reports the energy dependence of the ratio of inclusive cross sections of the charge current induced pion production and quasielastic reaction in neutrino reaction on polystyrene(@xmath4 ) target .
both of these reactions are influenced by the nuclear medium effects in the energy region of this experiment .
the monte carlo simulation of single pion production in this reaction uses the rein and sehgal model@xcite which does not include any nuclear medium effect in the production process .
the neut generator used in the analysis of k2k experiment@xcite , however , includes the nuclear effects arising due to final state interactions of pions with the nucleus like pion absorption and pion scattering @xcite . the quasielastic reaction , on the other hand , uses smith - moniz model @xcite which does not include the effect of nuclear medium arising due to nucleon - nucleon correlations but includes only the effect of pauli principle and fermi motion in a fermi gas model .
there are many calculations for the nuclear medium effects in the quasielastic reaction @xcite-@xcite , and quite a few calculations in the case of single pion production which take into account explicitly the nuclear medium and final state interaction effects in various energy region @xcite-@xcite . in this paper
we present a study of nuclear medium effects in these reactions in the energy region of 0.4 - 3 gev averaged over the neutrino flux of the k2k experiment .
the single pion production in this region is dominated by the resonance production in which a @xmath5 resonance is excited and decays subsequently to a pion and a nucleon .
when this process takes place inside the nucleus then there are two possibilities , the target nucleus remains in the ground state leading to coherent production of pions or is excited and/or broken up leading to incoherent production of pions . in the present paper , we have considered both the production processes in the @xmath5 resonance model in the local density approximation to calculate single pion production from nuclei .
the effect of nuclear medium on the production of @xmath5 is treated by including the modification of @xmath5 properties in the medium .
once pions are produced , they undergo final state interactions with the final nuclei . for the incoherent pion production final state interaction
has been treated using a monte carlo code for pion nucleus interaction @xcite , while for the coherent case , the distortion of the pion plane wave is calculated by using the eikonal approximation @xcite . in the case of quasielastic reaction
, the effects of pauli principle and fermi motion are included through the lindhard function calculated in local density approximation @xcite .
the details of these calculations are given in earlier publications @xcite,@xcite,@xcite .
in this paper we report the results for the nuclear medium effects on total cross section @xmath6 , @xmath7 distribution @xmath8 and momentum distribution @xmath9 for the muon , in the case of quasielastic and single pion production induced by charged currents . we compare our results of the ratio @xmath10 , with the recent observations of k2k collaboration .
the results for the pion momentum and angular distributions have also been shown for the incoherent pion production process . in secs .
ii & iii , we briefly describe the various formula used in the present calculations for quasielastic and inelastic reactions and present the results and discussion in sec.iv .
the basic reaction for the quasielastic process is a neutrino interacting with a neutron inside the nucleus i.e. @xmath11 the cross section for quasi - elastic charged lepton production is calculated in the local density approximation by taking into account the fermi motion and the pauli blocking effects through the imaginary part of the lindhard function for the particle hole excitations in the nuclear medium .
the renormalization of the weak transition strengths are calculated in the random phase approximation(rpa ) through the interaction of the p - h excitations as they propagate in the nuclear medium using a nucleon - nucleon potential described by pion and rho exchanges @xcite .
the effect of the coulomb distortion of muon in the field of final nucleus is also taken into account using a local version of the modified effective momentum approximation(mema ) @xcite . process . ]
the total cross section @xmath12 for the charged current neutrino induced reaction on a nucleon inside the nucleus in a local fermi gas model is written as @xcite : @xmath13 where @xmath14 and @xmath15 , calculated with rpa correlations in nuclei .
the leptonic current @xmath16 and the hadronic current @xmath17 are given by @xmath18 @xmath19u(p).\end{aligned}\ ] ] induced charged current quasielastic process in @xmath1 . ]
induced incoherent charged current one pion production process in @xmath1 . ]
induced coherent charged current one pion production process in @xmath1 . ]
induced charged current one pion production process and charged current quasielastic process in polystyrene .
experimental points have been taken from ref.@xcite for k2k results in scibar and from refs.@xcite , @xcite for anl results . ]
distribution for @xmath20 induced charged current quasielastic process in @xmath1 averaged over the k2k spectrum . ] where @xmath21 is the four momentum transfer , m is the mass of the nucleon , @xmath22 is the fermi coupling constant and @xmath23 is the lepton angle .
@xmath24 is the lindhard function for the particle hole excitation @xcite .
the form factors @xmath25 , @xmath26 , @xmath27 and @xmath28 are isovector electroweak form factors . for our numerical calculations
we have used the parameterization of bradford et al .
@xcite for the vector form factors @xmath29 and @xmath30 with dipole mass @xmath31=0.84gev and a dipole form for @xmath32 with dipole mass @xmath33=1.1gev . inside the nucleus
, the q - value of the reaction and coulomb distortion of outgoing lepton are taken into account by modifying the lindhard function @xmath34 by @xmath35 , where @xmath36 is the coulomb potential of the final nucleus .
furthermore , the renormalization of weak transition strength in the nuclear medium in a random phase approximation(rpa ) is taken into account by considering the propagation of particle hole(ph ) as well as delta - hole(@xmath37 ) excitations @xcite .
these considerations lead to modified hadronic tensor components @xmath38 involving the bilinear terms , for which expressions are given in refs .
@xcite,@xcite .
induced charged current quasielastic process in @xmath1 averaged over the k2k spectrum . ] distribution for @xmath20 induced incoherent charged current one pion production process in @xmath1 averaged over the k2k spectrum . ]
the basic reaction for the inelastic one pion production in nuclei , for a neutrino interacting with a nucleon n inside a nuclear target is given by @xmath39 in the case of incoherent pion production in @xmath5 dominance model , the weak hadronic currents interacting with the nucleons in the nuclear medium excite a @xmath5 resonance which decays into pions and nucleons .
the pions interact with the nucleus inside the nuclear medium before coming out .
the final state interaction of pions leading to elastic , charge exchange scattering and the absorption of pions lead to reduction of pion yield .
the nuclear medium effects on @xmath5 properties lead to modification in its mass and width which have been discussed earlier by oset et al .
@xcite and applied to explain the pion and electron induced pion production processes from nuclei . in the local density approximation the expression for the total cross section for the charged current one pion production ( kinematics is shown in fig.1 ) is written as @xmath40 where the proton density @xmath41 and the neutron density @xmath42 with @xmath43 as the nuclear density taken as 3-parameter fermi density @xcite . induced incoherent charged current one pion production process in @xmath1 averaged over the k2k spectrum . ]
induced incoherent charged current one pion production process in @xmath1 averaged over the k2k spectrum . ]
induced incoherent charged current one pion production process in @xmath1 averaged over the k2k spectrum . ]
the transition matrix element @xmath44 is given by @xmath45 where @xmath46 is the leptonic current defined by eq.([lep_curr ] ) , and @xmath47 is the @xmath48 transition operator taken from lalakulich et al
. @xcite .
@xmath49 is the weak mixing angle .
@xmath50 is the @xmath5 propagator in momentum space and is given as : @xmath51 where @xmath52 is the spin-3/2 projection operator and the delta decay width @xmath53 is taken to be an energy dependent p - wave decay width @xcite : @xmath54 @xmath55 is the pion momentum in the rest frame of the resonance and w is the center of mass energy . inside the nuclear medium , the mass and width of delta are modified which in the present calculation are taken by considering the following effects .
+ in nuclear medium @xmath5s decay mainly through the @xmath56 channel .
the final nucleons have to be above the fermi momentum @xmath57 of the nucleon in the nucleus which leads to a modification in the decay width of delta .
the modified delta decay width @xmath58 has been taken from the works of oset et al .
@xcite which is given by @xmath59 where @xmath60 @xmath61 , @xmath62 is the @xmath5 momentum and @xmath63 .
furthermore , in the nuclear medium there are additional decay channels open due to two and three body absorption processes like @xmath64 and @xmath65 through which @xmath5 disappears in nuclear medium without producing a pion , while a two body @xmath5 absorption process like @xmath66 gives rise to some more pions .
these nuclear medium effects on the @xmath5 propagation are included by modifying the mass and the decay width of @xmath5 in nuclear medium as @xmath67 the expressions for the real and imaginary part of the @xmath5 self energy are taken from oset et al .
@xcite : @xmath68 in the above equation @xmath69 accounts for the @xmath70 process , @xmath71 for the two - body absorption process @xmath72 and @xmath73 for the three - body absorption process @xmath65 .
the coefficients @xmath69 , @xmath71 , @xmath73 and @xmath74 , @xmath75 and @xmath76 are taken from ref .
@xcite .
the pions produced in this process are scattered and absorbed in the nuclear medium .
this is treated in a monte carlo simulation which has been taken from ref .
@xcite .
@xmath20 induced coherent one pion production on @xmath1 target is given by @xmath77 for which the basic reaction is given by eq.([inelastic_reaction ] ) .
the total cross section ( kinematics is shown in fig.1 ) is given by @xmath78 where the matrix element for the reaction is given by @xmath79 @xmath16 is the leptonic current given by eq.(3 ) and @xmath80 is the hadronic current given by @xmath81 where f(@xmath82 ) ( p = p+q ) is @xmath83 form factor given by@xcite : @xmath84 with @xmath85 .
@xmath86 is the nuclear form factor , given by @xmath87e^{-i({\vec q}-{\vec k}_\pi).{\vec r}}\ ] ] when pion absorption effects are taken into account using the eikonal approximation then the nuclear form factor @xmath88 is modified to @xmath89 , which is taken to be @xcite : @xmath90 where @xmath91 @xmath92 and @xmath93 are the longitudinal and transverse component of the pion momentum and the pion self - energy @xmath94 is given by @xmath95 with w as the center of mass energy in the @xmath5 rest frame .
for the quasielastic reaction , the numerical results are obtained from eq.([sigma_quasi ] ) using vector form factors @xmath25 and @xmath26 given by bradford et al .
@xcite with vector dipole mass @xmath31=0.84gev and a dipole form for @xmath32 with axial dipole mass @xmath33=1.1gev , and are shown in fig.2 .
our results for the total scattering cross section @xmath96 without the rpa effects ( dotted line ) in the case of charged current quasielastic scattering is in fair agreement with the calculation done in fermi gas model @xcite,@xcite,@xcite in the energy region of 0.4 - 3 gev .
when rpa correlations are included in our model , we see that they give rise to a reduction in the total cross section @xmath6 ( solid line ) which is 12 - 15@xmath97 in the energy region of 0.4 - 3 gev . in the case of
charged current induced incoherent and coherent pion productions the results for the total cross sections are obtained from eq.([sigma_inelastic ] ) and eq.([sigma_coherent ] ) respectively using the n-@xmath5 transition form factors given by lalakulich et al.@xcite with @xmath98=1.1 gev .
the numerical results are shown in fig.3 and fig.4 for the incoherent and coherent pion productions .
we see that the effect of nuclear medium in the pion production process as well as the final state interaction of pions both reduce the cross section . in the case of incoherent production of pions ,
the reduction due to nuclear medium effects in the production process is larger than due to final state interaction while in the case of coherent pion production , the reduction due to final state interaction is quite large as compared to the reduction due to the nuclear medium effects .
however , the contribution of coherent process to the total pion production is small(5 - 6@xmath97 ) in the energy region of @xmath99 , therefore , the reduction in the total cross section is mainly given by the reduction in the cross section of the incoherent process .
in fig.5 , we show the results for the ratio of total one pion production and the quasielastic production cross sections for @xmath20 induced reaction in polystyrene(@xmath100 ) , i.e. @xmath10 ( solid line ) as a function of neutrino energy .
a comparison with the experimental results of k2k @xcite and anl @xcite , @xcite for r(e ) are also shown in the figure . in this figure
, we also show the ratio r(e)(dashed line ) when no nuclear medium effect is taken into account either in the quasielastic process or in the inelastic process . when the cross sections are calculated without any nuclear medium modifications except the fermi motion and pauli principle for the quasielastic process and without any modification of @xmath5 properties in the nuclear medium in case of inelastic process , the results for r(e )
are shown by dashed - double dotted line .
this should be compared with the monte carlo predictions used in k2k analysis@xcite .
we see from figs
. 2 - 5 that the nuclear medium modifications play an important role in quasielastic and inelastic reactions and tend to reduce the cross section ratio r(e ) .
we have also studied the effect of the uncertainties in the ratio r(e ) due to the use of various vector form factors given by budd et al .
@xcite , bosted @xcite and dipole form factors@xcite in the quasielastic case and the n-@xmath5 transition form factors given by schreiner & von hippel@xcite and paschos et al.@xcite in the inelastic case .
the effect of varying the axial dipole mass @xmath101 in the parameterization of axial form factor in quasielastic as well as in the inelastic processes in the region of @xmath102 gev on the ratio r(e ) has also been studied .
these effects lead to a small change of about 4 - 6@xmath97 in the ratio r(e ) . in figs .
6 & 7 , we show the effect of nuclear medium on @xmath7 distribution and the momentum distribution of the leptons averaged over the neutrino flux at k2k in the case of neutrino induced charged current quasielastic scattering on @xmath1 target . a reduction in @xmath7 distribution in the peak region
is found to be about 30@xmath97 when rpa effects are taken into account while this reduction is about 12@xmath97 in the momentum distribution of leptons in the peak region . in the case of
incoherent one pion production process , the @xmath7 distribution as well as the momentum distribution of leptons are shown in figs.8 & 9 . in these reactions charge pions are also produced and we show in figs.10 & 11 , the angular distribution as well as momentum distribution of pions and the effect of nuclear medium on these distributions .
we find that the reduction in @xmath7 distribution in the peak region is around 30@xmath97 when nuclear medium effects are taken into account which further reduces by about 14@xmath97 when pion absorption effects are incorporated .
for the pion momentum distribution the reduction due to nuclear medium effects is around 40@xmath97 which further reduces by about 15@xmath97 when pion absorption effects are taken into account .
we have not shown these results for the coherent process as the contribution to the total pion production events have been found to be quite small . however , these results are interesting in their own right and will be reported elsewhere . to conclude ,
we have studied in this paper the effect of nuclear medium on the inclusive quasielastic and one pion charged current production in nuclei and compared our results with the recent experimental results from the k2k experiment @xcite .
the nuclear medium effects are important in this energy region and reduce the ratio by 30@xmath97 around the energy region of 1gev which becomes 20 - 25@xmath97 in the neutrino energy region of 2 - 3gev .
the results for nuclear medium effects on the @xmath7 and the momentum distribution of lepton in the quasielastic as well as inelastic reaction and the results for the nuclear medium effects on the angular and momentum distribution of pions have also been presented .
one of the authors(m .
s. a. ) is thankful to t. kajita and y. hayato ( i. c. r. r. , university of tokyo ) for many useful discussions and the warm hospitality provided during his stay at icrr where part of this work was done .
s. c. is thankful to the jawaharlal nehru memorial fund for the doctoral fellowship . 0 a. rodriguez et al .
( k2k collaboration ) , phys . rev . *
d 78 * , 032003 ( 2008 ) .
m. o. wascko ( miniboone collaboration ) , nucl .
proc . suppl . *
159 * , 50 ( 2006 ) .
y. obayashi ( t2k collaboration ) , arxiv : hep - ex/0807.4012v2(2008 ) , y. itow et al .
, arxiv : hep - ex/0106019(2001 ) .
d. s. ayres et al .
( no@xmath3a collaboration ) , arxiv : hep - ex/0503053(2005 ) .
d. rein and l. m. sehgal , annals .
phys . * 79 * , 133 ( 1981 ) ; nucl . phys .
* b 223 * , 29 ( 1983 )
. y. hayato , nucl .
. suppl . *
b 112 * , 171 ( 2002 ) .
r. a. smith and e. j. moniz , nucl .
* b 43 * , 605 ( 1972 ) .
c. h. llewellyn smith , phys .
rep . * 3 * , 261 ( 1972 ) .
t. k. gaisser and j. s. oconnell , phys . rev . *
d 34 * , 822 ( 1986 ) .
t. kuramoto , m. fukugita , y. kohyama and k. kubodera , nucl . phys . * a 512 * , 711 ( 1990 ) . s. k. singh and e. oset , nucl
. phys . * a 542 * , 587 ( 1992 ) ; phys . rev .
* c 48 * , 1246 ( 1993 ) .
h. kim , j. piekarewicz and c. j. horowitz , phys .
* c 51 * , 2739 ( 1995 ) .
n. auerbach , n. van giai and o. k. vorov , phys . rev .
* c 56 * , 2368 ( 1997 ) .
a. c. hayes and i. s. towner , phys
* c 61 * , 044603 ( 2000 ) .
j. marteau , j. delorme and m. ericson , nucl .
. meth . * a 451 * , 76 ( 2000 ) .
o. benhar , n. farina , h. nakamura , m. sakuda and r. seki , phys . rev . *
d 72 * , 053005 ( 2005 ) .
j. nieves , j. e. amaro and m. valverde , phys .
* c 70 * , 055503 ( 2004 ) ; j. nieves , m. valverde and m. j. vicente vacas , phys . rev . * c 73 * , 025504 ( 2006 ) .
a. meucci , c. giusti and f. d. pacati , nucl . phys . * a 739 * , 277 ( 2004 )
. m. sajjad athar , s. ahmad and s. k. singh , nucl . phys . * a 764 * , 551 ( 2006 ) . c. giusti , a. meucci and f.d .
pacati , arxiv : nucl - th/0607037(2006 ) .
m. c. martinez , p. lava , n. jachowicz , j. ryckebusch , k. vantournhout , j.m .
udias , phys . rev . * c 73 * , 024607 ( 2006 ) .
j. a. caballero , j. e. amaro , m. b. barbaro , t. w. donnelly and j. m. udias , phys .
* b 653 * , 366 ( 2007 ) .
a. v. butkevich and s. a. kulagin , arxiv : nucl - th/0705.1051(2007 ) .
m. martini , g. co , m. anguiano and a. m. lallena , phys . rev . * c 75 * , 034604 ( 2007 ) . j. e. amaro , m. b. barbaro , j. a. caballero , t. w. donnelly and j. m. udias , phys .
* c 75 * , 034613 ( 2007 ) .
m. sajjad athar , s. ahmad and s.k.singh , eur .
j. * a 24 * , 459 ( 2005 ) ; s. ahmad , m. sajjad athar and s. k. singh , phys .
d 74 * , 073008 ( 2006 ) .
t. leitner , l. alvarez - ruso and u. mosel , phys . rev .
* c 73 * , 065502 ( 2006 ) .
o. benhar and d. meloni , nucl . phys . * a 789 * , 379 ( 2007 ) .
h. kim , s. schramm and c. j. horowitz , phys . rev .
* c 53 * , 3131 ( 1996 ) . s. k. singh , m. j. vicente vacas and e. oset , phys .
* b 416 * , 23 ( 1998 ) . j. marteau , eur . phys
. j. * a 5 * , 183 ( 1999 ) .
e. a. paschos , j. y. yu and m. sakuda , phys . rev . *
d 69 * , 014013 ( 2004 ) .
w. cassing , m. kant , k. langanke and p. vogel , phys .
* b 639 * , 32 ( 2006 ) .
e. a. paschos , i. schienbein and j .- y .
yu , arxiv : hep - ph/0704.1991 ( 2007 ) .
c. praet , o. lalakulich , n. jachowicz and j. ryckebusch , arxiv : nucl - th/0804.2750 ( 2008 ) .
a. kartavtsev , e. a. paschos and g. j. gounaris , phys . rev . *
d 74 * , 054007 ( 2006 ) .
s. k. singh , m. sajjad athar and s. ahmad , phys .
lett . , * 96 * , 241807 ( 2006 ) .
l. alvarez - ruso , l. s. geng , s. hirenzaki and m. j. vicente vacas , phys . rev . * c 75 * , 055501 ( 2007 ) . m. j. vicente vacas , m. kh .
khankhasayev and s. g. mashnik , arxiv : nucl - th/9412023(1994 ) .
r. c. carrasco , j. nieves and e. oset , nucl .
phys . * a 565 * , 797 ( 1993 ) .
h. toki and w. weise , phys .
rep . * 83 * , 281 ( 1982 ) .
r. bradford , a. bodek , h. budd and j. arrington , nucl .
. suppl . *
159 * , 127 ( 2006 ) .
e. oset , p. fernandez de cordoba , l. l. salcedo and r. brockmann , phys
. rep . * 188 * , 79 ( 1990 ) ; e. oset , d. strottman , h. toki and j. navarro , phys . rev .
c * 48 * , 2395 ( 1993 ) .
e. oset and l. l. salcedo , nucl
. phys . * a 468 * , 631 ( 1987 ) ; c. garcia recio , e. oset , l. l. salcedo , d. strottman , m. j. lopez , nucl .
phys . * a 526 * , 685 ( 1991 ) . c. w. de jager , h. de vries and c. de vries , at .
data nucl .
data tables * 14 * , 479 ( 1974 ) .
o. lalakulich , e. a. paschos and g. piranishvili , phys . rev .
* d 74 * , 014009 ( 2006 ) . g. penner and u. mosel , phys . rev . * c 66 * , 055211 ( 2002 ) ; m. post , ph.d .
thesis , universitat giessen , 2004(unpublished ) .
g. m. radecky et al . , phys . rev .
* d 25 * , 1161 ( 1982 ) .
s. j. barish , a. engler , r. w. kraemer , k. miller and b. j. stacey , phys . rev .
* d 16 * , 3103 ( 1977 )
. h. budd , a. bodek and j. arrington , arxiv : hep - ex/0308005 .
e. bosted , phys . rev .
* c 51 * , 409 ( 1995 ) .
s. galster , h. klein , j. moritz , k. h. schmidt , d. wegener and j. bleckwenn , nucl . phys .
* b 32 * , 221 ( 1971 ) .
p. a. schreiner and f. von hippel , nucl .
* b 58 * , 333 ( 1973 ) . | in this paper , we present the results of a study made for the effect of nuclear medium in the charged current induced quasielastic lepton production(ccqe ) and the incoherent and coherent one pion production ( cc1@xmath0)processes from @xmath1 in the @xmath2 energy region of 0.4 - 3gev .
the theoretical results are compared with the recent experimental results for the ratio of charged current @xmath2 induced one pion production cross section to the quasielastic lepton production cross section reported by k2k collaboration .
we also present the results for the angular and momentum distributions of leptons and pions produced in these processes . |
nonsequential double ionization ( nsdi ) in intense laser pulses has been experimentally observed in measurements of ion yields as a function of the laser intensity , which deviate from the yields expected from a sequential ionization scenario , forming the so - called nsdi `` knee '' ( see @xcite for recent reviews ) .
in fact , the multiple ionization yields are typically enhanced by several orders of magnitude . with ionization yields being rather integrated observables the mechanism behind nsdi
could not be unequivocally resolved until the measurement of ion spectra @xcite and correlated photoelectron distributions has become feasible ( see @xcite for early and , e.g. , @xcite for very recent work ) .
meanwhile nsdi is understood in terms of a recollision process : one electron is emitted but oscillates back to its parent ion to knock out the next electron . if the return energy is not sufficient for collisional ionization , the next electron might be excited and later emitted owing to the laser field @xcite .
the described recollision scenario poses a huge challenge for general many - body methods when applied to such a few - electron test case .
for example , in time - dependent hartree - fock ( tdhf ) or time - dependent density functional theory ( tddft ) applied to he starting from the singlet ground state there is only one spatial orbital describing both electrons ( one spin - up , the other spin - down ) .
not surprisingly , it was found that such methods are not capable of describing nsdi @xcite , although formally for different reasons .
tdhf , as a mean - field approach , does not incorporate correlation by definition .
tddft is in principle exact but only in the sense that it gives the exact time - dependent electron density .
however , even if the exact time - dependent electron density was known from a tddft calculation employing the exact exchange - correlation potential @xcite , the exact double - ionization probability could still not be calculated because this observable is unknown as an explicit functional of the electron density , and simple approximations to it do not reproduce the nsdi knee @xcite .
solving the full time - dependent schrdinger equation ( tdse ) for he in full dimensionality and in strong , long - wavelength ( i.e. , @xmath0 nm ) laser fields is still beyond what is possible with current super computer technology @xcite .
therefore it is essential to develop practicable time - dependent many - electron methods beyond linear response that account for correlation .
time - dependent configuration interaction ( tdci ) @xcite , multiconfigurational time - dependent hartree ( mctdh ) @xcite or multiconfigurational tdhf ( mctdhf ) @xcite are exact in principle .
however , the crucial question in practice is how many configurations or determinants are required to recover a certain strong correlation feature such as the nsdi knee .
general conclusions are difficult to draw , as different tdci and mctdhf approaches may vary in the single - particle basis functions chosen .
it has been shown that for he ( or h@xmath1 ) a time - dependent variational approach using a wavefunction ansatz with two different single - particle orbitals ( time - dependent extended hartree - fock ) @xcite or an _ ad hoc _ ansatz with an `` inner '' and an `` outer '' outer orbital @xcite at least generate kneelike structures in the double - ionization yield .
however , they are only in poor agreement with the exact numerical results available for low - dimensional models . to the best of our knowledge
there are no systematic tests of computational approaches that demonstrate a convergence toward the exact nsdi knee .
in fact , we are not aware of any work that accurately reproduces the nsdi knee using a many - body method that overcomes the `` exponential wall '' @xcite . in this work
, we will provide such an analysis for our recently introduced time - dependent renormalized - natural - orbital theory ( tdrnot ) @xcite .
before we present results on the nsdi knee ( sec .
[ sec : ioniyields ] ) and on correlated photoelectron spectra ( sec .
[ sec : temd ] ) , we briefly introduce the he model , review the essentials of tdrnot , particularly when applied to a two - electron system , and discuss the observables to be calculated .
atomic units ( a.u . ) are used unless otherwise indicated .
the widely applied one - dimensional helium model atom @xcite in a laser field has the hamiltonian @xmath2 where upper indices indicate the action on either electron @xmath3 , electron @xmath4 , or both .
the single - particle hamiltonian reads @xmath5 , with @xmath6 @xmath7 ( dipole approximation and velocity gauge with the @xmath8 term transformed away ) , and the electron - electron interaction @xmath9 the electron - ion smoothing parameter @xmath10 is chosen such that the groundstate energy of @xmath11 @xmath12 is recovered .
the electron - electron smoothing parameter @xmath13 is tuned to yield the neutral - he energy @xmath14 .
the hamiltonian ( [ hamiltonian ] ) does not act on the spin , which in the two - particle case allows one to factorize the wavefunction , @xmath15 here @xmath16 and @xmath4 are shorthand notations for position and spin @xmath17 and @xmath18 , respectively .
the two- and one - body density matrices read @xmath19 @xmath20 natural orbitals ( nos ) @xmath21 are defined as eigenvectors of @xmath22 : @xmath23 the corresponding eigenvalues @xmath24 $ ] are called occupation numbers ( ons ) .
nos and ons were introduced a long time ago ( see , e.g. , @xcite ) , but only recently have their usefulness for time - dependent few- and many - body problems been studied @xcite .
the coupled equations of motion for the ons and the nos can be unified by introducing renormalized nos ( rnos ) @xcite @xmath25 so that @xmath26 and @xmath27 the two - body density matrix expanded in rnos reads @xmath28 the equation of motion ( eom ) for the rnos is @xcite @xmath29 with @xmath30 @xmath31 , \label{eom - rno - b } \end{split}\end{aligned}\ ] ] and @xmath32 one observes that the effective hamiltonian in the tdse - like equation consists of the usual one - body operator @xmath33 , a diagonal part @xmath34 , the part @xmath35 which couples rnos , and the operator @xmath36 , which also couples rnos . as the effective hamiltonian in is hermitian ,
the corresponding time evolution of the rnos is unitary .
in general , there are infinitely many nos required to describe a correlated quantum system , even if it contains only two particles . ordered decreasingly according to their ons
, the number of rnos taken into account in an actual numerical implementation of is necessarily truncated , which introduces errors in the propagation .
the effect of this truncation will be seen in the results in sec .
[ sec : results ] below . in the two - particle case
the expansion coefficients @xmath37 are exactly known @xcite , @xmath38}}{2\sqrt{n_i(t)n_k(t)}}\delta_{i , j{^\prime}}\delta_{k , l{^\prime}}. \label{eq : gamma2tilde}\ ] ] here , the `` prime operator '' acts on the positive integer @xmath39 according to @xmath40 and the phase factors are @xcite @xmath41 in the spin - singlet and -triplet case , respectively . note that the eom for the rnos is given here for phase - including nos @xcite so that @xmath42 and @xmath43 in do not depend on time . employing the factorization ( [ wf - factorization ] )
we can write @xmath44 @xmath45 where @xmath46 here and in the following , spatial rnos and quantities calculated from them ( e.g. , @xmath47 ) will be indicated by underlining them . how the rnos can be written as a factorization in the spatial and the spin part
is discussed in detail in @xcite . in this work
we will only consider results for the singlet configuration where the rnos with @xmath48 can be arranged as latexmath:[\[\langle x | \tilde k(t)\rangle = \left\ { \begin{array}{ll } |\!\uparrow\rangle\,{\underline{\tilde{\phi}}}_{k'/2}(x;t ) & \mathrm{if}\ k\ \mathrm{odd } \\ so that any consecutive @xmath39-odd and @xmath50-even rnos share the same spatial component @xmath51 .
we are interested in the double - ionization probability of the model he atom as a function of the laser intensity and in correlated photoelectron spectra , i.e. , the probability to find one electron being emitted with momentum @xmath52 and the other with @xmath53 , for laser intensities where nsdi occurs .
both should in principle be calculated via the projection of the wavefunction after the laser pulse on two - electron continuum states of asymptotic momenta @xmath52 and @xmath53 .
however , this approach is numerically unfeasible .
we will shortly explain how the yields and spectra are calculated in a less rigorous but sufficiently accurate manner in this work . an efficient way to calculate ionization probabilities from the two - electron wavefunction @xmath54 after the laser pulse is based on the integration of the probability density @xmath55 over certain spatial regions , @xmath56 where we made use of the fact that @xmath57 .
the parameter @xmath58 should be sufficiently large such that the probabilities @xmath59 and @xmath60 are negligible for the groundstate and singly - excited eigenstates .
on the other hand , @xmath61 should not be too large so that the probability density describing ionization does not need too much time to leave the neutral - he region @xmath62 . for our model
we chose @xmath63 .
as for a two - electron system @xmath64 , eqs . and are read in terms of rnos : @xmath65 note that in the two - electron case this form is equivalent to first reconstructing the wavefunction ( which is possible for two electrons @xcite ) and then using eqs . and .
a numerically efficient method to calculate correlated double - ionization photoelectron spectra is to multiply the two - electron wavefunction by a mask function @xmath66 , which removes the parts representing he@xmath67 and neutral he : @xmath68 here , we chose @xmath69 , with @xmath70 and @xmath71 @xcite .
after fourier transforming @xmath72 to momentum space , @xmath73 the double - ionization photoelectron spectrum is obtained as @xmath74 in our tdrnot treatment we proceed analogously by first defining @xmath75 whose fourier transform is @xmath76 .
then , @xmath77 which can be written as @xmath78 where @xmath79 we thus have an explicit construction for @xmath80 in terms of rnos .
note that in tddft such a construction in terms of kohn - sham orbitals is unknown @xcite .
spatial rnos ( red , + ) are compared with the exact tdse result ( black , dots ) and with the ionization probability reconstructed using the first @xmath81 exact nos calculated from the exact tdse wavefunction ( green , @xmath82 ) . ]
we consider an @xmath83-nm ( @xmath84 ) linearly polarized @xmath85-cycle @xmath86-shaped laser pulse of duration @xmath87 .
the vector potential in dipole approximation reads @xmath88 and zero otherwise .
the numerical grids for both the tdse - benchmark and tdrnot calculations covered @xmath89 a.u . in the spatial directions .
figure [ fig : p2 ] shows the double - ionization probability @xmath60 as a function of the laser intensity for different numbers of spatial rnos @xmath81 between @xmath16 ( upper - left panel ) and @xmath90 ( lower - right panel ) . for comparison , the exact tdse result is included in black in all panels .
the nonmonotonic behavior of the first derivative of this exact @xmath60 curve in the region around @xmath91w/@xmath92 gives rise to the celebrated nsdi knee .
a tdrnot calculation with @xmath93 rno per spin yields a featureless @xmath60 curve , as seen in the upper - left panel of fig .
[ fig : p2 ] .
in fact , in the case of a two - electron spin - singlet system , a single no per spin is equivalent to tdhf or tddft in exchange - only approximation , for which it is already known that the nsdi knee is not reproduced @xcite .
truncating the number of rnos in a tdrnot calculation introduces an error in the propagation of the rnos @xcite .
this error should be distinguished from the error that arises alone due to the fact that only a finite number of nos is taken into account for the calculation of an observable .
we do this by determining all exact nos from the exact tdse wavefunction but consider only the @xmath81 most dominant of them to calculate the observable @xmath60 .
the respective results are also shown in fig . [
fig : p2 ] . for @xmath93
this procedure gives a result very different from the tdrnot with @xmath93 .
there is even already a knee in the tdse-1-no result , albeit a quantitatively wrong one .
both tdrnot with a single rno and the tdse-1-no curve show a wrong slope in the limit of low laser intensity . for @xmath94 nos per spin ( upper - right panel )
a knee appears also in the tdrnot result .
it is exaggerated and jaggedly structured , and underestimates the @xmath60 yield .
a similar behavior with two orbitals was observed in extended hartree - fock treatments @xcite and with the so - called `` crapola '' model @xcite , where an `` inner '' and an `` outer '' orbital is postulated . with increasing @xmath81
the agreement between tdrnot results and tdse improves . for @xmath95
the truncation error in the propagation of the rnos is small enough to give almost the same probability @xmath60 as if it was calculated with the first @xmath95 exact nos . at @xmath96 .
[ fig : p2dim_tdse ] ] clearly , our tdrnot approach is only attractive if @xmath81 can be kept reasonably small .
we have shown in refs .
@xcite how , with a few rnos , doubly excited states , autoionization , and rabi flopping can be described using tdrnot .
unfortunately , nsdi is more demanding in @xmath81 , meaning that nsdi is highly correlated , and thus many more nos than particles are required . moreover ,
note that although nsdi is a huge effect on the @xmath60 level , it is a small effect compared to the probability for single - ionization @xmath59 , and small effects on an absolute scale are captured by nos with small ons .
the dominant nos are mainly `` responsible '' for single ionization , or no ionization at all . in that respect
it is remarkable to achieve an agreement such as the one shown for @xmath95 spatial rnos in fig .
[ fig : p2 ] .
we are not aware of any tdci or tdmchf calculation that achieved such an agreement , let alone with only ten basis functions . at @xmath96 obtained from the @xmath97
dominant , exact spatial nos calculated from the exact tdse wavefunction after the laser pulse . ]
correlated photoelectron spectra contain more information than ionization probabilities .
in general , the `` more differential '' an observable is , the harder it is to reproduce by some approximate method because the dynamic range to be accurately covered increases .
an additional , conceptual challenge arises with tddft because @xmath80 is an unknown functional of the single - particle density , and simple approximations fail @xcite .
figure [ fig : p2dim_tdse ] shows the tdse benchmark result for @xmath80 at @xmath96 , i.e. , in the nsdi intensity regime .
the butterfly structure indicating electrons emitted into the same direction is characteristic of nsdi @xcite and has been essential to identify rescattering as its origin . from the tdse benchmark
we know that the first thousand exact nos have ons @xmath98 .
the question is how many nos are needed to recover the butterfly structure seen in fig .
[ fig : p2dim_tdse ] .
figure [ fig : p2dim ] shows that with the first @xmath99 exact nos from the tdse simulation the butterfly structure of fig .
[ fig : p2dim_tdse ] emerges , but details are still not accurately represented over the 4 orders of magnitude dynamic range shown . however , it is sufficient for the purpose of validating tdrnot with a reasonably small number of rnos .
up to @xmath100 , mainly uncorrelated , gridlike horizontal and vertical structures are visible . from @xmath101 on , however , clear correlated structures appear , first in the first quadrant @xmath102 . at @xmath96 obtained from tdrnot with @xmath97 spatial rnos .
] figure [ fig : p2dim_tdrot ] shows the corresponding tdrnot result with @xmath81 rnos per spin propagated .
again , the differences between the benchmark results in fig . [ fig : p2dim ] and tdrnot in fig . [ fig : p2dim_tdrot ] are due to the truncation error in the number of propagated rnos .
this truncation error severely spoils the correlation structure in the first quadrant ; only for @xmath103 does it start to emerge . in order to reproduce ,
say , the lower - right spectrum in fig .
[ fig : p2dim ] , one would need to propagate about @xmath104 rnos in tdrnot .
this is prohibitive with our current implementation of solving the nonlinear eom .
we found , for instance , that apart from the expected increase of the numerical effort there is the additional complication that the time step needs to be reduced with increasing @xmath81 .
because of the truncation error , the @xmath81th of the ( according on ordered ) @xmath81 dominant spatial rnos is expected to be most defective .
thus it may make sense to propagate more rnos than are actually used to calculate observables .
figure [ fig : p2dim_tdrot _ ] shows results where @xmath103 rnos per spin were propagated but only @xmath100 and @xmath105 were used for the calculation of the photoelectron spectra .
one sees that the agreement with the two corresponding middle - row spectra in fig .
[ fig : p2dim ] is much better than in fig .
[ fig : p2dim_tdrot ] . at @xmath96 obtained from tdrnot with @xmath103 spatial rnos propagated but only @xmath100 and
@xmath105 used to calculate @xmath80 . ]
the computational time @xmath106 required for a tdrnot propagation using @xmath81 nos on @xmath107 spatial grid points scales as @xmath108 for a fixed time step .
the first term on the right - hand side arises from the calculation of the potentials in using the fast fourier transform , and the second term from the evaluation of the required matrix elements in @xcite .
the computational costs of the corresponding operations are taken into account by the constant factors @xmath109 and @xmath110 .
the computational times @xmath111 required for one laser intensity using @xmath16 ( equivalent to tdhf ) , @xmath4 , and @xmath105 nos were @xmath112 , @xmath113 , and @xmath114 , respectively , on a single core of an i5 - 3570 processor .
this shows that the @xmath115 term in is dominating .
compared to the tdse calculation , where @xmath116 , tdrnot thus performs faster by a factor of @xmath117 , @xmath118 , and @xmath119 , respectively .
unfortunately , the time step in our present tdrnot implementation needs to be decreased with increasing @xmath81 to achieve converged results .
for example , for @xmath95 nos , @xmath120 was used , whereas in the tdse simulation @xmath121 was sufficient .
this leads to a total computation time of @xmath122 and @xmath123 . hence , improving our tdrnot scheme to allow for larger time steps is desirable .
however , note that for more than two particles the small tdrnot time step is harmless anyway compared to the exponential scaling of the tdse wavefunction .
in summary , we reproduced the nonsequential double - ionization knee for a he - model atom starting from the spin - singlet ground state using the recently introduced time - dependent renormalized - natural - orbital theory .
the equations of motion for the renormalized - natural - orbitals are exact in the two - electron case .
this is because the expansion of the time - dependent two - body density matrix in natural orbitals is known exactly .
only the practical limitation in the number of spatial orbitals @xmath81 forces us to restrict ourselves to @xmath124 .
correlated structures in the photoelectron spectra are also reproduced . however
, a quantitative agreement with the benchmark spectra obtained from the solution of the corresponding two - electron time - dependent schrdinger equation can only be achieved with more orbitals .
current work is devoted to the application of time - dependent renormalized - natural - orbital theory to he in full dimensionality , to more electrons , and to the mitigation of the truncation error via better - suited boundary conditions .
this work was supported by the sfb 652 of the german science foundation ( dfg ) .
38ifxundefined [ 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 link:\doibase 10.1103/revmodphys.84.1011 [ * * , ( ) ] link:\doibase 10.1080/09500340.2010.543958 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.84.443 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.84.447 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevlett.87.043003 [ * * , ( ) ] http://stacks.iop.org/1367-2630/16/i=3/a=033008 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.93.253001 [ * * , ( ) ] link:\doibase 10.1103/physreva.64.023405 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.94.143003 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevlett.97.203001 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.96.133001 [ * * , ( ) ] link:\doibase 10.1103/physreva.74.043420 [ * * , ( ) ] link:\doibase 10.1103/physreva.89.033415 [ * * , ( ) ] link:\doibase 10.1103/physreva.80.013412 [ * * , ( ) ] http://stacks.iop.org/0953-4075/37/i=4/a=004 [ * * , ( ) ] link:\doibase 10.1103/physreva.71.012712 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/j.amc.2005.04.027 [ * * , ( ) ] http://stacks.iop.org/1742-6596/220/i=1/a=012019 [ * * , ( ) ] link:\doibase 10.1103/physreva.73.032708 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.78.1884 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.71.1253 [ * * , ( ) ] link:\doibase 10.1103/physreva.88.052514 [ * * , ( ) ] link:\doibase 10.1103/physreva.90.012518 [ * * , ( ) ] link:\doibase 10.1103/physreva.56.3028 [ * * , ( ) ] link:\doibase 10.1088/0953 - 4075/31/6/001 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.85.4707 [ * * , ( ) ] link:\doibase 10.1103/physreva.76.023409 [ * * , ( ) ] link:\doibase 10.1103/physrev.97.1474 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.35.668 [ * * , ( ) ] @noop _ _ , springer lecture notes in chemistry 72 ( , , ) link:\doibase 10.1103/physreva.75.012506 [ * * , ( ) ] _ _ , @noop ph.d .
thesis , ( ) _ _ , @noop ph.d .
thesis , ( ) http://stacks.iop.org/0295-5075/92/i=2/a=23001 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/j.chemphys.2011.06.010 [ * * , ( ) ] \doibase http://dx.doi.org/10.1063/1.3687344 [ * * , ( ) ] note that , thanks to the sparsity of @xmath125 in , not all combinations of no indices contribute . | recently introduced time - dependent renormalized - natural - orbital theory ( tdrnot ) is tested on nonsequential double ionization ( nsdi ) of a numerically exactly solvable one - dimensional model he atom subject to few - cycle , 800-nm laser pulses .
nsdi of atoms in strong laser fields is a prime example of non - perturbative , highly correlated electron dynamics . as such
, nsdi is an important `` worst - case '' benchmark for any time - dependent few and many - body technique beyond linear response .
it is found that tdrnot reproduces the celebrated nsdi `` knee , '' i.e. , a many - order - of - magnitude enhancement of the double ionization yield ( as compared to purely sequential ionization ) with only the ten most significant natural orbitals ( nos ) per spin .
correlated photoelectron spectra as `` more differential '' observables
require more nos . |
the quality of the model predictions , concerning meson states , was discussed in ref .
figure 1 shows the lowest nucleon and @xmath0 resonances predicted by the model . in the same figure
are shown the calculated penta- and heptaquark low lying states .
for each state we indicate the spin , parity ( @xmath83 ) , and the quark and gluon content ( @xmath84 , @xmath85 ) .
the quantity @xmath84 is the total number of quarks and antiquarks , which is equal to the number of valence quarks ( 0 for mesons , 3 baryons ) plus the number of quarks and antiquarks of the @xmath86-pairs , and @xmath85 gives the number of gluons .
as shown in the figure , nucleonic states contain on the average about half an additional quark - antiquark pair ( equivalent to one extra quark ) , and approximately 2.8 gluons .
this implies a content of 59@xmath87 in the quark sector and of 41@xmath87 in the gluon sector .
the theoretical roper resonance ( first excited nucleon resonance ) lies near the experimental energy of 1.44 gev .
this is a nice feature of the model , which is shared only with the constituent quark model @xcite .
the first negative parity state with @xmath88 appears at 1.51 gev , also in good agreement with the data .
resonances ( second group ) , pentaquarks ( third group ) and heptaquarks ( fourth group ) . on the right side of each level
are given the assigned spin and parity ( @xmath89 ) , and the total quark and anti - quark ( @xmath84 ) and gluon ( @xmath85 ) contents ( see the text),width=340 ]
the simplest way to obtain a @xmath0 resonance is to couple the three valence quarks in the ( 3,0)@xmath90 configuration with @xmath91 pairs in a ( 0,0)@xmath92 configuration .
this scheme leads to the @xmath93 resonance .
the quark - antiquark and gluon content of the calculated @xmath93 turns out to be lower than that of the nucleon , while the structure of the state at 1.57gev can be compared with the roper resonance . concerning negative parity states , fig .
1 shows a @xmath94 state at 1.79 gev . @xmath0
resonances can also be obtained by coupling the three valence quarks in the @xmath95 configuration with @xmath96 @xmath86 states .
the lowest state of this type is at 1.51 gev , and it should be compared with the experimental value ( @xmath0 resonance ) at 1.62 gev @xcite .
in the present calculation the minimal representation of pentaquark - like states includes the following configurations : ( 0,0)@xmath97 , ( 1,1)@xmath97 , ( 3,0)@xmath97 , ( 0,3)@xmath97 and ( 2,2)@xmath97 . only the ( 0,3)@xmath97 and ( 2,2)@xmath97 configurations contains hypercharge and isospin combinations which can not be obtained in a pure @xmath98 coupling scheme , like @xmath99 , @xmath100 in ( 0,3 ) and @xmath101 , @xmath100 in ( 2,2 ) . within the model , the lowest pentaquark state @xmath2 @xcite
is interpreted as a coupling of the three valence quarks in ( 1,1)@xmath102 with the @xmath86 background in ( 1,1)@xmath103 to the final irrep ( 1,1)@xmath97 .
thus , within our model , the calculated pentaquark state at 1.51 gev may correspond to the observed @xmath2 state @xcite .
anther predicted pentaquark state is shown in figure . 1 . within the model ,
the lowest pentaquark has negative parity in accordance with qcd sum - rules and lattice gauge calculations @xcite .
if the orbital spin @xmath104 is included , pentaquark states with positive parity may exist with @xmath104=1 .
however , these states include an orbital excitation and should appear at higher energies .
the model contains heptaquarks , characterized by two @xmath86 pairs added to the three valence quarks .
the lowest state has an energy of approximately 2.5 gev and it has a content of 3.9 @xmath86 pairs of the type @xmath105 coupled to the three valence quarks in the configuration ( 1,1)@xmath102 . this coupling scheme yields three exotic flavor irreps : ( 3,3)@xmath102,@xmath106 , ( 4,1)@xmath102,@xmath106 and ( 1,4)@xmath102,@xmath106 .
the lowest heptaquark state contains , basically , three ideal valence quarks , four @xmath91 pairs and 4.6 gluons .
this implies a quark content of 70@xmath87 and a gluon content of 30@xmath87 .
the model predicts other heptaquark states at higher energies , which are obtained by coupling the three ideal valence quarks with the @xmath107 and @xmath108 @xmath86 configurations .
this leads to exotic flavor irreps like ( 4,1 ) , ( 1,4 ) and ( 3,3 ) with spin @xmath102 and @xmath106 .
the coupling of the three valence quarks with a @xmath86 irrep ( 2,2 ) @xmath22= 0 , 1 , 2 leads to exotic flavor irreps of the type ( 3,3 ) , ( 1,4 ) with spin - parity @xmath102 , @xmath106 and @xmath109 @xcite .
in this letter we do not go further into the discussion of all possible states with the structure @xmath110 , since the number of these configurations increases with the energy .
a more complete overview of these states will be presented in ref .
@xcite with its complete classification of states .
0.5 cm to conclude , we have applied a schematic model based on qcd degrees of freedom , to describe nucleon and @xmath0 resonances and more exotic penta- and heptaquark states .
the basis states were classified by applying group theoretical methods .
the hamiltonian , used in the calculations , was tested to the mesonic spectrum , nucleon and @xmath0 resonances .
after fixing the parameters in this manner , we have investigated the appearance of penta- and heptaquark states .
the results of the calculations show that the model predicts reasonably well the @xmath111(1540 ) resonance .
the lowest pentaquark state is obtained at an energy of approximately 1.5 gev and it has @xmath89=@xmath97 .
the lowest heptaquark state has an energy of approximately 2.5 gev and @xmath89 either @xmath102 or @xmath106 .
in addition , other penta- and heptaquark states are predicted to appear at higher energies .
the model allows for a complete classification of many quark - antiquark and gluon systems .
as we have shown , the exotic configurations which appear in our classification scheme can not be obtained in a simple constituent quark picture .
the overall agreement with the experimental data supports the claim about the suitability of the procedure .
this work obtained financial support from the dgapa project in119002 and from conacyt , conicet .
sj thanks the _ german academic interchange service _ ( daad ) for financial support . | a schematic model for hadronic states , based on constituent quarks and antiquarks and gluon pairs , is discussed .
the phenomenological interaction between quarks and gluons is qcd motivated .
the obtained hadronic spectrum leads to the identification of nucleon and @xmath0 resonances and to pentaquark and heptaquark states .
the predicted lowest pentaquark state ( @xmath1 ) lies at the energy of 1.5 gev and it is associated to the observed @xmath2 state . for heptaquarks ( @xmath3 )
the model predicts the lowest state at 2.5 gev . in a series of previous publications @xcite a schematic model for qcd
was developed .
the model was used to test the meson spectrum of qcd . in spite of its schematic nature
the model seems to contain the relevant degrees of freedom , as it was shown in the comparison between calculated and experimental meson spectra @xcite .
this letter is devoted to the extension of the model to accommodate baryonic features .
particularly , we shall concentrate on the appearance of exotic baryonic states , like pentaquark and heptaquark states @xcite
. the essentials of the model were discussed in detail in ref .
@xcite .
it consists of two fermionic levels in the quark ( @xmath4 ) and antiquark ( @xmath5 ) sector and a gluonic ( @xmath6 ) state containing pairs of gluons .
these are the elementary degrees of freedom of the model .
the interaction among these degrees of freedom is described by excitations of pairs of quarks and antiquarks mediated by the exchange of pairs of gluons .
the pairs of quarks are classified in a flavor - spin coupling scheme .
the pairs of gluons are kept in the angular momentum ( @xmath7 ) , parity ( @xmath8 ) and charge conjugation ( @xmath9 ) state @xmath10 .
the strength of various channels of the interaction , as well as the constituent masses , are taken from a phenomenological analysis .
the model describes meson ( @xmath11 ) states and baryonic ( @xmath12 ) states . among these states we focus on @xmath13 states ( pentaquarks ) and @xmath14 states ( heptaquarks ) , where the configurations indicated represent the leading terms in an expansion over many quark - antiquark and gluon states .
the basis states are classified using group theoretical methods @xcite .
the interaction of quark - antiquark pairs with gluon pairs is particle non - conserving .
the above described model belongs to a class of exactly solvable models of coupled fermion and boson systems @xcite .
alternative descriptions of pentaquark states were proposed in ref .
@xcite , enforcing particle number conservation . in what follows
we shall classify the basis states and solve the hamiltonian in the framework of the boson expansion method @xcite .
finally , we shall compare the results of the calculations with recently published experimental data @xcite the model hamiltonian is written the distance between the fermion levels is @xmath16=0.66 gev , @xmath17=1.6 gev is the energy of the glue ball , @xmath18 and @xmath19 are the number operators for fermion and gluon pairs , respectively , @xmath20 is the strength of the interaction in the flavor@xmath21 ) and spin ( @xmath22 ) channel .
the actual values @xmath23 = 0 , 1 refer to flavor ( 0,0 ) and ( 1,1 ) configurations , while the spin channel is @xmath22=0 or 1 .
the adopted values are : @xmath24=0.0337 gev , @xmath25=0.0422 gev , @xmath26=0.1573 gev , and @xmath27=0.0177 gev @xcite .
the operators @xmath28 and @xmath29 are boson images of quark - antiquark pairs @xcite .
the products which appear inside brackets in ( [ hamiltonian ] ) are scalar products .
the factor @xmath30 results from the boson mapping @xcite .
the mapping is exact for the channel @xmath31=[0,0]$ ] and simulates the effect of the boson mapping for the other channels . the operator @xmath32 ( @xmath33 ) creates ( annihilates ) gluon pairs with spin - color zero , and @xmath34 is the number operator of di - quarks coupled to flavor - spin @xmath35 .
the parameters @xmath36 and @xmath37 are adjusted to the nucleon and @xmath0 resonances .
the corresponding terms describe the interaction between valence quarks and glueballs .
the hamiltonian ( [ hamiltonian ] ) does not contain terms which distinguish between states with different hypercharge and isospin .
it does not contain flavor mixing terms , either .
therefore , the predicted states have to be corrected in the way described in @xcite to allow a comparison with data .
the adopted values of @xmath36 and @xmath37 are : @xmath38 - 1.442gev , @xmath39=-0.4388gev , @xmath40=-1.1873gev and @xmath41=-0.3622gev .
the hamiltonian contains all relevant degrees of freedom requested by qcd .
the complete classification of quark - antiquark configurations was given in ref .
@xcite . the unperturbed ground state is composed by 18 quarks occupying the lowest fermionic level .
the baryonic states are described by three quarks in the upper fermionic level to which we add @xmath42 states .
the group chain which describes these states is @xmath43 & [ h]=[h_1h_2h_3 ] & [ h^t ] \nonumber \\
u(4\omega ) & \supset u(\frac{\omega}{3 } ) \otimes & u(12 ) \label{group1}\end{aligned}\ ] ] where @xmath44=9 accounts for three color and three flavor degrees of freedom .
the irreducible representation ( irrep ) of @xmath45 is completely antisymmetric , and @xmath46 $ ] is the transposed young diagram of @xmath47 $ ] @xcite . for @xmath48 particles , and due to the antisymmetric irrep
@xmath49 $ ] of @xmath50 , the irreps of @xmath51 and @xmath52 are complementary and the irrep of @xmath51 is the color group , which is reduced to @xmath53 with the color irrep @xmath54 .
the @xmath52 group is further reduced to where @xmath56 is the flavor irrep and @xmath57 $ ] denotes the possible @xmath58 irreps .
the group reduction is done using the methods exposed in ref .
@xcite .
the basis is spanned by the states @xmath59 where @xmath48 is the number of particles , @xmath60 is the hypercharge and ( @xmath61,@xmath62 ) denotes the isospin and its third component , @xmath63 and @xmath64 are the multiplicities of the flavor and spin representations .
the color labels @xmath54 are related to the @xmath65 via @xmath66 and @xmath67 . to obtain the values of @xmath65 one
has to consider all possible partitions of @xmath68 , which fixes the color .
for colorless states we have @xmath69 .
each partition of @xmath48 appears only once .
the irrep @xmath70 $ ] of @xmath71 fixes the irrep of @xmath52 , as indicated in ( [ group1 ] ) . for @xmath72 and color ( 0,0 ) the irrep of @xmath52
is given by @xmath73 $ ] for mesons , and by @xmath74 $ ] for baryons .
as an example , table [ table1 ] shows the relevant irreps for mesonic states .
( more details are given in ref .
@xcite ) . .
flavor irreps coupled to the quark - antiquark content of some different @xmath58 irreps .
shown are the irreps which contain , at most , two quarks and two antiquarks .
the number of quarks ( antiquarks ) in a given configuration are denoted by @xmath75 ( @xmath76 ) . [ cols="<,<,<,<,<,<,<,<,<",options="header " , ] in the boson representation ,
the states are given by the direct product of one- , three- , eight and 24-dimensional harmonic oscillators @xcite .
for each harmonic oscillator the basis states are given by @xmath77 where @xmath78 is the number of bosons of type @xmath79 $ ] , @xmath80 is the corresponding seniority and @xmath81 is a normalization constant .
the seniority is defined as the number of uncoupled bosons .
the quantity @xmath82 represents the other quantum numbers needed to specify a particular harmonic oscillator . |
electro - negative gases , such as @xmath0 , @xmath1 , @xmath2 , and @xmath3 , play an important role in plasma - assisted materials processing .
most notably @xmath4 , either as the main feedstock gas or as an admixture to halogen- or silicon - based gases , is of vital importance for a large variety of etching and thin - film deposition techniques @xcite .
the requirements on the controllability and predictability of these processes are so high that further advancement of this technology will depend on modeling tools which go beyond the macroscopic , fluid - type approximations , which , for instance , can not reliably predict the velocity distributions of the species .
although electro - negative gas discharges have been studied for a long time @xcite , with significant progress made during the last two decades @xcite , a complete quantitative description is still lacking , in particular , with respect to discharge profiles , structuring , and operation regimes @xcite .
this is not surprising because the occurrence of negative ions leads to abrupt changes in the ion density ( density fronts @xcite ) which in most cases force the discharge to stratify into a central quasi - neutral ion - ion plasma and a peripheral electro - positive electron - ion plasma @xcite .
the transition between the two is rather subtle
. it can be , for instance , accompanied by a double layer ( internal sheath @xcite ) .
electro - negative gas discharges are thus rather complex and the investigation of the spatio - temporal structure of the discharge as a function of external control parameters ( current , voltage , frequency , pressure , and geometry ) is a great theoretical @xcite and experimental @xcite challenge . in addition , electro - negative processing gases are reactive molecular gases , with internal degrees of freedom , which lead to a quite involved plasma chemistry .
similar to the kinetic description of a mixture of reacting gases @xcite , the modeling of a gas discharge could be based on a coupled set of boltzmann equations for the distribution functions of the species which need to be treated kinetically .
this set has to be augmented by maxwell s equations , or parts of it , depending on how the discharge is electrically driven . even for simple reactive gas discharges , with only two negatively and one positively charged species ,
this approach is not practical .
more promising are methods which track representative samples of simulated particles subject to ( non - reactive and reactive ) collisions as well as electromagnetic fields , which are again determined from the relevant parts of maxwell s equations .
although these approaches are closely related to the direct simulation method successfully employed for the modeling of the flow of rarefied gases @xcite , in the context of gas discharges @xcite , it is more common to refer to them as particle - in - cell monte carlo collision ( pic - mcc ) methods @xcite , because without collisions the methods collapse to the pic approach @xcite for the solution of the vlasov problem .
this is the first paper in a series of three where we report the results of a combined experimental and theoretical study of the interplay between plasma - chemistry and electrodynamics in capacitively coupled radio - frequency ( rf ) discharges in oxygen .
oxygen is a weakly electro - negative gas whose plasma chemistry is strongly affected by vibrational , rotational and meta - stable states . up
to 75 reaction and scattering processes have been listed to potentially affect the properties of the discharge @xcite .
not all of them can be equally important .
the challenge is therefore to identify ( via comparison between experiment and modeling ) the subset of collisions responsible for the experimental findings under consideration . in the present paper ,
we describe such a reduced model .
it is tailor - made for the investigation of the charged species of the discharge , including the formation of ion density fronts , internal sheaths , and the resulting stratification of the discharge .
we critically assess the cross sections used to characterize the selected collision channels and point out some inconsistencies in previously used cross section data .
the validity of our model is verified by a comparison of calculated bulk densities with experimentally measured ones @xcite .
the following two papers discuss , from the experimental @xcite and theoretical @xcite point of view , respectively , the sheath region of the discharge focusing on the @xmath5 double emission layer in front of the powered electrode and a comparison of experimental and simulated distribution functions for positive ions .
the next section describes the main features of our pic - mcc implementation @xcite .
it is an extension of a code , which was originally designed for the particle - based modeling of sheaths in magnetized plasmas @xcite , to the case of electrically driven discharges in reactive molecular gases .
we focus , in particular , on the handling of collision processes , which deviates from the treatments usually employed in the plasma context @xcite and is closer to the simulation approaches used in rarefied gas dynamics @xcite .
the collisions included in our model are discussed in section iii .
where necessary , we combine measured cross sections @xcite with model cross sections @xcite to characterize collisions and fix the free parameters through a comparison with experimental data .
in section iv we show that once the parameters are fixed , good agreement between simulation and experiment can be achieved in the bulk of the discharge .
section v summarizes the essentials of our model and concludes with a short outlook .
schematic geometry of the rf discharges used in refs .
we use a planar , one - dimensional model to simulate the central axial part of the discharge ( thick solid line ) . ]
we are only interested in the central axial part of the rf discharges described in refs .
@xcite . ignoring the electric asymmetry between the powered and grounded electrode ,
this part of the discharge can be simulated by the planar , one - dimensional model shown in figure [ geometry ] .
the planar model retains only one spatial coordinate @xmath6 , with @xmath7 , where @xmath8 is the distance between the electrodes , but it keeps the full three - dimensional velocity space for the particles ( 1d3v model ) .
one of the electrodes is driven by the rf voltage @xmath9 , with @xmath10 and @xmath11 ranging from @xmath12 to @xmath13 , while the other is grounded .
both electrodes are assumed to be totally absorbing ; secondary electron emission is neglected . to mimic the constant oxygen flow through the discharge chamber ,
we enforce in the simulation volume a constant oxygen pressure ranging from @xmath14 to @xmath15 .
the simulation volume of the planar model is of course not well defined because the lateral `` cross - section '' a is a free parameter .
this parameter actually controls the weight of the simulated particles , that is , the number of real particles represented by one simulated particle .
we arbitrarily choose @xmath16 with @xmath17 the electron debye length of a reference electron system ( res ) which we use to initially set - up the length and time scales of the simulation ( see below ) .
the res is specified by a density @xmath18 and a temperature @xmath19 .
both have to be adjusted to the particular experimental conditions . for the experiments
@xcite , @xmath20@xmath21 and @xmath22 , resulting in approximately @xmath23 simulated particles .
the complex plasma - chemistry of oxygen gives rise to a large number of non - reactive and reactive collisions @xcite . in table
[ reactions ] we show the ones with the largest cross sections .
they are included in our model and will be discussed in detail in section iii .
we treat only three species kinetically : electrons ( @xmath24 ) , negatively charged oxygen atoms ( @xmath25 ) , and positively charged oxygen molecules ( @xmath26 ) .
neutral particles appearing either as educts or products in table [ reactions ] are not explicitly simulated . molecular oxygen in its ground state , @xmath27 ( feedstock gas ) , is modelled as a reservoir characterized by a pressure @xmath28 and a temperature @xmath29 . whereas atomic oxygen , @xmath30 , vibrationally excited oxygen molecules , @xmath31 , the rydberg state of the molecular oxygen , @xmath32 , and the molecular meta - stables , @xmath33 and @xmath34 , are only indirectly accounted for in as far as their production results in an energy loss for electrons .
the meta - stable @xmath33 requires special attention because it also appears as an educt in the entrance channel for associative detachment ( 17 ) .
its concentration is therefore important , and we should actually build - up the @xmath33 distribution function , that is , we should also simulate @xmath33 particles . in that case , however , not only their volume production process ( 11 ) but also their volume and surface loss processes should be included .
this is beyond the one - dimensional model . to take associative detachment ,
which is known to be an important process @xcite , nevertheless into account , we use instead a simple model with one free parameter , which can be interpreted as the @xmath33 to @xmath27 density ratio ( see below ) .
the three - species pic - mcc model describes the physics of an @xmath4 discharge by tracing the spatio - temporal evolution of an ensemble of charged particles ( @xmath24 , @xmath35 , and @xmath36 ) . as usual , to track the velocities and positions of the simulated particles , the simulation domain @xmath37 and the time are discretized and free flights are decoupled from collisions ( see figure [ alg ] ) . in order to have stable free flights of charged particles ( in the mean electric field given by the poisson equation ) , the spatial resolution @xmath38 has to be less then the electronic debye length , @xmath17 , and the time step @xmath39 should resolve the electron plasma oscillation whose frequency is @xmath40 @xcite .
both @xmath17 and @xmath40 are set by the res .
the time interval over which collisions and free flights are decoupled is of the order of the mean free time , which , in general , is much larger then the fundamental time step @xmath39 enforced by the electric field .
we distinguish two groups of collisions , each of which characterized by an average mean free time .
the first group of collisions , occurring on a time scale @xmath41 comprises all collisions of table [ reactions ] except coulomb collisions between charged particles , which take place on a different time scale @xmath42 .
graphical representation of our implementation of the pic - mcc scheme . ] to build up a representative sample for the charged particles , we first sample the initial distribution functions for @xmath24 , @xmath35 , and @xmath36 and then repeat the following procedure @xcite ( see figure [ alg ] ) : * for @xmath43 time steps , the coordinates and velocities of all simulated particles are moved according to newton s equations of motion .
the force acting on particles with charge @xmath44 is given by @xmath45 where @xmath46 is the electric potential satisfying the poisson equation with the boundary conditions specified above .
force interpolation and charge assignment are done within a linear interpolation scheme . *
after @xmath47 time steps , a representative sample of the first class of collisions is executed , with post - collision velocities stochastically determined in accordance with energy and momentum conservation , followed by free flights with @xmath48 time steps , where @xmath49 , after which a representative random sample of coulomb collisions is performed , again consistent with energy and momentum conservation .
after typically @xmath50 rf cycles quasi - stationarity with respect to the rf cycle is achieved , that is , within the statistical noise , macroscopic quantities , for instance , density and potential profiles , do not change anymore when averaged over a rf cycle .
our treatment of collisions utilizes a number of concepts originally developed for the direct simulation monte carlo of rarefied gas flows @xcite . since it differs from the handling of collisions in the standard pic - mcc schemes , which is based on the null - collision method @xcite , we give some more details . after a free flight
is completed , we sort the simulated particles into cells ( see figure [ alg ] ) .
simulated and feedstock gas particles located in one and the same cell ( collisions are local ) have then the opportunity to sequentially perform all possible types of collisions with each other .
simulated particles produced in the exit channel of a collision are readmitted for the next collision in the list .
the ordering of particles into cells has to be repeated after each collision producing or annihilating simulated particles .
thus , as in the real system , simulated particles may perform more than one type of collision within the time interval over which the simulation decouples free flights and collisions .
the probability of a simulated particle @xmath51 to make a collision of type @xmath52 is given by @xmath53 it depends on the density @xmath54 of the collision partners which act as the targets , the total cross section @xmath55 for the collision , the relative velocity @xmath56 between particle @xmath51 and a particular collision partner , and the mean free time @xmath57 . as can be seen in table [ reactions ] , except for associative detachment ( 17 ) , which we discuss in more detail in the next section , the three species model recruits collision partners either from the simulated particles or from the feedstock gas @xmath3 . in the first case ,
@xmath54 is the number of simulated particles of the respective type in the considered cell divided by the cell volume , whereas in the second case , @xmath58 , with @xmath28 the gas pressure and @xmath29 the room temperature .
we do not calculate the mean free times @xmath57 for all the collisions listed in table [ reactions ] .
instead , we set @xmath59 for coulomb collisions ( 1 ) ( 3 ) and @xmath60 for the remaining collisions ( 4 ) ( 19 ) .
in order to verify that our code treats collisions correctly , we performed various test runs and compared the results either with analytical results @xcite or with results obtained from the bit1 code @xcite .
the latter was also used to check discrepancies in the cross section data .
