article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
the single - nucleon transfer reactions , induced by light as well as heavy ions , have been established as a useful tool in probing the single - particle components of the wave functions of stable nuclei ( see e.g. @xcite ) .
the theory of these reactions developed within the framework of the distorted wave born approximation ( dwba ) @xcite has been widely used to analyze the absolute magnitudes and shapes of measured cross sections and to deduce the structure information including angular momentum assignments , occupation probabilities and spectroscopic factors of the ground as well as excited states of the residual nuclei .
nonetheless , transfer reactions are not yet routinely used in probing the structure of exotic nuclei near the neutron and proton drip lines , even though the first theoretical feasibility study @xcite for such investigations with transfer reactions and the first experimental results @xcite for the @xmath1be(p , d)@xmath4be reaction have been already reported . with the currently available experimental techniques ,
the measurements of these reactions involving drip line nuclei are performed in the inverse kinematics with low intensity projectile beams .
this puts severe experimental restrictions as the corresponding cross sections are usually low .
furthermore , the theoretical analysis of these data in terms of the dwba gets complicated as the usual well - depth search method to calculate the wave function of the transfered particle becomes unreliable @xcite , and the methods such as skyrme hartree - fock theory need to be invoked @xcite for a proper description of these wave functions .
recently , an alternative new and more versatile technique for investigating the spectroscopy of nuclei near the drip line has been developed @xcite . in this method , referred to as the ( a , b@xmath0 ) reaction in the following , one nucleon ( usually the valence or halo ) is removed from the projectile ( a ) in its breakup reaction within the field of a target nucleus .
the states of the core fragment ( b ) populated in this reaction are identified by their gamma ( @xmath0 ) decay .
the @xmath0-ray intensities are used to determine the partial breakup cross sections to different core states .
the signatures of the orbital angular momentum @xmath5 associated with the relative motion of core states with respect to the valence nucleon ( removed from the projectile ) are provided by the measured parallel momentum distributions @xcite .
this method improves the experimental conditions for working with projectiles of low beam intensities because of : ( i ) large partial cross sections for transitions to various bound states of the core fragment , even in experiments done with high - energy projectiles , ( ii ) possibility of using thick targets , and ( iii ) strong forward focusing .
these features may be contrasted with those of the corresponding transfer reactions .
in addition , while , in the case of transfer reactions , the angular distributions of the ejectile lose their characteristic @xmath5-dependence at high energies @xcite , the longitudinal momentum distributions of the core states in the breakup reactions continue to show a strong dependence on @xmath5 .
most of the studies of the ( a , b@xmath0 ) reaction performed so far involve a light @xmath6be target , where the breakup process is governed almost entirely by only the nuclear interaction between the projectile fragments and the target . since , this reaction is essentially inclusive in nature ( as the measurements are performed only for the heavy core fragment ) , the nuclear partial cross sections have contributions from both elastic ( also known as diffraction dissociation ) and inelastic ( also known as stripping or breakup - fusion ) breakup modes @xcite .
several attempts have been made to calculate the elastic and inelastic nuclear breakup cross sections of halo nuclei and they were either based on the semiclassical methods @xcite or on the eikonal approximation @xcite .
the fragment - target interactions are dealt with differently in these two approaches which could be important for the light targets @xcite .
data of refs .
@xcite have been analyzed in terms of an eikonal model @xcite with core - target and neutron - target interactions treated in the black disc approximation and in the optical limit of the glauber theory , respectively . in order to extract unambiguous spectroscopic informations from the ( a , b@xmath0 ) type of measurements performed on a light target , it is quite desirable to develop the calculations of nuclear breakup reactions within the dwba theory as has been done for the breakup of stable projectiles @xcite . however , currently a full quantum mechanical theory of the pure coulomb breakup reaction , formulated within the framework of the post - form distorted - wave born - approximation , is well established and has been applied successfully to investigate the breakup of halo nuclei @xcite .
finite range effects are accounted for in this theory which can be applied to projectiles of any ground - state orbital angular - momentum structure .
moreover , an alternative theory of the coulomb breakup reactions within the framework of an adiabatic model has also been formulated @xcite .
the expressions for the breakup amplitude within this theory are very similar to those of the finite - range dwba theory , although the two have been derived under quite different assumptions . in the adiabatic model , it is assumed that the excited states of the projectile are degenerate with the ground state . in the studies of the breakup reactions done so far ( where the core fragments were assumed to remain in their ground states ) , the two theories produced almost identical results @xcite . however , with the excitation of the core , the one - neutron separation energies increase significantly .
it would , therefore , be interesting to see if the two models lead to different results in these cases .
there are no adjustable parameter in either of the theories of pure coulomb breakup reaction . assuming that the processes , in which the mutual excitation of the target nucleus takes place due to the coulomb interaction , contribute negligibly ,
the inelastic breakup mode is absent in the pure coulomb breakup reactions .
this is an added advantage as there is some ambiguity regarding the calculation of this mode which dominates the partial cross sections for the excited core states in the nuclear breakup process @xcite . in this paper
, we present calculations of the pure coulomb breakup contributions to the partial cross sections and longitudinal momentum distributions of the ground as well as excited states of the core fragments , @xmath4be and @xmath7c , in the ( a , b@xmath0 ) type of reaction induced by @xmath1be and @xmath2c projectiles , respectively on a @xmath3pb target at the beam energy of 60 mev / nucleon .
we assume that the states of the core fragments are the same as those seen in the similar reactions studied on the @xmath6be target .
our aim is to determine if there are quantitative differences in the @xmath8 populations of the core states in the pure coulomb breakup mechanism , as compared to those observed in the pure nuclear breakup process .
we shall also look whether there are differences in the predictions of the finite range dwba and adiabatic models of the breakup reactions leading to the core excited states .
we want to make it clear from the very beginning that it is not our intention to imply that the nuclear breakup contributions are negligible for the reactions investigated by us .
our results should be viewed as complementing contributions from the nuclear breakup process ; in any complete theory both contributions must be considered on an equal footing . in the next section
we briefly present the formalism of the coulomb breakup reactions .
the results of our calculations and discussions are presented in section iii . a summary and the conclusions of our work
are given in section iv .
we consider the reaction @xmath9 , where the projectile @xmath10 breaks up into fragments @xmath11 ( charged ) and @xmath12 ( uncharged ) in the coulomb field of a target @xmath13 .
the chosen coordinate system is shown in fig . 1 .
the position vectors satisfy the following relations : @xmath14 the starting point of both the finite - range distorted - wave born approximation ( frdwba ) and of the adiabatic model of the coulomb breakup is the post - form @xmath15-matrix of the reaction given by @xmath16 the functions @xmath17 are the distorted waves for the relative motions of @xmath11 and @xmath12 with respect to @xmath13 and the center of mass ( c.m.)of the @xmath18 system , respectively .
the functions @xmath19 are the internal state wave functions of the concerned particles dependent on the internal coordinates @xmath20 .
the function @xmath21 is the exact three - body scattering wave function of the projectile with a wave vector @xmath22 satisfying outgoing boundary conditions .
the vectors @xmath23 and @xmath24 are the jacobi wave vectors of @xmath11 and @xmath12 , respectively , in the final channel of the reaction .
the function @xmath25 represents the interaction between @xmath11 and @xmath12 .
as we concentrate only on the pure coulomb breakup , the function @xmath26 is taken as the coulomb distorted wave ( for a point coulomb interaction between the charged core b and the target ) satisfying incoming wave boundary conditions , and the function @xmath27 is just a plane wave as there is no coulomb interaction between the target and the neutral fragment @xmath12 . in the distorted wave born approximation ( dwba ) , we write @xmath28 the assumption inherent in eq .
( 4 ) is that the breakup channels are very weakly coupled and hence this coupling needs to be treated only in the first order . in this equation the dependence of @xmath29 on @xmath30 describes the relative motion of the fragments @xmath11 and @xmath12 in the ground state of the projectile .
the function @xmath31 is the coulomb distorted scattering wave describing the relative motion of the c.m . of the projectile with respect to the target , satisfying outgoing wave boundary conditions .
it may be noted that the particular case of the pure coulomb breakup of a projectile involving one uncharged fragment , where the choice of the coordinate * r * may appear more natural to describe the relative motion between the projectile and the target , follows from this expression as discussed below . the integration over the internal coordinates @xmath20 in the @xmath15-matrix gives @xmath32 with @xmath33 in eq .
( 6 ) , @xmath5 ( the orbital angular momentum for the relative motion between fragments @xmath11 and @xmath12 ) is coupled to the spin of @xmath12 and the resultant channel spin @xmath34 is coupled to the spin @xmath35 of the core @xmath11 to yield the spin of @xmath10 ( @xmath36 ) .
the @xmath15-matrix can now be be written as @xmath37 where @xmath38 with @xmath39 being the reduced @xmath15-matrix and with @xmath40 .
equation ( 8) involves a six dimensional integral which makes the computation of @xmath39 quite complicated .
the problem gets further acute because the integrand has a product of three scattering waves that exhibit an oscillatory behavior asymptotically . therefore ,
approximate methods have been used , such as the zero range approximation ( zra ) ( see e.g. @xcite ) , in which the product @xmath41 is replaced by a delta function , or the baur - trautmann approximation @xcite , where the projectile c.m .
coordinate is replaced by that of the core - target system ( i.e. @xmath42 ) .
both these approximations lead to a factorization of the reduced amplitude into two independent parts , which reduces the computational complexity . however , the application of both these methods to the reactions of halo nuclei is questionable @xcite .
the zra necessarily restricts the relative motion between @xmath11 and @xmath12 in the projectile to @xmath43-state only . even for such cases ,
this approximation may not be valid for heavier projectiles and at higher beam energies ( see e.g. @xcite ) .
the baur - trautmann approximation is justified if the c.m . of the @xmath11+@xmath12 system
is shifted towards @xmath11 ( which is indeed the case if @xmath44 ) . however , since * r*@xmath45 occurs in association with the wave vector * k*@xmath46 , whose magnitude is quite appreciable at the higher beam energies , the neglected piece of * r*@xmath45 ( i.e. @xmath47 * r*@xmath48 ) may still contribute substantially . in the frdwba theory , the coulomb distorted wave of particle @xmath11 in the final channel is written as @xcite @xmath49 equation ( 9 ) represents an exact taylor series expansion about @xmath50 if @xmath51 is treated exactly .
however , instead of doing this we employ a local momentum approximation @xcite , where the magnitude of momentum @xmath52 is taken to be @xmath53 here @xmath54 is the reduced mass of the @xmath55 system , @xmath56 is the energy of particle @xmath11 relative to the target in the c.m .
system and @xmath57 is the coulomb potential between @xmath11 and the target separated by @xmath58 .
thus , the magnitude of the momentum of @xmath59 is evaluated at some separation @xmath58 which is held fixed for all the values of @xmath60 .
the value of @xmath58 was taken to be equal to 10 fm . for reactions under investigation in this paper ,
the magnitude of @xmath61 remains constant for distances larger than 10 fm @xcite . due to the peripheral nature of the breakup reaction
, the region @xmath62 fm contributes maximum to the cross section . in fact , the calculated cross sections change by only about 5@xmath63 if @xmath58 is varied between 5 to 10 fm and with a further increase in r the change is less than 1@xmath63 .
furthermore , the results of the calculations for these reactions , at the beam energies under investigation , are almost independent of the choice of the direction of momentum @xmath52@xcite .
therefore , we have taken the directions of @xmath59 and @xmath64 to be the same in all the calculations presented in this paper .
it may be remarked here that in ref .
@xcite an approximation similar to eq .
( 9 ) was applied to the coulomb distorted wave of the incident channel .
that procedure brings in two difficulties .
firstly , the choice of the direction of the local momentum is somewhat complicated as directions of the both fragments in the final channel will have to be brought into consideration .
secondly , the procedure may produce a deviation from the exact dwba approximation . on substituting eq .
( 9 ) into eq .
( 8) , we obtain the following factorized form of the reduced amplitude @xmath65 \nonumber \\ & \times & \left [ \int d{{\bf r}_i}\chi^{(-)*}_b({{\bf k}_b},{{\bf r}_i } ) e^{-i\delta{{\bf k}_c}.{{\bf r}_i } } { \chi^{(+)}}_a({{\bf k}_a},{{\bf r}_i } ) \right ] .\end{aligned}\ ] ] this amplitude differs from those in earlier studies @xcite since it includes the interaction @xmath66 to all orders .
recently , an alternative theory of the coulomb breakup has been developed within the adiabatic ( ad ) model @xcite .
this theory assumes ( i ) that one of the fragments ( the valence nucleon ) is neutral so that the projectile interacts with the target only through the coulomb interaction @xmath67 of the core fragment and the target nucleus , and ( ii ) that the relative excitation energy @xmath68 of the b - c system is much smaller than the total incident energy so that @xmath68 can be replaced by the constant separation energy of the fragments in the projectile ground state . following ( ii )
the continuum spectrum of the @xmath69 system is assumed to be degenerate with the ground state . under the above assumptions , the wave function @xmath21 is found @xcite in the form @xmath70 it is clear that substitution of eq .
( 12 ) to eq .
( 3 ) will lead to a factored form ( similar to eq .
( 11 ) ) of the reduced breakup amplitude .
however , one limitation of this procedure should be brought into attention . for larger values of @xmath71 , the wave function @xmath72 vanishes due to the presence of the factor @xmath73 , whereas there may still be contributions to the breakup from this region .
it has been argued @xcite that , due the presence of the interaction @xmath74 , the post form breakup amplitude may not be sensitive to the domain where @xmath72 is inaccurate .
however , since the wave functions for the relative motion of the fragments for @xmath75 values have a large spatial extension , the application of this model to such cases may test the need for the non - adiabatic corrections to the theory .
the reduced amplitude in the adiabatic model is given by , @xmath76 \nonumber \\ & \times & \left [ \int d{{\bf r}_i}\chi^{(-)*}_b({{\bf k}_b},{\bf { { \bf r}_i } } ) e^{-i\delta{{\bf k}_c}\cdot { \bf { { \bf r}_i}}}{\chi^{(+)}}_a({{\bf k}_a},{\bf { { \bf r}_i } } ) \right ] \end{aligned}\ ] ] it is obvious that this amplitude differs from that of the frdwba , eq . (
11 ) , only in the form factor part ( the first of the factors ) , which is evaluated here at the momentum transfer of @xmath77 .
equation ( 13 ) can also be obtained in the dwba model by making a local momentum approximation to the coulomb distorted wave in the initial channel of a reaction and by evaluating the local momentum at @xmath78 with the momentum direction being the same as that of the projectile . in both of the theories , the coulomb interaction between the fragments @xmath11 and
the target is treated non - perturbatively .
the adiabatic model does not make the weak coupling approximation of the dwba .
however , it necessarily requires one of the fragments ( in this case @xmath12 ) to be neutral .
in contrast , the frdwba model can , in principle , be applied to the cases where both of the fragments @xmath11 and @xmath12 are charged @xcite .
furthermore , calculation of the nuclear breakup in the adiabatic model is not as comparatively trivial @xcite , as it is in the case of frdwba .
the triple differential cross section of the reaction is given by @xmath79 where @xmath80 is the appropriate @xcite three - body phase space factor .
on substituting the coulomb distorted waves , @xmath81 into eqs . ( 11 ) and ( 13 ) , one gets for the triple differential cross section : @xmath82 in eqs . ( 1517 ) , @xmath83 s are the coulomb parameters for the respective particles . in eq .
( 17 ) , @xmath84 is the bremsstrahlung integral @xcite which can be evaluated in the closed form : @xmath85 } \
, , \end{aligned}\ ] ] where @xmath86},\end{aligned}\ ] ] @xmath87 with @xmath88 the factor @xmath89 contains the projectile structure information and is given by @xmath90 with @xmath91 , and @xmath92 for the cases of frdwba and adiabatic model , respectively .
the total pure coulomb one - nucleon removal cross section for a given @xmath93 configuration of the valence nucleon is obtained by integrating eq .
( 13 ) over angles and energy of fragment @xmath11 and over angles of the valence nucleon . here , @xmath94 is the principal quantum number and @xmath5 and @xmath34 are as defined in eq .
( 5 ) . for calculating the total cross section into a given core - fragment final - state ,
the projectile ground state is described as having a configuration in which a valence nucleon , with single particle quantum numbers @xmath93 and an associated spectroscopic factor @xmath95 , is coupled to a specific core state designated with @xmath35 in eq .
the total cross section @xmath96 is the sum @xcite of the cross sections calculated with configurations ( having non - vanishing spectroscopic factors ) corresponding to all the allowed values of the channel spin @xmath34
the one - neutron removal reaction of the type @xmath6be(@xmath1be,@xmath4be@xmath0)x has been recently studied @xcite at the beam energy of 60 mev / nucleon .
partial cross sections have been measured for four states of the core fragment @xmath4be : 0@xmath97 , 2@xmath97 , 1@xmath98 , and 2@xmath98 .
the data were analyzed in terms of an eikonal model of the nuclear breakup reactions @xcite , with the spectroscopic factors taken from @xcite .
it has been concluded in this study that about @xmath99 of the total partial cross section went into the excited states , and that the ground state of @xmath1be consists of an admixture of the 1@xmath43 and 0@xmath100 single particle neutron configurations with the spectroscopic factors of 0.74 and 0.18 , respectively .
we have calculated the pure coulomb partial cross sections @xmath96 to the four @xmath4be final states in the @xmath3pb(@xmath1be,@xmath4be@xmath0)x reaction at the beam energy of 60 mev / nucleon . the ground ( 0@xmath97 ) and excited ( 3.368 mev ) ( 2@xmath97 ) states were assumed to correspond to the configurations [ 1@xmath101 @xmath102 0@xmath97(@xmath4be ) ] and [ 0@xmath103 @xmath102 2@xmath97(@xmath4be ) ] , respectively , where @xmath104 represents a relative neutron state .
the corresponding @xmath95 values for these two configurations were taken @xcite to be 0.74 and 0.20 , respectively , i.e. the same as those used in @xcite .
the excited 1@xmath98 ( 5.956 mev ) and 2@xmath98 ( 6.256 mev ) states were assumed to stem from the configurations [ 0@xmath105 @xmath102 1@xmath98(@xmath4be ) ] and [ 0@xmath105 @xmath102 2@xmath98(@xmath4be ) ] , respectively , with the corresponding @xmath95 values of 0.69 and 0.58 .
these states could , in principle , also result from the stripping of a 1@xmath106 neutron from the @xmath4be(0@xmath97 ) core of the @xmath1be ground state , producing a [ 1s@xmath107 @xmath102 @xmath6be@xmath1081@xmath98 and [ 1s@xmath107 @xmath102 @xmath6be@xmath1082@xmath98 types of @xmath4be@xmath109 core .
in the nuclear breakup case , the cross sections to 1@xmath98 and 2@xmath98 states calculated with the latter configurations were found @xcite to be about 10 - 15@xmath63 smaller than those obtained with the preceding ones .
we have carried out our calculations with the former configuration for these states .
the one - neutron separation energy for the ground state of @xmath1be , with the configuration in which @xmath4be remains in its ground state , is taken to be @xmath110 mev .
for an excited state , the respective separation energy ( se ) is assumed to be the sum of @xmath111 and the excitation energy of that state with respect to the ground state . in each case , the neutron single particle wave function is calculated in a central woods - saxon well of radius 1.15 fm and diffuseness 0.50 fm .
the depth of this well is adjusted to reproduce the corresponding value of se . by this procedure
the root mean square ( rms ) radius of the ground state of @xmath1be comes out to be 2.91 fm for the assumed rms radius of the @xmath4be core of 2.28 fm .
our results for the partial cross sections are shown in table i. it is evident from this table that in the case of pure coulomb breakup of a projectile with a halo ground state , most of the cross section goes to the ground state ( 0@xmath97 ) of the core .
the sum of the partial .calculated partial cross sections to the final states of @xmath4be in the coulomb breakup of @xmath1be on the @xmath3pb target at the beam energy of 60 mev / nucleon .
@xmath112 represents the spin and parity of the populated states of the @xmath4be core .
[ cols="^,^,^,^,^,^ " , ] due to their non-@xmath43-wave nature , the wave functions for the excited - state neutron - core motion peak at larger values in the @xmath113-space yielding possibly a significance to the regime where the asymptotic form of the adiabatic wave function [ eq . ( 12 ) ] becomes inadequate .
it would , therefore , be interesting to investigate the importance of the non - adiabatic corrections @xcite to the theory , for these cases .
the full width at half maximum ( fwhm ) of the calculated lmd for the ground state of @xmath4be is 44 mev / c which is consistent with the experimental value of ( 47.5 @xmath114 6 ) mev / c seen in the measurements on a @xmath6be target @xcite .
this reconfirms that lmds are independent of the reaction mechanism and provide a very clean way of determining the existence of halo structure in nuclei .
the lmds for the excited states of @xmath4be are broad which is also consistent with the observations made in @xcite .
this indicates that the respective states have a non - halo structure .
table ii displays results of our calculations of the pure coulomb partial cross sections @xmath96 for transitions to ground and three excited bound states of @xmath7c core in the @xmath3pb(@xmath2c,@xmath7c@xmath0)x reaction at the beam energy of 60 mev / nucleon .
these states have recently been seen @xcite in the @xmath6be(@xmath2c,@xmath7c@xmath0)x reaction at the same beam energy .
the states of the @xmath7c core ( with excitation energies of 0.0 mev , 1.6 mev , 4.0 mev and 4.9 mev ) are assumed to have the configurations , [ 1@xmath101 @xmath102 0@xmath97(@xmath7c ) ] , [ 0@xmath103 @xmath102 2@xmath97(@xmath7c ) ] , [ 1@xmath101 @xmath102 0@xmath97(@xmath7c ) ] , and [ 0@xmath103 @xmath102 i@xmath115(@xmath7c ) ] , respectively .
the corresponding @xmath95 values were taken @xcite to be 0.58 , 0.48 , 0.32 and 2.44 , respectively , which are the same as those used in @xcite .
the value of @xmath111 for the ground state was taken to be 0.530 mev .
we see that in this case too the ground state of @xmath7c is predominantly excited .
the partial cross sections to the excited states are somewhat larger than those seen in the case of @xmath4be , since the excited @xmath116 state of @xmath7c core can have an @xmath43-wave neutron relative motion . yet , these contributions represent no more than about 2@xmath63 of the cross section to the ground state . the lmd for each of the @xmath7c core states is shown in fig .
the solid and dashed lines show the results of the frdwba and adiabatic models , respectively . in this case
too we note that the predictions of the two models differ for the excited states of the core , while for the ground state they agree very well with each other .
.3 in the value of @xmath111 for @xmath2c is still an unsettled issue .
the weighted average of the atomic mass measurements carried out at los alamos and ganil @xcite suggests a value of 0.16 @xmath114 0.11 mev .
however , from the analysis @xcite of the data on the coulomb dissociation of @xmath2c , a higher value of 0.530 mev has been extracted .
the interpretation of the recent data @xcite on the @xmath6be(@xmath2c,@xmath7c@xmath0)x reaction also suggests a higher value of [email protected] mev .
obviously , any conclusion drawn from the breakup data strongly depends on the reaction mechanism and on the theory used for the calculation of the breakup cross sections . with this precaution , we would like to show here that the pure coulomb breakup has some advantages over the nuclear breakup process in this regard . it should be mentioned here that one of the reasons for the uncertainty in the value of @xmath111 in ref .
@xcite is the fact that , due to the low beam intensity in this experiment , the statistical errors associated with the measured lmd for the ground state of @xmath7c are large .
the data do not allow to distinguish between the nuclear breakup calculations of the lmd done within the range @xmath111 = 800 @xmath114 300 kev .
the difference in the peak value of the nuclear lmd @xcite calculated with @xmath111 = 1100 kev and 500 kev is only about 1.6 .
in contrast , the peak values of the corresponding pure coulomb lmd calculated with the same values of @xmath111 differ by a factor of about 4 , as can be seen in fig . 4 .
this result is unlikely to be altered by the presence of the nuclear breakup effects , as they tend to show up in the tail regions of the lmds .
thus a(@xmath2c,@xmath7c(g.s))x type of reactions on a heavy target may offer a better chance to put more definite constraint on the value of @xmath111 for @xmath2c .
.3 in further insight into the value of @xmath111 from these reactions can be obtained from the full width at half maximum ( fwhm ) of the lmd . in fig . 5
, we show the @xmath111 dependence of the fwhm of the lmd for the @xmath7c(g.s ) in the same reaction as in fig .
it can be seen that fwhm increases from 42 mev / c to about 65 mev / c as @xmath111 increases from 500 kev to 1100 kev .
the variation of the corresponding fwhm in the nuclear breakup case is relatively weaker .
in this paper we calculated the pure coulomb breakup contributions to the partial cross sections and to the longitudinal momentum distributions for the ground and excited states of the core fragments observed in @xmath3pb(@xmath1be,@xmath4be@xmath0)x and @xmath3pb(@xmath2c,@xmath7c@xmath0)x types of one - neutron removal reactions , at the beam energy of 60 mev / a .
these reactions have recently been studied at the michigan state university but on the light @xmath6be target ; hence , these data are dominated by the nuclear breakup effects
. one of our aims was to see in what way the coulomb dominated reaction mechanism was different and could supplement the conclusions derived from the pure nuclear breakup studies of the nuclei .
the advantage of the pure coulomb break up process is that the corresponding theory has no free adjustable parameter , and assuming that the mutual excitation of the target by the coulomb force is negligible , the inelastic breakup mode does not contribute to this process . as in the previous studies
@xcite , we assumed that the coupling between the core states is weak and that there is no dynamical excitation of these states .
thus , the reaction can only populate those core states which have a non - zero spectroscopic factor for a given neutron - core configuration in the projectile ground state .
we employed both the finite range dwba and adiabatic model of the coulomb breakup theory in our calculations . in earlier studies @xcite of the inclusive coulomb dissociation cross sections , the two theories produced nearly identical results for the momentum distributions of heavy fragments .
we found that in reactions of the type a(a , b@xmath0)x on a heavy target , the core ground state is predominantly excited ; higher energy states account for only a few percent of the total cross section .
this finding is in contrast to the results obtained on similar reactions on a light target , where about a quarter of the total breakup cross section could be due to transitions to core excited states .
our finding is supported by a recent measurement @xcite of the @xmath117au(@xmath118,@xmath119b@xmath0)x reaction at the beam energy of 60 mev .
the reason for this difference is attributed to the fact that pure coulomb breakup cross sections drop very strongly as the separation energy increases . on the other hand ,
the nuclear breakup cross sections decrease slowly with increasing se
. therefore , such reactions on a heavy target are potentially a more useful tool for investigating the properties of the ground state of the core fragments .
a rather interesting result of our study is that the finite - range dwba and the adiabatic theories of coulomb breakup lead to very different longitudinal momentum distributions for the excited states of the core fragments .
it is probably for the first time that such a large difference is seen in the predictions of two theories for the momentum distributions .
this should provide some impetus to look for the non - adiabatic corrections to the adiabatic approximation which may come under some pressure for the excited states .
coulomb dominated breakup reactions may provide a better way for resolving the uncertainty associated with the one neutron separation energy of @xmath2c .
the peak value and the full width at half maximum of the longitudinal momentum distributions for the ground state of @xmath7c core are more sensitive to the one - neutron separation energy in the coulomb breakup process than in the nuclear breakup . in the latter case
, the dependence could be so weak that the data with limited statistics may not allow to distinguish between the values of these quantities calculated with quite different one - neutron separation energies .
angela a. bonaccorso and d.m .
brink , phys .
c * 38 * , 1776 ( 1988 ) ; @xmath120 phys .
c * 58 * , 2864 ( 1998 ) ; a. bonaccorso , phys . rev .
c * 60 * , 054604 ( 1999 ) ; a. bonaccorso and f. carstoiou , phys .
c * 61 * , 034605 ( 2000 ) .
ber96 k. henken , g.f .
bertsch , and h. esbensen , phys .
c * 54 * , 3043 ( 1996 ) ; g.f .
bertsch , k. henken and h. esbensen , phys .
c * 57 * , 1366 ( 1998 ) ; h. esbensen and g. bertsch , phys .
c * 59 * , 3240 ( 1999 ) . | within the pure coulomb breakup mechanism , we investigate the one - neutron removal reaction of the type a(a , b@xmath0)x with @xmath1be and @xmath2c projectiles on a heavy target nucleus @xmath3pb at the beam energy of 60 mev / nucleon . our intention is to examine the prospective of using these reactions to study the structure of neutron rich nuclei .
integrated partial cross sections and momentum distributions for the ground as well as excited bound states of core nuclei are calculated within the finite range distorted wave born approximation as well as within the adiabatic model of the coulomb breakup .
our results are compared with those obtained in the studies of the reactions on a light target where the breakup proceeds via the pure nuclear mechanism .
we find that the transitions to excited states of the core are quite weak in the coulomb dominated process as compared to the pure nuclear breakup . |
the observation of quantum phenomena , such as quantum entanglement @xcite , has been mainly limited to systems of only few particles .
one of the main open challenge for an experimental test in systems of large size is the construction of suitable criteria for the detection of entanglement in bipartite macroscopic systems .
much effort has been devoted in the last few years in this direction . some of them , such as the partial transpose criterion developed by peres in ref.@xcite , require the tomographic reconstruction of the density matrix , which for system of a large number of particles becomes highly demanding from an experimental point of view . in order to avoid the necessity of the complete reconstruction of the state , a class of tests where only few local measurements are performed has been introduced under the name of `` entanglement witness '' @xcite . for bipartite systems of a large number of particles ,
this approach has been further investigated considering the possibility to exploit collective measurements on the multiparticle state . within this context , duan et al .
proposed a general criterion in ref.@xcite based on continuous variable @xcite observables .
this general criterion was subsequently applied to the quantum extension of the stokes parameters @xcite to obtain an entanglement bound for such kind of variables @xcite .
other approaches have been developed based on spin variables @xcite or pseudo - pauli operators @xcite .
an experimental application of this criteria based on collective spin measurements has been performed in a bipartite system of two gas samples @xcite .
however , an experimental realization of most of these criteria in the quantum optical domain requires photon - number resolving detectors with unitary efficiency , which is beyond the current technology .
a feasible approach for the analysis of multiphoton fields has been developed in the last few years , and is based on the deliberate attenuation of the analyzed system up to the single photon level . in this way , standard single - photon techniques and criteria can be used to investigate the properties of the field .
the verification of the entanglement in the high - loss regime is an evidence of the presence of entanglement before the attenuation , since no entanglement can be generated by local operations .
such approach has been exploited in @xcite to demonstrate the presence of entanglement in a high gain spontaneous parametric down - conversion source up to 12 photons .
an analogous conclusion has been theoretically obtained in ref.@xcite on the same system by exploiting symmetry considerations of the source .
the attenuation method has been also applied to a different system , making it possible to obtain an experimental proof of the presence of entanglement between a single photon state and a multiphoton state generated through the process of optical parametric amplification in an universal cloning configuration up to 12 photons @xcite . in this paper
we discuss recent experimental results of a micro - macro entanglement test @xcite , where the system under investigation is realized through the process of optical parametric amplification @xcite of an entangled photon pair .
the exploited entanglement criterion is an extension of the spin - based single - particle criterion of ref.@xcite .
such an extension requires a supplementary assumption which will be clarified in the remaining part of this paper . in sec.[sec : micro_macro ] we briefly review the properties of the micro - macro system realized in ref.@xcite .
then , in sec.[sec : entanglement_test ] we discuss in details the performed entanglement test . in particular , we focus on the conditions adopted in order to justify the exploited entanglement criterion . finally , in sec.[sec : general_criteria ] we perform a theoretical analysis of the micro - macro system based on the parametric amplification of an entangled pair .
several approaches for the verification of the entanglement property of the system will be addressed , showing that a substantial fraction @xmath0 of the original entanglement survives even in high losses condition .
let us first briefly review the micro - macro system of ref.@xcite .
the system under investigation is given by the following micro - macro source .
an entangled pair of two photons in the singlet state @xmath1=@xmath2 is produced through spontaneous parametric down - conversion ( spdc ) by crystal 1 ( c1 ) pumped by a pulsed uv pump beam : fig.[fig : experimental_setup ] .
there @xmath3 and @xmath4 stands , respectively , for a single photon with horizontal and vertical polarization @xmath5 while the labels @xmath6 refer to particles associated respectively with the spatial modes @xmath7and @xmath8 .
the photon belonging to @xmath8 , together with a strong uv pump beam , is injected into an optical parametric amplifier ( opa ) consisting of a non - linear crystal 2 ( c2 ) pumped by the beam @xmath9 .
the crystal 2 is oriented for collinear operation , i.e. , emitting pairs of amplified photons over the same spatial mode which supports two orthogonal @xmath10 modes , respectively horizontal and vertical .
then , fringe patterns are recorded by varying the analyzed polarization in the single - photon site on the equatorial plane of the bloch sphere , and keeping fixed the analyzed polarization in the @xmath8 mode .
the interaction hamiltonian of the opa device is given by @xmath11 .
in this collinear configuration , the _ quantum injected _ optical parametric amplifier ( qiopa ) acts as an _ optimal phase - covariant cloning machine _ @xcite , thus performing the optimal quantum cloning process for all single photon states belonging to the equatorial plane of the polarization bloch sphere , defined as @xmath12 . for all equatorial basis
, the interaction hamiltonian presents the same form : @xmath13 .
the overall micro - macro state after the amplification process reads : @xmath14 the output multiphoton states @xmath15 , where @xmath16 labels the injection of single - photon state with equatorial polarization , is given by the following expression : @xmath17 where @xmath18 , @xmath19 , @xmath20 , with @xmath21 nonlinear gain of the amplifier .
hereafter , the state @xmath22 labels a fock state with @xmath23 photons with @xmath24 polarization and @xmath25 photons with @xmath26 polarization . for a detailed discussion on the properties of such states we refer to refs.@xcite .
symmetry considerations based on the _ rotational invariance _ of the overall micro - macro _ singlet _ photon pair @xmath27 and of the _ phase - covariant _ and _ information preserving _ properties of the adopted qi - opa , lead to conclude that the two fringe pattern recorded in two different equatorial basis , @xmath28 ( @xmath29 ) and @xmath30 ( @xmath31 ) are identical , in the sense that the micro and macro states adopted in both cases , present the same fock - space expansion . in practice , the experimental visibilities of the fringe patterns in such two equatorial basis have been found equal by @xcite , within the statistical errors .
in this section we discuss a recent entanglement test performed in ref.@xcite .
the system under investigation is the micro - macro source discussed in the previous section .
we focus our analysis on the exploited entanglement criterion , obtained as the extension of a spin - based criterion for a bipartite microscopic - microscopic system @xcite .
first , the criterion and its experimental implementation is introduced in details , including an analysis of the regions of the system s hilbert space filtered by the detection strategy and a numerical analysis of the effects of a lossy process .
then , the assumptions on the source necessary for the validity of the test are discussed .
for a two - photon state @xmath32 , defined on two different modes @xmath33 and @xmath34 , the entanglement is demonstrated by applying the following criterion . for any separable state ,
the following inequality holds @xcite : @xmath35 where @xmath36 are the pauli operators and @xmath37 stands for the average on the state @xmath32 .
the same criterion can be extended to a micro - macro scenario , by measuring the pseudo spin operators @xmath38 on the macro state , obtained through an unitary transformation upon the micro - micro state . here , the @xmath38 operators are the time evolution of the pauli operators according to @xmath39 , where @xmath40 is the time evolution operator of the amplifier @xmath41 .
the following inequality holds : @xmath42 where @xmath43 refer to the polarization basis @xmath44 , @xmath45 , @xmath46 .
since the operators @xmath38 are built from the unitary evolution of eigenstates of @xmath47 , they satisfy the same commutation rules of the single - particle 1/2-spin : @xmath48 = 2 \imath \epsilon_{ijk } \hat{\sigma}_{k}$ ] , where @xmath49 is the levi - civita tensor density .
indeed we have for @xmath50 : @xmath51 where @xmath52 .
for @xmath53 we have : @xmath54 where @xmath55 and @xmath56 . in equation ( [ eq : micro_macro_criterion ] )
the state @xmath57 is obtained by the amplification of the state @xmath32 over the single spatial mode @xmath58 , and can be identified as two - qubit state of micro and macro systems . in the ideal case ,
the following map holds : @xmath59 where @xmath40 is the unitary amplification operator . _
( a ) implementation of the pseudo - pauli operators : the o - filter . _ since measurements of eqs.([eq : pauli_1]-[eq : pauli_23 ] ) , which require the perfect discrimination of the number of photons present in the detected state , are out of reach by current technology , we have adopted another strategy which is based on the o - filter ( of ) device , shown in fig.[fig : ofilter]-(a ) .
this method is based on a probabilistic discrimination of the macro - states @xmath60 and @xmath61 , which exploits the macroscopic features present in their photon - number distributions .
such measurement is implemented by an intensity measurement in the @xmath62 basis , followed by an electronic processing of the signal . if @xmath63 , the ( + 1 ) outcome is assigned to the event , if @xmath64 the ( -1 ) outcome is assigned to the event . if @xmath65 , an inconclusive outcome ( 0 ) is assigned to the event .
the action of the o - filter is described by the following measurement observables , applied on the multiphoton state _
after _ losses : @xmath66 the state after losses is no more a macro - qubit living in a two dimensional hilbert space , but in general it is represented by the density matrix @xmath67 .
such density matrix is obtained by applying to the macroqubit @xmath60 the map that describes the action a lossy channel with transmittivity @xmath68 : @xmath69 = \sum_{p } \gamma_{p } \hat{a}^{p } \hat{\rho } \hat{a}^{\dag \ , p } \gamma^{\dag}_{p}$ ] where @xmath70 @xcite . in order to describe the measurement results , all the average values of the measurement operators
must be calculated with the density matrix of the state after losses @xmath67 .
nevertheless the large difference in the photon number distribution along the distribution s tails present in the macro - qubits before losses is present also in the distribution of the macro - states after losses .
a detailed discussion on the properties of the macrostates after losses in both the fock - space and the phase - space is reported in refs .
@xcite . by exploiting this feature of our system ,
this probabilistic detection method allows us to infer the generation _ before _ losses of a @xmath60 or a @xmath71 state by exploiting the information encoded in the unbalancement of the number of photons present in the state _ after _
losses @xmath67 .
an analogous measurement scheme is shown in fig.[fig : ofilter]-(b ) .
the field is analyzed in polarization , and each branch is equally divided among a set of single - photon detectors ( apd ) .
coincidences between the output ttl signals are recorded for each analyzed polarization , and the ( + 1 ) or the ( -1 ) outcomes are assigned depending on which of the two analyzed sets of apds record the @xmath72-fold coincidence . if no @xmath72-fold coincidences are recorded , the ( 0 ) inconclusive outcome is assigned to the event .
this scheme performs the measurement of the @xmath72-th order correlation function of the field , where @xmath72 is the number of detectors .
we note that the o - filter based and the multi - detector based schemes select analogous regions of the fock space .
_ ( b ) filtering of the detected state . _
the entanglement test performed on our system in ref.@xcite is given by eq.([eq : micro_macro_criterion ] ) where the @xmath73 operators are replaced with the @xmath74 operators of the o - filter : @xmath75 it is worth noting that , in general , the resulting eq.([eq : micro_macro_criterion_of ] ) is no longer an entanglement witness . as discussed in the remaining of the paper , the bound of eq.([eq : micro_macro_criterion_of ] ) can be recovered as an entanglement witness by making a supplementary assumpion on the micro - macro source .
on one side , we note that the measurement of the correlations for the entanglement test of eq.([eq : micro_macro_criterion_of ] ) , are performed in the same basis for alice and bob s sites .
however , care should be taken when a filtering of the detected state is performed .
as shown in fig.[fig : ofilter]-(a ) , the o - filter detection scheme corresponds to a fock space filtering of the output state .
the measurements performed on different polarization basis select _ different _ regions of the fock space , corresponding to different portions of the density matrix .
this is shown in fig.[fig : filtering ] , where the photon number distribution of a @xmath76 fock state with @xmath77 in the @xmath28 and @xmath78 polarization bases is reported . when measured with the o - filter device , such state generates a conclusive ( + 1 ) outcome in the @xmath28 basis , since a strong unbalancement is present between the two polarizations . on the contrary , in the @xmath78 basis with high probability
the state generates an inconclusive outcome ( 0 ) and is filtered out .
this feature has no counterpart in the micro - qubit formalism : indeed the hilbert space of the original photon is only two - dimensional , so there is no risk of different subspaces being detected for different choices of measurement basis . indeed
the presence of losses enlarges the dimension of the hilbert space in which the macro - qubit lives , and the criterion of eq.([eq : micro_macro_criterion_of ] ) requires an auxiliary assumptions on the micro - macro state . with @xmath77 in the @xmath28 polarization basis .
( b ) photon number distribution for a fock state @xmath79 with @xmath77 in the @xmath78 polarization basis.,scaledwidth=50.0% ] _ ( c ) violation of the entanglement bound for a micro - macro separable state .
_ without any assumption on the investigated system the inequality ( [ eq : micro_macro_criterion_of ] ) , that is , the original pseudo - pauli criterion ( [ eq : micro_macro_criterion ] ) where the @xmath80 operators have been replaced by the @xmath81 ones , does not represent anymore a bound for entangled states .
it is satisfied by separable states of the form @xcite : @xmath82 where @xmath83 is a rotation of the whole system polarization around the @xmath84 axis by an angle @xmath85 .
+ despite the previous considerations , the of based strategy allows us to discriminate between different macro states in a probabilistic way .