.[reactions ] collisions included in our model [ cols= " < , < " , ] collisions strongly influence the particle concentration and the energy balance in the discharge . for oxygen ,
an overwhelming number of elastic , inelastic , and reactive collisions is possible .
restricting , however , the description to a three species plasma ( @xmath24 , @xmath35 , and @xmath36 ) many processes can be ignored and others can be treated approximately . in table
[ reactions ] we show the collisions defining our three species model for an oxygen discharge , and classify them into three groups : elastic scattering , electron energy loss scattering , and electron and ion production and loss reactions .
the respective cross sections for these processes are shown in figures [ elasticxs ] [ prodlossxs ] as a function of the relative energy of the two particles in the entrance channel of the respective process .
our collection of cross sections is semi - empirical , combining measured data , which is usually available only in a finite energy range , with simple models for the low - energy asymptotic , which in most cases is not very well known from experiments . in general , the high - energy asymptotic has to be also determined from models , but it is less critical because the distribution functions usually decay sufficiently fast at high energies . if not stated otherwise , we extrapolated therefore the values of the cross sections for the largest energies shown in the plots to all energies above it .
some of the cross sections significantly deviate from the ones previously used @xcite .
our simulations indicate , however , that the modifications are essential for obtaining bulk densities in accordance with experiments @xcite .
elastic collisions ( 1 ) ( 6 ) are particle number conserving .
thus , when a collision takes place , only the post - collision velocities of the scattering partners have to be determined whereas the list of simulated particles remains unaltered . the first three scattering processes ( 1 )
( 3 ) are intra - species coulomb collisions .
they are not very important for the experiments @xcite we analyze in this series of papers . for other parameter regimes , however , the _ local _ ion density in an electro - negative gas discharge can be rather high @xcite and it can not be ruled out that ion - ion coulomb collisions affect , for instance , the width of these ion density peaks . since the computational burden is moderate , we kept intra - species coulomb scattering in all our simulations .
if not stated otherwise , inter - species coulomb collisions were however neglected . for coulomb collisions , we used a binary collision model @xcite with a uniformly distributed azimuth angle @xmath61 , and a gaussian distribution for @xmath62 , where @xmath63 is the scattering angle .
the second moment of the gaussian distribution is given by @xcite @xmath64 with @xmath65 the collision time for coulomb scattering , @xmath66 the coulomb logarithm , @xmath67 the magnitude of the relative velocity in the center - of - mass frame , and @xmath68 , @xmath69 , @xmath70 , and @xmath71 the local density , the charge , the reduced mass of the charged species under consideration , and the dielectric constant of the vacuum , respectively .
cross sections for elastic scattering of electrons ( 4 ) , @xmath35 ions ( 5 ) , and @xmath36 ions ( 6 ) on oxygen molecules .
the dashed line and the dotted line indicate , respectively , the cross section for elastic ( @xmath72 ) scattering and the cross section for ( @xmath73 ) charge exchange scattering used in ref . @xcite . ]
figure [ elasticxs ] shows the cross sections for elastic scattering of electrons ( 4 ) and @xmath35 ions ( 5 ) on @xmath4 molecules together with the cross section for charge exchange scattering ( 6 ) of @xmath36 ions on @xmath4 molecules .
since little is known about the angle dependence of these processes , we assume the collisions to be isotropic , with a uniform distribution of the azimuth angle @xmath61 and a uniform distribution of @xmath74 , where @xmath63 is again the scattering angle .
the cross section for elastic scattering of electrons on @xmath4 @xcite is the same as in ref .
the other two cross sections are different . for charge exchange scattering
, we constructed the cross section as follows .
below @xmath75 and above @xmath76 , we used empirical data for momentum transfer scattering @xcite , together with the expression @xmath77 where @xmath78 and @xmath79 denote , respectively , the momentum and charge exchange cross section . for energies in between , we employed a linear interpolation . with this cross section , we could reproduce experimentally measured @xmath36 velocity distribution functions very well @xcite . with the charge exchange cross section given in ref .
@xcite ( dotted line in figure [ elasticxs ] ) , on the other hand , we could not obtain the correct distribution functions neither with our pic - mcc code nor with the bit1 code @xcite , which we used in addition to exclude possible mistakes in our collision routines .
the cross section for elastic scattering of @xmath35 on @xmath4 originates from a model which assumes polarization scattering between the two , induced by a central potential @xmath80 with @xmath81 .
it is given by @xmath82}\big)^{2/n}+\sigma_0\big ] \cdot cm^2~ , \label{okada}\end{aligned}\ ] ] with @xmath83 and @xmath84 @xcite .
here , we added a constant shift @xmath85 to match the cross section at high energies as given in ref .
@xcite .
cross sections for the @xmath86 and @xmath87 vibrational excitation of @xmath4 molecules ( 7 ) . ]
electron energy loss occurs due to inelastic collisions , ( 7 ) ( 12 ) , in which the oxygen molecule is either promoted into an excited state or dissociated into neutral fragments . in the three species model , where the spatio - temporal evolution of the neutral fragments and the excited oxygen molecules is not traced , we can treat this group of processes in the spirit of `` test particle collisions '' where the electron is the `` test particle '' , suffering momentum as well as energy transfer , and the @xmath4 molecule is the `` field particle '' ( with internal degrees of freedom ) . drawing the field particles from a maxwell distribution characterized by the gas temperature @xmath29 and the gas density @xmath88 , the same binary collision model can be used as for elastic scattering @xcite .
the only difference is that now the magnitude of the post - collision relative velocity in the center - of - mass system is given by @xmath89 where @xmath67 is the magnitude of the pre - collision relative velocity in the center - of - mass frame , @xmath90 is the reduced mass of the scattering partners , which , in general , have not the same mass , and @xmath91 is the excitation energy of the process . due to lack of angle - resolved scattering cross sections , we again assume the collisions to be isotropic .
cross sections for rydberg excitation ( 8) , @xmath92 ( 9 ) and @xmath93 ( 10 ) dissociative excitation , @xmath94 excitation ( 11 ) , and @xmath95 excitation ( 12 ) of @xmath4 molecules . ]
the most important electron energy losses are due to vibrational and electronic excitation and impact dissociation of the oxygen molecule .
the cross sections for these processes @xcite are the same as the ones used in ref .
@xcite . for completeness
, they are plotted in figures [ inelasticxs_v ] and [ inelasticxs_e ] .
we consider vibrational excitations with @xmath96 , and @xmath87 , excitation of the @xmath4 rydberg state , @xmath97 and @xmath98 dissociation , and the excitation of the meta - stable states @xmath99 and @xmath100 .
in contrast to ref .
@xcite , we ignore @xmath101 dissociation , because , in the energy range of interest , its cross section is one order of magnitude smaller than for @xmath97 and @xmath98 dissociation , and rotational excitation which is almost elastic ( small energy transfer ) and thus dominated by the much more efficient elastic scattering of electrons on @xmath4 .
electron and ion production and loss reactions ( 13 ) ( 19 ) are inelastic collisions which change the number of charged particles .
therefore , the list of simulated particles has to be up - dated .
all collisions obey conservation laws and are again assumed to be isotropic .
the details of the modeling depend on the process .
the two recombination channels ( 13 ) and ( 14 ) result simply in the annihilation of the two oppositely charged particles which participate in the process . binary collisions with
one charged particle in the exit channel , such as dissociative attachment ( 15 ) and associative detachment ( 17 ) , are treated within the modified binary collision model described in the previous subsection .
impact ionization ( 18 ) and impact detachment ( 19 ) are modelled as follows : first , an inelastic binary collision is performed , in which the parent electron looses the ionization ( detachment ) energy .
then , the post - collision @xmath4 ( @xmath35 ) particle is split into an electron and a @xmath36 ( @xmath102 ) particle .
finally , an elastic binary collision is applied to distribute energy among the two charged particles .
modeling three particle processes as a sequence of two binary collisions with a particle splitting in between guarantees energy and momentum conservation which is critical for the stability of simulations @xcite .
although direct detachment ( 16 ) could be modelled in the same spirit , we adopted a simpler approach . from experiments @xcite we know that the energy of the released electron is approximately one - tenth of the energy of the primary @xmath103 ion . to obtain the energy distribution of the ejected electron we hardwired therefore this ratio in the collision routine for direct detachment .
cross sections for dissociative recombination ( 13 ) , neutralization ( 14 ) , dissociative attachment ( 15 ) , direct detachment ( 16 ) , associative detachment ( 17 ) , impact ionization ( 18 ) , and impact detachment ( 19 ) .
the dashed line indicates the experimentally determined high - energy asymptotic of the neutralization cross section and the dotted line is the cross section for neutralization used in ref .
@xcite . ]
the cross sections for electron and ion production and loss reactions are shown in figure [ prodlossxs ] .
for impact ionization @xcite , dissociative attachment @xcite , and impact detachment @xcite we used empirical cross sections throughout , putting , in the case of impact detachment a higher confidence level to more recent data @xcite , whereas for dissociative recombination , ion - ion neutralization , and detachment on neutrals we combined empirical data with analytical models .
first , we discuss dissociative recombination ( 13 ) . for this process we used an effective cross section , @xmath104 where @xmath85 is
the cross section for non - resonant dissociative recombination @xcite and @xmath105 is a scaling factor , which is used to fit experimental data @xcite .
a comparison of the effective cross section with a calculated cross section @xcite shows that @xmath106 can be also considered as a simple device to approximately take the effect of resonant dissociative recombination channels into account which originate from vibrationally excited states of @xmath107 @xcite . for ion - ion neutralization ( 14 )
, we constructed a cross section from a two - channel landau - zener model @xcite , with one free parameter , which we adjusted to obtain the correct high - energy asymptotic of the cross section @xcite .
ion - ion neutralization occurs because the adiabatic energy of the @xmath108 configuration decreases when the ions approach each other . at a certain distance @xmath109 ,
the energy of the @xmath108 configuration falls below the energy of the @xmath110 configuration @xcite .
the landau - zener theory estimates the probability for switching from one configuration to the other at the distance @xmath109 and leads to a cross section , @xmath111 which approaches a constant @xmath112 for @xmath113 .
fitting @xmath114 and thus @xmath109 to empirical data at high energies @xcite , we obtain @xmath115}\big)\cdot 10^{-14}\cdot cm^2~. \label{xsection_recomb}\end{aligned}\ ] ] this cross section deviates dramatically from the one used in ref .
@xcite ( dotted line in figure [ prodlossxs ] ) .
the most severe modification we made was for detachment of @xmath35 on neutrals .
it takes place through direct detachment ( 16 ) , resulting in an atomic oxygen , an oxygen molecule and an electron , and through associative detachment ( 17 ) , which leads to an @xmath116 molecule and an electron .
the latter is rather surprising because there is no evidence for it in beam experiments ( where only direct detachment is observed @xcite ) . yet , experimental studies of @xmath4 discharges @xcite ( as well as general theoretical considerations @xcite ) strongly suggest that associative detachment is possible in an oxygen discharge because of the presence of meta - stable @xmath117 .
in contrast to direct detachment , which has a threshold around @xmath118 @xcite ( see figure [ prodlossxs ] ) , associative detachment has usually no threshold .
thus , it may be a rather important loss channel for cold @xmath35 ions .
since we could not find an empirical cross section for associative detachment , we employed a model , which describes the detachment ( electron loss ) as the `` inverse '' of a classical langevin - type electron capture into an attractive auto - detaching state of @xmath119 . assuming the polarizability for @xmath117 to be the same as for @xmath4 , the cross section is then given by @xcite @xmath120}}~. \label{xsection_ad}\end{aligned}\ ] ] as can be seen in figure [ prodlossxs ] , associate detachment is already the dominant detachment process for energies below @xmath121 .
voltage dependence of the quasi - stationary , cycle - averaged electron and ion densities at @xmath122 ( bulk densities ) for a discharge with @xmath123 , @xmath124 , and @xmath10 . filled and open symbols denote , respectively ,
results of the simulation and experimentally measured densities @xcite .
solid and dashed lines guide the eye . ] from the cross section alone , of course , we can not determine the probability @xmath125 for associative detachment .
we also need the density of @xmath126 , which is unknown in the three species model . however , it should be of the order of the @xmath4 density .
therefore , we write @xmath127 , with @xmath128 , and obtain @xmath129 , where @xmath114 is a fit parameter which can be adjusted to experiments .
the plasma model described in the previous sections contains three parameters : @xmath85 , @xmath106 , and @xmath114 .
the first two we fixed by a direct comparison of the model cross section with the measured cross section .
the parameter @xmath114 , in contrast , can not be determined in this way , because it is not linked to a cross section .
it denotes instead the fraction of @xmath4 molecules in the meta - stable @xmath126 state which in turn depends on the particular set - up of the discharge .
we consider therefore @xmath114 as a free parameter which can be adjusted to the discharge to be modelled . to validate our model we focus now on the oxygen discharge described in ref .
@xcite . in order to determine @xmath114 , we simulated the discharge for @xmath124 , @xmath130 , and @xmath10 , and tuned @xmath114 to reproduce the quasi - stationary , cycle - averaged negative and positive ion densities at @xmath122 ( bulk densities ) ; the bulk electron density matches then also because of quasi - neutrality in the bulk of the discharge .
we obtained @xmath131 , implying that roughly one sixth of the @xmath4 molecules is in the meta - stable state . in figure
[ central ] we plot the bulk ion and electron densities for @xmath124 over a wide voltage range , using however for all voltages @xmath131 , the value determined for @xmath130 . as can be seen ,
the agreement between simulation and experimental data @xcite is rather good , indicating that our model captures the essential processes in the bulk of an oxygen discharge .
the precise value of @xmath114 should not be taken too serious because it is based on a rather crude model for associative detachment .
more important is that without this process ( @xmath132 ) , the simulation could not reproduce the measured densities .
this can be seen in figure [ profiles ] , where we plot the quasi - stationary , cycle - averaged density profiles without ( thin lines ) and with associative detachment ( thick lines ) taken into account .
( for these runs , we also included inter - species coulomb collisions but the profiles without it are basically the same . ) neglecting associative detachment , the bulk ion densities turned out to be almost one order of magnitude too high . remarkably , without associative detachment , the density profiles have the parabolic shape expected from ambipolar drift - diffusion models @xcite . quasi - stationary , cycle - averaged electron and ion density profiles for @xmath133 without ( thin lines ) and with ( thick lines ) associative detachment taken into account ; @xmath28 , @xmath8 , and @xmath134 are the same as in figure [ central ] .
open circles and open squares denote , respectively , the measured electron and @xmath35 density profiles @xcite .
note , the runs producing the shown data included inter - species coulomb scattering . ] to understand why associative detachment is so crucial for modeling the discharge of ref .
@xcite , we plot in figure [ fminus ] for @xmath124 and @xmath135 the ( marginal ) velocity distribution function for negative ions , @xmath136 with @xmath137 .
clearly , the majority of @xmath35 ions is cold ; the distribution function spreads out only in the region where the bulk plasma merges with the peripheral plasma at @xmath138 .
thus , loss processes whose cross sections are large for small energies ( see figure [ prodlossxs ] ) , that is , ion - ion neutralization and associative detachment , will be very efficient .
direct detachment , on the other hand , whose cross section has a threshold , will be suppressed . at low energies ,
both ion - ion neutralization and associative detachment have cross sections which increase with decreasing energy .
which process dominates depends therefore also on the collision probability @xmath139 . for ion - ion neutralization @xmath139
is proportional to the @xmath36 density while for associative detachment it is proportional to the @xmath126 density . since for the discharge of ref .
@xcite , the @xmath36 density is much smaller than the @xmath126 density , which we estimated to be @xmath140 of the @xmath4 density , associative detachment has to be the main loss process for negative ions in this experiment .
quasi - stationary , cycle - averaged negative ion velocity distribution function @xmath136 with @xmath137 for @xmath135 ; @xmath28 , @xmath8 , and @xmath134 as in figure [ central ] . ] additional support for this claim comes from the weak voltage dependence of the @xmath35 density between @xmath12 and @xmath141 indicating that both @xmath35 production and @xmath35 loss are in this voltage range rather insensitive to voltage changes .
we will now show that this strongly suggests that associative detachment is the main loss process for @xmath35 ions .
first , we conclude from figure [ felectron ] , which shows the attachment cross section ( in arbitrary units ) together with bulk electron velocity distribution functions for different voltages , that the energy range between @xmath142 and @xmath143 will be the most important one for the attachment process ; for energies below @xmath142 the production is zero because the cross section is zero while for energies above @xmath143 the number of electrons available for attachment is too low .
we also see that in the voltage range of interest , the electron velocity distribution function does not change much in this energy range .
thus , the production process is almost voltage - independent and the observed near constancy of the @xmath35 density has to be taken as a signature of the loss process .
as already mentioned , direct detachment can be ruled out .
ion - ion neutralization , on the other hand , would lead for @xmath144 , that is , in the relevant energy range , to an energy - resolved rate coefficient @xmath145 which strongly increases with decreasing energy .
thus , even small , voltage - induced changes of the energy of negative ions would suffice to lead to noticeable modifications of the loss rate .
what remains is associative detachment .
the energy - resolved rate coefficient for this process is @xmath146 . at the same time , the distribution function for @xmath103 ions does not change much between @xmath12 to @xmath141 .
thus , the loss rate due to associative detachment is nearly independent of voltage .
together with the voltage independence of dissociative attachment , this leads to the observed constancy of the @xmath103 density .
quasi - stationary , cycle - averaged electron velocity distribution function @xmath147 with @xmath148 and @xmath122 for @xmath149 , and @xmath141 ; @xmath28 , @xmath8 , and @xmath134 as in figure [ central ] .
the dotted line indicates the cross section for dissociative attachment ( 15 ) in arbitrary units . ]
our simulation reproduces the absolute values of bulk densities fairly well .
the ( axial ) density profiles of the simulation decay however too fast compared to the experimental ones ( open symbols in figure [ profiles ] ) .
we attribute this to the fact that the simulations assume a constant @xmath117 density , whereas , in reality , there should be a density profile , because of the interplay of volume and surface loss and generation processes for @xmath117 molecules . in particular surface losses , that is , the decay of @xmath117 molecules when they hit the boundary of the discharge should play an important role because they lead to a depletion of @xmath117 molecules in the vicinity of the electrodes and thus to a reduction of associative detachment .
a detailed investigation of the `` drag '' the @xmath117 density is expected to exert on the @xmath103 density is however beyond the scope of the present paper .
we presented a planar , one - dimensional pic - mcc model for a capacitively coupled rf discharge in oxygen which is capable to quantitatively describe experiments .
its main features are : ( i ) only electrons , @xmath35 ions , and @xmath36 ions are treated kinetically .
( ii ) neutral particles are only incorporated in as far as they affect the particle and energy balance of simulated charged particles .
( iii ) a direct simulation monte carlo model for collisions has been used , with two groups of collisions , which differ in the collision times .
the elementary processes of our three species model are shown in table [ reactions ] .
we took the processes with the largest cross sections into account .
when possible , we implemented empirical cross sections . in some cases , however , we combined them with analytical model cross sections to obtain the correct low - energy asymptotic .
in particular , we used a polarization - type scattering model for elastic @xmath150 scattering , a landau - zener - type model for @xmath151 neutralization , and an inverse classical , langevin - type capturing model for @xmath117-induced associative detachment . to describe associative detachment due to @xmath117 without explicitly calculating the @xmath117 density
, we furthermore introduced a parameter @xmath114 , which should be fitted to the particular oxygen discharge under consideration and could be interpreted as the fraction of @xmath4 molecules in the meta - stable @xmath117 state .
as a first application of our model , we simulated the discharge of ref . @xcite . after we adjusted the parameter @xmath114 to reproduce the bulk ion densities for @xmath124 and @xmath130 , we also obtained the correct bulk densities for other voltages .
the parameter @xmath131 , that is , approximately one sixth of the oxygen molecules are in the meta - stable state .
we also pointed out that the weak voltage dependence of the bulk densities for fixed pressure indicates that @xmath35 losses due to associative detachment dominate the ones due to ion - ion neutralization . although we could reasonably well describe electron and ion densities in the bulk , the ( axial ) ion density profiles of the simulation are too narrow compared to the experimental ones .
most probably this is because we assumed a constant @xmath117 density . in reality , however , @xmath117 molecules decay when they hit the boundary of the discharge .
the @xmath117 density should therefore decrease in the vicinity of the electrodes and with it associative detachment . to take this effect into account
requires however a model which not only treats electrons , @xmath35 ions , and @xmath36 ions kinetically , but also @xmath117 molecules .
our concern in the present paper , which is the first in a series of three , was the calibration and validation of a three species model for capacitively coupled rf discharges in oxygen .
for that purpose , we used published experimental data for the bulk of the discharge @xcite .
the following paper @xcite describes the results of an experimental investigation of the sheath region of such a discharge , focusing , among others , on the spatial and temporal evolution of the @xmath5 double emission layer in front of the powered electrode . in the third paper @xcite ,
finally , we will use the three species model to simulate the sheath region . besides a comparison of simulated with measured distribution functions for positive ions
, we will also present a model for the @xmath5 double emission layer .
the three species model , including its approximate treatment of associative detachment , is sufficient for that purpose , because negative ions are negligible in the region from which the emission originates .
negative ions are only needed for the overall charge balance of the discharge .
their density profile per se is not crucial for the explanation of the @xmath5 double emission layer as long as it features a vanishingly small density in front of the electrode .
this is already accomplished by the three species model .
+ support from the sfb - tr 24 `` complex plasmas '' is greatly acknowledged .
we thank b. bruhn , h. deutsch , k. dittmann , and j. meichsner for valuable discussions and critical reading of the manuscript .
k. m. and r. s. acknowledge funding by the initiative and networking fund of the helmholtz association and f. x. b. acknowledges special funding 0770/461.01 by the state mecklenburg - vorpommern . | in this series of three papers we present results from a combined experimental and theoretical effort to quantitatively describe capacitively coupled radio - frequency discharges in oxygen . the particle - in - cell monte - carlo model on which the theoretical description is based will be described in the present paper .
it treats space charge fields and transport processes on an equal footing with the most important plasma - chemical reactions .
for given external voltage and pressure , the model determines the electric potential within the discharge and the distribution functions for electrons , negatively charged atomic oxygen , and positively charged molecular oxygen .
previously used scattering and reaction cross section data are critically assessed and in some cases modified . to validate our model , we compare the densities in the bulk of the discharge with experimental data and find good agreement , indicating that essential aspects of an oxygen discharge are captured . |
interference effects are ubiquitous in quantum mechanics . at the beginning of modern quantum theory dirac enunciated a criterion for the existence of interference effects between photons @xcite : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ each photon interferes only with itself .
interference between two different photons never occurs . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ originally dirac formulated the principle only for photons , but soon it was extended to the case of massive particles .
later , following the analysis of some new developments in quantum theory it was realized that the criterion should be interpreted in a more subtle way .
these developments refer to situations where the wave function describing a multiparticle system can not be separated into the product of the wave functions of the particles composing the complete system .
there are two well - known scenarios where these states appear .
one is by preparing the particles in an initially entangled state . in the language of wave functions
this property translates into the impossibility of separating the wave function of the complete system .
the other scenario is related to the spin - statistics connection .
the multiparticle wave function of bosons or fermions must be symmetrized or antisymmetrized resulting in non separable wave functions . in both scenarios
it is impossible to speak about the properties of any of the particles as an independent entity .
one is tempted to interpret the interference effects associated with these systems as interferences between different particles . however , as the properties of the individual particles are only defined within a larger entity , the complete multiparticle system ( the only entity that is defined from the quantum point of view ) , we really observe interferences of the complete system . according to silverman , this extended dirac s criterion can be expressed as @xcite : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ a system interferes only with itself .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the extended formulation reflects the fundamental property of entangled systems of loosing the individuality of the particles within the multiparticle system , the only _ single entity _ from a quantum point of view .
recently , it has been discussed in the literature other experimental scenario where the extended interpretation must be used , in spite of the fact that the situation refers to non - entangled photons emitted by independent sources @xcite .
two sources emit independently photons , which are detected at two different positions .
when the composed paths ( the paths of the two photons ) between the two sources and the two detectors are indistinguishable we must add the probability amplitudes obtaining interferences in the joint detection probability .
once more we do not face interferences between both photons , since the probability amplitudes that we add are those of composed paths .
the composed paths are not properties of the individual photons , but a property of the two - photon system .
we present in this paper a new example of multiparticle system for which it is necessary the extended interpretation .
it is based on a recently proposed arrangement @xcite , in which two sources independently emit particles in non - entangled multimode states . when the particles have common modes the probability of detecting only one of the two particles shows interference effects that must be associated with the complete multiparticle system .
we briefly describe the ideal arrangement considered in this paper , which has been presented in detail in ref .
4 ( see fig 1 ) . two different sources independently emit indistinguishable particles . by the matter of simplicity
we restrict our considerations to bosons ( the case of fermions can be developed along very similar lines , but needs from a more elaborated and lengthy discussion of some technical aspects @xcite ) . in the region of overlapping of the two beams we place at a fixed position a detector .
we study the detections that occur at a given time @xmath0 .
we concentrate on the cases where only one of the two particles is detected , disregarding the events when the two particles are detected simultaneously .
we assume , by simplicity , that the detector can distinguish between one- and two - particle detection events ( see ref . 4 for a realistic arrangement with this property ) . and @xmath1
both beams come together and mix at the position of the detector.,width=453,height=680 ] the particles are emitted by the sources in the state @xmath2 where @xmath3 and @xmath4 represent the momenta of the particles , @xmath5 and @xmath6 are the complex distributions of momenta , @xmath7 is the creation operator of a particle with momentum @xmath3 and @xmath8 refers to the vacuum state in fock s space .
moreover , by simplicity , we have assumed both particles to be in the same state of spin .
the extension to particles in different states is straightforward .
the form of eq .
( [ eq : uno ] ) follows directly from the independent nature of the two sources .
when the momenta distributions are non - trivial , i. e. , when they are different from zero for several values of the momentum we have a multimode distribution .
we assume the modes to be plane waves , i. e. , the mode of momentum @xmath3 is @xmath9 .
we assume both mode distributions to be normalized to unity : @xmath10 using the commutation relations ( both particles are identical bosons ) , @xmath11 , and the property of the annihilation operators @xmath12 , we have @xmath13 to obtain this relation we have also used the normalization conditions ( [ eq : dos ] ) and the usual normalization of the vacuum , @xmath14 .
the probability of detecting only one of the two particles at point @xmath15 and time @xmath0 is given by @xcite : @xmath16 where @xmath17 is the schrdinger field operator , which in the plane wave representation for the modes is given by @xmath18 where @xmath19 is the energy of the mode @xmath20 . equation ( [ eq : cuat ] ) can be evaluated in the same way done for the denominator ( eq .
( [ eq : tres ] ) ) .
interchanging the parameters @xmath21 , @xmath4 and @xmath3 where necessary the result is @xmath22 we rearrange these expressions in a more tractable way .
we introduce the following notation : @xmath23 from the definition and eq .
( [ eq : dos ] ) they follow easily relations @xmath24 and @xmath25 .
these coefficients correspond to the integrals in eq .
( [ eq : siet ] ) that do not contain any dependence on @xmath15 or @xmath0 . on the other hand ,
we introduce the time and position dependent functions : @xmath26 note that @xmath27 . combining all these expressions ,
we obtain for the detection probability : @xmath28 using the properties of the @xmath29 and @xmath30 above remarked we have finally the equation @xmath31 where @xmath32 denotes the real part of the complex expression @xmath33 .
equation ( [ eq : once ] ) shows the existence of an interference phenomenon .
the detection probability is composed of three terms . @xmath34 and @xmath35 represent the contributions , up to normalization factors , that one would obtain if only one of the two sources would emit particles .
on the other hand , the third term @xmath36 has the typical form of an interference term .
it introduces a non trivial deviation with respect to the probability one would obtain if the detection events would be independent for both sources : the probability is not the sum of the probabilities corresponding to each source .
we emphasize that the interference phenomenon is observed at any fixed point @xmath15 .
we are not dealing with spatially extended interferences fringes , for whose observation we would need to move the detector .
the interference effects manifest at every point by the deviation of the detection probability from that corresponding to the two sources emitting independently .
the form of the third term depends on the product of the two mode distributions .
the interference phenomenon is clearly dependent on the existence of common modes : when there are no common modes the product @xmath37 is zero , becoming null the function @xmath38 and the interference term .
the interference effects found in the previous section are explained in terms of the existence of common modes .
the detector can not distinguish if the common modes belong to one or other of the particles .
in presence of alternatives that can not be distinguished quantum mechanics leads to interference effects .
this is a novel manifestation of the extended dirac s principle .
the interference effects can not be associated with one or other of the particles or with the interaction between them , but with the existence of common modes , which is not a property of the particles composing the system , but of the complete system .
we face a self - interference effect of the complete system .
we note that the example proposed in this paper can not be interpreted as a new manifestation of the spin - statistics connection .
although this connection has been used in the commutation relations it is clear that , in absence of common modes , the connection does not produce interference effects .
we also remark that in our example the state is a non - entangled one , as in ref .
the fundamental difference between both arrangements is that in that case the interference is associated with indistinguishable two - photon paths , whereas in our example the cause of the interferences is the existence of common modes . | the extended dirac s principle describes the interference between different particles as an effect of the multiparticle system with itself . in this paper
we present a novel example , based on the detection of particles emitted in multimode states by independent sources , which illustrates in a simple way the necessity of extending the original dirac s criterion . |
definite evidences from recent experiments @xcite have confirmed that the neutrinos are massive and mixed .
part of the intrinsic neutrino properties , such as the neutrino mass hierarchy and the 1 - 3 mixing angle , however , remain undetermined .
the need for a better knowledge of the neutrino physics has motivated plenty of phenomenological analyses on whether and how the unknown neutrino properties can be reliably probed with the next generation experiments involving neutrinos from both the astrophysical and the terrestrial sources @xcite .
the core - collapse supernovas represent a unique type of neutrino source in that they emit neutrinos of all three flavors with a characteristic energy range in a time scale distinct from that of the neutrinos emitted from the sun , the atmosphere , and the terrestrial sources .
the rich physical content involved in the environment further makes the supernova an important platform for the study of neutrino properties .
in particular , the neutrino burst from a supernova has long been considered as a promising tool for probing the undetermined neutrino parameters , despite the fact that the complexity arising from the unavoidable astrophysical uncertainties may result in ambiguity in the interpretation of the observed events .
the standard msw formulation @xcite suggests that as the supernova neutrinos propagate outward , the variation of density profile can have an impact on the resonant flavor conversion .
one of the difficulties in probing the unknown neutrino properties with the supernova neutrinos arises from lack of knowledge to the matter density profile in a supernova . the inverse power - law density , @xmath2 , of the progenitor star is usually adopted in the literature as the density distribution through which the supernova neutrinos propagate outwards .
however , whether the global density distribution @xmath3 with a fixed power can provide a justifiable connection between the dynamics of resonant flavor conversion and the expected neutrino events at the detectors remains uncertain .
in addition , one may easily question the validity of this over - simplified model if , among others , the time - dependent matter distribution driven by the shock waves and the associated effects ( see , @xmath4 , ref .
@xcite ) are considered .
it is shown @xcite that near the msw resonance , even a relatively small deviation of the power from @xmath5 could have a sizable impact on the interpretation of the neutrino parameters .
furthermore , no evidence suggests that the density distributions near both the higher resonance and the lower resonance should be described by a similar density profile .
another major source of astrophysical uncertainty in analyzing the supernova neutrinos would be the spectral parameters .
distinct predictions to the average energy and the luminosity of neutrinos exist among models .
thus , the possible impact due to the uncertain spectral parameters also needs to be addressed in an analysis .
independent and variable power - law functions are assumed in this work as the matter profiles near the locations of msw resonance .
in addition , physical observables derived from the expected neutrino events at a water cherenkov detector are proposed .
it is the aim of this work to analyze the outcomes due to the uncertain density profile and the choice of spectral models , and to investigate whether and how the experimental observables would provide hints for the determination of the unknown neutrino parameters .
the analysis is also extended to include the earth matter effect .
this work is organized as follows .
section ii summarizes the general features of the neutrino fluxes emitted by a core - collapse supernova and that of the relevant detection processes at a water cherenkov detector . in section iii , the variable power - law function is proposed to account for the possible uncertainty in the matter profile near the msw resonance . in addition
, observables are constructed and examined .
focus has been placed on whether and how these observables would resolve the various scenarios arising from the uncertainty in the matter profile , the possible neutrino mass hierarchies , and the undetermined mixing angle @xmath6 . in section iv
, the general formulation is expanded to incorporate the regeneration effect due to the earth matter . in section
v , the observables proposed in section iii are re - analyzed with the earth matter effect included .
section vi is devoted to analyzing the impact due to the choice of spectral models .
we then summarized this work in section vi .
a core - collapse supernova emits neutrinos of all three flavors on a typical time scale of roughly ten seconds . within a factor of two ,
the total luminosity is approximately equi - partitioned into each flavor .
the mean energy of each neutrino flavor is determined by its interacting strength with matter .
it is believed that the neutrino spectra are not exactly thermal ; they are usually modeled by the pinched fermi - dirac distributions .
the flux of flavor @xmath7 reaching the earth surface can be written as @xmath8}+1},\ ] ] where @xmath9 is the luminosity of the neutrino flavor @xmath7 , @xmath10 is the distance to a supernova , @xmath11 is the effective temperature of @xmath7 inside the respective neutrinosphere , @xmath12 is the energy of @xmath7 , and @xmath13 is the pinching parameter .
the normalization factor @xmath14 is given by @xmath15 with the polylogarithm function @xmath16 defined as @xmath17 note that the temperature @xmath11 is related to the mean energy @xmath18 by @xmath19 as an illustration , the spectral parameters published by the lawrence livermore ( ll ) group @xcite will be adopted in the analysis : @xmath20 with @xmath21 mev , and @xmath22 = 2.0 , @xmath23 , where @xmath24 , @xmath25 , @xmath26 , and @xmath27 .
the pinching parameters @xmath28 are assumed . as for the neutrino parameters , we choose @xmath29 ev@xmath30 , @xmath31 ev@xmath30 , and @xmath32 .
note that the ongoing work by the garching group @xcite predicts distinct neutrino fluxes and average energies from that of the ll group . the comparison will be performed and discussed in section vi . in a water cherenkov detector , such as the super - kamiokonde ( see , @xmath4 , @xcite ) , the upgraded hyper - kamiokonde @xcite , and other facilities that are under planning or construction @xcite , the event rate induced by the isotropical inverse @xmath33-decay process @xmath34 dominates over that of other channels .
in this analysis , the isotropical cc processes : @xmath35 @xmath36 and the highly directional events induced by @xmath37 ( @xmath38 ) and @xmath39 scattering will also be included .
the neutrinos would likely go through multiple flavor transitions in a supernova at distinct density scales . a single power - law function , @xmath2 or @xmath3 in general , is usually adopted in the literature as the density profile of the supernova matter .
when the neutrinos travel through a complex environment such as the matter of a supernova , however , it would seem more plausible in an analysis to allow independent and variable local density profiles near the resonant locations .
if the matter distribution is modeled as layers of matter along the path of neutrinos in a specific direction , a simple power - law function can be adopted as the local profile within a layer : @xmath40 where @xmath41 stands for the @xmath41-th layer , @xmath42 describes the matter profile within the @xmath41-th layer , and @xmath43 denotes the magnitude of the component @xmath42 . the density profile near the location of resonance
can then be written as @xmath44 where the @xmath45-th layer stands for the resonance layer of the flavor conversion .
note that in this approach the power can vary from layer to layer to account for the uncertainty of profile originating from various types of sources .
the thickness of the layers , which do not take part in the formulation of this analysis , can be uneven and unrestricted , with the assumption that a specific power @xmath46 properly describes the density profile within a specific layer .
furthermore , the variable power and thickness also allow the deviation of the matter density from spherical symmetry and the inhomogeneity of matter distribution along different directions .
one may also employ functions of more than one polynomial term .
the extra polynomial terms would provide more details about the density profile in a global scale .
however , the trade - off is that they also introduce more undetermined coefficients for the extra terms .
since one only concerns the local density profile near the msw layer , the detailed shape and variation of the density in a global scale would be unimportant , and the extra terms with their undetermined coefficients would become redundant in the analysis .
thus , the power - law function with an adjustable @xmath46 would be suitable and much simpler for the analysis on how the variation of local density profile , rather that the global profile , would impact the msw resonance . in the following ,
one may simply focus on the density shapes of matter in the resonance layers : @xmath47 for the lower resonance , and @xmath48 for the higher resonance .
the variation of @xmath49 and @xmath50 would be expected to alter the efficiency of neutrino flavor transition . along this approach
, the crossing probabilities at the resonance can be modified as @xmath51 where @xmath52 or @xmath53 , @xmath54 is the mixing angle , @xmath55 is the correction factor to a non - linear profile , and the adiabaticity parameter @xmath56 takes the form @xmath57 where @xmath58 , @xmath59 is the fermi constant , @xmath60 is the neutrino energy , @xmath61 is the electron number per baryon , @xmath62 is the baryon mass , and the scale @xmath43 varies very weakly with @xmath10 over the range @xmath63 .
note that @xmath64 , and that @xmath65 can be obtained directly from @xmath66 by replacing @xmath54 with @xmath67 .
the probabilities that @xmath0 and @xmath68 survive the flavor conversion are given respectively by @xmath69 @xmath70 for the normal hierarchy , and @xmath71 @xmath72 for the inverted hierarchy . eqs .
( [ eq : pc ] ) and ( [ eq : gamma ] ) imply that the variation of neutrino energy plays an insignificant role @xcite , as compared to the variations of @xmath73 and the 1 - 3 mixing angle , in affecting the survival probabilities .
in addition , the two extreme values of @xmath74 : @xmath75 ( non - adiabatic ) and @xmath76 ( adiabatic ) , are given by the conditions : @xmath75 if @xmath77 , and @xmath76 if @xmath78 , where the function @xmath79 relates @xmath50 and @xmath6 : @xmath80 furthermore , eqs .
( [ eq : pc ] ) and ( [ eq : gamma ] ) also suggest that the arbitrary @xmath49 gives rise to only two distinct values of @xmath66 ( @xmath65 ) : @xmath81 ( @xmath82 ) for @xmath83 , and @xmath84 ( @xmath85 ) for @xmath86 .
as summarized in table 1 , the variation of @xmath73 ( @xmath87 ) , the undetermined @xmath6 , and the possible mass hierarchies would lead to six distinct scenarios in terms of the survival probabilities . in the first row of table 1 ,
the quantities in square brackets denote the variables , @xmath88 , @xmath66 and @xmath65 are determined by the variable @xmath49 ; @xmath74 is determined by @xmath50 and @xmath79 ; and the total survival probabilities , @xmath89 and @xmath90 , are determined by all @xmath49 , @xmath50 , @xmath79 , and the mass hierarchy .
the last column labels the six different scenarios in terms of the possible combinations formed by @xmath89 and @xmath90 .
note that the variation of @xmath73 would impact the probabilities differently according to whether @xmath91 or @xmath92 .
furthermore , the outcomes due to @xmath2 , which is usually assumed in the literature , correspond to those due to ( @xmath83 , @xmath93 ) in table 1 .
some qualitative observations are outlined as follows .
( i ) no information on the mass hierarchy or on @xmath6 is available if either scenario i or scenario iv is observed .
( ii ) if either ii or iii is observed , then the lower bound on @xmath6 would be available .
( iii ) likewise , if either v or vi is observed , then the lower bound on @xmath6 would be available .
the neutrino events may be categorized into isotropical ( eqs .
( [ eq : beta ] ) , ( [ eq : f ] ) , ( [ eq : n ] ) ) and directional ( neutrino scattering ) events at the water cherenkov detector .
the expected time - integrated number spectra for both the isotropical and the directional events are shown in figs .
1(a ) and 1(b ) , respectively .
it can be seen from fig .
1(a ) that three groups of nearly degenerate spectra would form for the isotropical events : ( i , v ) , ( ii , iv ) , and ( iii , vi ) .
thus , the isotropical events alone fail to provide all the needed clues for separating the six scenarios . on the other hand , while the directional event rates shown in fig .
1(b ) amount to only a tiny fraction of the isotropical ones , they could provide the rest of the clues for lifting the above degeneracy . as a result , combining both the isotropical and the directional events is inevitable in the construction of feasible observables .
the @xmath0 and @xmath68 fluxes arriving at the surface of earth , @xmath94 and @xmath95 , are related to the original ones , @xmath96 and @xmath97 , through the survival probabilities @xmath89 and @xmath90 for @xmath0 and @xmath68 , respectively : @xmath98 @xmath99 the expected total event number @xmath100 induced by the flavor @xmath7 at the detector is given by @xmath101 de,\ ] ] where @xmath102 is the number of targets at the detector , @xmath10 is the distance to the supernova , and @xmath103 is the cross section for @xmath7 in a particular reaction channel @xmath104 .
the detection efficiency is assumed to be one .
note that the regeneration effect due to the earth matter will be expected to alter the survival probabilities , and thus the observed event rates .
we shall first concentrate on building the observables without considering the earth regeneration effect . a more detailed analysis incorporating
the earth effect will follow .
statistically , the predominant isotropical channels can provide the most satisfying events for the analysis .
1(a ) suggests that one may obtain the optimized results by examining @xmath105 , which is defined as the ratio of the event rates in the medium and low energy range ( @xmath106 mev ) to that in the high energy tail ( @xmath107 mev ) : @xmath108 this ratio can be reduced to @xmath109 this expression leads to the following possible values for @xmath110 , where @xmath111 labels the six scenarios : ( i ) @xmath112 . (
ii ) @xmath113 , ( iii ) @xmath114 .
it is seen that the values of @xmath105 for the group ( iii , vi ) are smaller than the others by a factor of @xmath115 and can be easily singled out . from a different point of view , one notes that the inverted mass hierarchy implies the following possible results : ( i ) @xmath116 , which predicts @xmath86 , @xmath117 , while the information of @xmath6 is unavailable .
( ii ) @xmath118 , which predicts @xmath83 , @xmath93 , @xmath119 , and thus the upper bound @xmath120 .
( iii ) @xmath121 , which predicts @xmath83 , @xmath122 , and thus the lower bound @xmath120 . on the other hand ,
the following would be possible if the mass hierarchy is normal : ( i ) @xmath118 , which predicts @xmath83 , while @xmath50 and @xmath6 are undetermined .
( ii ) @xmath121 , which only predicts @xmath86 . to establish the @xmath89-sensitive observables that resolve the degenerate scenarios in the same group
, we now examine the @xmath0-dominating directional events . to optimize the results ,
one may define the ratio of the directional events for @xmath123 mev to that for @xmath124 mev as @xmath125 : @xmath126 this ratio can be reduced to @xmath127 which leads to the following : @xmath128 , @xmath129 , @xmath130 , @xmath131 , @xmath132 , and @xmath133 .
it is seen that the six scenarios lead to distinct observable results in terms of the combination ( @xmath105 , @xmath125 ) .
the observable @xmath105 would reduce the six scenarios into three degenerate groups : ( i , v ) , ( ii , iv ) , and ( iii , vi ) . observing @xmath125 would further remove the degeneracy between the scenarios in each group .
thus , each scenario can be characterized by a unique set of observables @xmath134 , which should act as a feasible reference in removing the degeneracy among the six scenarios . in searching for useful observables other than @xmath105 and @xmath125 ,
one notes from fig .
1 that in terms of the height and the width , the possible spectra exhibit the most pronounced differences in the region near the peaks . the ratio formed by the peak value of a specific spectrum ( @xmath135 ) and the width at half peak value ( @xmath136 ) deserves some attention .
one may introduce the double ratio involving both the isotropical and the directional events : @xmath137 the six scenarios lead to the following possible values : @xmath138 , @xmath139 , @xmath140 , @xmath141 , @xmath142 , and @xmath143 furthermore , one may define another observable @xmath144 involving both the isotropical the directional events in the medium energy range , @xmath145 : @xmath146 which leads to @xmath147 and @xmath148 , @xmath149 , @xmath150 , @xmath151 , @xmath152 , and @xmath153 if the neutrino energy spectrum is reconstructed , one may examine yet another observable @xmath154 , which involves the average energies of both the isotropical and the directional spectra : @xmath155 . the combined average energy @xmath156 is defined as @xmath157 } { \sum_{\alpha } n_{\alpha i}},\ ] ] where @xmath104 stands for isotropical or directional , @xmath158 is the time - integrated number spectrum , @xmath159 is the event number , and the summation over @xmath160 includes all the possible isotropical or directional channels induced by @xmath7
. further reduction of @xmath154 yields @xmath161 and the values of @xmath154 for the six scenarios are given by : @xmath162 , @xmath163 , @xmath164 , @xmath165 , @xmath166 , and @xmath167 .
the uncertainty arising from the density profile can leave observable signatures on the neutrino events through the resonant flavor transition . for a given set of input spectral parameters
, the proposed experimental observables may be combined to form useful references for removing the degeneracy among scenarios .
however , it should be noted that whether the unknown neutrino parameters could be definitely determined remains as an open question as other possible sources of uncertainty may blur or even wash out the signatures .
we shall discuss this issue in sections vi and vii .
the neutrinos emitted from a supernova arrive at the earth surface in mass eigenstates .
since the matter density of earth is approximately the same order of magnitude as that near the lower resonance in a typical supernova , the neutrinos could oscillate again if they cross earth prior to the detection .
this regeneration effect in the earth matter would be expected to alter the neutrino fluxes arriving at the detector .
the observable effects due to this regeneration process have been intensively studied @xcite .
it is realized that in general the earth effects are signaled by an oscillatory modulation on the neutrino spectrum that is expected when the regeneration effect is missing . given the uncertainty in matter density profile of a supernova , we intend here to further examine the feasibility of the proposed observables under the influences of the earth regeneration effect .
the regeneration effect depends on several factors : the neutrino parameters , the neutrino energy , the density shape of earth matter , and the location of the detectors ( or , the incident direction of the neutrinos ) .
the @xmath0 and @xmath68 fluxes arriving at a detector @xmath168 can be expressed as @xmath169 and @xmath170 , respectively : @xmath171 + f^{0}_{x}[1-p_{h}(p_{l}+p_{2e}(1 - 2p_{l}))],\ ] ] @xmath172 + f^{0}_{x}[\bar{p}_{l}+\bar{p}_{2e}(1 - 2\bar{p}_{l})]\ ] ] for the normal hierarchy , and @xmath173 + f^{0}_{x}[1-(p_{l}+p_{2e}(1 - 2p_{l}))],\ ] ] @xmath174\ ] ] for the inverted hierarchy , where @xmath175 ( @xmath176 ) represents the probability that a @xmath177 ( @xmath178 ) arriving at the earth surface is detected as a @xmath0 ( @xmath68 ) at the detector .
the earth matter density encountered by the electron neutrinos is given by @xmath179 , where @xmath59 is the fermi constant and @xmath180 , with @xmath181 the density profile , @xmath182 the proton mass , and @xmath62 the neutron mass .
the probability @xmath175 is given by @xcite : @xmath183,\ ] ] where @xmath184 is the entry point , @xmath185 is the exit point , and @xmath186 is the distance into the earth .
the function @xmath187 is defined as @xmath188 where @xmath189^{2}+\sin^{2}2\theta_{12}},\ ] ] and the parameter @xmath190 is given by @xmath191 . note that @xmath192 and @xmath193 the density profile of the earth matter can be approximated by a simple step function @xcite : @xmath194 @xmath195 for @xmath196 ( mantle ) , and @xmath197 @xmath195 for @xmath198 ( core ) , where @xmath199 is the earth radius .
for a specific detector , the incident angle of the neutrino determines the path length through earth and the magnitude of @xmath175 .
because of the chosen density model , it is necessary to calculate @xmath175 separately for both @xmath200 ( mantle only ) and @xmath201 ( mantle+core+mantle ) , where @xmath202 is the incident angle in general , and @xmath203 denotes the incident angle when the path is tangent to the sphere of radius @xmath204 , at which radius the density changes . for @xmath200 , as shown in fig .
2(a ) , the probability @xmath175 takes the form @xmath205,\ ] ] where the total path length @xmath206 inside the earth is given by @xmath207 with @xmath208 the depth of the detector . for @xmath201 ,
the path inside the earth consists of three segments : @xmath209 , @xmath210 , and @xmath211 , as shown in fig .
2(b ) : @xmath212},\ ] ] @xmath213},\ ] ] @xmath214}.\ ] ] the probability @xmath175 in this case can be written as @xmath215 where @xmath216 + \epsilon^{c}(x)[1-\cos(\delta^{c } x_{bc})]+ \epsilon^{m}(x)[1-\cos(\delta^{m } x_{cd})],\ ] ] and the superscripts @xmath217 and @xmath218 stand for mantle and core , respectively .
it is clear from the oscillatory terms that a longer path length in matter would give rise to a higher frequency of @xmath175 .
note that for the probability @xmath176 , the parameter @xmath219 is replaced by @xmath220 : @xmath221
for one of the six scenarios @xmath111 , the expected neutrino flux of flavor @xmath160 ( @xmath222 ) at the surface of earth , @xmath223 , and at the detector , @xmath224 , are determined by the type of mass hierarchy , the values of @xmath74 , @xmath66 ( @xmath65 ) , and the original fluxes @xmath225 for a given model .
we list the expressions of @xmath223 and @xmath224 in appendix a. the effective survival probabilities for @xmath0 and @xmath68 at the detector , @xmath226 and @xmath227 , can be read off from @xmath228 and @xmath229 , respectively : @xmath230 and @xmath231 in principle , the flux variation due to the earth matter effect can be conveniently illustrated by the ratios @xmath232 and @xmath233 , for @xmath0 and @xmath68 , respectively .
the ratios @xmath234 and @xmath235 , as given in appendix b , suggest the following : \(i ) the values of @xmath236 , @xmath237 , @xmath238 , and @xmath239 vanish , @xmath88 , including the earth matter effect would not alter the resultant @xmath0 flux for ii and v , and the resultant @xmath68 flux for iii and vi .
this can be understood by the fact that the crossing probability at the higher resonance approaches zero ( @xmath240 ) for each of the above scenarios , as indicated by table 1 .
if the neutrinos travel through the earth matter before reaching the detector , table 1 and eq.(46 ) imply that @xmath241 for scenarios ii and v. likewise , table 1 and eq.(47 ) imply that @xmath242 for scenarios iii and vi .
\(ii ) the two ratios , @xmath243 and @xmath244 , are identical because of @xmath245 .
in addition , one may also conclude that @xmath246 , @xmath247 , and @xmath248 .
\(iii ) the scenarios in each degenerate pair , ( ii , iv ) or ( i , vi ) , will not be distinguished from one another by the observation of @xmath68 events alone even if the earth matter effect is included .
thus , constructing the observables that are sensitive to the @xmath0 events would be beneficial to the analysis .
likewise , the @xmath68-sensitive observables would be needed to lift the degeneracy between iii and iv , as well as that between i and iv . in summary , the values of @xmath249 and @xmath250 may act as a useful qualitative and quantitative reference in analyzing the distinct scenarios .
for example , scenario ii and scenario iv can be distinguished by examining whether the earth matter effect is missing ( scenario ii ) or observed ( scenario iv ) in the @xmath0 flux . likewise , scenario iii and scenario iv can be separated by examining whether the earth matter effect is observed ( scenario iv ) or not ( scenario iii ) in the @xmath68 flux .
similar qualitative observations can be applied to separating i and v , or i and vi .
we now reexamine the expected ratio @xmath251 at a detector when the earth matter effect is included .
the impact on @xmath252 from the earth effect can be examined through the plot @xmath253 in fig .
as expected , scenarios iii and vi can be easily singled out by their low values , @xmath254 .
note that the values of @xmath252 for iii and vi are nearly independent of the incident angle of neutrinos .
in addition , a jump of @xmath252 occurs at @xmath255 for all the scenarios .
this is realized by the fact that the length of neutrino path in mantle begins to increase significantly when @xmath256 .
it is also seen that an even more significant jump occurs at @xmath257 , beyond which angle the neutrino flux starts to pass through the earth matter that is formed by the combination mantle+core+mantle before reaching the detector .
note that the values of @xmath105 for the four scenarios : i , ii , iv , and v , become almost indistinguishable when @xmath258 .
as discussed in section iii , the neutrino events from the @xmath89-sensitive channels are needed to remove the degeneracy that the observable @xmath105 fails to resolve .
thus , the @xmath89-sensitive ratios @xmath259 as functions of @xmath202 for all the six scenarios are reexamined and shown in fig .
4 . to explain the obvious step - up behavior of iii and iv , and the step - down behavior of i , ii , v , and vi in fig . 4 ,
one first notes from eq .
( [ eq : pec ] ) that @xmath175 increases with @xmath260 .
while the path length determines the frequency of the oscillatory term , the mean value of @xmath175 is determined by the sum @xmath261 for @xmath262 , and by @xmath263 for @xmath264 . since the parameter @xmath265 is proportional to the density , so that @xmath175 increases with the density . likewise , eq . ( [ eq : p2ea ] ) implies that @xmath176 decreases when the density increases .
it then follows from eq .
( [ eq : effp ] ) that @xmath266 and @xmath267 decrease when the density increases @xmath268 , while @xmath269 and @xmath270 increase when the density increases . on the other hand , eq .
( [ eq : aeffp ] ) implies that all @xmath271 , @xmath272 , @xmath273 , and @xmath274 increase when the density increases .
given the conditions that @xmath275 , @xmath276 , @xmath277 , and @xmath278 are vanishing , and that the coefficients of the @xmath89-term dominate over that of the @xmath90-term in both the denominator and the numerator of eq .
( [ r2 ] ) , the stepping down behavior of @xmath279 , @xmath280 , @xmath281 , and @xmath282 , as well as the stepping up of @xmath283 and @xmath284 at greater @xmath202 become apparent .
it should be pointed out , as mentioned earlier , that the earth matter effect is manifested by an oscillatory modulation on top of the neutrino spectra that would have been observed without the regeneration effect .
practically , it would be very difficult , if not impossible , to extract from @xmath285 the information that is significantly different from that obtained by @xmath286 .
the ratio of the isotropical events to the directional events , @xmath287 , and the ratio formed by the average energy of the isotropical to that of the directional events , @xmath288 , are shown in fig . 5 and fig . 6 , respectively .
it is seen that the six scenarios are in principle distinguishable by examining @xmath252 , @xmath289 , @xmath290 and @xmath291 .
however , as can be seen from the figures , the distinct scenarios become less distinguishable at a larger incident angle , @xmath262 .
given the potential theoretical and experimental uncertainties , ambiguity may arise in interpreting the proposed observables if the earth matter becomes pronounced .
we have shown how the original neutrino fluxes would be modified by the msw effect in a supernova even if the details of the density gradient of matter are unavailable .
the input spectral parameters are adopted from the ll model . as compared to the ll model , the parameters predicted by the ongoing work of the garching group @xcite , labeled as g1 and g2 , suggest smaller differences in both the flux and in the average energy among neutrinos of different flavors : g1 : @xmath292 mev , @xmath21 mev , @xmath293 mev , @xmath294 , g2 : @xmath292 mev , @xmath21 mev , @xmath295 mev , @xmath296 . to examine the impact due to the choice of models , the expected neutrino spectra for the six possible scenarios using the parameters of g1 and g2
are shown in figs . 7 and 8 , respectively .
note that the spectral shapes predicted by the ll model in fig .
1 exhibit a very distinct nature from that predicted by g1 and g2 .
it is seen that the peak locations of the isopropical events for iii and vi mark the most significant difference between g1 and g2 .
the peak locations for other scenarios are relatively insensitive to the choice of g1 or g2 . as for the directional events , the six scenarios are barely distinguishable under the g1 model . in general
, the g2 model predicts directional spectra that are sharper than that predicted by the g1 model .
the inherent complexity that leads to the entanglement in the spectral shapes , the peak heights , and the peak locations among scenarios and models may render a thorough , quantitative analysis very difficult .
however , as can be seen from figs .
7 and 8 , the most significant difference among scenarios for a given model appears in the spectra near the peaks .
thus , the observed @xmath252 and @xmath289 might be able to reflect the distinct characters of each model .
we show the plots of @xmath253 for g1 and g2 in fig .
it is seen that g1 gives rise to distinct values of @xmath297 ranging from @xmath298 to @xmath299 . as for g2 ,
the values of @xmath297 are confined within @xmath300 for @xmath301 , and @xmath302 for @xmath303 . fig .
9 suggests that if the spectral parameters of g1 are adopted , the six different scenarios may still cause @xmath252 to vary by roughly a factor of two . on the other hand for g2
, the ratio @xmath297 appears to be insensitive to the uncertainties represented by the distinct scenarios at small incident angle . as a comparison
, one recalls from fig .
3 that the possible values of @xmath297 for the ll model varies from @xmath304 to @xmath305 , which is in a region very different from that for g1 and g2 .
we conclude that the observable @xmath252 would reflect the distinct characters of the three different model .
the ratio @xmath306 can be analyzed in a similar way .
we show the plots of @xmath306 for g1 and g2 in fig .
it is clear that the values of @xmath289 predicted by the three different models fall in distinct regions : @xmath307 for ll , @xmath308 for g1 , and @xmath309 for g2 .
in addition , it should be pointed out that we also perform calculations on the other ratios , @xmath285 , @xmath290 , and @xmath291 , using the input parameters predicted by g1 and g2 . however , unlike @xmath252 and @xmath289 , the results show that the expected values for each of these ratios would all tangled up unless either a particular model is specified or the uncertainty due to the density profile can be significantly reduced . in summary ,
the choice of models would lead to distinct signatures in the values of @xmath297 and @xmath306 .
the degeneracy among the six scenarios for either ll or g1 could be removed . on the other hand ,
the very nature of g2 would lead to only very little differences among the scenarios , @xmath88 , it would be difficult for them to be quantitatively distinguished .
however , if the correct supernova parameters are the ones predicted by the g2 model , the analysis based on the g2 model could be relatively free from the disturbing uncertainties , such as that originating from the density profile , the mass hierarchy , and the mixing angle @xmath6 .
while a satisfactory knowledge of the supernova mechanism is still far from complete , the neutrino flux emitted from the core - collapse supernova could be a promising tool for probing the unknown neutrino properties .
it is the intention of this work to search for possible hints of the unknown neutrino properties using the experimental observables derived from the supernova neutrino signals .
given the uncertain density profile and the choice of input spectral models , the aim is set to exploring the unknown neutrino parameters when the neutrino fluxes are modified by the msw resonant conversion in both the supernova and in crossing earth . although the msw mechanism has been well established in the literature , this analysis shows that the phenomenological outcomes due to the possible variation of density profile still provide certain new physical insight to the nature of the unknown neutrino properties . in stead of following the traditional approach based on the global density profile @xmath2 , or the profile with a fixed power in general : @xmath3
, this analysis allows the density profiles near the locations of resonance to vary independently as arbitrary power - law functions : @xmath47 for the lower resonance , and @xmath48 for the higher resonance .
it is shown that the variation of @xmath73 ( @xmath310 or @xmath311 ) would cause different degrees of impact on the survival probability for @xmath0 ( @xmath68 ) according to whether @xmath91 or @xmath92 .
as far as the survival probabilities are concerned , there exists six possible scenarios , which are characterized by possible combinations of the density profile , the mass hierarchy , and the mixing angle @xmath6 .
we propose observable quantities that are derived from the expected supernova neutrino events at the water cherenkov detector , and explore the feasibility of probing the neutrino parameters with these observables . as an illustration ,
the set of spectral parameters predicted by the lawrence livermore group ( ll ) is adopted . to investigate how the choice of supernova models would affect the outcomes ,
the analysis based on the input parameters predicted by the graching model is also presented as a comparison .
given the unknown density profile of the supernova and the choice of input spectral parameters , we show how the unknown mass hierarchy and the mixing angle @xmath6 would leave observable signatures on the neutrino events . the difficulties and limits for this type of analysis
are also discussed .
it is shown that the two observables , @xmath105 and @xmath125 , can combine to act as the most effective pair of observables for probing the neutrino parameters .
other observables , @xmath286 , @xmath144 , and @xmath154 , may be considered as the supplementary references . a more realistic analysis incorporating the earth regeneration effect is also included .
3 , 4 , 5 , and 6 show how each of the observables varies with the incident angle of neutrinos when the earth matter induces additional flavor conversion prior to the detection of neutrinos .
the general formulation can be further expanded to accommodate different models of the earth density profile and can be applied to analyzing the effects if the neutrinos are registered by other types of terrestrial detectors .
it should be pointed out that certain non - msw effects may occur in the complex environment of a supernova at different space and time scales .
these effects may be independent of the density profile and may be totally irrelevant to the msw conversion . if the proper conditions are met , recent simulations suggest that effects such as the neutrino self coupling in dense media ( see , @xmath4 , ref . @xcite and the references therein ) and the neutrino flavor de - polarization associated with the after shock turbulence ( see , @xmath4 , ref .
@xcite and the references therein ) could occur .
these effects are expected to impact the efficiency of neutrino flavor transition .
it is suggested that both the shock reheating and the hot bubble epochs could provide the proper conditions for two classes of large - scale collective effects : synchronization and bipolar flavor transformation , which arose if neutrinos themselves form significant background .
the neutrino flavor transformation due to this neutrino coupling effect shares no common origin with that due to the msw resonance .
in addition , a fluctuating density in the post - shock region could be created by the turbulent convective motions behind the forward moving shock and the reverse shock .
such fluctuations could erase part of the shock wave imprint on the neutrino spectra even with relatively small amplitudes .
it is suggested that if the amplitude of density fluctuations in the turbulence is large , the average neutrino survival probability would saturate to 1/2 .
the information about the initial state is then lost , and the final state is depolarized .
this depolarization could leave unique signatures that are distinct from those left by the shock and other density effects . as concluded by this work ,
the variation of density profile , which plays a determinant role in the msw flavor transition , could modify the spectra significantly and lead to variation of the proposed observables by a factor of two or more . whether the flavor transition effect arising from one origin would dominate over that from the others depends on the spectral parameters , and on the neutrino intrinsic properties such as the mass hierarchy and the magnitude of @xmath6 .
in addition , it also depends on the physical conditions that the neutrinos would encounter at different stages of the supernova environment and in crossing the earth matter .
given the undetermined neutrino intrinsic properties and the existing astrophysical uncertainties , it would be intriguing to further investigate , under all possible and relevant physical conditions , how the non - msw effects would act as a modification to the msw effect or even dominate over the msw effect . the more detailed analyses with reduced assumptions would certainly help suggest how some of the existing msw - based paradigms for the neutrino flavor conversion in a supernova should be reconstructed . this work is supported in part by the national science council of taiwan .
the neutrino flux of flavor @xmath160 ( @xmath312 ) for the six scenarios at the surface of earth , @xmath223 , and at the detector , @xmath224 , are related to the original fluxes @xmath225 by the following : @xmath313,\ ] ] @xmath314 + f^{0}_{x}[1-(\cos^{2}\theta_{12}+p_{2e}(1 - 2\cos^{2}_{12}))],\ ] ] @xmath315,\ ] ] @xmath316 + f^{0}_{x}[\sin^{2}\theta_{12}+\bar{p}_{2e}(1 - 2\sin^{2}\theta_{12})],\ ] ] @xmath317 @xmath318 @xmath319 @xmath320 @xmath321 @xmath322 @xmath323 @xmath324 @xmath325 @xmath326 @xmath327 @xmath328,\ ] ] @xmath329+f^{0}_{x}[\sin^{2}\theta_{12}+ \bar{p}_{2e}(1 - 2\sin^{2}\theta_{12})],\ ] ] @xmath330,\ ] ] @xmath331 + f^{0}_{x}[\sin^{2}\theta_{12}-p_{2e}(1 - 2\cos^{2}\theta_{12})],\ ] ] @xmath332
expressions for @xmath234 and @xmath235 are given as the following : .types of scenarios that are categorized by various combinations of @xmath346 .
each of the quantity in the square bracket in the first row represents the variable that leads to the probability to its right .
the total survival probabilities , @xmath89 and @xmath90 , are determined by the specific @xmath49 , @xmath50 , @xmath6 , and the mass hierarchy . [ cols="^,^,^,^,^,^,^,^,^,^",options="header " , ] | the flavor conversion of supernova neutrinos can lead to observable signatures related to the unknown neutrino parameters .
as one of the determinants in dictating the efficiency of resonant flavor conversion , the local density profile near the msw resonance in a supernova environment is , however , not so well understood . in this analysis , variable power - law functions are adopted to represent the independent local density profiles near the locations of resonance .
it is shown that the uncertain matter density profile in a supernova , the possible neutrino mass hierarchies , and the undetermined 1 - 3 mixing angle would result in six distinct scenarios in terms of the survival probabilities of @xmath0 and @xmath1 .
the feasibility of probing the undetermined neutrino mass hierarchy and the 1 - 3 mixing angle with the supernova neutrinos is then examined using several proposed experimental observables .
given the incomplete knowledge of the supernova matter profile , the analysis is further expanded to incorporate the earth matter effect . the possible impact due to the choice of models , which differ in the average energy and in the luminosity of neutrinos ,
is also addressed in the analysis . |
now that the observation of the quark - gluon matter in relativistic heavy ion collisions is established for small baryon chemical potential ( @xmath0 ) @xcite , a challenging experimental subject following this achievement is to reveal the global structure of the qcd phase diagram on the temperature ( @xmath1 ) and @xmath0 plane . in particular , finding the qcd critical point(s ) , whose existence is predicted by various theoretical studies @xcite , is one of the most intriguing problems . since the @xmath0 of the hot medium created by heavy - ion collisions can be controlled by varying the collision energy per nucleon pair , @xmath2 , the @xmath0 dependence of the nature of qcd phase transition should be observed as the @xmath2 dependence of observables .
an experimental project to explore such signals in the energy range @xmath3 , which is called the energy scan program , is now ongoing at the relativistic heavy ion collider ( rhic ) @xcite . experimental data which will be obtained in future experimental facilities designed for lower beam - energy collisions will also provide important information on this subject @xcite .
observables which are suitable to analyze bulk properties of the matter around the phase boundary of qcd in heavy ion collisions are fluctuations @xcite .
experimentally , fluctuations are measured through event - by - event analyses @xcite .
theoretically , it is predicted that some of them , including higher - order cumulants , are sensitive to critical behavior near the qcd critical point @xcite , and/or locations on the phase diagram , especially on which side the system is , the hadronic side or the quark - gluon side @xcite . among the fluctuation observables ,
those of conserved charges are believed to possess desirable properties to probe the phase structure in relativistic heavy ion collisions .
one of the advantages of using the conserved charges is that the characteristic times for the variation of their local densities are longer than those for non - conserved ones , because the variation of the local densities of conserved charges are achieved only through diffusion @xcite .
the fluctuations of the former thus can better reflect fluctuations generated in earlier stages of fireballs , when the rapidity coverage is taken sufficiently large . from a theoretical point of view , an important property of the conserved charges
is that one can define the operator of a conserved charge , @xmath4 , as a noether current .
moreover , their higher - order cumulants , @xmath5 , are directly related to the grand canonical partition function @xmath6 as @xmath7 with @xmath8 and @xmath9 being the hamiltonian and the chemical potential associated with @xmath4 , respectively .
these properties make the analysis of cumulants of conserved charges well defined and feasible in a given theoretical framework . for example
, they can be measured in lattice qcd monte carlo simulations @xcite .
the relation eq .
( [ eq : q ] ) also provides an intuitive interpretation for the behavior of higher - order cumulants of conserved charges .
for instance , the third - order cumulant of the net baryon number , @xmath10 , satisfies @xmath11 .
this formula means that @xmath12 changes its sign around the phase boundary on the @xmath1-@xmath0 plane where the baryon number susceptibility @xmath13 has a peak structure @xcite .
the change of the sign of observables like this will serve as a clear experimental signal @xcite .
qcd has several conserved charges , such as baryon and electric charge numbers and energy . among these conserved charges ,
theoretical studies suggest that the cumulants of the baryon number have the most sensitive dependences on the phase transitions and phases of qcd . in order to see this feature ,
let us compare the baryon number cumulants with the electric charge ones .
first , the baryon number fluctuations show the critical fluctuations associated with the qcd critical point more clearly .
although the baryon and electric charge number fluctuations diverge with the same critical exponent around the critical point , it should be remembered that this does not mean similar clarity of signals for the critical enhancement in experimental studies .
fluctuations near the critical point are generally separated into singular and regular parts , and only the former diverges with the critical exponent .
the singular part of the electric charge fluctuations is relatively suppressed compared to the baryon number ones , because the formers contain the isospin number fluctuations which are regular near the critical point @xcite .
the additional regular contribution makes the experimental confirmation of the enhancement of the singular part difficult , and this tendency is more pronounced in higher - order cumulants @xcite . while it is known that the proton number fluctuations in the final state also reflect the critical enhancement near the critical point @xcite , as we will show later the baryon number fluctuations are superior to this observable , too , in the same sense .
second , the ratios of baryon number cumulants @xcite behave more sensitively to the difference of phases , i.e. , hadrons , or quarks and gluons .
this is because the ratios are dependent on the magnitude of charges carried by the quasi - particles composing the state @xcite , while the charge difference between hadrons and quarks is more prominent in the baryon number . experimentally , however , the baryon number fluctuations are not directly observable , because chargeless baryons , such as neutrons , can not be detected by most detectors .
proton number fluctuations can be measured @xcite , and recently its cumulants have been compared with theoretical predictions for baryon number cumulants .
indeed , in the free hadron gas in equilibrium the baryon number cumulants are approximately twice the proton number ones , because the baryon number cumulants in free gas are simply given by the sum of those for all baryons , and the baryon number is dominated by proton and neutron numbers in the hadronic medium relevant to relativistic heavy ion collisions . in general , however , these cumulants behave differently .
in fact , we will see later that the non - thermal effects which exists in baryon number cumulants are strongly suppressed in the proton number ones . in heavy ion collisions , because of the dynamical evolution the medium at kinetic freezeout is not completely in the thermal equilibrium .
the original ideas to exploit fluctuation observables as probes of primordial properties of fireballs @xcite are concerned with this non - thermal effect encoded in the final state as a hysteresis of the time evolution . to observe such effects , it is highly desirable to measure baryon number cumulants that is expected to retain more effects of the phase transition and the singularity around the critical point .
the experimental determination of baryon number cumulants also makes the comparison between experimental and theoretical studies more robust , since many theoretical works including lattice qcd simulations are concerned with the baryon number cumulants , not the proton number ones . in ref .