.,scaledwidth=40.0% ] when @xmath86 the mean value of the @xmath87 operators , calculated over the real state @xmath88 , tends to the mean value of the pauli pseudo - spin operators , calculated over the ideal macro - qubit one @xmath89 . indeed , for asymptotically high values of the threshold @xmath90 , the measurement of the @xmath87 operators on the @xmath67 allows perfect , although probabilistic in the spirit of positive operator - valued measurements ( povm ) , discrimination of orthogonal states , as the pseudo - spin operator does for the macro - qubits @xmath91 .
in other words , if the @xmath92 state is generated , the measurement with the @xmath38 operator in the @xmath93 basis never leads to the ( -1 ) outcome . at the same time
, the measurement of the @xmath94 state after losses with the @xmath87 operator in the @xmath93 basis does not generate the ( -1 ) outcome if @xmath95 is large enough . according to these considerations
, we can infer the presence of the macro - qubit _ before _ losses and _ after _ the amplifier and then apply the original micro - macro inequality of eq.([eq : micro_macro_criterion ] ) .
this inference implies an assumption on the micro - macro system : the macro state has to be generated by an amplification process upon a micro - micro entangled pair .
the difference between the general case of a micro - macro entangled setup and the one here described is pointed out in fig.[fig : figure_assumption]-(a ) and fig.[fig : figure_assumption]-(b ) respectively .
therefore the entanglement test performed by the of scheme allows us to infer the presence of entanglement _ at least before _
losses , and to demonstrate the capability of amplifying an entangled pair in a coherent way . indeed the class of separable state in eq.([eq : separable ] ) can not be generated by a coherent amplification process , and the coherence of amplification is furthermore demonstrated by the presence of interference fringes in two different polarization bases . in order to conclude our analysis on the o - filtering measurement technique , we calculate theoretically how the visibility of the fringe pattern obtained in the micro - macro amplified scheme scales with the amount of losses when it is measured with this detection strategy .
the visibility is defined as : @xmath96 where @xmath97 and @xmath98 are the probability of obtaining respectively the ( + 1 ) and the ( -1 ) outcome , and are calculated as : @xmath99 \\
p(-1 ) & = & \mathrm{tr } \left [ \hat{\rho}_{\eta}^{\phi } \left(\sum_{m = k}^{\infty } \sum_{n=0}^{m - k } \vert n \vec{\pi}_{i } , m \vec{\pi}^{\bot}_{i } \rangle \langle n \vec{\pi}_{i } , m\vec{\pi}^{\bot}_{i } \vert \right ) \right]\end{aligned}\ ] ] we consider separately two different cases .
( 1 ) for a threshold @xmath100 , no filtering is performed on the analyzed state since the complete fock space is selected by the o - filter device . in this case , the visibility is a _ decreasing _ function of the losses parameter @xmath101 , where @xmath68 is the overall quantum efficiency of the channel . in fig.[fig : visibility_ofilter_a ] the trend of the visibility as a function of @xmath102 for @xmath100 and a gain value @xmath103 is reported .
we note that the visibility _ decreases _ with @xmath102 since a larger amount of losses is responsible for the cancellation of a larger amount of entanglement .
this point will be clarified later in this paper . for a threshold @xmath100 ; @xmath104.,scaledwidth=35.0% ] \(2 ) for a threshold @xmath105 ,
the o - filter device performs a filtering of the detected state as discussed in the previous paragraphs . in this case
, the visibility is an _ increasing _ function of the losses parameter @xmath101 .
in fig.[fig : visibility_ofilter_b ] the trend of the visibility as a function of @xmath102 for a gain value of @xmath103 , corresponding to an average number of generated photons @xmath104 , and several values of the threshold @xmath95 .
the visibility _ increases _ with the losses parameter since , for the same value of the threshold @xmath95 , a tighter filtering of the detected wave function is performed for higher @xmath102 due to the reduced average number of photons present in the state . + for @xmath106 ( solid line ) , @xmath107 ( dashed line ) and @xmath108 ( short - dashed line ) .
all lines correspond to @xmath104.,scaledwidth=35.0% ]
in this section we analyze different approaches for an entanglement test in a microscopic - macroscopic system , and we discuss the application to the system of ref.@xcite obtained by optical parametric amplification of an entangled photon pair . as a first step , we develop a generalized entanglement witness criterion based on dichotomic measurements .
we then apply this criterion for the specific case of the pauli pseudo - spin operators previously introduced , showing the fragility under losses of this detection strategy . as a second step ,
we consider a different approach based on the deliberate attenuation of the multiphoton field . such technique allows us to theoretically demonstrate the presence of entanglement in the investigated micro - macro system for any value of the amplifier gain and of the losses @xmath68 .
the same conclusion can be drawn with a different approach based on the quantum stokes operators already developed in @xcite and applied to our micro - macro system in refs.@xcite .
for a set @xmath109 of dichotomic operators , without making any supplementary assumption , the bound to be violated in order to demonstrate the entanglement of the overall micro - macro system must be modified with respect to eq.(2 ) , and a necessary condition for separable states is given by the following inequality : @xmath110 details over the derivation of this criterion are reported in app.[app : general_criterion ] .
such criterion presents the interesting feature of not requiring any knowledge of the hilbert space where the analyzed states live . indeed , in the derivation of the bound ( [ eq : general_criterion ] ) the only necessary assumption concerns the measurement operators , which can have only two possible outcomes @xmath111 . for the specific choice of the pauli pseudo - spin operators @xmath112 as measurement operators @xmath113 as a function of the detection losses @xmath68 , calculated for several values of the gain of the amplifier .
the upper horizontal solid line corresponds to the bound for separables states of the general criterion ( [ eq : general_criterion ] ) , while the lower horizontal solid line corresponds to the bound for separable states ( [ eq : micro_macro_criterion ] ) where the commutation properties of the operators have been exploited and a standard assumption on the hilbert space is necessary
. dashed lines : from top to bottom , curves correspond respectively to a nonlinear gain @xmath114 , @xmath115 , @xmath116 , @xmath117 , @xmath118 and @xmath119,scaledwidth=40.0% ] we then applied the obtained criterion to evaluate the quantity @xmath120 for the micro - macro state generated through the process of optical parametric amplification , for the specific choice of the pauli pseudo - spin operators @xmath121 ( [ eq : pauli_1]-[eq : pauli_23 ] ) as the measurement operators .
more specifically , we evaluated the value of @xmath120 as a function of the transmission efficiency @xmath68 of the multiphoton mode @xmath8 for several values of the gain @xmath122 [ fig.[fig : witness_pauli_micro_macro_p=1 ] ] .
the value of @xmath120 is then compared to the bound for separable states @xmath123 .
we observe that this entanglement measurement is fragile under losses , since the value of @xmath120 falls below the bound for separable states when the number of lost photons is @xmath124 .
such result is expected since the pauli operators allows to distinguish the @xmath60 states exploiting the well - defined parity in the number of photon generated by the amplifier depending on the polarization of the input states . in presence of losses ,
such well - defined parity is quickly cancelled , thus not allowing to discriminate among the macro - states with this kind of measurement .
this feature of the macro - states generated through the process of optical parametric amplification is reported and discussed in refs.@xcite .
an alternative approach can be used to demonstrate the presence of entanglement in our micro - macro configuration .
the macroscopic field is deliberately attenuated up to the single - photon regime and detected through an apd .
such method has been exploited to demonstrate the entanglement up to 12 photons in a spontaneous parametric down conversion source @xcite , or in a micro - macro configuration @xcite .
the average number of photons impinging onto the detector in this regime is then @xmath125 , where @xmath68 is the overall quantum efficiency of the channel . in this condition
, the probability of detecting more then one photon becomes negligible .
the density matrix of the macroscopic state can be reduced to a 1-photon subspace , and the joint micro - macro system is defined in a @xmath126 polarization hilbert space spanned by the basis vectors @xmath127 .
the complete state @xmath128 can be then evaluated by applying the map describing a lossy channel @xcite to the micro - macro amplified state @xmath129 $ ] .
we obtain the following expression : @xmath130 where : @xmath131 in fig.[fig : concurrence_p1]-(a ) we show the density matrix of the joint micro - macro system for a value of @xmath132 and @xmath133 , showing the presence of the off - diagonal terms even in the high losses regime .
this system is entangled for any value of the nonlinear gain @xmath122 .
this property can be tested by application of the peres criterion or by direct calculation of the concurrence , which reads : @xmath134 this quantity is always positive , as plotted in fig.[fig : concurrence_p1]-(b ) , showing the presence of entanglement for any value of the gain . since no entanglement can be generated with local operations ( such as a lossy process ) @xcite
, the presence of entanglement in the highly attenuated regime is due to the presence of entanglement in the micro - macro system before losses . + and a value of the losses parameter @xmath133 . ( b ) plot of the concurrence @xmath135 as a function of the parameter @xmath136
we note the persistance of the off - diagonal terms and entanglement for all values of @xmath122 and @xmath68.,scaledwidth=50.0% ] this criterion allows us to discuss an important feature of the micro - macro system based on optical parametric amplification .
the entanglement of this system is generated in the micro - micro source , where the singlet polarization state @xmath27 is produced .
the action of the amplifier is to broadcast the properties of the injected seed to the multiparticle state .
in particular , the entanglement present in the original photon pair after the amplification process is transfered and shared among the generated particles ( see fig.[fig : entanglement ] ) .
if a certain amount of losses is introduced in the macro - state and @xmath0 is the percentage of photons that survive such decoherence process , the amount of entanglement detected after losses is reduced of a factor @xmath0 but drops to 0 only if all particles are lost .
analytically , this feature is obtained by analyzing the expression ( [ eq : conc ] ) for @xmath135 . in the high gain limit ( @xmath137 ) ,
the concurrence of our system in the highly attenuated regime becomes : @xmath138 being directly proportional to @xmath68 , that is , the fraction of detected photons .
+ to conclude these considerations , we extend the analysis of the micro - macro amplified system in this highly attenuated scenario to the case where the injection of the single - photon in the optical parametric amplifier occur with a non unitary efficiency @xmath139 . such parameter represents the amount of matching ( spectral , spatial , and temporal ) between the optical mode of the amplifier and the optical mode of the injected single - photon . to model this source of experimental imperfection , the joint state between the two modes @xmath7 and @xmath8 before amplification is described by @xmath140 , where @xmath141 stands for a completely mixed polarization state and @xmath142 represents the vacuum input state . by following the same procedure described for the @xmath143 case , the density matrix of the joint micro - macro system after amplification and losses in the highly attenuated regime reads : @xmath144 where @xmath145 is the opportune normalization constant . in fig.[fig : concurrence_pmin1 ] ( a)-(b ) we show the density matrix for a gain value @xmath146 , for @xmath147 and injection probabilities of @xmath148 and @xmath149 .
the effect of a decreasing injection probability @xmath150 is the reduction of the off - diagonal terms and hence of the coherence terms .
the application of the peres criterion on this density matrix gives a critical value of the injection probability @xmath151 . for @xmath152 , the micro - macro system in this highly attenuated regime
is entangled , while for @xmath153 the system is separable .
the same result is confirmed by the calculation of the concurrence , which reads : @xmath154 in fig.[fig : concurrence_pmin1 ] ( c ) we report the plot of the concurrence as a function of the gain @xmath122 for several values of the injection probability @xmath150 and @xmath147 . for decreasing @xmath150 , the concurrence drops to @xmath155 for a lower value of the gain .
furthermore , in fig.[fig : concurrence_pmin1 ] ( d ) we report the plot of the critical injection probability @xmath156 as a function of the gain @xmath122 and the transmission efficiency @xmath68 .
as the gain @xmath122 is increased , the value of the critical injection probability increases up to a value close to @xmath157 .
this means that , for high values of the gain , an high injection efficiency is requested to detect the entanglement with such measurement strategy .
.,scaledwidth=50.0% ] and a value of the losses parameter @xmath133 .
( a ) injection probability @xmath148 and ( b ) injection probability of @xmath158 .
( c ) concurrence @xmath159 as a function of the gain @xmath122 for @xmath147 .
red solid line corresponds to an injection probability @xmath143 , green long dashed line to @xmath148 , blue short dashed line to @xmath149 and black dotted line to @xmath160 .
( d ) 3-dimensional plot for the critical injection probability @xmath156 as a function of the gain @xmath122 and the transmission coefficient @xmath68.,scaledwidth=50.0% ] another approach for a micro - macro entanglement test is based on the detection of the quantum stokes operators , defined as : @xmath161 .
for a micro - macro system , the following inequality , found by simon et al . in ref.@xcite , holds for any separable state : @xmath162 where @xmath163 is the photon number operator . for the micro - macro configuration under investigation ,
the following result @xcite holds : @xmath164 thus violating the bound for separable states .
again , some entanglement survives for any value of the gain and of the losses parameter @xmath68 , and the amount of entanglement is proportional to the number of detected photons .
however such criterion is not feasible from an experimental point of view since the measurement of the stokes operators requires perfect discrimination in the photon - number , as in the pseudo - spin operator case .
in this paper we analyzed several classes of entanglement criteria for bipartite systems of a large number of particles .
in particular , we addressed a specific joint microscopic , i.e. composed by a single particle , and macroscopic system based on optical parametric amplification of an entangled photon pair .
a first experimental entanglement test on this system has been recently reported in ref.@xcite .
we analyzed in details the conclusions that can be drawn on this experiment .
the adopted entanglement criterion in that paper allowed to infer the presence of entanglement after the amplification process before losses in the detection apparatus .
the validity of the test however requires a specific assumption on the system that generates the micro - macro pair .
an a priori knowledge of the source is necessary in order to exclude a class of separable states that can reproduce the obtained experimental results .
one of the reason for the necessity of this assumption is given by the exploited detection strategy , which presents the feature of a povm with an inconclusive outcome which depends on the measurement basis .
such problem has been recently investigated in ref.@xcite within the context of a nonlocality test in macroscopic systems . in that paper
it was shown that a bell test based on a generalized dichotomic measurement , i.e. with an inconclusive outcome , allows to exclude only a class of local hidden variables ( lhv ) models .
a more general approach to the micro - macro entanglement problem in the investigated system is addressed in the rest of the paper .
we discussed different entanglement criteria which do not require any supplementary assumption on the source , and applied these approaches to the micro - macro system based on optical parametric amplification .
we first derived a general bound for an entanglement criterion based on dichotomic operators .
then , an approach based on deliberate attenuation of the multiphoton field to the single - photon regime , already introduced in ref.@xcite , has been applied to our system .
this analysis allowed us to show that a fraction @xmath0 of the original entanglement of the entangled photon pair exists even in presence of losses , where @xmath0 is proportional to the number of particles that survive the lossy process .
as a further perspective , the system based on parametric amplification can lead to the investigation of entanglement in a bipartite macroscopic macroscopic system @xcite .
we acknowledge support by the `` futuro in ricerca '' project hyteq , and progetto dateneo of sapienza universit di roma .
in this appendix we demonstrate the inequality of eq.([eq : general_criterion ] ) , which gives a generalized bound for an entanglement test in a micro - macro bipartite system and dichotomic measurements .
the proof is divided into two parts .
first , a general treatement of dichotomic measurements is applied to the derivation of an entanglement criterion with no auxiliary assumption on the system under investigation .
then , the obtained results are applied to a micro - macro scenario .
the density matrix of a separable state , composed by two subsystems @xmath165 and @xmath166 , can be written as : @xmath167 we restrict our attention to the set of dichotomic measurements , i.e. @xmath111 valued upon each subsystem @xmath168 and @xmath169 respectively .
the average value of a generic measurement operator @xmath170 is given by @xmath171 , where the superscript @xmath172 refers to a specific choice of the operator @xmath173 .
the average value of the @xmath174 component of the decomposition of the density matrix reads : @xmath175 the average value @xmath176 can then be reexpressed as : @xmath177 the following inequality holds : @xmath178 since @xmath179 .
the sum of the average value over three different operators @xmath173 , where @xmath180 , is given by the following expression : @xmath181 by definition ( [ eq : visibility_i ] ) , we obtain for the @xmath174 component of the decomposition for a separable state : @xmath182 since @xmath183 the following inequality holds : @xmath184 hence for a generic separable state the following inequality holds : @xmath185 where the @xmath186 term is evaluated over the density matrix @xmath187 for subsystem @xmath165 .
the latter can be always decomposed as : @xmath188 where the set @xmath189 of probabilities satisfied the normalization condition @xmath190 .
we can then derive the following inequality : @xmath191 substituting this result in eq.([eq : separable_inequality_intermediate ] ) , by exploiting the normalization conditions for the coefficients @xmath192 and @xmath193 we obtain , due to convexity the following inequality for all bipartite separable states : @xmath194 where the maximization is performed over all possible states of system @xmath165 .
we now specialize the result of previous section in the microscopic - macroscopic states , i.e. when system @xmath165 is a single spin-@xmath195 particle .
let us make a specific choice for the measurement operators @xmath196 .
for a single spin-@xmath195 particle , we choose the pauli operators @xmath197 .
hereafter , we remove the subscript @xmath165 in all the equations for simplicity of notation .
the entanglement criterium ( [ eq : entanglement_criterium_separable ] ) for this choice of the system and operators then reads : @xmath198 we now need to maximize the righthand side of the latter equation over all possible choices of single particle states @xmath199 . by applying the lagrange multiplier method
, the upper bound for the average of the pauli operators reads : @xmath200 finally , we can write the following inequality for the joint microscopic - macroscopic system : @xmath201 | we discuss the conclusions that can be drawn on a recent experimental micro - macro entanglement test [ f. de martini , f. sciarrino , and c. vitelli , phys .
rev .
lett . * 100 * , 253601 ( 2008 ) ] .
the system under investigation is generated through optical parametric amplification of one photon belonging to an entangled pair .
the adopted entanglement criterion makes it possible to infer the presence of entanglement before losses , that occur on the macrostate , under a specific assumption .
in particular , an _ a priori _ knowledge of the system that generates the micro - macro pair is necessary to exclude a class of separable states that can reproduce the obtained experimental results .
finally , we discuss the feasibility of a micro - macro `` genuine '' entanglement test on the analyzed system by considering different strategies , which show that in principle a fraction @xmath0 , proportional to the number of photons that survive the lossy process , of the original entanglement persists in any losses regime . |
the gaas quantum dots are grown by molecular beam epitaxy .
a low density ( @xmath2 @xmath3 ) of self - assembled nanoholes approximately 10 nm deep and 100 nm wide were created by arsenic debt epitaxy in a gaas ( 100 ) substrate [ s1 ] .
deposition of 21 repeats of 0.4 monolayers ga / 0.15 monolayers gaas lead to droplets of excess gallium on the surface . a 5 minute interrupt under excess arsenic at a substrate temperature of 520@xmath4c results in the formation of nanoholes surrounded by a raised gaas ring due to the transfer of arsenic from the substrate to the gallium droplets [ s2 ] .
a 10 nm bottom al@xmath5ga@xmath6as barrier was then grown over the nanoholes , followed by a 2 nm gaas layer , an 8 nm top al@xmath5ga@xmath6as barrier and a final 2 nm gaas capping layer .
the substrate temperature was then reduced from 520@xmath4c to approximately 200@xmath4c , and one monolayer of gallium was deposited to ensure a gallium terminated surface followed by @xmath7 @xmath3 silicon to passivate the surface states [ s3 ] . during the growth of the algaas layers the shape of the nanoholes is well preserved . however , due to the greater migration length of gallium compared to aluminium , there is accumulation of gallium at the bottom of the nanohole during deposition of the gaas layer leading to gaas quantum dot formation [ s4 ] . at the end of the growth
, there is still a 2 nm deep depression in the sample surface marking the position of the quantum dot in the self - assembled nanohole .
the elevated feature corresponds to the gaas ring while the depression is at the position of the quantum dot .
dark - field scattering spectra of the gold nanoparticles were obtained using a home - built dark - field microspectroscopy setup .
light from a halogen lamp was used to illuminate the sample through a dark - field objective ( olympus mplanfl n 100@xmath8/0.90 bd ) .
the sample was positioned using a piezoelectric positioning system .
the light back - scattered by a nanoparticle was collected with the microscope objective and directed to a 500 mm focal length grating spectrometer equipped with a liquid nitrogen cooled charge coupled device camera .
a reference spectrum was collected from an area on the sample without nanoparticles and subtracted from the scattering spectrum of the particle . finally , to compensate for the wavelength dependent transmission of the optical train , the emission spectrum of the lamp , and the wavelength dependent sensitivity of the camera , a bright - field reflection spectrum of the sample from an area without nanoparticles was measured .
this spectrum was used to normalize the dark - field spectra to obtain the scattering spectra of the nanoparticles . | we demonstrate how the controlled positioning of a plasmonic nanoparticle modifies the photoluminescence of a single epitaxial gaas quantum dot .
the antenna particle leads to an increase of the luminescence intensity by about a factor of eight .
spectrally and temporally resolved photoluminescence measurements prove an increase of the quantum dot s excitation rate .
the combination of stable epitaxial quantum emitters and plasmonic nanostructures promises to be highly beneficial for nanoscience and quantum optics .
coupling optical quantum emitters to plasmonic nanostructures is a particularly topical area of quantum optics of single nanoobjects as well as plasmonics @xcite .
the localized nature of the electromagnetic field associated with particle plasmons is equivalent to a small effective mode volume of the optical field , leading to an increase in the light - matter coupling constant .
many fascinating experiments coupling emitters and plasmons have been proposed or performed , such as enhanced spontanous emission into free space @xcite , spatially directed emission by an optical nanoantenna @xcite , or the use of the emitter s nonlinearity as an optical transistor @xcite .
all these experiments require a quantum emitter that is stable in position and emission rate , and well defined in emission frequency and orientation of its dipole axis relative to the plasmonic structure .
epitaxially grown self - assembled semiconductor quantum dots have proven to be bright and non - blinking single photon sources @xcite , fulfilling all the above requirements .
however , because of the high refractive index of the semiconductor substrate ( @xmath0 for gaas ) , the coupling of the quantum dot to the free - space optical field is weak .
a plasmonic nanoantenna close to a quantum dot promises relief of this restriction . in this letter
we demonstrate to our knowledge for the first time the controlled coupling of a _ single _ epitaxial quantum dot to a plasmonic nanoantenna .
excellent control over the coupling is a prerequisite for realizing more complex plasmonic devices making use of several single emitters , where one can not rely anymore on random placement of emitters relative to the plasmonic structures . to achieve this degree of control ,
emitters can be chemically conjugated to the plasmon resonant structure @xcite .
alternatively atomic force microscopy ( afm ) based nanomanipulation can be used to approach emitter and metal nanostructure @xcite .
however , especially the orientation of the emitter s dipole axis remains difficult to control .
the epitaxial quantum dots used in this work have two energetically almost degenerate transitions with well defined , orthogonal dipole moments which are parallel to the sample surface and are oriented along specific crystal directions @xcite .
placing of the plasmonic structure in a well - defined distance to the quantum dot is straight forward as the dot s position is visible by afm and electron microscopy due to a characteristic structure on the sample surface .
our exclusively solid - state approach is very well suited for developing nanometer - scale optical circuits for future photonic networks .
the quantum dots in our sample are lens - shaped gaas inclusions of about 10 nm diameter and 5 nm height embedded in a algaas barrier of larger band - gap .
the two optically active dipole transitions of the neutral excitons show splittings of less than 90 @xmath1ev which can be ascribed to slight dot - shape anisotropies @xcite . with a ground state exciton emission wavelength around 760 nm ,
these quantum dots are well suited for si - based single photon detectors .
the quantum dots were grown using molecular beam epitaxy . in a first step
a template of nanoholes was created by arsenic debt epitaxy on a gaas ( 100 ) substrate @xcite . during
further growth steps the sample topography is preserved and the holes are visible by atomic force microscopy on the sample surface ( [ fig : nanoman]a ) .
the quantum dots are formed by first depositing the algaas barrier material followed by a layer of gaas . during this growth step , gaas accumulates in the holes and forms the quantum dots , while outside the holes only a thin gaas wetting layer remains @xcite .
a detailed description of sample growth is given in the supplementary material .
the quantum dots are located 10 nm beneath the sample surface .
the shallow depth of the emitters enables near - field coupling with plasmonic structures placed on the sample surface .
here we use gold nanoparticles positioned above the quantum dots to act as optical antennas .
a schematic of the studied structure is shown in [ fig : nanoman]b .
the quantum dots can be excited resonantly , i.e. , by absorption of a photon by the quantum dot , or non - resonantly by carriers that are created outside the dot and diffuse into it . here
we excite the quantum dot non - resonantly by creating electrons and holes in the barrier and wetting layer using an optical parametric oscillator .
the oscillator produces 400 fs pulses with a photon energy that can be tuned between 1.8 and 2.4 ev .
the photoluminescence is collected in a home - built low - temperature confocal laser scanning microscope .
the spatial resolution is limited by the objective s numerical aperture ( na = 0.7 ) to approximately 600 nm . the photoluminescence can be spectrally analyzed using a spectrometer equipped with a charge coupled device camera .
the spectral resolution of the setup is around 100 @xmath1ev . for time - resolved measurements we employ
time - correlated single photon counting with a temporal resolution of approximately 30 ps , which is much shorter than the measured exciton decay time of about 350 ps .
all the optical measurements were performed at a temperature of 10 k. as optical antennas we use spherical gold nanoparticles with a diameter of 90 nm as specified by the manufacturer ( nanopartz inc . ) .
the nanoparticles were deposited on the sample surface by drop casting from an aqueous solution .
the sample had been patterned with markers so that the same area can be studied with afm and optical microscopy .
an afm was used to identify regions of the sample with a low concentration of gold particles .
ten of the nanoparticles were moved on top of nearby quantum dots using the afm tip .
the optical properties of nanoantennas were characterized by recording the scattering spectrum of single antennas using dark - field microspectroscopy @xcite .
details of the dark - field spectroscopy are given in the supplementary material .
[ fig : nanoman]c shows a raster - scanned photoluminescence image of the region of the sample containing antenna - coupled emitters . for each bright , diffraction - limited spot in the luminescence image
one can find a corresponding surface depression in an afm image of the same area of the sample .
the majority of the quantum dots are comparable in brightness .
the emitters decorated with a gold nanoparticle , however , appear markedly brighter as seen in [ fig : nanoman]c .
to exclude a coincidence , we removed the particle from three quantum dots ( marked by red circles in [ fig : nanoman]d ) after characterizing their optical properties .
the photoluminescence of these three quantum dots changed to approximately the ensemble average value . to quantify the influence of the nanoparticle on the emission properties of the quantum dots
, we analyzed the photoluminescence of 228 emitters by fitting a two - dimensional gaussian function to each emission spot in the luminescence image .
the thus obtained brightness distribution is shown in [ fig : histo ] .
the center of the distribution was determined as 194 counts/30 ms ( 6470 counts / s ) by fitting a gaussian function to the histogram .
the brightness of the 10 emitters coupled to a nanoantenna was determined in a similar fashion .
we observe an increase in the photoluminescence intensity of up to a factor of 6 compared to the ensemble average .
all our antenna - coupled quantum dots have a brightness that is greater than the ensemble average value .
one advantage of our approach is the possibility to compare the properties of an emitter - antenna complex directly with those of the same single emitter without the antenna .
here we did this by removing the gold nanoparticle from the quantum dot for the three dots encircled in red in [ fig : nanoman]d .
the change in the photoluminescence intensity is shown in [ fig : histo]b .
both the reference quantum dots and the unmodified emitter - antenna complexes remain at a similar brightness , as expected .
the brightness of the three emitters from which the antennas were removed decreased to a value well within the ensemble distribution .
for these quantum dots we deduce photoluminescence enhancement factors of 3.5 , 3.7 , and 6.1 when compared to their emission without antenna .
the spread in the enhancement can be explained by a variation of the relative position of the antennas with respect to the quantum dots . to determine the origin of the enhancement in photoluminescence intensity let us compare the emission spectra of the quantum dots with and without antenna as shown in [ fig : emissiondecay]a .
we remark that the photoluminescence spectrum displayed in [ fig : emissiondecay]a for the quantum dot coupled to the antenna was shifted in energy by 1.4 mev to compensate for spectral shifts most likely due to surface charges in the vicinity of the quantum dot . at 10 nw excitation power
the emission spectrum of the quantum dot without antenna is dominated by the exciton transition ( [ fig : emissiondecay]a ) . at 50
nw excitation power the exciton line starts to saturate and a second emission line appears .
compatible with previous reports on gaas / algaas dots @xcite , we identify this as one of the trions , i.e. , charged exciton , transition from its super - linear excitation power dependence and its lifetime .
the antenna - enhanced quantum dots show already at 10 nw excitation power an emission spectrum that resembles that of the unenhanced emitter at higher excitation power ( [ fig : emissiondecay]a ) . from this
we conclude that the emitter in the quantum dot - nanoparticle complex experiences an about 5 times higher excitation power density due to the gold particle .
the field enhancement around the plasmonic nanoparticle locally increases the probability that an electron - hole pair is created .
the photoluminescence enhancement as determined from luminescence images is 6.1 for this emitter .
this is in good agreement with the ratio of the excitation powers for which the emission spectra of the quantum dot with and without antenna is similar .
also for the other emitters the agreement between the enhancement values determined with the two methods was good .
we can thus attribute the brightness enhancement predominantly to an increase of the excitation rate .
this is also supported by measurements of spectrally integrated luminescence decay . in [ fig : emissiondecay]b we show the luminescence decay traces corresponding to the spectra displayed in [ fig : emissiondecay]a for 10 nw excitation power and for 50nw with and without antenna , respectively .
we observe no significant changes in the decay dynamics due to the antenna .
we note that the emission lines of the antenna - quantum dot complexes appear slightly narrower than the emission of the same dots without antenna at comparable emission intensity . in the usual far - field
excitation scheme carriers are generated in a larger volume than in the presence of the antenna .
this leads to fluctuating charges , broadening the dot s emission by spectral diffusion . to further elucidate the effect of the plasmonic antenna on the optical properties of the quantum dots in our structure we have studied the influence of the excitation wavelength on the photoluminescence . for non - resonant excitation ,
the luminescence rate of the emitters depends on the carrier density in their vicinity and their diffusion into the quantum dot . assuming a dipole transition , for low excitation power the carrier density is proportional to the local intensity of the electromagnetic field .
the excitation rate of the quantum dots will thus be maximized when the excitation photon energy is tuned to the antenna resonance . for the spread at 2.06 ev ) . ]
[ fig : spectral_enh ] shows the photoluminescence enhancement of the quantum dots coupled to a nanoparticle antenna as a function of excitation energy for an average excitation power of about 50 nw . the enhancement was determined by comparing the brightness of the antenna - enhanced emitters to an ensemble of reference quantum dots that were not coupled to an antenna .
we observe a peak in the measured photoluminescence enhancement at approximately 2.0 ev energy . at the plasmon resonance the electromagnetic field
is enhanced close to the particle and the excitation rate of electron - hole pairs in the vicinity of the quantum dot is thus increased .
the enhancement maximum occurs at a slightly lower energy than the plasmon resonance as seen from the scattering spectrum shown in [ fig : spectral_enh ] .
this is in agreement with the report of bryant _ et al . _
@xcite that the near - field resonance occurs red - shifted of the far - field scattering resonance .
the emission of the quantum dots is not modified ( [ fig : emissiondecay]b ) as it occurs far off the antenna s plasmon resonance . in summary ,
we have introduced single epitaxial gaas quantum dots as emitters for plasmonics by demonstrating the controlled enhancement of the excitation efficiency . placing a gold nanoparticle in the vicinity of
the emitter resulted in an increase in the photoluminescence intensity of about a factor of eight . in the near - field of the antenna close to the quantum dot more carriers
are generated so that the excitation rate is increased .
our conclusion is supported by the observation that the decay dynamics of the emitters is unaltered by the coupling to the antenna .
varying the wavelength of the incident light clearly demonstrates that the photoluminescence enhancement originates from the plasmon resonance of the antenna .
our scheme can be applied to more complex antennas and more than one emitter .
for example , rod - like plasmon resonant metal particles can be positioned in the same way as here using afm .
moreover , the topography of the quantum dots can be used to first determine their position and subsequently fabricate antenna structures by electron - beam lithography on top of them .
the gaas quantum dots provide a stable , solid - state source of photons that , when coupled to plasmonic nanostructures , forms an attractive system in which to study light - matter interactions and to use as a basis for nanophotonic applications .
we thank a. vlandas for assistance with atomic force microscopy , l. wang , f. ding and b. eichler for the growth and characterization of test samples , and the deutsche forschungsgemeinschaft dfg for funding ( research unit for 730 ) .
99 chang , d. e. ; srensen , a. s. ; hemmer , p. r. ; lukin , m. d. _ phys .
rev .
lett . _ * 2006 * , _ 97 _ , 053002 .
akimov , a. v. ; mukherjee , a. ; yu , c. l. ; chang , d. e. ; zibrov , a. s. ; hemmer , p. r. ; park , h. ; lukin , m. d. _ nature ( london ) _ * 2007 * , _ 450 _ , 402 .
fedutik , y. ; temnov , v. v. ; schps , o. ; woggon , u. ; artemyev , m. v. _ phys .
rev .
lett . _ * 2007 * , _ 99 _ , 136802 .
shimizu , k. t. ; woo , w. k. ; fisher , b. r. ; eisler , h. j. ; bawendi , m. g. _ phys .
rev .
lett . _ * 2002 * , _ 89 _ , 117401 .
liu , y. ; blair , s. _ opt .
lett . _ * 2003 * , _ 28 _ , 507 .
farahani , j. n. ; pohl , d. w. ; eisler , h .- j . ; hecht , b. _ phys .
rev .
lett . _ * 2005 * , _ 95 _ , 017402 .
mertens , h. ; polman , a._appl .
phys .
lett .
_ * 2006 * , _ 89 _ , 211107 .
tam , f. ; goodrich , g. p. ; johnson , b. r. ; halas , n. j. _ nano lett .
_ * 2007 * , _ 7 _ , 496 .
ringler , m. ; schwemer , a. ; wunderlich , m. ; nichtl , a. ; krzinger , k. ; klar , t. a. ; feldmann , j. _ phys .
rev .
lett . _ * 2008 * , _ 100 _ , 203002 .
bek , a. ; jansen , r. ; ringler , m. ; mayilo , s. ; klar , t. a. ; feldmann , j. _ nano lett . _ * 2008 * , _ 8 _ , 485 .
kinkhabwala , a. ; yu , z. ; fan , s. ; avlasevich , y. ; mllen , k. ; moerner , w. e. _ nature photonics _ * 2009 * , _ 3 _ , 654 .
schietinger , s. ; barth , m. ; aichele , t. ; benson , o. _ nano lett . _ * 2009 * , _ 9 _ , 1694 .
taminiau , t. h. ; stefani , f. d. ; segering , f. b. ; van hulst , n. f. _ nature photonics _ * 2008 * , _ 2 _ , 234 .
chang , d. e. ; srensen , a. s. ; demler , e. a. ; lukin , m. d. _ nature phys .
_ * 2007 * , _ 3 _ , 807 .
michler , p. ; kiraz , a. ; schoenfeld , w. v. ; petroff , p. m. ; zhang , l. ; hu , e. ; imamolu , a. _ science _ * 2000 * , _ 290 _ , 2282 .
plumhof , j. d. ; kpek , v. ; wang , l. ; schliwa , a. ; bimberg , d. ; rastelli , a. ; schmidt , o. g. _ phys . rev .
b _ * 2010 * , _ 81 _ , 121309 .
alonso - gonzlez , p. ; martn - snchez , j. ; gonzlez , y. ; aln , b. ; fuster , d. ; gonzlez , l. _ crystal growth and design _ * 2009 * , _ 9 _ , 2525 .
wang , l. ; kpek , v. ; ding , f. ; horton , f. ; schliwa , a. ; bimberg , d. ; rastelli , a. ; schmidt , o. g. _ phys .
rev .
b _ * 2009 * , _ 80 _ , 085309 .
schultz , s. ; smith , d. r. ; mock , j. j. ; schultz , d. a. _ proc .
natl .
acad .
u.s.a . _ * 2000 * , _ 97 _ , 996 .
snnichsen , c. ; franzl , t. ; wilk , t. ; von plessen , g. ; feldmann , j. ; wilson , o. ; mulvaney , p. _ phys .
rev .
lett . _ * 2002 * , _ 88 _ , 077402 .
bryant , g. w. ; de abajo , f. g. ; aizpurua , j. _ nano lett .
_ * 2008 * , _ 8 _ , 631 . |
the large hadron collider ( lhc ) has so far not seen any signal of new physics beyond the standard model ( bsm ) . on the other hand atlas and cms
have recently observed a new state consistent with a standard - model - like higgs boson at @xmath4 gev @xcite , which is within the range for it to be consistent with the lightest higgs in supersymmetric models . in the minimal supersymmetric standard model ( mssm ) the light higgs mass at tree - level is bounded from above by the @xmath5 boson mass ( @xmath6 ) .
the large radiative contributions from stops needed to raise it to the observed value typically imply very large fine - tuning .
conventional @xmath1 inspired susy models involve both a singlet generated @xmath7 term , denoted @xmath8 , and a massive @xmath0 gauge boson at the tev scale .
such models can increase the tree level physical higgs boson mass above the @xmath6 limit of the mssm , due to both f - term contributions of the singlet and the d - term contributions associated with the @xmath0 , allowing lighter stop masses and hence reducing fine - tuning due to stop loops . the exceptional supersymmetric standard model ( e@xmath9ssm ) @xcite is an example of such a model , inspired by the @xmath1 group .
it involves an extra singlet responsible for @xmath8 and an extra @xmath10 gauge symmetry at low energy , giving both new f - term and d - term contributions at tree level to the light higgs mass , which is larger than both the mssm and the next - to - minimal supersymmetric standard model ( nmssm)@xcite . in the e@xmath9ssm the light higgs mass is given by , @xmath11 where @xmath12 is the ratio between the two higgs doublets vacuum expectation values ( vevs ) , @xmath13 is the yukawa coupling of the singlet field to the higgs doublets , the extra @xmath3 gauge group has a gauge coupling @xmath14 and @xmath15 represents loop corrections .
eq.[mh ] exhibits two extra terms proportional to @xmath16 , relative to the mssm , which contribute at tree level to the higgs mass squared .
this means that the e@xmath9ssm permits lower stop masses than in the mssm ( or the nmssm ) corresponding to lower values required for the radiative correction term @xmath15 .
however , as we shall discuss , one of the minimisation conditions of the e@xmath9ssm can be written in the form , [ mz0 ] c = -_^2 + + d , where @xmath17 are functions of @xmath18 which are of order @xmath19 , @xmath20 are soft higgs mass squared parameters , @xmath21 and @xmath22 arise from the singlet vev @xmath23 . written in this form it is clear that there is a new source of tree - level fine - tuning , due to the @xmath0 mass squared term in eq.[mz0 ] , which will increases quadratically as @xmath24 , eventually coming to dominate the fine - tuning for large enough values of @xmath25 .
this tree - level fine - tuning can be compared to that due to @xmath26 which typically requires this quantity to be not much more than 200 gev , and similar limits also apply to @xmath25 . with the current cms experimental mass limit for the @xmath0 in the e@xmath9ssm of @xmath27 tev
@xcite it is clear that there is already a significant , perhaps dominant , amount of fine - tuning due to the @xmath0 mass limit , and furthermore this source of fine - tuning increasing quadratically with @xmath25 will rapidly overtake the logarithmic fine - tuning due to the stop mass limits , as the experimental mass limits of both types of particles increases in the future .
this was first pointed out in @xcite and has been discussed quantitatively @xcite in the framework of the constrained e@xmath9ssm @xcite , where it has been verified that this new source of fine - tuning dominates over all other sources . in this paper
we propose a new class of models called little @xmath0 models which differ from the usual class of @xmath1 models by having a reduced gauge coupling @xmath14 leading to the possibility of lower mass @xmath0 bosons .
such a reduction in the gauge coupling @xmath14 at the unification scale has some motivation from f - theory constructions @xcite .
we show that reducing @xmath14 relaxes the experimental limit on the @xmath0 mass , allowing a lighter value and hence reducing the tree - level fine - tuning associated with @xmath1 models .
we show that , although for sufficiently small values of @xmath14 the new source of fine - tuning due to the @xmath0 mass can be essentially eliminated , it does so at the expense of increasing the singlet vacuum expectation value , leading to overall fine - tuning similar to that in the minimal supersymmetric standard model .
we emphasise that the main prediction of little @xmath0 models is the presence of weakly coupled low mass @xmath0 resonances , perhaps as low as 200 gev .
the layout of the remainder of the paper is as follows . in section 2
we briefly review the e@xmath9ssm , followed by a discussion of the electroweak symmetry breaking ( ewsb ) conditions and the impact of the @xmath0 mass on fine - tuning in section 3 .
little @xmath0 models are introduced in section 4 , where the experimental limits on such a boson are studied as a function of its mass and ( reduced ) gauge coupling .
section 5 concludes the paper .
at low energies , the group structure of the exceptional supersymmetric standard model ( e@xmath9ssm ) is that of the standard model ( sm ) , along with the additional @xmath28 symmetry , @xmath29 the matter content of the model is contained in the complete 27-dimensional representation which decomposes under @xmath30 to , @xmath31 ordinary quarks and leptons are contained in the representations : @xmath32 and @xmath33 the higgs doublets and exotic quarks are contained in @xmath34 and @xmath35 the singlets are contained in @xmath36 , and finally the right handed neutrinos are included in @xmath37 moreover , the model requires three 27 representations , hence @xmath38 , in order to ensure anomaly cancellation .
this means that there are three copies of each field present in the model .
however , only the third generation ( by choice ) of the two higgs doublets , and the sm singlet acquire vevs .
the other two generations are called : inert . furthermore , in order to keep gauge coupling unification , non - higgs fields that come from extra incomplete @xmath39 representations are added to the model . as a result , a @xmath40 term , which is not necessary related to the weak scale , is present in the model .
the full superpotential consistent with the low energy gauge structure of the e@xmath9ssm contains includes both @xmath1 invariant invariant terms and @xmath1 breaking terms , full details of which are given in @xcite . to prevent proton decay and flavour changing neutral currents a discrete @xmath41 symmetry is imposed .
all superfields except the third generation higgs doublets and singlet are odd under this symmetry .
the @xmath41 invariant superpotential then reads , @xmath42 & & + \dfrac{1}{2}m_{ij}\hat{n}^c_i\hat{n}^c_j+\mu'(\hat{h}'\hat{\overline{h'}})+ h^{e}_{4j}(\hat{h}_d \hat{h}')\hat{e}^c_j+h_{4j}^n ( \hat{h}_{u } \hat{h}')\hat{n}_j^c \nonumber \\[2 mm ] & & + w_{\rm{mssm}}(\mu=0 ) , \label{eq : suppot}\end{aligned}\ ] ] where the indices @xmath43 and @xmath44 denote the generations .
@xmath45 is the sm singlet field , @xmath46 and @xmath47 are the higgs doublet fields corresponding to the up and down types .
exotic quarks and the additional non - higgs fields are denoted by @xmath48 and @xmath49 respectively . finally to ensure that only third generation higgs like fields get vevs a certain hierarchy between the yukawa couplings must exist .
defining @xmath50 , we impose @xmath51 .
the scalar higgs potential is , @xmath52 + \delta_{\text{loops } } } \end{split}\ ] ] where , @xmath53 , @xmath54 , and @xmath55 are the gauge couplings of @xmath56 ( gut normalized ) , and the additional @xmath28 , respectively . @xmath57 and @xmath58 are effective @xmath28 charges of @xmath59 and @xmath45 , respectively .