@xcite , the authors of the present paper have argued that , whereas the baryon number cumulants are not the direct experimental observables as discussed above , they can be determined in experiments by only using the experimentally measured proton number fluctuations for @xmath14 .
the key idea is that isospins of nucleons in the final state are almost completely randomized and uncorrelated , because of reactions of nucleons with thermal pions in the hadronic stage , as will be elucidated in sec .
[ sec : b ] .
this leads to the conclusion that , when @xmath15 nucleons exist in a phase space of the final state , the probability that @xmath16 nucleons among them are protons follows the binomial distribution .
more generally , the probability distribution that @xmath16 protons , @xmath17 neutrons , @xmath18 anti - protons , and @xmath19 anti - neutrons are found in the final state in a phase space is factorized as @xmath20 where the nucleon and anti - nucleon numbers are @xmath21 and @xmath22 , respectively , and @xmath23 is the binomial distribution function with probabilities @xmath24 and @xmath25 .
the function @xmath26 describes the distribution of nucleons and anti - nucleons and the correlation between them in the final state , which are determined by the dynamical history of fireballs .
using the factorization eq .
( [ eq : p ] ) , one can obtain formulas to represent the ( anti-)nucleon number cumulants by the ( anti-)proton number ones , and vice versa ; whereas the neutron number is not determined by experiments , this missing information can be reconstructed with the knowledge for the distribution function , eq .
( [ eq : p ] ) .
the ( anti-)nucleon number in eq .
( [ eq : p ] ) can further be promoted to the ( anti-)baryon number in practical analyses with a simple treatment for strange baryons to a good approximation .
these formulas enable to determine the baryon number cumulants solely with the experimentally measured proton number fluctuations , and , as a result , to obtain insights into the present experimental results on the proton number cumulants .
the main purpose of the present paper is to elaborate the discussion in ref .
@xcite with some extensions . in ref .
@xcite the formulas are derived only for isospin symmetric medium . in the present study
we extend them to incorporate cases with nonzero isospin densities . with the extended relations
, it is shown that the effect of nonzero isospin density is well suppressed for @xmath14 .
the procedures of the manipulations and discussions omitted in ref .
@xcite are also addressed in detail . in the next section ,
we show that the factorization eq .
( [ eq : p ] ) is well applied to the nucleon and baryon distribution functions in the final state in heavy ion collisions .
we then derive formulas to relate baryon and proton number cumulants in sec . [
sec : cumulant ] . in sec .
[ sec : discussion ] , we discuss the recent experimental results at star @xcite using the results in sec .
[ sec : cumulant ] , and possible extensions of our results .
the final section is devoted to a short summary . throughout this paper
, we use @xmath27 to represent the number of particles @xmath28 leaving the system after each collision event , where @xmath29 , @xmath30 , n , and b represent proton , neutron , nucleon , and baryon , respectively , and their anti - particles , @xmath31 , @xmath32 , @xmath33 , and @xmath34 .
the net and total numbers are defined as @xmath35 and @xmath36 , respectively .
in this section , we discuss the time evolution of the proton and neutron number distributions in the hadronic medium generated by relativistic heavy ion collisions , and show that the nucleon distribution in the final state in a phase space is factorized as in eq .
( [ eq : p ] ) at sufficiently large @xmath2 . in sec .
[ sec : b : free ] , as a preliminary example we show that eq .
( [ eq : p ] ) is applicable to the equilibrated free hadron gas in the ranges of @xmath1 and @xmath0 relevant to relativistic heavy ion collisions .
we then extend the argument to the distribution function in the final state in relativistic heavy ion collisions in sec .
[ sec : b : hic ] .
let us first consider nucleons in the free hadron gas in equilibrium .
for @xmath1 and @xmath0 which are relevant to relativistic heavy ion collisions , the nucleon mass @xmath37 satisfies @xmath38 .
one thus can apply the boltzmann approximation for the distribution functions of nucleons .
the number of particles in a phase space , @xmath39 , which obey boltzmann statistics is given by the poisson distribution , @xmath40 with the average @xmath41 .
accordingly , the probability to find @xmath16 ( @xmath18 ) protons ( anti - protons ) and @xmath17 ( @xmath19 ) neutrons ( anti - neutrons ) in the phase space is given by the product of the poisson distribution functions , @xmath42 the product of two poisson distribution functions satisfies the identity , @xmath43 where @xmath44 is the binomial distribution function eq .
( [ eq : b ] ) . using eq .
( [ eq : pp = pb ] ) , eq .
( [ eq : p_hg1 ] ) is rewritten as @xmath45 where @xmath21 and @xmath46 are the nucleon and anti - nucleon numbers , respectively , and @xmath47 and @xmath48 .
equation ( [ eq : p_hg ] ) shows that the distribution of nucleons in the free hadron gas is factorized using binomial functions as in eq .
( [ eq : p ] ) with @xmath49 the appearance of the binomial distribution functions in eq .
( [ eq : p_hg ] ) is understood as follows . when one finds a nucleon in the hadron gas , the probability that the nucleon is a proton is @xmath50 .
the isospins of all nucleons found in the phase space , moreover , are not correlated with one another as a consequence of boltzmann statistics and the absence of interactions . once @xmath15 nucleons are found in the phase space , therefore ,
the probability that @xmath16 particles are protons is a superposition of independent events with probability @xmath50 , i.e. , the binomial distribution .
we note that the above discussion is not applicable when the condition @xmath38 , required for boltzmann statistics , is not satisfied .
when quantum correlations of nucleons arising from fermi statistics are not negligible , the isospin of each nucleon can no longer be independent . as long as we are concerned with the range of @xmath1 and @xmath0 which can be realized by relativistic heavy ion collisions , however , the condition for the boltzmann approximation is well satisfied except in very low energy collisions @xcite .
next , we consider the nucleon distribution functions in the final state in heavy ion collisions .
we show that the nucleon distribution in this case is also factorized as in eq .
( [ eq : p ] ) , by demonstrating that the isospins of all nucleons in the final state are random and uncorrelated .
the key ingredient to obtain the factorization eq .
( [ eq : p ] ) in the final state in relativistic heavy ion collisions is @xmath52 reactions in the hadronic stage mediated by @xmath51 resonances having the isospin @xmath53 . as we will see later ,
this is the most dominant reaction of nucleons in the hadronic medium .
this is because i ) the cross section of @xmath54 reactions exceeds @xmath55 and is comparable with @xmath56 and @xmath57 reactions for @xmath58 @xcite , and ii ) the pion density dominates over those of all other particles in the ranges of @xmath1 and @xmath0 accessible with heavy ion collisions at @xmath59 ; at the top rhic energy , the density of pions is more than one order larger than that of nucleons .
we shall show below that these reactions frequently take place even after chemical freezeout in the hadronic medium during the time evolution of the fireballs .
the @xmath52 reactions through @xmath60 contain charge exchange reactions , which alter the isospin of the nucleon in the reaction .
the reactions of a proton to form @xmath60 are : @xmath61 among these reactions , eqs .
( [ eq : d+ ] ) and ( [ eq : d0 ] ) are responsible for the change of the nucleon isospin .
the ratio of the cross sections of a proton to form @xmath62 , @xmath63 , and @xmath64 is @xmath65 , which is determined by the isospin su(2 ) symmetry of the strong interaction .
the isospin symmetry also tells us that the branching ratios of @xmath63 ( @xmath64 ) decaying into the final state having a proton and a neutron are @xmath66 ( @xmath67 ) . using these ratios , one obtains the ratio of the probabilities that a proton in the hadron gas forms @xmath63 or @xmath64 with a reaction with a thermal pion , and then decays into a proton and a neutron , respectively , @xmath68 and @xmath69 , as @xmath70 provided that the hadronic medium is isospin symmetric and that the three isospin states of the pion are equally distributed in the medium . because of the isospin symmetry of the strong interaction one also obtains the same conclusion for neutron reactions : @xmath71 similar results are also obtained for anti - nucleons .
equations ( [ eq : p_pn ] ) and ( [ eq : p_np ] ) show that these reactions act to randomize the isospin of nucleons during the hadronic stage . of a rest nucleon to form @xmath63 or @xmath64 in the hadronic medium as a function of temperature @xmath1 .
] next , let us estimate the mean time of these reactions .
assuming that pions are thermally distributed , the mean time @xmath72 of a nucleon at rest in the medium to undergo a reaction eq .
( [ eq : d+ ] ) or ( [ eq : d0 ] ) is given by @xmath73 with the bose distribution function @xmath74 , the pion momentum @xmath75 , the pion velocity @xmath76 , @xmath77 , and the pion mass @xmath78 .
@xmath79 is the sum of the cross sections for n@xmath80 reactions producing @xmath63 and @xmath64 for the center - of - mass energy @xmath81^{1/2}$ ] with the nucleon mass @xmath37 . for the cross section @xmath79
, we assume that the peak corresponding to @xmath51 resonance is well reproduced by the breit - wigner form , @xmath82 which is a sufficient approximation for our purpose . here
, we use the value of the parameters determined by the n@xmath80 reactions in the vacuum , @xmath83 , @xmath84 , and @xmath85 @xcite .
the medium effects on the cross section will be discussed later .
substituting @xmath86 and @xmath87 , one obtains the @xmath1 dependence of the mean time @xmath72 presented in fig .
[ fig : tau ] .
the figure shows that the mean time is @xmath88fm for @xmath89 .
one can confirm that the mean time hardly changes even for moving nucleons in the range of momentum @xmath90 by extending eq .
( [ eq : tau ] ) to cases with nonzero nucleon momentum .
the lifetime of @xmath60 resonances is @xmath91 .
the mean time evaluated above is much shorter than the lifetime of the hadronic stage in relativistic heavy ion collisions . according to a dynamical model analysis for collisions at rhic ,
nucleons in the hadron phase continue to interact for a couple of tens of fm on average at midrapidity @xcite . as a result , at the rhic energy
each nucleon in a fireball has chances to undergo the charge exchange reactions several times in the hadronic stage .
two remarks are in order here .
first , the above result on the time scales shows that the reactions to produce @xmath60 proceed even after chemical freezeout .
these reactions do not contradict the success of the statistical model , which describes the chemical freezeout @xcite , because chemical freezeout is a concept to describe ratios of particle abundances such as @xmath92 and the above reactions do not alter the average abundances in the final state .
the success of the model , on the other hand , indicates that creations and annihilations of ( anti-)nucleons hardly occur after chemical freezeout .
second , we note that the dynamical model in ref .
@xcite uses an equation of states having a first order phase transition in the hydrodynamic simulations for the time evolution above the critical temperature @xmath93 .
recently , dynamical simulations have been carried out with more realistic equations of states obtained by lattice qcd simulations @xcite .
the lifetime of hadronic stage evaluated in these studies is more relevant to this argument .
we , however , note that the qualitative behavior of the time evolution seems not sensitive to the difference in equations of states @xcite . while n@xmath80 reactions frequently take place even below the chemical freezeout temperature , @xmath94 , @xmath95 annihilatios and productions almost terminate at @xmath94 .
this is necessary for the success of the thermal model . for @xmath96
the cross section of the @xmath95 pair annihilation is largest among all nn and n@xmath33 reactions . if nucleons and anti - nucleons are distributed without correlation , therefore , all nn and n@xmath33 reactions cease to take place at @xmath94 .
this conclusion is , of course , obtained also by evaluating the mean time for each reaction using the cross sections @xcite as in eq .
( [ eq : tau ] ) . after chemical freezeout
, the only inelastic reactions nucleons go through are thus eqs .
( [ eq : d+ ] ) and ( [ eq : d0 ] ) , and after each reaction the nucleon loses its initial isospin information . only after repeating the reactions eq .
( [ eq : p_pn ] ) twice , the ratio becomes @xmath97 , which is almost even .
if medium effects on the formations and decays of @xmath60 are negligible , therefore , irrespective of the nucleon distribution at the chemical freezeout , the isospin of nucleons at the kinetic freezeout can be regarded random and uncorrelated . on the other hand
, the nucleon number distribution can have a deviation from the boltzmann distribution reflecting the dynamical history of fireballs .
because of the absence of correlations between isospins of nucleons in the final state , once @xmath15 ( @xmath98 ) nucleons ( anti - nucleons ) exist in a phase space in the final state , their isospin distribution is simply given by the binomial one .
this conclusion leads to the factorization eq .
( [ eq : p ] ) for proton and neutron number distribution in the final state for an arbitrary phase space .
in particular , the final state proton and anti - proton number distribution is written as @xmath99 unlike in the simple example in sec .
[ sec : b : free ] , the nucleon distribution function @xmath26 in this case is determined by the time evolution of fireballs and is not necessary of a thermal or separable form as in eq .
( [ eq : f_hg ] ) ; no specific form for @xmath26 is assumed here or will be assumed in the analyses in sec .
[ sec : cumulant ] .
what we have used here is the fact that the time scale for the exchange of isospins between nucleons and pions is sufficiently short compared to the lifetime of hadronic stage after the chemical freezeout . on the other hand , the time scale for the variation of a conserved charge in a phase space depends on the form of the phase space , and can become arbitrary long by increasing the spatial volume . when the time scale is long , the information of the physics of the early stages is encoded in @xmath26 .
next , let us inspect the possibility of medium effects on the formation and decay rates of @xmath60 . in medium ,
the decay rate of @xmath60 acquires the statistical factor , @xmath100 where @xmath101 is the fermi distribution function and @xmath102 and @xmath103 are the energies of the nucleon and pion produced by the decay , respectively . the first term in eq .
( [ eq : f ] ) represents the pauli blocking effect . at the rhic energy ,
since the boltzmann approximation is well applied to nucleons , the pauli blocking effect is suppressed .
the bose factor @xmath104 in eq .
( [ eq : f ] ) , on the other hand , has a non - negligible contribution since @xmath105 .
as long as all @xmath106 for the three isospin states of the pion are the same , however , this factor does not alter the branching ratios eqs .
( [ eq : p_pn ] ) and ( [ eq : p_np ] ) , while the factor enhances the decay of @xmath60 .
a possible origin for the variation of @xmath106 is the isospin density of nucleon number ; since the isospin density is locally conserved , the isospin density of pions is affected by the nucleon isospin .
this effect on @xmath106 is , however , well suppressed since the density of pions is much larger than that of nucleons below @xmath94 .
another possible source which gives rise to a different pion distribution is the event - by - event fluctuation of the isospin density in the phase space at the hadronization .
it is , however , expected that the effect is well suppressed , again because of the large pion density .
one , therefore , can conclude that the medium effect hardly changes the branching ratios eqs .
( [ eq : p_pn ] ) and ( [ eq : p_np ] )
. the same conclusion also applies to the formation rate of @xmath60 , since the medium effect on the probabilities of a nucleon to undergo reactions eqs .
( [ eq : d++ ] ) - ( [ eq : d0 ] ) depends only on @xmath106 .
after all , all medium effects on the ratios eqs .
( [ eq : p_pn ] ) and ( [ eq : p_np ] ) are negligible . when the system has a nonzero isospin density , probabilities eqs .
( [ eq : p_pn ] ) and ( [ eq : p_np ] ) receive modifications because the three isospin states of the pion are not equally distributed , although this effect is not large as will be shown in sec .
[ sec : cumulant : nid ] . even in this case , however , the only modification to the above conclusion is to replace the probabilities @xmath50 and @xmath107 with appropriate values , since the reactions eqs .
( [ eq : d++ ] ) - ( [ eq : d0 ] ) still act to randomize the nucleon isospins with the modified probabilities determined by the detailed balance condition . here , we emphasize that the large pion density in the hadronic medium is responsible for the validity of eq .
( [ eq : gg ] ) in the final state . in the hadronic medium ,
there are so many pions which can be regarded as a heat bath when the nucleon sector is concerned , while nucleons are so dilutely distributed that they do not feel other ones existence .
so far , we have limited our attention to reactions mediated by @xmath51 .
interactions of nucleons with other mesons , however , can also take place in the hadronic medium , while they are much less dominant .
it is also possible that @xmath60 interacts with thermal pions to form another resonance before its decay @xcite .
all these reactions with thermal particles , however , proceed with certain probabilities determined by the isospin su(2 ) symmetry as long as they are caused by the strong interaction .
each reaction of a nucleon thus makes its isospin random , and act to realize the factorization eq .
( [ eq : p ] ) . the factorization eq .
( [ eq : gg ] ) is fully established for the rhic energy . at very low beam energy , however ,
pions are not produced enough and the duration of the hadron phase below @xmath94 becomes shorter .
nucleons , therefore , will not undergo sufficient charge exchange reactions below @xmath94 .
when the reactions hardly occur , the isospin correlations generated at the hadronization remain until the final state . at low beam energy , also
the density of the nucleon becomes comparable to that of pions , and pions can no longer be regarded as a heat bath to absorb isospin fluctuations of nucleons .
the requirements to justify the factorization eq .
( [ eq : gg ] ) , therefore , eventually breaks down as the beam energy is decreased .
this would happen when @xmath108 , since the abundance of pions is responsible for all of the above conditions . from the @xmath2 dependence of the chemical freezeout line on the @xmath1-@xmath0 plane @xcite , the factorization eq .
( [ eq : gg ] ) should be well - satisfied in the range of beam energy @xmath109gev .
so far , we have limited our attention to nucleons . since baryons in the final state in heavy ion collisions are dominated by nucleons , the nucleon number , which is not a conserved charge , is qualitatively identified with the baryon one .
it is , however , important to recognize the difference between these two fluctuation observables especially in considering higher - order cumulants .
the difference predominantly comes from strange baryons @xmath110 and @xmath111 . in this subsection , we argue a practical method to include the effect of these degrees of freedom in our factorization formula .
strange baryons produced in the hadronic medium decay via the weak or electromagnetic interaction outside the fireball .
@xmath110 decays via the weak interaction into @xmath112 and @xmath113 with the branching ratio @xmath114 on the other hand , branching ratio of @xmath115 is @xmath116 while @xmath117 always decays into @xmath118 .
@xmath119 decays into @xmath110 via the electromagnetic interaction and then decays with eq .
( [ eq : p_l ] ) @xcite .
if the @xmath110 and @xmath111 multiplets are created with an equal probability , the production ratio of @xmath120 and @xmath30 from their decays is given by @xmath121 actually , because of the mass splitting between @xmath110 and the @xmath111 triplets , @xmath122 , the production of the @xmath111 triplets are a bit suppressed compared to that of @xmath110 .
this makes the above ratio even closer to even .
if one can assume that the correlations between strange baryons emitted from the fireball are negligible , therefore , the number of nucleons produced by these decays can be incorporated into @xmath16 and @xmath17 in eq .
( [ eq : p ] ) . the nucleon number in eq .
( [ eq : p ] ) , then , is promoted to that of baryons .
the same argument holds also for @xmath123 and @xmath124 . in short , by simply counting all protons observed by detectors in the event - by - event analysis , @xmath15 and @xmath98 in eq .
( [ eq : p ] ) are automatically promoted to the baryon and anti - baryon numbers , respectively .
in this section , we focus on the cumulants of the baryon and proton numbers , and derive formulas to relate these cumulants on the basis of the factorization eq .
( [ eq : p ] ) . with these relations
the cumulants of the baryon number , which is a conserved charge , are calculated from experimentally observed proton number ones . in this section ,
we change the variables in the probability distribution function in eq .
( [ eq : p ] ) as @xmath125 where we have replaced the neutron numbers with the baryon ones , @xmath126 and @xmath127 .
it is understood that the prescription discussed in sec .
[ sec : b : s ] is adopted for @xmath110 , @xmath111 , and their antiparticles . before deriving formulas to relate the baryon and proton number cumulants , in this subsection we first remark that the distribution functions of these degrees of freedom satisfy a linear relation under the factorization eq .
( [ eq : p ] ) .
this relation explains why the baryon number cumulants can be represented by the proton number cumulants and vice versa .
let us start with the final state proton and anti - proton number distribution function , eq .
( [ eq : gg ] ) , @xmath128 with @xmath129 equation ( [ eq : g = fm ] ) shows that the distribution functions @xmath130 and @xmath131 satisfy a linear relation . since @xmath132 has the inverse , @xmath133 , @xmath131 is given in terms of @xmath130 as @xmath134 the specific form of @xmath135 is easily obtained by using the fact that the matrix eq .
( [ eq : kbb ] ) has a triangular structure , in the sense that @xmath132 takes nonzero values only for @xmath136 and @xmath137 . using eq .
( [ eq : f = gm ] ) , the baryon number distribution function @xmath131 @xcite is in principle determined by @xmath130 . in practice , however , this analysis does not work efficiently since the elements of @xmath138 are rapidly oscillating , which results in large errorbars in @xmath131 determined in this way . in the following , instead of
the distribution functions themselves , we concentrate on the cumulants of @xmath131 and @xmath130 . the moments and cumulants of a distribution function are defined in terms of their generating functions .
the moment generating function for the proton and anti - proton numbers with the probability @xmath139 is given by @xmath140 and the corresponding cumulant generating function reads @xmath141 derivatives of eq .
( [ eq : g ] ) give moments of @xmath139 , @xmath142 as long as the sum in eq .
( [ eq : g ] ) converges , while cumulants of the proton and anti - proton numbers are defined with eq .
( [ eq : k ] ) as @xmath143 the first - order cumulant is the expectation value of the operator @xmath144 while the second- and third - order cumulants are moments of fluctuations , such as , @xmath145 and so forth , with @xmath146 . substituting the explicit form of @xmath139 in eq .
( [ eq : gg ] ) for @xmath147 , one obtains @xmath148 where @xmath149 is the cumulant generating function for two independent binomial distribution functions . with eq .
( [ eq : k ] ) , one easily finds that this function satisfies @xmath150 and @xmath151 for positive integers @xmath30 and @xmath152 , with the cumulants of the binomial distribution function normalized by the total number @xmath153 and the same formulas for the anti - particle sector . imposing eqs .
( [ eq : kk ] ) - ( [ eq : k3 ] ) as the structure of @xmath147 , cumulants of net proton and baryon numbers , @xmath154 and @xmath155 , respectively , are calculated to be @xmath156 and @xmath157 a detailed description of the procedure to obtain these results is given in appendix [ app : cumulant ] .
we emphasize that no explicit form of @xmath131 is assumed in deriving these results .
moreover , in appendix [ app : cumulant ] we only use eq .
( [ eq : kk ] ) for the structure of @xmath147 and eqs .
( [ eq : k1 ] ) - ( [ eq : k3 ] ) for properties of @xmath158 to derive eqs . ( [ eq : np1x ] ) - ( [ eq : nb4x ] ) . therefore , these results hold for any distribution functions satisfying these conditions with the appropriate choice for the values of @xmath159 and @xmath160 . in hot medium produced by heavy ion collisions , ( anti-)proton and ( anti-)neutron number densities are in general different because of the isospin asymmetry of colliding heavy nuclei . in relativistic heavy ion collisions at sufficiently large @xmath2 and small impact parameters , however , the isospin density is negligibly small because a large number of particles having nonzero isospin charges ( mainly pions ) are created and most of the initial isospin density is absorbed by these degrees of freedom ( see , appendix [ app : isospin ] ) . when the isospin density vanishes , @xmath50 and @xmath107 are to be set at @xmath161 in the binomial distribution functions in eq .
( [ eq : p ] ) .
substituting @xmath162 into eqs .
( [ eq : np1x ] ) - ( [ eq : nb4x ] ) , which are obtained by putting @xmath163 in eq . ( [ eq : xi ] ) , one obtains @xmath164 and @xmath165 which are the results given in ref .
@xcite . here
a note is in order about the terms on rhss of eqs .
( [ eq : nb2])-([eq : nb4 ] ) .
each term on rhs of these equations is not necessarily uncorrelated with each other . in particular ,
generally @xmath166 is not separable , i.e. , it can not be written as @xmath167 . if there is such correlation ,
the statistical fluctuations of these terms are not independent but mutually correlated .
thus , an appropriate care need to be taken in estimating the statistical error for lhss of eqs .
( [ eq : nb2])-([eq : nb4 ] ) .
as the collision energy is lowered , the effect of nonzero isospin density eventually gives rise to non - negligible contribution to the above relations .
to investigate this effect , we first assume that the isospins of nucleons , anti - nucleons , and pions in the final state are in chemical equilibrium , as is indicated by the fast n@xmath80 reactions discussed in the previous section .
because the nucleon distribution is well approximated by the boltzmann distribution , the numbers of ( anti-)protons and ( anti-)neutrons in the final state are given with the isospin chemical potential @xmath168 and the temperature @xmath1 as @xmath169 where @xmath170 and @xmath171 are constants determined by the chemical freezeout condition such as the volume of the system , the rapidity coverage , and so on .
these relations lead to @xmath172 and thereby @xmath173 .
one thus can parametrize @xmath50 and @xmath107 as @xmath174 with the negative isospin density per nucleon @xmath175 @xmath176 assumes a positive value in heavy ion collisions . when the value of @xmath176 is small , @xmath177 , the effects of nonzero isospin density on eqs .
( [ eq : nb1x ] ) - ( [ eq : nb4x ] ) are well described by the taylor series with respect to @xmath176 .
substituting eq .
( [ eq : ralpha ] ) in these equations , up to the first order in @xmath176 eqs .
( [ eq : nb1 ] ) - ( [ eq : nb4 ] ) become @xmath178 next , let us estimate the value of @xmath176 in relativistic heavy ion collisions . under the chemical equilibrium condition ,
the ratio of the charged pion numbers , @xmath179 and @xmath180 , having isospin charges @xmath181 , is given by @xmath182 where we have adopted boltzmann statistics for pions , since the effect of bose - einstein correlation on the pion density is about @xmath183 for @xmath184 and does not affect our qualitative conclusion .
the experimental result for @xmath185 in the final state is almost unity for high energy collisions in accordance with the approximate isospin symmetry .
substituting eq .
( [ eq : n_pi_pm ] ) in eq .
( [ eq : alpha ] ) and using @xmath186 , one obtains @xmath187 the value of @xmath176 , as well as @xmath188 , grows as @xmath2 is lowered . in order to see how these parameters become non - negligible for small @xmath2 , we focus on the @xmath189gev collision at the sps ( @xmath190gev ) . for this collision ,
the experimental value of @xmath185 is @xmath191 @xcite . substituting the worst value within @xmath192 , @xmath193 , in eq .
( [ eq : alpha - pi ] ) , one obtains @xmath194 . on the other hand , below the top sps energy the production of anti - nucleons
is well suppressed and one can replace all @xmath195 and @xmath196 in eqs .
( [ eq : nb1i ] ) - ( [ eq : nb4i ] ) with @xmath197 to a good approximation . equation ( [ eq : nb4i ] ) , for example , then becomes @xmath198 this result shows that for @xmath199 the corrections of nonzero isospin density to eqs .
( [ eq : nb1 ] ) - ( [ eq : nb4 ] ) are less than @xmath183 in magnitude .
the effect is smaller in relations for the lower - order cumulants , eqs .
( [ eq : nb1i ] ) - ( [ eq : nb3i ] ) , and formulas for proton number cumulants , eqs .
( [ eq : np1 ] ) - ( [ eq : np4 ] ) . with these results , one can conclude that the formulas for the isospin symmetric case , eqs .
( [ eq : nb1 ] ) - ( [ eq : nb4 ] ) , can safely be used to the analysis of the baryon number cumulants for @xmath200 with a precision of less than @xmath183 . because the production of isospin charged particles increases as @xmath2 goes up , the value of @xmath176 , and hence the effect of nonzero isospin density on eqs .
( [ eq : nb1 ] ) - ( [ eq : nb4 ] ) are more suppressed for higher energy collisions .
as @xmath2 is lowered , the value of @xmath176 grows and eventually approaches the one in the colliding heavy nuclei , @xmath201 . for @xmath202 , the first - order correction in eq .
( [ eq : nb4alpha ] ) is comparable with the zeroth - order one .
relations for the isospin symmetric case , eqs .
( [ eq : nb1 ] ) - ( [ eq : nb4 ] ) , therefore , are no longer applicable . for such collision energies , however , conditions required for the factorization eq .
( [ eq : p ] ) themselves break down as discussed in sec .
[ sec : b : hic ] . before closing this subsection , we recapitulate that the suppression of the isospin density in the nucleon sector , and hence @xmath176 , in the final state is caused by the production of the large number of particles having isospin charges , especially charged pions . in appendix [ app :
isospin ] , we present an analysis for this effect .
as emphasized in the previous sections , the cumulants of the proton and baryon numbers are in general different .
one , therefore , has to be careful when comparing theoretical predictions on baryon number cumulants with experimental proton number ones . in this subsection , we show that the deviation from the thermal distribution in baryon number cumulants becomes difficult to measure in proton number cumulants using relations obtained in the previous section with some additional assumptions . in general , it is possible that , while the net baryon number fluctuations in the final state have a considerable deviation from the grand canonical ones reflecting the hysteresis of fireballs and/or the global charge conservation , baryon and anti - baryon numbers separately follow the thermal ( boltzmann ) distributions .
for example , if the net baryon number fluctuations above @xmath93 survive until the final state , the net baryon number fluctuations remain small compared to the thermal ones in the hadronic medium , while baryon and anti - baryon number fluctuations separately follow the thermal one .
generally , cumulants of net numbers can not take arbitrary values ; for instance , the second - order cumulant is constrained by the cauchy - schwartz inequality : @xmath203 the values of net baryon number cumulants satisfying these constraints are not forbidden .
suppose that , as an extreme case , the net baryon number fluctuations completely vanish and the left equality in eq .
( [ eq : cs ] ) is realized .
a baryon and anti - baryon distribution function @xmath204 which is a constrained baryon and anti - baryon number distribution following the canonical distribution , constitutes such an example .
the distribution function @xmath205 for free gas in the grand canonical ensemble , i.e. , an unconstrained case , on the other hand , is given by eq .
( [ eq : f_hg ] ) .
now , let us consider the difference between the net baryon and net proton number cumulants when the baryon and anti - baryon number distributions follow boltzmann statistics while the net baryon number does not . because of the boltzmann nature of @xmath206 and @xmath207 , distributions of @xmath16 and @xmath18 are also poissonian from eq .
( [ eq : p ] ) .
thus , cumulants of the baryon and proton numbers satisfy @xmath208 and the same for the anti - baryon and anti - proton numbers , where @xmath209 is the expectation value for free hadron gas ( hg ) composed of mesons and nucleons at @xmath94 , i.e. , a simplified version of the hrg model @xcite . the factors two in front of the proton number cumulants in eq .
( [ eq : hg ] ) are understood from eq .
( [ eq : pp = pb ] ) . using eq .
( [ eq : hg ] ) , the second terms in eqs .
( [ eq : np2 ] ) and ( [ eq : np3 ] ) are transformed as @xmath210 where in the last equalities we have used the fact that the proton and anti - proton numbers do not have correlations in the free gas , i.e. , @xmath211 . substituting eqs .
( [ eq : hg1 ] ) and ( [ eq : hg2 ] ) in eqs .
( [ eq : np2 ] ) and ( [ eq : np3 ] ) , respectively , one obtains @xmath212 these results show that the second terms on the rhss , which come from the binomial distributions of the nucleon isospin , have significant contribution to the cumulants of the proton number , and the contribution of the net baryon number cumulants , @xmath213 , are relatively suppressed .
since the second terms give the thermal fluctuations , these results show that the deviation of @xmath213 from the thermal value is hard to be seen in the proton number cumulants .
although one can not transform the fourth - order relation eq .
( [ eq : np4 ] ) to a simple form as in eqs .
( [ eq : np2hg ] ) and ( [ eq : np3hg ] ) , from the factor @xmath214 in front of @xmath215 in eq .
( [ eq : np4 ] ) it is obvious that the direct contribution of this term to experimentally measured @xmath216 is more suppressed compared to the lower - order cumulants , and that its experimental confirmation is more difficult .
these analyses strongly indicate that , even if the baryon number cumulants have considerable deviation from the thermal values , they are obscured in the experimentally measured proton number cumulants due to the redistribution in isospin space .
such a tendency seems to become more prominent for higher - order cumulants .
it is known that higher - order cumulants of the baryon number have large critical exponents and thus can have significant enhancement in the vicinity of the critical point @xcite .
the above result , however , indicates that such enhancement is suppressed by a factor @xmath217 and difficult to measure in experiments in proton number cumulants .
the analysis of the baryon number cumulants with eqs .
( [ eq : nb1 ] ) - ( [ eq : nb4 ] ) enables to remove the thermal contribution in the proton number cumulants and makes the direct experimental observation of signals in @xmath218 possible .
the @xmath2 dependences of proton number cumulants are recently measured by star collaboration at rhic @xcite .
the experimental result shows that ratios between net proton number cumulants follow the prediction of the hrg model within about @xmath183 precision . we , however , emphasize that one should not conclude from this result that baryon number cumulants also follow the prediction of the hrg model within @xmath183 precision .
as demonstrated above , the binomial nature of isospin distribution makes proton number cumulants close to the ones in the hrg model . in this sense , it is interesting that the experimental results for skewness and kurtosis nevertheless have small but significant deviations from the hrg predictions @xcite . the deviation , for example , in skewness , can be a consequence of @xmath219 in eq .
( [ eq : np3hg ] ) , which possibly reflects the properties of the matter in the early stage . a remark on eqs .
( [ eq : np2hg ] ) and ( [ eq : np3hg ] ) is in order .
these formulas are obtained with the assumption that baryon and anti - baryon number distributions are poissonian , while the net baryon number is not .
when one further assumes that the net baryon number cumulants also follow the thermal distribution in these results , these formulas simply reproduce the free gas result @xmath220 as they should do .
this is easily checked by substituting @xmath221 in eqs .
( [ eq : np2hg ] ) and ( [ eq : np3hg ] ) . in this subsection , we considered the experimental results on proton number cumulants using the results in sec .
[ sec : cumulant ] . more direct application of these formulas , i.e. , to determine baryon number cumulants from experimental results on proton number cumulants with eqs .
( [ eq : nb1 ] ) - ( [ eq : nb4 ] ) , is to be done .
the baryon number cumulants obtained in this way are to be compared with various theoretical predictions .
so far , we have considered the reconstruction of the missing information for the neutron number in experiments using eq .
( [ eq : p ] ) .
it is possible to extend this argument to infer different information on the event - by - event analysis .
an example is the evaluation of the effect of efficiency and acceptance of detectors .
the experimental detectors usually do not have @xmath222 acceptance .
moreover , protons entering a detector are identified with some efficiency .
if one can assume that protons ( anti - protons ) in the final state is detected by the detector with a fixed probability @xmath223 ( @xmath224 ) independent of momentum , multiplicity , and so on , and the efficiency for each particle does not have correlations , the distribution function @xmath225 for the observed proton and anti - proton numbers , @xmath226 and @xmath227 , respectively , are related to the one for all particles entering the detector , @xmath16 and @xmath18 , as @xmath228 or substituting this result in eq .
( [ eq : p ] ) and using the property of the binomial distribution one obtains @xmath229 eq .
( [ eq : pobs ] ) indicates that when the deviations of @xmath223 and @xmath224 from the unity become large , they affect cumulants with different orders differently . the effect of efficiency , therefore , can not be canceled out by taking the ratio between cumulants . in particular , as @xmath223 and @xmath224 become smaller , @xmath230 approach the product of independent poisson distributions irrespective of the form of @xmath131 .
this would be another reason of the present experimental results on proton number cumulants @xcite , which is consistent with the hrg model .
other experimental artifacts which have not taken into account yet in experimental analyses are background and misidentified protons . in particular , according to ref .
@xcite , the contamination from knockout protons is not negligible . by their nature , they give poissonian contribution and make observed proton number cumulants approach the poissonian values .
indeed , the hijing + geant simulation in ref .
@xcite shows that these effects are considerable .
the most important results of the present paper is summarized in eqs .
( [ eq : np1 ] ) - ( [ eq : np4 ] ) and eqs .
( [ eq : nb1 ] ) - ( [ eq : nb4 ] ) , which are formulas relating baryon and proton number cumulants in the final state in heavy ion collisions . the baryon number cumulants are a conserved charge , and one of the fluctuation observables which is most widely analyzed by theoretical studies .
our results enable to determine the baryon number cumulants with experimental results in heavy ion collisions , and hence make the direct comparison between theoretical predictions and experiments possible .
such a comparison will provide significant information on the qcd phase diagram .
the results eqs .
( [ eq : np1 ] ) - ( [ eq : nb4 ] ) are obtained on the basis of the binomial nature of the nucleon and anti - nucleon number distributions in isospin space , which is justified for @xmath59 . although these results are obtained for isospin symmetric medium , the effect of nonzero isospin density in relativistic heavy ion collisions is well suppressed in this energy range because of the abundance of the created pions .
the authors thank stimulating discussions at the workshop `` fluctuations , correlations and rhic low energy runs '' held at the brookhaven national laboratory , u.s.a . , oct 3rd through 5th , 2011 .
this work is supported in part by grants - in - aid for scientific research by monbu - kagakusyo of japan ( no .
21740182 and 23540307 ) .
in this appendix , we derive eqs . ( [ eq : np1x ] ) - ( [ eq : nb4x ] ) . to obtain these relations , we start from the cumulant generating function eq .
( [ eq : kk ] ) , @xmath231 , \label{eq : ap : kk}\end{aligned}\ ] ] where @xmath232 is a shorthand notation for @xmath233 . in this appendix
, we also suppress the subscript in @xmath158 .
we require the following four conditions for the properties of @xmath234 : @xmath235 for positive integers @xmath30 and @xmath152 .
( [ eq : ap : k0 ] ) is satisfied for probability distribution functions normalized to unity .
( [ eq : ap : k1 ] ) - ( [ eq : ap : k3 ] ) are eqs .
( [ eq : k1 ] ) - ( [ eq : k3 ] ) in the text . all calculations in this appendix
are based only on these constraints on @xmath147 . using @xmath147 ,
the net proton number cumulants are given by @xmath236 to proceed the calculation of eq .
( [ eq : ap : net ] ) , it is convenient to use the cumulant expansion of eq .
( [ eq : ap : kk ] ) @xmath237^m_c \nonumber \\ = & 1 + \sum_f k(\theta,\bar\theta ) + \frac12 \sum_f ( \delta k(\theta,\bar\theta ) ) ^2 \nonumber \\ & + \frac1{3 ! } \sum_f ( \delta k(\theta,\bar\theta ) ) ^3 + \frac1{4 ! } \sum_f ( \delta k(\theta,\bar\theta ) ) ^4_c \nonumber \\ & + \cdots .
\label{eq : ap : cum}\end{aligned}\ ] ] each term on the far right hand side defines each cumulant , @xmath238^m_c$ ] , up to the fourth order , with @xmath239 because of eq .
( [ eq : ap : k0 ] ) , all @xmath234 and @xmath240 in a term in eq .
( [ eq : ap : cum ] ) must receive at least one differentiation so that the term gives nonzero contribution to eq .
( [ eq : ap : net ] ) .
this immediately means that the @xmath152-th order term in eq .
( [ eq : ap : cum ] ) can affect eq .
( [ eq : ap : net ] ) only if @xmath241 . the first - order net - proton number cumulant , eq .
( [ eq : np1x ] ) , is calculated to be @xmath242 with @xmath243 and @xmath244 . in the third equality in eq .
( [ eq : ap : np1c ] ) , we have used eqs .
( [ eq : ap : k1 ] ) and ( [ eq : ap : k2 ] ) .
the second - order relation , eq .
( [ eq : np2x ] ) , is obtained as follows : @xmath245 ^ 2 \nonumber \\ & = \xi_2 \langle n_{\rm b } \rangle + \bar\xi_2 \langle n_{\bar{\rm b } } \rangle + \langle ( \xi_1 \delta n_{\rm b } - \bar\xi_1 \delta n_{\bar{\rm b } } ) ^2 \rangle.\end{aligned}\ ] ] to obtain the third line , we have used eqs .
( [ eq : ap : k3 ] ) and ( [ eq : ap : k0 ] ) for the first and second terms , respectively .
the factor two in the second term comes from the number of the outcomes of the application of the two derivatives to the two @xmath240 in the second line . eqs .
( [ eq : ap : k1 ] ) and ( [ eq : ap : k2 ] ) are used in the last equality .
similar manipulations lead to eqs .
( [ eq : np3x ] ) and ( [ eq : np4x ] ) .
we note that the relation , @xmath246 ^ 4_c,\end{aligned}\ ] ] makes the calculation for the fourth - order cumulant more concise . to obtain eqs .
( [ eq : nb1x ] ) - ( [ eq : nb4x ] ) , we start from the following relation for the net baryon number cumulants , @xmath247^n_c \nonumber \\ & \equiv \sum_f \left [ \partial_\xi k \right]^n_c , \label{eq : ap : b}\end{aligned}\ ] ] with @xmath248 .
we suppress arguments in @xmath249 and @xmath250 throughout this subsection . the manipulation of eq .
( [ eq : ap : b ] ) for @xmath251 is trivial . for @xmath252 , eq .
( [ eq : ap : b ] ) is calculated to be @xmath253 in the second line , we introduced a symbol , @xmath254 and used the relation , @xmath255 where we have used eqs .
( [ eq : ap : k1 ] ) - ( [ eq : ap : k3 ] ) . the last equality in eq .
( [ eq : ap : nb2 ] ) comes from the definition of @xmath256 . to proceed to @xmath257 , we first introduce the following notation , @xmath258 for positive integers @xmath30 , @xmath152 , and @xmath259 .
@xmath260 for @xmath261 is also defined as in eqs .
( [ eq : ap : d1 ] ) , ( [ eq : ap : d2 ] ) , and ( [ eq : ap : d3 ] ) .
one easily finds i ) @xmath262 are invariant under the permutations of the subscripts , for example , @xmath263 , and ii ) when a subscript is one , it can be eliminated , e.g. , @xmath264 , while @xmath265 . with this notation ,
derivatives of @xmath266 are written as @xmath267 and so forth . using these relations , for example , eq .
( [ eq : ap : b ] ) for @xmath268 is calculated as @xmath269 which leads to eq .
( [ eq : nb3x ] ) . in the second and last equalities , we used @xmath270 a similar manipulation for @xmath271 leads to @xmath272 which gives eq .
( [ eq : nb4x ] ) .
in this appendix , we demonstrate that the isospin density of nucleons in the final state of heavy ion collisions is suppressed owing to the abundant production of particles having nonzero isospin charges . to simplify the calculation , we consider a gas composed of nucleons and pions in chemical equilibrium , and assume that pions and ( anti-)nucleons obey boltzmann statistics , since this approximation does not alter the qualitative conclusion in this appendix . under these assumptions , the ratios between the numbers of ( anti-)protons and ( anti-)neutrons in a phase space are given in terms of @xmath168 and @xmath1 as @xmath273 with @xmath274 , and the ratio of the numbers of @xmath275 and @xmath276 is given by @xmath277 with these relations , the total isospin in the phase space is calculated to be @xmath278 with the number of charged pions @xmath279 . in the initial state of heavy ion collisions ,
the isospin asymmetry of the colliding heavy nuclei @xmath280 is approximately @xmath281 . assuming that this isospin asymmetry equally distributes along the rapidity direction in the final state , one has @xmath282 . with eq .
( [ eq : n_i ] ) , one then obtains @xmath283 the term in the parentheses is larger than unity , and becomes larger as more charged pions and anti - nucleons are produced .
equation ( [ eq : ap : alpha ] ) thus shows that the value of @xmath176 is more suppressed than @xmath280 owing to the production of these particles .
if the contribution of other particles with nonzero isospin charges is taken into account , the value of @xmath176 is further suppressed .
i. arsene _ et al . _
[ brahms collaboration ] , nucl .
a * 757 * , 1 ( 2005 ) ; b. b. back _ et al . _
[ phobos collaboration ] , nucl .
phys . a * 757 * , 28 ( 2005 ) ; j. adams _ et al . _ [ star collaboration ] , nucl . phys . a * 757 * , 102 ( 2005 ) ; k. adcox _ et al . _ [ phenix collaboration ] , nucl .
phys . a * 757 * , 184 ( 2005 ) .
m. asakawa and k. yazaki , nucl .
a * 504 * , 668 ( 1989 ) .
m. a. stephanov , pos * lat2006 * , 024 ( 2006 ) [ arxiv : hep - lat/0701002 ] . m. m. aggarwal _ et al . _ [ star collaboration ] , phys .
lett . * 105 * , 022302 ( 2010 ) [ arxiv:1004.4959 [ nucl - ex ] ] .
b. mohanty [ star collaboration ] , j. phys .
g * 38 * , 124023 ( 2011 ) [ arxiv:1106.5902 [ nucl - ex ] ] ; s. kabana [ for the star collaboration ] , arxiv:1203.1814 [ nucl - ex ] .
m. bleicher , arxiv:1107.3482 [ nucl - th ] .
v. koch , arxiv:0810.2520 [ nucl - th ] .
m. a. stephanov , k. rajagopal , and e. v. shuryak , phys .
* 81 * , 4816 ( 1998 ) [ arxiv : hep - ph/9806219 ] ; phys .
d * 60 * , 114028 ( 1999 ) [ arxiv : hep - ph/9903292 ] .
y. hatta and m. a. stephanov , phys .
lett . * 91 * , 102003 ( 2003 ) [ erratum - ibid . *
91 * , 129901 ( 2003 ) ] [ arxiv : hep - ph/0302002 ] .
m. a. stephanov , phys .
* 102 * , 032301 ( 2009 ) [ arxiv:0809.3450 [ hep - ph ] ] . c. athanasiou , k. rajagopal and m. stephanov , phys .
d * 82 * , 074008 ( 2010 ) [ arxiv:1006.4636 [ hep - ph ] ] .
e. s. fraga , l. f. palhares and p. sorensen , phys .
c * 84 * , 011903 ( 2011 ) [ arxiv:1104.3755 [ hep - ph ] ] . m. asakawa , u. w. heinz , and b. mller , phys .
lett . * 85 * , 2072 ( 2000 ) [ arxiv : hep - ph/0003169 ] . s. jeon and v. koch , phys .
lett . * 85 * , 2076 ( 2000 ) [ arxiv : hep - ph/0003168 ] .
v. koch , a. majumder and j. randrup , phys .
lett . * 95 * , 182301 ( 2005 ) [ nucl - th/0505052 ] .
s. ejiri , f. karsch and k. redlich , phys .
b * 633 * , 275 ( 2006 ) [ hep - ph/0509051 ] .
m. asakawa , s. ejiri , and m. kitazawa , phys .
lett . * 103 * , 262301 ( 2009 ) [ arxiv:0904.2089 [ nucl - th ] ] .
b. friman , _ et al .
_ , eur .
j. c * 71 * , 1694 ( 2011 ) [ arxiv:1103.3511 [ hep - ph ] ] .
m. a. stephanov , phys .
lett . * 107 * , 052301 ( 2011 ) [ arxiv:1104.1627 [ hep - ph ] ] .
r. v. gavai and s. gupta , phys .
b * 696 * , 459 ( 2011 ) [ arxiv:1001.3796 [ hep - lat ] ] .
c. schmidt , prog .
suppl . * 186 * , 563 ( 2010 ) [ arxiv:1007.5164 [ hep - lat ] ] .
s. mukherjee , j. phys .
g g * 38 * , 124022 ( 2011 ) [ arxiv:1107.0765 [ nucl - th ] ] .
s. borsanyi , _ et al .
_ , jhep * 1201 * , 138 ( 2012 ) [ arxiv:1112.4416 [ hep - lat ] ] .
a. bazavov
_ et al . _
[ hotqcd collaboration ] , arxiv:1203.0784 [ hep - lat ] .
m. kitazawa and m. asakawa , phys .
c * 85 * , 021901r ( 2012 ) [ arxiv:1107.2755 [ nucl - th ] ] .
j. cleymans and k. redlich , phys .
* 81 * , 5284 ( 1998 ) [ arxiv : nucl - th/9808030 ] . the review of particle physics , k. nakamura , _ et al . _
( particle data group ) , j. phys .
g * 37 * , 075021 ( 2010 ) . c. nonaka and s. a. bass , phys .
c * 75 * , 014902 ( 2007 ) [ arxiv : nucl - th/0607018 ] .
p. braun - munzinger , k. redlich and j. stachel , arxiv : nucl - th/0304013 .
b. schenke , s. jeon and c. gale , phys .
c * 82 * , 014903 ( 2010 ) [ arxiv:1004.1408 [ hep - ph ] ] .
y. pang , t. j. schlagel , and s. h. kahana , phys . rev
. lett . *
68 * , 2743 ( 1992 ) .
p. braun - munzinger , b. friman , f. karsch , k. redlich and v. skokov , phys .
c * 84 * , 064911 ( 2011 ) [ arxiv:1107.4267 [ hep - ph ] ] ; nucl .
a * 880 * , 48 ( 2012 ) [ arxiv:1111.5063 [ hep - ph ] ] .
f. karsch and k. redlich , phys .
b * 695 * , 136 ( 2011 ) [ arxiv:1007.2581 [ hep - ph ] ] .
b. i. abelev _ et al . _ [ star collaboration ] , phys .
c * 79 * , 034909 ( 2009 ) [ arxiv:0808.2041 [ nucl - ex ] ] . | we explore the relation between proton and nucleon number fluctuations in the final state in relativistic heavy ion collisions .
it is shown that the correlations between the isospins of nucleons in the final state are almost negligible over a wide range of collision energy .
this leads to a factorization of the distribution function of the proton , neutron , and their antiparticles in the final state with binomial distribution functions . using the factorization
, we derive formulas to determine nucleon number cumulants , which are not direct experimental observables , from proton number fluctuations , which are experimentally observable in event - by - event analyses . with a simple treatment for strange baryons ,
the nucleon number cumulants are further promoted to the baryon number ones . experimental determination of the baryon number cumulants makes it possible to compare various theoretical studies on them directly with experiments .
effects of nonzero isospin density on this formula are addressed quantitatively .
it is shown that the effects are well suppressed over a wide energy range . |
objective - prism surveys revolutionized our knowledge of active galactic nuclei ( agns ) in the local universe .
in particular , many of the known nearby seyfert galaxies were discovered by markarian in the first byurakan survey ( fbs ; see khachikian this volume . )
galaxies were selected for inclusion in the fbs based on the presence of a uv excess .
this selection criterion is biased against the discovery of more heavily reddened agn such as seyfert 2 s and liner galaxies .
a number of line - selected surveys have been carried out over the past 20 years in order to select well - defined samples of local agns . in this paper , we review existing line - selected surveys for local agns carried out using objective - prism techniques ( see stepanian , this volume , for a discussion of the uv+line selected second byurakan survey ) .
we then present a discussion of a new , modern digital objective - prism survey for emission - line galaxies ( elgs ) , the kpno international spectroscopic survey ( kiss ) .
kiss probes several magnitudes deeper than existing surveys , and due to its selection via h@xmath0 emission , is less biased against seyfert 2 and liner galaxies than previous surveys .
kiss provides one of the best , well - defined samples of local agns for comparison to higher redshift studies .
the use of line - selected objective - prism surveys to detect active and star - forming galaxies in the local universe has a long and fruitful history ( e.g. , smith 1975 ; smith et al . 1976 ) .
the university of michigan ( um ) survey ( macalpine et al . 1977 ; macalpine & lewis 1981 ) selected objects via their [ oiii]@xmath1 emission line .
349 elgs were detected in 667 sq .
this is a number density of 0.52 per square degree . in lists 4 & 5 of the um survey ( for which complete follow - up spectroscopy is available from salzer et al .
1989 ) 166 elgs are found in 325 sq .
approximately 10% of the elgs detected in the um survey are agns : 9 seyfert 1 s and 7 seyfert 2 s .
wasilewski ( 1983 ) also utilized the [ oiii]-line as his selection criterion and found 96 elgs in 825 sq .
degrees , or 0.18 per square degree .
wasilewski found that 8% are agns with 1 seyfert 1 and 7 seyfert 2 galaxies being detected . the case survey ( pesch & sanduleak 1983 ; stephenson et al . 1992 ) used two selection criteria , both uv - excess and [ oiii]-line emission and cataloged 1440 elgs in 1551 sq .
degrees or 0.94 per sq . degree .
salzer et al .
( 1995 ) obtained follow - up spectroscopy for a complete subsample of 176 of the case galaxies and found that only 6% are agns : 2 seyfert 1 s , 7 seyfert 2 s , and 2 liners .
most recently the universidad complutense de madrid ( ucm ; zamaraono et al .
1994 , 1996 ) survey utilized selection via h@xmath0-emission to discover 263 elgs in 471 sq .
degrees or 0.56 per square degree .
follow - up spectroscopy by gallego et al .
( 1997 ) found that 14 ( or 5% ) of these are agns , including 5 seyfert 1 s and 9 seyfert 2 s .
the completeness limits of all of the above surveys range from @xmath2 to 17 .
most of the previous line - selected surveys discussed above have been done in the blue ( selecting via [ oiii]@xmath1 emission ) and all used photographic plates .
there are a number of biases introduced by these methods .
first , selection in the blue limits the sensitivity of the surveys to more heavily reddened seyfert 2 s and low - ionization liner galaxies .
second , because the standard iiia - j emulsions used in most photographic surveys cut off at @xmath3 , the redshift depth of these surveys is limited to @xmath4 .
a modern , digital survey using ccds to substantially improve survey depth ( to fainter magnitudes and higher redshifts ) plus selecting via a redder line ( h@xmath0 ) would greatly reduce these biases .
such a survey is now available , the kpno international spectroscopic survey ( kiss ) .
the kpno international spectroscopic survey ( kiss ) was initiated in 1994 as a collaborative effort between astronomers from russia , ukraine , and the us who shared a common interest in the study of galaxian activity .
the survey was envisioned as being the next generation markarian survey " for active and star - forming galaxies . by combining the traditional objective - prism survey method for emission line detection with modern detector technology and computer - based analysis ,
kiss is able to discover seyfert galaxies ( sy 1/sy 2 ) , liner s , starburst nucleus galaxies ( sbns ) , hii galaxies , and blue compact dwarfs ( bcds ) at least 2 magnitudes fainter than previous work .
the survey is thus a unique resource for probing the nature of star - forming and active galaxies in the nearby universe .
a full description of the survey can be found in salzer et al .
( 2000 , 2001 ) .
the kiss survey is conducted with ccd detectors on the 24-inch burrell schmidt on kitt peak .
for the first survey strip , a 2048@xmath5 ccd was used , which afforded a field - of - view of 70 arcmin ( 1.18@xmath6 ) square , with a scale of 2.03 arcsec / pixel .
this one - degree - wide strip was chosen to coincide with the century redshift survey ( geller et al .
1997 ) covering a strip at a constant declination of 29@xmath6 from ra = 12@xmath7 15@xmath8 to 17@xmath7 which encompasses an area of 62 deg@xmath5 .
the second survey strip goes through the botes void at a constant declination of 43@xmath6 from ra = 12@xmath7 to 16@xmath7 15@xmath8 covering an area of 66 deg @xmath5 . for the second strip a @xmath9 ccd with 1.45 arcsec / pixel resolution was used .
the kiss survey data consists of broadband @xmath10 and @xmath11 images ( for astrometric and photometric calibration ) , and spectral ( objective prism ) data .
the spectra were taken in the red ( using a blocking filter to restrict the wavelength range to 6400 - 7200 ) to detect h@xmath0 emission .
one of the important features of the kiss project is its objective criteria : the search and measurement process is carried out entirely by computer .
a complete package of iraf scripts and executables analyzes every source in the field ( typically 5000 7000 objects ) automatically ; the output of these routines is a catalog of photometric and astrometric data , and an extracted spectrum for every object in the field .
in addition , a list of elg candidates is produced .
our survey technique has proven to be quite successful . in our first red spectral strip
we have cataloged 1128 elg candidates .
this is about _ 18.1 per square degree_. in the second survey strip , we have detected 1030 elg candidates or 15.6 per square degree . for comparison ,
the entire markarian survey cataloged 1500 uv - excess galaxies ( 0.1 galaxy per square degree ) , and had a completeness limit of m@xmath12 = 15.2 ( mazzarella & balzano 1986 ) , while the deeper um ( median m@xmath12 = 16.9 ; macalpine et al .
1977 ) and ucm ( median m@xmath12=16.5 ; zamarano et al . 1994 ) surveys both detected only 0.5 elgs per square degree .
we have obtained follow - up spectroscopy for 725 of 1128 elg candidates selected via h@xmath0 emission from the first survey strip using the wiyn 3.5-m , the kpno 2.1-m , the mdm 2.4-m , the apo 3.5-m , and the lick 3-m telescopes as well as the hobby eberly telescope .
these spectra provide redshifts , h@xmath0equivalent widths , and line fluxes for various important emission lines including h@xmath13 , [ o iii]@xmath144959,5007 , hei@xmath155876 , [ o i]@xmath156300 , h@xmath0 , [ n ii]@xmath146548,6583 , and [ s ii]@xmath146717,6731
. measurements of the balmer decrement allow us to directly measure the extinction ( @xmath16 ) for each of these galaxies .
the follow - up spectra also allow us to identify agns in our sample : seyfert 1s are identified via their broad permitted emission lines , while seyfert 2 s and liners are distinguished primarily via the [ oiii]/h@xmath13 line ratio with liners having [ oiii]/h@xmath13 @xmath17 3 ( e.g. , veilleux & osterbrock 1987 )
. representative spectra of agn in the kiss sample are shown in figures 1 & 2 .
we find that 91% ( 663 ) of the kiss candidate elgs are confirmed emission - lines galaxies .
about 11% of the galaxies are active galactic nuclei with 67 ( 10 seyfert 1 s , 19 seyfert 2 s , and 38 liner galaxies ) detected via h@xmath0 with redshifts less than 0.095 , and an additional 10 ( 1 seyfert 1 , 5 seyfert 2 s and 4 qso s ) detected at higher redshifts where another line has redshifted into our filter .
a line diagnostic diagram plotting the logarithm of [ o iii]@xmath155007/h@xmath13 against the logarithm of [ n ii]@xmath156583/h@xmath0 is shown in figure 3 .
the solid line represents the hii sequence with low metallicity , high - ionization elgs in the upper left and high metallicity , low - ionization objects in the lower right part of the sequence .
the star - forming elgs follow the hii sequence allowing us to classify them from bcds in the upper left to starburst nuclei in the lower right .
agns are clearly separated from star - forming galaxies .
we find many more high metallicity , low - ionization objects than found with traditional [ o iii]@xmath155007-selected surveys .
the data make it clear that h@xmath0-selected surveys are much more effective at detecting the full range of ionization and metallicity present in star - forming galaxies than [ o iii ] or [ o ii ] observations .
we will concentrate here on the 67 h@xmath0-selected agn which constitute a well - defined local sample of agn .
note that because our survey has a sharp upper - wavelength cutoff due to the blocking filter used , the sample is _ volume - limited _ for the more luminous objects and all but 8 of the agns fall in this category .
figure 4 shows the integrated absolute magnitude distributions of the agns in the local kiss sample .
kiss detects agns with @xmath18 ranging from -21.8 to -18 ( assuming @xmath19 km s@xmath20 mpc@xmath20 ) .
the seyfert 1 , seyfert 2 , and liner luminosity distributions look similar .
figure 5 shows the @xmath21 color distributions of the kiss agns .
as expected , seyfert 2 and liner distributions are significantly redder ( median @xmath21 = 0.92 and 0.93 respectively ) than the seyfert 1 color distribution ( median @xmath22 ) . because of their redder colors ,
an h@xmath0-selected survey is much more sensitive to seyfert 2 s and liners than a blue ( e.g. , uv excess or [ oiii]-selected ) survey .
this is borne out by comparing the 2:1 liner / sy 2 ratio in the kiss survey to the complete lack of liners discovered in the [ oiii]-selected surveys discussed in section 2 .
we have also calculated the @xmath10-band luminosity function for the seyfert galaxies in our sample .
because our direct images lack the angular resolution required to resolve the nucleus , this is an _ integrated _ luminosity function .
however , it is directly comparable to the integrated local seyfert luminosity function of huchra & burg ( 1992 ) .
figure 6 shows the blue luminosity function for both the kiss sample and that of huchra & burg .
the two lfs are consistent , with the kiss lf extending to fainter luminosities .
note that for this calculation we have not accounted for the possibility that there are agns in the fraction of our sample for which we have no follow - up spectra . because we have preferentially obtained spectra for the more luminous galaxies in our sample
, we expect that such a correction would be small , but it could increase our volume densities slightly
. further discussion of the lf and its implications for agn evolution when compared to other local and higher redshift agn lfs will be discussed in a future paper ( sarajedini , gronwall & salzer 2002 ) .
we have also cross - correlated our h@xmath0-selected sample of elgs from both our first and second survey strips with surveys done in the x - ray , radio , and the far - infrared . a comparison of the kiss sample with the rosat all - sky survey ( voges et al .
1999 ) finds that 17 of the 2158 kiss elgs are detected by rosat .
these are all agns : 13 sy 1 s , 2 sy 2 s , and 2 liners .
the x - ray detected agns are primarily the most luminous agns detected by kiss . a cross - correlation of the kiss sample with the 1.4 ghz first survey ( white et al .
1997 ) done with the vla reveals that 178 of the 2158 kiss elgs are detected by first .
we have follow - up spectra for all of these sources and find 5 sy 1 s , 28 sy 2 s , and 30 liners .
similarly , 154 of the 2158 kiss elgs are found in the iras faint source catalog ( moshir et al .
1992 ) and follow - spectra for these sources have detected 1 sy 1 , 9 sy 2 s , and 12 liners .
the kiss elgs follow the well - known radio - ir correlation discovered by condon et al .
this correlation is thought to be due to the fact that both the radio and far - ir emission are dominated by star formation .
somewhat surprisingly , we find that the majority of the agns in our sample follow the same radio - ir correlation as the star - forming galaxies , implying that the dominant ionizing source in these objects might be young stars , not the active nucleus .
we are currently investigating this possiblity .
we also plan to use these data to test the validity of ir color diagnostics for agn activity .
the kiss sample represents a substantial step forward in our understanding of the agn population in the local universe .
it is essentially volume - limited and less biased against redder seyfert 2 and liner galaxies , and provides a well - defined sample of nearby ( @xmath23 ) agn with which to study the statistical properties of agn . in the first kiss survey strip covering 62 square degrees , 10 seyfert 1 s , 19 seyfert 2 s , and 38 liner s were discovered .
kiss provides an essential database of local agns for comparison to higher redshift samples .
cg is grateful to the american astronomical society and the international astronomical union for travel grants which allowed her to participate in this meeting .
she would also like to thank the local organizing committee for their wonderful hospitality . | objective - prism surveys for uv - excess and emission - line objects , especially the first and second byurakan surveys , have been central to the study of active galactic nuclei ( agns ) .
we review previous line - selected surveys for agns and discuss their contribution to our understanding of the agn phenomena .
in addition , we present results from the kpno international spectroscopic survey , a modern digital objective - prism survey for emission line objects .
this survey is discovering substantial numbers of new agns , in particular low - luminosity agns and liners . |
the simultaneous study of photospheric and chromospheric active regions on the sun allows us to trace the emersion of magnetic flux tubes .
recently , a tight spacial association between spots and plages has been observed in the young solar type star hd 206860 by means of strmgren photometry and caii h@xmath2k and h@xmath0 chromospheric emissions ( @xcite ) . the spot / plage association has been also monitored in some very active rs cvn binaries ( @xcite ) . in this work we show that the spatial spot / plage association is also observed in other g - k main sequence stars younger than the sun .
we detected evidence of photospheric inhomogeneities from light curves and temperature measurements obtained by means of the line - depth ratios ( ldrs ) method ( @xcite ) .
the chromospheric inhomogeneities have been detected from the variation of the h@xmath0 line equivalent width .
the cases of @xmath3 eri ( hd 22049 , k2v , @xmath4=0.88 ) , hd 166 ( k0v , @xmath4=0.75 ) , @xmath51 ori ( hd 39587 , g0v , @xmath4=0.59 ) and @xmath61 cet ( hd 20630 , g5v , @xmath4=0.68 ) are considered .
the photometric observations have been carried out in the standard strmgren _ ubvy _
system with the t4 0.75 m automatic photoelectric telescope at fairborn observatory in southern arizona ( usa ) , equipped with an emi 9124qb photomultiplier detector . a complete discussion of photometry with this telescope can be found in @xcite .
we analyzed data acquired from november 2000 to january 2001 , taken contemporaneously with the spectroscopic observations .
spectroscopic observations have been obtained during the same time interval as the photometry with the reosc chelle spectrograph fed by the 91-cm telescope at catania astrophysical observatory - _ m. g. fracastoro _ station ( serra la nave , mt . etna ) .
the spectral resolving power of about 14000 has been obtained in the cross - dispersed configuration with the 79-lines / mm chelle grating as a main dispersing element .
the spectra were recorded on a ccd camera equipped with a thinned back - illuminated site ccd of 1024@xmath71024 pixels ( size 24@xmath724 @xmath8 m ) .
the detector allows us to record five orders in each frame , spanning approximately 5850 to 6700 . in this spectral region
there are the h@xmath0 line and several line pairs , whose depth ratios are suitable for effective temperature determination ( @xcite ) . the average signal - to - noise ratio ( _ s / n _ ) at continuum in the spectral region of interest was 200 - 500 for the very bright standard stars and 100 - 200 for the target stars .
the spectra extraction was performed by using the echelle task of iraf following the usual steps : background subtraction , division by a flat field spectrum ( given by a halogen lamp ) , wavelength calibration using the emission lines of a thorium - argon lamp , aperture extraction and continuum fitting with a low order polynomial .
detailed information about the data reduction can be found in @xcite .
temperature determinations of our target stars have been made measuring the depth ratio of several line pairs , selected in the spectral interval 6190 - 6280 .
calibrations of individual ldr into temperature scale have been made through the observation of non variable stars of different spectral type .
more information about this technique is given in @xcite .
excess emission in the h@xmath0 line that contributes to filling in the line cores of the target stars have been extracted by using the spectral synthesis " method ( e.g. @xcite ) . for each active star , we have used the spectrum of an appropriate inactive star of the same spectral type that was rotationally broadened and subtracted from each individual spectrum of the target star . the net h@xmath0 equivalent width has been measured in such difference spectra by integrating the net emission profile .
rotational modulation of photospheric line fluxes in solar - type main sequence stars is not always evident . in our small sample ,
however , we have a different situation , and we describe our results on individual stars below .
@xmath3 eri and hd 166 are two examples where the variations induced by the presence of spots and plages at photospheric and chromospheric levels , as shown by the effective temperature and the h@xmath0 line flux , are evident .
@xmath3 eri is a bright ( @xmath9=373 , k2v ) , nearby ( 3.3 pc ) single late - type main sequence star that shows variability attributed to magnetic activity .
we have acquired spectra of this star from november 2000 to january 2001 .
phases are computed from the initial epoch of 2451856.0 corresponding to the first observing date ( november 7 , 2000 ) and a rotational period of @xmath10 = 11@xmath1168 ( @xcite ) .
the rotational period is somewhat variable because spots appear at different latitudes and rotation is latitude - dependent ( @xcite ) .
we have used the average photometric period for the epoch of our observations .
we find a fairly well - defined anti - correlation between the photospheric temperature curve ( @xmath12 ) , deduced by averaging nine ldrs , and the net h@xmath0 equivalent width curve , derived with the spectral synthesis method using 54 psc ( hd 3651 , k0v , @xmath4=0.849 ) as template .
the full amplitude of the @xmath12 variation is only 50 k , i.e. about 1% , while the equivalent width excursion is about 33% .
hd 166 ( @xmath9=613 , k0v ) is a young solar - type star belonging to the local association with an age between 20 and 150 myr .
the data plotted in fig . [ kbiazzof2 ] refer to spectra acquired from october 2000 to november 2000 .
the @xmath12 and @xmath1 variations are reported as a function of the rotational phase , computed from the following ephemeris : hjd@xmath13 = 2449540.0 + 6@xmath1123 @xmath14 , taken from @xcite .
the template used for the spectral synthesis method is @xmath15 cet ( hd 10700 , g8v , @xmath4=0.727 ) .
also in this case , the anti - correlation between photosphere and chromosphere is rather good .
the full amplitude of @xmath12 variation is 48 k , i.e. about 1% , while the @xmath1 excursion is about 34% .