@xmath60 is the mass of the singlet field , and @xmath61 .
the higgs field and the sm singlet acquire vevs at the physical minimum of this potential , @xmath62 + it is reasonable exploit the fact that @xmath63 , which will help in simplifying our master formula for fine - tuning as will be seen in section 4 .
then , from the minimisation conditions , @xmath64 the electroweak symmetry breaking ( ewsb ) conditions are , @xmath65 @xmath66 @xmath67 where @xmath68 and @xmath69 .
eq.[mz ] can be written , @xmath70 where @xmath71 and @xmath72 if one takes @xmath73 we have @xmath74 and the factor in front of @xmath75 in is equal to one and we recover the well known mssm relation between @xmath76 , @xmath77 and the soft higgs masses @xmath78 , @xmath79 .
written in this form , which may be compared to eq.[mz0 ] but with the coefficients @xmath17 explicitly given , it is clear that fine - tuning will increase quadratically as @xmath25 increases . to avoid any fine - tuning we would like to keep @xmath80 gev or less
this motivates the main idea of this paper , namely to relax the cms experimental mass limit of @xmath27 tev @xcite down to @xmath81 gev by reducing its gauge coupling @xmath14 . indeed , as we shall see , such a low value of @xmath81 gev may be made consistent with the experimental limit by choosing @xmath82 and @xmath83 tev . in order to keep @xmath84 close to the electroweak scale
this requires a very small value of @xmath85 . in fig .
[ fig : finetuning ] the contribution @xmath86 to fine - tuning from @xmath87 is plotted , where @xmath86 is defined as follows .
@xmath88 mass . ]
we emphasise that the appearance of @xmath25 in the tree - level minimisation condition is characteristic of all susy @xmath0 models where the usual higgs doublets carry @xmath3 charges ( e.g. it applies to all @xmath1 models but not , for example the @xmath89 model . )
this provides a motivation for little @xmath0 models in which the extra gauge coupling @xmath14 is reduced and the experimental lower bound on @xmath25 may be relaxed .
in general little @xmath0 models can be defined by the gauge group su(3)_csu(2)_lu(1)_y u ( 1)^[low ] where the standard model is augmented by an additional @xmath90 gauge group with a gauge coupling @xmath14 which is significantly smaller than the hypercharge gauge coupling @xmath91 .
the @xmath2 gauge group is broken at low energies giving rise to a massive @xmath92 gauge boson with couplings to a sm fermion @xmath93 given by @xcite : @xmath94 the values of @xmath95 depend on the particular choice of @xmath90 and on the particular fermion @xmath93 .
we assume universality amongst the three families .
more explicitly , this assumption implies that @xmath96 , @xmath97 , @xmath98 , and @xmath99 for up - quarks , down - quarks , charged leptons and neutrinos respectively .
the axial couplings , @xmath100 , behave accordingly . in a given model
there are eight model dependent couplings of the extra @xmath92 boson to sm fermions , that is @xmath101 with @xmath102 .
these are fixed by group theory , so can not be changed for a given model
. however the low energy @xmath3 gauge coupling @xmath14 is fixed by a unification condition .
e.g. in e@xmath9ssm @xmath103 which is approximately equal to the ( gut normalised ) hypercharge gauge coupling .
if unification of @xmath14 with the other gauge couplings is relaxed , then @xmath14 becomes a free parameter . in this paper
we are interested in taking it to be smaller than the gut prediction , namely we shall consider @xmath104 , keeping @xmath95 fixed at their model predictions . specializing to the charged lepton pair production cross - section relevant for the first runs at the lhc , the cross - section may be written at the leading order ( lo ) as @xcite : _
^+^-^lo = [ eq : ll ] where the coefficients @xmath105 and @xmath106 are given by : c_u=(g_v^u^2+g_a^u^2)br(^+^- ) , c_d=(g_v^d^2+g_a^d^2)br(^+^- ) , [ eq : cucd ] and @xmath107 and @xmath108 are related to the parton luminosities @xmath109 and @xmath110 and therefore only depend on the collider energy and the @xmath0 mass .
all the model dependence of the cross - section is therefore contained in the two coefficients , @xmath105 and @xmath106 .
these parameters can be calculated from @xmath111 and @xmath14 , assuming only sm decays of the @xmath0 boson .
note that the cross - section is proportional to @xmath112 and will therefore be reduced in little @xmath0 models in which @xmath104 . a given model such as the e@xmath9ssm @xcite appears as a point in the @xmath113 plane , assuming that the low energy @xmath3 gauge coupling @xmath14 is fixed by a unification condition .
if we relax the unification condition then the point will become a line in the @xmath113 plane , since each of @xmath105 and @xmath106 are proportional to @xmath112 and the points on the line will approach the origin as @xmath114 . for example in the e@xmath9ssm we have : c_u = 5.9410 ^ -4 ^ 2 , c_d = 1.4810 ^ -3 ^ 2 .
[ cucdgp ] since the experimental @xmath0 mass contours in the @xmath113 plane are fixed for a given limit on the cross - section , the effect of reducing @xmath14 will not change those contours
. the only effect of reducing @xmath14 is to move the model point in the @xmath113 plane closer to the origin , resulting in a reduced experimental limit on the @xmath0 mass .
see for example @xcite where this approach is followed for conventional @xmath0 models .
although this provides a simple way to understand qualitatively why the experimental limits are relaxed by reducing @xmath14 , it turns out that for the lower mass @xmath0 signal regions backgrounds and other constraints become more important for this reason we shall not present our results in the @xmath113 plane . in the e@xmath9ssm
the @xmath0 mass is given to good approximation by : m^2_z=g^2_1 v^2(_1 ^ 2 ^ 2+_2 ^ 2 ^
2)+g^2_1 ^ 2_s s^2 g^2_1 ^ 2_s s^2 , [ zpmass ] where the charges are @xmath115 , @xmath116 , @xmath117 .
the last approximation in eq.[zpmass ] assumes @xmath118 where we can neglect the terms involving the electroweak vev @xmath119 gev .
what is the effect of reducing @xmath14 in this model ?
on the one hand , reducing @xmath14 will reduce @xmath25 in direct proportion , since @xmath120 for a fixed value of @xmath23 .
on the other hand , reducing @xmath14 will reduce the cross - section since @xmath121 ( see eq.[cucdgp ] ) . in fig .
[ fig : cs - scan - m ] we show the cross section for lepton ( @xmath122 ) pair - production via @xmath0 at lhc at @xmath123 tev in the @xmath14
@xmath87 plane for the little @xmath0 models with charges corresponding to the e@xmath9ssm .
the horizontal , dashed line indicates the standard gut predicted @xmath14 value .
the dash - dotted lines are cross section limits on the e@xmath9ssm @xmath124 from d@xmath125@xcite , cms@xcite and atlas@xcite that have been converted to limits on the coupling @xmath126 .
the estimated indirect exclusion limits from electro - weak precision tests ( ewpt)@xcite , mostly by lep , on the ratio @xmath127 are plotted with red crosses .
these are not available for the e@xmath9ssm but the limit for the @xmath128 @xmath92 is @xmath129 and the limit for the @xmath130 @xmath92 is @xmath131 as an estimate of these limits for the e@xmath9ssm , we plot with red crosses an intermediate limit @xmath132 figure [ fig : cs - scan - m ] also shows contours of constant values of the singlet vev @xmath23 , so it is possible to read off exclusions limits on @xmath23 . at large masses the limits from atlas and cms
follow the cross section contours well but in the low mass regime the standard model background is large which weakens the limits on the cross section . in this region , just above 200 gev , the direct searches by the lhc and tevatron experiments place the strongest bounds on the coupling and all place limits of about @xmath133 .
it is obvious from fig .
[ fig : cs - scan - m ] that it is possible to lower the limit on @xmath134 by decreasing the coupling @xmath126 but by doing this the value of the singlet vev , @xmath23 , generally has to increase .
the limit on @xmath23 is however strongest for @xmath134 of about 500 - 800 gev and gets slightly relaxed in the lowest mass region , 200 - 500 gev .
examples of how the limits on @xmath134 , @xmath23 and the fine - tuning with respect to @xmath134 changes as the coupling @xmath55 decreases are tabulated in tab .
[ tab : g1p ] . )
pair - production via @xmath0 at lhc at @xmath123 tev in the @xmath14
@xmath87 plane for little @xmath0 models with charges corresponding to the e@xmath9ssm .
the horizontal , dashed line indicates the standard gut predicted @xmath14 value .
exclusion limits from direct searches are plotted with dash - dotted lines in magenta , black and blue for d@xmath125 , cms and atlas respectively .
indirect constraint on the mass - coupling ratio from electro - weak precision tests are plotted with red crosses and coincides with the contour for the singlet vev @xmath135 tev . ]
.scenarios with different values of @xmath126 for the little @xmath0 models with charges corresponding to the e@xmath9ssm . the @xmath124 mass and thus its source of fine - tuning , @xmath136 , can be reduced by reducing @xmath55 at the cost of increasing the singlet vev , @xmath23 . because experimental limits on the cross section get weaker in the low mass region the limit on @xmath23 gets slightly weaker , hence the weaker limit on @xmath23 in the case of @xmath137 . [ cols="^,^,^,^,^,^",options="header " , ]
the current experimental limits from the lhc on the @xmath0 boson mass of 2 - 3 tev raises the fine - tuning in @xmath1 supersymmetric models to undesirably high levels .
this is a generic property of susy models where the higgs doublets carry the @xmath2 charge . in order to solve this problem
we have proposed a new class of models called _ little @xmath0 models _ involving a weakly coupled lower mass @xmath0 .
these models can originate from supersymmetric @xmath1 inspired supersymmetric models where the spontaneously broken extra @xmath3 gauge group has a reduced gauge coupling .
we have shown that reducing the value of the extra gauge coupling relaxes these limits , leading to the possibility of low mass @xmath0 resonances , for example down to about 200 gev , thereby reducing fine - tuning due to the @xmath0 mass down to acceptable levels .
such a reduced extra gauge coupling does not affect conventional gauge coupling unification of the strong , weak and electromagnetic gauge couplings and in fact is well motivated in certain classes of f - theory models .
we emphasise the main experimental prediction of such little @xmath0 models which is the appearance of a low mass weakly coupled @xmath0 which may yet appear in future lhc searches .
although the source of tree level fine - tuning due to the @xmath0 mass is reduced in little @xmath0 models , it does so at the expense of increasing the singlet vacuum expectation value , leading to overall fine - tuning similar to that in the minimal supersymmetric standard model .
sfk acknowledges partial support from the stfc consolidated st / j000396/1 and eu itn grants unilhc 237920 and invisibles 289442 .
ps is thankful for support from the next institute , a part of sepnet .
b * 634 * ( 2006 ) 278 [ hep - ph/0511256 ] .
u. ellwanger , m. rausch de traubenberg and c.a .
savoy , phys .
b * 315 * ( 1993 ) 331 , z. phys .
c * 67 * ( 1995 ) 665 and nucl .
b * 492 * ( 1997 ) 307 ; u. ellwanger , phys .
b * 303 * ( 1993 ) 271 ; p. pandita , z. phys .
c * 59 * ( 1993 ) 575 ; t. elliott , s.f .
king and p.l .
white , phys .
d * 49 * ( 1994 ) 2435 ; s.f .
king and p.l .
white , phys .
d * 52 * ( 1995 ) 4183 ; f. franke and h. fraas , int .
j. mod .
phys . a * 12 * ( 1997 ) 479 .
. lett .
b * 681 * ( 2009 ) 448 [ arxiv:0901.1192 [ hep - ph ] ] ; p. athron , s. f. king , d. j. miller , s. moretti and r. nevzorov , phys .
d * 86 * ( 2012 ) 095003 [ arxiv:1206.5028 [ hep - ph ] ] .
j. c. callaghan and s. f. king , arxiv:1210.6913 [ hep - ph ] . | we propose a new class of models called _ little @xmath0 _ models in order to reduce the fine - tuning due to the current experimental limits on the @xmath0 mass in @xmath1 inspired supersymmetric models , where the higgs doublets are charged under the extra @xmath2 gauge group .
the proposed _
little @xmath0 _ models allow a lower mass @xmath0 due to the spontaneously broken extra @xmath3 gauge group having a reduced gauge coupling .
we show that reducing the value of the extra gauge coupling relaxes the experimental limits , leading to the possibility of low mass @xmath0 resonances , for example down to 200 gev , which may yet appear in lhc searches .
although the source of tree level fine - tuning due to the @xmath0 mass is reduced in little @xmath0 models , it typically does so at the expense of increasing the vacuum expectation value of the @xmath2-breaking standard model singlet field , reducing the fine - tuning to similar levels to that in the minimal supersymmetric standard model .
* little @xmath0 models * + alexander belyaev , s. f. king , patrik svantesson + _ school of physics and astronomy , university of southampton , + southampton , so17 1bj , u.k . |
in part i , the double well potential problem ( dwp ) is defined by @xmath0 where @xmath1 is an @xmath2 real symmetric matrix , @xmath3 is an @xmath4 real matrix , @xmath5 , @xmath6 and @xmath7 . by introducing a continuous variable transformation @xmath8 , the double well potential problem ( dwp )
can be transformed into the following equivalent quadratic program over one nonhomogeneous quadratic constraint ( qp1qc ) : @xmath9 the dual problem of ( qp1qc ) and the dual of the dual were studied in part 1 ( theorem 1 ) in order to find a global minimum solution to problem ( dwp ) . for practical applications , knowing only the global minimum of a double well potential function may not be sufficient .
for example , the double well potential model can be used to describe the ion - molecule reactions , where the intermediate molecule complexes must go across the energy barrier to cause reactions @xcite .
researchers have to know the potential difference between the energy wells ( caused by local minima ) and energy barrier ( caused by local maximum ) .
the understanding of all types of critical points of a double well function is thus necessary .
mathematically , we are motivated by the pioneering work of martnez @xcite which showed that a trust - region subproblem ( trs ) @xcite of the following form @xmath10 ( with @xmath11 being a positive scalar ) has at most one local , but non - global , minimizer .
please notice that , on one hand , problem ( qp1qc ) can be regarded as an extension of problem ( trs ) towards the nonhomogeneous and possibly singular case . on the other hand ,
the penalty version of the trust - region subproblem , namely , @xmath12 ( with the penalty parameter @xmath13 being sufficiently large ) is clearly a special case of the double well potential problem ( dwp ) .
therefore , our approach to analyzing the local non - global minimizer of a double well potential problem extends the results of @xcite .
moreover , when restricted to problem ( trs ) , our approach simplifies the proof provided in @xcite .
although , in general , a double well potential problem may have infinitely many local , but non - global , minimizers ( see figure 1 ) , we ll show that , after taking the space reduction technique developed in section 2 of part i , the reduced nonsingular problem has at most one local non - global minimizer and at most one local maximizer .
[ fig:0 ] we remark that characterizing the local maximizer of the trust - region subproblem ( [ tr:1])-([tr:2 ] ) can be reduced to the problem of finding a local minimizer of ( [ tr:1 ] ) with @xmath1 being replaced by @xmath14 .
however , due to the non - symmetric nature , it is no longer the case for the double well potential problem ( [ d - well ] )
. hence martnez s approach may not be able to characterize the local maximizer for a general ( dwp ) problem . in the rest of the paper , a characterization of the local , but
non - global , minimizer of a double well function is provided in section 2 .
then , a characterization of the global minimizer of a double well function is given in section 3 , while the local maximizer is characterized in section 4 .
computational algorithms for each type of the optimizers of a double well potential function are proposed in section 5 with some illustrative examples .
some concluding remarks are given in section 6 .
here we define some notations to be used throughout the paper .
let @xmath15 be the set of all @xmath16-dimensional symmetric real matrices , @xmath17 be the set of all @xmath16-dimensional positive semi - definite matrices , and @xmath18 be the set of all n - dimensional positive definite matrices .
for any @xmath19 , @xmath20 means that matrix @xmath21 and @xmath22 means that matrix @xmath23 .
we sometimes write @xmath24 for @xmath20 and @xmath25 for @xmath22 .
the @xmath26th smallest eigenvalue of @xmath27 is denoted by @xmath28 and the determinant of @xmath29 by @xmath30 .
the @xmath16-dimensional identity matrix is denoted by @xmath31 . for a vector @xmath32
, @xmath33 ( @xmath34 ) means that each component of @xmath35 is nonnegative ( positive ) and diag@xmath36 is an @xmath16-dimensional diagonal matrix with diagonal components being @xmath37 .
moreover , for a number @xmath38 , sign@xmath39 if @xmath40 , otherwise sign@xmath41 .
following the space reduction technique developed in part i , without loss of generality , we may assume that , in problem ( bwp ) , @xmath42 is positive definite such that , matrices @xmath1 and @xmath42 are simultaneously diagonalizable via congruence , i.e. , there is a nonsingular matrix @xmath29 such that @xmath43 with @xmath44 and @xmath45 .
it follows immediately that @xmath46 .
let @xmath47 then we have @xmath48 and @xmath49 for simplicity , we define @xmath50 and @xmath51 .
by dropping the constant terms , we can rewrite problem ( dwp ) defined in ( [ d - well ] ) as @xmath52 recall that the canonical primal problem defined in ( 19 ) of part i is to minimize @xmath53 the form in ( [ d - well:4 ] ) is a further simplified version of form ( [ d - well:19 ] ) by setting @xmath54 . in this way
, the third order term in problem ( dwp ) is eliminated and the complexity is decreased for analysis .
it s interesting to note that , in the finite deformation theory , the diagonal matrix @xmath55 represents the material constants , the first order coefficient vector @xmath56 stands for the external forces , and the cauchy - green strain @xmath57 measures the square of local changes in distance due to deformation . as we shall observe below , the first order and the second order necessary conditions of ( [ d - well:4 ] ) ( see @xcite ) are highly related to the term of @xmath58 , which is the sum of the cauchy - green strain and the material constants .
our first result of lemma [ lem:2 ] will show that , at a local minimum of the double well potential function , the cauchy - green strain can not be too small , at least no smaller than the negative of the second smallest material constant .
[ lem:1 ] assume that @xmath59 is a local minimizer of ( [ d - well:4 ] ) .
it holds that @xmath60 [ lem:2 ] assume that @xmath61 and @xmath59 is a local minimizer of ( [ d - well:4 ] ) .
it holds that @xmath62 furthermore , if @xmath63 , then @xmath64 suppose that the statement ( [ lem : eq ] ) is false , then @xmath65 .
hence @xmath66 .
let @xmath67 and @xmath68 . if @xmath69 , by the necessary condition ( [ con:2 ] ) , we have @xmath70 which causes a contradiction . on the other hand , if @xmath71 , then , by ( [ con:2 ] ) again , we have @xmath72 it again causes a contradiction .
therefore , the statement ( [ lem : eq ] ) must be true .
when @xmath63 , suppose that the statement ( [ lem : ineq ] ) is false , then we have @xmath73 by ( [ lem : eq_1 ] ) , we know that the second order necessary condition ( [ con:2 ] ) becomes @xmath74 since the first two leading principal minors of the matrix in ( [ x_2 ] ) are nonnegative , we have @xmath75 and @xmath76+\left[\begin{array}{cc}\alpha_1-\alpha_2&0\\0&0 \end{array}\right ] \right\}= ( \alpha_1-\alpha_2)\underline{w}_2^{2}\ge 0.\ ] ] remember that @xmath77 , inequality ( [ x_0 ] ) implies that @xmath78 .
moreover , inequality ( [ x1 ] ) implies that @xmath79 .
together with ( [ con:1 ] ) , we obtain that @xmath80 and @xmath81 without loss of generality , we assume that @xmath82 , and hence @xmath83 .
this implies , from ( [ lem : eq_1 ] ) and the fact that @xmath79 , we have @xmath84 consider the following parametric curve in @xmath85 : @xmath86 where @xmath87 , i.e. , @xmath88 passes through @xmath59 at @xmath89 .
evaluating @xmath90 on @xmath88 , we have @xmath91 it is not difficult to see that @xmath89 is a local minimum point of @xmath92 since @xmath59 is a local minimizer of @xmath90
. however , this conclusion contradicts to the fact that @xmath93 therefore , the statement ( [ lem : ineq ] ) must be true , if @xmath63 .
the next result lemma shows that any critical point of the double well potential function having a sufficiently large cauchy - green strain ( larger than the negative of all the material constants ) must be a global minimum point .
[ lem:3 ] let @xmath94 be a critical point of the function @xmath90 in problem ( [ d - well:4 ] ) with @xmath95 . if @xmath96 then @xmath94 is a global minimizer of problem ( [ d - well:4 ] ) .
in particular , a local minimizer @xmath59 of problem ( [ d - well:4 ] ) satisfying condition ( [ con:3 ] ) must be a global minimizer .
define @xmath97 . by the assumption that @xmath98 , it follows that @xmath99 $ ] , and @xmath100 is positive semidefinite .
then , @xmath101 since @xmath102 , the lower bound function expressed in ( [ fun:1 ] ) is a convex quadratic function .
its global minimum is attained at any @xmath103 satisfying @xmath104 .
since @xmath94 is a critical point of ( [ d - well:4 ] ) , by equation ( [ con:1 ] ) in lemma [ lem:1 ] , it is a global minimizer of the lower bound function in ( [ fun:1 ] ) and thus inequality ( [ fun:2 ] ) holds .
finally , ( [ fun:3 ] ) becomes true by substituting @xmath94 into ( [ fun:0 ] ) .
[ theorem1 ] the double well potential problem ( [ d - well:4 ] ) has at most one local , but non - global , minimizer .
let us assume that @xmath61 first .
lemmas [ lem:2 ] and [ lem:3 ] imply that any local , but non - global , minimizer @xmath59 of problem ( [ d - well:4 ] ) exits only if @xmath63 and @xmath105 .
consequently , we know the matrix @xmath106 is nonsingular with its first diagonal element being negative and others positive .
therefore , @xmath59 can be uniquely determined by equation ( [ con:1 ] ) with @xmath107.\label{sol}\ ] ] from ( [ con:2 ] ) , we have @xmath108 since @xmath109 , we know that @xmath110 putting all @xmath111 together , we have @xmath112 in other words , the norm square of the local minimizer , i.e. , @xmath113 , must be the root of the following secular function on a specific open interval : @xmath114 notice that each root of @xmath115 can only correspond to one local non - global minimizer of problem ( [ d - well:4 ] ) using ( [ sol ] ) .
taking a simple calculation of ( [ varphi ] ) , we have @xmath116 therefore , the secular function @xmath117 is a strictly convex function on @xmath118 with at most two roots . furthermore , since @xmath119 is nonsingular , the second order necessary condition ( [ con:2 ] ) implies that @xmath120 where @xmath121 since its determinant is nonnegative , we have @xmath122 in other words , if @xmath59 is a local minimizer of problem ( [ d - well:4 ] ) , then it must satisfy the second order necessary condition ( in matrix form ) whose determinant is the first derivative of the secular function at @xmath123 .
however , a strictly convex function has at most one root with a nonnegative first derivative .
thus we have shown the theorem for @xmath124 . when @xmath125 , it amounts to setting @xmath126 in the above analysis , and the proof follows .
the next corollary provides some simple sufficient conditions for having no local non - global minimizer .
[ cor:1 ] when one of the following conditions is met : * @xmath127 ( in this case @xmath90 is convex ) ; * @xmath128 ; * @xmath129 ; * @xmath130 , @xmath131 and @xmath132 ; any local minimizer of the double well potential problem ( [ d - well:4 ] ) is globally optimal .
\(i ) if @xmath127 , then @xmath133 for any @xmath134 and @xmath135 $ ] . using the second derivative of @xmath90 in ( [ con:2 ] ) , we have @xmath136 which shows that @xmath90 is indeed convex and any local optimum becomes globally optimal .
\(ii ) if @xmath59 is a local minimizer and @xmath128 , then inequality ( [ lem : eq ] ) in lemma [ lem:2 ] and lemma [ lem:3 ] imply that @xmath59 is indeed the global minimizer of problem ( [ d - well:4 ] ) .
\(iii ) if @xmath129 , then equation ( [ con:1 ] ) leads to either @xmath137 or @xmath138 . by property ( [ con:2 ] )
, @xmath137 further implies that @xmath139 . using lemma [ lem:3 ] ,
both cases lead @xmath59 to be a global minimizer .
\(iv ) in this case , the secular function @xmath117 actually does not have any solution in its domain .
the key result of establishing a necessary and sufficient condition for local , non - global minimizer is provided below .
[ cor:0 ] the double well potential problem ( [ d - well:4 ] ) has a local - nonglobal minimizer if and only if there is a @xmath140 such that the secular function @xmath141 and @xmath142 .
moreover , when it exists , the local non - global minimizer is given by @xmath143 suppose that @xmath144 with @xmath140 . for the @xmath59 defined by ( [ tw ] ) , we have @xmath145 and @xmath59 satisfies the first order necessary condition ( [ con:1 ] ) . moreover , the diagonal matrix @xmath146 is nonsingular with positive diagonal elements except for the first one . by weyl s inequality
( see @xcite , theorem 4.3.1 ) , we can estimate the largest @xmath147 eigenvalues of the second order matrix @xmath148 by @xmath149 since @xmath150 , by ( [ determinant1 ] ) and ( [ determinant2 ] ) , we have @xmath151 where @xmath152 is defined in ( [ gamma:1 ] ) . combining ( [ new:1 ] ) with ( [ new:2 ] ) , we know that the smallest eigenvalue of the second order matrix must be positive , i.e. , @xmath153 this is a sufficient condition to guarantee that @xmath59 is a local minimizer of problem ( [ d - well:4 ] ) . on the hand side ,
let @xmath59 be a local , non - global minimizer of problem ( [ d - well:4 ] ) , which is unique quaranteed by theorem [ theorem1 ] .
let @xmath145 . by the proof of theorem [ theorem1 ] , we know @xmath140 .
moreover , @xmath59 can be expressed by @xmath154 as in ( [ tw ] ) because @xmath59 satisfies the first order necessary condition ( [ con:1 ] ) .
also we have @xmath141 and @xmath155 .
it remains for us to show that @xmath156 .
suppose that , by contradiction , @xmath157 .
from ( [ new:2 ] ) , we have @xmath158 and thus there is a @xmath159 such that @xmath160 from ( [ new:000 ] ) , we can write @xmath161 then , @xmath162 implies that @xmath163 consider the double well potential function along the direction @xmath164 defined by @xmath165 .
it is routine to verify that @xmath166 by the first order necessary condition ( [ con:1 ] ) , we have @xmath167 . by ( [ con:2 ] ) and ( [ new:000 ] ) , we further have @xmath168
. however , ( [ nzero ] ) implies that @xmath169 this result contradicts to the fact that @xmath59 is a local minimizer of problem ( [ d - well:4 ] ) .
therefore , @xmath156 and the proof is complete .
in this section , we try to characterize different aspects of the global minimizer of the double well potential problem .
we first observe that the double well potential function tends to @xmath170 as @xmath171 .
therefore , the global minimizer of problem ( [ d - well:4 ] ) always exists .
our first result is that each component of the global minimizer must be of the same sign as the corresponding component of the external force ( i.e. , the first - order term vector ) .
[ lem:4 ] if @xmath172 is the global minimizer of ( [ d - well:4 ] ) , then @xmath173.\ ] ] let @xmath174 .
since the only odd - order term in @xmath90 is the linear term , we have @xmath175 hence we know @xmath176 .
a similar argument applies for any other components .
the next result shows that the sufficient condition @xmath98 in lemma [ lem:3 ] is indeed necessary for a critical point to become the global minimizer .
[ thm : suf - nec ] @xmath172 is a global minimizer of ( [ d - well:4 ] ) if and only if @xmath177 and @xmath178 the sufficiency is clear from lemma [ lem:3 ] .
in addition , we can observe that the necessity of ( [ glob : s2 ] ) follows immediately from equation ( [ con:1 ] ) .
it remains to show that ( [ glob : s1 ] ) is also a necessary condition . to avoid triviality
, we may assume that @xmath63 . otherwise , by substituting @xmath128 into ( [ lem : eq ] ) , we can obtain the result at once .
suppose that @xmath179 then ( [ con:2 ] ) implies that @xmath180 hence we have @xmath181 . using ( [ glob : s2 ] ) , we have @xmath182 this causes a contradiction to lemma [ lem:4 ] and the proof follows .
an immediate consequence of theorem [ thm : suf - nec ] is that the sign of the first component of the local non - global minimizer , if it exists , must be opposite to that of the first component of a global minimum solution .
[ cor:3 ] if @xmath59 be the local non - global minimizer and @xmath172 is a global minimizer of @xmath90 of problem ( [ d - well:4 ] ) , then @xmath183 since both @xmath59 and @xmath172 are critical points , theorem [ thm : suf - nec ] implies that @xmath184 and @xmath185 .
it follows from condition ( iii ) of corollary [ cor:1 ] that @xmath186 and , from ( [ con:1 ] ) , @xmath187 consequently , @xmath188 in section 4 of part i , we have shown that the dual of the dual of the canonical primal problem ( p ) ( see equation ( 19 ) of part i ) is equivalent to only a portion of ( p ) subject to @xmath16 linear constraints ( see equation ( 35 ) of part i ) .
moreover , that portion contains the global minimizer . in the simplified version here , we have the third order term coefficient @xmath189 , which reduces the dual of the dual problem in part i to the following problem : @xmath190 the portion of ( p ) corresponding to ( [ dualofdual ] ) becomes @xmath191 under the nonlinear one - to - one map : @xmath192 from lemma [ lem:4 ] , we know that the portion specified by ( [ sdcdwpconstraint ] ) contains the global minimizer @xmath94
. however , due to the opposite sign behavior on the first component , corollary [ cor:3 ] implies that the local non - global minimizer @xmath59 is not in that portion .
the mapping ( [ x - y ] ) was used to reveal the hidden convexity of ( qp1qc ) in part i , but the local non - global minimizer is definitely excluded from the transformation .
the missing of the local non - global minimizer can been seen clearly in examples 1 and 2 of part i.
it is not difficult to see that the global maximum of problem ( [ d - well:4 ] ) goes to @xmath170 as @xmath193 grows without a bound .
hence there is no global maximizer of the problem . in this section ,
we provide an analytic study of the local maximizer of the simplified problem ( [ d - well:4 ] ) .
[ lem:1n ] if @xmath194 is a local maximizer of ( [ d - well:4 ] ) , then @xmath195 the proof is easy .
moreover , it follows directly from ( [ con:2n ] ) that @xmath196 in other words , at the local maximizer , the value of the cauchy - green strain is smaller than the negative value of all material constants .
[ max : re ] if @xmath194 is a local maximizer of ( [ d - well:4 ] ) , then @xmath197 it follows from ( [ con:1n ] ) that @xmath198 if @xmath199 , then @xmath200 . on the other hand ,
if @xmath200 , it implies from @xmath201 that either @xmath202 or @xmath199 ( or both ) .
suppose that @xmath202 .
it follows from ( [ con:2n ] ) that @xmath203 therefore , @xmath204 must be also @xmath205 , and the proof follows .
[ mc:1 ] if @xmath206 , then the double well potential problem ( [ d - well:4 ] ) has no local maximizer .
if @xmath207 , then ( [ a2 ] ) can not be true and we have the conclusion .
now , assume that @xmath208 .
if ( [ d - well:4 ] ) has a local maximizer @xmath194 , then it follows from ( [ a2 ] ) that @xmath209 or , equivalently , @xmath210 by lemma [ max : re ] , we have @xmath211 .
it is routine to verify that @xmath212 since @xmath213 and @xmath211 , we have @xmath214 notice that @xmath215 consequently , @xmath213 is not a local maximizer and we reached a contradiction .
this completes the proof .
[ mc:2 ] if @xmath216 and @xmath211 , then the double well potential problem ( [ d - well:4 ] ) has a unique local maximizer @xmath213 .
since @xmath217 and @xmath218 , we have @xmath219 therefore , @xmath213 is a local maximizer of problem ( [ d - well:4 ] ) .
lemma [ max : re ] further guarantees that @xmath213 is the unique local maximizer .
[ mc:3 ] if @xmath216 and @xmath220 , then the double well potential problem ( [ d - well:4 ] ) has at most one local maximizer .
suppose @xmath194 is a local maximizer of ( [ d - well:4 ] ) .
since @xmath220 , we let @xmath221 $ ] be the largest nonzero index in @xmath222 such that @xmath223 in addition , let @xmath224 be the identity matrix of order @xmath225 and @xmath226 from equation ( [ con:1n ] ) , we have @xmath227 . since @xmath228 , by lemma [ max : re ] , we know @xmath229 , which implies that @xmath230 . from inequality ( [ a2 ] ) , we further know that @xmath231 moreover , we have @xmath232 consequently , matrix @xmath233 is negative definite and , once @xmath234 is computed , @xmath235 can be uniquely determined by the following system of equations : @xmath236 since @xmath237 implies that @xmath238 , it follows that any local maximizer @xmath194 must satisfy that @xmath239 from ( [ a3 ] ) and ( [ a5 ] ) , we know @xmath240 is a root of the following convex secular function : @xmath241 since matrix @xmath233 is negative definite , from ( [ con:2n ] ) , we know that @xmath242 where @xmath243 and @xmath244 then , @xmath245 since a strictly convex function can have at most one root with its first derivative being non - positive , based on ( [ sol : n ] ) , we can conclude that there is at most one local maximizer . combining lemmas [ mc:1 ] , [ mc:2 ] and [ mc:3 ] together
, we have the next result . [ theorem2 ] the double well potential problem ( [ d - well:4 ] )
has at most one local maximizer .
the above result can be further extended to obtain a necessary and sufficient condition under which a local maximum exists .
[ cor:00 ] the double well potential problem ( [ d - well:4 ] ) has a local maximizer if and only if @xmath246 and there is a @xmath247 such that @xmath248 and @xmath249 for the secular function defined in ( [ equ : n ] ) .
moreover , if it exists , the local maximizer @xmath194 is given by @xmath250 \(i ) ( if part ) when @xmath211 , lemma [ mc:2 ] assures that @xmath213 is the unique local maximizer of ( [ d - well:4 ] ) , which can be expressed as ( [ tw1 ] ) . now , consider @xmath220 .
let @xmath251 be defined as in ( [ nonzero ] ) and @xmath194 as in ( [ tw1 ] ) .
since @xmath248 and @xmath238 , we have @xmath252 then we see that @xmath194 satisfies the first order necessary condition ( [ con:1n ] ) .
moreover , we have @xmath253 let @xmath243 . using weyl s inequality
( see @xcite , theorem 4.3.1 ) , we have @xmath254 therefore , the first @xmath255 eigenvalues of the matrix @xmath256 are negative .
it follows from ( [ determinant3 ] ) , ( [ determinant4 ] ) and the assumption of @xmath257 that @xmath258 if @xmath225 is even , then @xmath259 which implies that the @xmath260 eigenvalue of matrix @xmath261 is negative .
if @xmath225 is odd , then @xmath262 which says that the @xmath260 eigenvalue of matrix @xmath261 is again negative .
in other words , @xmath263 . from ( [ ne ] ) , we have @xmath264\prec 0,\ ] ] where @xmath265 .
consequently , @xmath194 satisfies the second order sufficient condition and becomes a local maximizer .
\(ii ) ( only if part ) let @xmath194 be the unique local maximizer of ( [ d - well:4 ] ) . by lemma [ mc:1 ] , we have @xmath246 . if @xmath211 , lemma [ mc:2 ] implies that @xmath213 .
in this case , since @xmath266 , there is a unique @xmath267 such that @xmath268 and @xmath269 .
the expression ( [ tw1 ] ) follows immediately .
assume that @xmath220 with @xmath228 and @xmath237 , and let @xmath270 . from ( [ equ : n ] ) and ( [ determinant4 ] ) , we know @xmath271 and @xmath272 .
then the expression for @xmath194 in ( [ tw1 ] ) follows from ( [ sol : n ] ) and @xmath238 . in the rest of the proof
, we shall show a stronger result of which @xmath247 and @xmath273 first , if @xmath274 , then @xmath275 . from ( [ a2 ] ) , we know @xmath276 $ ] .
suppose that @xmath277 , similar to ( [ gw_n ] ) , we can verify that @xmath278 and @xmath279 this is a contradiction to the fact that @xmath194 being a local maximizer of probem ( [ d - well:4 ] ) .
therefore , @xmath247 .
next , we show that @xmath280 , if not so , we consider @xmath281 by ( [ det:41 ] ) , we have @xmath282 consequently , matrix @xmath283 is singular and there exists a @xmath284 such that @xmath285 equivalently , @xmath286 since @xmath162 , we know @xmath287 define @xmath288 then , we have @xmath289 similar to the proof of theorem [ cor:0 ] , we can consider @xmath290 .
it follows that @xmath291 since @xmath194 satisfies the first order necessary condition ( [ con:1n ] ) , @xmath167 . by ( [ second-0 ] )
, we have @xmath168 .
moreover , ( [ nzero:1 ] ) implies that @xmath292 consequently , @xmath205 is not a local maximizer of @xmath293 and @xmath194 is not a local maximizer , which causes a contradiction .
therefore , we know @xmath257 .
this completes the proof .
the next result shows that , when it exists , the unique local maximizer is `` surrounded '' by all local ( non - global and global ) minimizers .
[ theorem4 ] if @xmath59 is a local minimizer and @xmath194 is the local maximizer of the double well potential problem ( [ d - well:4 ] ) , then @xmath294 \(i ) @xmath61 : if @xmath63 , following lemma [ lem:2 ] and ( [ a2 ] ) , we have @xmath295 otherwise , @xmath128 and we assume that @xmath296 . applying lemma [ lem:2 ] and ( [ a2 ] ) again , we have @xmath297 since both @xmath59 and @xmath194 are critical points of @xmath90 , lemma [ lem:3 ] implies that both of them are global minimizers , which is impossible .
therefore , @xmath298 .
\(ii ) @xmath125 : if @xmath129 , then the first order necessary condition ( [ con:1n ] ) implies that either @xmath213 or @xmath299 . since @xmath194 can not be a global minimizer , the latter case is eliminated and thus @xmath213
. to prove ( [ relat ] ) , it is sufficient to show that @xmath300 .
suppose that @xmath301 , then the second order necessary condition ( [ con:2 ] ) implies that @xmath302 by corollary [ cor:1 ] ( i ) , @xmath90 is convex and hence the local maximizer @xmath194 does not exist , which causes a contradiction to the setting of the theorem .
if @xmath131 , then @xmath303 for any local minimizer or maximizer @xmath304 . therefore , @xmath305 and @xmath306 are two solutions to the following equation : @xmath307 from the proofs of theorem [ theorem1 ] and theorem [ theorem2 ] , we have @xmath308 since @xmath117 is strictly convex , it has two distinct solutions satisfying the above first order conditions only when @xmath309 this completes the proof .
according to corollaries [ cor:0 ] and [ cor:00 ] , the local , non - global minimizer and the local maximizer of the simplified version of ( [ d - well:4 ] ) , if they exist , are closely related to the convex secular function @xmath117 over different intervals .
the convex secular function @xmath117 is a convenient substitute for the first order necessary condition , while the intervals capturing the root of @xmath117 reflect the second order necessary condition .
the sign of the first derivative of @xmath117 at the root provides necessary and sufficient conditions for the type of a local extremum , namely , positive sign for the local , but non - global , minimizer ; negative sign for the local maximizer .
the necessary and sufficient condition for the global minimum @xmath94 in theorem [ thm : suf - nec ] can also be expressed in terms of the secular function @xmath117 . from ( [ glob : s1 ] ) , we have @xmath310 . if @xmath311 , ( [ glob : s2 ] ) implies that @xmath312 .
moreover , by ( [ a4 ] ) , @xmath313 it implies that @xmath117 is monotonically decreasing on @xmath314 and the unique root @xmath315 must recover @xmath94 . otherwise ,
if @xmath316 , the secular function @xmath117 is singular at @xmath315 and there could be multiple global minimum solutions . in this case , let @xmath317 be the index such that @xmath318 .
the first order necessary condition @xmath319 can be solved by letting @xmath320 , where @xmath321 denotes the moore - penrose generalized inverse ; @xmath322 are free parameters and @xmath323 is the @xmath26-th column of @xmath31 .
then , we can establish the following _ generalized secular equation _
@xmath324 from which we try to find solution(s ) @xmath325 .
since the vector @xmath326 is perpendicular to each vector of @xmath327 , we have @xmath328 if @xmath329 and @xmath330 , there are exactly two global optimal solutions .
if @xmath331 and @xmath330 , there are infinitely many global solutions which form a @xmath225-dimensional sphere .
the result coincides with theorem 1 of part i. if @xmath332 , the optimal solution set degenerates to a singleton @xmath333 . in summary
, we provide three algorithms for finding the global minimizers , local non - global minimizer and local maximizer , respectively .
[ alg1 ] ( finding global minimizers ) + * solve the equation of one variable @xmath334^{-1 } \psi \|^2-t=0,~ t\in ( 2\nu-2\alpha_1,\infty).\ ] ] if there is a solution @xmath335 , stop ! the unique global minimizer of the double well potential problem ( [ d - well:4 ] ) is @xmath336^{-1}\psi.\ ] ] otherwise , go to step 2 . *
if @xmath63 and @xmath337=1 , solve equation ( [ secular - eq : root ] ) for at most two solutions : @xmath338 if @xmath339 , the double well potential problem ( [ d - well:4 ] ) has exactly two global minimizers of the form @xmath340 if @xmath341 , @xmath342 is the unique global minimizer . *
if @xmath343 , the double well potential problem ( [ d - well:4 ] ) has one or infinitely many global minimizers : @xmath344 where @xmath345 are obtained by solving ( [ secular - eq : root ] ) .
+ if @xmath346 @xmath342 is the unique optimal solution .
+ otherwise , the global optimal solutions form a sphere centered at @xmath347 with the radius of @xmath348 .
[ alg2 ] ( finding local non - global minimizer ) + solve the equation @xmath334^{-1 } \psi \|^2-t=0,~ t\in ( \max\{2\nu-2\alpha_2,0\},2\nu-2\alpha_1).\ ] ] if there is a solution @xmath154 such that @xmath349 , the unique local non - global minimizer of the double well potential problem ( [ d - well:4 ] ) is @xmath350^{-1}\psi.\ ] ] otherwise , declare that there is no local non - global minimizer .