@xmath51 ori ( @xmath9=441 , g0v ) is indeed a single - lined spectroscopic binary with a long period of @xmath16 = 5156@xmath117 ( @xcite ) .
it is a relatively rapid rotator because it is a young star belonging to the ursa major cluster with an age of 300 myr . the simultaneous temperature , light and h@xmath0 emission curves of this magnetically active star
are reported in fig .
[ kbiazzof3 ] .
the data have been folded in phase with the ephemeris hjd@xmath13 = 2451856.0 + 5@xmath1124 @xmath14 , where the rotational period is taken from @xcite .
the averaged effective temperature and the net h@xmath0 equivalent width do not give clear evidence of rotational modulation , while the @xmath17 photometry displays a very low amplitude ( @xmath180.02 mag ) modulation barely visible within the photometric noise . since @xmath51 ori is a rather active star , as denoted by its large h@xmath0 excess emission , we suppose that we have observed it at an epoch when the active regions were evenly distributed in longitude , which would give rise to the very low amplitude @xmath19 light and no h@xmath0 modulation .
the presence of significant h@xmath0 filling and a mean magnitude lower than the historical maximum supports this picture . applying the analytical approach proposed by @xcite , from the temperature variation amplitude
we estimate a minimum spot coverage ( with respect to the stellar surface ) of about 0.0327 and 0.0275 for @xmath3
eri and hd 166 , respectively , where these two values have been obtained in correspondence of a ratio between spot temperature and photospheric temperature of 0.825 .
@xmath61 cet ( @xmath9=483 , g5v ) is a single member of the hyades moving group with an estimated age of @xmath20750 myr . for this star we obtained simultaneous light and temperature curves , allowing us to make a spot model for a rough reconstruction of the photospheric inhomogeneities .
we have also developed a plage model to be applied to the h@xmath0 equivalent width curve .
all curves show asymmetric shapes , so that all the solutions required at least two active longitudes . the spot model that we used to reproduce the temperature and light curves is based on fixed geometric parameters of the cool spots ( longitudes , latitudes ) and
solves for the spot relative area @xmath21 , taking as a free parameter the ratio between the spot temperature and the photospheric temperature ( @xmath22 ) . in this way ,
we have obtained two grids of solutions , one for the @xmath17 curve and the other for the @xmath23 curve ( fig .
[ kbiazzof4 ] ) , as we did in our previous work on rs cvn stars ( @xcite ) .
their intersection provides the best values of the spot temperature @xmath24 and the projected area of the spots relative to the stellar surface @xmath21 .
moreover , for the evaluation of continuum flux , we have used the black body approximation and two atmospheric models developed by @xcite and @xcite . for the h@xmath0 curve we have considered a `` bright spot '' model with an emission flux ratio between plages and quiet chromosphere @xmath25=3 , that is the typical value of the brightest solar plages . in fig .
[ kbiazzof5 ] the @xmath12 variation , the @xmath17 photometry and the h@xmath0 line flux of @xmath61 cet are displayed as a function of the rotational phase ( dots ) .
the synthetic curves obtained by using the kurucz model solutions are also shown .
the ephemeris is hjd@xmath13 = 2451856.0 + 9@xmath1120 @xmath14 , where the rotational period is taken from @xcite .
both the temperature and light curves have a regular trend with the rotational phase and appear anti - correlated with the chromospheric flux curve .
this implies a good spatial correlation between the stellar spots and the chromospheric plages . as a matter of fact ,
the photospheric and chromospheric maps have no appreciable longitude difference between the photospheric ( spots ) and chromospheric ( plages ) active regions .
this result is similar to that obtained by @xcite for the young solar type star hd 206860 .
the contemporaneous spectroscopic and photometric observations of some solar - type stars here reported have given remarkable results in the study of the connection between stellar photospheric and chromospheric active regions .
the analysis of the rotational modulation of effective temperature and h@xmath0 line equivalent width of our sample of active solar - like stars has really shown evidence of spacial correlation among the active regions , i.e. cool spots in photospheres and bright faculae in chromospheres in relatively low - activity stars .
moreover , from the combined analysis of contemporaneous temperature and light curve variations , we have obtained unique solutions of the spot temperature and the spot coverage factor for the solar - type star @xmath61 cet by using a method already tested in some rs cvn stars ( @xcite ) .
also , for @xmath3 eri and hd 166 , we have found a valuation of these two parameters .
the spot temperatures we derived are closer to solar spot penumbrae rather than umbrae , probably due to the larger size of stellar penumbra , which has a greater weight in determining the spectral line depths .
in addition , the relative spot coverage in these stars is much smaller than in the spotted rs cvn stars ( @xcite ) and more similar to the largest solar spot groups .
catalano s. , biazzo k. , frasca a. , marilli e. 2002 , a&a , 394 , 1009 catalano s. , biazzo k. , frasca a. et al .
2002 , an , 323 , 260 catalano s. , lanza a.f . , brekke p. et al .
1998 , in asp conf .
ser . : the 10th cambridge workshop on cool stars , stellar systems and the sun , eds .
donahue r.a . &
bookbinder j.a .
154 , p. 584
catalano s. , rodon m. , cutispoto g. et al .
2000 , in kluwer academic publishers : variable stars as essential astrophysical tools , ed .
ibanoglu c. , vol .
544 , p. 687
donahue r.a . , saar s.h . , baliunas s.l .
1996 , apj , 466 , 384 frasca a. , biazzo k. , catalano s. et al .
2004 , a&a , in press frasca a. & catalano s. 1994 , a&a , 284 , 883 frasca a. , freire ferrero r. , marilli e. , catalano s. 2000 , a&a , 364 , 179 frey g.j . , grim b. , hall d.s .
aj , 102 , 1813 gaidos e.j . ,
henry g.w .
, henry s.m .
2000 , aj , 120 , 1006 han i. & gatewood g. 2002 , pasp , 114 , 224 hauschildt p.h . , allard f. , ferguson j. et al .
1999 , apj , 525 , 871 henry g.w . 1999 , pasp , 111 , 845 kurucz r.l .
1993 , atlas9 stellar atmosphere programs and 2 km s@xmath26 grid , ( kurucz cd - rom no . 13 ) messina s. , rodon m. , guinan e.f .
2001 , a&a , 366 , 215 | as part of our program to study stellar photospheric and chromospheric activity , we have examined several young solar type stars with activity levels intermediate between the sun and the very active rs cvn binaries .
we have analysed contemporaneous spectroscopic data obtained at catania observatory ( serra la nave station , mt . etna ) and photometric data acquired in the strmgren bands with an automatic photometric telescope ( apt ) at fairborn observatory ( arizona , usa ) .
surface inhomogeneities have been detected from the rotational modulation of stellar brightness as well as from the modulation of several photospheric line - depth ratios ( ldrs ) .
the presence of chromospheric plage - like regions has been inferred from the rotational modulation of the h@xmath0 line equivalent width ( @xmath1 ) evaluated with the spectral synthesis method .
the most relevant results are the strong correlation between the brightness and temperature curves derived respectively from photometry and the ldrs as well as the striking anti - correlation between brightness and h@xmath0 emission .
this suggests a close spatial association of spots and plages , as frequently observed for the largest sunspot groups ( e.g. @xcite ) and for some very active rs cvn systems ( @xcite ) .
moreover , a simple spot / plage model applied to the observed flux curves allows a rough reconstruction of photospheric and chromospheric features of young main sequence stars . |
resonant excitation of single qds has recently gained a lot of interest @xcite , in part because this type of coherent excitation is promising for the generation of single photons with excellent coherence properties @xcite .
the techniques developed for effective laser stray light suppression have enabled the collection of resonance fluorescence from a single qd with high signal - to - noise ratio .
resonance fluorescence ( rf ) emission below saturation of the quantum emitter has revealed close - to - fourier limited single photons with record - high emission coherence and two - photon interference visibility @xcite .
recent experiments have even been able to beat the fourier limit for single - photon emission coherence in the heitler regime , e.g. , with excitation strengths well below those to saturate the quantum emitter @xcite .
another major achievement with respect to rf is the demonstration of single- and cascaded photon emission between the mollow sidebands above saturation of the qd @xcite .
recent investigations of single qd rf have revealed distinct differences of their emission coherence properties @xcite which need to be theoretically treated beyond a simple two - level description usually used for atoms .
one of the main consequences of the solid - state character of these quantum emitters is the consideration of specific dephasing channels primary caused by carrier - phonon scattering .
dephasing of a resonantly driven qd system has been theoretically studied in detail with respect to electron - phonon interaction on the dynamics of an optically driven system @xcite .
these studies anticipated excitation - induced dephasing ( eid ) for moderate rabi frequencies . however , non - monotonic behavior was predicted for rabi frequencies larger than a cut - off frequency defined by the material parameters and the qd size @xcite .
non - monotonic behavior is also predicted for cavity structures with suitable cavity coupling @xcite .
experimental evidence of eid effects has recently been observed as oscillation damping in pulsed photocurrent measurements on a resonantly driven qd @xcite .
this damping was found to have a clear quadratic dependence on the effective rabi frequency @xmath5 .
the effect of eid has also been observed under strictly resonant continuous wave excitation of a qd in a microcavity in terms of mollow - triplet sideband broadening @xcite .
these experiments reveal good agreement with a theoretical description based on a polaron master equation approach to multi - phonon and multi - photon effects in a cavity - qed system @xcite . in the work of ulrich _
et al . _
@xcite , the phenomenon of spectral mollow sideband _ narrowing _ in dependence of laser - excitation detuning from the bare emitter resonance had to be left open for further in - depth theoretical analysis .
motivated by these findings and with the aim of fundamental interpretation , the focus of the current work lies on a detailed study of the detuning - dependent dressed state emission of a single qd without cavity coupling . our theory is based on a polaron master equation approach from which we develop a fully analytical description of the emission spectrum .
we are able to distinguish between different regimes of spectral broadening or narrowing of the mollow sidebands , under strong influence by pure dephasing and phonon - induced scattering .
the comparison of detuning - dependent resonance fluorescence data reveals very good agreement with the theoretical model .
the planar sample employed for the measurements in this work is grown by metal - organic vapor epitaxy ( movpe ) .
the self - assembled in(ga)as qds are embedded in a gaas @xmath6-cavity , sandwiched between 29 ( 4 ) periods of @xmath7-thick alas / gaas layers as the bottom ( top ) distributed bragg reflectors ( dbrs ) . for our experimental investigations ,
the sample is kept in a helium flow cryostat providing high temperature stability @xmath8
k. suppression of parasitic laser stray - light is achieved by use of an orthogonal geometry between qd excitation and emission detection .
in addition , polarization suppression and spatial filtering via a pinhole is applied in the detection path .
resonant ( tunable ) qd excitation is achieved by a narrow - band ( @xmath9 khz ) continuous - wave ( cw ) ti : sapphire ring laser . for high - resolution spectroscopy ( hrpl ) of micro - photoluminescence ( @xmath10-pl )
we employ a scanning fabry - prot interferometer with @xmath11ev as described earlier @xcite ) .
in our experiments we apply pump powers well above the saturation of the quantum emitters . in this high - field regime ,
the excitation - induced rabi rotation of the two - level emitter system becomes much faster than the spontaneous decay rate @xmath12 .
the incoherent spectrum of the resulting dressed state is the characteristic mollow triplet @xcite . under strictly resonant excitation ,
i.e. for a laser detuning @xmath13 from the qd exciton resonance ( see fig .
1(a ) , green center trace ) , the spectrum is composed of the central _ rayleigh line _
`` r '' at the bare emitter energy @xmath14 and two symmetric satellite peaks , i.e. the _ three photon line _
`` t '' and the _ fluorescence line _ `` f '' at @xmath15 , respectively . @xmath16 denotes the effective rabi frequency including renormalization effects from the phonon bath as discussed in the theory section below .
laser detuning ( @xmath17)-dependent mollow triplet spectra taken at a constant excitation strength of @xmath18 ( @xmath19 const . )
are depicted in fig .
according to theory ( see , e.g. ref .
@xcite ) , the laser - detuning between the driving field and the bare emitter resonance @xmath14 modifies the dressed emission . besides the center transition at @xmath20 the two sideband frequencies become @xmath21 , where @xmath22 denotes the generalized rabi frequency at a given excitation strength .
the extracted spectral positions for the red- and blue - shifted mollow sidebands and the central rayleigh line are depicted in fig .
1(b ) with a corresponding fit to the data . in addition , observations on the detuning - dependent mollow triplet series depicted in fig .
1(a ) reveal distinct broadening of the mollow sidebands with increasing @xmath17 , accompanied by a change in the relative sideband intensities . in order to explain these observations
, we will develop a theoretical description in terms of a polaron master equation formalism and derive an analytical expression for the incoherent spectrum .
the main theoretical findings used to interpret our experimental observations are described in the following section .
we model the qd as a effective two - level system interacting with a coherent pump field and an acoustic phonon reservoir . in a frame rotating with respect to the laser pump frequency @xmath23 , the model hamiltonian ( excluding qd zero - phonon line decay )
is @xmath24 where @xmath25 are the annihilation and creation operators of the phonon reservoir , @xmath26 ( and @xmath27 ) are the pauli operators of the exciton .
@xmath28 is the exciton pump rate , and @xmath29 ( assumed real ) is the coupling strength of the electron - phonon interaction . in order to include electron - phonon scattering nonperturbatively ,
we transform the above hamiltonian to the polaron frame .
consequently , we derive a polaron master equation ( me ) @xcite which is particularly well suited for studying quantum optical phenomena such as resonance fluorescence spectra . in the following
we will closely follow ( and extend where necessary ) the theoretical formalisms described in refs .
@xcite , except we can neglect cavity terms .
defining @xmath30 , then the polaron transformed hamiltonian , @xmath31 @xcite , consists of a system part , reservoir part , and an interaction part , respectively : @xmath32 with the coherent displacement operators @xmath33 defined as @xmath34,$ ] and @xmath35 and @xmath36 .
the polaron shift is @xmath37 where @xmath38 denotes the characteristic phonon spectral function that describes the la - phonon interaction resulting from deformation potential coupling .
the thermally - averaged bath displacement operator is defined through @xcite @xmath39 $ ] , with @xmath40 , at a bath temperature @xmath41 . for convenience
, we will assume that the polaron shift is implicitly included in our definition of @xmath42 below .
the operators @xmath43 and @xmath44 are defined through @xmath45 and @xmath46 we next present the time - local ( or time - convolutionless ) me for the reduced density operator @xmath47 of the qd - bath system in the second - order born approximation of the system - reservoir coupling . in the interaction picture , we consider the exciton - photon - phonon coupling @xmath48 in the born approximation and trace over the phonon degrees of freedom .
the _ full polaron me _ takes the following form @xcite : @xmath49 + \frac{{\gamma}}{2}{\cal l}[\sigma^- ] + \frac{{\gamma'}}{2}{\cal l}[\sigma_{11 } ] % + \frac{{\gamma}}{2}(2\sigma^{-}\rho\sigma^{+ } % -\sigma^{+}\sigma^{-}\rho-\rho\sigma^{+}\sigma^{- } ) % + \frac{{\gamma'}}{2}(2\sigma_{11}\rho\sigma_{11 } % -\sigma_{11}\sigma_{11}\rho-\rho\sigma_{11}\sigma_{11 } ) \nonumber \\ & -\frac{1}{\hbar^{2}}\int^{t}_{0}d\tau\sum_{m = g , u } \bigg ( g_{m}(\tau ) \left [ { x}_{m},e^{-ih_{s}^{\prime}\tau/\hbar}{x}_{m}e^{ih_{s}^{\prime}\tau/\hbar}\rho(t)\right ] + { \rm h.c . } \bigg ) , \label{me1}\end{aligned}\ ] ] where @xmath50 and the time - dependent function @xmath51 is derived as follows : @xmath52 - 1 \right ) $ ] and @xmath53 $ ] @xcite , with the phonon correlation function @xmath54 $ ] .
the lindblad operators @xmath55=2o\rho o^\dagger -o^\dagger o\rho - \rho o^\dagger o$ ] describe dissipation through zero - phonon line ( zpl ) radiative decay ( @xmath12 ) and zpl pure dephasing ( @xmath56 ) , where the latter process is known to increase as a function of temperature @xcite . without the coherent pump term and the residual zpl broadenings ,
this me formally recovers the independent boson model @xcite . for continuous wave ( cw ) excitation , the integration appearing in eq .
( [ me1 ] ) can have the upper time limit @xmath57 , resulting in a markovian me where the scattering rates are computed as a function of @xmath58 @xcite .
such an approach is valid since the acoustic phonon lifetimes are very fast , i.e. on a few ps timescale .
as was shown earlier @xcite , for the pump strengths we consider , one can neglect the pump - dependence of @xmath58 appearing in the exponential phase terms above ( which we will further justify below ) to derive an _ effective phonon me _ as @xmath59 + \frac{\gamma}{2}{\cal l}[\sigma^- ] + \frac{\gamma'}{2}{\cal l}[\sigma_{11}]+\frac{\gamma_{\rm ph}^{\sigma^+}}{2}{\cal l}[\sigma^+ ] + \frac{\gamma_{\rm ph}^{\sigma^-}}{2}{\cal l}[\sigma^- ] \nonumber \\ & - { { \gamma{\rm}^{\rm cd}_{\rm ph}}}(\sigma^+\rho\sigma^++\sigma^-\rho\sigma^- ) \quad . \label{me2}\end{aligned}\ ] ] here the pump - driven incoherent scattering processes , mediated by the phonon bath , are obtained from @xmath60 , \nonumber\\ \label{rates}\end{aligned}\ ] ] and @xmath61 \quad .\end{aligned}\ ] ] the classical rabi frequency of the exciton pump , including renormalization effects from the phonon bath , is given by @xmath62 ( in contrast to the bare rabi frequency @xmath63 ) .
the scattering term @xmath64 is a cross - dephasing rate that only affects the off - diagonal components of the resulting optical bloch equations .
similar terms appear when a system is driven by a broadband squeezed light reservoir @xcite and are sometimes referred to as `` anomalous correlations '' . as might be expected , the excitation - dependent rates depend upon the phonon correlation function , the coherent pump rate , and the laser - exciton detuning .
the @xmath65 process corresponds to an _ enhanced radiative decay _ , while the @xmath66 process represents an _
incoherent excitation _
process @xcite .
we stress that these mechanisms are quite different to simple pure dephasing models , which are frequently used to describe weak ( i.e. , perturbative ) electron - phonon scattering @xcite .
note that @xmath16 can be significantly smaller than @xmath67 , even at low temperatures .
for example , using inas qd parameters that closely represent our experimental samples @xcite and a phonon bath temperature @xmath68k , then @xmath69 , and this value decreases ( increases ) with increasing ( decreasing ) temperature . for a pump field strength of @xmath70ev ,
example phonon scattering rates are shown in fig .
[ thfig1 ] . within the zoomed region of laser detunings
@xmath71ev the relevant rates can be assumed to be constant .
therefore , these values will be treated as constant in the following to compute the analytical mollow triplets . to get better insight into the underlying physics , it is desirable to derive an analytical form for the mollow triplet spectrum , which we derive from the effective phonon me
. thus we will first investigate how good the approximation is to replace the phonon scattering terms in the _ full _ polaron me [ eq .
( [ me1 ] ) ] by the ones appearing in the _ effective _ phonon me [ eq . ( [ me2 ] ) ] . in fig .
[ thfig2 ] , a direct comparison between the numerically calculated mollow triplet based on the full polaron and the effective phonon me is shown , revealing excellent agreement even for large detunings @xmath72ev and high field strengths of @xmath73ev . the main reason that one can neglect the pump - dependence of the phase terms in eq .
( [ me1 ] ) is that for the pump values we consider one ps timescale , phonon correlation times are much faster than the inverse rabi oscillation .
we highlight that a cw rabi field of @xmath74ev is already close to the highest achievable experiments to date , and thus for our purposes can be considered the high - field regime .
however , we note that the polaron approach , although nonperturbative , can break down if extremely high field strengths are used such that @xmath16 becomes comparable to ( or greater than ) the phonon cut - off frequency . in this case ,
other approaches exists such as a variational master equation approach @xcite and path integral techniques @xcite . since out
maximum rabi field strengths are much less than @xmath75 , as shown by mccutcheon _
et al . _
@xcite , the polaron me should be rigorously valid for the field strengths that we model .
one of our main theoretical goals is to derive a useful analytical expression that will allow one to fit the experimental mollow triplets over a wide range of parameters , including the laser - exciton detuning . from the effective phonon me [ eq . ( [ me2 ] ) ] and @xmath76 @xcite , we obtain the following optical bloch equations : @xmath77 where we define the polarization decay @xmath78 and the population decay @xmath79 , as well as @xmath80 . for notational convenience ,
we have also defined @xmath81 .
the incoherent spectrum can be computed from an integration of the appropriate two - time correlation function @xcite : @xmath82 where @xmath83 and @xmath84 is a geometrical factor .
the coherent spectrum can be derived in a similar way . by exploiting the quantum regression theorem and equation set ( [ eq : m])-([eq : z ] ) ,
it is possible to derive the spectrum analytically , e.g. , using laplace transform techniques .
we first define the steady - state expectation values @xmath85 , @xmath86 , and @xmath87 , and keep the explicit laser - exciton detuning dependence in the solution . defining @xmath88 , we can obtain the spectrum from @xmath89 , \label{eq : sexact}\end{aligned}\ ] ] where @xmath90 + \frac{i\omega_r}{2 } \left [ -h(0 ) + \frac{i g(0)\omega_{r}}{i\delta\omega-(\gamma_{\rm pol}-i\delta ) } \right ] } { \left[i\delta\omega-(\gamma_{\rm pol}+i\delta)\right](i\delta\omega-\gamma_{\rm pop})+\frac{{\omega_r^2 } \left[i\delta\omega-\gamma_{\rm pol } - \gamma_{\rm cd}/2\right]}{i\delta\omega-(\gamma_{\rm pol}-i\delta)}}. \label{lineshape}\end{aligned}\ ] ] the steady - state inversion and polarization components are @xmath91}{(\gamma_{\rm pol}^2 + \delta^2-\gamma_{\rm cd}^{2 } ) } = ( \braket{\sigma^-}_{ss})^*,\end{aligned}\ ] ] from which we can obtain the following steady - state values for @xmath92 , and @xmath93 : @xmath94 these equations are used with eq .
( [ lineshape ] ) to obtain @xmath95 .
we stress that the resulting spectrum is _ exact _ within the stated model assumptions .
the full - width at half - maximum ( fwhm ) of spectral resonances can be obtained from eq .
( [ lineshape ] . though these are complicated to write down analytically .
however , as we have verified , one can simply fit the analytical spectrum to a sum of lorentzian line shapes ( see discussion of fig . 4 ) and extract the broadening parameters .
+ > from the analytical spectra above , we can discern when the mollow triplets will become asymmetric and whether the detuning dependence will exhibit broadening or narrowing of the three resonances .
we define @xmath96 as the following ratio : @xmath97,\end{aligned}\ ] ] where @xmath98 .
worth to note , for off - resonant driving and @xmath99 , a completely symmetric mollow triplet is expected _ only _ if all phonon terms are neglected .
thus , phonon coupling causes an asymmetry for off - resonant driving . under systematic increase of the excitation - detuning @xmath17 ,
spectral broadening or narrowing can be achieved depending upon the value of @xmath96 . in figs .
[ thfig4a ] and [ thfig4b ] we plot the mollow triplet as a function of @xmath17 , and extract the fwhm of the sidebands for three values of @xmath96 .
as can be seen , @xmath100 ( for a suitably small @xmath56 ) leads to spectral sideband narrowing , whereas the effect of spectral broadening occurs for values @xmath101 .
in the following , we show a detailed modeling of experimentally derived results from detuning - dependent high - resolution pl measurements with the above presented theory .
the measurements are performed on a qd in a planar sample structure with negligible cavity coupling .
to reproduce the experimental results with the theoretical model we do not use fitting parameters apart from the electron - phonon coupling strength ( which is obtained by fitting the observed eid ) , but derive all characteristic values for the calculation via independent measurements or analysis . for the cut - off frequency and electron - phonon coupling strength , we use @xmath102mev and @xmath103 .
the deformation potential constant is somewhat higher compared to the value used in refs .
however , the value for @xmath4 and the dimensionless huang - rhys parameter @xmath104 ( with @xmath105 the speed of sound and @xmath106 the and electron / hole confinement length ) reported in the literature ( i.e. @xmath107 ) covers a large range and there are no well - accepted numbers to date .
additionally , @xmath108 has been shown to be enhanced in zero - dimensional qds compared to bulk material , for which different explanations are proposed , e.g. , in terms of non - adiabatic effects or the influence of defects @xcite . in the experiment
the sample temperature has been measured as @xmath109k .
the qds in the planar sample are found to have rather similar radiative lifetimes due to no purcell - like enhancement .
the radiative decay rate @xmath12 is extracted from time - correlated photon counting measurements that reveal a typical radiative lifetime of ( 750 - 850 ps ) , yielding @xmath110ev .
the rabi field @xmath111ev is derived from the mollow center - to - sideband splitting at zero laser - detuning @xmath112 .
the sum of the main phonon scattering rates @xmath113ev , which is constant in the detuning range accessible in our measurements , is calculated according to fig . 1 .
the cross - dephasing term has been extracted from the same graph as @xmath114ev . to carefully extract the pure dephasing rate @xmath56 , spectra of a power - dependent mollow triplet series of the qd under investigation at @xmath112
have been modeled with a constant @xmath56 as the only free parameter .
the extracted fwhm can be well reproduced with a pure dephasing rate of @xmath115ev ( equivalent to a pure dephasing time of 192 ps ) . a direct comparison between the extracted fwhm of the experimental data and the theoretical model is shown in fig . [ fig6 ] : the expected linear increase [ slope : @xmath116 in the fwhm with @xmath117 shows very good agreement with the experiment [ slope : @xmath118 .
ev , showing theoretical predictions versus experimental results for a system with @xmath119 .
( a ) mollow triplet spectra for increasing negative detunings , @xmath17 , the spectra are plotted with respect to the energetic laser position set to zero .
( b ) fwhm of the blue and red mollow sideband reveal distinct sideband broadening with increasing laser - detuning .
( c ) change of the relative mollow sideband area with @xmath17 .
the theoretically expected trend can be seen . ] with all parameters at hand , the experimentally measured detuning - dependent mollow triplet series is compared with the theoretical expectations in terms of sideband broadening and the change in the relative sideband areas @xmath120 . figure [ fig7](a ) shows a direct comparison of the mollow triplet spectra for increasing negative detuning @xmath121 , from which the fwhm and relative intensities are extracted .
the discrepancy between the expected and measured central mollow line intensity results from contributions of scattered laser stray - light to the true qd emission that can experimentally not be differentiated due to the equal emission frequency . for the detuning @xmath122 the spectral resolution of the high - resolution spectroscopy is not sufficient to distinguish between laser - excitation and qd rayleigh line emission .
the gray shaded peaks in fig . [ fig7](a )
( lower panel ) belong to a higher order interference of the _ fabry prot _ interferometer .
the extracted fwhm values are depicted in fig .
[ fig7](b ) . for the system under investigation
, @xmath96 is calculated to be around @xmath123 , and therefore an increase in the sidebands width is expected according to the theoretical model .
indeed , we observe a systematic increase with increasing negative detuning @xmath121 .
additionally , the relative sideband areas @xmath124 in dependence on @xmath17 are plotted in fig . [ fig7](c ) .
as becomes already visible from the mollow spectra , for positive detunings the blue sideband gains intensity whereas the red sideband area decreases , and vice versa . the crossing between relative intensities
is expected to occur at @xmath112 .
interestingly , we observe crossings at moderate negative laser - detuning values for all different qds under study .
a detailed interpretation of the physics behind this effect has to be left for further on - going in - depth analysis ( and may involve the inclusion of more excitons ) .
all detuning dependent mollow triplet series have revealed values of @xmath101 and therefore a spectral broadening of the mollow sidebands due to impurities present in the sample .
the regime of distinct sideband narrowing has also been experimentally observed by ulrich _ _ et al.__@xcite ( see fig .
3(d ) of their paper ) on a qd embedded in a micropillar cavity structure , grown by molecular beam epitaxy .
even though the effect can be qualitatively understood from the theoretical model discussed above , a direct comparison to theory requires the inclusion of qed - cavity coupling into the polaron me approach .
previous numerical studies were performed in this regime @xcite , but did not obtain this behaviour , suggesting that further work is required to explain the significant narrowing effects that are observed in the experiment .
in conclusion , we have presented a combined theoretical - experimental study on the impact of pure dephasing and phonon - induced scattering on the excitation detuning - dependence of mollow triplet sidebands .
based on a polaron master equation approach , supplemented by an analytical solution for the mollow triplet spectra , it is possible to clearly distinguish different regimes of spectral broadening or narrowing , defined by the ratio of different dephasing contributions . for the case of
experimentally observed distinct sideband broadening , we have found excellent agreement with the predictions of theory .
we have derived general formulas which are broadly applicable to reproduce resonance fluorescence spectra of single quantum dots without adopting a simplistic atomic model .
during the final preparation of this work we became aware of similar results obtained independently for off - resonant mollow sideband narrowing @xcite . we would like to thank dara mccutcheon and ahsan nazir for bringing these to our attention and for useful discussions .
s. weiler acknowledges financial support by the carl - zeiss - stiftung .
the work of s. hughes was supported by the national sciences and engineering research council of canada . | we present both experimental and theoretical investigations of a laser - driven quantum dot ( qd ) in the dressed - state regime of resonance fluorescence .
we explore the role of phonon scattering and pure dephasing on the detuning - dependence of the mollow triplet and show that the triplet sidebands may spectrally broaden or narrow with increasing detuning . based on a polaron master equation approach which includes electron - phonon interaction nonperturbatively
, we derive a fully analytical expression for the spectrum . with respect to detuning dependence
, we identify a crossover between the regimes of spectral sideband narrowing or broadening .
a comparison of the theoretical predictions to detailed experimental studies on the laser detuning - dependence of mollow triplet resonance emission from single in(ga)as qds reveals excellent agreement . 99 a. muller , e. b. flagg , p. bianucci , x. y. wang , d. g. deppe , w. ma , j. zhang , g. j. salamo , m. xiao , and c. k. shih , `` resonance fluorescence from a coherently driven semiconductor quantum dot in a cavity , '' phys . rev
. lett . * 99 * , 187402 , ( 2007 ) .
e. b. flagg , a. muller , j. w. robertson , s. founta , d. g. deppe , m. xiao , w. ma , g. j. salamo , and c. k. shih , `` resonantly driven coherent oscillations in a solid - state quantum emitter , '' nat .
phys . * 5 * 203 - 207 ( 2009 ) .
a. nick vamivakas , yong zhao , chao - yang lu , and mete atatre , `` spin - resolved quantum - dot resonance fluorescence , '' nat .
physics * 5 * , 198 - 202 ( 2009 ) .
a. kiraz , m. atatre , and a imamolu , `` quantum - dot single - photon sources : prospects for applications in linear optics quantum - information processing , '' phys .
rev .
a * 69 * , 032305 ( 2004 ) .
s. ates , s. m. ulrich , s. reitzenstein , a. lffler , a. forchel , and p. michler , `` post - selected indistinguishable photons from the resonance fluorescence of a single quantum dot in a microcavity , '' phys .
rev .
lett . * 103 * , 167402 ( 2009 ) . c. matthiesen , a. n. vamivakas , and atatre , `` subnatural linwidth single photons from a quantum dot , '' phys
. rev . lett . * 108 * , 093602 ( 2012 ) .
h s. nguyen , c voisin , p. roussignol , c. diedrichs , and g cassabois , `` ultra - coherent single photon source , '' app .
. lett . * 99 * , 261904 ( 2011 ) .
a. ulhaq , s. weiler , s. m. ulrich , r. robach , m. jetter , and p. michler , `` cascaded single - photon emission from resonantly excited quantum dots , '' nat .
photonics * 6 * , 238 ( 2012 ) . c. roy , and s. hughes , `` influence of electron - acoustic - phonon scattering on intensity power broadening in a coherently driven quantum - dot - cavity system , '' phys .
rev .
x * 1 * , 021009 ( 2011 ) . c. frstner , c. weber , j. danckwerts , and a. knorr , `` phonon - assisted damping of rabi oscillations in semiconductor quantum dots , '' phys .
rev .
lett . * 91 * , 127401 ( 2003 ) .
p. machnikowski , and l jacak , `` resonant nature of phonon - induced damping of rabi oscillations in quantum dots , '' phys .
rev .
b * 69 * , 193302 ( 2004 ) .
k. j. ahn , j. frstner , and a. knorr , `` resonance fluorescence of semiconductor quantum dots : signatures of the electron - phonon interaction , '' phys .
rev .
b * 71 * , 153309 ( 2005 ) . a. vagov , m. d. croitoru , v. m. axt , t.kuhn , and f. m. peeters , `` nonmonotonous field dependence of damping and reappearance of rabi oscillations in quantum dots , '' phys . rev . lett . * 98 * , 227403 ( 2007 ) .
a. nazir , `` photon statistics from a resonantly driven quantum dot , '' phys .
rev .
b * 78 * , 153309 , ( 2008 ) .
c roy , and s. hughes , `` phonon - dressed mollow triplet in the regime of cavity quantum electrodynamics : excitation - induced dephasing and nonperturbative cavity feeding effects , '' phys .
rev .
lett . * 106 * , 247403 ( 2011 ) . c. roy , h. kim , e. waks , and s. hughes , `` anomalous phonon - mediated damping of a driven quantum dot embedded in a high - q microcavity , '' photon nanostruct : fundam
. appl .
* 10 * 359 ( 2012 ) .
a. j. ramsay , a. v. gopal , e. m. gauger , a. nazir , b. w. lovett , a. m. fox , and m. s. skolnick , `` damping of exciton rabi rotations by acoustic phonons in optically excited ingaas / gaas quantum dots , '' phys .
rev .
lett .
* 104 * , 017402 ( 2010 ) .
s. m. ulrich , s. ates , s. reitzenstein , a. lffler , a. forchel , and p. michler , `` dephasing of mollow triplet sideband emission of a resonantly driven quantum dot in a microcavity , '' phys .
rev .
lett .
* 106 * , 247402 , ( 2011 ) .
b. r. mollow , `` power spectrum of light scattered by two - level systems , '' phys .
rev .
* 188 * , 169 - 175 ( 1969 ) .
i. wilson - rae and a. imamolu ,
quantum dot cavity - qed in the presence of strong electron - phonon interactions , phys .
rev .
b * 65 * , 235311 ( 2002 ) . d. p. s. mccutcheon and a.nazir , quantum dot rabi rotations beyond the weak exciton - phonon coupling regime , new j. phys .
* 12 * , 113042 ( 2010 ) .
g. d. mahan , _ many - particle physics _ ,
plenum , new york , 1990 .
l. besombes , k. kheng , l. marsal , and h. mariette , acoustic phonon broadening mechanism in single quantum dot emission , phys .
rev .
b * 63 * , 155307 ( 2001 ) .
p. borri , w. langbein , s. schneider , u. woggon , r. l. sellin , d. ouyang , and d. bimberg , ultralong dephasing time in ingaas quantum dots , phys .
rev .
lett .
* 87 * , 157401 ( 2001 ) .
e. a. muljarov and r. zimmermann , dephasing in quantum dots : quadratic coupling to acoustic phonons , phys .
rev .
lett .
* 93 * , 237401 ( 2004 ) .
m. bayer and a. forchel , temperature dependence of the exciton homogeneous linewidth in @xmath0 self - assembled quantum dots , phys .
rev .
b * 65 * , 041308 ( 2002 ) .
g. ortner , d. r .
yakovlev , m. bayer , s. rudin , t. l. reinecke , s. fafard , z .
wasilewski , and a. forchel , temperature dependence of the zero - phonon linewidth in inasgaas quantum dots , phys .
rev .
b * 70 * , 201301(r ) ( 2004 ) .
b. krummheuer , v. m. axt , and t. kuhn , theory of pure dephasing and the resulting absorption line shape in semiconductor quantum dots , phys .
rev .
b * 65 * , 195313 ( 2002 ) .
j. frstner , c. weber , j. danckwerts , and a. knorr , phonon - assisted damping of rabi oscillations in semiconductor quantum dots , phys .
rev .
lett .
* 91 * , 127401 ( 2003 ) .
c. roy and s. hughes , `` polaron master equation theory of the quantum - dot mollow triplet in a semiconductor cavity - qed system , '' phys rev b * 85 * , 115309 ( 2012 ) . in order to derive the phonon scattering rates , we use parameters for inas / gaas qds , which are @xmath1mev and @xmath2 , where @xmath3 is the high frequency cutoff proportional to the inverse of the typical electronic localization length in the qd and @xmath4 is a material parameter ( extracted from our experiments ) that accounts for the difference between the deformation potential constants between electrons and holes .
g. s. agarwal and r. r. puri , `` cooperative behavior of atoms irradiated by broadband squeezed light , '' phys .
rev .
a * 41 * , 3782 ( 1990 ) anders moelbjerg , per kaer , michael lorke , and jesper mrk `` resonance fluorescence from semiconductor quantum dots : beyond the mollow triplet , '' phys .
rev .
lett . * 108 * , 017401 ( 2012 ) .
d. p. s. mccutcheon , n. s. dattani , e. m. gauger , b. w. lovett , and a. nazir , `` a general approach to quantum dynamics using a variational master equation : application to phonon - damped rabi rotations in quantum dots , '' phys . rev .
b 84 , 081305(r ) ( 2011 ) .
e.g. , see m. glssl , a. vagov , s. lker , d. e. reiter , m. d. croitoru , p. machnikowski , v. m. axt , and t. kuhn , `` long - time dynamics and stationary nonequilibrium of an optically driven strongly confined quantum dot coupled to phonons , '' phys .
rev .
b * 84 * , 195311 ( 2011 ) .
e.g. , see h. j. carmichael , statistical methods in quantum optics 1 : master equations and fokker - planck equations , springer ( 2003 ) .
m. bissiri , g. baldassarri hger von hgersthal , a. s. bhatti , m. capizzi , a. frova , p. figeri , and s. franchi , optical evidence of polaron interaction in inas / gaas quantum dots , phys . rev .
b * 62 * , 46428 ( 2000 ) .
s. hughes , p. yao , f. milde , a. knorr , d. dalacu , m. mnaymneh , v. sazonova , p. j. poole , g. c. aers , j. lapointe , r. cheriton , and r. l. williams , influence of electron - acoustic phonon scattering on off - resonant cavity feeding within a strongly coupled quantum - dot cavity system , phys .
rev .
b * 83 * , 165313 ( 2011 ) .
dara p. s. mccutcheon , and ahsan nazir , `` emission properties of a driven artificial atom : increased coherent scattering and off - resonant sideband narrowing , '' arxiv:1208.4620v1 |
the top quark and the higgs boson ( together with the @xmath7 lepton ) are the only elementary standard model ( sm ) particles that remain unobserved so far in nuclear collisions .
their production cross sections in hadronic collisions are dominated by gluon - gluon fusion processes ( @xmath8 ) , computable today at the highest degree of theoretical accuracy : nnlo+nnll for the top quark @xcite , and n@xmath3lo for the h boson @xcite .
the study of their yield modifications in heavy - ion compared to pp collisions would provide novel extremely well calibrated probes of the initial and final pa and aa states .
both elementary particles , the heaviest known , have very different decay channels and lifetimes , and can thereby be used to uniquely probe various aspects of strongly - interacting matter in nuclear collisions . on the one hand
, the top - quark decays very rapidly before hadronizing ( @xmath9 fm / c , much shorter than typical @xmath10(1 fm / c ) qgp formation times ) into @xmath11 with @xmath12100% branching ratio , with the w themselves decaying either leptonically ( @xmath13 , 1/3 of the times ) or hadronically ( @xmath14 , 2/3 of the times ) .
the kinematical distributions of the charged leptons ( @xmath15 unaffected by any final - state interactions ) from @xmath16 @xmath5 decays provide accurate information @xcite on the underlying nuclear gluon distribution function in the unexplored high-@xmath17 region where `` antishadowing '' and `` emc '' effects are supposed to modify its shape compared to the free proton case @xcite . on the other hand ,
the higgs boson has a lifetime of @xmath18 50 fm / c and , once produced , traverses the produced medium and scatters with the surrounding partons , resulting in a potential depletion of its yields compared to the pp case @xcite .
the amount of higgs boson suppression can be thereby used to accurately determine the final - state density of the produced qgp .
the perspectives of t - quark @xcite and higgs boson @xcite measurements in nuclear collisions at current and future colliders are summarized here . [ cols="<,^,^,^,^,^,^,^,^,^ " , ] the theoretical cross sections and yields in ppb and pbpb are obtained with mcfm at nlo ( v.6.7 ) for top - quarks @xcite , and at nnlo ( v.8.0 ) for the higgs boson @xcite , using ct10 proton pdfs @xcite and eps09 npdfs ( including its 30 eigenvalues sets ) for the pb ion @xcite .
the following mcfm processes are run : @xmath19 for @xmath5 , @xmath20 for single - top in the @xmath21-,@xmath22-channels and associated with w @xcite ; and @xmath23 for gluon - fusion @xmath24 ( plus @xmath25 for the corresponding @xmath26 backgrounds ) , as well as @xmath27 corresponding to the rest of higgs production channels ( vector - boson - fusion , and associated with w , z and top ) to obtain the total h cross sections @xcite .
all numerical results have been obtained with the latest sm parameters @xcite , and fixing the default renormalization and factorization scales at @xmath28 for @xmath5 and single - top , @xmath29 gev for tw , and @xmath30 for higgs .
these calculations reproduce very well the top and higgs cross sections measured in pp collisions at @xmath31 = 7 , 8 , 13 tev at the lhc .
the collision energy dependence of the total top - quark and higgs boson cross sections are shown in fig .
[ fig : xsec ] .
the cross sections increase by a factor of @xmath65590 for @xmath5 and @xmath620 for the higgs boson between lhc and fcc energies .
the cross sections at the nominal lhc and fcc energies are listed in table [ tab : yields ] .
compared to the corresponding pp results at each c.m .
energy , antishadowing npdf modifications increase the total top - quark yields by 28% , whereas the higgs cross sections are just slightly enhanced ( depleted ) by @xmath123% at the lhc ( fcc ) .
the pdf uncertainties , obtained adding in quadrature the ct10 and eps09 uncertainties , are around ( below ) 10% for the top
( higgs ) case .
the @xmath32 theoretical @xmath33 scale uncertainties , not quoted either , cancel out in the ratios of ( ppb , pbpb)/(pp ) cross sections at the same c.m . energy .
top quarks decay almost exclusively to a b quark and a w boson and , in a heavy - ion environment , it is the w leptonic decays that can be best resolved from the backgrounds . the estimated measurable yields using realistic luminosities and analysis cuts ( b - jets : anti-@xmath34 algorithm with @xmath35 , @xmath36 gev / c , @xmath37(lhc , fcc ) , 50% b - tagging efficiency ; charged leptons : @xmath38 , @xmath39 gev / c , @xmath37(lhc , fcc ) , @xmath40 gev , @xmath41 gev ; neutrinos : missing transverse energy @xmath42 gev ) are listed in table [ tab : yields ] . after branching ratios ( br @xmath43 5% for @xmath5 , 22% for single - t ) and acceptance and efficiency losses ( 4050% for @xmath5 , 2030% for single - t ) , we expect about 500 and 800 leptonic top - quark events in pbpb and ppb collisions at the lhc ( with controllable backgrounds in the @xmath5 case ) . at fcc , the higher c.m .
energy and luminosities will yield very large data samples with 150- and 400-thousand pairs in pbpb and ppb .
the resulting nuclear modification ratios of @xmath5 decay leptons yields as a function of rapidity , @xmath44 and @xmath45 , are shown in fig .
[ fig : ttbar ] ( left ) for ppb at the lhc ( top ) and pbpb at the fcc ( bottom ) .
the effect of antishadowing ( shadowing ) in the npdf results in small enhancements ( deficits ) at central ( forward ) rapidities @xmath46 0 ( @xmath47 3 ) .
the impact that these pseudodata would have in an eps09 global fit of the nuclear pdfs is quantified via the hessian reweighting technique @xcite .
the expected improvements of the gluon pdf nuclear modification factor @xmath48 at @xmath49 are shown in fig .
[ fig : ttbar ] ( right ) .
the uncertainties on the nuclear gluon pdf are reduced by more than 10% ( 50% ) at the lhc ( fcc ) mostly in the antishadowing and emc @xmath17 regions @xcite .
the discovery of the higgs boson in pp collisions at the lhc was carried out in the clean diphoton and four - lepton channels with very low branching ratios but small and/or controllable backgrounds @xcite .
a first measurement in heavy - ion collisions will certainly exploit the same final states .
table [ tab : yields ] lists the estimated measurable yields for nominal luminosities after accounting for typical atlas / cms analysis cuts ( photons : @xmath50 gev / c , @xmath51(lhc , fcc ) , @xmath38 ; charged leptons : @xmath52 gev / c , @xmath51(lhc , fcc ) , @xmath38 ) . after branching ratios ( br = 0.23% for @xmath53 , 0.012% for @xmath54 ) and signal losses from acceptance and efficiency ( 4560% for diphotons , 50% for 4-leptons ) , we expect about 10 ( 1000 ) higgs bosons in pbpb and ppb collisions at the lhc ( fcc ) , on top of the corresponding @xmath53 and @xmath55 non - resonant backgrounds @xcite . in the @xmath53 case ,
the backgrounds include the irreducible qcd diphoton continuum plus 30% of events coming from misidentified @xmath56-jet and jet - jet processes .
for the nominal lhc luminosities , the significance of the diphoton signal ( s ) over the background ( b ) , computed via s/@xmath57 at the higgs peak , is 0.5,0.6@xmath58 in pbpb and ppb collisions , and thus one needs a factor of @xmath63040 larger integrated luminosities to reach @xmath59 evidence ( which would be further enhanced by including the @xmath60 results ) .
such a large @xmath61 increase would require , first , running one full - year ( instead of the nominal heavy - ion month ) plus increasing the instantaneous luminosity by a factor of 34 . whereas the feasibility of such a setup seems difficult in the pbpb case , it s not unrealistic in the ppb mode during the high - luminosity lhc phase ( hl - lhc )
. figure [ fig : higgs ] shows the expected diphoton invariant mass distributions for ppb at the lhc ( with @xmath62 pb@xmath63 , left ) and in pbpb collisions at the fcc .
the significance of the h@xmath64 peaks are 3@xmath58 and 5.5@xmath58 respectively . | the perspectives for measuring the top quark and the higgs boson in nuclear collisions at the lhc and future circular collider ( fcc ) are summarized .
perturbative qcd calculations at ( n)nlo accuracy , including nuclear parton distribution functions , are used to determine their cross sections and visible yields after standard analysis cuts in pbpb and ppb collisions at the lhc ( @xmath0 = 5.5 , 8.8 tev ) and fcc ( @xmath0 = 39 , 63 tev ) . in their `` cleanest '' decay channels , @xmath1 and @xmath2 ,
about 10@xmath3 ( @xmath4 ) top - quark and 10 ( 10@xmath3 ) higgs - boson events are expected at the lhc ( fcc ) for their total nominal integrated luminosities . whereas the @xmath5 observation is clearcut at both colliders , evidence for higgs production , perfectly possible at the fcc , requires integrating @xmath630 more luminosities at the lhc .
top quark , higgs boson , heavy - ions , lhc , fcc |
the first white dwarf known to have a dusty debris disk or cloud was giclas 29 - 38 , discovered by zuckerman & becklin ( 1987 ) to have pronounced excesses in the @xmath4 , @xmath5 , and @xmath6 bands relative to the white dwarf s photosphere .
initially , it was unclear whether the ir excess indicated a substellar companion ( zuckerman & becklin 1987 ) or particulate debris .
eventually the near - ir to mid - ir ( tokunaga , becklin , & zuckerman 1990 ; chary , zuckerman , & becklin 1999 ) spectral energy distribution ( sed ) , time - resolved photometry in the visible and near - ir that constrained the dust geometry ( graham et al .
1990 ; patterson et al .
1991 ) , and limits on the presence of companions from pulsation - timing studies ( kleinman et al .
1994 ) and speckle imaging ( kuchner et al .
1998 ) , argued in favor of particulate debris in a disk .
@xmath7 m spectroscopy ( reach et al .
2005b ) shows a strong 10 @xmath7 m emission feature in the g 29 - 38 excess that can only be caused by small silicate dust particles .
if g 29 - 38 has any planets , they are currently undetected .
while we do not have strong evidence regarding the distribution of the dust around g 29 - 38 , we do have a few constraints : the near - ir pulsation data suggest a non - spherical dust distribution ( graham et al .
1990 ; patterson et al .
1991 ) and the particles we see in emission radiate at temperatures of 290 to 890 k ( reach et al .
2005b ) . given the luminosity of the white dwarf ( 2 @xmath8 @xmath9 ) , these temperatures places the dust particles at approximately 0.15 to a few solar radii ( zuckerman & becklin 1987 ; jura 2003 ; reach et al .
2005b ) ; the outer radius depends on whether we assume the dust is in an optically thin or thick component .
g 29 - 38 is a relatively normal , h - atmosphere ( da ) wd currently passing through its instability strip , it has @xmath10 11800 k ( kepler & nelan 1993 ; bergeron , wesemael , & beauchamp 1995 ) , log(g ) = 7.9 ( koester & wilken 2006 ) to 8.14 ( bergeron et al .
1995 ) , and an implied mass = 0.56 to 0.69 m@xmath11 .
it has been cooling as a wd for @xmath30.5 gyr .
the initial - final mass relation of weidemann ( 2000 ) suggests that this range of wd masses corresponds to a zero age main sequence ( zams ) mass of 1.2 to 3.1 m@xmath11 .
g 29 - 38 was therefore once an f or a star ( drilling & landolt 2000 , table 15.8 ) .
g 29 - 38 has photospheric absorption lines from heavy elements ( ca , mg , fe ; koester , provencal , & shipman 1997 ) , which makes it a spectral type daz .
additionally , the photospheric ca ii k line varies in strength on timescales as short as 15 days ( von hippel & thompson 2007 ) , indicating episodic accretion .
becklin et al .
( 2005 ) and kilic et al .
( 2005 ) recently discovered a second white dwarf with circumstellar dust debris .
this object , gd 362 , was found to have a broad - band , @xmath12 sed that could be fit by a debris disk ( becklin et al .
its near - ir spectroscopy shows a strong @xmath4-band excess without any of the near - ir spectral features of a brown dwarf ( kilic et al .
gd 362 is an h - rich atmosphere wd with metals observed in its photosphere ; it is in fact the most metal - rich wd known .
it is slightly cooler than g 29 - 38 ( @xmath13 = 9740 k , gianninas , dufour , & bergeron 2004 ; kawka & vennes 2004 ) , yet is substantially more massive at 1.24 m@xmath11 ( gianninas et al .
2004 ) , which placed it at @xmath14 7 m@xmath11 on the main sequence ( weidemann 2000 ) .
a third wd with circumstellar dust , gd 56 , was just reported by kilic et al .
( 2006b ) .
this star was discovered via its pronounced excess at wavelengths longer than 1.6 microns in low resolution near - ir spectroscopy .
gd 56 is also a daz and has @xmath13 = 14400 k and log(g ) = 7.8 ( koester et al .
these reported atmospheric parameters correspond to a low mass wd ( 0.52 m@xmath11 ) , which is expected only for a halo star or a remnant from binary star common envelope evolution . on the other hand ,
if we assume a slightly higher surface gravity , the inferred mass becomes normal for a galactic disk wd , e.g. , for log(g ) = 7.9 , mass = 0.566 m@xmath11 ( bergeron et al .
1995 ) . in this paper
we report the fourth daz with circumstellar dust , wd2115@xmath0560 .
this wd has @xmath13 = 9700 k and log(g ) = 8.1 ( koester et al .
2005 ) , corresponding to a mass of 0.66 m@xmath11 .
we discovered this object in our _ spitzer _ irac 4.5 and 8 @xmath7 m survey of 124 wds ( mullally et al .
we conducted this survey to improve upon the ground - based @xmath4-band survey of zuckerman and becklin ( 1992 ) and farihi , becklin , & zuckerman ( 2005 ) , and the 6.75 @xmath7 m space - based , isocam survey of chary et al .
( 1999 ) of 11 wds , none of which found any new wds with circumstellar dust .
our survey also aimed to study the photospheres of cool wds ( see kilic et al .
2006a ) , which are of great importance to wd cosmochronology , as well as to search for giant planets and brown dwarfs ( see mullally et al . 2007 ) .
our discovery of the fourth wd with circumstellar dust leads us to suggest a new class of white dwarfs , the `` dazd '' white dwarfs , where `` d '' indicates circumstellar dust .
we choose a lower case `` d '' to avoid confusion with double white dwarfs , known as double degenerates and abbreviated `` dd '' .
the classification notation for wds is based on spectroscopic features , and up to now all have been notated by capital letters . yet
these features are all thought to be photospheric , whereas the dust causing the ir excess is circumstellar ; a lower case `` d '' emphasizes the difference between photospheric and circumstellar spectroscopic properties . using the photometry and discovery statistics from our survey , the detailed properties of g 29 - 38 reported by reach et al .
( 2005b ) , and some simple models , we explore the nature and cause of particulate debris in dazd wds as a group .
our _ spitzer _ survey initially targeted all wds in the mccook and sion ( 1999 , 2003 ) catalog brighter than @xmath15=15 as measured by 2mass ( skrutskie et al .
2006 ) , rejecting known binaries and planetary nebulae , yielding a total of 134 wds .
all of these wds have well - determined positions and most have well - determined proper motions , facilitating cross - referencing between catalogs .
we removed one object to avoid a conflict with the _ spitzer _ reserved observations catalog and the time allocation committee removed three other wds in common with a program focused on daz wds .
our observations for six of these wds failed to provide quality photometry , mostly because of blending problems , and in the end , we obtained good data for a sample of 124 wds with @xmath13 = 5000 to 60,000 k , cooling ages of 1 myr to 7 gyr , and representing most wd spectral types .
our observations consisted of five dithers of 30 seconds each ( 150 seconds total integration time ) for every wd in our survey .
we used the products of the _ spitzer _ science center pipeline , the basic calibrated data ( bcd ) frames and the post - bcd frames ( mosaics ) , for our analysis .
we used iraf phot as well as a custom idl package to perform aperture photometry on individual bcd frames .
we also attempted psf - fitting photometry but found that the poorly defined point response function for the irac instrument with its large pixels ( 1.213 and 1.220 arcseconds at 4.5 and 8 microns , respectively ) meant that we obtained better results with the idl and iraf aperture photometry than with psf - fitting photometry .
both the idl and iraf approaches gave identical results , within the errors . in order to maximize the signal - to - noise ratio
, we used 5 pixel apertures for bright , isolated objects , and 2 or 3 pixel apertures for faint objects or objects in crowded fields .
we corrected the resultant fluxes by the aperture correction factors determined by the irac team ( see the irac data handbook ) . for each object , we compared the photometry from 2 , 3 , and 5 pixel apertures ; we found them all to be consistent within the errors .
following the standard irac calibration procedure , we made corrections for the observed array location of each wd before averaging the fluxes of the five frames in each irac band .
we also performed photometry on the mosaic images and found the results to be consistent with the photometry from individual frames . based on the calibrations of reach et al .
( 2005a ) , we expect that our irac photometry is calibrated to an accuracy of 3% .
we estimated the photometric error from the observed scatter in the five images ( corresponding to 5 dither positions ) plus the 3% absolute calibration error , added in quadrature .
see mullally et al .
( 2007 ) for further details of our reduction procedure . our flux - limited survey of 124 single wds detected circumstellar dust around two wds , wd2115@xmath0560 and g 29 - 38 , which met the criteria for inclusion in our sample .
these two detections among the general population of white dwarfs indicate a detection rate of 1.6% in our flux limited wd sample . however , other surveys have shown higher circumstellar dust detection rates with more limited samples .
kilic et al .
( 2006b ) found the third circumstellar dust wd candidate , gd 56 , in a small survey of daz wds ; they estimated that @xmath39% of all dazs have circumstellar dust detectable at @xmath4 .
these two samples differ both in wavelength coverage and target selection .
the _ spitzer _
survey was more sensitive to cooler circumstellar dust .
the irtf survey of kilic et al .
( 2006b ) selected wd targets with high atmospheric metal content , the dazs .
the greater sensitivity to wds with circumstellar dust in the _ spitzer _ survey did not yield more such systems ; the daz survey at the irtf had a @xmath35.6 times higher yield . this result , along with the fact that our new circumstellar dust system , wd2115@xmath0560 , turned out to be a daz ( a fact we did not know when we assembled the sample ) , suggests a deep connection between detectable circumstellar dust around wds and the daz phenomenon .
table 1 summarizes the properties of wd2115@xmath0560 , along with the three other wds hosting circumstellar dust already presented in the literature .
these wds span a modest temperature range of 9700 to 14400 k , though their mass range is wide for wds , running from @xmath30.56 to @xmath31.24 m@xmath11 .
this mass and @xmath13 range mean their cooling ages range from @xmath30.2 to @xmath16 2.5 gyr .
all four display photospheric metal lines .
we list the temperature range of their circumstellar dust from a simple model fit , which we discuss below . rounding out the known properties of these wds , becklin et al .
( 2005 ) and reach et al .
( 2005b ) find that the fractional luminosity over all wavelengths in the dust is @xmath33% in gd 362 and g 29 - 38 .
the minimum dust mass around g 29 - 38 , assuming an optically - thin cloud of dust that radiates efficiently in the thermal infrared , is @xmath17 g , approximately equivalent to the mass of a 10 km asteroid .
perhaps g 29 - 38 , the first known and best studied dazd , is a good template for understanding this emerging class . using archival iue spectra , optical photometry from the literature , and _
irac , irs , and mips data , reach et al .
( 2005b ) fit an optically thin dust cloud model to the g 29 - 38 sed .
reach et al .
( 2005b ) showed that a simple model consisting of two modified blackbodies , with temperatures of 290 and 890 k , provided a good fit to the infrared excess .
their _ spitzer _ spectra revealed a strong 10 @xmath7 m emission feature , well fit by a combination of forsterite , mg@xmath18sio@xmath19 , and olivine , ( mg , fe)@xmath18sio@xmath19 .
the assumptions of an optically - thin dust cloud and blackbody dust place the detectable dust at 1 to 5 @xmath2 from the wd .
the poynting - robertson ( p - r ) timescale for particles to spiral onto the wd from a debris reservoir at 1 @xmath2 is only 4 @xmath20 years , where @xmath20 is the particle radius in microns .
the silicate emission feature observed by reach et al .
( 2005b ) can only be as strong as it is relative to the continuum ( contrast = 125% ) , if it is emitted by submicron particles .
the p - r timescale for such tiny particles is years or less .
radiation pressure can eject small grains in orbit around a star ( e.g. , wyatt & whipple 1950 ) . a spherical blackbody grain with radius @xmath21 and density @xmath22 ( in g @xmath23 ) will be ejected from bound orbit around a wd by radiation pressure if @xmath24 , where @xmath25 however , for a typical wd luminosity , @xmath26 , and wd mass , @xmath27 , the blowout size is 0.002 @xmath7 m , much smaller than the grains producing the 10 @xmath7 m feature . an optically - thin dust cloud would have a lifetime no longer than the short p - r time scale .
therefore , it seems likely that the dust cloud is optically thick , shielding much of the dust from the stellar radiation that causes p - r drag . with this idea in mind
, we fit the seds of each of the four wd debris systems using standard optically - thick flat disk models ( friedjung 1985 ; jura 2003 ) .
we derived the input spectra for these simple disk models by fitting blackbody curves to the photospheric portions of the seds , and then used the models to derive disk inclinations and dust temperatures that correspond to the inner and outer radii of the disks .
for each wd , we created a grid of models with inner disk temperatures @xmath28 = 600 to 1250 k , outer disk temperatures @xmath29 = 200 to 1175 k , @xmath30 , and a range of inclination angles .
we then found the model that best fit the ir photometry of each wd ( j band and beyond ) by minimizing @xmath31 .
table 1 lists the temperature ranges and inclinations of the best fit models .
for gd 56 and gd 362 we included the irtf spectroscopy of kilic et al .
( 2005 , 2006b ) in the data to be fit .
for g 29 - 38 we performed one fit using our _ spitzer _ photometry and another fit using the irtf ( tokunaga et al .
1999 ) plus iso photometry ( chary et al .
1999 ) , though without the 10.5 @xmath7 m photometry , which is likely to be influenced by the strong silicate emission feature .
the fit to the older irtf and iso photometry was poorer than the fit to the _ spitzer _ data , and we report only the fit to the modern data in table 1 .
the difference between the _ spitzer _ and older data sets most likely reflects calibration errors that plagued the ground - based and particularly the iso photometry ( r. chary and b. zuckerman , private communication ) .
it is also possible that the g 29 - 38 debris disk has evolved substantially over the 15 year span of these observations .
figures 14 present our best fit debris disk models .
all four dazd wds show a pronounced ir excess relative to the expected wd photospheres , represented by the wd model atmospheres ( solid lines , kindly provided by detlev koester ) the debris disk components ( dashed lines ) fit the near - ir and mid - ir excess well . table 1 also lists the inner and outer radii of the best fit flat disk models .
they range from 0.15 to 0.58 @xmath2 = 13 to 49 white dwarf radii for the three normal mass wds , and 0.08 to 0.50 @xmath2 = 17 to 106 white dwarf radii for gd 362 , assuming the high mass value currently in the literature , which implies a much smaller radius .
the disk model has no explicit inner edge directly exposed to radiation from the wd , and thus the inner radii are significantly underestimated .
perhaps the observations of daz disks all point to a common simple picture originally suggested by jura ( 2003 ) ; the disks are analogs of planetary rings , physically thin disks of rocks and dust with optical depth @xmath32 1 .
our new data make this analogy much more compelling .
as shown above , the physically thin , optically - thick disk geometry implied by this picture can match the observed seds of dazd wds .
besides providing an excuse for that disk geometry , the planetary - ring analogy has the following virtues . \1 .
this picture naturally incorporates both small dust grains and larger , long - lived source particles .
standard models of planetary rings ( e.g. , cuzzi et al . 1979 )
contain a distribution of particle sizes .
the largest ring particles contain most of a ring s mass .
these particles occupy a `` monolayer '' a few particle - widths thick , while the smaller particles like the ones observed to produce the 10-micron emission feature in g29 - 38 ( reach et al .
2005b)are spread over a thicker poly - layer .
planetary rings may also consist of dusty rings interspersed with rings of larger bodies ( e.g. , de pater et al .
the small particles are continually produced by the collisions of the larger particles , and protected to some degree from radiation forces by the optical thickness of the ring .
this picture provides a disk lifetime consistent with the ages of the dazd wds .
an optically - thick disk of rocks has a lifetime set by the viscous spreading time of the ring , @xmath33 , where @xmath34 is the radius of the ring and @xmath35 is the kinematic viscosity . to estimate the disk lifetime ,
let us assume that the viscosity of the wd rings is similar to that of saturn s rings .
the viscosity of a monolayer ring is @xmath36 , where @xmath37 is the optical depth and @xmath38 is a representative particle size , weighted toward the larger particles ( goldreich & tremaine 1982 ) . the angular speed in the wd ring is similar to that in planetary rings : @xmath39 s@xmath40 . so assuming a similar viscosity for the wd debris disks and saturn s rings is like assuming a similar optical depth and particle size distribution . with this assumption ,
the wd ring lifetimes are @xmath3100 times longer than the lifetime of saturn s rings because of their larger radii .
this lifetime is easily @xmath41 years , comparable to the median post - main sequence lifetime of the wds in our survey . in this picture , wd rings do not require continual replenishment from a longer - lived source of small bodies .
this picture implies a metal accretion rate consistent with the photospheric abundances of metals in the dazd wds . in the context of this picture
, we can estimate the mass of a typical dazd disk as @xmath42 , using the surface density of saturn s rings ( @xmath43 g @xmath44 ; tiscareno et al .
2007 ) and an outer radius @xmath45 ( see item 5 , below ) . with these assumptions ,
we find a representative disk mass of @xmath32 @xmath46 .
if the accretion time is @xmath47 years , the resulting accretion rate onto the wd is @xmath37 @xmath48 m@xmath11 yr@xmath40 .
figure 5 shows the observed photospheric abundances derived by koester & wilken ( 2006 ) for a large sample of daz wds .
figure 6 shows accretion rates for this sample inferred from these photospheric abundances and the atmospheric residence times koester & wilken ( 2006 ) derived for these wds . these accretion rates from koester & wilken ( 2006 ) are based on an assumption that the accreted material has solar composition .
our model assumes that the material is mostly refractory , so to compare our accretion rate to figure 6 , we need to multiply it by a factor of 50 , giving us @xmath33 @xmath49 m@xmath11 yr@xmath40 .
this accretion rate sits right in the middle of the range of daz inferred accretion rates , and is only about an order of magnitude less than that of the known dazds .
it is easy to imagine a disk with slightly higher surface density or viscosity than saturn s rings that would precisely match the dazd accretion rates in figure 6 .
this picture explains the inner radii of dazd disks inferred from their seds . the radius where the equilibrium temperature of a dust grain exceeds the sublimation temperature , @xmath50 , for the grains should mark the inner boundary of the ring .
although most of the disk sees only stellar radiation at an oblique angle , the inner radius of the disk can see direct stellar radiation , so it should be hotter than the flat - disk model would predict , closer to the temperature of an optically - thin dust cloud , or a single grain in free space . with this geometry in mind , we find that for blackbody grains , the sublimation radius is @xmath51 where @xmath52 is the wd luminosity . for a typical wd luminosity , @xmath53 , and assuming a silicate sublimation temperature , @xmath54 k , equation 2 gives @xmath55 .
another component of the disks may be grains too small to efficiently emit at the blackbody peak wavelength .
these grains are slightly hotter than blackbody grains at the same location in the nebula , so they sublimate at larger radii .
e.g. , for the `` medium''-sized grains described by backman & paresce ( 1993 ) , the sublimation radius is @xmath56 where @xmath57 is the grain size measured in microns . for the same assumed wd luminosity and silicate sublimation temperature , and for a grain size of 1 @xmath7 m , @xmath58 .
table 1 , column 12 lists dust sublimation radii for each of the dazds calculated from equation 2 assuming @xmath59 k. this assumption produces a good match to the inner radii inferred from our models of the seds .
however , we did not incorporate the physics of direct heating of the inner rim of the disk into our sed modeling , so we can not compare sublimation radii calculated using the assumption of direct heating to the radii inferred from our flat disk models .
we can , however , compare the inner temperatures inferred from our flat disk models to the sublimation temperatures of typical dust materials and retain self - consistency .
table 1 , column 9 lists the temperatures of the inner and outer edges of the disks inferred from the flat disk models . the temperatures at the inner edge range from 8001150 k. \5 .
this picture explains the outer radii of dazd disks inferred from their seds .
a dynamically cold ring can not extend much beyond the roche radius ; beyond this radius , the ring particles coagulate into moons or in the case of the white dwarf rings asteroids or planets .
the roche radius for tidal disruption of an asteroid near a wd is given by jura ( 2003 ) : @xmath60 where @xmath61 is a parameter of order unity that depends on the asteroid s composition , orbit , and rotation , @xmath62 is the radius of the wd , @xmath63 is the density of the wd , and @xmath64 is the density of the asteroid . for g 29 - 38 , @xmath65 .
since this equation only includes density to the one - third power and wd masses span a narrow range ( @xmath30.55 to 1.2 m@xmath11 ) , all wds have about the same roche radii .
jura ( 2003 ) pointed out that the outer radius of the disk around g29 - 38 roughly matched the roche radius for this wd .
we find that the outer radii of all the dazd disks as inferred from physically - thin , optically thick disk models for the seds are consistent with the wd roche radii .
this picture may explain the absence of dust around luminous daz white dwarfs .
this _ spitzer _ survey and the survey of kilic et al .