[ alg3 ] ( finding the local maximizer ) + * if @xmath351 , declare that there is no local maximizer .
+ if @xmath352 and @xmath211 , then @xmath205 is the unique local maximizer .
+ otherwise , go to step 2 . *
solve the equation @xmath353^{+ } \psi \|^2-t = 0,~t\in[0 , 2\nu-2\alpha_n).\ ] ] if there is a solution @xmath354 such that @xmath355 , then the unique local maximizer of the double well potential problem ( [ d - well:4 ] ) is @xmath356^{+ } \psi.\ ] ] + otherwise , declare that there is no local maximizer . notice that each of the above three algorithms can be done in a polynomial time since
the main computation involved is to solve the secular equation in one variable . to illustrate their numerically behavior
, we use the same data set of ( @xmath357 ) of the three examples in part i of this paper and apply the space reduction in section 2 to convert the testing problems into the format of ( [ d - well:4 ] ) .
( example 1 of part i : ) let @xmath125 and @xmath358 the double well potential problem becomes @xmath359 the corresponding function @xmath90 is shown in figure 2 .
in this example , there are one global minimizer , one local non - global minimizer and one local maximizer .
the secular function @xmath360 is shown in figure 2 . by finding the root of ( [ m:1 ] ) in @xmath361 , algorithm [ alg1 ] provides a solution @xmath362 and we find the global minimizer @xmath363 with the value of @xmath364 . for the local non - global minimizer
, we apply algorithm [ alg2 ] to find the root of ( [ m:1 ] ) in @xmath365 .
algorithm [ alg2 ] returned @xmath366 with @xmath367 , which concluded that the local non - global minimizer is @xmath368 with the value of @xmath369 .
as for the local maximizer , by finding the root of ( [ m:1 ] ) in @xmath370 , algorithm [ alg3 ] returned @xmath371 with @xmath372 .
it led to the local maximizer @xmath373 with the value of @xmath374 .
notice that the signs of the two minimizers , @xmath363 and @xmath368 , are different , which demonstrates corollary [ cor:3 ] .
the numerical results also showed that the local maximizer @xmath373 locates between the two minimizers , which is claimed by theorem [ theorem4 ] .
[ fig:1 ] in example 1 ( @xmath125).,scaledwidth=50.0% ] [ fig:1:2 ] ) .,scaledwidth=50.0% ] ( example 2 of part i : ) applying the space reduction technique , we obtain the double well potential problem in the format of ( [ d - well:4 ] ) with the data @xmath375 and @xmath376,~ \psi = \left[\begin{array}{c } -22.0487\\-502.0209\end{array}\right].\ ] ] the corresponding function @xmath90 and its contour are shown in figure 4 .
its secular function becomes @xmath377 ( shown in figure 5 ) .
finding the root of ( [ m:2 ] ) on @xmath378 results in @xmath379 , algorithm [ alg1 ] gives the global minimizer @xmath380 $ ] with the value of @xmath381 .
similarly , finding the root of ( [ m:2 ] ) in @xmath382 results in @xmath383 .
since @xmath384 , algorithm [ alg2 ] provides the local non - global minimizer @xmath385 $ ] with the value of @xmath386 .
notice that the signs of the first component of the two minimizers are different , which demonstrates corollary [ cor:3 ] . finally , since @xmath387 , algorithm [ alg3 ] says that there is no local maximizer for this example . [ fig:2 ] in example 2 and its contour ( @xmath375).,title="fig:",scaledwidth=50.0% ] in example 2 and its contour ( @xmath375).,title="fig:",scaledwidth=50.0% ] [ fig:2:2 ] ) .,scaledwidth=50.0% ] ( maxican hat example ) in this example , @xmath388 which is already in the format of ( [ d - well:4 ] ) with @xmath375 , @xmath389,~ \psi = \left[\begin{array}{c } 0\\0\end{array}\right].\ ] ] the graph of the maxican hat function @xmath90 and its contour are shown in figure 6 .
since @xmath390 and @xmath391 , the secular function @xmath392 has a unique solution @xmath205 and it becomes singular at @xmath393 .
algorithm [ alg1 ] stopped at step 3 and claimed that @xmath394 is the set of global optimal solutions with the optimal value of @xmath205 .
since @xmath395 , algorithm [ alg2 ] returned an answer that there is no local non - global minimizer .
it is clear that ( [ m:3 ] ) has a unique root @xmath267 on @xmath396 and @xmath397 .
since @xmath211 , algorithm [ alg3 ] returned the unique local maximizer @xmath398 $ ] .
[ fig:3 ] in example 3 and its contour ( @xmath375).,title="fig:",scaledwidth=50.0% ] in example 3 and its contour ( @xmath375).,title="fig:",scaledwidth=50.0% ]
in this paper we have characterized the local minimizers and maximizers of the double well potential problem . by analyzing the first and the second order necessary conditions and through the study of the corresponding secular functions ,
we are able to estimate the number of local optimizers and locate each of them .
moreover , the convex secular functions ( equations ) are used to characterize sufficient and necessary conditions for all types of optimizers with explicit computational algorithms developed for finding them .
the ( dwp ) problem is a special case of the more general quadratic programming problem with one quadratic constraint ( qp1qc ) .
we expect that the analytical techniques developed in this paper can be extended to study ( qp1qc ) and other quadratic programming problems .
this research was undertaken while y. xia visited the national cheng kung university , tainan , taiwan .
sheu s research work was sponsored partially by taiwan nsc 98 - 2115-m-006 -010 -my2 and by national center for theoretic sciences ( the southern branch ) .
xia s research work was supported by national natural science foundation of china under grants 11001006 and 91130019/a011702 , and by the fund of state key laboratory of software development environment under grant sklsde-2013zx-13 .
xing s research work was supported by nsfc no . | in contrast to taking the dual approach for finding a global minimum solution of a double well potential function , in part ii of the paper , we characterize a local minimizer , local maximizer , and global minimizer directly from the primal side .
it is proven that , for a nonsingular " double well function , there exists at most one local , but non - global , minimizer and at most one local maximizer . moreover ,
when it exists , the local maximizer is surrounded " by local minimizers in the sense that the norm of the local maximizer is strictly less than that of any local minimizer .
we also establish some necessary and sufficient optimality conditions for the global minimizer , local non - global minimizer and local maximizer by studying a convex secular function over specific intervals .
these conditions lead to three algorithms for identifying different types of critical points of a given double well function .
* keywords * : double well potential , local minimizer , local maximizer , global minimum . |
one - dimensional ( 1d ) quantum spin systems have attracted much attention over the decades .
this is not only because these systems have been a good testing ground for various theoretical techniques and approximations but also because they exhibit a wealth of fascinating phenomena in their ground states and low - lying excitations .
these include quasi long - range - order ( lro ) , topological order and ground - state phase transitions , which are all purely quantum effects due to the low - dimensionality of the systems . among these quantum phenomena , the effects of randomness on quantum spin systems have been studied intensively by many groups .
these studies revealed the appearance of various exotic phases , which are realized neither in regular quantum systems nor classical random systems .
the interplay of randomness and quantum fluctuations plays an essential role in these phases .
systems which are gapless in the absence of randomness are unstable against weak randomness,@xcite while gapped systems , e.g. integer spin chains in the haldane phase and dimerized chains , are comparatively robust.@xcite a simplest model of the 1d random spin-@xmath0 heisenberg spin systems is described by the hamiltonian of the form @xmath3 where @xmath4 are @xmath5 spin operators and the exchange coupling constants @xmath6 are distributed randomly according to a probability distribution @xmath7 .
there are several quasi-1d systems which provide the realizations of the model hamiltonian ( [ eq : ham ] ) .
to our knowledge , the first example of such systems belongs to the class of the organic charge - transfer salts tetracyanoquinodimethanide ( tcnq).@xcite the low - temperature magnetic properties of these systems are successfully described by the model hamiltonian ( [ eq : ham ] ) with random antiferromagnetic ( af ) coupling,@xcite in which the couplings @xmath6 are restricted to take positive random values .
a more recently studied system@xcite is sr@xmath8cupt@xmath9ir@xmath10o@xmath11 .
while the pure compounds sr@xmath8cupto@xmath11 and sr@xmath8cuiro@xmath11 represent , respectively , af and ferromagnetic ( fm ) spin chains , sr@xmath8cupt@xmath9ir@xmath10o@xmath11 contains both af and fm couplings .
the fraction of fm bonds is simply related to the concentration @xmath12 of ir .
these compounds are modeled adequately by the hamiltonian ( [ eq : ham ] ) with @xmath13 , where @xmath14 is the probability of fm bonds .
the experimental data of the susceptibility of the compounds were in fact explained successfully by a theory based on the hamiltonian ( [ eq : ham ] ) with the above probability distribution of fm and af bonds.@xcite the model ( [ eq : ham ] ) is also realized in the low - temperature regime of randomly depleted 1d spin-@xmath0 even - leg ladders.@xcite in these systems , effective @xmath5 spins are induced in the vicinity of each depleted site.@xcite if the density of depleted spins is sufficiently small , the induced effective spins are the only degrees of freedom which are active in the low - energy limit .
the strength of the residual interaction between effective spins depends on the distance between the effective spins , and the coupling can be either antiferromagnetic or ferromagnetic , depending on whether or not two spins are depleted from the same sublattice .
thus , the low - energy physics of the randomly depleted spin-@xmath0 even - leg ladders is described as a random spin chain with both af and fm exchange couplings.@xcite for 1d random spin chains containing only af couplings , rather complete theoretical understanding on their low - energy properties has been achieved .
one of the most powerful techniques to study such systems is the real - space renormalization group ( rsrg ) method introduced by ma , dasgupta , and hu.@xcite the basic idea of this method is iterative decimation of spin degrees of freedom by integrating out the strongest bonds in the chain successively . for the random af chains this procedure keeps the form of the hamiltonian ( [ eq : ham ] ) and renormalizes the probability distribution @xmath7 .
fisher has given a solution to this rg equation of @xmath7 that becomes exact in the low - energy limit and shown that any normalizable initial distribution @xmath7 flows to a single universal fixed point distribution.@xcite the ground state of the chain which belongs to the fixed point is characterized by the `` random singlet phase , " where each spin forms a singlet with another spin which can be arbitrarily far away . from the intuitive picture of the random singlet phase , fisher has also shown that the random average of the two - spin correlation function in this phase decays algebraically like @xmath1 with @xmath15 being the distance between the two spins .
the main contribution to the average comes from the rare events that two spins separated by the distance @xmath15 form a singlet pair .
the @xmath1 power - law decay has been verified by a numerical calculation for the random @xmath16 spin-@xmath0 chains,@xcite which can be mapped to a disordered system of noninteracting fermions .
the random spin chains containing both af and fm couplings have also been studied theoretically for the last five years , from which a qualitative picture of the low - energy physics of such random fm - af chains has emerged .
et al . _
@xcite have adapted the rsrg scheme and shown through extensive numerical simulations that the distribution @xmath7 is renormalized to a single universal fixed - point distribution unless the initial distribution is highly singular around @xmath17 .
the chain at this fixed point can be viewed as an ensemble of weakly interacting large effective spins whose average size @xmath18 with @xmath19 for temperature @xmath20 .
these large effective spins are generated as a result of decimations of two spins coupled via a strong fm coupling into a larger spin .
the results of the rsrg simulations are supported by a recent calculation by frischmuth _
et al_. , @xcite who have used the continuous - time quantum monte carlo loop algorithm .
in contrast to the success in the quantitative calculations on the thermodynamic properties , less is known about the spin correlations in the random fm - af chains .
since the average size of the effective spins becomes very large in the low - energy ( long - distance ) limit , one can expect that the system should be close to a classical spin chain that can show the lro of the generalized staggered spin component .
one might also argue , however , that there should be no lro in 1d quantum spin chains .
it is therefore an interesting open question how the correlation function of the generalized staggered moment ( whose definition will be given in sec .
iii ) behaves in the ground state .
the purpose of this paper is in fact to present results of our extensive numerical calculation of the two - spin correlation function .
we find that it decays very slowly with the form close to @xmath2 , which is consistent with the naive argument that the ground state is extremely long - range correlated , but not really long - range ordered .
the outline of this paper is as follows .
section [ sec : conventionalrsrg ] is devoted to a brief review of the rsrg scheme of the generalized version which is applicable to both random af and fm - af case . using the wigner - eckart theorem,@xcite
we simplify the algorithm to allow for calculating the two - spin correlation function between original @xmath21 spins . in section [ sec : extension ] , in order to achieve higher accuracy , we extended the rsrg scheme to take into account the contributions of local excitations to the ground state of the whole system .
we perform these conventional and extended rsrg algorithm numerically on both random af and fm - af chains to calculate the `` staggered " spin correlations on their ground states .
the results on the random af case are shown in section [ sec : af ] . in section [ sec : fmaf ] , we show the results on the random fm - af case . analyzing the data obtained with the extended rsrg method ( which can be applied to larger systems than the dmrg ) ,
we conclude that the mean correlations on the random fm - af case decay very slowly with logarithmic dependence on @xmath15 .
we also discuss the distribution functions of the logarithm of the correlation functions in section [ sec : dis ] . we find that , in the random af case , the rare spin pairs , which are strongly correlated , dominate the mean correlation while such rare events are not essential in the random fm - af case .
finally , our results are summarized in section [ sec : conclusion ] .
ma , dasgupta and hu introduced a rsrg procedure to investigate the low - temperature properties of random af spin chains.@xcite the method has been generalized to the random fm - af case by westerberg _
et al_.@xcite in this section we explain our extension of this scheme to calculate the ground - state two - spin correlation functions .
we begin with a brief review of the rsrg method for the random fm - af case .
let us consider a random fm - af spin-@xmath0 chain described by the hamiltonian ( [ eq : ham ] ) .
the basic strategy of rsrg is to decimate spin degrees of freedom by combining two spins connected via a strong bond into one effective spin .
consequently , the system is described in terms of effective spins of various sizes coupled by random exchange interactions , although the original hamiltonian ( [ eq : ham ] ) consists of only @xmath22 spins .
we accordingly treat the effective hamiltonian , @xmath23 where both the coupling @xmath24 and the size of the effective spins @xmath25 are random .
we call the @xmath22 spins @xmath4 appearing in the hamiltonian ( [ eq : ham ] ) `` original " spins in the following to distinguish them from the effective spins @xmath26 .
we define @xmath27 as the energy gap between the ground - state multiplet and the first excited multiplet of the corresponding bond hamiltonian @xmath28 , @xmath29 and focus on the bond with the largest gap @xmath27 in the chain . the terms in the effective hamiltonian ( [ eq : heff ] ) which involve the effective spins @xmath26 and @xmath30 are @xmath31 where @xmath32
if @xmath27 is much larger than the gaps of the neighboring bonds , @xmath33 and @xmath34 , the spins @xmath26 and @xmath30 , to a good approximation , are frozen into the ground - state multiplet of the local hamiltonian @xmath35 .
we , therefore , replace the block composed of @xmath26 and @xmath30 by the single effective spin @xmath36 .
the wigner - eckart theorem @xcite then implies that both @xmath26 and @xmath30 are proportional to @xmath36 : @xmath37 where @xmath38 and @xmath39 can be obtained from the clebsch - gordan coefficients .
substituting eq .
( [ eq : we ] ) into the four - spin hamiltonian ( [ eq : h ] ) , we obtain , apart from a constant term coming from @xmath35 , @xmath40 where @xmath41 the case where @xmath24 is antiferromagnetic with @xmath42 needs a special treatment . in this case the two spins @xmath26 and @xmath30 form a singlet , and accordingly , we remove both spins from the effective hamiltonian . between the spins @xmath43 and @xmath44 ,
a coupling is generated through a second - order process that virtually breaks the singlet of @xmath26 and @xmath30 .
we obtain @xmath45 where @xmath46 by replacing the four - spin hamiltonian ( [ eq : h ] ) with @xmath47 [ eq . ( [ eq : hnew ] ) or eq .
( [ eq : hnew ] ) ] in the hamiltonian of the whole system , we obtain a new effective hamiltonian ( see fig . [
fig : conrsrg ] ) .
we note that this procedure preserves the form of the effective hamiltonian ( [ eq : heff ] ) but changes the distributions of the exchange couplings , @xmath24 , and the size of effective spins , @xmath25 .
we repeat this procedure of integrating out the strongest bonds in a chain successively until the distribution functions for @xmath24 and @xmath25 converge to a scaling form .
this rg flow has been investigated extensively for various initial distributions including both af and fm - af case .
@xcite in particular , for the random af @xmath22 chains fisher has solved the rg equation exactly .
@xcite it is also possible to calculate the correlation functions between original spins within the scheme described above .
here we make use of the fact that the original spin operator @xmath4 is , at every step in the rsrg procedure , proportional to the effective spin @xmath26 to which it belongs .
we can keep track of the coefficient for each original spin operator by applying eq .
( [ eq : we ] ) at each step . at the step
where the effective spins @xmath26 and @xmath30 is added into @xmath36 , we calculate the correlations between @xmath4 and @xmath48 in the ground state of the bond hamiltonian @xmath35 , @xmath49 where @xmath50 represents the expectation value in the ground state ; @xmath4 and @xmath48 belong to @xmath26 and @xmath30 , respectively ; @xmath51 and @xmath52 are the proportionality coefficients for each spin .
as seen in the previous subsection , the conventional rsrg procedure consists of `` diagonalizing the bond hamiltonian with the largest gap `` and `` projecting the low - energy states onto the lowest multiplet . ''
this means that we completely neglect the contribution of the excited multiplets of the local hamiltonian ( [ eq : h0 ] ) to the ground state of the whole system .
this approximation is valid only if the energy gap of @xmath35 is much larger than the ones of the neighboring bonds .
fisher s solution for the random af @xmath22 chain becomes asymptotically exact since this condition is satisfied near the fixed point of the rg flow .
however , the condition is often not satisfied , in particular , in the early stage of the rg , unless the initial distribution of energy gap is very broad . as a result
, we have no choice but to cut off the contributions of local excitations .
this is a poor approximation that has a serious effect especially on the calculation of the expectation values of microscopic operators such as the original spins .
the ground - state correlation functions between original spins calculated via the conventional rsrg in fact deviate largely from those obtained from the dmrg method and the extended rsrg algorithm described below ( see fig .
[ fig : afc ] and fig .
[ fig : fmc ] in section [ sec : results ] ) .
one possible prescription to avoid this error is `` to keep the local multiplet excitations " at each step .
of course , if we keep all eigenstates of the bond hamiltonian @xmath35 at every step of the rg , the calculation on @xmath35 in the final step is equivalent to the exact diagonalization on the hamiltonian of the whole system . in practice
what we should do is to extend the rsrg algorithm under the policy that we keep as many states as we can store in computer memories . in the original rsrg scheme a segment of original spins
are combined and represented by a single effective spin . in the extended scheme
we keep more states than the lowest multiplet and call the segment a ` block. each block consists of several original spins and is represented by ` block states, which are the @xmath53-lowest eigenstates of the block hamiltonian .
since we have to keep or discard all states of a multiplet to ensure the su(2 ) symmetry , the actual number of kept states for block @xmath54 is @xmath55 .
an original spin operator in block @xmath54 is represented by @xmath56 matrix on the set of the block states accordingly .
let us consider the effective hamiltonian @xmath57 where @xmath58 is a block hamiltonian of the @xmath54th block , diagonal in the block states ; @xmath59 is a coupling hamiltonian between the @xmath54th and @xmath60th blocks ; @xmath61 and @xmath62 are original spin operators on the right and left edge of the @xmath54th block , respectively ; @xmath63 is a coupling between @xmath61 and @xmath64 . in the extended rsrg scheme we renormalize the two - block hamiltonian @xmath65 with the largest gap @xmath27 into one block hamiltonian ( see fig .
[ fig : extrsrg ] ) . here
we define @xmath27 as the energy difference between the highest energy in the eigen - multiplets of @xmath66 which will be kept and the lowest energy in the multiplets which will be discarded after the decimation .
the basic scheme of the extended rsrg is the same as the conventional rsrg except the changes described above .
the algorithm is summarized as follows : 1 .
focus on the bond with the largest gap @xmath27 .
construct the two - block hamiltonian ( [ eq:2bh ] ) of the bond .
2 . diagonalize the two - block hamiltonian to find a set of eigenvalues and eigenstates .
at this point , we can calculate expectation values of various operators in the two blocks , such as the two - spin correlation functions between the original spins , using the ground state of the two - block hamiltonian .
3 . discard all but the lowest @xmath67 eigenstates in the block hamiltonian .