( 2006b ) examined eleven daz white dwarfs hotter than 15,000 k and found circumstellar dust around none of them ( see figures 5 and 6 ) .
the high surface abundances and very short atmospheric residence times for calcium ( see 4 ) in these eleven stars made them excellent candidates to host observable debris disks .
gnsicke et al .
( 2006 ) discovered a single wd with effective temperature @xmath322,000 k that hosts a calcium gas disk , but shows no sign of circumstellar dust .
these observations suggest that perhaps hotter , more luminous wds can not host dust .
the analogy between wd disks and planetary rings suggests a natural interpretation of these observations ; for hot wds , the sublimation zone reaches beyond the roche radius .
these wds , like main sequence stars , lack a circumstellar region where small bodies are stable against both sublimation and coagulation in dynamically - cold disks .
we can find the critical wd effective temperature above which this stable region disappears by setting @xmath66 , assuming blackbody dust grains : @xmath67 ignoring the slow dependence of this expression on the wd density , we find the critical wd temperature is @xmath68 . if instead we use the 1 @xmath7 m `` medium '' grains of backman & paresce ( 1993 ) , we find that the maximum wd temperature for hosting a dust disk is @xmath69 so @xmath70 . for grains with @xmath71
k , we find @xmath72 k , and @xmath73 k. we suggested above that the inner edges of the observed dazd disks match blackbody grains with @xmath74 k. this combination implies that there should be no disks around wds hotter than @xmath75 k , keeping in mind the caveats described above .
the observations described above suggest that the critical temperature should be roughly 15,00022,000 k. these observations broadly agree with the hypothesis that the grain sublimation temperature and the roche radius together limit the kinds of wds that can host dust disks , although clearly details remain to be better understood .
as we mentioned above , the presence of photospheric metals in dazs implies ongoing accretion of these metals . our physical model for the dazd disks provides a natural explanation for this accretion .
let us examine in more detail the connection between dazd disks and the daz phenomenon .
white dwarfs are classified as daz or dbz if their optical spectra contain metal lines along with much stronger hydrogen or helium lines , respectively .
white dwarfs showing metal lines may also be classified simply dz if they are too cool to show hydrogen or helium lines .
hybrid types , such as dabz , also exist . since the discovery of metal lines in wd atmospheres ( see weidemann 1958 ; greenstein 1960 ) , the origin of these metals has been debated ( e.g. , lacombe et al . 1983 ; aannestad et al . 1993 ; dupuis , fontaine , & wesemael 1993 ; zuckerman & reid 1998 ; zuckerman et al .
2003 ; koester & wilken 2006 ; kilic & redfield 2007 ) . the high surface gravity of wds should cause the denser atoms of the metals to sink in the atmosphere on timescales of days to millions of years , depending on the atmospheric constituents , effective temperature , and wd mass .
these timescales are all much shorter than the ages of most wds .
figure 7 shows the timescale for gravitational settling for da wds with log(g ) = 7.75 to 9.0 , roughly mass = 0.48 to 1.20 m@xmath11 , taken from koester & wilken ( 2006 ) . for comparison ,
the four dazd stars are plotted on this figure based on their atmospheric @xmath13 and log(g ) values .
figure 7 shows that for the wds hotter than 11,000 k , the settling times are substantially less than a year and even for our coolest , low surface gravity object , wd2115@xmath0560 , the settling time is still less than 200 years .
these short settling timescales imply that daz stars were very recently accreting heavy elements .
interstellar accretion , perhaps the most discussed scenario for supplying the heavy elements , seems to fail by orders of magnitude for many of the known daz wds ( aannestad et al .
1993 ; although see the counter - argument by koester & wilken 2006 )
. however , such accretion would be natural if daz wds typically harbor substantial circumstellar disks .
so far all debris wd systems with circumstellar disks have detectable photospheric metal lines .
is it possible that all metal - rich wds ( daz / dbz / dz ) accrete their metals via debris disks ?
one reason answering this question is difficult is that not all wds with photospheric metals can be recognized as daz / dbz / dzs .
this point is made in figure 5 , where the curve indicating a constant ca equivalent width of 15 m corresponds to much higher ca abundances for hotter wds than for cooler wds .
it is much more difficult to recognize a daz at higher stellar temperatures .
figure 6 shows another view of the same data , easier to grasp visually , but more model - dependent .
the maximum accretion rates necessary to explain the observed abundances are essentially independent of @xmath13 ; the short atmospheric residence timescales for the hot wds are balanced by the deeper convective reservoirs for the cooler wds .
the dazd wds are located near the top of figures 5 and 6 ; they are among the wds with the highest photospheric heavy element abundances and the highest inferred accretion rates , and therefore should harbor the most massive and easiest to detect of the debris disks .
it thus seems possible that the debris disks found to date display an observational selection effect , and that the majority of dazs harbor debris disks . for all dazd
wds except the possibly massive gd 362 , koester & wilken ( 2006 ) find accretion rates in the narrow range of 4.078.13 @xmath76 m@xmath11 yr@xmath40 .
koester & wilken ( 2006 ) assumed that the accreted material had solar abundances .
we assume instead that the accreted material lacks hydrogen and helium , reducing the total accretion rate by a factor of 50 .
this lower limit corresponds to 8 to 16 @xmath77 m@xmath11 yr@xmath40 or 1.6 to 3.2 @xmath78 g yr@xmath40 of refractory elements fed to the wd , a rate consistent with viscous accretion in our planetary - ring model .
if this accretion persists for at least 25% of the lifetime of the wds , commensurate with the fraction of das that are dazs ( zuckerman et al .
2003 ) , it corresponds to an accreted mass of @xmath79 of an earth mass accreted over 1 gyr .
we have discussed the basic properties of the debris systems found to date and we have connected the debris observable in the infrared to the photospheric metal lines observable in the optical . we have argued that dust debris is common around wds and that this dust must be steadily accreting onto the wd .
we have not yet examined possible sources for the debris around wds . here
we consider possible origins : wd mergers , debris left during the asymptotic giant branch ( agb ) or proto - planetary nebulae ( ppne ) phases , and accretion of dust derived from planetary system bodies .
merging wds are expected to leave behind disks ( livio , pringle , & wood 2005 ; hansen , kulkarni , & wiktorowicz 2006 ) to shed angular momentum . according to the calculations of livio et al .
( 2005 ) , such disks should be massive ( @xmath30.007 m@xmath11 ) , extend to 1 au or more , and predominantly composed of carbon and oxygen , assuming the merger of two wds each with mass @xmath80 0.7 m@xmath11 .
the minimum mass for most of these merged wds ought to be @xmath81 1.1 m@xmath11 , since it takes about 10 gyr to produce a 0.55 m@xmath11 wd . except for gd 362 , with a mass possibly equal to 1.24 m@xmath11 , these mass constraints rule out this explanation for the majority of dazd wds .
the other wds with circumstellar dust have masses ranging from 0.52 to 0.69 m@xmath11 .
additionally , the disk mass and radial extent appear inconsistent with all dazd stars discovered to date . while wd mergers must be taking place , and they may possibly be the source of the disk around gd 362 , they are not the source of the majority of the observed dazd systems . the outflows of late - stage asymptotic giant branch ( agb ) stars and proto - planetary nebulae ( ppne ) are often highly non - spherical
. this process is not understood , and a range of models have been proposed for the origin of the asymmetries , including stellar companions and magnetic fields ( see balick & adams 2002 , and references therein ) .
there is circumstantial evidence for the formation of disks and tori during the agb to post - agb star transition , albeit at much greater distances from the host star , and most ppne are producing or have recently produced dust particles ( balick & adams 2002 ) .
in particular , waters et al .
( 1998 ) observed olivines and pyroxenes around a pre - white dwarf .
while most of the outflows are ejected well above the stars escape velocities , at present it is not known whether some small fraction of these outflows remain in orbit after they are ejected .
remaining stellar material is a possible explanation for the dust disk discovered by su et al .
( 2007 ) around ( at 35150 au ) the helix planetary nebula . while su et al .
attribute the observed dust to disrupted kuiper belt objects or disrupted comets , it could also be the result of stellar material ejected with less than the escape speed .
this scenario predicts dusty environments for the youngest wds with less and less dust as wds age .
the observations , however , run counter to this .
all four of the known dazd systems have old host wds with ages of @xmath16 0.2 to @xmath16 2.5 gyr and no hot , dusty wds have been discovered .
the metal disk wd found by gnsicke et al .
( 2006 ) pushes this to slightly younger objects , @xmath30.1 gyr , but still , if the late stages of stellar evolution commonly leave debris disks at a range of radii , these debris disk should be detectable in our survey and in iue observations of hot wds . yet , to date , no such systems have been seen ( hansen et al .
2006 ; kilic et al .
2006b ) .
the dust around dazds may represent the remains of planetary systems .
roughly 5 - 10% of a - type stars have substantial debris disks ( e.g. , backman & paresce 1993 ) and precise doppler surveys find giant planets around roughly 5 - 10% of solar - type stars ( e.g. , fischer & valenti 2003 ) .
perhaps 5 - 10% or more of wd progenitors had planetary systems , and some small bodies in these systems survived the post - main sequence evolution of their host stars ( debes & siguardsson 2002 , jura 2003 ) .
our observations and models appear to be consistent with this picture for the origin of all the dazd disks .
we presented the properties of wd2115@xmath0560 , a new wd with infrared excess we discovered in our _ spitzer _ irac photometry , and the fourth daz seen to have an infrared excess well - described by circumstellar dust debris .
we examined the dusty daz wds as a class , which we refer to with the letters `` dazd '' . using a simple , flat , optically - thick disk model motivated by the geometry of saturn s rings , we find that the dust in these four dazd wds reside in disks ranging in inner temperature from @xmath3800 to 1150 k and outer temperature from @xmath3200 to 725 k , not taking into account the direct heating of the inner edges of the disks .
the outer disk temperatures are likely to be lower limits set by the sensitivity of the observations and the amount of dust radiating at that temperature .
these dust temperatures imply that the inner disk edges are at @xmath1 0.1 to 0.2 @xmath2 and the detectable extent of the outer disks are @xmath30.3 to 0.6 @xmath2 .
the close circumstellar locations of these debris disks are consistent with simple sublimation calculations and with the roche radius for tidal disruption of an asteroid near a wd ( jura 2003 ) .
we find that , taken together , the dazd wds point to a disk model analogous to planetary rings .
this model naturally explains the disk accretion rates and lifetimes and the temperature distribution of the dusty wds .
we favor the interpretation that most of these disks represent planetary systems remnants .
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 award project nbr 1269551 issued by jpl / caltech to the university of texas .
kilic , m. , von hippel , t. , leggett , s. k. , & winget , d. e. 2005 , , 632 , l115 kilic , m. , von hippel , t. , leggett , s. k. , & winget , d. e. 2006b , , 646 , 474 kilic , m. , von hippel , t. , mullally , f. , reach , w. t. , kuchner , m. j. , winget , d. e. , burrows , a. , & saumon , d. 2006a , , in press kleinman , s. j. et al .
1994 , , 436 , 875 reach , w. t. , et al . 2005a , , 117 , 978 reach , w. t. , kuchner , m. j. , von hippel , t. , burrows , a. , mullally , f. , kilic , m. , & winget , d. e. 2005b , , 635 , l161 tiscareno , m. s. , burns , j. a. , nicholson , p. d. , hedman , m. & porco , c. c. 2007 , icarus , in press , ( astro - ph/0610242 ) lcrccccccccc gd 56 & dazd & 14,400 & 0.520.57 & 0.91.4 & 0.20.4 & 4.2% & 67 & 1125525 & 0.210.58 & 45 & 0.36 + g 29 - 38 & dazd & 11,600 & 0.560.69 & 1.23.1 & 0.5 & 3.5% & 14 & 1150725 & 0.150.28 & 45 & 0.18 + gd 362 & dazd & 9740 & 1.24 & @xmath147.0 & @xmath142.5 & 6.1% & 24 & 800200 & 0.080.50 & 60 & 0.07 + wd2115@xmath0560 & dazd & 9700 & 0.66 & 2.8 & 0.9 & 0.9% & 22 & 900550 & 0.170.32 & 80 & 0.16 + | our mid - infrared survey of 124 white dwarfs with the _ spitzer _ space telescope and the irac imager has revealed an infrared excess associated with the white dwarf wd 2115@xmath0560 naturally explained by circumstellar dust .
this object is the fourth white dwarf observed to have circumstellar dust .
all four are daz white dwarfs , i.e. , they have both photospheric balmer lines and photospheric metal lines .
we discuss these four objects as a class , which we abbreviate `` dazd '' , where the `` d '' stands for `` dust '' . using an optically - thick , geometrically - thin disk model analogous to saturn s rings
, we find that the inner disk edges are at @xmath1 0.1 to 0.2 @xmath2 and that the outer disk edges are @xmath30.3 to 0.6 @xmath2 .
this model naturally explains the accretion rates and lifetimes of the detected wd disks and the accretion rates inferred from photospheric metal abundances . |
coherent actions of apparently distinct physical systems often provoke questions of their possible interactions .
such coherence in interacting systems is often a result of their synchronization @xcite .
it became a popular topic with the discovery of synchronization of non - identical chaotic oscillators @xcite .
over the years different types of synchrony were studied , notably phase synchronization @xcite .
there were also numerous attempts to study more complicated interactions under the names of generalized synchronization or interdependence @xcite .
in biological context synchronization is expected to play a major role in cognitive processes in the brain @xcite such as visual binding @xcite and large - scale integration @xcite .
various synchronization measures were successfully applied to electrophysiological signals @xcite . in this work
we concentrate on nonlinear interdependence @xcite . for an experimentalist it is often interesting to know how two systems synchronize during short periods of evoked activity @xcite .
such questions arise naturally in analysing data from animal experiments @xcite .
one measures there electrical activity on different levels of sensory information processing and aims at relating changes in synchrony to the behavioral contex , such as attention or arousal
. it may be the case that the stationary dynamics ( with no sensory stimulation ) corresponds to a fixed point .
for instance , when one measures the activity in the barrel cortex of a restrained and habituated rat , the recorded signals seem to be noise @xcite . on
the other hand transient activity evoked by specific stimuli seems to provide useful information .
for example , bending a bunch of whiskers triggers non - trivial patterns of activity ( evoked potentials , eps ) in both the somatosensory thalamic nuclei and the barrel cortex @xcite .
explorations described in this paper aim at solving the following problem .
suppose we have two pairs of transient signals , for example recordings of evoked potentials from thalamus and cerebral cortex in two behavioral situations @xcite .
can we tell in which of the two situations the strength of coupling between the structures is higher ?
thus we investigate if one can measure differences in the strength of coupling between two structures using nonlinear interdependence measures on an ensemble of eps .
since eps are short , transient signals , straightforward application of the measures motivated by studies of systems moving on the attractors ( stationary dynamics ) is rather doubtful and a more sophisticated treatment is needed @xcite .
our approach is similar in spirit to that advocated by janosi and tel for the reconstruction of chaotic saddles from transient time series @xcite .
( note that the transients we study should not be confused with the transient chaos studied by janosi and tel . )
thus we cut pieces of the recordings corresponding to well - localized eps and paste them together one after another .
since we are interested in the coupled systems , unlike janosi and tel , we obtain two artificial time - series to which we then apply nonlinear interdependence measures and linear correlations .
it turns out that this approach allows to extract the information about the strength of the coupling between the two systems .
we test our method on a population model of information processing in structure of the model of the thalamocortical loop used in the simulations.,scaledwidth=35.0% ] thalamocortical loop ( figure [ schfig ] ) consisting of two coupled wilson - cowan structures @xcite .
sensory information is relayed through thalamic nuclei to cortical fields , which in return send feedback connections to the thalamus . this basic framework of the early stages of sensory systems is to a large extent universal across different species and modalities @xcite . to check that the results are not specific to this particular system we also study evoked dynamics of two coupled rssler - type oscillators in non - chaotic regime .
the paper is organized as follows . in sec .
[ sec : measures ] we define the measures to be used . in sec .
[ sec : models ] we describe the models used to test our method .
our model of thalamocortical loop is discussed in sec .
[ sec : models1 ] and a system of two coupled rssler - type oscillators is described in sec . [ sec : models2 ] . in sec .
[ sec : results ] we present the results . in sec .
[ sec : results1 ] we show how various interdependence measures calculated on the transients are related to the coupling between the systems , while in sec .
[ sec : results2 ] we study how the resolution of our methods degrades with noise .
finally , in sec .
[ sec : results3 ] , we apply time - resolved interdependence measure @xmath0 @xcite and compare its utility with our approach . we summarize our observations in sec .
[ sec : concl ] .
in the present paper we mainly study the applicability of nonlinear interdependence measures on the transients .
these measures , proposed in @xcite , are non - symmetric and therefore can provide information about the direction of driving , even if the interpretation in terms of causal relations is not straightforward @xcite .
these measures are constructed as follows .
we start with two time series @xmath1 and @xmath2 , @xmath3 , measured in systems @xmath4 and @xmath5 .
we then construct @xmath6-dimensional delay - vector embeddings @xcite @xmath7 , similarly for @xmath8 , where @xmath9 is the time lag .
the information about the synchrony is inferred from comparing the size of a neighborhood of a point in @xmath6-dimensional space in one subsystem to the spread of its equal - time counterpart in the other subsystem .
the idea behind it is that if the systems are highly interdependent then the partners of close neighbors in one system should be close in the other system .
several different measures exploring this idea can be considered depending on how one measures the size of the neighborhood .
these variants include measures denoted by @xmath10 , @xmath11 @xcite , @xmath12 @xcite , @xmath13 @xcite .
we have studied the properties of most of these measures but for the sake of clarity here we report only the results for the `` robust '' variant @xmath11 and a normalized measure @xmath12 , as they proved most useful for our purposes .
let us , following @xcite , for each @xmath14 define a measure of the spread of its neighborhood equal to the mean squared euclidean distance : @xmath15 where @xmath16 are the time indices of the @xmath17 nearest neighbors of @xmath14 , analogously , @xmath18 denotes the time indices of the @xmath17 nearest neighbors of @xmath8 .
to avoid problems related to temporal correlations @xcite , points closer in time to the current point @xmath14 than a certain threshold are typically excluded from the nearest - neighbor search ( theiler correction ) .
then we define the @xmath19-conditioned mean @xmath20 where the indices @xmath16 of the nearest neighbors of @xmath14 are replaced with the indices @xmath18 of the nearest neighbors of @xmath8 .
the definitions of @xmath21 and @xmath22 are analogous .
the measures @xmath11 and @xmath12 use the mean squared distance to random points : @xmath23 and are defined as @xmath24 the interdependencies in the other direction @xmath25 , @xmath26 are defined analogously and need not be equal @xmath27 , @xmath28 .
such measures base on repetitiveness of the dynamics : one expects that if the system moves on the attractor the observed trajectory visits neigborhoods of every point many times given sufficiently long recording .
the same holds for the reconstructed dynamics .
however , if the stationary part of the signal is short or missing , especially if we observe a transient such as evoked potential , this is not the case .
still , if we have noisy dynamics , every repetition of the experiment leads to a slightly different probing of the neighborhood of the noise - free trajectory .
this observation led us to an idea of gluing a number of repetitions of the same evoked activity ( with different noise realizations ) together and using such pseudo - periodic signals as we would use trajectories on a chaotic attractor .
a similar idea was used by janosi and tel in a different context for a different purpose @xcite .
an example of a delay embedding of a signal obtained this way is presented in fig .
[ nbsfig ] .
note that artifacts may emerge at the gluing points .
this is discussed in @xcite , and some countermeasures are proposed . for simplicity
we proceed with just gluing as we expect that the artifacts only increase the effective noise level .
the influence of noise is studied in sec .
[ sec : results2 ] .
delay - vector embeddings ( shown in planes defined by the first two principal components ) of pseudo - periodic signals obtained by gluing 50 evoked potentials generated in a model of thalamocortical loop . on the left ( signal from `` thalamus '' ) a point is chosen ( black square ) and its 15 nearest neighbors are marked with red ( gray ) diamonds .
on the right ( `` cortex '' ) the equal - time partners of the marked points from the left picture are shown .
, scaledwidth=47.0% ] recently , time - resolved variants of the methods described above were studied @xcite . they are applied to ensembles of simultaneous recordings , each consisting of many different realizations of the same ( presumably short ) process .
let us denote the @xmath29-th state vector in @xmath30-th realization of the time - series by @xmath31 ( @xmath32 , respectively ) , @xmath33 .
the idea in @xcite is , for given @xmath31 to find one neighbor in each of the ensembles .
then a measure ( denoted @xmath34 ) based on distances to these neighbors is constructed .
the proposition of @xcite is to look not at the nearest neighbors of a given @xmath14 no matter what time they occur at , but rather at the spread of state - vectors at the same latency across the ensemble . in sec .
[ sec : results3 ] we study the measure @xmath0 as defined in @xcite .
let @xmath35 denote the ensemble index of the @xmath36-th nearest neighboor of @xmath37 among the whole ensemble @xmath38 .
define the quantities @xmath39 the time - resolved interdependence measure is further defined as @xmath40 analogously one can define @xmath41 and also time - resolved variants of other interdependence measures . in the numerical experiments described in this paper we use the following parameters of the nonlinear interdependence measures : time lag for construction of delay - vectors : @xmath42 , embedding dimension @xmath43 , number of nearest neighbors @xmath44 , theiler correction @xmath45 . to calculate the interdependencies we used the code by rodrigo quian quiroga and chee seng koh available at http://www.vis.caltech.edu/~rodri/synchro/synchro_home.htm . in case of the measure @xmath0 we use the same embedding dimension and time lag ; here @xmath46 . to calculate this measure we used the code provided in supplementary material to @xcite . to compare the linear and nonlinear analysis methods we calculated the cross - correlation coefficients using matlab .
while in numerical studies the correctness of reconstruction can often be easily checked by comparison with original dynamics , in analysis of experimental data it can be a complex issue .
correct reconstruction is a prerequisite for application of our technique . for technical details on best practices of delay embedding reconstructions , pitfalls and caveats ,
see @xcite .
our model of the thalamocortical loop was based on the wilson and cowan mean - field description of interacting populations of excitatory and inhibitory neural cells @xcite . in the simplest version , which we used ,
each population is described by a single variable standing for its mean level of activity @xmath47 the variables @xmath48 and @xmath49 are the mean activities of excitatory and inhibitory populations , respectively , and form the phase space of a localized neuronal aggregate .
the symbols @xmath9 , @xmath17 , @xmath50 , @xmath51 denote parameters of the model , @xmath52 are sigmoidal functions , @xmath53 and @xmath54 are input signals to excitatory and inhibitory populations , respectively .
these equations take into account the absolute refractory period of neurons which is a short period after activation in which a cell can not be activated again .
such models exhibit a number of different behaviors ( stable points , hysteresis , limit cycles ) depending on the exact choice of parameters @xcite . to relate the simulation results to the experiment @xcite we considered the observable @xmath55 , since the electric potential measured in experiments is related to the difference between excitatory and inhibitory postsynaptic potentials
( see the discussion in @xcite ) .
we studied a model composed of two such mutually connected aggregates , which we call `` thalamus '' and `` cortex '' ( figure [ schfig ] ) .
note that the parameters characterizing the two parts are different ( see the appendix [ appa ] for a complete specification of the model ) . specifically , there are no excitatory - excitatory nor inhibitory - inhibitory connections in the thalamus .
only the thalamus receives sensory input , and we assume that @xmath54 is always a constant fraction of @xmath53 .
the connections between two subsystems are excitatory only . to model the stimulus we assumed that the input ( @xmath56 ) switches at some point from 0 to a constant value ( @xmath57 ) , and after a short time ( on the time - scale of relaxation to the fixed point ) switches back to zero .
this is clearly another simplification , as the real input , which could be induced by bending a bunch of whiskers @xcite , would be a more complex function of time .
however , the transient nature of the stimulus is preserved . in this simple
setting we can understand that the `` evoked potential '' corresponds to a trajectory approaching the asymptotic solution of the `` excited '' system ( with the non - zero input @xmath57 ) , followed by a relaxation to the `` spontaneous activity '' in the system with null input .
the model parameters were chosen so that its response to brief stimulation were damped oscillations of @xmath58 both in the thalamus and the cortex similar to those observed in the experiments , both in terms of shape and time duration @xcite ( figure [ epfig ] ) .
however , apart from that , we exercised little effort to match the response of the model to the actual activity of somatosensory tract in the rat brain .
our main goal in the present work was establishing a method of inferring coupling strength from transients and not a study of the rat somatosensory system .
for this reason it was convenient to use a very simplified , qualitative model .
interestingly , the response of the model , measured for example as the activity of excitatory cells in the thalamus , extends in time well beyond the end of the stimulation ( figure [ epfig ] ) .
such behavior is not observed in a single aggregate and requires at least two interconnected structures @xcite
. `` evoked potentials '' ( @xmath59 ) , ( a ) , ( b ) and their delay - vector embeddings shown in a plane defined by the first two principal components ( c ) , ( d ) .
plots ( a ) and ( c ) : thalamus , ( b ) and ( d ) : cortex . the intervals above the ep indicate the duration of the non - zero stimulus .
black ( thick ) lines are solutions for the system without noise , blue ( thin ) curves are five different realisations of noisy dynamics.,scaledwidth=47.0% ] we performed numerical simulations in three modes : either stationary ( null or constant input ) , or not ( transient input ) .
the dynamics of the model is presented in figure [ dynfig ] . in case of transient input
the simulation was done for @xmath60ms .
we used the stimulus @xmath53 and @xmath54 which was 0 except for the time @xmath61 when it was @xmath62 and @xmath63 .
the system settled in the stationary state during the initial segment ( @xmath64 ) which was discarded from the analysis .
the noise was simulated as additional input to each of the four populations , see the appendix [ appa ] for the equations . for each population we used different gaussian ( mean @xmath65 , standard deviation @xmath66 )
white noise , sampled at 1khz and interpolated linearly to obtain values for intermediate time points . in case of stationary dynamics we simulated longer periods , @xmath67ms .
the signals were sampled at 100hz before the synchronization measures were applied . in case of constant or null stimulation
the system approaches one of the two fixed - point solutions which are marked by large dots in figure [ dynfig ] . for the amount of noise used here the dynamics of the system changes as expected : the fixed points become diffused clouds ( figure [ dynfig ] ) . during the transient
`` evoked potential '' the switching input forces the system to leave the null - input fixed point , approach the constant - input attractor , and then relax back to its original state ( figure [ dynfig ] ) .
of course , in the presence of noise the shape of the transient is affected ( figure [ dynfig ] ) . observe the similarity between the embedding reconstructions of the evoked potentials ( figure [ epfig ] , bottom row ) and the actual behavior in @xmath68-@xmath69 coordinates ( figure [ dynfig ] , bottom row ) .
while we are specifically interested in the dynamics of thalamocortical loop which dictated our choice of the studied system , we checked if our approach is not specific to this model . our second model of choice consisted of two coupled rssler - type oscillators @xcite @xmath70 we used the frequency detuning parameter @xmath71 and the maximum coupling constant @xmath72 .
the scaling parameter @xmath73 took values from @xmath74 to @xmath75 .
the stimulation parameter @xmath53 was @xmath74 except for @xmath76 where it was set to @xmath77 ; the noise inputs @xmath78
, @xmath79 were gaussian white noise with parameters as for the wilson - cowan model .
the simulation was done for @xmath80 $ ] and segments from @xmath81 to @xmath82 , sampled every @xmath83 , were used for the analysis of the transients .
the synchronization was measured between @xmath84 and @xmath85 .
parameters of the system were chosen so that asymptotically it moved into a stable fixed point ( note the signs in the equations for @xmath86 and @xmath87 ) for both @xmath88 and @xmath89 . therefore the transient dynamics ( fig . [ rosdynfig ] ) ( a ) , ( b ) : signals ( @xmath90 coordinate ) from coupled rssler - type oscillators ; ( c ) , ( d ) : their delay - vector embeddings , shown in a plane defined by the first two principal components .
the intervals in ( a ) and ( b ) indicate the duration of non - zero input @xmath53 .
black ( thick ) lines are solutions for the system without noise , blue ( thin ) curves are five different realisations of noisy dynamics.,scaledwidth=47.0% ] is of the same type as in the model of thalamocortical loop : the system switches briefly to the second stable point and then returns .
note that the level of noise in the second subsystem is quite high and the evoked activity is barely visible at the single trial level ( fig .
[ rosdynfig ] , right column ) .
we aim at solving the following problem : suppose we have two pairs of signals , for example recordings from thalamus and cerebral cortex in two behavioral situations @xcite
. can we tell in which of the two situations the strength of connections between the structures is higher ? thus we need to find a measure being a monotonic function of the coupling strength .
we have studied this problem in our model of thalamocortical loop ( section [ sec : models1 ] ) .
we scaled the strength of connections from thalamus to cortex by changing @xmath73 between 0 and 1 , and calculated synchrony measures on signals from these structures .
the strength of connections from cortex to thalamus was constant ( @xmath91 ) ; see the appendix [ appa ] for the details .
[ cols= " > " , ] consider first stationary signals with @xmath88 or @xmath92 . without noise
the system is in a fixed point and obviously it is impossible to obtain the connection strength . however , given the noise , in principle the dynamics in the neighborhood of the fixed point is also probed .
thus there is a possibility that the interdependence and the strength of the coupling could be established during stationary parts of the dynamics .
it turns out that for null stimulation neither the interdependence measures nor the linear correlations detect any changes in the coupling strength ( figure [ synchrofig ] , left column ) . for constant non - zero input
there is a connection between the coupling strength and the values of the measure but they are anti - correlated and the dependence is not very pronounced ( figure [ synchrofig ] , right column ) .
one must also bear in mind that while it is possible to have no stimulation , in brain studies prolonged and constant stimulation in the present sense can not be experimentally realized ( at least for most sensory systems ) because of the adaptation of receptors .
the natural stimuli are necessarily transient . to use the synchrony measures on the transient we cut out pieces of signal corresponding to the evoked potential , and pasted them one after another .
thus obtained pseudo - periodic signal contained the same underlying dynamics with each piece differing due to the noise .
we then applied the same measures as we did for the stationary signals . in the simulations we calculated
50 `` evoked potentials '' ( figure [ epfig ] ) for each value of @xmath73 .
plots in the middle column of figure [ synchrofig ] show the values of the synchronization measures evaluated for different coupling strengths . it can be seen that they are increasing functions of the coupling strength between the subsystems .
therefore , our approach is indeed a viable solution to the problem of data - based quantification of the coupling strength .
it is interesting to study the values of these interdependence measures in different cases .
observe that @xmath93 for @xmath88 .
the opposite is true for transients ( for small @xmath73 ) .
this is even more clearly visible for @xmath12 .
in all the cases linear correlations showed similar trends to the nonlinear measures @xmath94 . the asymmetry in the interdependence measures was originally intended to be used for inferring the direction of the coupling or driving
. however , the inference of specific driving structure in every case must follow a careful analysis of underlying dynamics ( see , for example , discussions in @xcite and @xcite ) .
let us consider the plots in the middle column of figure [ synchrofig ] . for small @xmath73
the dominant connections are from the cortex to the thalamus so one might expect that the state of the thalamus might be easier predictable from the states of the cortex than the other way round . thus one would intuitively expect @xmath93 . however , we observe the opposite .
the reason is that the measures used are related to the relative number of degrees of freedom @xcite . loosely speaking , as discussed @xcite
, the effective dimension of the driven system ( thalamus for small @xmath73 ) is usually higher than the dimension of the driver ( which means that the response the dynamics of the thalamus is `` more complex '' ) .
this effect is further enhanced by the fact that we stimulate the thalamus in moments unpredictable from the point of view of the cortex .
summarizing , the result is compatible with the analysis in @xcite .
what happens for higher @xmath73 when the two measures become equal is probably the coupling between the two subsystems becoming so strong that the quality of prediction in any direction is comparable . in the stationary case
the situation is different as we observe the asymptotic behavior .
it turns out that for @xmath88 for every @xmath73 , and for @xmath95 for small @xmath73 we have @xmath93 .
but it seems that another effect also plays a role here .
the noise in the cortex has a higher amplitude than in the thalamus and as a consequence it is easier to predict the state of the thalamus from that of the cortex than in the other direction .
the reason for this disparity in the amplitudes is the difference in the shape of the sigmoidal functions @xmath96 .
to summarize , here , the asymmetry of the measures reflects internal properties of the two subsystems and not the symmetry properties of the coupling between them .
figure [ rosfig ] shows similar results obtained for two coupled rssler - type systems . in stationary situation
the interdependence measures are very noisy .
although a weak trend is visible , one would not be able to reliably discriminate between , say , @xmath97 and @xmath98 .
the equality of the measures in two directions is due to the fact that the systems are almost identical and symmetrically coupled .
if the interdependence is quantified on transient parts of the dynamics , the situation improves considerably .
@xmath99 has a high slope and is a very good measure of the coupling strength between the systems .
although @xmath100 has a slope comparable to that in the stationary case for @xmath88 , the variability of the results is much smaller , compared to the size of the fluctuation in the ensemble mean in the stationary case .
the difference between @xmath99 and @xmath100 reflects the asymmetry of the driving ( which makes the dynamics of @xmath101 `` more complex '' than the dynamics of @xmath102 ) , not of the coupling ( which is symmetric ) .
the performance of the procedure described above depends on the level of noise present in the system . to study this dependence we performed the simulations of the thalamocortical model ( the case of transient dynamics ) for 25% , 50% , 100% and 200% of the original noise level .
we found that for increasing level of noise the dynamics of the system may change qualitatively : if the noise level is large enough the system may be kicked out of the basin of attraction of the fixed point and would not return there after @xmath53 is reset to @xmath74 .
instead it may fall into the basin of attraction of another stable orbit or switch between the basins repeatedly .
we observed such behavior only once for 2500 simulations performed with 200% of the original noise and this trial was excluded from the analysis .
such behavior becomes more frequent with increasing noise ( e.g. 400% ) and so we did not study this situation as it was very different from the original dynamics of the system .
as one would expect , the higher the noise , the less sensitive the measures are ( fig . [ noisefig ] ) .
however , even for twice the original level of noise a weak trend in the interdependence is clearly visible .
nonlinear interdependence @xmath11 in thalamocortical loop model for varying level of noise .
the curves represent the values of the measure on transients .
noise level are 25% ( a ) , 50% ( b ) , 100% ( c ) and 200% ( d ) of the original noise.,scaledwidth=47.0% ] since we are interested in the dynamics of non - autonomous systems one might wonder if time - resolved measures , such as @xmath0 introduced in @xcite , would not perform better in the problem of inferring connection strength .
we performed tests on cut - and - pasted transient signals .
this problem is different from the one studied in @xcite .
( a ) + time - resolved nonlinear interdependence measure @xmath103 ( dashed lines ) and @xmath104 ( solid lines ) for 3 different values of the coupling @xmath73 .
the means are shown with horizontal lines .
( a ) @xmath105 , ( b ) @xmath106.,title="fig:",scaledwidth=48.0% ] + ( b ) + time - resolved nonlinear interdependence measure @xmath103 ( dashed lines ) and @xmath104 ( solid lines ) for 3 different values of the coupling @xmath73 .
the means are shown with horizontal lines .
( a ) @xmath105 , ( b ) @xmath106.,title="fig:",scaledwidth=48.0% ] there , two lorenz systems were coupled for short periods of time and @xmath0 was shown to identify these times of coupling well . in our problem
the coupling is constant in time , it is only the input to the system that is varying . for
the problem at hand the values of @xmath0 do not seem to change with varying coupling constant @xmath73 ( fig .
[ hi1fig ] , ( a ) ) when @xmath107 is constant , @xmath105 .
the reason for this may be that even for @xmath108 the subsystems are coupled through the connections from cortex to thalamus .
this hypothesis can be tested in another experiment , where all the connections between the subsystems are scaled and @xmath109 .
mean value of the time - resolved interdependence measure @xmath103 ( dashed lines ) and @xmath104 ( solid lines ) .
( a ) @xmath105 , ( b ) @xmath106,scaledwidth=47.0% ] indeed , in this setup the measure @xmath0 is sensitive to the coupling strength ( fig .
[ hi1fig ] , ( b ) ; fig .
[ hi6fig ] ) .
one may also note that @xmath103 is on average higher than @xmath104 , exactly as for @xmath11 in case of @xmath88 and contrary to what is observed using @xmath11 on transients ( fig .
[ synchrofig ] ( a ) ) .
thus it seems that for the problem of inferring coupling strength between two systems the optimal approach is to use @xmath11 or @xmath12 , or linear correlations , on the transients , as described in section [ sec : results1 ] .
to summarize , we have proposed a general approach for inference of the coupling strength using transient parts of dynamics .
we have shown that our approach gives more information about the coupling between subsystems than the approach using the stationary part of dynamics in case when the asymptotic dynamics is on a fixed point .
we have checked the validity of this approach on a model of a thalamocortical loop of sensory systems and on two coupled rssler - type oscillators .
we showed that our method is quite robust with respect to increasing level of noise as long as the dynamics does not change qualitatively .
we have also shown that this method measures different aspects of coupling than a time - resolved measure @xmath0 and than linear correlations .
we believe that our approach will be of use in many other physical systems studied in the stimulus - response paradigm , especially in the experimental context . the results of section [ sec : results1 ] are compatible with our preliminary studies of data from real neurophysiological experiments @xcite .
there one can not discern coupling strength in two contextual situations basing on stationary recordings , but the analysis of transients leads to clear differences between two variants of experiment .
the results of this analysis will be published elsewhere .
we are grateful to ewa kublik , daniel wiejkowski and andrzej wrbel for discussion of these topics .
some phase - space embeddings were calculated with `` chaotic systems toolbox '' procedure phasespace.m by alexandros leontitsis .
this research has been supported by the polish ministry of science and higher education under grants n401 146 31/3239 , pbz / mnisw/07/2006/11 and cost/127/2007 .
s was supported by the foundation for polish science .
we use the following equations for the model of thalamocortical loop : @xmath110 @xmath111 @xmath112 @xmath113 where @xmath114 @xmath115 standing for @xmath116 and @xmath78 , @xmath117 are noise inputs .
the normalizing constants @xmath118 are defined as @xmath119 .
the strength of connections was scaled by @xmath121 $ ] . everywhere except in section [ sec : results3 ] we used @xmath105 . in section
[ sec : results3 ] we used either @xmath121 $ ] and @xmath105 , or @xmath121 $ ] and @xmath106 . | we propose an approach for inferring strength of coupling between two systems from their transient dynamics .
this is of vital importance in cases where most information is carried by the transients , for instance in evoked potentials measured commonly in electrophysiology .
we show viability of our approach using nonlinear and linear measures of synchronization on a population model of thalamocortical loop and on a system of two coupled rssler - type oscillators in non - chaotic regime . |
a bimodal distribution in galaxy properties ( star - formation rate , size , morphology ) has been observed in the local universe ( e.g. kauffmann et al .
this bimodality is made of blue , predominantly late - type galaxies , whose emission is dominated by young stellar populations and experiencing significant level of star formation , complemented by red , early - type ( elliptical or s0 ) galaxies dominated by an old stellar population with little or absent star formation . the bimodality has been observed all the way to z @xmath7 2 ( labb et al . 2005 , kriek et al .
2006 , ilbert et al .
2010 , brammer et al .
2011 , whitaker et al .
2013 ) .
specific star - formation rates ( ssfr ) from spectral energy distribution ( sed ) fitting and equivalent widths from emission lines indicate for quiescent galaxies very low values ( @xmath8 ) even at high redshift ( ciambur et al .
2013 , kriek et al .
2006 , whitaker et al .
2013 ) , suggesting that these galaxies really are dead .
these levels of star formation are much lower than expected .
even if the galaxy would have stopped accreting new gas from the intergalactic medium , some gas should always become available again for star formation due to gas recycled from evolved stages of stellar evolution ( e.g. leitner & kravtsov , 2010 ) .
if the low levels of star formation are confirmed , it could have important implications for gas recycling and the effectiveness of quenching at high redshift . alternatively
, it is possible that amounts of star formation have been overlooked in previous studies because of heavy obscuration by dust . in bc03 models .
it overpredicts the sfr by a factor of 20 at each redshift .
ssfrs of star - forming galaxies on the main sequence ( cyan ) and the hubble time ( dashed gray ) are shown as references . ] to address this question , in this paper we determine the obscured sfrs of quiescent galaxies up to redshift @xmath9 using their 24@xmath2 m emission . in section 2 , we discuss the data . in section 3 ,
we describe the selection of qgs , and compare their sfrs from optical and near - infrared ( ir ) sed fitting to the values expected from the recycling of gas from mass loss .
we additionally evaluate how much obscured star formation might be hidden in our selection : this proves the need of looking at a mid - ir indicator for sfr . in section 4 , we stack 24@xmath2 m thumbnails of qgs in order to measure their obscured sfr .
we evaluate possible contributions to the mid - ir fluxes of qgs in section 5 .
we discuss our findings in section 6 and summarize them in section 7 . through the paper
we assume a standard cosmology with @xmath10 , @xmath11 , and @xmath12 ( komatsu et al .
the 3d - hst survey ( van dokkum et al . 2011 ; brammer et al .
2012 ) is a 600 arcmin@xmath13 survey using the hubble space telescope ( hst ) to obtain complete , unbiased low - resolution near - ir spectra for thousands of galaxies .
( cycles 18 and 19 , pi : van dokkum ) .
, @xmath14 ) of the dusty sfgs seds to the qgs seds .
orange dots show composite seds on the uvj boundary , corresponding to @xmath15 % .,width=302 ] it targets five fields ( cosmos , goods - s , goods - n , aegis , uds ) where a wealth of ancillary multi - wavelength data are available ( @xmath16 band to 24@xmath2 m ) ; they are crucial for interpreting spectra that often contain a single emission line , if any .
the 3d - hst photometric catalogue is described in detail in skelton et al .
it contains @xmath17 sources , detected on a noise - equalized combination of the f125w , f140w and f160w images .
the completeness of 3d - hst as a function of magnitude is evaluated by comparing the number of detections in the catalog to those in a deeper image of goods - s : the two of them deviate at magnitudes fainter than @xmath18 mag .
the wfc3 grism spectra have been extracted with a custom pipeline , described in momcheva et al .
( 2014 , in prep ) .
redshifts have been measured via the combined photometric and spectroscopic information using a modied version of the eazy code ( brammer et al .
the precision of redshifts is proven to be @xmath19 ( brammer et al .
2012 , momcheva et al .
2013 ) .
accurate redshifts allow the derivation of accurate rest - frame fluxes : we interpolate rest - frame filters from the observed sed with the inter - rest code ( taylor et al . , 2009 ) , based on the algorithm by rudnick et al ( 2003 ) .
stellar masses have been determined using the fast code by kriek et al .
( 2009 ) , using bruzual & charlot ( 2003 ) models , and assuming exponentially declining star - formation histories ( with e - folding times @xmath20 ranging from @xmath21 to @xmath22 yr ) , solar metallicity and a chabrier ( 2003 ) imf .
the 3d - hst catalogs are evaluated to be 90@xmath23 complete in stellar mass down to @xmath24 at @xmath25 ( tal et al .
2014 ) . in this paper
we restrict the analysis to the goods - n and goods - s fields , for which very deep _ spitzer_-mips ( @xmath26 , 3@xmath27 ) data are available ( dickinson et al . , 2003 ) , necessary for inferring low levels of sf . the mips
24@xmath2 m beam has a fwhm of 6 arcsec , therefore confusion and blending effects are unavoidable in deep observations at this resolution .
we perform photometry using a source - fitting algorithm ( labb et al .
2006 , wuyts et al .
2007 ) that takes advantage of the higher resolution information contained in the f160w images , as described in the appendix .
this method produces a model psf for the image with the broader native psf ( mips-24@xmath2 m in our case ) , which is then used to estimate the flux distribution of each source identified in the detection image ( @xmath28 ) segmentation map output by sextractor . for each individual object , the flux from neighboring sources ( closer than 10 arcsec ) is fitted and subtracted , allowing for a reliable aperture flux measurement of individual objects .
total ir luminosities ( @xmath29 = @xmath30 ) were derived from the observed 24@xmath2 m fluxes , on the basis of a single template that is the average of dale & helou ( 2002 ) templates with 1 @xmath31 @xmath32 @xmath31 2.5 , following wuyts et al .
( 2008 ; see also franx et al .
2008 , muzzin et al . 2010 ) , and in good agreement with recent _
herschel_/pacs measurements by wuyts et al .
sfrs are determined from the ir emission as in bell et al .
( 2005 ) for a chabrier imf : @xmath33.yr and solar metallicity .
] this quantity accounts properly just for obscured sf .
we derive sfrs without using data at wavelengths longer than 24@xmath2 m because photometry from the pacs and spire instruments on _ herschel _ in the goods fields is not as deep as that from mips-24@xmath2 m .
we evaluate the potential for detecting low sfrs with the _ herschel _ instruments , by using the pacs and spire detection limits ( elbaz et al .
2011 ) to sfrs : we extrapolate total ir luminosities from monochromatic fluxes , and convert them to sfr as described above .
we find that at @xmath34 the pacs-100@xmath2 m and 160@xmath2 m photometry is able to detect sfrs higher than @xmath35 ( 1@xmath27 ) , and spire-250 higher than @xmath36 ( 1@xmath27 ) , while at the same redshift mips-24 @xmath2 m is one order of magnitude deeper ( @xmath37 ) .
the same conclusion holds true at @xmath38 , with detection limits for mips-24@xmath2 m , pacs-100@xmath2 m , pacs-160@xmath2 m and spire-250@xmath2 m being respectively @xmath35 , @xmath39 , @xmath40 , @xmath39 ( all 1@xmath27 limits ) . on the 3d - hst goods fields extremely deep x - ray data are also available , 4ms in cdf - south ( see xue et al .
2011 ) , and 2ms in cdf - north ( see alexander et al . 2003 ) , that we use for identifying bright agns .
in order to select quiescent galaxies ( qgs ) we use a color - color technique ( figure [ uvj ] ) , specifically rest - frame @xmath41 versus rest - frame @xmath42 ( hereafter : uvj diagram ) . this technique has been widely used to distinguish qgs from sfgs , including the heavily reddened sfgs ( labb et al .
2005 ; wuyts et al . 2007 ; williams et al .
2009 ; brammer et al .
2009 ; whitaker et al . 2010 ; patel et al . 2012 ; bell et al . 2012 ; gobat et al .
qgs are identified using the criteria @xmath43 , @xmath44 and @xmath45 , as in whitaker et al .
magnitude , which does not affect the analysis . ] effectively , this selection targets galaxies whose optical and near - ir light is dominated by an old stellar population .
we select galaxies more massive than @xmath46 and divide the sample in five redshift bins , centered on @xmath3 = 0.5 , 0.9 , 1.2 , 1.7 , 2.2 .
the sample is mass complete at @xmath47 level even at the highest redshift ( @xmath25 ) we consider ( tal et al .
2014 ) . at each redshift
the qg sample consists of at least 60 galaxies ( table 1 ) . in figure [ spectrastackplot2 ]
we show stacked optical spectra of qgs and sfgs from 3dhst in mass - selected samples ( @xmath46 ) .
sfgs are subdivided into blue sfgs ( @xmath48 1.5 ) and dust - reddened sfgs ( @xmath49 1.5 ) .
qgs are subdivided according to the presence of a mips 24@xmath2 m detection . as noted by other authors ( e.g. brammer et al .
2009 , barro et al .
2013 ) , approximately @xmath50 of the optically - selected qgs have a 24@xmath2 m detection , which is in apparent contrast with the red optical colors and the seds .
we also notice that 24@xmath2m - detected qgs do not lie preferentially in any locus of the uvj diagram ( green dots in figure 1 ) .
the spectra in figure [ spectrastackplot2 ] clearly show that the uvj selection is efficient in dividing the two populations ; the sfg selection includes the heavily dust - reddened sfgs , that despite red @xmath41 colors , show spectral features ( h@xmath32 , d4000 ) characteristic of sfgs .
it is also noteworthy to see that qgs with 24@xmath2 m detections have some h@xmath32 and [ oiii ] ( cfr .
whitaker et al . 2013 ) , that indicate the presence of low level star formation and/or nuclear activity .
figure [ composite_sed ] shows composite seds ( following the methodology of kriek et al .
2011 ) for sfgs ( divided into blue and dusty ) and qgs ( divided according to the presence of a 24@xmath2 m detection ) .
the sed shapes of star - forming galaxies , dusty star - forming galaxies and quiescent galaxies are clearly different .
the rest - frame optical and near - ir seds of qgs with and without 24@xmath2 m detection are instead very similar .
we first analyze the ssfr from the sed fits to the uv - optical and near - ir photometry ( see section 2 ) .
the values are shown against redshift for the quiescent galaxies in figure 4 ( gray dots and black line ) .
the median value is @xmath51 , and the correlation with redshift is weak . these values compare well with those of ciambur et al ( 2013 ) , who used a similar method . no significative difference in the sed - derived sfrs is seen if we split the quiescent population between galaxies with and without a 24@xmath2 m detection .
the low ssfrs can be compared to the stellar mass loss from evolved stellar populations ( parriott & bregmann , 2008 ; leitner & kravtsov , 2010 ) .
green dots in figure [ fastvsmassloss ] represent for the qg sample the ssfr expected from stellar mass loss , assuming that 100% of the gas expelled from old stars is recycled into star formation .
mass loss is computed directly from @xmath52 of the bc03 models at the best fit age of the galaxy , given the best fit @xmath53 model ( see section 2 , data ) .
the expected ssfr from gas recycling is @xmath54 , with a weak redshift dependence .
it overpredicts the ssfr from optical and near - ir sed fitting by more than one order of magnitude .
the discrepancy between the two values at each redshift tells us that one of the following options must hold true : * a mechanism able to prevent the cooling of gas expelled from old stars and therefore the fueling of new star formation exists , _ or _ * sfrs from optical and near - ir best fitting are underestimated ( and a lot of star formation shows up in the mid - ir ) .
in the rest of the paper we test the latter possibility measuring sfrs in the mid - ir , in order to prove the former .
we evaluate how much star formation a galaxy can hide ( with high dust obscuration ) , while still retaining red optical - nir colors .
we stack the rest - frame seds of qgs and dusty sfgs in different redshift bins ; to each qg sed we add a variable fraction ( @xmath14 , normalized in light at 6000@xmath55 ) of the dusty sfg sed .
figure [ uvjpercentage ] shows the position in the uvj diagram of the qgs stacks ( red ) , the dusty sfgs stacks ( purple ) and the sed with @xmath15% ( orange ) , on the uvj separation border . adding a 30% dusty sfg sed to our typical qg
sed would keep such a galaxy as quiescent under our selection criteria despite the non - negligible contribution of obscured star formation . since the sfr of the average sfg evolves with redshift
, @xmath15% corresponds to @xmath56 at @xmath57 and @xmath58 at @xmath59 .
this shows that with high dust content , a red ( optical and near - ir ) galaxy can hide a significant amount of sf .
it is therefore necessary to measure sfr from mir indicators in order to evaluate the sfrs of qgs .
there is also a potential for entirely obscured populations with @xmath60 , which are known to exist the the centers of local dusty star - bursting galaxies ( e.g. arp 220 , sturm et al .
in this section we discuss the sfrs determined from the ir emission with the methodology described in section 2 . in figure [ masssfr ]
we plot the relation between stellar mass and sfr for galaxies in the mass - selected sample . as already noticed by various authors using a variety of sfr indicators ( e.g. noeske et al .
2007 , damen et al .
2009 , whitaker et al .
2012 ) , sfrs and masses of sfgs are correlated ( light - blue dots ) , with a scatter of approximately 0.3 dex .
the vast majority of the qgs lies in this plane below the star - forming main sequence. most of the qgs are undetected in the mips 24@xmath2 m image at 3@xmath27 ( red dots ) , while some of them ( approximately 25 % , table 1 ) have a 24@xmath2 m detection ( green dots ) , placing them in the mass - sfr plane between sfgs and the detection limit . in order to measure the sfrs for qgs
, we stack 24@xmath2 m thumbnails .
we emphasize that in this step we stack _ cleaned _ images ( 20 wide ) , where neighboring sources identified in the high resolution @xmath28 band have been subtracted with the technique described in the appendix .
summing original 24@xmath2 m thumbnails would lead to a stack with a very poorly constrained background , raised by the presence of neighboring objects .
since the goal of this paper is to measure very low sfrs with accuracy , it is fundamental to perform photometry on a stacked image with small uncertainty on the background ( as shown in figure [ stack_procedure ] ) .
we perform an average - stacking in different redshift bins , for two samples : all qgs and only non-24@xmath2m - detected qgs .
photometry on the stack is performed within an aperture of 6 arcsec diameter , similar to the size of the 24@xmath2 m fwhm . to measure the total 24@xmath2 m flux
, we create a mips growth curve from several bright , isolated , and unsaturated point sources within each field .
these square postage stamps are 20 arcsec wide , and we derive an aperture correction of a factor of 2.57 from r =3 arcsec to r = 10 arcsec . to convert to total flux , we include an additional aperture correction for the 22% of the flux that falls outside 10 arcsec , derived from the mips handbook . , we color - correct fluxes by dividing them by a factor of 0.967 ( mips handbook , table 4.17 ) . this way flux densities correspond to those of sources with a flat spectrum .
] m , while yellow dots represent all qgs .
for the two samples , the average ratio is respectively @xmath61 and @xmath62 .
these ratios are possibly even lower because for qgs ir inferred sfrs can be significantly contaminated by other sources of dust heating ( section 5 )
. , width=321 ] we obtain mostly clear detections with signal - to - noise of 35 , and fluxes @xmath63 , corresponding to sfr @xmath64 0.5 - 5 @xmath65 .
we summarize the measured stacked fluxes in table 1 .
errors on the stacks are measured through bootstrapping of the sources .
errors on the stacks are measured through bootstrapping , as follows .
each sample of qgs is resampled 1000 times .
we stack the individual 24@xmath2 m images of galaxies belonging to each resampling and perform photometry on the new stacked images .
the dispersion of the flux values in the resampled stacks gives the uncertainty on the flux measurement . in figure [ masssfr ]
we overplot with large yellow dots the sfr obtained from all qgs ( big yellow dots ) and non-24@xmath2m - detected galaxies ( big red dots ) , representative of the deadest fraction of the galaxy population ( this definition of quiescent galaxy is the same as in bell et al .
2012 ) . despite the different sample selection ( all qgs or just qgs not detected at 24@xmath2 m ) , it is evident that at each redshift the average qgs has a sfr which is at least @xmath66 times lower than the ones on the star - forming sequence. in figure [ evolution_ssfr ] ( left panel ) we show the redshift evolution of sfrs of sfgs and qgs .
we plot ssfr since it is more mildly dependent on stellar mass than sfr itself . as noted by previous studies ( e.g. damen et al .
2010 , whitaker et al .
2012 , karim et al .
2012 , fumagalli et al .
2012 ) , the evolution of ssfr in redshift for star - forming galaxies is well fit by a power law @xmath67 where @xmath68 . at each redshift
qgs have a ssfr at least 20 times lower than sfgs .
the evolution with redshift of ssfr of qgs seems to resemble the evolution of sfgs . at redshift @xmath69 , qgs form 10 times more stars than at redshift @xmath70 . in figure [ ratio_ssfr ]
we show for our samples the ratio of the average sfr of qgs to the average sfr of sfgs at the same redshift . for quiescent galaxies undetected at 24@xmath2 m ,
the mean value of the ratio is @xmath71 while for the entire sample it is @xmath72 .
this confirms that at each redshift quenching of star formation is very efficient .
for qgs the sfrs inferred from the ir emission are generally an order of magnitude larger than those inferred from stellar population modeling ( black dashed line in figure [ evolution_ssfr ] , left panel ) . at the highest redshifts they are similar to the values predicted by the recycling of mass loss ( green dashed line in figure [ evolution_ssfr ] , left panel ) , while at redshift lower than 1.5 they are significantly lower than those .
strictly should be that the ir - inferred sfrs for qgs are upper limits , because of contributions of agns , agb stars and dust heating from old stellar populations to the ir fluxes .
we treat each of these components separately in the following subsections and compare their contributions to @xmath29 with the observed stacked values of @xmath29 in the discussion section .
we evaluate the possible contribution of agns by stacking x - ray thumbnails ( from the cdf - s 4ms and cdf - n 2ms ) of the qgs in different redshift bins .
mullaney et al .
2011 demonstrates ( equation 4 ) the existence of a linear relation between the x - ray luminosity @xmath73 and @xmath29 for a sample of local agns . after subtracting individually - detected x - ray point sources ( marked with orange stars in figure [ masssfr ] ) ,
we obtain marginal detections ( @xmath74 ) ranging from @xmath75 in the lowest redshift bin to @xmath76 in the highest redshift bin . converting the obtained x - ray luminosities to ir luminosities with the mullaney relation ,
we obtain the gray line in figure [ evolution_ssfr ] ( right panel ) .
it lies three orders of magnitude below the observed @xmath77 .
olsen et al . ( 2013 ) suggest that at redshift @xmath69 most qgs host a low - luminosity agn , comparing sfr inferred from ir and x - ray .
they find , for qg at @xmath78 of the same mass of that of our sample , a mean luminosity of @xmath79 , consistent with our study . even though most of qgs host a ( low - luminosity ) agn , we find that those weak agn can not account for the mir emission of the galaxies . other studies ( donley et al .
2008 ; kartaltepe et al .
2010 ) have also already pointed out that systems with 24@xmath2 m flux dominated by agns are not the dominant population at low @xmath29 , such as qgs .
agb stars are known to evolve embedded in a circumstellar dusty envelope ( e.g. bressan et al .
1998 , lancon & mouhcine 2002 , piovan et al . 2003 ) .
they are the dominant source of the rest - frame @xmath80-band luminosity between 0.1 and 1.5 gyr of age ( kelson & holden 2010 ) and significantly contribute to mir emission , but their dust contribution is not included in classical optical - near infrared sed fitting ( bruzual & charlot 2003 , maraston 2005 ) .
we evaluate the contribution to @xmath29 with the new charlot & bruzual 2010 model ( cb2010 ) of an ssp with solar metallicity ( private communication ) . given galaxy ages from the fast best fits ( see section 2 , and whitaker et al .
2013 , in press ) , for each galaxy in our qg sample we estimate the observed 24@xmath2 m flux from the cb2010 model and convert it to @xmath29 with the same relation of wuyts et al.(2008 ) ( figure 8 , right panel , orange line ) .
another possible contribution to @xmath29 is dust heating from old stellar populations .
salim et al . ( 2009 ) concludes that , for a sample of 24@xmath2m - detected galaxies in the deep2 survey ( 0.2 @xmath81 1.0 ) , the bulk of ir emission in red ( @xmath82 ) galaxies comes from the heating of diffuse cirrus dust by old stellar populations , rather than by dust heating in star - forming regions .
we test if this holds true for the galaxies in our sample as follows .
given the stellar population parameters from the fast best - fit to the seds ( age , @xmath53 , @xmath83 ) , we evaluate the luminosity absorbed at @xmath84 m by integrating the difference between the unattenuated and the attenuated synthetic sed , and assume it is re - emitted in the ir ( see charlot & fall 2000 , da cunha et al .
2008 ) .
we then compare the model @xmath29 predicted by the attenuated sed with the best fit sfr .
if @xmath29 originates in dust associated with star - forming regions , we expect the ratio @xmath29/sfr to be @xmath85 ( bell et al . , 2005 ) .
figure [ predicted ir ] shows that sfgs ( blue ) are consistent with this prediction . on the other hand , for qgs ( red points )
@xmath29 is systematically higher than the expectations from dust heating in star - forming regions .
this indicates that in qgs a significant contribution to @xmath29 comes from dust heated by old stellar populations . inferring sfr from @xmath29 ( and therefore from 24@xmath2 m fluxes )
overestimates the real sfr of qgs . for each galaxy in the qg sample we estimate the expected @xmath29 luminosity from circumstellar dust , and compute the mean value in different redshift bins ( figure 8 , right panel , green line ) . /sfr
, for qgs ( red ) and sfgs ( blue ) .
@xmath29 is reconstructed assuming that the light absorbed by dust at uv - optical wavelengths is re - emitted in the ir ( section [ contributions to lir ] ) . for sfgs
the ratio is comparable to the bell et al .
( 2005 ) relation ( black line ) , while for qgs @xmath29 are systematically higher than the expectations from sfr , meaning that for qgs most of dust heating comes from old stellar populations.,width=302 ]
as we have seen above , various processes other than star formation can contribute to the observed mid - ir flux .
we next discuss the impact on the derived sfrs .
moreover , we put constraints on the mass growth of qgs implied by the measured sfrs and on their size growth implied by the stellar mass loss . in figure [ evolution_ssfr ] ( right panel )
we show the approximate evolution of @xmath86 , for data ( dashed lines ) and models ( thick solid lines ) .
we saw earlier that observations of @xmath29 are based on the extrapolation of the single band 24@xmath2 m to @xmath87(ir ) assuming a template for dust heating by star - forming regions ( section 2 ) .
model predictions estimate that the agn contribution ( gray line ) to the @xmath29 is negligible for our sample , while the model expectation for @xmath29 from cirrus dust ( green ) and circumstellar dust ( orange ) is comparable to the observed values from stacking .
we note that qualitatively both of them decrease towards lower redshift , respectively because of higher @xmath83 and younger stellar ages at higher redshifts ( which leads to more absorbed optical light re - emitted in the ir in the younger universe ) and because of the aging of galaxies ( which leads to lower contribution of agb stars in the sed ) .
if sfrs from sed fitting are correct , their contribution to @xmath29 ( black line in figure [ evolution_ssfr ] , right panel ) would be 1 dex lower than the observed @xmath29 , while dust heated by old stellar population can account for the most of the observed luminosities .
all the measured values from 24@xmath2 m stacks must therefore be considered as upper limits to the sfr . at each redshift , the mean qg has a sfr at least @xmath720 - 40 times lower than that of a sfg at the same redshift .
these sfrs are significantly higher than estimates based on optical and near - ir model fits ( see section 3 and ciambur , kauffman & wuyts 2013 ) . in order to evaluate the growth of a qg via star formation we integrate the ssfr(ir)-z trend of figure 8 ( left )
this leads to estimate that the @xmath88 growth of a qg via star formation is 20% from redshift 2 to 0 . some authors ( e.g. van dokkum et al . 2010 , patel et al .
2012 ) have inferred that a present - day @xmath89 @xmath90 galaxy has to grow 60% of its mass from redshift @xmath91 to @xmath92 .
we show that star formation can not be responsible for the entire stellar growth of qgs , while other mechanisms must be in place , such as minor merging ( see , among others , hopkins 2009 ) .
the limit we compute on the mass increase via star formation is more stringent than that computed by prez - gonzlez et al .
( 2008 ) , who estimates that massive spheroid - like galaxies may have doubled ( at the most ) their stellar mass from redshift 2 to 0 .
the sfrs expected from stellar mass loss are probably much higher than the real sfrs of qgs , meaning that star formation from mass loss is inefficient .
if mass loss from evolved stars is not converted into stars and gas is expelled from the galaxy , an interesting consequence is that the potential of the system becomes shallower and the system expands ( zhao et al .
2002 , murray et al .
2010 ) . in brief
( following damjanov et al . , 2009 ) , if a system loses a fraction @xmath93 of its mass in a time longer than a dynamical timescale , it will expand its radius by a factor of @xmath94 .
the modeled mass losses for galaxies in our sample ( figure 4 ) integrated over the redshift range 0 to 2 give @xmath95 , which leads to @xmath96 .
the observed size growth of quiescent galaxies from redshift 2 to 0 amounts to a factor of 2 - 3 ( e.g. williams et al .
2010 , newman et al .
2012 , whitaker et al .
2012 ) , therefore mass loss can not be its unique cause but only one of the concurrent ones ( see also damjanov et al . , 2009 ) .
we note that the assumed mass loss depends on the absolute ages of galaxies at each redshift , which are very uncertain .
a possible caveat in the study comes from the fact that the 24@xmath2m - to - l(ir ) conversion relies on a single infrared template ( wuyts et al .
2008 ) , while the underlying sed for high - redshift quiescent galaxies is unknown . as explained in section 2 ( data ) ,
measuring the low fluxes of quiescent galaxies at herschel wavelengths is extremely challenging .
available observationally motivated far - ir seds of galaxies at high redshift are based on bright herschel sources ( e.g. elbaz et al .
2011 , magdis et al .
2014 ) and refer to galaxies on or above the star - forming main sequence .
we evaluate the possible bias introduced by our synthetic template by comparing the l(ir ) integrated under the elbaz et al .
2011 sed for main - sequence galaxies to that inferred from a simulated 24@xmath2 m observation of that sed , at different redshifts .
we obtain that at @xmath97 the inferred l(ir ) is a factor of 2 higher than the integrated l(ir ) , while at @xmath98 the bias is lower than 50% . the possibility that galaxies below the main - sequence have different infrared sed shapes must however be taken into account ( see also utomo et al .
2014 , hayward et al .
we compute the systematic uncertainty on l(ir ) arising from the unknown underlying sed as follows .
we compare the 24@xmath2m - to - lir conversion of wuyts et al .
2008 , obtained by averaging a vast array of infrared templates from dale & helou 2002 , with those obtained by using each single dale & helou 2002 template for different values of the ionization parameter @xmath32 .
the dispersion on the values is 0.3 dex , which we consider the systematic uncertainty on the conversion .
we conclude that the possible biases and uncertainties induced by inferring l(ir ) from a single band and a single template amount to a factor of 2 , and do not affect the conclusions of the paper .
we select quiescent galaxies at redshift @xmath99 in the 3d - hst survey from their rest - frame optical and near - ir colors . fitting their uv to near - ir photometry with stellar population models
, we find very low star - formation rates ( @xmath1 ) .
these values are much lower than the stellar mass loss rates predicted by the same models .
this suggests that the star formation is either missed because it is dust obscured , or that the gas from stellar mass loss is expelled from the galaxy or prevented from refueling star formation .
we put upper limits on the obscured star - formation rate of quiescent galaxies by stacking 24@xmath2 m images . including direct 24@xmath2 m detections , we find that @xmath100 . at each redshift
the ssfr of quiescent galaxies is @xmath7 20 - 40 times lower than the typical value on the main sequence of star - forming galaxies .
sfrs of quiescent galaxies are possibly even lower than this , because the ir luminosity can also be due to other sources , such as the presence of agb dust enshrouded stars and dust heating from older stellar populations .
stacks of longer wavelength data ( such as from _ herschel _ ) are necessary for constraining the dust temperature and therefore distinguishing between the different contributions to @xmath29 , however a large sample may be necessary to achieve adequate s / n ( e.g. viero et al .
we show nevertheless that dust heating from old stellar populations can account for most of the observed @xmath29 .
the observed sfr(ir ) are therefore upper limits to the real sfr , which are possibly one order of magnitude lower .
this means that there must be a mechanism that not only shuts down star formation , but also keeps the galaxy dead for a long period of time , preventing the ejected gas from cooling and forming new stars .
if gas from mass - loss is expelled from galaxies , we predict that it is responsible for a growth in stellar radii of 60% from redshift 2 to 0 .
we acknowledge funding from erc grant highz no .
this work is based on observations taken by the 3d - hst treasury program ( go 12177 and 12328 ) with the nasa / esa hst , which is operated by the association of universities for research in astronomy , inc .
, under nasa contract nas5 - 26555 .
akylas , a. , georgakakis , a. , georgantopoulos , i. , brightman , m. , & nandra , k. 2012 , , 546 , a98 alexander , d. m. , bauer , f. e. , brandt , w. n. , et al .
2003 , , 126 , 539 barro , g. , faber , s. m. , prez - gonzlez , p. g. , et al .
2013 , , 765 , 104 bell , e. f. , papovich , c. , wolf , c. , et al .
2005 , , 625 , 23 bell , e. f. , van der wel , a. , papovich , c. , et al .
2012 , , 753 , 167 brammer , g. b. , van dokkum , p. g. , & coppi , p. 2008
, , 686 , 1503 brammer , g. b. , whitaker , k. e. , van dokkum , p. g. , et al . 2009 , , 706 , l173 brammer , g. b. , whitaker , k. e. , van dokkum , p. g. , et al .
2011 , , 739 , 24 brammer , g. , van dokkum , p. , franx , m. , et al .
2012 , arxiv:1204.2829 parriott , j. r. , & bregman , j. n. 2008 , , 681 , 1215 bressan , a. , granato , g. l. , & silva , l. 1998 , , 332 , 135 bruzual , g. , & charlot , s. 2003 , , 344 , 1000 chabrier , g. 2003 , , 586 , l133 charlot , s. , & fall , s. m. 2000 , , 539 , 718 ciambur , b. c. , kauffmann , g. , & wuyts , s. 2013 , , 432 , 2488 da cunha , e. , charlot , s. , & elbaz , d. 2008 , , 388 , 1595 damen , m. , labb , i. , van dokkum , p. g. , et al .
2011 , , 727 , 1 damjanov , i. , mccarthy , p. j. , abraham , r. g. , et al .
2009 , , 695 , 101 dickinson , m. , & goods team 2004 , bulletin of the american astronomical society , 36 , # 163.01 donley , j. l. , rieke , g. h. , prez - gonzlez , p. g. , & barro , g. 2008 , , 687 , 111 elbaz , d. , dickinson , m. , hwang , h. s. , et al .
2011 , , 533 , a119 franx , m. , van dokkum , p. g. , schreiber , n. m. f. , et al .
2008 , , 688 , 770 fumagalli , m. , patel , s. g. , franx , m. , et al .
2012 , , 757 , l22 gobat , r. , strazzullo , v. , daddi , e. , et al .
2013 , arxiv:1305.3576 hopkins , p. f. , bundy , k. , murray , n. , et al .
2009 , , 398 , 898 ilbert , o. , salvato , m. , le floch , e. , et al .
2010 , , 709 , 644 karim , a. , schinnerer , e. , martnez - sansigre , a. , et al .
2011 , , 730 , 61 kartaltepe , j. s. , sanders , d. b. , le floch , e. , et al .
2010 , , 709 , 572 kauffmann , g. , heckman , t. m. , white , s. d. m. , et al .
2003 , , 341 , 54 kelson , d. d. , & holden , b. p. 2010
, , 713 , l28 kennicutt , r. c. , jr .
1998 , , 498 , 541 komatsu , e. , smith , k. m. , dunkley , j. , et al .
2011 , , 192 , 18 kriek , m. , van dokkum , p. g. , franx , m. , et al . 2006 , , 649 , l71 kriek , m. , van dokkum , p. g. , labb , i. , et al .
2009 , , 700 , 221 kriek , m. , van dokkum , p. g. , whitaker , k. e. , et al .
2011 , , 743 , 168 labb , i. , huang , j. , franx , m. , et al .
2005 , , 624 , l81 labb , i. , bouwens , r. , illingworth , g. d. , & franx , m. 2006 , , 649 , l67 lanon , a. , & mouhcine , m. 2002 , , 393 , 167 leitner , s. n. , & kravtsov , a. v. 2011 , , 734 , 48 maraston , c. 2005 , , 362 , 799 mullaney , j. r. , alexander , d. m. , goulding , a. d. , & hickox , r. c. 2011 , , 414 , 1082 muzzin , a. , van dokkum , p. , kriek , m. , et al . 2010 , , 725 , 742 murray , n. , quataert , e. , & thompson , t. a. 2010 , , 709 , 191 newman , a. b. , ellis , r. s. , bundy , k. , & treu , t. 2012 , , 746 , 162 noeske , k. g. , weiner , b. j. , faber , s. m. , et al .
2007 , , 660 , l43 olsen , k. p. , rasmussen , j. , toft , s. , & zirm , a. w. 2013 , , 764 , 4 patel , s. g. , holden , b. p. , kelson , d. d. , et al .
2012 , , 748 , l27 prez - gonzlez , p. g. , trujillo , i. , barro , g. , et al .
2008 , , 687 , 50 piovan , l. , tantalo , r. , & chiosi , c. 2003 , , 408 , 559 rudnick , g. , rix , h .- w . , franx , m. , et al .
2003 , , 599 , 847 salim , s. , dickinson , m. , michael rich , r. , et al .
2009 , , 700 , 161 skelton , r. e. , whitaker , k. e. , momcheva , i. g. , et al .
2014 , arxiv:1403.3689 sturm , e. , lutz , d. , genzel , r. , et al .
1996 , , 315 , l133 taylor , e. n. , franx , m. , van dokkum , p. g. , et al . 2009 , , 183 , 295 van dokkum , p. g. , whitaker , k. e. , brammer , g. , et al . 2010 , , 709 , 1018 viero , m. p. , moncelsi , l. , quadri , r. f. , et al .
2013 , arxiv:1304.0446 xue , y. q. , luo , b. , brandt , w. n. , et al .
2011 , , 195 , 10 whitaker , k. e. , van dokkum , p. g. , brammer , g. , et al .
2010 , , 719 , 1715 whitaker , k. e. , van dokkum , p. g. , brammer , g. , & franx , m. 2012 , , 754 , l29 whitaker , k. e. , van dokkum , p. g. , brammer , g. , et al .
2013 , , 770 , l39 williams , r. j. , quadri , r. f. , franx , m. , van dokkum , p. , & labb , i. 2009 , , 691 , 1879 wuyts , s. , labb , i. , franx , m. , et al .
2007 , , 655 , 51 wuyts , s. , labb , i. , schreiber , n. m. f. , et al .
2008 , , 682 , 985 wuyts , s. , frster schreiber , n. m. , lutz , d. , et al .
2011 , , 738 , 106 [ cols="<,<,<,<,<,<,<",options="header " , ] for different redshift bins : number of galaxies in the quiescent sample ( qg ) and quiescent sample without 24@xmath2 m detection ( qg , no24@xmath2
m ) , along with their stacked 24 m fluxes , and the implied sfr from ir emission .
[ appendix ]
the mips-24@xmath2 m beam has a fwhm of 6 arcsec , therefore confusion and blending effects are unavoidable in deep observations at this resolution .
we use a source - fitting algorithm designed to extract photometry from irac and mips images ( see , e.g. , labbe et al .
2006 ; wuyts et al .
the information on position and extent of the sources based on the higher resolution @xmath28 segmentation map is used to model the lower resolution mips-24@xmath2 m images .
local convolution kernels are constructed using bright , isolated , and unsaturated sources in the @xmath28 and mips-24@xmath2 m , derived by fitting a series of gaussian - weighted hermite functions to the fourier transform of the sources .
each source is extracted separately from the @xmath28 image and , under the assumption of negligible morphological k - corrections , convolved to the mips-24@xmath2 m resolution using the local kernel coefficients .
all sources in each mips-24@xmath2 m image are fit simultaneously , with the flux left as the only free parameter .
the modeled light of neighboring sources ( closer than 10 arcsec ) is subtracted , thereby leaving a `` clean '' mips-24@xmath2 m image to perform aperture photometry and stacking of faint sources .
the technique is illustrated in figure a1 and a2 , respectively for a bright and a faint source .
m fluxes for objects identified in the f160w detection image .
panel 1 shows the original 24@xmath2 m cutout for an object in the catalog .
panel 2 and 3 show the matching f160w detection image and segmentation map from sextractor .
the bottom row shows the modeled 24@xmath2 m flux for all objects in the region ( panel 4 ) , the residual image with all modeled fluxes removed ( panel 5 ) , and the flux for the central object alone ( panel 6).,width=377 ]
the paper is built on data from the goods - north and goods - south fields .
the two fields feature a similar large number of optical - near - ir observations included in the 3d - hst photometric catalog , and data quality in the 3d - hst fields is uniform ( see skelton et al . , 2014 ) .
the depths of mips-24@xmath2 m data are similar in the two fields ( dickinson et al .
we show in figure b1 the main result of the paper - the evolution of ssfrs of qgs - once the data are stacked separately in the two fields .
differences and errors are consistent with lower statistics .
m detection , for different fields and in the entire sample .
red / black dots are stacked values from goods - north / goods - south , and large blue dots are values from the combined sample .
errors are computed bootstrapping the sample .
mean redshifts have been shifted of @xmath101 0.05 for clarity .
differences and errors are consistent with lower statistics.,width=321 ] | we investigate star - formation rates ( sfr ) of quiescent galaxies at high redshift ( @xmath0 ) using 3d - hst wfc3 grism spectroscopy and _ spitzer _ mid - infrared data .
we select quiescent galaxies on the basis of the widely used uvj color - color criteria .
spectral energy distribution ( sed ) fitting ( rest - frame optical and near - ir ) indicates very low star - formation rates for quiescent galaxies ( @xmath1 ) .
however , sed fitting can miss star formation if it is hidden behind high dust obscuration and ionizing radiation is re - emitted in the mid - infrared .
it is therefore fundamental to measure the dust - obscured sfrs with a mid - ir indicator .
we stack the mips-24@xmath2 m images of quiescent objects in five redshift bins centered on @xmath3 = 0.5 , 0.9 , 1.2 , 1.7 , 2.2 and perform aperture photometry . including direct 24@xmath2 m detections
, we find @xmath4 @xmath5 .
these values are higher than those indicated by sed fitting , but at each redshift they are 20 - 40 times lower than those of typical star - forming galaxies .
the true sfrs of quiescent galaxies might be even lower , as we show that the mid - ir fluxes can be due to processes unrelated to ongoing star formation , such as cirrus dust heated by old stellar populations and circumstellar dust .
our measurements show that star - formation quenching is very efficient at every redshift .
the measured sfr values are at @xmath6 marginally consistent with the ones expected from gas recycling ( assuming that mass loss from evolved stars refuels star formation ) and well below that at lower redshifts . |
the first positive parity resonance of the nucleon , the @xmath2 or roper resonance , has been the subject of extensive interest since its discovery in 1964 @xcite .
this @xmath3-wave isospin-1/2 spin-1/2 ( @xmath1 ) pion - nucleon resonance has held the curiosity and imagination of the nuclear and particle physics community due to its surprisingly low mass . for example , in constituent quark models the lowest - lying odd - parity state occurs below the @xmath4 state @xcite whereas in nature , the negative parity @xmath5 @xmath6 state is almost 100 mev _ above _ the roper resonance .
this phenomenon has led to wide speculation on the possible exotic nature of the roper resonance .
for example , the roper resonance has been described as a hybrid baryon state with explicitly excited gluon field configurations @xcite , as a breathing mode of the ground state @xcite or a state which can be described in terms of a five quark ( meson - baryon ) state @xcite .
the elusive nature of this low - lying resonance is not constrained to model calculations alone .
there have been several investigations of the low - lying nucleon spectrum using the first - principles approach of lattice field theory .
the lattice approach to quantum chromodynamics ( qcd ) provides a non - perturbative tool to explore the properties of hadrons from the first principles of this fundamental quantum field theory .
numerical simulations of qcd on a space - time lattice with the light up , down and strange dynamical - quark masses similar to those of nature are now possible @xcite . as such , some long - standing problems in nuclear - particle physics are now being resolved .
for example , the ground - state hadron spectrum is now well understood @xcite . however
, gaining knowledge of the excited - state spectrum presents additional challenges .
the euclidean - time correlation function provides access to a tower of energy eigenstates in the form of a sum of decaying exponentials with the masses of the states in the exponents .
the ground state mass , being the lowest energy state , has the slowest decay rate , and is obtained through the analysis of the large - time behaviour of this function .
however , the excited states appear in the sub - leading exponentials . extracting excited state masses from these exponents
is intricate as the correlation functions decay quickly and the signal to noise ratio deteriorates rapidly .
in addition , the spectrum is composed of both single - particle states and multiple - particle states interacting and mixing in the finite physical volume of the lattice .
understanding the finite - volume dependence of these states and linking them to the resonances of nature is a long - term program of the lattice qcd community . in this letter
we report the excited - state energy spectrum of the nucleon in the light quark - mass regime of qcd for the first time .
of particular note is the identification of a new low - lying state associated with the roper resonance of nature .
several attempts have been made in the past to find the elusive low - lying roper state in the lattice framework @xcite .
the results reported therein are as much about the development of lattice techniques as they are about the nucleon spectrum . where the lattice techniques are regarded as robust , a low - lying roper state was not observed .
the difficulties lie in finding effective methods to isolate the energy eigenstates of qcd and in accessing the light quark mass regime of qcd .
the ` variational method ' @xcite is the state - of - the - art approach for determining the excited state hadron spectrum .
it is based on the creation of a matrix of correlation functions in which different superpositions of excited state contributions are linearly combined to isolate the energy eigenstates .
a diversity of excited - state superpositions is central to the success of this method .
early implementations of this method , using a variety of standard spin - flavor interpolating fields of fixed source - distribution size , were not successful in isolating energy eigenstates .
instead the putative eigenstates were superpositions of energy - states @xcite and the low - lying roper state was hidden by excited - state contaminations .
a solution to this problem was established in refs .
@xcite where a low - lying roper state was isolated .
key to this approach , used herein , is the utilization of a diverse range of fermion source and sink smearings in creating the matrix of correlation functions .
the diversity of smearings leads to a wide variety of superpositions of excited - state contributions , providing a suitable basis for constructing linear combinations which isolate the eigenstates .
these effective techniques @xcite were developed in the quenched approximation and we bring them to the dynamics of full qcd for the first time here .
the low - lying even - parity states of the nucleon are explored in full qcd using 2 + 1-flavor dynamical - qcd gauge - field configurations @xcite .
whereas other recent full qcd analyses @xcite report a first positive parity excited state that appears too high to be considered as the roper resonance , we will illustrate how the low - lying roper state of the present analysis approaches the physical mass of nature , displaying significant chiral curvature at the lightest quark mass ( corresponding to a pion mass of 156 mev , only slightly above the physical value of 140 mev ) .
these full qcd results , providing the world s first insight into the nucleon mass spectrum in the light - quark regime , are significantly different from those of quenched qcd and provide interesting insights into the dynamics of qcd . in constructing our correlation matrix for the nucleon spectrum ,
we consider the two - point correlation function matrix with momentum @xmath7 @xmath8 where dirac indices are implicit . here ,
@xmath9 and @xmath10 are the couplings of the interpolators @xmath11 and @xmath12 at the sink and source respectively and @xmath13 enumerates the energy eigenstates with mass @xmath14 .
@xmath15 projects the parity of the eigenstates .
since the only @xmath16 dependence comes from the exponential term , one can seek a linear superposition of interpolators , @xmath17 , such that @xmath18 for sufficiently large @xmath19 and @xmath20 .
multiplying the above equation by @xmath21^{-1}$ ] from the left leads to an eigenvalue equation @xmath22_{ij}\ , u^{\alpha}_{j } & = c^{\alpha}\ , u^{\alpha}_{i } , \label{eq : right_evalue_eq}\end{aligned}\ ] ] where @xmath23 is the eigenvalue .
similar to eq .
( [ eq : right_evalue_eq ] ) , one can also solve the left eigenvalue equation to recover the @xmath24 eigenvector @xmath25_{ij } & = c^{\alpha}\ , v^{\alpha}_{j}. \label{eq : left_evalue_eq}\end{aligned}\ ] ] the vectors @xmath26 and @xmath27 diagonalize the correlation matrix at time @xmath19 and @xmath20 and provide the projected correlator @xmath28 the parity projected , eigenstate projected correlator @xmath29 is then analyzed using standard techniques to obtain the masses of different states @xcite .
the pacs - cs @xmath31 flavor dynamical - fermion configurations @xcite made available through the ildg are used herein .
these configurations use the non - perturbatively @xmath32-improved wilson fermion action and the iwasaki - gauge action @xcite .
the lattice volume is @xmath33 , with @xmath34 providing a lattice spacing @xmath35 fm .
five values of the ( degenerate ) up and down quark masses are considered , with hopping parameter values of @xmath36 , corresponding to pion masses of @xmath37 = 0.702 , 0.572 , 0.413
, 0.293 , 0.156 gev @xcite ; for the strange quark @xmath38 we consider 350 configurations for the four heavier quarks , and 198 configurations for the lightest quark .
an ensemble of 750 samples for the lightest quark mass is created by using multiple fermion sources on each configuration , spaced to sample approximately independent regimes of each configuration .
our error analysis is performed using a second - order jackknife method , where the @xmath39 for projected correlator fits is obtained via a covariance matrix analysis .
our fitting method is discussed in refs .
@xcite .
each interpolator has a unique dirac structure giving rise to different spin - flavor combinations .
moreover , as each spinor has upper and lower components , with the lower components containing an implicit derivative , different combinations of zero and two - derivative interpolators are provided .
the local scalar - diquark nucleon interpolator , @xmath42 , is well known to have a good overlap with the ground state of the nucleon . also , this @xmath42 interpolator is able to extract a low - lying roper state in quenched qcd @xcite .
on the other hand , the @xmath43 interpolator , which vanishes in the non - relativistic limit , couples strongly to higher energy states .
the interpolator @xmath44 is the time component of the local spin-@xmath45 isospin-@xmath40 interpolator which also couples to spin-@xmath40 states .
it provides a different linear combination of zero- and two - derivative terms . in constructing our correlation matrices
, we first consider an extensive sample of different levels of gauge - invariant gaussian smearing @xcite at the fermion source and sink , including 4 , 9 , 16 , 25 , 35 , 50 , 70 , 100 , 125 , 200 , 400 , 800 and 1600 sweeps . these levels of smearing correspond to rms radii in lattice units ( @xmath46 fm ) of 1.20 , 1.79 , 2.37 , 2.96 , 3.50 , 4.19 , 4.95 , 5.920 , 6.63 , 8.55 , 12.67 , 15.47 and 16.00 .
fig.[fig : eff_kud_013770_013700_n_so_allsmear_x1x1 ] displays effective mass plots , @xmath47 for smeared - source to point - sink correlators of @xmath42 .
the variation in the superposition of excited state contributions is revealed in the different approaches of the effective mass to the ground state plateau where the results converge . from these plots , it is clear that the correlation matrix analysis for excited state contributions will be most effective in the regime @xmath48 where there is diversity in the curves . to explore the low - lying eigenstates of the nucleon spectrum ,
we construct several @xmath49 correlation matrices of @xmath42 as described in table [ table : varieties_of_4x4_matrices ] .
these matrices provide robust results for the lowest three energy eigenstates observed , with the highest energy level accommodating the fourth eigen - energy and any residual strength from higher states not eliminated via euclidean time evolution . , for each set of variational parameters @xmath19 ( major axis ) and @xmath50 ( minor axis ) .
this figure corresponds to the lightest quark mass , @xmath51 for which @xmath52 , and the 3rd basis of table [ table : varieties_of_4x4_matrices].,scaledwidth=46.0% ] the masses from the projected correlation functions obtained from the correlation matrix analysis are very consistent over the variational parameters ( @xmath19,@xmath50 ) as illustrated in fig .
[ fig : ma_kud_013781_4x4_x1x1 ] . careful examination of fig .
[ fig : ma_kud_013781_4x4_x1x1 ] reveals some systematic drift in the second excited state mass at small @xmath50 for @xmath53 this emphasises the preference for selecting larger values of ( @xmath19,@xmath50 ) as discussed in refs .
. however , larger uncertainties are evident for @xmath54 with large @xmath50 due to the suppression of excited states via euclidean time evolution .
we select @xmath19=18 , @xmath50=2 as providing the best balance between these systematic and statistical uncertainties @xcite .
these variational parameters also provide projected correlation functions having favorable @xmath55 in the effective mass fits .
the consistency of the extracted masses from all the @xmath49 matrices considered in table [ table : varieties_of_4x4_matrices ] is illustrated in fig .
[ fig:4x4_x1x1_kud01377000ks01364000 ] .
in particular , the ground and roper states are robust .
both lower and higher smearing radii are beneficial for spanning the space of states at all quark mass .
however , we avoid bases which include extreme smearing counts ( 400 and 800 ) as these often provide ill - defined correlation matrices . hence , we select basis number 3 as the focus of subsequent analysis . in fig .
[ fig : m.12x12 ] , results for the first 12 eigenstate energies are reported for the five quark masses available .
the scale is set via the sommer parameter @xcite .
the 12 states are drawn from three @xmath58 correlation - matrix analyses for pairs of @xmath59 , @xmath60 and @xmath61 .
the matrices are formed with each interpolator having four levels of smearing . whereas the @xmath59 , @xmath60 and @xmath61
, @xmath60 analyses reveal the same spectrum , four new states are revealed in the @xmath59 , @xmath61 analysis in place of the four states dominated by @xmath60 .
the colour coding and symbols for the states of fig . [ fig : m.12x12 ] illustrate the flow of the states from one quark mass to the next as identified by the eigenvector annihilating the state .
the structure of the eigenvectors isolating the states varies only slowly from one quark mass to the next making it easy to trace the propagation of the states from the heavy to the light quark - mass region .
we note that this analysis reveals the same number of energy states as in the physical resonance spectrum below 2.2 gev @xcite . among the most significant results of this investigation
is the manner in which the extracted roper state ( filled triangles ) approaches the physical value .
the significant curvature in the chiral regime indicates the important role played by mesonic dressings of the roper .
non - interacting @xmath3-wave @xmath62 ( @xmath63 ) and @xmath64-wave @xmath65 ( @xmath66 ) threshold scattering states are presented by the dashed and dotted lines , respectively .
for the two large quark masses , the results sit close to the @xmath3-wave @xmath67 scattering threshold whereas the masses for the lighter three quark masses sit much higher . while there is some evidence for the @xmath3- and @xmath64-wave scattering states for the two heavier quarks , there is no evidence of these states at light quark masses .
a possible explanation for this feature is that the attractive mass - dependent and spin - dependent forces which are necessary for the formation of a strong resonance only have sufficient strength at light quark masses .
for example , it is typical to encounter spin - dependent forces which are inversely proportional to the product of the quark masses undergoing gluon exchange . as there is no evidence for the threshold scattering states with back - to - back momenta of one lattice unit , @xmath68 , at the lightest three quark masses
, we find it unlikely that scattering states with the next back - to - back momenta , @xmath69 would suddenly appear in our spectrum . at light quark masses , resonant eigenstates dominated by single - particle states
dominate the spectral function whereas at heavy quark masses , only the multiparticle states have spectral strength sufficient to be seen in the spectrum .
in addition to this quark - mass effect , there is a well - known volume effect .
the couplings to the multi - particle meson - baryon states are suppressed by @xmath70 relative to states dominated by a single - particle state . on our relatively large volume ,
it is likely that multi - particle states will be suppressed and missed in our spectrum , particularly at lighter quark masses where the quark - mass effect also acts to suppress the spectral strength .
further analysis of finite volume effects @xcite on the spectrum is desirable .
future calculations should also investigate the use of five - quark operators to ensure better overlap with the multi - particle states .
this type of novel work using the stochastic laph method is in progress @xcite . indeed ,
complete knowledge of the spectrum is required for a definitive determination of the properties of the roper resonance .
[ fig : m.4x4_dyn_quen_pswave ] provides a comparison of our results in full qcd with earlier results in quenched qcd @xcite , where the effects of dynamical quark loops are not considered .
while the ground state of the nucleon in quenched ( open symbols ) and full qcd ( full symbols ) are in reasonable agreement , significant differences are observed for the roper in the light quark mass regime .
this is an important discovery emphasizing the role of dynamical fermion loops in the structure of the roper .
note , the difference between the @xmath3-wave @xmath71 threshold scattering state energies in full and quenched qcd is due to the difference in the lattice volume .
we have also compared @xcite our extracted nucleon spectrum with the hsc collaboration s
@xcite results at @xmath72 @xmath73 where multi - particle states are seen in our analysis . taking into account their small - volume of @xmath74 , which shifts the energy of the @xmath3-wave scattering states for example ,
both spectra are in good qualitative agreement . of particular note
is the identification of the same number of energy states below 3 gev .
this investigation is the first to illustrate the manner in which the roper resonance of nature manifests itself in today s best numerical simulations of qcd .
the quark mass dependence of the state revealed herein substantiates the essential role of dynamical fermions and their associated non - trivial light - mesonic dressings of baryons , which give rise to significant chiral non - analytic curvature in the roper mass in the chiral regime .
this research was undertaken on the nci national facility in canberra , australia , which is supported by the australian commonwealth government .
we also acknowledge eresearch sa for generous grants of supercomputing time .
this research is supported by the australian research council . | the low - lying even - parity states of the nucleon are explored in lattice qcd using the pacs - cs collaboration 2 + 1-flavor dynamical - qcd gauge - field configurations made available through the international lattice datagrid ( ildg ) .
the established correlation - matrix approach is used , in which various fermion source and sink smearings are utilized to provide an effective basis of interpolating fields to span the space of low - lying energy eigenstates . of particular interest
is the nature of the first excited state of the nucleon , the @xmath0 roper resonance of @xmath1 pion - nucleon scattering .
the roper state of the present analysis approaches the physical mass , displaying significant chiral curvature at the lightest quark mass .
these full qcd results , providing the world s first insight into the nucleon mass spectrum in the light - quark regime , are significantly different from those of quenched qcd and provide interesting insights into the dynamics of qcd .
roper resonance , dynamical fermions , lattice qcd 11.15.ha , 12.38.gc , 12.38.-t , 13.75.gx |
a specific model of nonlinear electrodynamics was proposed by born and infeld in 1934 @xcite founded on a principle of finiteness , namely , that a satisfactory theory should avoid physical quantities becoming infinite .
the born - infeld model was inspired mainly to remedy the fact that the standard picture of a point particle possesses an infinite self - energy , and consisted on placing an upper limit on the electric field strength and considering a finite electron radius .
later , plebaski presented other examples of nonlinear electrodynamic lagrangians @xcite , and showed that the born - infeld theory satisfies physically acceptable requirements .
a further discussion of these properties can be found in ref .
furthermore , a recent revival of nonlinear electrodynamics has been verified , mainly due to the fact that these theories appear as effective theories at different levels of string / m - theory , in particular , in d@xmath3branes and supersymmetric extensions , and non - abelian generalizations ( see ref .
@xcite for a review ) .
much interest in nonlinear electrodynamic theories has also been aroused in applications to cosmological models , in particular , in explaining the inflationary epoch and the late accelerated expansion of the universe @xcite . in this cosmological context , an inhomogeneous and anisotropic nonsingular model for the universe , with a born - infeld field
was studied @xcite , the effects produced by nonlinear electrodynamics in spacetimes conformal to bianchi metrics were further analyzed @xcite , and geodesically complete bianchi spaces were also found @xcite .
homogeneous and isotropic cosmological solutions governed by the non - abelian born - infeld lagrangian @xcite , and anisotropic cosmological spacetimes , in the presence of a positive cosmological constant @xcite , were also extensively analyzed .
in fact , it is interesting to note that the first _ exact _
regular black hole solution in general relativity was found within nonlinear electrodynamics @xcite , where the source is a nonlinear electrodynamic field satisfying the weak energy condition , and the maxwell field is reproduced in the weak limit .
it was also shown that general relativity coupled to nonlinear electrodynamics leads to regular magnetic black holes and monopoles @xcite , and regular electrically charged structures , possessing a regular de sitter center @xcite , and the respective stability of these solutions was further explored in ref .
@xcite .
recently , an alternative model to black holes was proposed , in particular , the gravastar picture @xcite , where there is an effective phase transition at or near where the event horizon is expected to form , and the interior is replaced by a de sitter condensate .
the gravastar model has no singularity at the origin and no event horizon , as its rigid surface is located at a radius slightly greater than the schwarzschild radius . in this context
, a gravastar model within nonlinear electrodynamics , where the interior de sitter solution is substituted with a born - infeld lagrangian , was also found .
this solution was denoted as a born - infeld phantom gravastar @xcite .
relatively to wormhole spacetimes @xcite , an important and intriguing challenge is the quest to find a realistic matter source that will support these exotic geometries .
the latter are supported by _ exotic matter _ , involving a stress energy tensor that violates the null energy condition ( nec ) , i.e. , @xmath4 , where @xmath5 is the stress - energy tensor and @xmath6 any null vector .
several candidates have been proposed in the literature , for instance , to cite a few , null energy condition violating massless conformally coupled scalar fields supporting self - consistent classical wormholes @xcite ; the extension of the morris - thorne wormhole with the inclusion of a cosmological constant @xcite ; and more recently , the theoretical realization that wormholes may be supported by exotic cosmic fluids , responsible for the accelerated expansion of the universe , such as phantom energy @xcite and the generalized chaplygin gas @xcite .
it is also interesting to note that an effective wormhole geometry for an electromagnetic wave can appear as a result of the nonlinear character of the field @xcite .
in ref @xcite , evolving @xmath0 and @xmath1dimensional wormhole spacetimes , conformally related to the respective static geometries , within the context of nonlinear electrodynamics were also explored .
it was found that for the specific @xmath1dimensional spacetime , the einstein field equation imposes a contracting wormhole solution and the obedience of the weak energy condition .
furthermore , in the presence of an electric field , the latter presents a singularity at the throat .
however , a regular solution was found for a pure magnetic field .
for the @xmath2dimensional case , it was also found that the physical fields are singular at the throat .
thus , taking into account the principle of finiteness , that a satisfactory theory should avoid physical quantities becoming infinite , one may rule out evolving @xmath1dimensional wormhole solutions , in the presence of an electric field , and the @xmath2dimensional case coupled to nonlinear electrodynamics . in this work
we shall be interested in exploring the possibility that nonlinear electrodynamics may support static , spherically symmetric and stationary , axisymmetric traversable wormhole geometries .
in fact , bronnikov @xcite showed that static and spherically symmetric @xmath1dimensional wormholes is not the case , and we shall briefly reproduce and confirm this result .
we further consider the @xmath2dimensional case , which proves to be extremely interesting , as the principle of finiteness is imposed , in order to obtain regular physical fields at the throat .
we shall next analyze the @xmath0 and @xmath1dimensional stationary and axisymmetric case @xcite coupled to nonlinear electrodynamics .
this paper is outlined in the following manner : in sec .
[ staticwh ] we analyze @xmath0 and @xmath1dimensional static and spherically symmetric wormholes coupled with nonlinear electrodynamics , and in sec .
[ rotwh ] rotating traversable wormholes in the context of nonlinear electrodynamics are studied . in sec .
[ conclusion ] we conclude .
in this section , we shall be interested in @xmath2dimensional general relativity coupled to nonlinear electrodynamics .
we will use geometrized units throughout this work , i.e. , @xmath7 .
the respective action is given by @xmath8\,d^3x \,,\ ] ] where @xmath9 is the ricci scalar and @xmath10 is a gauge - invariant electromagnetic lagrangian , which we shall leave unspecified at this stage , depending on the invariant @xmath11 given by @xmath12 .
@xmath13 is the electromagnetic field .
note that the factor @xmath14 , in the action , is maintained to keep the parallelism with @xmath1dimensional theory @xcite . in einstein - maxwell theory , the lagrangian is defined as @xmath15 , but here we consider more general choices of electromagnetic lagrangians , however , depending on the single invariant @xmath11 .
it is perhaps important to emphasize that we do not consider the case where @xmath16 depends on the invariant @xmath17 , where @xmath18 denotes the hodge dual with respect to @xmath19 .
varying the action with respect to the gravitational field provides the einstein tensor @xmath20 where @xmath21 .
clearly , the stress - energy tensor is given by @xmath22 where the einstein field equation is defined as @xmath23 .
the variation of the action with respect to the electromagnetic potential @xmath24 , yields the electromagnetic field equations @xmath25 where the semi - colon denotes a covariant derivative .
the spacetime metric representing a spherically symmetric and static @xmath2dimensional wormhole is given by @xmath26 where @xmath27 and @xmath28 are functions of the radial coordinate , @xmath29 .
@xmath27 is denoted as the redshift function , for it is related to the gravitational redshift ; @xmath28 is called the form function @xcite . the radial coordinate has a range that increases from a minimum value at @xmath30 , corresponding to the wormhole throat , to @xmath31 . for the wormhole to be traversable , one must demand the absence of event horizons , which are identified as the surfaces with @xmath32 , so that @xmath27 must be finite everywhere .
a fundamental property of wormhole physics is the flaring out condition , which is deduced from the mathematics of embedding , and is given by @xmath33 @xcite .
note that at the throat @xmath34 , the flaring out condition reduces to @xmath35 .
the condition @xmath36 is also imposed .
taking into account the symmetries of the geometry , we shall consider the following electromagnetic tensor @xmath37 note that the only non - zero terms for the electromagnetic tensor are the following @xmath38 and @xmath39 .
the invariant @xmath40 is given by @xmath41 \,.\end{aligned}\ ] ] the electromagnetic field equation , eq .
( [ em - field ] ) , provides the following relationships @xmath42 where the constants of integration @xmath43 and @xmath44 are related to the electric and magnetic charge , @xmath45 and @xmath46 , respectively .
the mathematical analysis and the physical interpretation will be simplified using a set of orthonormal basis vectors .
these may be interpreted as the proper reference frame of a set of observers who remain at rest in the coordinate system @xmath47 , with @xmath48 fixed .
now , the non - zero components of the einstein tensor , @xmath49 , in the orthonormal reference frame , are given by @xmath50 \label{gpp}\,.\end{aligned}\ ] ] the einstein field equation , @xmath51 , requires that the einstein tensor be proportional to the stress - energy tensor , so that in the orthonormal basis the latter must have an identical algebraic structure as the einstein tensor components , @xmath49 , i.e. , eqs .
( [ gtt])-([gpp ] ) .
recall that a fundamental condition in wormhole physics is the violation of the nec , which is defined as @xmath52 , where @xmath53 is _ any _ null vector . considering the orthonormal reference frame with @xmath54
, we have @xmath55 \,.\ ] ] using the flaring out condition of the throat , @xmath56 @xcite , and considering the finite character of @xmath27 , we verify that evaluated at the throat the nec is violated , i.e. , @xmath57 .
matter that violates the nec is denoted as _
exotic matter_. the only non - zero components of @xmath58 , taking into account eq .
( [ stress - energy ] ) , are @xmath59 we need to impose the conditions @xmath60 and @xmath61 as @xmath62 , to ensure the regularity of the stress - energy tensor components . note that the lagrangian may be obtained from the following relationship : @xmath63 , and using the einstein field equation , is given by @xmath64 -\frac{b'r - b}{2r^3 } \bigg\ } \,.\end{aligned}\ ] ] however , from the metric ( [ metricwormhole ] ) we verify the following zero components of the einstein tensor : @xmath65 , @xmath66 and @xmath67 .
thus , through the einstein field equation , a further restriction may be obtained from @xmath68 , i.e. , @xmath69 which imposes that @xmath70 or @xmath71 , considering the non - trivial case of @xmath72 non - zero .
it is rather interesting that both @xmath73 and @xmath74 can not coexist simultaneously , in the present @xmath2dimensional case . for the specific case of @xmath71 , from eqs .
( [ gtt])-([grr ] ) and eqs .
( [ tttnle])-([trrnle ] ) , we verify the following condition @xmath75 which may be integrated to yield the solution @xmath76 this corresponds to a non - traversable wormhole solution , as it possesses an event horizon at the throat , @xmath77 .
now , consider the case of @xmath70 and @xmath78 .
for this case we have @xmath79 , and the respective einstein tensor components , eqs .
( [ grr])-([gpp ] ) , provide the following differential equation @xmath80 considering a specific choice of @xmath28 or @xmath27 , one may , in principle , obtain a solution for the geometry .
equation ( [ diffeq ] ) may be formally integrated to yield the following general solution @xmath81 \,,\ ] ] where @xmath82 and @xmath83 are constants of integration .
for instance , consider a constant form function , @xmath84 , so that from eq .
( [ gen : phi ] ) , we deduce @xmath85\big\ } \,,\end{aligned}\ ] ] which at the throat reduces to @xmath86\,.\ ] ] to obtain a regular solution at the throat , we impose the condition : @xmath87 .
consider for instance @xmath88 , then eq . ( [ gen : phi ] ) provides the solution @xmath89\bigg\ } \,,\end{aligned}\ ] ] which at the throat , reduces to @xmath90\,.\ ] ] once again , to ensure a regular solution , we need to impose the following condition : @xmath91 .
note that these specific solutions are not asymptotically flat , however , they may be matched to an exterior vacuum spacetime , much in the spirit of refs .
@xcite .
however , a subtlety needs to be pointed out .
consider eqs .
( [ gtt])-([grr ] ) and ( [ tttnle])-([trrnle ] ) , from which we deduce @xmath92 now , taking into account eq .
( [ emf : magnetic ] ) , we find the following relationships for the magnetic field , @xmath74 , and for @xmath72 @xmath93 \,,\ ] ] and @xmath94 respectively . considering that the redshift @xmath95 be finite throughout the spacetime , one immediately verifies that the magnetic field @xmath74 is singular at the throat , which is transparent considering the first term in square brackets in the right hand side of eq .
( [ magnetic - field ] ) .
this is an extremely troublesome aspect of the geometry , as in order to construct a traversable wormhole , singularities appear in the physical fields .
this aspect is in contradiction to the model construction of nonlinear electrodynamics , founded on a principle of finiteness , that a satisfactory theory should avoid physical quantities becoming infinite @xcite .
thus , one should impose that these physical quantities be non - singular , and in doing so , we verify that the general solution corresponds to a non - traversable wormhole geometry .
this may be verified by integrating eq .
( [ dphi ] ) , which yields the following general solution @xmath96 we have considered the factor @xmath97 as @xmath98 , to ensure the regularity of the term in the exponential . however , this solution corresponds to a non - traversable wormhole solution , as it possesses an horizon at the throat , @xmath99 .
one may also prove the non - existence of @xmath2dimensional static and spherically symmetric traversable wormholes in nonlinear electrodynamics , through an analysis of the nec violation . in the context of nonlinear electrodynamics , and taking into account eqs .
( [ tttnle])-([trrnle ] ) , we verify @xmath100 which evaluated at the throat , considering the regularity of @xmath101 and @xmath72 , is identically zero , i.e. , @xmath102 .
the nec is not violated at the throat , so that the flaring - out condition is not satisfied , showing , therefore , the non - existence of @xmath2dimensional static and spherically symmetric traversable wormholes in nonlinear electrodynamics .
the action of @xmath1dimensional general relativity coupled to nonlinear electrodynamics is given by @xmath8\,d^4x \,,\ ] ] where @xmath9 is the ricci scalar and the gauge - invariant electromagnetic lagrangian , @xmath10 , depends on a single invariant @xmath11 @xcite , defined by @xmath103 , as before .
we shall not consider the case where @xmath16 depends on the invariant @xmath17 , as mentioned in the @xmath2dimensional case .
varying the action with respect to the gravitational field provides the einstein field equations @xmath23 , where the stress - energy tensor is given by @xmath104 taking into account the symmetries of the geometry , the only non - zero compatible terms for the electromagnetic tensor are @xmath105 and @xmath106 .
the spacetime metric representing a spherically symmetric and static @xmath1dimensional wormhole takes the form @xcite @xmath107 the non - zero components of the einstein tensor , given in an orthonormal reference frame , are given by @xmath108 \label{ptwh}\,.\end{aligned}\ ] ] its a simple matter to prove that for this geometry , the nec is identical to eq .
( [ necthroat ] ) , and is also violated at the throat , i.e. , @xmath57 .
the relevant components for the stress - energy tensor , regarding the analysis of the nec , are the following @xmath109 analogously with the @xmath2dimensional case , we will consider @xmath110 , as @xmath62 , to ensure that the stress - energy tensor components are regular . from eqs .
( [ rhowh])-([prwh ] ) and eqs .
( [ 4tttnle])-([4trrnle ] ) , we verify the following condition @xmath75 which may be integrated to yield the solution @xmath111 , rendering a non - traversable wormhole solution , as it possesses an event horizon at the throat , @xmath77 .
note that the nec , for the stress - energy tensor defined by ( [ 4dim - stress - energy ] ) , is identically zero for arbitrary @xmath29 , i.e. , @xmath112 .
in particular this implies that the flaring - out condition of the throat is not satisfied , showing , therefore , the non - existence of @xmath1dimensional static and spherically symmetric traversable wormholes coupled to nonlinear electrodynamics .
the analysis outlined in this section is consistent with that of refs .
@xcite , where it was pointed out that nonlinear electrodynamics , with any lagrangian of the form @xmath10 , coupled to general relativity can not support static and spherically symmetric @xmath1dimensional traversable wormholes .
the impediment to the construction of traversable wormholes may be overcome by considering a non - interacting anisotropic distribution of matter coupled to nonlinear electrodynamics .
this may be reflected by the following superposition of the stress - energy tensor @xmath113 where @xmath114 is given by eq .
( [ 4dim - stress - energy ] ) , and @xmath115 is provided by @xmath116 @xmath117 is the four - velocity and @xmath118 is the unit spacelike vector in the radial direction .
@xmath119 is the energy density , @xmath120 is the radial pressure measured in the direction of @xmath118 , and @xmath121 is the transverse pressure measured in the orthogonal direction to @xmath118 .
now , the nec takes the form @xmath122 \,,\end{aligned}\ ] ] which evaluated at the throat , reduces to the nec violation analysis of ref .
@xcite , i.e. , @xmath123 .
we now analyze nonlinear electrodynamics coupled to a stationary axisymmetric @xmath2dimensional wormhole geometry .
the stationary character of the spacetime implies the presence of a time - like killing vector field , generating invariant time translations .
the axially symmetric character of the geometry implies the existence of a spacelike killing vector field , generating invariant rotations with respect to the angular coordinate @xmath124 .
consider the metric @xmath125 where @xmath126 and @xmath127 are functions of @xmath29 .
@xmath128 may be interpreted as the angular velocity @xmath129 of a particle .
@xmath130 is the analog of the redshift function in eq .
( [ metricwormhole ] ) and is finite and nonzero to ensure that there are no event horizons .
we shall also assume that @xmath131 is a positive , nondecreasing function of @xmath29 that determines the proper radial distance @xmath9 , i.e. , @xmath132 to transform to an orthonormal reference frame , the one - forms in the orthonormal basis transform as @xmath133 .
the metric ( [ rwhmn ] ) can be diagonalized @xmath134 by means of the tetrad @xmath135 now , @xmath136 and @xmath137 is defined as @xmath138 \label{tranfs3}\,.\ ] ] from the latter transformation , one may deduce the orthonormal basis vectors , @xmath139 , given by @xmath140 using the fact that @xmath141 , we have @xmath142 . using the einstein field equation , @xmath143 and taking into account the null vector @xmath144
, we obtain the following relationship , at the throat @xmath145 we verify that the nec is clearly violated because the conditions @xmath146 and @xmath147 are imposed by construction @xcite , so that the metric ( [ rwhmn ] ) can describe a wormhole type solution .
now for nonlinear electrodynamics , we consider the stress energy tensor given by eq .
( [ stress - energy ] ) , where the nonzero components of the electromagnetic tensor are @xmath148 which are only functions of the radial coordinate @xmath29 .
then , using the orthonormal reference frame , the nec takes the following form @xmath149\frac{l_{f}}{r^2k^2n^2 } \,,\ ] ] which , at the throat , reduces to @xmath150 from this relationship , we verify that the nec is violated at the throat only if the derivative @xmath72 is positive and @xmath151 is nonzero .
the latter condition , @xmath152 , is imposed to have a compatibility of eqs .
( [ gnec ] ) and ( [ 2setnect ] ) . however , note that from the metric ( [ rwhmn ] ) we verify the following zero components of the einstein tensor : @xmath153 and @xmath154 , implying that @xmath155 and @xmath156 .
these stress - energy tensor components are given by @xmath157 from these conditions , considering that the derivative @xmath72 be finite and positive and the non - trivial case @xmath152 , we find that @xmath158 from the above imposition we deduce @xmath159 , implying the presence of an event horizon , showing the non - existence of @xmath2dimensional stationary and axially symmetric traversable wormholes coupled to nonlinear electrodynamics . now , consider the stationary and axially symmetric @xmath1dimensional spacetime , and analogously to the previous case , it possesses a time - like killing vector field , which generates invariant time translations , and a spacelike killing vector field , which generates invariant rotations with respect to the angular coordinate @xmath124 .
we have the following metric @xmath160\ ] ] where @xmath130 , @xmath161 , @xmath162 and @xmath163 are functions of @xmath29 and @xmath164 @xcite .
@xmath165 may be interpreted as the angular velocity @xmath129 of a particle that falls freely from infinity to the point @xmath166 . for simplicity
, we shall consider the definition @xcite @xmath167 which is well suited to describe a traversable wormhole .
assume that @xmath168 is a positive , nondecreasing function of @xmath29 that determines the proper radial distance @xmath9 , i.e. , @xmath169 and @xmath170 @xcite , as for the @xmath2dimensional case .
we shall adopt the notation that the subscripts @xmath171 and @xmath172 denote the derivatives in order of @xmath29 and @xmath173 , respectively @xcite .
we shall also write down the contravariant metric tensors , which will be used later , and are given by @xmath174 note that an event horizon appears whenever @xmath159 @xcite .
the regularity of the functions @xmath130 , @xmath127 and @xmath161 are imposed , which implies that their @xmath164 derivatives vanish on the rotation axis , @xmath175 , to ensure a non - singular behavior of the metric on the rotation axis .
the metric ( [ 3rwh ] ) reduces to the morris - thorne spacetime metric ( [ metricwormhole ] ) in the limit of zero rotation and spherical symmetry @xmath176 in analogy with the morris - thorne case , @xmath177 is identified as the wormhole throat , and the factors @xmath130 , @xmath161 and @xmath162 are assumed to be well - behaved at the throat .
the scalar curvature of the space - time ( [ 3rwh ] ) is extremely messy , but at the throat @xmath77 simplifies to @xmath178_r % + \frac{\sin^2\theta\,\omega_\theta^2}{2n^2 } \nonumber \\ % & & + \frac{2}{r^2k^4}\,(k^2+k_\theta^2 ) \,.\end{aligned}\ ] ] the only troublesome terms are the ones involving the terms with @xmath179 and @xmath180 , i.e. , @xmath181 note that one needs to impose that @xmath182 and @xmath183 at the throat to avoid curvature singularities .
this condition shows that the throat is located at a constant value of @xmath29 .
thus , one may conclude that the metric ( [ 3rwh ] ) describes a rotating wormhole geometry , with an angular velocity @xmath162 .
the factor @xmath161 determines the proper radial distance .
@xmath130 is the analog of the redshift function in eq .
( [ 4metricwormhole ] ) and is finite and nonzero to ensure that there are no event horizons or curvature singularities .
@xmath127 is the shape function which satisfies @xmath184 ; it is independent of @xmath164 at the throat , i.e. , @xmath182 ; and obeys the flaring out condition @xmath185 . in the context of nonlinear electrodynamics
, we consider the stress energy tensor defined in eq .
( [ 4dim - stress - energy ] ) .
the nonzero components of the electromagnetic tensor are @xmath186 which are functions of the radial coordinate @xmath29 and the angular coordinate @xmath164 .
the analysis is simplified using an orthonormal reference frame , with the following orthonormal basis vectors @xmath187 now the einstein tensor components are extremely messy , but assume a more simplified form using the orthonormal reference frame and evaluated at the throat .
they have the following non - zero components @xmath188 note that the component @xmath189 is zero at the throat , however , we have included this term , as it shall be helpful in the analysis of the stress - energy tensor components , outlined below . using the einstein field equation , the components @xmath190 and @xmath191 have the usual physical interpretations , and
in particular , @xmath192 characterizes the rotation of the matter distribution .
it is interesting to note that constraints on the geometry , placing restrictions on the stress energy tensor needed to generate a general stationary and axisymmetric spacetime , were found in ref .
@xcite . taking into account the einstein tensor components above
, the nec at the throat is given by @xmath193 rather than reproduce the analysis here , we refer the reader to ref .
@xcite , where it was shown that the nec is violated in certain regions , and is satisfied in others .
thus , it is possible for an infalling observer to move around the throat , and avoid the exotic matter supporting the wormhole .
however , it is important to emphasize that one can not avoid the use of exotic matter altogether . using the stress - energy tensor , eq .
( [ 4dim - stress - energy ] ) , we verify the following relationship @xmath194\frac{l_{f}}{r^2k^2n^2\sin^2\theta } \,,\end{aligned}\ ] ] which evaluated at the throat reduces to @xmath195 \frac{l_{f}}{r_0 ^ 2k^2n^2\sin^2\theta } \,.\ ] ] note that for this expression to be compatible with eq .
( [ nec ] ) , @xmath72 may be either positive , negative or zero .
the non - zero components of the einstein tensor are precisely the components expressed in eqs .
( [ rotgtt])-([rotgtphi ] ) , so that through the einstein field equation , we have the following zero components for the stress energy tensor in the @xmath1dimensional case : @xmath196 .
thus , taking into account this fact , and considering that @xmath72 is regular , we obtain the following relationships @xmath197 now , rewriting eqs .
( [ 1])-([2 ] ) in order of @xmath198 and @xmath199 , respectively , and introducing these in eq .
( [ 3 ] ) , we finally arrive at the following relationship @xmath200 from which we obtain that @xmath201 , implying that @xmath202 for all values of @xmath29 except at the throat . however , this restriction is in clear contradiction with the definition of a traversable wormhole , where the condition @xmath36 is imposed .
the same restriction can be inferred from the electromagnetic field equations @xmath203 , which provide the following relationships @xmath204_{,r}+f^{\theta t}[\ln(l_{f})]_{,\theta}=-f^{\mu t}{}_{;\mu } \ , , \\
\label{3eqlfphi } & & f^{r\phi}[\ln(l_{f})]_{,r}+f^{\theta\phi}[\ln(l_{f})]_{,\theta } = -f^{\mu\phi}{}_{;\mu } \,.\end{aligned}\ ] ] these can be rewritten in the following manner @xmath205_{,r}=\frac{f^{\theta t}f^{\mu\phi}{}_{;\mu}-f^{\theta \phi}f^{\mu t}{}_{;\mu}}{f^{rt}f^{\theta\phi}-f^{r
\phi}f^{\theta t } } \ , , \\ & & [ \ln(l_{f})]_{,\theta}=\frac{f^{r\phi}f^{\mu t } { } _ { ; \mu}-f^{rt}f^{\mu \phi}{}_{;\mu}}{f^{rt}f^{\theta\phi}-f^{r \phi}f^{\theta t } } \,.\end{aligned}\ ] ] now , a crucial point to note is that to have a solution , the term in the denominator , @xmath206 , should be non - zero , and can be expressed as @xmath207 \times \nonumber \\ & & \times ( f_{tr}f_{\phi\theta}-f_{t\theta}f_{\phi r } ) \,.\end{aligned}\ ] ] however , using the eqs .
( [ 1])-([4 ] ) , we may obtain an alternative relationship , given by @xmath208 confronting both relationships , we obtain that @xmath201 , implying that @xmath202 for all values of @xmath29 except at the throat , which , as before , is in clear contradiction with the definition of a traversable wormhole . if we consider the individual cases of @xmath209 , @xmath210 , @xmath211 or @xmath212 , separately , it is a simple matter to verify that eqs .
( [ 1])-([4 ] ) impose the restriction @xmath213 , which does not satisfy the wormhole conditions . for the specific case of @xmath214 ( with @xmath215 ) , the restrictions @xmath216 are imposed . taking into account these impositions one verifies that from eq .
( [ 3setnect ] ) , we have @xmath217 , which is not compatible with the geometric conditions imposed by eq .
( [ nec ] ) . considering @xmath218 ( with @xmath219 ) , one readily verifies from eqs .
( [ 1])-([4 ] ) the existence of an event horizon , i.e. , @xmath159 , rendering the wormhole geometry non - traversable .
the specific case of @xmath220 and @xmath221 also obeys eqs .
( [ 1])-([4 ] ) , and needs to be analyzed separately . to show that this case is also in contradiction to the wormhole conditions at the throat
, we shall consider the following stress - energy tensor components , for @xmath220 and @xmath221 @xmath222 an important result deduced from the above components is the following @xmath223 for all points in the geometry .
now , considering the einstein tensor components , eqs .
( [ rotgtt])-([rotgphiphi ] ) , evaluated at the throat , along the rotation axis , and using the einstein field equation , we verify the following relationship @xmath224 which is always positive .
we have taken into account that the functions @xmath130 and @xmath161 are regular , so that their @xmath164 derivatives vanish along the rotation axis , @xmath225 , as emphasized above .
therefore , through the einstein field equation , we verify that relationship ( [ einstein - cond ] ) is not compatible with condition ( [ setcond ] ) , and therefore rules out the existence of rotating wormholes for this specific case of @xmath220 and @xmath221 .
in this work we have explored the possibility of the existence of @xmath0 and @xmath1dimensional static , spherically symmetric and stationary , axisymmetric traversable wormholes coupled to nonlinear electrodynamics . for the static and spherically symmetric wormhole spacetimes ,
we have found the presence of an event horizon , and that the nec is not violated at the throat , proving the non - existence of these exotic geometries within nonlinear electrodynamics .
it is perhaps important to emphasize that for the @xmath2dimensional case we found an extremely troublesome aspect of the geometry , as in order to construct a traversable wormhole , singularities appear in the physical fields .
this particular aspect of the geometry is in clear contradiction to the model construction of nonlinear electrodynamics , founded on a principle of finiteness , that a satisfactory theory should avoid physical quantities becoming infinite @xcite .
thus , imposing that the physical quantities be non - singular , we verify that the general solution corresponds to a non - traversable wormhole geometry .
we also point out that the non - existence of @xmath1dimensional static and spherically symmetric traversable wormholes is consistent with previous results @xcite . for the @xmath2dimensional stationary and axisymmetric wormhole
, we have verified the presence of an event horizon , rendering a non - traversable wormhole geometry .
relatively to the @xmath1dimensional stationary and axially symmetric wormhole geometry , we have found that the field equations impose specific conditions that are incompatible with the properties of wormholes .
thus , we have showed that for the general cases of solutions outlined above the non - existence of traversable wormholes within the context of nonlinear electrodynamics . nevertheless , it is important to emphasize that regular magnetic time - dependent traversable wormholes do exist coupled to nonlinear electrodynamics @xcite . in the analysis outlined in this paper ,
we have considered general relativity coupled to nonlinear electrodynamics , with the gauge - invariant electromagnetic lagrangian @xmath10 depending on a single invariant @xmath11 given by @xmath226 .
an interesting issue to pursue would be the inclusion , in addition to @xmath11 , of another electromagnetic field invariant @xmath227 .
this latter inclusion would possibly add an interesting analysis to the solutions found in this paper .
we thank ricardo garca salcedo and nora bretn for extremely helpful comments and suggestions .
m. born , `` on the quantum theory of the electromagnetic field , '' proc .
a143 * , 410 ( 1934 ) ; m. born and l. infeld , `` foundations of the new field theory , '' proc .
roy . soc . * a144 * , 425 ( 1934 ) .
i. bialynicki - birula , `` nonlinear electrodynamics : variations on a theme by born and infeld , '' in _ j. lopuszanki s festschrift , quantum theory of particles and fields _ , eds
. b. jancewicz and j. lukierski , singapore , world scientific 31 , ( 1983 ) .
v. v. dyadichev , d. v. galtsov , a. g. zorin and m. yu .
zotov , `` non - abelian born infeld cosmology , '' phys .
d * 65 * , 084007 ( 2002 ) [ arxiv : hep - th/0111099 ] .
d. n. vollick , `` anisotropic born - infeld cosmologies , '' gen .
. grav . * 35 * , 1511 - 1516 ( 2003 ) [ arxiv : hep - th/0102187 ] .
e. ayn - beato and a. garca , `` regular black hole in general relativity coupled to nonlinear electrodynamics , '' phys .
lett . * 80 * , 5056 - 5059 ( 1998 ) [ arxiv : gr - gc/9911046 ] .
e. ayn - beato and a. garca , `` new regular black hole solution from non - linear electrodynamics , '' phys .
b * 464 * , 25 ( 1999 ) [ arxiv : hep - th/9911174 ] ; e. ayn - beato and a. garca , `` non - singular charged black hole solutions for non - linear source , '' gen .
* 31 * , 629 - 633 ( 1999 ) [ arxiv : gr - gc/9911084 ] .
k. a. bronnikov , `` regular magnetic black holes and monopoles from nonlinear electrodynamics , '' phys .
d * 63 * , 044005 ( 2001 ) [ arxiv : gr - qc/0006014 ] .
i. dymnikova , `` regular electrically charged structures in nonlinear electrodynamics coupled to general relativity , '' class . quant
* 21 * , 4417 - 4429 ( 2004 ) [ arxiv : gr - qc/0407072 ] .
n. breton , `` stability of nonlinear magnetic black holes , '' phys .
d * 72 * , 044015 ( 2005 ) [ arxiv : hep - th/0502217 ] .
p. o. mazur and e. mottola , `` gravitational condensate stars : an alternative to black holes , '' [ arxiv : gr - qc/0109035 ] ; g. chapline , e. hohlfeld , r. b. laughlin and d. i. santiago , `` quantum phase transitions and the breakdown of classical general relativity , '' int . j. mod .
a 18 3587 - 3590 ( 2003 ) [ arxiv : gr - qc/0012094 ] ; m. visser and d. l. wiltshire , `` stable gravastars - an alternative to black holes ?
, '' class .
* 21 * 1135 ( 2004 ) [ arxiv : gr - qc/0310107 ] ; c. cattoen , t. faber and m. visser , `` gravastars must have anisotropic pressures , '' class .
* 22 * 4189 - 4202 ( 2005 ) [ arxiv : gr - qc/0505137 ] ; g. chapline , `` dark energy stars , '' [ arxiv : astro - ph/0503200 ] ; f. s. n. lobo , `` stable dark energy stars , '' class .
* 23 * , 1525 ( 2006 ) [ arxiv : astro - ph/0508115 ] .
n. bili , g. b. tupper and r. d. viollier , `` born - infeld phantom gravastars , '' jcap 0602 , 013 ( 2006 ) [ arxiv : astro - ph/0503427 ] . c.
barcelo and m. visser , `` traversable wormholes from massless conformally coupled scalar fields , '' phys .
b * 466 * , 127 ( 1999 ) [ arxiv : gr - qc/9908029 ] .
j. p. s. lemos ,
f. s. n. lobo and s. q. de oliveira , `` morris - thorne wormholes with a cosmological constant , '' phys .
d * 68 * , 064004 ( 2003 ) [ arxiv : gr - qc/0302049 ] .
s. sushkov , `` wormholes supported by a phantom energy , '' phys .
d * 71 * , 043520 ( 2005 ) [ arxiv : gr - qc/0502084 ] ; f. s. n. lobo , `` phantom energy traversable wormholes , '' phys .
d * 71 * , 084011 ( 2005 ) [ arxiv : gr - qc/0502099 ] ; f. s. n. lobo , `` stability of phantom wormholes , '' phys .
d * 71 * , 124022 ( 2005 ) [ arxiv : gr - qc/0506001 ] .
f. s. n. lobo , `` chaplygin traversable wormholes , '' phys .
rev . d*73 * , 064028 ( 2006 ) [ arxiv : gr - qc/0511003 ] .
m. novello , v. a. de lorenci , e. elbaz and j. m. salim , `` closed lightlike curves in non - linear electrodynamics , '' [ arxiv : gr - qc/0003073 ] ; f. baldovin , m. novello , s. e. perez bergliaffa and j.m .
salim , `` a nongravitational wormhole , '' class .
* 17 * , 3265 - 3276 ( 2000 ) [ arxiv : gr - qc/0003075 ] .
j. p. s. lemos and f. s. n. lobo , `` plane symmetric traversable wormholes in an anti - de sitter background , '' phys .
d * 69 * , 104007 ( 2004 ) [ arxiv : gr - qc/0402099 ] ; f. s. n. lobo , `` surface stresses on a thin shell surrounding a traversable wormhole , '' class .
* 21 * , 4811 ( 2004 ) [ arxiv : gr - qc/0409018 ] ; f. s. n. lobo , `` energy conditions , traversable wormholes and dust shells , '' gen .
. grav . * 37 * 2023 - 2038 ( 2005 ) [ arxiv : gr - qc/0410087 ] ; f. s. n. lobo and p. crawford , `` stability analysis of dynamic thin shells , '' class .
* 22 * , 4869 ( 2005 ) , [ arxiv : gr - qc/0507063 ] . | in this work we explore the possible existence of static , spherically symmetric and stationary , axisymmetric traversable wormholes coupled to nonlinear electrodynamics . considering static and spherically symmetric @xmath0 and @xmath1dimensional wormhole spacetimes
, we verify the presence of an event horizon and the non - violation of the null energy condition at the throat . for the former spacetime ,
the principle of finiteness is imposed , in order to obtain regular physical fields at the throat .
next , we analyze the @xmath2dimensional stationary and axisymmetric wormhole , and also verify the presence of an event horizon , rendering the geometry non - traversable . relatively to the @xmath1dimensional stationary and axisymmetric wormhole geometry , we find that the field equations impose specific conditions that are incompatible with the properties of wormholes .
thus , we prove the non - existence of the general class of traversable wormhole solutions , outlined above , within the context of nonlinear electrodynamics . |
years after the serendipitous discovery of gamma - ray bursts ( grbs ) by the vela spacecraft ( klebesadel et al .
1973 ) , the nature of the progenitors responsible for generating the initial explosion remains uncertain , while an elegant set of ideas about the production and evolution of the afterglow has developed ( mszros & rees 1997 ; sari & piran 1997 ) .
aside from recent breakthroughs in x - ray spectroscopy of grbs ( piro et al .
2000 ) a large fraction of the observational information about the environments of grbs derives from optical spectroscopy and imaging of the host galaxies .
optical spectroscopy of the integrated light and calibrated emission lines has been used to derive the star - formation rate ( sfr ) for a number of the host galaxies ( see for instance bloom et al .
1998 ; djorgovski et al . 1998 ) . as a complement to spectroscopy , high - resolution optical imaging and astrometry have helped pinpoint two - dimensional locations for grbs with respect to the host galaxy .
the distribution of these grb locations has yielded a number of statistical constraints in the progenitor scenarios ( bloom , kulkarni , & djorgovski 2000 ) .
a promising technique that has been applied with less success so far is the study of time dependence in absorption lines from metal ions ( perna & loeb 1998 ; bttcher et al .
1999 ) or @xmath13 vibrational levels ( draine 2000 ) that might be excited by the uv radiation generated during the evolution of the burst .
the detection of absorption - line variability could provide important clues about the physical dimensions of the photoionized region and the density in the immediate environment of the grb .
absorption - line variability can be quantified by measuring changes in the equivalent widths as a function of time or by identifying the appearance of new absorption features .
vreeswijk et al .
( 2001 ) studied the time evolution of the mg ii doublet in the optical spectra of grb 990510 and grb 990712 but failed to find any significant changes . in this work we present a study of the time evolution of absorption systems in the optical spectrum of grb 010222 , and discuss possible implications for its progenitor environment
the outline of the paper is as follows : 2 describes the optical spectroscopy , 3 describes the absorption line identification and continuum fitting . in 4 we detail column density determinations , arguments in favor of the @xmath14 redshift for grb 010222 , kinematics of the host galaxy , and a study of time evolution of absorption lines .
a description of the photoionization code and results is given in 5 and 6 .
finally , the implications of our results and conclusions are presented in 7 and 8 .
grb 010222 was initially detected by bepposax on ut 2001 feb .
22.30799 ( piro 2001 ) .
an optical counterpart was reported only a few hours later by henden ( 2001a,2001b ) .
our group began spectroscopic observations at the position of the optical counterpart on ut 2001 february 22.66 with the dual - beam low resolution imaging spectrometer ( lris ) on the keck i telescope and obtained a second spectrum on ut 2001 february 23.66 . on both nights
we used a 300 lines / mm grating blazed at 5000 with a @xmath15 wide slit .
the effective spectral resolution fwhm varies from @xmath16 13 on the blue side to @xmath16 11 on the red side .
a total of 1800 s of exposure were obtained on each night .
the data were trimmed , bias - substracted and flat - fielded using standard procedures .
extraction of the spectra was performed using the iraf apall package .
telluric lines were corrected by fitting the continuum including the a and b atmospheric bands of a spectrophotometric standard .
the spectra of spectrophotometric standards feige 67 , hiltner 600 , and hz 44 were used for flux calibration and a hg - kr - ne - ar - xe comparison lamp provided the wavelength calibration .
the wavelength calibrations have a typical accuracy of @xmath17 as determined by the root - mean - square ( rms ) deviation of the individual lines from the smooth function used to fit them .
the large range of precision reflects the presence of fewer arc lines on the blue side .
finally , the lris - b ( blue channel ) and lris - r ( red channel ) spectra were connected at the crossover wavelength of the dichroic beam splitter .
a higher - resolution spectrum of grb 010222 was obtained on ut 2001 feb 23.61 using the echellette spectrograph and imager ( esi ) mounted on the keck ii telescope .
this mode spans ten orders with effective wavelength coverage from 3900 to 10900 .
the spectral resolution is 11.4 km s@xmath18 pix@xmath18 with typical dispersion in the instrument ranging from 0.16 pixel@xmath18 in order 15 to 0.30 pixel@xmath18 in order 6 . in the echellette mode
the instrument provides a wavelength resolution r @xmath16 10000 over the entire spectrum .
a total of 3 @xmath11 1200 second exposures were obtained and reduced using tom barlow s mauna kea echelle extraction program ( makee ) .
makee , originally designed for the high resolution echelle spectrometer ( hires ) , is an automated procedure that allows standard processing as well as wavelength and flux normalization . for the wavelength calibration we used a cuar cathode lamp which provides a large number of unsaturated lines to fit the spectrum with rms @xmath16 0.09 .
the individual spectra at a common epoch were average - combined after running through the pipeline extraction .
we have identified three absorption systems ( bloom et al . 2001 ; castro et al .
2001a ) which were reported independently by jha et al .
the systems are located at @xmath0 , @xmath1 , and @xmath2 , where the redshift reported is the average of individual lines for each system .
the line identification corresponds to a cross - correlation of features in the spectrum with a listing of known absorption lines ( verner , barthel , & tytler 1994 ) assuming possible redshift systems at @xmath19 , @xmath20 , and @xmath21 . in order to weed out spurious features possibly due to sky subtraction ,
only those lines present in three spectra are reported .
moreover , we demanded @xmath22[1.0 + z(1-@xmath23 ) ] @xmath24 @xmath25 @xmath24 @xmath22[1.0 + z(1+@xmath23 ) ] and the corresponding oscillator strength ratios for identified lines . in the previous , @xmath23 stands for a tolerance factor in the line identification and accounts for uncertainties in the wavelength calibration .
figure 1 shows the lris spectra including line identification for each absorption system . the higher - resolution esi spectrum with identified absorption features is shown in figure 2 . in the highest redshift system we find kinematic evidence for two distinct components at @xmath26=1.47590 and @xmath27=1.47688 respectively , with a rest - frame separation of @xmath28 km s@xmath18 .
the kinematic structure of this system will be discussed at greater length in 4.3 .
tables 1 and 2 summarize the line identifications for each setup and list the measured vacuum wavelengths , observed wavelengths converted into vaccum and corrected for heliocentric motion , estimated redshift , oscillator strengths @xmath29 , equivalent widths ( @xmath30 ) in the rest frame for both nights , and error estimates in their equivalent widths .
we have adopted the compilation of wavelengths and oscillator strengths by morton ( 1991 ) with the revisions proposed by savage & sembach ( 1996 ) .
the main source of uncertainty in the equivalent width measurements derives from the identification of the continuum level . in order to compute the errors in the equivalent width for each line we used the iraf splot task which allows error estimates based on a poisson model for the noise . for blended lines ,
iraf splot fits and deblends each line separately using predetermined line profiles .
error estimates for blended lines are computed directly in splot by running a number of monte - carlo simulations based on preset instrumental parameters .
the flux - calibrated spectra were corrected for galactic foreground extinction assuming @xmath31 ( schlegel , finkbeiner , & davis 1998 ) and the extinction curve of cardelli , clayton & mathis ( 1989 ) . after correcting for galactic extinction , the spectra were converted to @xmath32 and @xmath33 units and the continuum divided in equal size bins , excluding regions containing absorption lines .
the statistical 1@xmath34 error was determined by estimating the standard deviation within each bin .
next we found that a power law of the form @xmath35 produced an adequate fit to the optical spectrum .
the best fitted model changed slightly from @xmath36 on feb .
22.66 to @xmath37 on feb .
the continuum points along with the best fitted models are shown in figure 3 . despite the apparent steepening of the spectra ,
there is one important caveat , the difficulty of doing spectrophotometry through a narrow slit . because we have used a 1.5 slit width the throughput of the spectrograph is reduced .
spectrophotometry might also be affected by atmospheric differential refraction ( filippenko 1982 ) . on feb 22.66
this effect was minimized by aligning the slit at the parallactic angle for the grb as well as for the spectrophotometric standards .
moreover all the objects were observed at similar airmasses . during feb 23.66 only one
slit angle was used throughout the night and the airmasses for the spectrophotometric standards were slightly higher at the time of acquisition .
therefore , the uncertainty in the spectrophotometry during the second night still allows the possibility that the spectral index did not change between the two nights .
the difficulty of achieving accurate spectrophotometry in this case gives more weight to the results obtained from simultaneous multi - band photometry .
our derived spectra index @xmath36 appears in agreement with the reported single power law @xmath38 obtained from _
bvri _ optical data ( stanek et al .
moreover the spectral behavior of the lightcurves does not show any steepening of the index @xmath39 as a function of time ( stanek et al .
the previous result is also consistent with @xmath36 derived from a low - resolution spectrum obtained five hours after the burst ( jha et al .
but our spectral index is slightly shallower than the index @xmath40 reported by masetti et al .
one possible cause for the steepening in the masetti et al .
( 2001 ) results is the inclusion of the _ u _ band data in the overall fit ( stanek et al .
the bulk of the evidence seems to argue for a single power law spectral index for the duration of the burst .
hereafter we adopt @xmath36 .
our optical spectral index is similar within uncertainties to the measured x - ray spectral index @xmath41 0.05 reported for the whole bepposax observation ( int zand et al .
2001 ) . however , as pointed out by these authors , the optical flux is too faint or the x - ray flux is too bright to accommodate a simple synchrotron power - law .
one possible explanation for this mismatch is an x - ray excess above the standard synchrotron spectrum due to inverse compton ( ic ) scattering ( sari & esin 2001 ) .
harrison et al . (
2001 ) have recently presented broad - band observations of grb 000926 that are consistent with this ic interpretation .
an alternative explanation for the difference between the optical and x - ray flux is reddening of the optical spectrum .
the steep slope of the continuum fit to the optical spectra in addition to the absence of a 2175 `` bump '' ubiquitous in a galactic extinction curve implies that the extinction curve intrinsic to grb 010222 would have to be approximately flat in order to not redden the spectra dramatically .
this in turn suggests that the dust content at the burst location is relatively low or that any dust present mimics the gray variety ( aguirre 1999 ) .
ir and submillimeter observations of grb 010222 by frail et al .
( 2001b ) concluded that the majority of the reprocessed emission seen at the host galaxy is from a separate starburst component . by their estimation
thermal radiation or dust scattering near the burst site can not reproduce the bulk of the observed sub - millimeter energy . considering the possibility that the host is a starburst galaxy , we estimated the intrinsic reddening using the extinction curve of calzetti et al .
dust corrections are particularly dangerous given the complexity of the dust geometry in any particular line of sight ( witt et al .
nonetheless , even a small color excess , @xmath42 , requires a flatter intrinsic power law for the afterglow with @xmath43 .
this color excess is required to satisfy @xmath44 for @xmath45 in the optical band and @xmath46 at frequencies @xmath47 in the x - ray regime . here
@xmath48 is the index of the power - law electron energy distribution , and @xmath49 is the `` cooling frequency '' at which the electron energy loss time scale is equal to the age of the shock .
the spectrum , corrected for such extinction at the host galaxy , is illustrated in figure 4 .
alternatively , it is possible that @xmath49 has moved below the optical band at the time of our first spectrum , 8.45 hr after the burst .
if this is the case , the reddening argument does not apply ; instead , an ic scattering interpretation is possible .
the statistical properties of metallic absorption systems have been studied extensively at various spectral resolutions ( steidel & sargent 1992 ; churchill et al .
these studies have provided insight into the equivalent width distribution , evolution , and cosmological clustering of absorbers . in the particular case of grb 010222 ,
the equivalent widths of the mg ii doublet in the identified systems indicate that they are unusually strong in comparison to the population of absorbers studied by steidel & sargent ( 1992 ) .
the mg ii doublet ratios @xmath30(mg ii 2796.35)/@xmath30(mg ii 2803.53 ) are 1.06 @xmath50 0.06 , 1.17 @xmath50 0.19 , and 1.46 @xmath50 0.12 for the @xmath19=1.47688 , @xmath20=1.15628 and @xmath21=0.92747 systems , respectively . in each case
the ratio differing from 2 implies that the mg ii lines are strongly saturated .
in addition to the saturation of mg ii lines , we shall show evidence for saturation effects in the majority of fe ii lines in the esi spectrum .
we avoided using these lines for direct column density measurements given the likelihood of a non - standard curve of growth under saturated conditions ( morton & bhavsar 1979 ) . instead , resolved absorption lines in the esi spectrum that are not strongly saturated
are more appropriate for column density determinations .
the mn ii triplet at 2576.88 , 2594.50 , and 2606.46 in the @xmath19=1.47688 system meets this requirement and appears particularly well suited to studying the highest column absorption system at @xmath27 in which it resides .
these lines fall in the region of intermediate optical depth , @xmath51 , _
i.e. _ , on the `` flat '' part of the curve of growth . here
the observed broadening of the mn ii lines is assumed to be due to the random motion of the gas , described by a maxwellian velocity distribution . in this regime
the equivalent width becomes sensitive to the doppler parameter @xmath52 , which can be determined directly from the broadening of the absorption lines .
the measured doppler parameters @xmath52 are listed in table 3 . in order to derive column density @xmath53 for any metal @xmath54
, we used the general prescription for a standard curve of growth ( spitzer 1978 ) @xmath55 @xmath56 where @xmath53 is written as a function of equivalent width @xmath57 , rest wavelength @xmath58 , oscillator strength @xmath29 , optical depth @xmath59 , and its dependence on the maxwellian velocity distribution f(@xmath59 ) .
metallic column densities log(@xmath53 ) as well as equivalent neutral hydrogen column densities log(@xmath60 ) assuming solar abundances ( anders & grevesse 1989 ) are given in table 3 . notice the overall agreement in the column densities derived using three different mn ii lines .
the column densities listed in the fifth and sixth columns of table 3 have not been corrected for depletion into dust grains ( jenkins 1987 ) , which is an uncertain quantity .
waxman & draine ( 2000 ) have shown that dust may be sublimated by a prompt optical / uv flash at any distance closer than @xmath16 10 pc from the burst .
fruchter , krolik & rhoads ( 2001 ) also considered the possibility of grain charging as a mechanism for dust destruction .
this could take place as far away as @xmath16 100 pc from the birthsite .
thus , the combined effect of both destruction mechanisms , temperature of the medium , and the return of dust depleted metals to the gas phase make the depletion correction largely uncertain . for completeness
we make the simplest assumption ignoring any immediate return of metals into the gas phase and adopt a depletion correction of @xmath61 dex for mn@xmath62 based on goddard high - resolution spectrograph ( ghrs ) measurements towards @xmath63 oph obtained by savage & sembach ( 1996 ) .
the resulting corrected values are shown in the last column of table 3 .
given the uncertainty in the amount of depletion it appears safe to regard the uncorrected hi column densities as lower limits .
we note that the weighted average of mn@xmath62 corrected for depletion implies log@xmath64 @xmath65 in the @xmath26 system , which is similar to the redshift - corrected column density log@xmath66 @xmath65 derived by in t zand et al .
( 2001 ) from the x - ray afterglow .
the results presented so far assume solar abundances , alternatively , less than solar abundances for mn ii might accommodate the difference between metal abundances and the values reported by in t zand et al .
( 2001 ) and salamanca et al .
( 2001 ) .
we have completed the estimates of column densities @xmath53 by including blended absorption lines .
the term `` blended '' is used here to refer to the lines where the @xmath26 and @xmath27 systems can not be separated cleanly in the esi spectrum . in particular , by comparing the equivalent widths and the oscillator strength ratios for zn ii ( 2026.14 , 2062.66 ) and cr ii ( 2056.25 , 2062.23 , 2066.16 ) lines we see that they fall on the linear portion of the curve of growth . for these lines
we can obtain a direct measurement of the column density . on the other hand , the fe@xmath62 and mg@xmath62 lines , although structured , are strongly saturated .
for that reason we decided to fit gaussian profiles at the positions of the identified @xmath26 and @xmath27 components while running the deblending routines . in this case the linear part of the curve of growth only provides a lower limit to the column density @xmath53 through the equation @xmath67 the resulting column densities are listed in tables 3 and 4 . in each case
the density corresponds to the average of various single transitions . as a consistency check , the hi column density derived assuming solar abundances for the generally undepleted zn@xmath62 is in agreement with the values cited in table 3 for the resolved mn@xmath62 lines .
the absence of any emission lines in the spectrum slightly complicates the unequivocal identification of this system with the host galaxy .
however , there is information in the optical spectrum that supports this interpretation .
first , the non - detection of a lyman-@xmath68 break places an upper limit for the grb host galaxy of z @xmath69 1.96 .
second , salamanca et al . (
2001 ) have reported evidence of a red wing of ly-@xmath68 absorption at @xmath70 1.476 .
third , the measured equivalent widths of the mg ii doublet and the ratio of mg i to mg ii in this system are also among the highest known for grb afterglows and for the sample of mg ii absorbers compiled by steidel & sargent ( 1992 ) at similar spectral resolution .
fourth , we can quantify the probability of detecting three intervening galaxies in the line of sight .
the lris setup used during the first two nights spans the 36009205 wavelength range with a rest equivalent width threshold limit @xmath71 0.3 , which matches the resolution and equivalent width detection limit of the steidel & sargent survey .
this wavelength coverage can be translated into a redshift range for mg ii lines of @xmath72 , or a redshift interval @xmath73 .
thus , the number of absorbers per unit redshift is n/@xmath74 , which is less stringent than an earlier estimate of this quantity ( jha et al .
2001 ) , but is still larger than the expected number of mg ii absorbers per unit redshift @xmath75 = 0.97 @xmath50 0.10 for a rest equivalent width threshold of 0.3 ( steidel & sargent 1992 ) . the probability for the chance superposition of three absorbers in this line of sight is non - negligible but small .
lastly , if the highest redshift of an mgii system is @xmath76=1.47688 , then , in 80% of the cases , the source could have originated from no further than @xmath77 .
if it originated from a larger distance then , given the rapid increase in the number of metal absorption systems with increasing redshift , a source redshift higher than z=1.8 would be unlikely ( bloom et al .
although not definitive proof , these arguments are most consistent with the @xmath19=1.47688 absorption system as the host galaxy of grb 010222 .
the advantage of using higher spectral resolution is the capability to resolve multiple components along the line of sight . in this instance , we identify two kinematic components at @xmath26=1.47590 and @xmath27=1.47688 in the esi spectrum .
the presence of multiple components immediately raises the question of structure within the host galaxy .
we might be probing a galactic disk accompanied by an additional kinematic component introduced by absorbing clouds in the halo ( disk - halo scenario ) , or instead the two components could be due to absorption arising in distinct regions of a galactic disk ( disk scenario ) . in order to test these scenarios
we also used information about element abundances and dust depletion .
in particular , the ratios [ zn / cr ] and [ zn / mn ] provide a reasonable measure of the gas - to - dust ratio and the temperature of the absorbing medium .
the main line of argument lending support to a disk - halo scenario is the absence of a @xmath26 kinematic component in the resolved set of mn@xmath62 lines .
this is better illustrated in figure 5 , which shows the kinematic components for different metallic lines in velocity space using @xmath4 , which is assumed to be the systemic redshift , as the velocity zero point .
mg@xmath78 , mn@xmath62 and mg@xmath62 , are present in the @xmath27 system , while the @xmath26 component lacks mn@xmath62 but contains mg@xmath62 and some of the structured but likely saturated fe@xmath62 lines .
since the ionization potential of mn@xmath62 is almost identical to that of mg@xmath62 , a possible reason for the absence of mn@xmath62 at @xmath26 is an underabundance of mn in a metal - poor environment like the halo .
this is based on the conjecture that type ia supernovae are an important source of mn ( samland 1998 , nakamura et al .
1999 ) , or that there is a metallicity dependence yield of mn in massive stars ( timmes , woosley , & weaver 1995 ) , as discussed in the case of damped ly-@xmath68 systems by pettini et al .
an early appearance of type ia sne during the halo phase in milky way formation models has been questioned ( chiappini , mateucci , & romano 2001 ) .
if the latter authors are correct , our @xmath26 component might arise in a halo cloud , separated from the velocity zero point by @xmath79 km s@xmath18 , with signatures of mg@xmath62 and fe@xmath62 but underabundant in mn@xmath62 .
the possibility of a disk - halo line of sight makes the abundance ratios for each kinematic component that much more crucial in probing the ism at the host galaxy .
zinc is a fairly reliable metallicity indicator since it resides more than any other metal in the gas phase .
on the other hand , metals like cr and mn are more easily depleted onto grains and a large fraction of their abundance is contained in metallic dust grains in the galaxy . if the @xmath27 component is an absorption feature of the halo then a higher gas - to - dust ratio than in the disk is expected .
unfortunately , the [ zn / mn ] ratio in the @xmath27 component is only an upper limit given that zn@xmath62 lines from both velocity components are blended , as are the cr@xmath62 lines .
we note however that the mn and cr abundances are comparable .
in fact the observed [ zn / cr ] @xmath80 is less than 1.61 @xmath10 [ zn / cr ] @xmath10 1.04 observed towards @xmath63 oph ( savage & sembach 1996 ) and 1.85 @xmath10 [ zn / cr ] @xmath10 0.75 measured in a sample of 20 sightlines observed with hst ( roth & blades 1995 ) . although the [ zn / cr ] ratio is less than in any measured galactic sightline it does not represent a clear discriminant between a halo and a disk or a damped lyman absorber since the spread in [ zn / cr ] is large for either sample ( roth & blades 1995 ) . our results might be an indication of low dust content but not at the extreme reported by salamanca et al .
it seems that their zn ii measurements are systematically lower than those presented here or any other published results for grb 010222 ( see jha et al .
2001 , masetti et al .
in addition , zn ii(2026.14 ) listed at @xmath81 in table 2 of salamanca et al .
( 2001 ) appears to be a misidentification .
we find that line is more consistent with fe ii(2600.17 ) at the @xmath21 system while zn ii should be around @xmath82 at the higher redshift .
for completeness we mention that we have not detected any of the unidentified lines listed in table 3 of salamanca et al .
( 2001 ) in the region covered by our spectra . finally , the reduced [ zn / cr ] ratio in the spectrum might be hinting that we have detected a warm medium in the @xmath27 component where depletion for cr and mn converge ( savage & sembach 1996 ) or else that there is little dust to deplete onto in this environment .
a warmer medium would result in different dust depletion corrections and lower the inferred values of @xmath60 assuming standard solar abundances .
these results are in agreement with the depletion patterns discussed by savaglio et al . 2002 .
we now consider the alternative of a disk scenario . in general , any neutral mg present in a galactic disk must be shielded from external photoionizing flux .
a feasible scenario is that we are probing a line of sight passing through two distinct regions within the disk containing neutral gas
. these regions could be underdepleted molecular clouds cocooning mg in its neutral form .
a possible consequence of uneven photoionization in the disk is the marginal detection of a blue `` wing '' on the @xmath26 component of the mg ii ( 2796.35 ) absorption line ( figure 5 ) .
this feature would argue for a stronger photoionizing source near the closer side of the disk and against a galactic halo .
this option would eliminate the necessity for a halo component , but it does not explain the absence of mn@xmath62 in the @xmath26 system , assuming similar enrichment processes for different parts of the disk . instead , there may be a strong metallicity gradient along the disk , as has been observed along the disk of the milky way ( van steenberg & shull 1988 ) . assuming that a disk scenario is correct and that the [ zn / mn ] ratio peaks towards the inner part of the host galaxy ; the velocity systems might be tracing a fraction of galactic rotation .
independent of the interpretation , the presence of a galactic disk appears necessary to explain the properties of the strongest absorption system .
this furthers the connection of long - duration grbs with a disk progenitor .
the impact of the associated uv flash on @xmath5 molecules in the vicinity of a grb has been determined by draine ( 2000 ) .
the direct observable consequence of strong uv radiation from a grb on the surrounding @xmath5 is the production of vibrationally excited levels which could create strong line absorption in the region 912 @xmath69 @xmath83 @xmath69 1650 , and reradiated fluorescent emission in a similar range . for a redshift of z
= 1.477 this corresponds to the region @xmath84 @xmath69 4087 .
unfortunately the region where the absorption effect is strongest at @xmath85 falls outside our spectral coverage .
thus , we are only able to probe a fraction of the absorption region near the blue end of the spectrum where the quantum efficiency is low and the noise is high .
the strength of the absorption depends on the amount of @xmath5 available for vibrational excitation . at the lris resolution
we find no definite evidence of @xmath5 absorption for @xmath84 @xmath69 4087 to make a proper determination of @xmath86 .
however , our spectra probably rule out excited states of @xmath5 at @xmath86 @xmath10 @xmath87 @xmath65 based on the transmission plots of draine ( 2000 ) .
if the line absorption occurs sufficiently close to the burst site , the afterglow emission could manifest itself through photoionization of the grb environment , thus , variability of the absorption features . to detect this time - dependent effect various snapshots of the absorption spectrum
are required . in this particular instance
we take advantage of the low - resolution spectra obtained over two consecutive nights to search for the predicted effect .
the measured equivalent widths for each night as well as the statistical uncertainties are listed in table 1 .
comparison between the lris equivalent widths on consecutive nights shows no strong evidence for time dependence of the absorption lines within statistical uncertainties .
there is also an overall agreement with the esi measurements .
we deduce that discrepancies in any two measurements are mainly due to uncertainties in the continuum level in the vicinity of each line
. there does seem to be a systematic trend of larger equivalent widths on the second night below 5403.80 .
however , we also note that it is clear from the spectrum that the equivalent widths of lines on the blue side are relatively `` weak '' with respect to well detected lines on the red side where such trend is not detected .
hence , at this point we can not attribute the systematic trend entirely to a physical effect since the relative line flux measurements might have been affected by the continuum fitting during the second night as discussed in 3.1 and 3.2 .
although the equivalent width depends on the overall characteristics of the continuum , its main uncertainty is defined locally by the error in determining the proper level of the continuum on either side of the line of interest . as discussed in
3.1 the errors quoted for equivalent widths correspond to a poisson model of the noise in the region of interest .
such an assumption might be oversimplified and not account for true variations in the continuum , but we believe that we safeguard against strong systematic effects by using the same , consistent method in all measurements . hence based on our analysis
we do not find statistically significant changes in our measurements . taking this comparison
further there are published equivalent width measurements for this burst elsewhere ( jha et al .
2001 , masetti et al 2001 , salamanca et al .
although the duration of observation , spectral resolution and brightness of the burst varies in each case , the reported measurement techniques are similar .
direct comparison finds that our results are consistent with previous measurements .
nevertheless some differences are present in a few cases such as c iv , cr ii and zn ii that were discussed in 4.3 .
however , we do not find systematic and statistically significant changes as a function of time in the published data that should be present if either photoionization or dust destruction in the circumburst medium are actively dominant for similar species .
we refrain from reporting a measurement for al iii(1862.79 ) during the second lris spectrum in table 1 because of the reduced signal - to - noise in that region of the spectrum . for completeness
we note that the line has not disappeared and it is clearly detected in the esi spectrum obtained at a comparable time ( table 2 ) .
in order to interpret our observations under contraints imposed by the lack of absorption - line variability , we simulated the grb environment using a standard photoionization model .
the photoionization evolution has been described by perna & loeb ( 1998 ) and bttcher et al .
the photoionization rate can be written in the following form : @xmath88 where @xmath89 is the density for species @xmath54 , @xmath90 is the flux at radius @xmath91 , @xmath92 is the elapsed time in the grb frame and @xmath93 is the photoionization cross section at frequency @xmath94 using the analytical fits provided by verner et al .
we neglect recombination processes since the densities we will be considering are typically @xmath95 @xmath96 @xmath8 , where the corresponding recombination timescales are much longer than the duration of the grb and our observations ( perna & loeb 1998 ) .
in addition the full effect of dust destruction and the return of depleted metals into the gas phase has been ignored in the code .
next we constructed the input flux @xmath90 for this particular afterglow from the observed time decay ( stanek et al .
2001 ) and the measured optical spectral index .
the function has two forms to accommodate the observed break in the light - curve at @xmath97=0.72 day . thus for t @xmath69 0.72 day @xmath98 otherwise @xmath99 where @xmath100 is the luminosity distance to the burst at @xmath14 ( assuming @xmath101 km s@xmath18 mpc@xmath18 , @xmath102 , @xmath103 ) , and @xmath104 = @xmath105 cm is the inner radius of the photoionized region set as a boundary condition . we have used the observed flux and ignored any intrinsic dust corrections . in a previous section we showed that intrinsic reddening following a calzetti et al .
( 2000 ) extinction law for starbursts would dim and stepeen an intrinsically flat spectrum .
the effect of using an input flux uncorrected for intrinsic extinction is to slow down the time - dependence in our models ; therefore , the integration times presented here should be considered upper limits .
the input flux has been integrated in the uv range up to a cutoff energy of 0.2 kev .
the reasoning behind this choice is partly arbitrary given the failure to locate a precise spectral break at the `` cooling frequency '' @xmath106 between the optical and x - ray ranges .
the selected cutoff energy of 0.2 kev represents roughly the region where the photoionization cross section for the species of interest has dropped sharply . above 0.5 kev ,
x - rays are mostly absorbed by heavy elements .
we start the photoionization of a pristine neutral medium at initial time @xmath107 that follows the detection of prompt emission in the bepposax energy band .
the transition between the prompt grb emission and the onset of the afterglow is not a very well defined point observationally .
variability arguments set the scale for internal shocks to take place in a region of size @xmath108 cm @xmath109 .
external shocks responsible for the afterglow become significant at @xmath110 cm from the source ( piran 1999 ) . for numerical reasons we have selected an initial time of @xmath107 = 10 seconds for the onset of photoionization . the photoionization process can be pictured as ocurring inside a sphere of outer radius @xmath111 , subdivided into thin spherical shells of width @xmath112 so that each shell @xmath113 is optically thin . in each shell the initial incident flux spectrum described above will be absorbed and will also follow a @xmath114 law as r @xmath115 @xmath111 .
hence the incident spectrum at any shell @xmath113 will be given by : @xmath116 we ignore bound - bound processes therefore @xmath117 stands for the photoionization optical depth which is estimated within each shell @xmath113 as : @xmath118\ ] ] where @xmath119 . for our models we included h , he , and mg and neglected other elements .
the initial number density @xmath89 for each species is largely unknown but we decided to use standard solar - system abundances . for simplicity
we also assumed that initially mg is equally divided into mg@xmath78 and mg@xmath120 .
since the depletion correction / abundance and therefore the total @xmath121 near the burst site is not well determined , we examined the time dependence of equivalent widths for two different hydrogen column densities .
the first , log@xmath122 @xmath65 represents the inferred hi column ignoring any dust depletion correction and takes into account the possibility of complete circumburst dust destruction .
the second , log@xmath123 @xmath65 is appropriate for standard milky way depletion onto dust or perhaps less than solar metal abundances .
the latter is also consistent with measurements of the red wing of a possible ly@xmath68 feature ( salamanca et al .
2001 ) , and the bepposax afterglow results ( int zand et al .
2001 ) . in each case we chose four different combinations of number densities @xmath124 and @xmath111 to reproduce the simulated hydrogen column density @xmath125 .
the parameters for each model are listed in table 5 .
the denser , more compact models represent the conditions in dense molecular clouds which have been postulated as a potential ambient environment for the progenitors of grbs . on the other hand ,
the low density and extended environments are compatible with low density molecular clouds , young stellar clusters , and possibly galactic superbubbles ( scalo & wheeler 2001 ) .
we considered it relevant to trace the time evolution of two particular absorption lines present in our spectra , mg i ( 2852.96 ) and mg ii ( 2796.35 ) .
mg i is an appealing line given the possibility of encountering mg in its neutral form in regions of the galaxy shielded from photoionization .
mg ii(2796.35 ) is of interest because of its detection in a number of grb spectra .
an additional observational reason for choosing these particular lines to test our models is the consistency of the line flux measurements on the red side of the spectrum where the effects of differential refraction appear of less importance . in order to model our observations
, we integrated to a maximum time @xmath126 hours after the grb , when the second lris spectrum was obtained .
the equivalent width at different stages was estimated using equations ( 1 ) and ( 2 ) which return @xmath57 . throughout our calculations
we have assumed a constant doppler parameter @xmath52 representative of the system ( see table 3 ) .
this appears to be a fairly robust assumption since the time dependence for the lines is mainly a function of the photoionization equations .
figure 6 illustrates the time evolution of the equivalent width for mg i ( 2852.96 ) and mg ii ( 2796.35 ) in a log@xmath127 @xmath65 environment .
the times when the lris spectra were obtained are indicated .
notice the distinct differences in the line evolution in each case .
the denser , more compact models are ionized rather rapidly since most of the gas is close to the grb and gets impacted by a large flux . on the other hand ,
less dense models representative of thin galactic environments require longer for noticeable effects to take place .
the @xmath128 @xmath8 , @xmath129 pc model appears to mark the line where our low - resolution spectroscopy would have been able to detect time dependence with this column density .
denser environments than this appear to be disfavored according to our model .
figure 7 shows the time dependence for a log@xmath130 @xmath65 environment .
similarly compact , high density environments are photoionized rather quickly while extended , less dense ones would evolve with a shallower decline .
the @xmath131 @xmath8 , @xmath132 cm model is close to our detection limit for this denser environment .
the results in this case are more dependent on the choice of the inner radius @xmath104 because the scale of some of the models becomes comparable to @xmath104 . a different choice for the inner radius would either accelerate or delay the photoionization timescale , therefore , compact models with this column density are more tentative .
however , it appears that very dense , compact environments would have been more highly ionized than is observed in our initial spectrum .
one interpretation of x - ray and optical afterglow data from grb 010222 invokes a transition from a relativistic to a non - relativistic regime , and consequently the deceleration of the shock , to account for the break in the light curve ( in t zand et al .
this scenario requires an environment with @xmath133 @xmath8 and size @xmath110 cm .
although our simulations argue against the presence of a dense @xmath134 cm neutral gas slab between the two observations , we can not rule out the presence of a large density slab at the time of the burst itself .
however our modelling also presents an alternative scenario for more extended @xmath135 absorption in the host galaxy rather than a single slab .
recent modelling of the multiwavelength observations of grb 010222 , including the radio emission , finds that a jet model expanding into homogeneous external medium with density @xmath136 @xmath8 is consistent with the broadband data ( panaitescu & kumar 2001 ) . a jet and low - density interpretation
is also required to satisfy the standard - energy output with beaming corrections of frail et al .
( 2001a ) .
one final point in this discussion is the shape of the @xmath137-ray burst light curve and baryon contamination .
a very dense nearby medium gives rise to high optical depth @xmath59 where baryons could prevent relativistic expansion .
the overall effect could lead to `` dirty '' fireballs and smearing of the @xmath137-ray profile .
instead the burst profile appears highly structured ( in t zand et al .
2001 ) .
consider now the actual column density to this burst if the host galaxy has a dust - to - gas ratio similar to that of the milky way .
a color excess of @xmath138 mag is estimated for log@xmath123 @xmath65 and assuming a galactic conversion @xmath139 @xmath65 mag@xmath18 ( savage & mathis 1979 ) .
this is significantly larger than @xmath140 mag for a log@xmath122 @xmath65 environment .
it is possible that the uv emission from the burst is responsible for destruction of neighboring dust in the high - density case ( galama & wijers 2001 ) , but further extinction at larger distances is expected along the line of sight unless the burst is near the earth - facing side of the host galaxy ( reichart & price 2002 ) .
we must keep in mind , however , the possibility that we are probing different scales in two separate energy bands .
grb 010222 is a good candidate to explore whether we can place constrains on the progenitor environment .
the kinematic components resolved in the high - resolution spectrum give support to a location of this burst in a galactic disk .
high - resolution spectroscopy of grb 000926 ( castro et al .
2001b ) favors a similar interpretation for that burst .
the slope of the spectral continuum in this instance suggests that the burst occurred in a dust - free environment or was surrounded by a screen of gray dust ( aguirre 1999 ) .
near - ir and submillimeter observations by frail et al .
( 2001b ) also appear to rule out the possibility of strong dust scattering or thermal emission near the grb site .
however it is difficult to model _ a priori _ the properties of dust at this redshift since the distribution and properties of galactic dust as a function of redshift are largely unknown .
popular models of gamma - ray burst progenitors include young massive stars ( woosley 1993 , vietri & stella , 1998 ) .
this makes the interior of molecular clouds a potential birthplace for grbs .
molecular clouds tend to comprise large amounts of dust and clumpy pockets of gas ideal for star formation .
if grb 010222 took place in a molecular cloud , the slope of the optical continuum indicates that either the nearby dust was destroyed rather immediately by the burst through a combination of sublimation by an optical / uv flash ( waxman & draine 2000 ) and grain charging ( fruchter , krolik & rhoads 2001 ) or that the surrounding dust content was low prior to the explosion .
the absence of measurable signatures of metals being returned to their gas phase as a by - product of dust destruction on such a short timescale does not allow direct distinction between these two possibilities .
however , a high fraction of dust results in a larger @xmath5 formation rate since molecular hydrogen tends to form in reactions taking place on grains .
the absence of strong absorption features due to vibrational excited levels of @xmath5 expected from strong uv emission impacting the surrounding @xmath5 that might permeate a progenitor in a dusty molecular cloud ( draine 2000 ) is conspicuous .
moreover , assuming a mean density of @xmath12 @xmath8 and the typical size of a dense molecular cloud , the absorption lines detected should have weakened over the period of our observations
. a massive and dense molecular cloud might also give rise to reflection at 6.4 kev as seen near the galactic center ( murakami , koyama , & maeda 2001 ) .
nevertheless , it is still viable that a molecular hydrogen detection has been made blueward of our spectral limit by lee et al .
( 2001 ) and galama et al .
as reported by these authors the flux deficit in the @xmath141 band which extends below 3500 does not fit the smc or lmc - like extinction models .
a depression signature is also expected if there is absorption by a lyman-@xmath68 wing ( salamanca et al .
the absence of molecular hydrogen and reduced depletion resulting from the [ zn / mn ] and [ zn / cr ] ratios still allows the location of the grb to be near the edge of a molecular cloud prior to the explosion .
thus , at the time of the event the fraction of gas to dust can be larger than at the center of the cloud , inhibiting the formation of @xmath5 . immediately after the onset of the uv emission from the burst
, any remaining dust could have been destroyed efficiently by the expanding uv radiation pulse , minimizing reddening of the optical continuum .
panaitescu & kumar ( 2001 ) in their multiwavelength analysis of well - studied afterglows found that the ambient density for some bursts can be as low as @xmath142 @xmath8 ; however , some arguments favor higher ambient densities once the complete radio data set is included in the analysis of individual bursts ( berger et al .
the presence of low densities led scalo & wheeler ( 2001 ) to argue for active star formation regions in superbubbles as possible environments for grbs .
our photoionization models show that low - density environments with the typical sizes of superbubbles would not produce strong time - dependence in absorption lines .
moreover , the interior of a superbubble could be partionally ionized prior to the grb event .
the prior ionization might be accomplished via overlapping ionization bubbles generated by supernovae in clusters of massive stars .
a pre - ionized environment decreases the amount of neutral material seen by the expanding photoionization front and prevents the full effect of time - dependent absorption .
but it also enables the detection of time dependence in high - ionization species .
we detect high - ionization c iv absorption in our low - resolution spectrum , but it is not clear that equivalent width variations are seen . in general c iv is a tracer of additional diffuse warm and hot gas that could originate in a superbubble environment .
unfortunately , our high - resolution spectrum does not cover this absorption line to firmly establish if it comes from the disk of the host galaxy or rather from a gaseous halo .
we note however that the equivalent width of c iv ( 1548.20 ) is comparable to that of the mg ii absorption lines . an additional signature of prior ionization could be the attenuation of the x - ray continuum by an ionized absorber like those seen in agn spectra . the absence of strong absorption features except for a marginal fe xiv - xviii edge suggested by in t zand et al .
( 2001 ) prevents a straightforward connection from being made between x - ray and uv absorption features .
an advantage of the superbubble scenario is the availability of a dust destruction mechanism and a favorable geometry for observation .
the combination of winds and sn explosions would tend to blow the gas and dust away leaving filamentary or shell - like patterns like the one reported in the galactic orion - eridanus bubble ( guo et al .
in addition , winds and radiation pressure in the interior of a superbubble might selectively destroy small grains , a mechanism for gray dust production locally .
however , this selective process is balanced by rapid destruction of large grains by the burst and earlier destruction by x - rays emitted inside the superbubble that might lead to grain charging ( fruchter , krolik , & rhoads 2001 ) . in combination , these two processes could counteract the expulsion of small grains and suppress the formation of gray dust .
either dust destruction or its gray nature are consistent with the slope of the observed continuum and the lack of strong dust reprocessing near the burst site ( frail et al .
the only practical problem with this interpretation is the mismatch between the optical and x - ray spectra .
it is possible that the cooling frequency has moved out of the optical band . in this case
we propose that the observed x - ray excess ( in t zand et al .
2001 ) is mostly due to inverse compton scattering rather than intrinsic reddening near the burst location but if true the ambient density can not be below a few @xmath143 or inverse compton scattering would be suppressed ( harrison et al .
2001 ) . if the effects of dust depletion are less pronounced than in the milky way , a log@xmath122 @xmath65 environment with @xmath144 @xmath8 and @xmath145 pc is suggested for grb 010222 .
this is consistent with the best - fitted parameter @xmath146
@xmath8 resulting from jet modelling ( panaitescu & kumar 2001 ) . the recently discovered arches cluster ( nagata et al .
1995 ; cotera et al .
1996 ; serabyn , shupe , & figer 1998 ; figer et al .
1999 ) consists of more that 100 young stars with masses greater than 20 within a @xmath147 pc radius .
radio observations of the arches stellar cluster have confirmed the presence of powerful ionized winds with mass - loss rates @xmath148 yr@xmath18 coincident with massive stars in the cluster ( lang , goss , & rodriguez 2001 ) .
if a low metallicity massive star that has lost its envelope turns out to be the prototype for the collapsar model of grbs ( macfadyen & woosley 1999 ) , a young stellar cluster environment appears to be a perfect cradle for progenitors as prelude to grb formation .
a compact young cluster environment is consistent with our observations and models but requires low dust content or rapid destruction near the burst site , as well as photoionization of the gas in the region by hot young stars prior to the grb .
the complexity of wind interaction and shocks might provide favorable dust viewing geometries and dust distribution similar to those discussed in the superbubble scenario .
in addition , the uv absorption lines and their lack of time dependence might be attributed to the medium well outside the stellar cluster where the photoionizing flux is not strong enough to imprint substantial changes .
one problem with this interpretation is the type of metallicities required by the macfadyen & woosley ( 1999 ) model which can be thought of as lmc - like ( a.i .
macfadyen 2002 , private communication ) .
the abundances in the grb 010222 host are larger than the abundances reported in lmc line of sights ( welty et al .
in addition , the absence of strong signatures of ionized absorption and pronounced emission lines in the x - ray spectrum ( int zand et al .
2001 ) might argue against a pre - ionized medium .
an alternative grb progenitor is a coalescing compact object ( neutron star - neutron star , black hole - neutron star ) ( paczyski 1986 ) .
even though the distribution of birth kick velocities for compact objects has not been fully surveyed , typical velocities around a few hundred km s@xmath18 could place a fraction of potential progenitors at high - latitude above the disk
. the apparent presence of a galactic disk in this case seems to rule out the possibility of a halo origin for grb 010222 .
even though a location in the halo implies a low - density environment , the ionized regions of interest would be relatively compact clouds in the halo . in this case , it is difficult to reproduce the measured neutral column densities and observed time evolution in a halo event .
photoionization in halo events might be more effective in erasing absorption features entirely if the line - of - sight does not intercept the disk of the host galaxy .
current ideas suggest that the hard - short class of grbs are due to coalescing compact objects ( janka 1999 , fryer 1999 ) .
rapid localization of the hard - short grbs might allow the detection of absorption - line time dependence in high - ionization species following coalescence in a galactic halo .
the use of high - resolution spectroscopy has resolved the host galaxy of grb 010222 into multiple components .
we conclude that the grb took place in a galactic disk that gives rise to the strongest absorption system .
the power - law index of the optical continuum indicates that the dust content at the burst place is low or that the reddening follows a gray dust model . under these conditions
the bulk of the x - ray excess ( in t zand et al .
2001 ) can be attributed to inverse compton scattering .
low resolution spectroscopy obtained over the span of two days failed to reveal any significant evidence for time dependence of the observed absorption lines .
signatures of strong @xmath5 absorption and/or fluorescence are also absent from our spectra .
photoionization models constructed to test a range of initial conditions for this burst show that dense , compact media would be photoionized fairly quickly while less dense media with large radii evolve slower with time .
we argue that prior to a grb , an environment partially ionized either by the progenitor or by nearby sources would prevent the full manifestation of predicted absorption - line evolution .
the physical attributes of grb 010222 are consistent with a low - density molecular cloud , superbubble , or young stellar cluster as the possible environment of the grb . in the future
some consideration should be given to a prompt optical - uv flash like the one accompanying grb 990123 .
the intensity of such an event could in principle lead to dust destruction , but it might also be responsible for the photoionization of a fraction of the neutral gas .
the full effect of such a flash has been ignored in our models .
preferentially , the observation of time - dependent absorption lines would work best in bursts occurring in the earth - facing side of the host galaxy and in compact regions .
our analysis suggests that this technique applied with high - resolution spectroscopy could potentially distinguish the metallicity of progenitor environments . in that regard shortening
the interval between burst localization and initial spectroscopy becomes crucial to exploring earlier , less photoionized times .
the next generation of high - resolution imagers might directly resolve the environment of grbs in a relatively nearby host galaxy .
a prospect raised by the observations of grb 010222 is the study of the properties of galactic dust at high - redshift using bursts as beacons to illuminate the dust nearby .
further exploration of this issue is encouraged as spectroscopy is obtained for a larger sample of grbs .
we would like to thank r. -p .
kudritzki and f. bresolin for obtaining the first set of spectra as well as h. spinrad , d. stern , a. dey , a. bunker , and s. dawson for taking the second set of lris spectra .
we also acknowledge eric gotthelf for allowing us to use his new alpha computer and robert uglesich for tips on optimization in fortran .
this work was supported by the national science foundation under grant no .
ast-00 - 71108 .
jsb is supported as a fannie and john hertz fellow .
cccccc si ii(1526.71 ) & 3782.12 & 1.4773 & 1.10@xmath149 & 1.51 @xmath50 0.13 & 1.72 @xmath50 0.47 + c iv(1548.20
) & 3835.36 & 1.4773 & 1.91@xmath149 & 2.55 @xmath50 0.23 & 2.67 @xmath50 0.29 + + c iv(1550.78 ) & & & 9.52@xmath150 & & + fe ii(1608.45 ) & 3984.61 & 1.4773 & 6.19@xmath150 & 0.40 @xmath50 0.04 & 0.52 @xmath50 0.30 + al ii(1670.79 ) & 4138.77 & 1.4771 & 1.83 & 1.48 @xmath50 0.09 & 1.58 @xmath50 0.12 + si ii(1808.01 ) & 4478.51 & 1.4770 & 5.53@xmath151 & 0.60 @xmath50 0.08 & 0.81 @xmath50 0.15 + al iii(1854.72 ) & 4592.77 & 1.4763 & 5.60@xmath149 & 0.79 @xmath50 0.06 & 0.94 @xmath50 0.13 + fe ii(2382.77 ) & 4592.77 & 0.9275 & 3.01@xmath149 & 1.01 @xmath50 0.08 & 1.21 @xmath50 0.17 + al iii(1862.79 ) & 4614.50 & 1.4772 & 2.79@xmath149 & 0.69 @xmath50 0.08 & . . .
+ zn ii(2026.14 ) & 5018.95 & 1.4771 & 4.89@xmath149 & 0.71 @xmath50 0.07 & 1.03 @xmath50 0.13 + + mg i(2026.48 ) & & & 1.15@xmath149 & & + cr ii(2056.25 ) & 5093.54 & 1.4771 & 1.05@xmath149 & 0.3 @xmath50 0.04 & 0.45 @xmath50 0.08 + zn ii(2062.66 ) & 5109.62 & 1.4772 & 2.56@xmath149 & 0.7 @xmath50 0.04 & 0.52 @xmath50 0.09 + + cr ii(2062.23 ) & & & 7.80@xmath150 & & + fe ii(2382.77 ) & 5137.88 & 1.1563 & 3.01@xmath149 & 0.64 @xmath50 0.08 & 0.81 @xmath50 0.11 + mg ii(2796.35 ) & 5390.80 & 0.9278 & 6.12@xmath149 & 1.04 @xmath50 0.10 & 1.06 @xmath50 0.13 + mg ii(2803.53 ) & 5403.80 & 0.9275 & 3.05@xmath149 & 0.71 @xmath50 0.07 & 0.91 @xmath50 0.08 + mg i(2852.96 ) & 5493.01 & 0.9254 & 1.83 & 0.67 @xmath50 0.13 & 0.63 @xmath50 0.15 + fe ii(2600.17 ) & 5605.97 & 1.1560 & 2.24@xmath149 & 1.32 @xmath50 0.07 & 1.21 @xmath50 0.11 + fe ii(2344.21 ) & 5807.30 & 1.4773 & 1.10@xmath149 & 1.97 @xmath50 0.06 & 1.86 @xmath50 0.06 + fe ii(2374.46 ) & 5881.78 & 1.4771 & 3.26@xmath150 & 1.67 @xmath50 0.08 & 1.80 @xmath50 0.22 + fe ii(2382.77 ) & 5902.60 & 1.4772 & 3.01@xmath149 & 2.37 @xmath50 0.09 & 2.10 @xmath50 0.12 + mg ii(2796.35 ) & 6030.31 & 1.1565 & 6.12@xmath149 & 2.07 @xmath50 0.13 & 2.08 @xmath50 0.16 + cccccc mg ii(2803.53 ) & 6044.97 & 1.1562 & 3.05@xmath149 & 1.77 @xmath50 0.14 & 1.79 @xmath50 0.15 + mn ii(2576.88 ) & 6383.63 & 1.4773 & 3.51@xmath149 & 0.62 @xmath50 0.07 & 0.50 @xmath50 0.10 + fe ii(2586.65 ) & 6407.39 & 1.4771 & 6.84@xmath150 & 1.53 @xmath50 0.08 & 1.5 @xmath50 0.15 + fe ii(2600.17 ) & 6441.14 & 1.4772 & 2.24@xmath149 & 2.48 @xmath50 0.08 & 2.2 @xmath50 0.13 + mg ii(2796.35 ) & 6926.28 & 1.4769 & 6.12@xmath149 & 3.06 @xmath50 0.04 & 3.00 @xmath50 0.11 + mg ii(2803.53 ) & 6944.62 & 1.4771 & 3.05@xmath149 & 2.88 @xmath50 0.04 & 2.90 @xmath50 0.09 + mg i(2852.96 ) & 7067.07 & 1.4771 & 1.83 & 1.22 @xmath50 0.05 & 1.31 @xmath50 0.09 + ccccc al ii(1670.79 ) & 4138.41 & 1.47692 & 1.83 & 1.87 @xmath50 0.43 + si ii(1808.01 ) & 4478.18 & 1.47686 & 2.18@xmath151 & 0.76 @xmath50 0.1 + al iii(1854.72 ) & 4592.39 & 1.47606 & 5.60@xmath149 & 0.41 @xmath50 0.20 + fe ii(2382.77 ) & 4592.39 & 0.92733 & 3.01@xmath149 & 0.53 @xmath50 0.26 + al iii(1854.72 ) & 4594.04 & 1.47695 & 5.60@xmath149 & 0.80 @xmath50 0.18 + fe ii(2382.77 ) & 4594.04 & 0.92803 & 3.01@xmath149 & 1.03 @xmath50 0.23 + al iii(1862.79 ) & 4614.18 & 1.47703 & 2.79@xmath149 & 0.49
@xmath50 0.07 + fe ii(2600.17 ) & 5011.47 & 0.92736 & 2.24@xmath149 & 0.41 @xmath50 0.11 + zn ii(2026.14 ) & 5018.98 & 1.47711 & 4.89@xmath149 & 0.91 @xmath50 0.12 + + mg i(2026.48 ) & & & 1.15@xmath149 & + cr ii(2056.25 ) & 5093.10 & 1.47689 & 1.05@xmath149 & 0.44 @xmath50 0.08 + zn ii(2062.66 ) & 5108.62 & 1.47671 & 2.56@xmath149 & 0.69
@xmath50 0.16 + + cr ii(2062.23 ) & & & 7.80@xmath150 & + cr ii(2066.16 ) & 5117.54 & 1.47684 & 5.15@xmath150 & 0.27 @xmath50 0.12 + fe
ii(2382.77 ) & 5138.48 & 1.15652 & 3.01@xmath149 & 1.12 @xmath50 0.11 + mg ii(2796.35 ) & 5389.83 & 0.92745 & 6.12@xmath149 & 0.96 @xmath50 0.14 + mg ii(2803.53 ) & 5404.04 & 0.92758 & 3.05@xmath149 & 1.08 @xmath50 0.16 + mg i(2852.96 ) & 5498.99 & 0.92747 & 1.83 & 0.51 @xmath50 0.21 + fe ii(2260.78 ) & 5597.47 & 1.47590 & 2.44@xmath151 & 0.10 @xmath50 0.08 + fe ii(2260.78 ) & 5599.68 & 1.47688 & 2.44@xmath151 & 0.42 @xmath50 0.16 + fe ii(2600.17 ) & 5606.38 & 1.15616 & 2.24@xmath149 & 1.15 @xmath50 0.16 + ccccc fe
ii(2344.21 ) & 5804.03 & 1.47590 & 1.10@xmath149 & 0.37 @xmath50 0.11 + fe ii(2344.21 ) & 5806.33 & 1.47688 & 1.10@xmath149 & 1.51 @xmath50 0.23 + fe ii(2374.46 ) & 5878.93 & 1.47590 & 3.26@xmath150 & 0.30
@xmath50 0.06 + fe ii(2374.46 ) & 5881.25 & 1.47688 & 3.26@xmath150 & 1.24 @xmath50 0.10 + fe ii(2382.77 ) & 5899.50 & 1.47590 & 3.01@xmath149 & 0.56 @xmath50 0.09 + fe ii(2382.77 ) & 5901.84 & 1.47688 & 3.01@xmath149 & 2.21 @xmath50 0.16 + mg ii(2796.35 ) & 6028.98 & 1.15602 & 6.12@xmath149 & 2.49 @xmath50 0.08 + mg ii(2803.53 ) & 6044.44 & 1.15601 & 3.05@xmath149 & 2.14 @xmath50 0.60 + mg i(2852.96 ) & 6152.90 & 1.15667 & 1.83 & 0.54 @xmath50 0.31 + mn ii(2576.88 ) & 6382.58 & 1.47686 & 3.51@xmath149 & 0.58 @xmath50 0.12 + fe
ii(2586.65 ) & 6404.29 & 1.47590 & 6.84@xmath150 & 0.64 @xmath50 0.13 + fe ii(2586.65 ) & 6406.82 & 1.47688 & 6.84@xmath150 & 1.72 @xmath50 0.20
+ mn ii(2594.50 ) & 6426.18 & 1.47685 & 2.71@xmath149 & 0.68 @xmath50 0.12 +
fe ii(2600.17 ) & 6437.76 & 1.47590 & 2.24@xmath149 & 0.49 @xmath50 0.05 + fe ii(2600.17 ) & 6440.31 & 1.47688 & 2.24@xmath149 & 2.12 @xmath50 0.19
+ mn ii(2606.46 ) & 6455.96 & 1.47691 & 1.93@xmath149 & 0.56
@xmath50 0.12 + mg ii(2796.35 ) & 6923.66 & 1.47596 & 6.12@xmath149 & 1.63 @xmath50 0.24 + mg ii(2796.35 ) & 6926.76 & 1.47707 & 6.12@xmath149 & 1.51 @xmath50 0.19 + mg ii(2803.53 ) & 6940.97 & 1.47580 & 3.05@xmath149 & 1.39 @xmath50 0.15 + mg
ii(2803.53 ) & 6944.06 & 1.47690 & 3.05@xmath149 & 1.61 @xmath50 0.11 + mg i(2852.96 ) & 7063.74 & 1.47593 & 1.83 & 0.42 @xmath50 0.24 + mg i(2852.96 ) & 7066.33 & 1.47684 & 1.83 & 1.21 @xmath50 0.22 + cccccccc mn@xmath62 & 2576.88 & 0.3508 & 22.8 & 14.00 @xmath50 0.44 & 20.47 @xmath50 0.44 & 21.92 @xmath50 0.44 & 0.50 + & 2594.50 & 0.2710 & 35.3 & 13.95 @xmath50 0.17 & 20.42 @xmath50 0.17 & 21.87 @xmath50 0.18 + & 2606.46 & 0.1927 & 29.6 & 14.00 @xmath50 0.20 & 20.47 @xmath50 0.20 & 21.92 @xmath50 0.21 + mg@xmath62 & 2803.53 & 3.05@xmath149 & . .
& @xmath10 13.88 & @xmath10 18.30 & @xmath10 19.85 & @xmath69 2.65 + fe@xmath62 & 2382.77 & 3.01@xmath149 & . .
& @xmath10 14.16 & @xmath10 18.65 & @xmath10 20.92 & @xmath69 2.3 + cccc @xmath152 & 0.089 & 1000.0 & i + @xmath152 & 0.89 & 100.0 & ii + @xmath152 & 89 & 1.0 & iii + @xmath152 & @xmath153 & 6.48 @xmath11 @xmath154 & iv + @xmath155 & 3.2 & 1000.0 & v + @xmath155 & @xmath156 & 0.19 & vi + @xmath155 & @xmath157 & 6.48 @xmath11 @xmath154 & vii + @xmath155 & @xmath96 & 3.56 @xmath11 @xmath154 & viii + | = 100 we present sequential optical spectra of the afterglow of grb 010222 obtained one day apart using the low resolution imaging spectrometer ( lris ) and the echellette spectrograph and imager ( esi ) on the keck telescopes .
three low - ionization absorption systems are spectroscopically identified at @xmath0 , @xmath1 , and @xmath2 .
the higher resolution esi spectrum reveals two distinct components in the highest redshift system at @xmath3 and @xmath4 .
we interpret the @xmath4 system as an absorption feature of the disk of the host galaxy of grb 010222 .
the best fitted power - law optical continuum and [ zn / cr ] ratio imply low dust content or a local gray dust component near the burst site .
in addition , we do not detect strong signatures of vibrationally excited states of @xmath5 . if the grb took place in a superbubble or young stellar cluster , there are no outstanding signatures of an ionized absorber , either .
analysis of the spectral time dependence at low resolution shows no significant evidence for absorption - line variability .
this lack of variability is confronted with time - dependent photoionization simulations designed to apply the observed flux from grb 010222 to a variety of assumed atomic gas densities and cloud radii .
the absence of time dependence in the absorption lines implies that high - density environments are disfavored .
in particular , if the grb environment was dust free , its density was unlikely to exceed @xmath6@xmath7 @xmath8 .
if depletion of metals onto dust is similar to galactic values or less than solar abundances are present , then @xmath9 @xmath10 2 @xmath11 @xmath12 @xmath8 is probably ruled out in the immediate vicinity of the burst . |
the born - oppenheimer approximation@xmath0 is fundamental to studies of a wide variety of molecular systems .
however , violations of the born - oppenheimer approximation@xmath1 are ubiquitous in many molecular and condensed phase phenomena , and are responsible for several relaxation processes . despite their importance , fully quantum treatments of nonadiabatic corrections to born oppenheimer dynamics are limited and no perturbative formulation exists to account for them .
it is practically impossible to compute the full quantum nonadiabatic dynamics for realistic systems due to the prohibitive computational cost associated with such a calculation . to circumvent this difficulty , a class of mixed quantum classical methods have been invented@xmath2 .
these methods can be broadly classified into mean - field@xmath3 and trajectory surface hopping methods@xmath4 .
mean field methods originate from work by ehrenfest@xmath5 , where the nuclear degrees of freedom , treated as classical mechanical variables , experience a potential that is the expectation value of the electronic hamiltonian with respect to the instantaneous electronic wavefunctions obtained by solving the electronic schrodinger equation . on the other hand , surface hopping methods@xmath4 treat the nuclear dynamics as being localized to a single born - oppenheimer surface at any time , with nonadiabatic couplings between surfaces leading to instantaneous jumps , or `` hops '' between born - oppenheimer surfaces .
common to all mixed quantum - classical methods is the identification of the nuclear ( or slow ) degrees of freedom as behaving as essentially classical quantities , while interacting with a quantum system consisting of the electronic degrees of freedom .
this identification is usually justified by a mixture of mathematical and phenomenological arguments . as a result , while these methods have been investigated by intuitive tests and computational simulations over the years , there is no rigorous theory to understand situations where this separation into classical and quantum subsystems breaks down and quantum effects on the classical degrees of freedom have to be considered .
this problem is also evident in studies of the dynamics of molecular systems on higher - dimensional born - oppenheimer surfaces , where geometric features , like conical intersections between surfaces , have a significant physical effect on the dynamics .
thus , given the somewhat ad - hoc nature of most mixed quantum classical methods , a central problem is to develop a framework which would incorporate these methods as an approximation and would enable the systematic study of corrections to the mixed quantum classical picture of nonadiabatic dynamics .
this work aims to systematically construct mixed quantum classical methods as semiclassical limits of quantum nonadiabatic dynamics .
such an approach enables the systematic improvement of mixed quantum classical methods to include further quantum corrections as well as the development of new and hybrid mixed quantum classical methods .
exact path integral descriptions of the quantum nonadiabatic dynamics are provided and stationary phase approximations for the dynamics are derived .
the resulting dynamical picture is a consistent means of obtaining a mixed quantum classical approximation to the nonadiabatic dynamics .
it is demonstrated that both the ehrenfest approach and the surface hopping approach correspond to stationary phase approximations of the quantum nonadiabatic dynamics , when studied in complementary representations of the electronic system .
furthermore , the approximate equations of motion obtained in the ehrenfest picture contain additional contributions to the effective nuclear force .
these contributions are a direct , and nontrivial consequence of the geometry of the born - oppenheimer potential energy surfaces . for systems with only one nuclear dimension , these contributions are zero , and the dynamics is of pure ehrenfest type , while for higher dimensional surfaces , they are non - zero , and represent corrections to the quasiclassical dynamics due to the topological features of the surface , like the presence of conical intersections and degeneracies . in the complementary picture from which surface hopping methods are obtained as an approximation , quantum corrections to the nonadiabatic nuclear dynamics can be built in by considering semiclassical expansions of the nuclear propagator in the born - oppenheimer approximation .
in addition to this , the approach outlined in this paper provides a new interpretation of nonadiabatic effects as a consequence of the geometry of the electronic hilbert space , and thus as arising due to a generalization of berry s phase for parametrized quantum systems .
the outline of the paper is as follows . in sec.ii , a path integral description of the quantum nonadiabatic dynamics
is introduced .
issues concerning path integral approaches to nonadiabatic dynamics are discussed and the derivation of the approach used in this paper are outlined in sec.ii.a . in sec.ii.b ,
the path integral description is derived .
it is based on integrating out the electronic degrees of freedom using a coherent state basis .
furthermore , the path integral so obtained is compared to other path integral formulations constructed within a born - oppenheimer basis .
it is shown in this section that nonadiabatic couplings between states are due to the geometry of the electronic hilbert space , and can be regarded as generalizations of berry s phase . in sec.iii , the path integral approach is analysed and stationary phase approximations are derived .
sec.iii.a demonstrates that the stationary phase approximation to the coherent state path integral corresponds to the ehrenfest formulation of mixed quantum classical dynamics .
furthermore , geometric corrections to the nuclear force in the ehrenfest approach are obtained .
these corrections are related to the phase coherence between different states , and are most significant when different born oppenheimer energy surfaces become close in energy and are strongly coupled due to nonadiabatic effects .
sec.iii.b studies stationary phase approximations to the complementary path integral description constructed using a born - oppenheimer basis for the electronic system . it is shown that the equations of motion so obtained correspond in the semiclassical limit , to the fewest switches surface hopping approach to nonadiabatic dynamics .
the theory in sec.iii.b also points to the construction of semiclassical propagators that can extend descriptions of nonadiabatic dynamics beyond the mixed quantum - classical limit .
sec.iii.c describes semiclassical quantization rules to approximate the eigenvalue spectrum for the full quantum dynamics .
the paper then concludes with a discussion of the results and future work in sec.iv .
the problems with path integral approaches to quantum nonadiabatic dynamics are briefly discussed , and a short outline of the consequent derivation of the path integral based theory is given .
the detailed mathematical treatment of a path integral approach based on coherent states is then subsequently discussed .
it should also be noted that this text uses the units convention where planck s constant is set to one , i.e @xmath6 in this paper .
the nonadiabatic corrections to the born oppenheimer approximation can be viewed as a result of a coupling between a set of `` fast '' or bath degrees of freedom and slow or `` system '' degrees of freedom . however , the mechanisms that result from this coupling are different from the usual problem of a system interacting with a dissipative bath , since the hilbert space enclosing the bath degrees of freedom in the born oppenheimer theory is parametrized by the slow degrees of freedom . as a consequence ,
the time dependent evolution of such a system involves complex geometric phase effects@xmath7 .
pechukas@xmath8 first developed a path integral prescription for nonadiabatic dynamics and derived a set of nonlocal equations of motion for the semiclassical nuclear path .
later work@xmath9 on a path integral description clarified the interpretation of the nonadiabatic coupling vector potentials as nonabelian gauge fields .
however , these descriptions are not easily amenable to approximations which can be studied computationally . in the pechukas formulation
, the effect of the electronic degrees of freedom on system dynamics enters through a nonlocal potential which is a function of the paths taken by the system , and similarly in the formulation of moody , shapere and wilczek@xmath10 , nonadiabatic couplings between born - oppenheimer states enter as path ordered matrix exponentials .
thus , an approach that adopts the path integral framework for the exact nonadiabatic dynamics , but which is also amenable to systematic and computationally tractable approximations would be of considerable theoretical and practical utility . the problems with regards to nonlocality and path ordered exponentials encountered in a path integral description of the quantum nonadiabatic dynamics can be partially resolved by an appropriate choice of the path integral . to this end
, a local basis of continuous , @xmath11-particle coherent state slater determinants is chosen to describe the electronic system .
the use of coherent states enables the straightforward construction of stationary phase approximations to the quantum dynamics and corresponding semiclassical quantization rules .
furthermore , since the propagator can be written as a functional integral over a continuous coherent state space , issues relating to path ordering and nonlocality become considerably simplified .
the coherent state basis is parametrized by the nuclear coordinates and is overcomplete .
since the basis is parametrized by the nuclear coordinates , a local hilbert space is defined at each point on a given nuclear path .
thus , intuitively , motion along a nuclear path involves the following : the first contribution is due to the propagation of the nuclear part of the hamiltonian , and the second is due to a rotation of the hilbert space from one point into the next one on the path ( fig 1 ) .
this rotation takes basis states at one point into the basis states at the next , and hence changes the electronic contribution to the effective nuclear hamiltonian .
in addition , the rotation of the basis in general involves `` twists '' and skews resulting in torsional contributions to the force acting on the nuclear degrees of freedom . to explicitly describe this , the derivation of the coherent state path integral proceeds as follows .
the matrix elements of the time propagation operator for the full ( nuclear plus electronic ) hamiltonian between initial and final nuclear positions are written down as a path integral .
the time for propagation is broken up into a large number of ( infinitesimal ) time segments and the propagation operator is written in a trotter product expansion .
complete sets of nuclear position eigenstates are then introduced for each time segment , and an infinitesimal propagator between successive nuclear positions is written down . the resulting expression is a path integral of an operator acting on the electronic degrees of freedom over each nuclear path . for a given nuclear path , this quantity simply corresponds to the electronic propagation operator parametrized by the nuclear path .
the matrix elements of the quantum nonadiabatic propagation operator between initial and final electronic states and nuclear positions are then evaluated through the matrix elements of this path dependent quantity between initial and final electronic states .
this is done by adding overcomplete sets of electronic @xmath11 particle coherent state slater determinants at each time slice and evaluating matrix elements of the infinitesimal propagation operator .
the overlap between these basis states belonging to consecutive timesteps gives rise to the nonadiabatic coupling contribution to the dynamics .
the resulting path integral is then further analysed .
the full hamiltonian for a system with nuclear and electronic degrees of freedom can be written as @xmath12 the nuclear hamiltonian @xmath13 is given by @xmath14 the coordinates @xmath15 denote the @xmath16 nuclear degrees of freedom with momenta given by @xmath17 .
the potential energy function , @xmath18 is the bare nuclear potential .
the electronic hamiltonian can be similarly defined and the electronic coordinates , @xmath19 , are elements of the vector @xmath20 .
the electronic potential energy @xmath21 is parametrically dependent on the nuclear positions .
@xmath22 for simplicity of presentation , the nuclei and electrons are assumed to be spinless .
this assumption is reasonable for the nonadiabatic effects considered here , and the treatment that follows can be extended to include spin effects .
also it is assumed that the nuclear masses , @xmath23 are all equal and have the value @xmath24 .
this is solely for notational convenience and does not significantly affect the conclusions of this work .
a born oppenheimer basis of eigenstates and eigenvalues is defined by @xmath25\mid\psi_{\mu};{\bf r}\rangle = e_{\mu}({\bf{r}})\mid\psi_{\mu } ; { \bf{r}}\rangle\ ] ] the unitary time propagation operator for this hamiltonian can be expanded in a trotter product expansion as described below .
@xmath26 } \exp{\big[-i\epsilon\frac{{\bf p}^{2}}{2m}\big]}\exp{\big[-i\frac{\epsilon}{2}(h_{e } + u)\big]}}\end{aligned}\ ] ] the propagation operator for a given set of initial and final nuclear positions @xmath27 and @xmath28 can be written as @xmath29}\langle{\bf r}_{k}\mid\exp{\big[-i\epsilon\frac{{\bf p}^{2}}{2m}\big ] } \mid{\bf r}_{k+1}\rangle}}\nonumber\\ \exp{\big[-\frac{i\epsilon}{2}\{h_{e}({\bf r , r}_{k+1})+u({\bf r}_{k+1})\}\big ] } & & \end{aligned}\ ] ] where @xmath30 and @xmath31 .
it is important to note that the matrix elements of the propagation operator described in eq.(6 ) are still operators acting on the electronic hilbert space .
this is because the matrix elements are a partial trace with respect to the eigenstates @xmath32 and @xmath33 of the nuclear position operator @xmath34 .
this partial trace can be rewritten as a path integral over the set of nuclear paths @xmath35 .
@xmath36 } g_{na}[{\bf r}(\tau)]}\ ] ] @xmath37 $ ] is the bare nuclear action corresponding to the potential @xmath18 .
the quantity @xmath38 is an operator quantity in the electronic degrees of freedom , and is a nonlocal functional of the nuclear paths .
physically , the operator @xmath39 $ ] includes the `` connection '' between electronic hilbert spaces defined at each point on a given nuclear path @xmath40 .
thus , a complete solution of the nonadiabatic problem would require that electronic matrix elements of @xmath38 be evaluated . to perform such an evaluation , overcomplete sets of antisymmetric @xmath11 electron slater determinant coherent state wavefunctions
are introduced at each nuclear position along a given nuclear path@xmath41 . a general @xmath11
electron antisymmetric wavefunction can be written as a linear combination of slater determinants .
this basis of @xmath11 electron slater determinants , @xmath42 is overcomplete and has the closure property : @xmath43 here , a local basis of @xmath11 electron coherent state slater determinants@xmath41 is defined at each point on a given nuclear path .
hence , this local electronic basis is parametrized by the nuclear position to account for the fact that a local electronic hilbert space exists at each nuclear position .
for such a basis defined for a nuclear position @xmath44 , the completeness relation reads : @xmath45 for notational convenience , the integration measure in the overcompleteness relation will be written as @xmath46 = \frac{d\phi^{*}({\bf r})d\phi({\bf r})}{(2\pi i)^{n}}e^{-\mid\phi({\bf r})\mid^{2}}\ ] ] with the differentials and the absolute value of @xmath47 defined by @xmath48 the parametrization of the electronic hilbert space by the nuclear position is responsible for geometric phase effects .
as will be shown below , these effects enter through the definition of a local basis of wavefunctions at each nuclear position .
the operator @xmath38 can be explicitly defined to be @xmath49=\prod_{k=1}^{p}{\exp{[-i\epsilon h_{e}({\bf r , r}_{k})]}}\ ] ] this operator is a path ordered quantity because the electron hamiltonian does not commute with itself when evaluated at different points on a nuclear path . to evaluate the matrix elements of this operator , overcomplete sets of @xmath11-electron slater determinant wavefunctions
are introduced at each timestep .
on doing so , the matrix elements of @xmath38 between an initial electronic state @xmath50 and a final state @xmath51 can be written as @xmath52d\mu[\phi_{j}({\bf r}_{j})]\langle\phi^{0};{\bf q ' } \mid\psi({\bf q}')\rangle\langle\psi({\bf q}')\mid e^{-i\epsilon h_{e}({\bf r , q } ' ) } \mid\phi({\bf q}')\rangle}}\nonumber\\ \langle\phi({\bf q}')\mid\psi_{1}({\bf r}_{1})\rangle\langle\psi_{1}({\bf r}_{1})\mid e^{-i\epsilon h_{e}(\bf{r , r}_{1})}\mid\phi_{1}({\bf r}_{1})\rangle ... \langle\phi_{p}({\bf r}_{p})\mid\phi^{f};{\bf q}\rangle\nonumber\\\end{aligned}\ ] ] two sets of overcomplete basis states are introduced at each timestep , because the time evolution operator is in general not diagonal in this basis of states , and the overlap between basis states corresponding to different positions on the nuclear path is not unity .
more specifically , the coherent states inserted at successive timesteps , @xmath53 , are parametrized by different sets of nuclear positions , @xmath54 .
since there is no constraint on the basis states at different nuclear positions , all possible matrix elements of the time evolution operator between these basis states will have to be considered .
furthermore , to evaluate this off - diagonal matrix element , an additional set of basis states parametrized by the position @xmath55 is introduced at time @xmath56 .
since the path integral is discontinuous in the coherent state space , all possible overlaps of this new basis element with the basis element parametrized by @xmath57 will have to be considered , and the matrix elements of the time evolution operator are not constrained to be diagonal .
this lack of constraint is unlike the cases encountered with more traditional coherent state path integrals where only diagonal matrix elements between coherent states suffice , and is a consequence of the parametrization by the nuclear coordinates .
it is clear that the above path integral involves matrix elements of the form @xmath58 = \langle\psi({\bf r}_{k})\mid\exp{(-i\epsilon h({\bf r , r}_{k}))}\mid\phi({\bf r}_{k})\rangle\ ] ] if the exponent on the right hand side of eq.(14 ) is expanded in powers of @xmath59 and terms higher than the first order are neglected , the matrix elements @xmath60 can be evaluated to be approximately @xmath58 = \langle\psi({\bf r}_{k})\mid\big[1 - i\epsilon h({\bf r , r}_{k})\big]\mid\phi({\bf r}_{k})\rangle\ ] ] making the same assumption as before , the right hand side of eq.(15 ) can be reexponentiated to yield @xmath58 = \langle\psi({\bf r}_{k})\mid\phi({\bf r}_{k})\rangle\exp{\big[-i\epsilon\frac{\langle \psi({\bf r}_{k})\mid h({\bf r , r}_{k})\mid\phi({\bf r}_{k})\rangle}{\langle\psi({\bf r}_{k } ) \mid\phi({\bf r}_{k})\rangle}\big]}\ ] ] in addition to the matrix elements @xmath60 , the path integral contains `` connection '' factors of the form @xmath61 these connection factors are geometric in character and are due to the parallel transport of the local electronic basis from one nuclear position to the next .
thus , these factors correspond to a generalization of the canonical , or berry s phase for simple , parameter dependent hamiltonians@xmath62 . to obtain a more familiar form for this phase dependence ,
the connection factors can be rewritten as follows : @xmath63 the difference @xmath64 is the difference between the wavefunctions @xmath65 evaluated at successive time steps .
@xmath66 on exponentiating and making the assumption that the basis wavefunctions @xmath65 vary continuously , the connection factor takes the form @xmath67 where the time derivative acts on the wavefunction @xmath65 .
thus , on multiplying the matrix elements and correction factors together , the following path integral expression for the electronic contribution @xmath38 can be obtained as below :
@xmath68d[\phi^{*}({\bf r}(\tau)),\phi({\bf r}(\tau ) ) ] \langle\psi^{0};{\bf q}'\mid\psi({\bf
q}')\rangle}\nonumber\\ \langle\phi({\bf q})\mid\psi^{f};{\bf q}\rangle\exp{\big[i\int_{0}^{t } { \frac{\langle\phi({\bf r}_{\tau})\mid i\overleftrightarrow{\partial}_{\tau } - h_{e } \mid\psi({\bf r}_{\tau})\rangle}{\langle\phi({\bf r}_{\tau})\mid\psi({\bf r}_{\tau})\rangle } d\tau}\big]}\end{aligned}\ ] ] the quantities @xmath69 = \prod_{j=1}^{p}{d\mu[\psi_{j}({\bf r}_{j})]}$ ] are the measures for the coherent state path integral .
the time derivative is symmetrized by integrating by parts and ignoring boundary terms .
thus , the full path integral between two states @xmath70 and @xmath71 is given by a total action which of the form : @xmath72 = \int_{0}^{t}{\{l_{n}({\bf r},\dot{\bf r } ) + \frac{\langle\phi({\bf r})\mid i\overleftrightarrow{\partial_{t } } - h_{e}\mid\psi({\bf r})\rangle}{\langle\phi({\bf r})\mid\psi({\bf r } ) \rangle}\}dt}\ ] ] here the time derivative , @xmath73 , is the symmetrized derivative defined as : @xmath74\ ] ] the purely nuclear contribution to the action integral is contained in the lagrangian @xmath75 which is defined as follows : @xmath76 here the quantities @xmath44 correspond to the set of nuclear coordinates , and @xmath24 is the mass of the nuclei .
@xmath18 is the bare nuclear potential as defined earlier in eq.(2 ) .
the path integral in terms of this action can be written as @xmath77\big]}\nonumber\\\end{aligned}\ ] ] since , in general , off diagonal matrix elements of the time evolution operator in the coherent state basis are being considered , the boundary conditions on the path integral are that the coherent states defined at the end points @xmath78 and @xmath27 are allowed to vary without constraint . however , in the semiclassical limit , which corresponds to choosing only the subset of continuous paths in coherent state space , the coherent states at the boundaries are constrained to have unit overlaps with the initial and final electronic wavefunctions respectively .
as noted earlier , the necessity of choosing two sets of coherent states at each time slice in the path integral means that the paths in general are discontinuous .
this is because , in the limit where the discrete time slicing in the path integral is made infinitesimally small , the choice of different states at each such time slicing corresponds to a discontinuous jump in the space of coherent states .
however , the stationary phase limit of the path integral involves continuous paths in state space .
hence , it is reasonable to reduce the unconstrained sum over all ( discontinuous and continuous ) paths to one where the coherent states are constrained to be similar to each other .
a simple way of doing this is to assume the overlap , @xmath79 , to be nearly unity .
this is a reasonable approximation also because contributions to the path integral when this overlap is small are negligible .
this can be seen from the discrete version of the path integral , eq.(13 ) and eq.(16 ) , from which it is clear that as the overlap goes to zero , the exponent in the path integral becomes highly oscillatory , and in addition is multiplied by the small overlap .
the action when the overlap is assumed to be unity can be considered to be a zeroth order approximation to the full action , when expanded in terms of the overlap . with this simplification ,
the path integral action becomes : @xmath72 = \int_{0}^{t}{\big[l_{n}({\bf r},\dot{\bf r } ) + \langle\phi({\bf r})\mid i\overleftrightarrow{\partial}_{\tau } - h_{e}\mid\psi({\bf r})\rangle\big]d\tau}\ ] ] this action has a simple interpretation .
the time derivative part of the action corresponds to the nonabelian generalization of berry s phase , while the other parts correspond to the dynamical evolution of the system .
it is now possible to show that the generalized berry s phase in this action corresponds to nonadiabatic coupling terms between born - oppenheimer energy levels . to show this , the slater determinant states @xmath65 and
@xmath80 are expanded in a local basis of born oppenheimer electronic eigenstates as below : @xmath81\mid\psi_{\mu};{\bf r}}\rangle\\ \mid\phi({\bf r})\rangle = \sum_{\mu}{w_{\mu}[\phi,{\bf r}]\mid\psi_{\mu};{\bf r}}\rangle\end{aligned}\ ] ] the states @xmath82 are eigenstates of the electron hamiltonian @xmath83\mid\psi_{\mu};{\bf r}\rangle = e_{\mu}({\bf r})\mid\psi_{\mu};{\bf r } \rangle\ ] ] in this basis , the phase of the connection factor over the entire path , @xmath84 can be rewritten as @xmath85 = \sum_{\mu\nu}{\int_{0}^{t}{d\tau\frac{1}{2}(w^{*}_{\mu}\langle\psi_{\mu};{\bf r}\mid i\partial_{\tau}[w_{\nu}\mid\psi_{\nu};{\bf r}\rangle ] - i\partial_{\tau}[w^{*}_{\mu } \langle\psi_{\mu};{\bf r}\mid]w_{\nu}\mid\psi_{\nu};{\bf r}\rangle)}}\ ] ] this connection factor is the phase obtained by multiplying the connection factors @xmath86 , in eq.(20 ) over the entire path .
a little algebra can be used to demonstrate that the connection factor can be further simplified into the following form : @xmath85 = \int_{0}^{t}{d\tau\sum_{\mu}{\frac{1}{2}(w^{*}_{\mu}[\phi,{\bf r } ] i\dot{w}_{\mu}[\psi,{\bf r } ] + h.c ) } } - \int_{c_{{\bf r}}}{d{\bf r}\cdot \sum_{\mu\nu}{{\bf a}_{\mu\nu}({\bf r})w^{*}_{\mu}[\phi,{\bf r}]w_{\nu}[\psi,{\bf r}]}}\ ] ] here , the coupling term @xmath87 is the familiar nonadiabatic coupling vector in the born oppenheimer basis : @xmath88 the second term that contributes to the connection factor has the form of an integral over the gauge potential @xmath87 .
in contrast to the case of a simple berry phase , this `` gauge '' potential is an infinite dimensional matrix and corresponds to a nonabelian geometric phase .
thus the nonadiabatic coupling between born oppenheimer surfaces has a purely geometric interpretation as arising due to the parallel transport of a local basis of electronic wavefunctions along a nuclear path .
it is therefore a natural generalization of the berry phase for parameter dependent quantum systems@xmath62 .
this generalized geometric phase also depends on the off - diagonal elements of the electronic density matrix , and consequently couples the slow , nuclear and fast , electronic degrees of freedom . in the limit where an infinite number of adiabatic states corresponding to the fast modes can be accessed
, this phase represents a dissipative coupling between the system of slow nuclear modes and the bath of fast electronic modes@xmath89 .
it should be noted that generalizations of berry s phase to the case of nondegenerate electronic levels have previously been obtained by mead@xmath90 and by zygelman @xmath91 .
a related path integral description of the nonadiabatic dynamics has also previously been derived by moody , shapere and wilczek@xmath92 . in their work
, a path integral was obtained by introducing born - oppenheimer electronic eigenstates at each nuclear position , in the expression for @xmath93 $ ] given by eq.(12 ) .
the electronic matrix elements of @xmath93 $ ] between initial and final born oppenheimer eigenstates @xmath94 and @xmath95 are given by @xmath96\mid\psi_{n};{\bf q}\rangle = \sum_{m_{1},m_{2}, .. }{\exp[-i\epsilon e_{m}({\bf q}')]\langle\psi_{m};{\bf q}'\mid \psi_{m_{1}};{\bf r}_{1}\rangle\exp[-i\epsilon e_{m_{1}}({\bf r}_{1})]}\nonumber\\ \langle\psi_{m_{1}};{\bf r}_{1}\mid\psi_{m_{2}};{\bf r}_{1}\rangle .... \langle \psi_{m_{n}};{\bf r}_{n}\mid\psi_{n};{\bf q}\rangle\nonumber\\ \end{aligned}\ ] ] this expression contains connection factors of the form : @xmath97 it is easy to see that this connection factor can be manipulated as with the coherent state case to give @xmath98\ ] ] with @xmath99
. this expression can be rewritten in terms of the nonadiabatic coupling to give @xmath100\ } _ { m_{k},m_{k+1}}\ ] ]
this connection factor is similar to the connection factor , eq.(20 ) derived in the coherent state electronic basis .
it is however harder to handle for the purpose of approximations and computations , due to the discrete nature of the summation over born - oppenheimer eigenstates .
although the approach presented here is physically equivalent to that of moody , shapere and wilczek , the use of electronic coherent states enables semiclassical approximations to the nonadiabatic quantum dynamics .
this is possible because the restricted sum over only born oppenheimer eigenstates in eq.(33 ) is replaced with an integral over continuous paths in the coherent state space .
this provides a practical advantage , in that computationally tractable semiclassical approximations to the full path integral become possible .
it is shown later in this work that stationary phase approximations to the nonadiabatic path integral so obtained , are closely related to the ehrenfest formulation of mixed quantum - classical dynamics .
the path integral representation derived here is an exact reformulation of the quantum dynamics of the system .
however , due to the continuum nature of the coherent state basis , it is difficult to numerically evaluate the path integral propagator and obtain the full quantum dynamics .
the path integral derived in the previous section can be used to develop a semiclassical formulation of the nonadiabatic dynamics . in this section , stationary phase approximations to the path integral are derived and a semiclassical quantization rule is described for the nonadiabatic dynamics .
the stationary phase equations so obtained form a natural quasiclassical limit for the full quantum dynamics . for the coherent state path integral , this approximation is shown to yield the ehrenfest equations of motion with an additional correction term .
this correction term corresponds to the geometric contribution to the equations of motion , and is non - zero in general when the nuclear subspace is three dimensional or higher .
for the sake of completeness , the stationary phase approximation to the coherent state path integral is compared to a generalized stationary phase type approximation to the nonadiabatic path integral when written out in the basis of electronic born - oppenheimer eigenstates .
it is shown that a generalization of the stationary phase approximation in this case yields a quasiclassical theory which is the target of the fewest switches trajectory surface hopping method
. a natural semiclassical extension of this approximation yields a generalization of the fewest switches method which can account for hops in classically forbidden regions .
finally , the derivation of general mixed quantum classical methods that combine the ehrenfest and surface hopping methods is discussed .
thus , this section clarifies and lays down the neccesary theoretical foundations for the systematic improvement of mixed quantum classical methods .
a stationary phase approximation to the path integral can be made to yield quasiclassical equations of motion for the coupled dynamics of the nuclear and electronic degrees of freedom .
such approximations are derived here to obtain quasiclassical equations of motion for the system .
it is shown that the equations so obtained are a generalization of ehrenfest s equations of motion .
a density matrix for the electronic subsystem can be defined as @xmath101 = w^{*}_{\beta}[\phi,{\bf r}]w_{\alpha}[\psi,{\bf r}]\ ] ] where the quantities @xmath102 are as defined in eq.(27 ) and eq.(28 ) . to express the nonadiabatic action in terms of this density matrix , the explicit form of the action from eqs.(25 ) and ( 30 )
is written to be @xmath103 = \int_{0}^{t}{d\tau\big[l_{n}({\bf r},\dot{\bf r } ) + \frac{i}{2 } [ w_{\mu}^{*}\dot{w}_{\nu } - \dot{w}_{\mu}^{*}w_{\nu } ] - \dot{\bf r}\cdot\sum_{\mu\nu}{{\bf a}_{\mu\nu}w_{\nu}[\phi , t]w_{\mu}^{*}[\psi , t]}\big]}\ ] ] with this form of the action , the stationary phase approximation corresponds to the condition @xmath104 = 0 .\ ] ] a detailed derivation of the stationary phase equations is given in appendix a. at the end of this work . on performing the appropriate manipulations ,
the stationary phase condition for the electronic density matrix , @xmath105 , can be shown to be the following equation of motion : @xmath106}\ ] ] here , the energies @xmath107 are the eigenvalues of the born oppenheimer eigenstates as defined in eq.(29 ) .
the corresponding equation of motion for the nuclear coordinates , @xmath44 with mass @xmath24 and the bare nuclear potential @xmath18 can be written to be @xmath108 the stationary phase equation for the nuclear coordinates , @xmath44 , contains a term that is proportional to the time derivative of the off diagonal electron density matrix @xmath109 .
in addition to the other terms which are proportional to the off - diagonal density matrix , this is an explicit contribution to the force acting on the nuclei due to the time variation of the electronic coherence .
hence , this accounts for changes in the effective nuclear force due to electronic dephasing .
eq.(41 ) can be further simplified and written explicitly in terms of the elements of the electronic density matrix , @xmath105 and the nonadiabatic couplings @xmath110 as , @xmath111 if a nonabelian `` vector potential '' is defined as an infinite dimensional matrix with elements @xmath112_{\mu\nu } = { \bf a}_{\mu\nu}\ ] ] then the corresponding gauge invariant , nonabelian `` magnetic field '' is given by @xmath113 from these definitions and eq.(42 ) , the stationary phase equations for the nuclear coordinates can be written in the form @xmath114 eq.(40 ) corresponds to the familiar liouville equation for the electronic density matrix , while eq.(45 ) is different because it contains corrections due to the nonadiabatic coupling .
the first three force terms correspond to the effective force in the ehrenfest formulation of mixed quantum classical dynamics , while the last term is a geometric correction to the ehrenfest equations .
this term therefore constitutes a new addition to the equations of motion and will need to be retained for consistency with the full quantum dynamics . when this term is ignored , the equations of motion so obtained correspond to the ehrenfest equations of motion .
the equation , eq.(45 ) , for the nuclear coordinates satisfies the invariance under general unitary transformations of the electronic basis . to see this , note that the nonadiabatic coupling matrix @xmath115 is defined through the nonadiabatic coupling matrix elements as @xmath112_{\mu\nu } = { \bf a}_{\mu\nu}({\bf r } ) = -i\langle\psi_{\mu}\mid\nabla_{\bf r}\psi_{\nu}\rangle\ ] ] to write this in a more transparent form
, infinite dimensional column vectors @xmath116 can be constructed with elements given by the electronic basis wavefunctions @xmath117 .
the matrix can then be rewritten as a product of column vectors @xmath118 it is straightforward to show that under a unitary transformation , @xmath119 of the electronic wavefunctions , @xmath120 the nonadiabatic coupling matrix transforms as @xmath121 the `` magnetic field '' , @xmath122 with matrix elements @xmath123 correspondingly transforms as @xmath124 the density matrix elements @xmath125 also transform similarly under the unitary transformation , and hence the `` magnetic field '' contributions to the equations of motion are invariant under such transformations .
this invariance is a general version of the the usual invariance under gauge transformations , of vector potentials and magnetic fields in classical electromagnetism .
it should be noted that the unitary transformations under which the electronic wavefunction basis transforms are infinite dimensional matrices , and hence the group , @xmath126 that these unitary transformations belong to is the group of symmetries for the nonadiabatic corrections . on the other hand , the terms in the equations of motion which are gradients of the electronic energy eigenstates break this symmetry ,
since the energy eigenvalues are , in general , not completely degenerate . when degeneracies exist for a subset of the electronic energy levels , the contribution to the nuclear force from the degenerate states is symmetric under unitary transformations corresponding to the degenerate subspace .
the group corresponding to this symmetry is @xmath127 , where @xmath128 is the degree of the degeneracy .
when this symmetry group exists , the time evolution of the degenerate electronic subspace picks up a berry phase@xmath129 which contributes to the nuclear force in eq.(45 ) in a fashion identical to that of the nonadiabatic coupling between nondegenerate states .
thus , the equation of motion for the nuclei , eq.(45 ) , is valid irrespective of whether the states in the electronic subspace are degenerate or not .
for example , in the case of a conical intersection between two potential energy surfaces , the wavefunction becomes multivalued . to resolve this multivaluedness , a geometric phase due to a vector potential
can be added to the wavefunctions .
the geometric forces in eq.(45 ) are the classical analog of this procedure .
it should be noted that the geometrical , `` magnetic field '' contributions in eq.(45 ) are zero for one dimensional nuclear potential energy surfaces .
this can be intuitively understood through the discussion in sec.ii.a .
this magnetic contribution is due to the `` rotation '' of the local electronic hilbert space along the nuclear path .
for one dimensional nuclear potential energy surfaces , this `` rotation '' does not have any torsional components , and is simply a pure parallel transport of the basis .
thus , for single nuclear dimensions , these forces are zero and do not contribute to the dynamics . however , for potential energy surfaces parametrized by higher dimensional nuclear coordinates , these forces are non - zero and of importance to the nonadiabatic dynamics .
these couplings therefore cause an effective magnetic force contribution to be added to the usual ehrenfest force , in the mixed quantum classical dynamics .
the magnetic forces are the quasiclassical analog of the additional berry - mead phase which the electronic wavefunctions are multiplied by , to maintain their single valuedness . a complementary description of the full nonadiabatic quantum propagation of the nuclei can be obtained by writing the nonadiabatic propagation operator @xmath130 $ ] from eq.(12 ) in a basis of electronic born oppenheimer eigenstates .
as mentioned previously such a path integral expression for the propagator has been written down by moody , shapere and wilczek .
explicitly , if the wavefunction for the combined system is given by @xmath131 then the time evolution of the nuclear states @xmath132 can be written as @xmath133 the propagator @xmath134 is the matrix element of the time evolution operator when evaluated between an initial electronic born - oppenheimer state , @xmath135 , defined for an initial position @xmath27 and a final electronic born - oppenheimer state , @xmath136 , defined for a final nuclear position @xmath28 .
this matrix element can be evaluated using the trotter product formula to slice the time @xmath137 into infinitesimal time intervals , and introducing a basis of electronic born oppenheimer states at each time slice .
the resulting propagator is the same as the propagator defined in eq.(33 ) .
explicitly , this propagator is the matrix element @xmath138 if the born - oppenheimer limit is assumed to be valid , then the propagator is diagonal in the indices @xmath139 .
@xmath140 here , @xmath141 is the nuclear propagator in a potential given by @xmath142 @xmath143 the nonadiabatic nuclear propagation can be thought of as a sequence of propagation along born - oppenheimer paths with the propagator defined above in eq.(54 ) , and `` hops '' occurring in between to switch between born - oppenheimer paths .
this is illustrated in fig.2 .
detailed considerations of the structure of the paths that contribute to the transition amplitude @xmath134 thus point to the origins of surface hopping type approximations . to derive explicitly quasiclassical techniques based on the discrete path integral ,
consider the effective propagation operator defined in eq.(12 ) , which corresponds to matrix elements of the full propagation operator between initial and final nuclear position eigenstates .
to recapitulate , the matrix elements are given as an integral of a path ordered exponential @xmath144 $ ] .
for a given nuclear path @xmath35 the matrix elements of @xmath38 between an initial born - oppenheimer electronic state @xmath135 and a final state @xmath136 are given by @xmath145\mid{\bf q};\psi_{n}\rangle = \langle{\bf q}';\psi_{m}\mid e^{-ih({\bf q}')\epsilon}e^{-ih({\bf r}_{1})\epsilon} ...
e^{-ih({\bf q } ) \epsilon}\mid{\bf q};\psi_{n}\rangle\ ] ] from this relationship , the quantum nonadiabatic propagator between an initial born oppenheimer electronic wavefunction @xmath95 defined at an initial nuclear position @xmath27 and a final born oppenheimer electronic wavefunction , @xmath94 defined for a position @xmath28 can be written as @xmath146\mid{\bf q};\psi_{n}\rangle}\ ] ] if a nonadiabatic transition takes place during the interval @xmath147 for sufficiently small values of @xmath59 , the propagator has the composition property @xmath148 the overlap between electronic wavefunctions at @xmath149 and @xmath150 is the connection factor , @xmath151 , discussed in the previous section , and is responsible for the nonadiabatic coupling between levels .
as before , it can be rewritten as : @xmath152 with the nonadiabatic coupling @xmath153 given by eq.(32 ) as before .
this composition law for the exact nonadiabatic propagator can be used to derive a born series expansion for the propagator in terms of the born oppenheimer propagator .
the physics behind this is illustrated in fig.2 .
@xmath154 here , the propagator is written as a sum over paths where there are successively higher number of nonadiabatic transitions . to obtain a classical nuclear propagation scheme ,
the nuclei can be propagated along the stationary phase solutions for those pieces of the nuclear path where the propagator is of the born oppenheimer type . in other words ,
the born - oppenheimer limit of the nonadiabatic propagator can be approximated with a semiclassical expansion around the stationary phase path as discussed below .
it is easy to see that along a born oppenheimer path , @xmath155 furthermore , if a single nonadiabatic transition occurs from @xmath156 on a sufficiently short timescale @xmath59 , the amplitude @xmath132 changes by @xmath157 putting the two equations together , the total time derivative of the wavefunction @xmath158 is recovered to be : @xmath159 it is important to note that this equation is valid under the assumption that nonadiabatic transitions occur on a very short timescale in comparison to the timescales that govern the rest of the dynamics . in general , there is a nonzero probability for nonadiabatic transitions occuring over longer timescales and accompanied by finite changes , @xmath160 in the nuclear coordinates .
this equation is derived assuming that the contribution from such transitions is small , and that dominant features of the nonadiabatic dynamics can be captured by considering transitions that occur on infinitesimal timescales . to derive a semiclassical limit for the nuclear evolution , the nuclear wavefunctions @xmath161 are approximated as gaussian wavefunctions centered around the instantaneous classical nuclear path on the born oppenheimer surface , according to the prescription of heller@xmath162 .
@xmath163\ ] ] the functions @xmath164 are instantaneous gaussian wavepackets centered around the classical born - oppenheimer trajectories given by @xmath165 furthermore , this gaussian wavepacket approximately satisfies ( when the born oppenheimer potential energy , @xmath166 , is expanded to quadratic order in @xmath167 ) the time dependent schrodinger equation : @xmath168 substituting the ansatz , eq.(64 ) , into eq.(63 ) , one obtains for the time evolution of the coefficients , @xmath169 @xmath170 making use of the approximation in eq.(66 ) , the time evolution of @xmath169 becomes , in the semiclassical limit : @xmath171 multiplying by @xmath172 and integrating over the coordinates @xmath167 gives @xmath173 this can be rewritten in terms of a density matrix @xmath174 as @xmath175}\ ] ] furthermore , if the gaussians @xmath176 are assumed to be delta functions in the positions @xmath177 , the above equations simplifies further into : @xmath178}\ ] ] where the variable @xmath44 and its velocity , @xmath179 are classical quantities corresponding to the center of the gaussian function @xmath176 . since the functions @xmath176 are gaussians centered on the classical nuclear born - oppenheimer trajectories , the classical variables @xmath44 satisfy newton s laws : @xmath180\ ] ] when the trajectory being considered is that belonging to the @xmath181 born - oppenheimer surface .
further , the rate of change of the diagonal elements , @xmath182 is given by @xmath183}\ ] ] it is clear from this equation and the fact that the nuclear propagation always occurs on a single born - oppenheimer surface , that these equations describe the surface hopping method@xmath4 , for which the fewest switches algorithm is a monte carlo solution .
thus , the fewest switches surface hopping ( fssh ) approach@xmath184 corresponds to a semiclassical approximation to the exact quantum nonadiabatic dynamics , similar to the ehrenfest method described in sec iii.a .
it is also interesting to note that eq.(70 ) can be used instead of eq.(73 ) to generate the surface hopping criterion .
eq.(70 ) provides for a nonzero hopping probability even in regions where transitions are classically forbidden , unlike the usual fssh method .
a similar extension has also been proposed previously by neria and nitzan@xmath185 , on physical grounds through the construction of a semiclassical golden rule .
in addition to this , it is clear from this discussion that semiclassical propagation of the nuclear wavefunction , either using the gaussian form , eq.(65 ) and eq.(66 ) proposed by heller@xmath162 , or by the use of a different basis , lead to generalized schemes for nonadiabatic dynamics that go beyond the mixed quantum classical limit . the use of frozen gaussians within this framework correspond to a scheme for nonadiabatic dynamics that is similar to the multiple spawning method proposed by martinez and co - workers@xmath186 .
the approximation of hops occurring over infinitesimal timescales that is required to derive both eq.(70 ) and eq.(73 ) causes phase changes in the wavefunction due to the hops to be neglected .
furthermore , hops need not occur on infinitesimal timescales , and the quantum nuclear paths which contribute to the dynamics are in general not differentiable , rendering the nuclear velocity to be a poorly defined quantity . in circumstances where such paths contribute significantly to the dynamics surface hopping type methods
will be invalid .
additionally , in the case where two ( or more ) electronic potential energy surfaces become degenerate , the nonadiabatic coupling also picks up nonzero diagonal elements which could act on the dynamics along significant sections of the nuclear path . in such situations ,
an approximate means of treating the dynamics would be to add the diagonal terms in the nonadiabatic coupling to the nuclear hamiltonian and perform the nuclear propagation on each born oppenheimer surface with an additional magnetic force due to the diagonal terms in the nonadiabatic coupling .
the off diagonal nonadiabatic couplings can then still be treated as acting instantaneously to cause hops .
it is likely that such modifications could be of value in improving the accuracy of the fssh method in such situations . finally , it is useful to discuss the relative utility of the discrete basis path integral presented here versus the coherent state approach that leads to the ehrenfest equations in the previous section .
the natural stationary phase limit of the coherent state path integral is the mean field ehrenfest theory , while the discrete path integral has a corresponding stochastic limit .
these two limits can be thought of as corresponding to two different types of statistics in the hopping events .
the coherent state path integral is a better starting point for approximations when there is significant nonadiabatic coupling between multiple born - oppenheimer surfaces , i.e when hops can occur with high frequency .
in such a situation , a mean field treatment is a good first approximation to the dynamics .
however , the discrete basis approach is better suited to studying the dynamics in the presence of nonadiabatic dynamics between a few levels , and when born oppenheimer surfaces begin to diverge .
in such a situation , the hopping events are relatively infrequent , and nonadiabatic behaviour is sensitive to the detailed structure of neighbouring born - oppenheimer surfaces . the discrete basis approach , for which the surface hopping algorithm is an approximation is a good starting point for approximation in this situation . the nonadiabatic action eq.(38 ) can be used to describe a semiclassical quantization condition when the system is in a bound state .
this condition is derived in a fashion similar to that used to study the semiclassical quantization of coherent state path integrals@xmath187 , and in quantum field theories@xmath188 .
if an assumption is made that the system executes periodic orbits , then a quantization condition can be written . to derive this quantization condition ,
the trace of the propagator is written as @xmath189\ ] ] the trace is related to the density of states as below .
@xmath190 = i\int_{0}^{\infty}{e^{iet}k_{na}(t ) dt}\ ] ] if the trace corresponding to @xmath191 is expanded in terms of the eigenstates of the entire ( the combined nuclear and electronic ) system , @xmath191 can be written in terms of the system energy eigenvalues @xmath192 as @xmath193 it is clear from this that the poles of the function @xmath191 correspond to the energy eigenvalues of the system .
hence , the poles of the fourier transform of the propagator trace are related to the eigenstates of the combined system in this fashion .
the fourier transform of the trace of the propagator can be written in terms of the nonadiabatic coherent state path integral as follows : @xmath194d[\psi(t ) ] \exp{(is_{t}[{\bf r},\psi,\phi ] + et)}}}\ ] ] here , the quantities @xmath195 are coherent state basis elements and the action @xmath196 is the same as defined in eq.(26 ) . to obtain an approximate quantization rule , only diagonal coherent state matrix elements in eq.(26 )
are considered .
eq.(75 ) which relates the fourier transform @xmath191 to the energy spectrum of the system also implies that @xmath191 is singular when the energy @xmath197 is equal to any of the system eigenenergies .
this implies that the path integral of the propagator in eq.(77 ) is a maximum for such values of the energy @xmath197 .
the propagator also acquires its maximal values when its phase is stationary .
consequently , the density of states , @xmath191 , is dominated by contributions to the path integral from stationary phase paths .
furthermore , if it is assumed that periodic orbits exist for the combined dynamics of the system , then the density of states in eq.(75 ) can be approximately written as @xmath198 = \sum_{p.o}{c_{k}\exp{[is[t_{k}(e ) ] + iet_{k}(e)]}}\ ] ] where the time @xmath199 is the time period for the classical traversal of the @xmath200th periodic orbit . since
, multiple traversals of a given periodic orbit also constitutes a periodic orbit , this sum can be simplified into a sum over periodic orbits @xmath201 which can not be reduced further into smaller sets of periodic orbits : @xmath202 + iet_{\mu}(e)]}}{1 - \exp{[is[t_{\mu}(e ) ] + iet_{\mu}(e)]}}}\ ] ] this expression has poles given by @xmath203\ ] ] this quantization condition can be explicitly written out to be @xmath204 } = 2\pi(n + \frac{1}{2})\ ] ] it should be noted that this quantization condition is valid under the assumption that periodic orbits corresponding to the effective classical system defined by the action @xmath205 exist . for nuclear motion coupled to a large number of electronic levels , finding the periodic orbits of the system is a difficult task . however , for small values of the nonadiabatic coupling , this quantization rule provides an approximate means of estimating the total energy spectrum of the system , given an initial set of born - oppenheimer surfaces .
furthermore , if the nonadiabatic transitions couple only a finite set of electronic born - oppenheimer surfaces , this rule may have validity because periodic orbits for the combined nuclear and electronic density matrix variables could exist .
this rule can be written in terms of the total connection factors @xmath206 $ ] for the closed path @xmath207 as @xmath208 = 2\pi(n + \frac{1}{2})\ ] ] a special case of the born oppenheimer problem , for which the semiclassical quantization rule is useful , is that of a single vibrational coordinate , @xmath209 , coupled to a set of electronic levels .
in this situation , the quantization rule reads @xmath210 in the absence of the nonadiabatic couplings , @xmath211 , the quantization rule gives the exact vibrational energy levels ( assuming that the vibrations are purely harmonic ) .
the contribution due to nonzero nonadiabatic coupling can be interpreted as a shifting of the energy levels of the oscillator to account for energy exchange with the electronic bath .
if an orbit is executed , nonadiabatic transitions between vibrational states can occur , which leads to energy transfer between the oscillator and the electronic system .
this work has presented an exact first principles path integral formulation of quantum nonadiabatic dynamics .
the path integral approach developed here has the advantage that it enables computationally tractable stationary phase approximations to the full path integral .
previous path integral based approaches to this problem have stationary phase approximations that lead to nonlocal theories of mixed quantum classical dynamics@xmath8 , or require considering a constrained set of paths for the purposes of approximation@xmath10 .
the method presented here therefore represents an advance over previous approaches in this regard .
a second consequence of this work is that corrections to the quasiclassical nuclear dynamics have been obtained .
these corrections reflect underlying geometric features of the electronic hilbert space .
the corrections so obtained have the advantage that they can be calculated through straightforward extensions of current mixed quantum - classical simulation methodologies .
applications of this approach to study the dynamics of real systems in the presence of conical intersections and other degeneracies in the electronic energy levels are currently being considered .
finally , semiclassical quantization rules have been proposed for nonadiabatic dynamics in the quasiclassical limit .
these rules are potentially of utility in determining approximate energy spectra when the born - oppenheimer approximation is not valid .
a third consequence of this work has been to provide a link between exact quantum theories of nonadiabatic dynamics and semiempirical surface hopping methods .
this work establishes a rigorous theoretical basis for the surface hopping method .
it has been demonstrated that the fewest switches surface hopping approach can be regarded as a generalized stationary phase approximation to the exact quantum dynamics , in complement to the ehrenfest formulation of mixed quantum classical dynamics .
natural extensions of the surface hopping method to include classically forbidden nonadiabatic transitions , as well as the inclusion of nuclear quantum effects through the construction of rigorous multiple spawning type algorithms have been demonstrated .
this work also allows for the construction of semiclassical golden rules to describe nonadiabatic transitions .
the present work also provides a clear perspective on issues concerning equilibrium between the quasiclassical nuclear and quantum , electronic subsystems .
it is clear from the theory presented here that traditional ehrenfest methods do not satisfy equilibrium conditions if only single mean field trajectories are considered .
each such trajectory corresponds to a stationary phase solution for the nonadiabatic propagator at a given set of initial conditions .
however , thermal properties of the system enter through sampling of initial conditions .
hence , if appropriately sampled ensembles of mean field trajectories are chosen , a mixed quantum - classical equilibrium can be established .
correct sampling of mean field trajectories will require an initial sampling of the reduced density matrix for the nuclei and should also correctly account for initial phase correlations according to eq.(26 ) .
it is possible that observed violations of mixed quantum classical equilibrium by the ehrenfest method are due to an incorrect sampling of either of these contributions .
since the surface hopping approximation is obtained by looking at the propagation of the system in each of the states occupied by it , an ensemble averaging of surface hopping trajectories will sample the initial density matrix of the system , and hence this approach and any systematic generalizations based on it can be expected to reproduce the correct mixed quantum classical equilibrium .
this has been demonstrated through a separate set of arguments@xmath212 .
this work provides a different conceptual perspective on the nature of nonadiabatic dynamics .
nonadiabatic couplings can be regarded as geometrical objects which are generalizations of berry s geometric phase .
thus , they can be regarded on the same footing as the geometrical phases that arise due to conical intersections@xmath213 .
this similarity between nonadiabatic couplings and geometric phases that arise in degenerate systems has also been noted by previous authors@xmath214 . in principle , the methods described in this work do not differentiate between systems which have degeneracies in energy and those which are nondegenerate .
therefore , the nonadiabatic path integral formulation presented here is of considerable generality and may be of practical utility in computational studies of nonadiabatic dynamics .
a potentially useful approach to study exact quantum nonadiabatic dynamics is to consider quantum monte carlo schemes to evaluate the coherent state path integral .
direct applications of quantum monte carlo schemes are likely to face difficulties due to the sign problem .
however , simplifications could be made by using an effective single electron ansatz to replace the matrix elements in the action eq.(26 ) .
such an ansatz could be rigorously constructed by using the runge - gross theorem@xmath215 which relates the exact one electron time dependent density to the expectation value of the electronic hamiltonian in eq.(26 ) .
extensions of the time dependent density functional approach to one electron density matrices can then be used to accurately evaluate the path integral due to the action in eq.(26 ) .
this has the added advantage of naturally incorporating time dependent density functional theory with the nonadiabatic propagation of a quantum system .
this approach will be explored in later work . although the exact coherent state path integral presented here is a challenge for numerical computation , it is possible to develop numerically more tractable semiclassical propagators that describe the nonadiabatic dynamics as an approximation to the coherent state path integral .
one such approach would involve a stationary phase evaluation of the electronic dynamics ( the ehrenfest equations for the electron density matrix ) , while propagating the nuclear motion using a semiclassical initial value representation .
more promising applications of this work are in developing better approximate methods to describe quantum nonadiabatic dynamics .
one potential application is to develop methods which systematically combine ehrenfest and surface hopping type algorithms . to enable this
the propagation of the nuclear dynamics can be divided into sections wherein different basis sets are used depending on physical considerations regarding the nature of nonadiabatic interactions along the nuclear paths .
another natural extension is in refining the surface hopping method to include classically forbidden nonadiabatic transitions by approximating the nuclear wavefunction with a gaussian wavepacket .
alternatively , more general semiclassical propagation schemes can be used to propagate nuclear dynamics in the discrete representation of the path integral .
this would facilitate the inclusion of nuclear quantum effects as well as more accurate treatment of dynamics in the presence of strong nonadiabatic coupling between born - oppenheimer surfaces .
the use of coherent states , and the derivation of the path integral presented here is analogous to path integral treatments of quantum spin systems .
this analogy with quantum spin systems also forms the basis for classical mapping techniques@xmath216 that have been developed to study mixed quantum classical dynamics .
finally , the path integral methods presented here are apparently amenable to perturbative approximations around its stationary phase value .
however difficulties remain due to the nature of the nonadiabatic corrections to born oppenheimer dynamics .
the nonadiabatic coupling between born oppenheimer levels consists of matrix elements of the nuclear momentum operator .
the nuclear momentum operator is unbounded in the hilbert space , and further , the nuclear paths that contribute to the path integral for the system are in general not differentiable .
hence , a perturbation expansion in the nonadiabatic coupling element is not well defined in the path integral framework@xmath92 . in the case of ehrenfest dynamics
, this issue also manifests through the assumptions made in constructing the coherent state path integral .
the assumption made in coherent state path integrals is usually that the paths chosen in the coherent state space are continuous .
this is not valid in general for the problem of nonadiabatic dynamics .
similarly , extensions of surface hopping type methods will need to account for non - differentiable nuclear paths where the nuclear velocity is not defined .
a means of overcoming this problem for the coherent state path integral is to regularize the path integral by considering higher order time derivatives in the action , in analogy to the procedure followed for spin systems@xmath217 .
a perturbative treatment of nonadiabatic dynamics that takes these issues into account is being worked on and will be presented at a later date .
the major part of this work was done when the author was at yale university . financial support from a yale university graduate teaching fellowship is gratefully acknowledged .
the author thanks professor j.c tully for his support and encouragement , and for discussions that motivated this work .
the author acknowledges dr.a.v madhav and professor j. simons for critical comments and suggestions regarding the manuscript .
the author is grateful to an anonymous referee for remarks and criticism which greatly improved and clarified the content of this paper .
the derivation of the stationary phase equations in sec.iii a. is described here .
the effective action in the coherent state representation is given by @xmath218 with a lagrangian @xmath219 given by @xmath220 } - \dot{\bf r}\cdot\sum_{mn}{{\bf a}_{mn}w^{*}_{m}w_{n}}\ ] ] for the stationary phase path , the action is stationary when varied with respect to the variables @xmath44 and @xmath221 which determine the path .
in other words the stationary phase condition : @xmath222 implies the equations : @xmath223 the first of the three equations above is the usual lagrangian equation of motion for newtonian mechanics which governs the classical dynamics of the variable @xmath35 , while the next two equations render the action stationary with respect to variations of the electronic density matrix .
the equations for the variables @xmath224 are obtained by using the appropriate unsymmetrized form of their kinetic energy term ( the fourth term in the right hand side of eq.(a2 ) ) .
on applying these equations of motion to the lagrangian , one obtains @xmath225 the second and third equations , eq.(a7 ) and eq.(a8 ) , can be combined to give the liouville von - neumann equations for the electronic density matrix , @xmath226 : @xmath227}\ ] ] further , the identity @xmath228 can be used to simplify the last term in eq.(a6 ) to @xmath229 using this , the equation of motion for the nuclear coordinates becomes @xmath230 } -\sum_{m}{\sigma_{mm}\nabla_{\bf r}e_{m}({\bf r } ) } -\nabla_{\bf r}u({\bf r})\nonumber\\ -\sum_{m , n}{\sigma_{nm}[(\dot{\bf r}\cdot\nabla){\bf a}_{mn } + \dot{\bf r}\times(\nabla\times{\bf a}_{nm})]}\end{aligned}\ ] ] the time derivative of @xmath110 can be rewritten as @xmath231 and hence cancels with the corresponding term on the right hand side .
thus , the equation of motion simplifies to : @xmath232}\ ] ] furthermore , the time derivative of the electronic density matrix , @xmath233 , can be replaced by using the liouville - von neumann equations of motion for the electronic density matrix , to obtain @xmath234_{nm}\ } } - \sum_{mn}{\sigma_{nm}\dot{\bf r}\times(\nabla_{\bf r}\times{\bf a}_{mn})}\end{aligned}\ ] ] the first three terms are the standard ehrenfest force acting on the nuclear variables , while the last two terms are new .
further , using the vector identity @xmath235 the penultimate term on the right hand side can be simplified and rewritten as @xmath236_{nm}\ } = -i\sigma_{nm}\dot{\bf r } \times({\bf a}\times{\bf a})_{mn}\ ] ] putting this together with eq.(a14 ) and defining an effective magnetic field by @xmath237 the equation of motion for the nuclear coordinate is obtained as @xmath238 this is the equation of motion discussed in the text of the paper , and is the ehrenfest equation of motion with an additional contribution from the `` magnetic field '' , @xmath239 . | this work identifies geometric effects on dynamics due to nonadiabatic couplings in born oppenheimer systems and provides a systematic method for deriving corrections to mixed quantum - classical methods .
specifically , an exact path integral formulation of the quantum nonadiabatic dynamics of born oppenheimer systems is described .
stationary phase approximations to the propagator for full quantum dynamics are derived .
it is shown that quantum corrections to mixed quantum classical methods can be obtained through stationary phase approximations to the full quantum dynamics .
a rigorous description of the quantum corrections due to electronic nonadiabatic coupling on the nuclear dynamics within the ehrenfest framework is obtained .
the fewest switches surface hopping method is shown to be obtained as a quasiclassical approximation to the dynamics and natural semiclassical extensions to include classically forbidden nonadiabatic transitions are suggested .
.4 in .1 in -2 in 6.35 in 8.75 in |
throughout , @xmath11 is a ring with identity element 1 .
suppose that @xmath12 , @xmath10 and @xmath13 ranging over the positive integers , are in @xmath11 and that @xmath14 whenever @xmath3 lies outside a fixed finite set
. then if @xmath15 is the infinite matrix @xmath16 , one may speak of @xmath17 for all @xmath18 , and one gets a generating function @xmath19 in @xmath20 $ ] , where @xmath21 is the ( 1,1 ) entry in the matrix @xmath17
. we shall prove : theorem i [ theorem1.1 ] suppose that @xmath14 if @xmath22 , and that @xmath23 unless @xmath24 .
suppose further that @xmath25 , @xmath1 a field , so that @xmath26 may be viewed as an @xmath27 by @xmath27 matrix with entries in @xmath28 $ ]
. then these matrix entries are algebraic over @xmath9 .
corollary [ corollary1.2 ] let @xmath1 be a field and @xmath2 , @xmath10 and @xmath13 ranging over the positive integers , be in @xmath1 .
suppose : 1 .
@xmath29 whenever @xmath3 lies outside a fixed finite set .
2 . for fixed @xmath30 in @xmath31
, @xmath32 is an eventually periodic function of @xmath10
. then if @xmath0 is the matrix @xmath33 , the generating function @xmath34 is algebraic over @xmath9 . to derive the corollary we choose @xmath27 so that : 1 .
@xmath29 whenever @xmath35 and @xmath36 or @xmath37 and @xmath38 .
@xmath39 whenever @xmath40 .
we then write the initial @xmath41 by @xmath41 block in @xmath0 as @xmath42 with @xmath43 , @xmath44 , @xmath45 , @xmath46 in @xmath47 .
our choice of @xmath27 tells us that @xmath0 is built out of @xmath27 by @xmath27 blocks , where the blocks along the diagonal are a single @xmath46 , followed by @xmath44 s , those just below a diagonal block are @xmath43 s , those just above a diagonal block are @xmath45 s , and all other entries are 0 .
now let @xmath25 and @xmath48 where @xmath49 , @xmath50 , @xmath51 , @xmath52 for @xmath53 , and all other @xmath12 are 0 .
view @xmath26 as an @xmath27 by @xmath27 matrix with entries in @xmath28 $ ] .
one sees easily that @xmath34 is the ( 1,1 ) entry in this matrix , and theorem i applied to @xmath15 gives the corollary .
remark [ remark1.3 ] when @xmath2 only depends on @xmath3 , the above corollary is due to gessel .
( when the matrix entries of @xmath0 are all @xmath54 s and @xmath55 s the result is contained in corollary 5.4 of @xcite .
the restriction on the matrix entries is nt essential in gessel s proof , as one can use a generating function for walks with weights . )
our proof of theorem i is easier than gessel s proof of his special case of the corollary .
the reason for this is that by working over @xmath11 rather than over @xmath1 we are able to restrict our study to walks with step - sizes in @xmath56 .
( a complication , fortunately minor , is that the weights must be taken in the non - commutative ring @xmath11 . )
our proof is well - adapted to finding an explicit polynomial relation between @xmath34 and @xmath57 ; we ll work out a few examples .
this paper would not have been possible without ira gessel s input .
i thank him for showing me tools of the combinatorial trade .
[ def2.1 ] if @xmath58 , an ordered @xmath59-tuple @xmath60 of integers is a ( motzkin ) walk of length @xmath61 if each of @xmath62 is in @xmath56 .
we say that the start of the walk is @xmath63 , the finish is @xmath64 , and that @xmath65 is a walk from @xmath63 to @xmath64 .
[ def2.2 ] if @xmath65 and @xmath66 are walks of lengths @xmath67 and @xmath68 , the concatenation @xmath69 of @xmath65 and @xmath66 is the walk @xmath70 of length @xmath71 .
now let @xmath11 be a ring with identity element 1 , and @xmath43 , @xmath44 , @xmath45 , @xmath46 lie in @xmath11 .
to each walk @xmath65 we attach weights @xmath72 and @xmath73 in @xmath11 : [ def2.3 ] if @xmath74 , @xmath75 . if @xmath76 , @xmath77 where @xmath78 , @xmath44 or @xmath45 according as @xmath79 is @xmath80 , @xmath54 , or @xmath55 .
the definition of @xmath73 is the same with one change : if @xmath81 then @xmath82 rather than @xmath44 .
evidently @xmath83 .
furthermore @xmath84 whenever @xmath65 and @xmath66 are walks from @xmath54 to @xmath54 .
[ def2.4 ] @xmath65 is `` standard '' if each @xmath85 . note that a walk from @xmath54 to @xmath54 is standard if and only if each @xmath86 . [ def2.5 ] @xmath65 is `` primitive '' if @xmath76 , @xmath87 and no @xmath88 with @xmath89 is @xmath63 .
note that a standard walk from @xmath54 to @xmath54 is primitive if and only if @xmath76 and each @xmath88 , @xmath89 , is @xmath90 . [ def2.6 ] 1 .
@xmath91 , the sum extending over all standard walks from @xmath54 to @xmath54 .
@xmath92 is the sum extending over all primitive standard walks from @xmath54 to @xmath54 .
2 . @xmath93 and @xmath94 are defined similarly , using @xmath73 in place of @xmath72 .
[ lemma2.7 ] let @xmath95 , @xmath96 .
then , in @xmath20 $ ] : 1 .
@xmath97 2 .
@xmath98 every standard walk from @xmath54 to @xmath54 of length @xmath90 is either primitive or uniquely a concatenation of two or more primitive standard walks from @xmath54 to @xmath54 .
the multiplicative property of @xmath99 now gives ( 1 ) . to prove (
2 ) note that the primitive standard walk @xmath100 has @xmath101 . and a primitive standard walk from @xmath54 to @xmath54 of length @xmath102 is a concatenation of @xmath103 , a standard walk , @xmath66 , from @xmath54 to @xmath54 of length @xmath104 and @xmath105
. then @xmath106 .
since @xmath107 gives a 11 correspondence between primitive standard walks of length @xmath67 from @xmath54 to @xmath54 and standard walks of length @xmath104 from @xmath54 to @xmath54 , we get the result .
[ corollary2.8 ] if @xmath95 , then @xmath108 in @xmath20 $ ] . by ( 1 ) of lemma [ lemma2.7 ] , @xmath109 .
substituting @xmath98 gives the result .
[ theorem2.9 ] suppose that @xmath110 , @xmath1 a field , so that @xmath111 may be viewed as an @xmath27 by @xmath27 matrix with entries in @xmath28 $ ] .
then these matrix entries , @xmath112 , are algebraic over @xmath9 .
let @xmath113 be an @xmath27 by @xmath27 matrix of indeterminates over @xmath1 , and @xmath114 be the @xmath115 entry in @xmath116 .
the @xmath114 are degree 2 polynomials in @xmath117 with co - efficients in @xmath118 $ ] . by corollary [ corollary2.8 ] , @xmath119 .
now @xmath120 where the @xmath121 are polynomials with co - efficients in @xmath1 .
it follows that the jacobian matrix of the @xmath114 with respect to the @xmath122 , evaluated at @xmath123 , is congruent to @xmath124 in the @xmath125 by @xmath125 matrix ring over @xmath28 $ ] , and so is invertible .
thus @xmath126 is an isolated component of the intersection of the hypersurfaces @xmath127 , and so its co - ordinates , @xmath128 , are algebraic over @xmath9 .
[ lemma2.10 ] @xmath129 .
the proof of lemma [ lemma2.7 ] ( 1 ) shows that @xmath130 with @xmath94 as in definition [ def2.6 ] .
so it suffices to show that @xmath131 .
now for a primitive walk @xmath65 of length @xmath132 from @xmath54 to @xmath54 one can not have @xmath133 , and so @xmath134 . on the other hand , for the primitive walk @xmath100 , @xmath101 and @xmath135 .
this gives the lemma .
combining lemma [ lemma2.10 ] with theorem [ theorem2.9 ] we get : [ theorem2.11 ] if @xmath136 the matrix entries of the @xmath27 by @xmath27 matrix @xmath93 are algebraic over @xmath9
. now let @xmath48 where @xmath49 , @xmath50 , @xmath51 , @xmath52 for @xmath53 , and all the other @xmath14 .
in view of theorem [ theorem2.11 ] the proof of theorem i will be complete once we show that @xmath137 where @xmath138 is the weight function of definition [ def2.3 ] .
the key to this is : [ lemma2.12 ] for @xmath139 let @xmath140 be @xmath141 , the sum extending over all standard walks of length @xmath6 from @xmath142 to @xmath54 . then
@xmath143 or @xmath54 according as @xmath144 or @xmath145 .
2 . @xmath146 .
if @xmath145 , @xmath147 .
lemma [ lemma2.12 ] has the following immediate corollaries , with the first proved by induction on @xmath6 .
[ corollary2.13 ] the first column vector in @xmath17 is @xmath148 .
[ corollary2.14 ] the @xmath149 co - efficient of @xmath17 is @xmath141 , the sum extending over all standard walks of length @xmath6 from @xmath54 to @xmath54 .
so @xmath137 .
it remains to prove lemma [ lemma2.12 ] .
( 1 ) is evident .
let @xmath65 be a standard walk of length @xmath6 from @xmath54 or @xmath55 to @xmath54 .
then @xmath150 is a standard walk of length @xmath151 from @xmath54 to @xmath54 , and @xmath152 is @xmath153 in the first case and @xmath154 in the second .
also each standard walk @xmath66 of length @xmath151 from @xmath54 to @xmath54 arises in this way from some @xmath65 ; explicitly @xmath155 . summing over @xmath66 we get ( 2 ) . similarly , suppose that @xmath145 and that @xmath65 is a standard walk of length @xmath6 from @xmath156 , @xmath142 or @xmath157 to @xmath54
. then @xmath158 is a standard walk of length @xmath151 from @xmath142 to @xmath54 and @xmath159 in the first case , @xmath160 in the second , and @xmath154 in the third . also , each standard walk @xmath66 of length @xmath151 arises from such an @xmath65 ; explicitly @xmath161 . summing over @xmath66
we get ( 3 ) , completing the proof .
[ remark2.15 ] to calculate the matrix entries of @xmath26 explicitly as algebraic functions of @xmath57 by the method of theorem [ theorem2.9 ] involves solving a system of @xmath125 quadratic equations in @xmath125 variables .
this is nt practical when @xmath162 ; in the next section we give a different proof of theorem [ theorem2.9 ] that is often better adapted to explicit calculations .
[ theorem3.1 ] @xmath163 , the sum extending over all length @xmath6 walks ( not necessarily standard ) with finish @xmath54 , is the element @xmath164 of @xmath165 $ ] .
denote the sum by @xmath166 . since @xmath167 it s enough to show that @xmath168 .
let @xmath169 be the co - efficient of @xmath170 in @xmath166 .
then @xmath171 , the sum extending over all length @xmath6 walks from @xmath157 to @xmath54 .
the proof of ( 3 ) of lemma [ lemma2.12 ] , using all walks rather than all standard walks , shows that @xmath172 for all @xmath157 in @xmath31 , giving the result .
[ def3.2 ] * @xmath173 , the sum extending over all @xmath54 to @xmath54 walks .
* @xmath174 is the sum extending over all @xmath80 to @xmath54 ( or @xmath54 to @xmath55 ) walks . *
@xmath175 is the sum extending over all @xmath55 to @xmath54 ( or @xmath54 to @xmath80 ) walks .
we ll generally omit the @xmath99 and just write @xmath176 , @xmath177 or @xmath175 .
[ corollary3.3 ] suppose that @xmath178 , @xmath80 or @xmath55 . then @xmath179 is the co - efficient of @xmath180 in the element @xmath181 of @xmath165[[z]]$ ] .
[ def3.4 ] @xmath182 is @xmath183 , the sum extending over all primitive @xmath54 to @xmath54 walks .
[ theorem3.5 ] 1 .
2 . @xmath185 .
\(1 ) follows from the multiplicative property of @xmath99 , as in the proof of lemma [ lemma2.7 ] .
so @xmath186 , and ( 2 ) asserts that @xmath187 .
if @xmath65 is a walk from @xmath54 to @xmath54 let @xmath188 be the number of ways of writing @xmath65 as a concatenation of a walk from @xmath54 to @xmath80 and a walk from @xmath80 to @xmath54 . also let @xmath189 be the number of ways of writing @xmath65 as a concatenation of a walk from @xmath54 to @xmath80 , a primitive walk from @xmath80 to @xmath80 and a walk from @xmath80 to @xmath54 .
the multiplicative property of @xmath99 shows that @xmath190 , the sum extending over all walks from @xmath54 to @xmath54 .
if @xmath65 is standard , @xmath191 .
if @xmath65 is not standard there is an @xmath10 with @xmath192 .
let @xmath193 be those @xmath10 with @xmath192 .
one sees immediately that @xmath194 and that @xmath195 .
so @xmath196 is the sum over the standard walks from @xmath54 to @xmath54 of @xmath197 , and this is precisely @xmath111 .
suppose now that @xmath110 , @xmath1 a field , so that @xmath176 , @xmath175 and @xmath177 may be viewed as @xmath27 by @xmath27 matrices with entries in @xmath28 $ ] .
theorem [ theorem3.5 ] , ( 2 ) , will give a new proof of theorem [ theorem2.9 ] once we show that these matrix entries are algebraic over @xmath9 .
the facts about the matrix entries of @xmath176 , @xmath175 and @xmath177 follow from a standard partial fraction decomposition argument well give our own version .
the algebraic closure of the field of fractions of @xmath28 $ ] is a valued field with value group @xmath198 .
let @xmath199 be the completion of this field and @xmath200 be the @xmath201 function in @xmath199 .
let @xmath202 consist of formal power series @xmath203 with @xmath204 and @xmath205 as @xmath206 .
@xmath202 has an obvious multiplication and is an overring of @xmath207[[z]]$ ] .
@xmath208 , @xmath209 and @xmath210 are the @xmath199-linear maps @xmath211 taking @xmath212 to @xmath213 , @xmath214 and @xmath215 .
note that @xmath216 , the algebraic closure of @xmath9 , imbeds in @xmath199 .
[ lemma3.6 ] suppose @xmath217 with @xmath218 .
then the element @xmath219 of @xmath202 is invertible , and for all @xmath139 , @xmath220 in @xmath202 with the @xmath221 in @xmath216 . in particular , @xmath208 , @xmath209 and @xmath210
take each @xmath222 to an element of @xmath216 . if @xmath223 , @xmath224 has inverse @xmath225 , while if @xmath226 , @xmath227 has inverse @xmath228 .
[ lemma3.7 ] let @xmath229 and @xmath230 be elements of @xmath231 $ ] .
suppose that @xmath232 for some @xmath27 .
then @xmath230 has an inverse in @xmath207[[z]]$ ] and the co - efficients of @xmath233 , @xmath234 and @xmath235 in the element @xmath236 of @xmath207[[z]]$ ] all lie in @xmath216 .
write @xmath230 as @xmath237 with @xmath238 in @xmath239 $ ]
. then @xmath240 is the desired inverse of @xmath230 .
if @xmath241 in @xmath199 has @xmath242 then @xmath243 has @xmath242 and can not be @xmath54 .
so when we factor @xmath230 in @xmath244 $ ] as @xmath245 with @xmath246 in @xmath118 $ ] and @xmath247 in @xmath216 , no @xmath248 can be @xmath54 .
view @xmath236 as an element of @xmath249 .
as such it is an @xmath216 linear combination of powers of @xmath250 and powers of the @xmath251 .
since @xmath208 , @xmath209 and @xmath210 are @xmath199-linear they are @xmath216-linear .
lemma [ lemma3.6 ] then tells us that @xmath236 , viewed as an element of @xmath202 , is mapped by each of @xmath208 , @xmath209 and @xmath210 to an element of @xmath216 .
this completes the proof .
[ lemma3.8 ] let @xmath43 , @xmath44 and @xmath45 be in @xmath47 and @xmath252[[z]]$ ] be an entry in the matrix @xmath253 .
then the co - efficients of @xmath233 , @xmath234 and @xmath235 in @xmath254 all lie in @xmath216 .
@xmath254 may be written as @xmath255 where @xmath229 and @xmath230 are in @xmath231 $ ] and @xmath256 .
then @xmath232 , and we apply lemma [ lemma3.7 ] . [ corollary3.9 ] if @xmath136 , @xmath1 a field , then the matrix entries of @xmath176 , @xmath175 and @xmath177 are algebraic over @xmath9 .
( so by theorem [ theorem3.5 ] the same is true of the matrix entries of @xmath111 . )
@xmath257 , and we combine lemma [ lemma3.8 ] with corollary [ corollary3.3 ] .
[ example4.1 ] for @xmath10 , @xmath13 positive integers define @xmath2 by : 1 .
@xmath258 if @xmath259 .
2 . @xmath258 if @xmath260 and @xmath10 is odd .
3 . all other @xmath2 are @xmath54 .
we calculate @xmath34 where @xmath261 .
if we take @xmath262 , ( 1 ) and ( 2 ) in the corollary to theorem i are satisfied , and @xmath263 , @xmath264 , @xmath265 .
let @xmath266 .
@xmath8 is a @xmath267 by @xmath267 matrix @xmath268 with entries in @xmath28 $ ] , and @xmath269 . by corollary [ corollary2.8 ] , @xmath270 .
two of the four equations this gives are : @xmath271 solving the first equation for @xmath272 and substituting in the second we find that @xmath273 is a root of : [ example4.2 ] for @xmath10 , @xmath13 positive integers define @xmath2 by : 1 .
@xmath258 if @xmath259 .
2 . @xmath258 if @xmath260 and @xmath10 is even . 3 .
all other @xmath2 are @xmath54 .
we calculate @xmath274 where @xmath275 .
since @xmath276 , condition ( 1 ) of the corollary to theorem i is not met when @xmath262 , and we instead take @xmath277 .
now @xmath278 let the entries in the first column of the @xmath279 by @xmath279 matrix @xmath95 be @xmath280 , @xmath281 , @xmath282 and @xmath283 .
examining the entries in the first column of the matrix equation @xmath284 we see : @xmath285 using maple to eliminate @xmath281 , @xmath282 , and @xmath283 from this system we find that @xmath286 is a root of : @xmath287 [ example4.3 ] for @xmath10 , @xmath13 positive integers define @xmath2 by : 1 .
@xmath258 if @xmath288 .
2 . @xmath258 if @xmath289 and @xmath290 .
3 . all other @xmath2 are @xmath54 .
we calculate @xmath34 where @xmath261 .
take @xmath291 . then : @xmath292 the determinant of the matrix @xmath293 is @xmath294 .
the splitting field of this polynomial over @xmath9 is the extension of @xmath9 generated by @xmath295 .
the arguments of section [ section3 ] show that @xmath176 , @xmath175 and @xmath177 have entries in this extension field .
it s not hard to write down these matrices explicitly using the partial - fraction decomposition argument .
theorem [ theorem3.5 ] and a maple calculation then show that the @xmath149 entry in @xmath111 is @xmath296 . since @xmath297 , @xmath298 , and this @xmath149 entry is the desired @xmath34 .
[ def5.1 ] suppose that @xmath110 , @xmath1 a field , and that @xmath43 , @xmath44 , @xmath45 , @xmath46 are in @xmath11 . then @xmath299 the field of fractions of @xmath28 $ ] is the extension field of @xmath9 generated by the matrix entries of the @xmath176 , @xmath175 and @xmath177 of definition [ def3.2 ] .
[ remark5.2 ] as we ve seen @xmath300 contains the matrix entries of @xmath111 and @xmath93 and is finite over @xmath9 .
indeed the proofs of lemmas [ lemma3.7 ] , [ lemma3.8 ] and corollary [ corollary3.9 ] show that @xmath299 a splitting field over @xmath9 of the polynomial @xmath301 .
one can say a bit more .
the above polynomial splits into linear factors in @xmath302 $ ] , and one may view its splitting field as a subfield of the valued field @xmath199 . by examining the partial - fraction decomposition
one finds that @xmath300 is fixed elementwise by each automorphism of the splitting field that is the identity on @xmath9 and permutes the roots that have positive @xmath201 among themselves .
the goal of this section is to show that some generating functions related to @xmath111 also have their matrix entries in @xmath300 .
these results will be used in a sequel to show the algebraicity ( under a conjecture ) of certain hilbert - kunz series and hilbert - kunz multiplicities . a standard walk from @xmath157 to @xmath54
can be written in just one way as the concatenation of a standard walk from @xmath157 to @xmath157 , the walk @xmath305 and a standard walk from @xmath142 to @xmath54 .
we introduce new weight functions @xmath313 and @xmath314 as follows .
replace @xmath11 , @xmath43 and @xmath45 by @xmath315 $ ] , @xmath316 and @xmath317 , and let @xmath313 and @xmath314 be the new @xmath99 and @xmath138 that arise . if @xmath318 is a walk from @xmath157 to @xmath54 then there are @xmath319 more steps of size @xmath80 in the walk than there are steps of size @xmath55 .
it follows that @xmath320 and @xmath321 are @xmath322 and @xmath323 .
in particular , @xmath324 and @xmath325 . applying lemma [ lemma5.6 ] in this new situation we find : @xmath326 in particular , the co - efficient of @xmath327 in the left - hand side of the above equation is @xmath54 . evaluating this co - efficient we get the lemma .
[ theorem5.8 ] let @xmath328 be elements of @xmath1 .
suppose there is a polynomial function whose value at @xmath13 is @xmath329 for sufficiently large @xmath13 .
let @xmath330 .
then all the matrix entries of @xmath331 lie in @xmath300 .
corollary [ corollary5.4 ] shows that the generating function arising from any single @xmath332 entry has matrix entries in @xmath300 .
so we may assume that @xmath333 is a polynomial function .
since any polynomial function is an @xmath1-linear combination of the functions @xmath334 , @xmath335 we may assume @xmath336 .
but then @xmath331 is @xmath337 , and we use lemmas [ lemma5.6 ] , [ lemma5.7 ] and induction .
suppose further that @xmath328 are in @xmath1 and that for each @xmath10 , @xmath340 , there is a polynomial function agreeing with @xmath341 for large @xmath157 .
let @xmath342 be the @xmath343 entry in @xmath5 .
then @xmath344 is in @xmath300 .
construct @xmath15 as in the proof of the corollary to theorem i. as the first column of @xmath17 is @xmath345 it follows that @xmath346 is just the @xmath343 entry in the @xmath27 by @xmath27 matrix @xmath347 .
theorem [ theorem5.8 ] shows that for each @xmath10 with @xmath340 , @xmath348 is in @xmath300 . summing over @xmath10 we get the result .
[ lemma5.10 ] let @xmath349 be a finite dimensional vector space over @xmath1 , @xmath350 and @xmath351 linear maps and @xmath352 a sequence in @xmath349 .
let @xmath0 and @xmath27 be as in corollary [ corollary5.9 ] .
suppose that for each @xmath10 , @xmath340 , each co - ordinate of @xmath353 with respect to a fixed basis of @xmath349 is an eventually polynomial function of @xmath157 .
define @xmath354 inductively by @xmath355 , @xmath356see corollary [ corollary5.9 ] for the definition of @xmath342 .
then @xmath357 is in @xmath300 .
@xmath358 . by corollary [ corollary5.9 ] , all the co - ordinates of @xmath359 with respect to a fixed basis of @xmath349 lie in @xmath300 .
since @xmath360 is a non - zero element of @xmath361 , the same is true of the co - ordinates of @xmath362 , giving the lemma .
[ theorem5.11 ] suppose @xmath363 is a vector space over @xmath1 , @xmath349 is a finite dimensional subspace , @xmath364 is linear with @xmath365 , and @xmath366 lie in @xmath363 .
suppose further that @xmath367 , where @xmath261 is as in lemma [ lemma5.10 ] and @xmath352 is a sequence in @xmath349 satisfying the condition of lemma [ lemma5.10 ] .
then if @xmath368 is linear with each @xmath369 , the power series @xmath370 is in @xmath300 .
[ example5.12 ] suppose @xmath374 and @xmath375 are a basis of @xmath349 , that @xmath376 and that @xmath377 , @xmath145 .
suppose further that @xmath378 , @xmath379 , @xmath380 , and that @xmath381 for @xmath145 .
suppose @xmath368 takes @xmath374 to @xmath55 , and @xmath375 and each @xmath382 to @xmath54 .
we shall calculate the power series @xmath383 explicitly . in the above situation ,
@xmath384 and all other @xmath2 are @xmath54 .
so we can take @xmath385 , @xmath386 and @xmath387 . since @xmath385 , @xmath388 .
it follows from this and the definition of the @xmath389 that @xmath390 .
now the matrix of @xmath350 on the basis @xmath391 is @xmath392 .
it follows that the matrix of @xmath393 is @xmath394 with inverse @xmath395 .
since @xmath396 is the co - efficient of @xmath374 in @xmath397 , the last paragraph shows that @xmath398 .
it only remains to calculate @xmath93 , @xmath399 and @xmath400 .
lemma [ lemma2.7 ] and corollary [ corollary2.8 ] show that @xmath401 , and @xmath402 .
so @xmath111 and @xmath92 are @xmath403 and @xmath404 .
lemma [ lemma2.10 ] then shows @xmath405 . making use of lemmas [ lemma5.6 ] and [ lemma5.7 ]
we find that @xmath399 and @xmath400 are @xmath406 and @xmath407 .
a brief calculation then gives the explicit formula : @xmath408 | let @xmath0 be an infinite matrix with rows and columns indexed by the positive integers , and entries in a field @xmath1 .
suppose that @xmath2 only depends on @xmath3 and is 0 for @xmath4 large .
then @xmath5 is defined for all @xmath6 , and one has a `` generating function '' @xmath7 .
ira gessel has shown that @xmath8 is algebraic over @xmath9 .
we extend his result , allowing @xmath2 for fixed @xmath3 to be eventually periodic in @xmath10 rather than constant .
this result and some variants of it that we prove will have applications to hilbert - kunz theory . |
Subsets and Splits