4 . express operators , such as the block hamiltonian itself and the original spin operators in the new block , in terms of the new block states .
rewrite the coupling hamiltonians between the new block and its neighboring blocks in terms of the new @xmath68 and @xmath69 . diagonalize them to update the distribution of the energy gap @xmath70 .
return to the step ( i ) .
we obtain all two - spin correlation functions by repeating this procedure until the whole chain is finally represented by one block .
using both the conventional rsrg and the extended rsrg with various values of @xmath53 , we calculated the correlation functions @xmath71 in the ground state of open chains for both the random af case and the random fm - af case .
the maximum size of the chains used for the conventional and the extended rsrg simulations is @xmath72 and @xmath73 , respectively . in the fm - af case ,
the random average of the spin correlation , @xmath71 , always decays exponentially@xcite and is not an appropriate quantity to characterize the spin correlation because the sign of @xmath71 can be either positive or negative depending on the number of af bonds between @xmath4 and @xmath48 .
instead we introduce the `` generalized staggered " correlation function @xmath74 where @xmath75 for @xmath76 . for random af case , @xmath77 is the usual staggered spin correlation .
we take the random average of @xmath77 and @xmath78 , which represents the mean correlations and the logarithm of the typical correlations , respectively .
note that it is impossible to take the random average of @xmath78 numerically in the conventional rsrg algorithm , within which @xmath79 for two spins that do not form a singlet pair . to check the results of the rsrg methods
, we also calculated the correlation functions on @xmath80 open chains using the dmrg method @xcite with the improved algorithm proposed by white .
@xcite the number of kept states in the dmrg calculation was up to 100 and 200 for the random af and fm - af case , respectively . in both cases ,
the mean and typical correlations calculated by the extended rsrg are in excellent agreement with those by dmrg ( see below ) .
we note that the systems we have treated are much bigger than those in the earlier work by hida,@xcite in which the dmrg method is applied for the fm - af chains . as a typical probability distribution of the random af case ,
we choose the box - type bond distribution @xmath7 , @xmath81 where the cutoff , @xmath82 , is taken as energy unit . for the random af case
, it is known that any normalizable initial distribution flows to a single universal fixed point .
@xcite hence , the results we obtained for the initial distribution ( [ p(j);af ] ) are generic .
the number of sample chains we used for each method are shown in table [ tb1 ] . before discussing our numerical results , we briefly comment on the @xmath53-dependence of the data of the extended rsrg . as noted in the last section
, the extended rsrg can provide more accurate results as the number of kept states , @xmath53 , increases . in the random af case ,
the ground - state multiplet of a block hamiltonian is either singlet or doublet , depending on the number of original spins belonging to the block .
the degeneracy of each low - lying excited multiplet is , therefore , always small . as a result
, we can keep considerably many multiplets even if @xmath53 is rather small .
we estimate the @xmath53-dependence of the results of the extended rsrg from the numerical data with @xmath83 .
figures [ fig : afc ] and [ fig : aflnc ] show , respectively , the numerical data of the mean correlations , @xmath84 , and the average of the logarithm of correlations , @xmath85 , where @xmath86 represents random average .
it is clear that the data of the extended rsrg converge already at @xmath87 , and we can regard those data with @xmath87 as those in the limit @xmath88 . in fig .
[ fig : afc ] the data of the extended rsrg with @xmath87 are in good agreement with those of dmrg for @xmath89 , where the dmrg data are considered to be exact and free from finite - size effects . on the other hand
, the conventional rsrg largely underestimates the correlations , but its data are on a line parallel to the data of the extended rsrg in the log - log plot .
we conclude from these observations that the results of the extended rsrg are quantitatively reliable , whereas the conventional rsrg can be used to estimate the exponent of the power - law decay .
encouraged by the agreement between the extended rsrg and the dmrg data , we anticipate that the extended rsrg provide quantitatively reliable data even for @xmath90 where the reliable dmrg data are not available . from the data of the extended rsrg with @xmath91 for @xmath92 , where the data are expected not to be hampered by finite - size effects
, we estimate the asymptotic form of the average correlations to be @xmath93 for the average of @xmath78 , we also rely on the data obtained from the extended rsrg scheme .
figure [ fig : aflnc ] gives @xmath94 both results ( [ eq : fs1 ] ) and ( [ eq : fs2 ] ) agree with fisher s theory @xcite and can be considered as a numerical verification of his solution on the mean and typical correlations for the heisenberg case .
for the random @xmath16 chains@xcite and for a related model of the random transverse - field ising model,@xcite the power - law behavior ( [ eq : fs1 ] ) and ( [ eq : fs2 ] ) is observed numerically . to our knowledge ,
[ fig : afc ] and [ fig : aflnc ] are first numerical results confirming fisher s theory for the heisenberg case .
westerberg _ et al_. @xcite showed that the rg trajectories of the random fm - af chains flow towards a single universal fixed point under the conventional rsrg procedure , unless the initial distribution of couplings is more singular than @xmath95 , @xmath96 . in this section
we investigate the spin correlations at this fixed point with the extended rsrg scheme . for this purpose
we assume the box - type bond distribution @xmath7 , @xmath97 with @xmath82 as energy unit , as a representative of distributions with no singularity at @xmath98 .
we expect that our results obtained for the initial distribution ( [ eq : p ] ) should capture the universal behavior for the random fm - af chains which are in the basin of the universal fixed point found in ref . .
we have numerically calculated spin correlations using the conventional rsrg , the extended rsrg with @xmath99 and the dmrg method . in the random fm - af chains the degeneracy of the lowest multiplet of a block becomes larger on average as the size of the block grows . at the point when the degeneracy exceeds the number of kept states , @xmath53 , determined beforehand in the algorithm , the extended rsrg scheme breaks down because we need to keep all the degenerate states in the lowest multiplet to preserve the su(2 ) spin symmetry .
in fact , we could complete the extended rsrg procedure without exceeding the limit of the number of kept states @xmath87 ( @xmath100 ) for 30% ( 75% ) of the sample chains ( @xmath101 ) .
we then used only those samples for which the rsrg could be completed to take the random average .
although this sorting out of sample chains may lead to a systematic underestimate on the average values , we believe that we can correct it by carefully checking the @xmath53-dependence of the data .
the number of sample chains we used to take the random average for each rg scheme is shown in table [ tb2 ] .
the numerical results for the mean correlations , @xmath102 , are shown in fig .
[ fig : fmc ] in a log - log plot .
it is clear that the data of the extended rsrg have almost converged with @xmath100 for @xmath103 , where the data are expected to be free from the effect of the open boundaries .
the data with @xmath104 are also in excellent agreement with the data of dmrg .
we , therefore , regard the results of the mean correlation with @xmath105 as essentially converged .
we notice that the curve in the log - log plot are bent upward , indicating that the mean correlations @xmath106 decay more slowly than a power - law . to analyze the @xmath15-dependence of the mean correlations , we plot the inverse of @xmath102 as a function of @xmath107 in the inset of fig .
[ fig : fmc ] .
we find that the data lie on a straight line in this plot , from which we speculate that the mean correlations decay with the logarithmic form , @xmath108 where @xmath109 and @xmath110 are constants .
we note , however , that eq . (
[ eq : log ] ) is not the only form that can account for the numerical data.@xcite in any case the mean correlations show very weak @xmath15-dependence , probably through the form of @xmath111 , certainly different from power - law .
figure [ fig : fmlnc ] shows the numerical results of @xmath85 .
the data of the extended rsrg with @xmath105 exhibit a similar behavior to the logarithm of the mean correlations ; the typical correlation @xmath112 $ ] decays again more slowly than a power - law .
this leads us to plot the inverse of the typical correlations as a function of @xmath107 ( see the inset of fig . [
fig : fmlnc ] ) .
although the curve at @xmath105 seems almost linear for @xmath113 , we can not determine the @xmath15-dependence of the typical correlations from the figure due to the slow convergence of the data with increasing @xmath53 .
this slow convergence arises from the fact that the numerical estimate of @xmath78 is very sensitive to small fluctuations of @xmath77 especially when the value of @xmath77 is extremely small .
calculations with much larger values of @xmath53 would be needed to obtain the data accurate enough to determine the @xmath15-dependence of @xmath85 . in order to make a distinction between characteristics of ground - state correlation functions in the random af chains and in the random fm - af chains , we analyze the distributions of the logarithm of the correlations , @xmath114 , where @xmath115 .
henelius and girvin@xcite have shown numerically that for large @xmath15 the distribution function of @xmath16 chains with random af couplings scales to a fixed - point distribution of the form @xmath116 with @xmath117 where @xmath118 is a positive constant .
the scaling function @xmath119 satisfies the normalization condition @xmath120 .
we will demonstrate that the distribution @xmath114 in random heisenberg chains also exhibits the scaling behavior eq .
( [ eq : scl ] ) for both the random af and the random fm - af case .
figure [ fig : asc ] shows the data of @xmath114 for the random af case obtained using the extended rsrg with @xmath91 . according to eqs .
( [ eq : fs2 ] ) and ( [ eq : scl2 ] ) , we can set @xmath121 and @xmath122 .
the data points ( circles and squares ) in fig .
[ fig : asc ] collapse on a single curve , indicating that the scaling ( [ eq : scl ] ) applies.@xcite in fig .
[ fig : asc ] we also plot @xmath123 which measures the contribution to the mean value of the correlation @xmath124 .
although the curves are rather rough due to a statistical error , it is clear that @xmath125 has a considerable weight in the range where @xmath114 is very small .
this means that a very few spin pairs that have much stronger correlations than typical ones give dominant contribution to the mean correlation @xmath124 .
we also find that , as @xmath15 increases , the region of @xmath126 where @xmath125 is large moves towards @xmath127 .
indeed this is the behavior expected from eq .
( [ eq : fs1 ] ) ; the value of @xmath126 where @xmath125 takes a large weight should change as @xmath128 as @xmath129 .
thus we regard the results shown in fig .
[ fig : asc ] as a further support for the random - singlet picture of the ground state of the random af heisenberg chains , where each spin forms a singlet pair with another spin that can be arbitrarily far away .
the mean correlation @xmath124 is dominated by the contribution from the rare case in which two spins separated by distance @xmath15 form a spin singlet .
figure [ fig : fsc ] shows the numerical data of @xmath114 of the extended rsrg with @xmath105 for the random fm - af chains . here
we set @xmath130 and @xmath131 .
the data points ( circles and squares ) for @xmath132 , where we may ignore the boundary effect , clearly lie on a single scaling curve . in this range of @xmath15
, however , @xmath133 changes by several percent only , from @xmath134 to @xmath135 . to establish the scaling behavior for wide range of @xmath133
, the calculations on much ( exponentially ) larger systems might be necessary .
such large - scale calculations are obviously unfeasible with computers available at present , and thus we can only conclude that our results shown in fig . [
fig : fsc ] are consistent with the scaling hypothesis ( [ eq : scl ] ) . in the following discussion
, we regard @xmath136 in fig .
[ fig : fsc ] as a fixed - point form of the distribution function .
figure [ fig : fsc ] clearly shows that the distribution function of the random fm - af chains has a quite different form from that of the random af chains .
it is essentially zero for @xmath137 and increases approximately linearly at @xmath138 . the weight function @xmath125 representing the contribution to the mean correlation is also shown in fig .
[ fig : fsc ] .
in contrast to the random af case , @xmath125 has most of its weight in the region where @xmath114 is not negligible .
this feature highlights the different nature of the spin correlations in the random fm - af chains .
as shown in the rsrg analysis , many spins in the random fm - af chains correlate and form a large effective spin .
this suggests that a spin is correlated with many other spins that belong to the same large effective spin , and , therefore , the mean value of the spin correlation function is not at all dominated by the `` rare events that two spins , far apart from each other , form a singlet pair . from the observation that @xmath136 of the random fm - af chains is negligible for @xmath139 ( @xmath140 in fig .
[ fig : fsc ] ) and has an approximately linear dependence for @xmath141 until it takes a maximum value , we can also draw a conclusion that @xmath142 and @xmath143 should have a similar dependence on @xmath15 , in agreement with figs .
[ fig : fmc ] and [ fig : fmlnc ] .
let us assume that the scaling function has the form @xmath144 where @xmath145 , @xmath146 and @xmath147 are positive constants . since @xmath148 in the limit @xmath129 , the mean correlation @xmath102 is dominated by the contribution from the region of small @xmath149 .
thus , we are allowed to use eq .
( [ eq : f ] ) for calculating @xmath102 , even though this form is not correct for large @xmath149 .
using eqs .
( [ eq : scl ] ) , ( [ eq : scl1 ] ) and ( [ eq : f ] ) , we calculate the mean correlation : @xmath150 ^ 2 } e^{-a g(r)},\ ] ] yielding @xmath151 in the limit @xmath152 .
this result is in agreement with our observation that the curves in figs .
[ fig : fmc ] and [ fig : fmlnc ] look very similar .
it is also consistent with the fact that @xmath125 in fig .
[ fig : fsc ] stays almost in the same region of @xmath126 with increasing @xmath15 .
equation ( [ eq : ln = ln ] ) is not changed qualitatively even when @xmath153 has an algebraic form @xmath154^\alpha$ ] , as far as @xmath155 .
note that eq .
( [ eq : ln = ln ] ) does not hold in the random af chains , for which we should take @xmath156 and @xmath157 .
we have studied the spin correlations in spin-@xmath0 random heisenberg chains for both the random af and the random fm - af case .
one of the most powerful methods for the study of random spin chains is the real - space rg in which the spin degrees of freedom are decimated by integrating out the strongest bonds successively . in order to calculate the two - spin correlation functions between original @xmath21 spins
, we modified the algorithm of the conventional rsrg using the wigner - eckart theorem .
we also extended the rsrg scheme by keeping the local excited multiplets as block states to keep track of quantum effects .
we demonstrated that the extended rsrg algorithm is very powerful , in the sense that it can be applied to rather large systems , and provides accurate numerical data that are in excellent agreement with those of the dmrg method .
the numerical data of the mean and typical correlations for the random af chains verified fisher s prediction [ eqs .
( [ eq : fs1 ] ) and ( [ eq : fs2 ] ) ] for the heisenberg case , which can not be mapped to noninteracting fermions . for the random fm - af chains
, we found that the mean correlation , @xmath102 , decays very slowly with the logarithmic @xmath15-dependence .
we , therefore , conclude that the generalized staggered spin correlation in the random fm - af chains has no lro , although it shows very long - range correlations decaying slower than power - law behavior .
investigating the distribution of the logarithm of the spin correlation functions , we also clarified the different nature of the ground states of the random af and fm - af chains . in both cases , our data of distributions
satisfy the scaling hypothesis , eq .
( [ eq : scl ] ) . analyzing the form of the scaling function
, we found that the `` rare spin pairs `` , which are correlated much stronger than typical ones , dominate the mean correlation @xmath102 in the random af case .
it is also shown that the @xmath15-dependence of the mean correlations in this case is quite different from that of the typical ones .
these features strongly support the fact that the `` random singlet phase '' is realized in the ground state of the random af heisenberg chains . on the other hand ,
the scaling form of the distribution functions for the random fm - af chains suggests that such `` rare spin pairs " do not play an important role in these chains .
we also deduced that @xmath158 and @xmath85 have very similar @xmath15-dependence , using the scaling hypothesis and the specific feature of the scaling function we obtained numerically .
this result is non - trivial and consistent with our numerical data ( figs .
[ fig : fmc ] and [ fig : fmlnc ] ) .
ht was supported by the research fellowship of the japan society for the promotion of science for young scientists .
the work of af and ms was in part supported by a grant - in - aid from the ministry of science , education and culture of japan .
numerical calculations were carried out at the yukawa institute computing facility .
r. a. hyman , kun yang , r. n. bhatt , and s. m. girvin , phys .
* 76 * , 839 ( 1996 ) ; r. a. hyman and kun yang , phys .
lett . * 78 * , 1783 ( 1997 ) ; kun yang and r. n. bhatt , phys . rev
. lett . * 80 * , 4562 ( 1998 ) .
l. n. bulaevskii , a. v. zvarykina , yu .
s. karimov , r. b. lyuubovskii , and i. f. shchegolev , zh .
. fiz . * 62 * , 725 ( 1972 ) [ sov .
jetp * 35 * , 384 ( 1972 ) ; l. j. azevedo and w. g. clark , phys . rev .
b * 16 * , 3252 ( 1977 ) .
a. furusaki , m. sigrist , p. a. lee , k. tanaka , and n. nagaosa , phys .
* 73 * , 2622 ( 1994 ) ; a. furusaki , m. sigrist , e. westerberg , p. a. lee , k. b. tanaka , and n. nagaosa , phys . rev .
b * 52 * , 15930 ( 1995 )
. some data points at @xmath163 are off the scaling curve in fig .
[ fig : asc ] .
we estimate that the numerical error in @xmath77 in our calculation is of order @xmath164 .
the scattered data are those with small correlations of order @xmath165 .
thus the failure of the data collapse at @xmath163 does not mean the violation of the scaling . | we calculate the ground - state two - spin correlation functions of spin-@xmath0 quantum heisenberg chains with random exchange couplings using the real - space renormalization group scheme .
we extend the conventional scheme to take account of the contribution of local higher multiplet excitations in each decimation step .
this extended scheme can provide highly accurate numerical data for large systems .
the random average of staggered spin correlations of the chains with random antiferromagnetic ( af ) couplings shows algebraic decay like @xmath1 , which verifies the fisher s analytic results . for chains with random ferromagnetic ( fm ) and af couplings , the random average of generalized staggered correlations
is found to decay more slowly than a power - law , in the form close to @xmath2 .
the difference between the distribution functions of the spin correlations of the random af chains and of the random fm - af chains is also discussed . |
fredholm matrices appear naturally in graph theory .
they arise most prominently in the * chebotarev - shamis forest theorem * @xcite which tells that @xmath26 is the number of rooted forests in a graph @xmath0 , if @xmath27 is the kirchhoff laplacian of @xmath0 .
this forest theorem follows readily from the * generalized cauchy - binet formula * @xcite @xmath28 which holds for any pair @xmath29 of @xmath30 matrices and were the right hand side is a dot product of the minor vector giving all possible minors of the matrix , defined by the index set @xmath31 which can be empty in which case @xmath32 .
the forest theorem uses this with @xmath33 , where @xmath34 and @xmath35 are poincar s incidence matrices .
the fredholm determinant @xmath26 is then @xmath36 , which directly counts rooted forests .
+ similarly , any fredholm determinant @xmath37 of a matrix @xmath1 can be written as @xmath38 which implies the well known formula @xmath39 where @xmath40 is the @xmath41th exterior power of @xmath1 . that expansion is at the heart of extending determinants to fredholm determinants in infinite dimensions , in particular if @xmath1 is trace class @xcite . in this paper
, we look at the adjacency matrix @xmath1 , which unlike the laplacian @xmath27 of a graph is not positive semi - definite so that one can not write @xmath42 for some other matrix . indeed , the determinant of an adjacency matrix @xmath43 of a graph @xmath0 appears rather arbitrary in the strong sense that experiments show that a limiting distribution emerges when looking at the random variables @xmath44 on probability spaces of graphs .
see figure ( [ distribution ] ) .
+ given a finite simple graph @xmath0 with adjacency matrix @xmath1 , we call @xmath45 the * fredholm adjacency matrix * and @xmath46 the * fredholm determinant * of @xmath0
. the fredholm determinant of a graph can be pretty arbitrary as the following examples show : for complete graphs @xmath47 , where the eigenvalues of the adjacency matrix are @xmath48 with multiplicity @xmath49 and @xmath49 with multiplicity @xmath50 and the fredholm determinant is @xmath21 for @xmath51 . for cyclic graphs
@xmath52 , the fredholm determinant is @xmath53-periodic in @xmath49 , zero for @xmath54 , and @xmath55 for @xmath56 not divisible by @xmath55 and @xmath57 for @xmath58 not divisible by @xmath53 . for wheel graphs
@xmath59 with @xmath60 vertices , it is @xmath61 if @xmath49 is not divisible by @xmath55 and @xmath21 else .
measuring the statistical distribution of fredholm determinant on classes of random graphs suggests that the normalized distribution of the adjacency or fredholm determinants produces an absolutely continuous limit with singularity .
also this appears not yet explored theoretically but it illustrates that the fredholm determinant of a general graph can be pretty arbitrary .
+ we can think of @xmath62 as a partition function or a path integral `` , in which the underlying paths are fixed - point - free permutations of the vertices of the graph .
the determinant generates therefore * derangements * and @xmath63 defines the oriented paths . on the other hand ,
the fredholm determinant @xmath46 is a partition function for all oriented paths in the graph as @xmath64 can now also have pairs @xmath65 as transpositions and vertices @xmath66 as fixed points .
one can see the effect of changing from determinants to fredholm determinants well when replacing the determinants with the * permanent * , the bosonic analogue of the determinant : @xmath67 is the number * derangements * of the vertex set of the graph while @xmath68 is the number of all * permutations * of the vertex set honoring the connections . for the complete graphs @xmath69 in particular , @xmath70 generates the * permutation sequence * @xmath71 while the permanent of the adjacency matrix @xmath72 generates the * derangement sequence * @xmath73 .
it is therefore not surprising that fredholm determinants are natural .
+ the distribution of fredholm determinants changes drastically if we evaluate them on the set of * connection graphs * , graphs which have the set of simplices of a graph @xmath0 as subgraphs and where two simplices are connected if they intersect .
connection graphs are also defined for abstract finite simplicial complexes or even finite cw complexes .
if we talk about a graph @xmath0 , we usually understand it equipped with the whitney complex , the set of complete subgraphs of @xmath0 .
but any simplicial complex structure or cw complex structure on the graph works . it does not even have to come from simplices .
the * graphic matroid * of a graph is an example where the connection graph has forests in @xmath0 as vertices and has two forests connected if some trees in it share a common branch .
+ for the smaller * barycentric refinement * @xmath74 of @xmath0 , two simplices are connected only if and only if one is contained in the other .
the graph @xmath74 has the same vertices than @xmath75 but is a subgraph of @xmath75 .
connection graphs are in general much higher dimensional than the graph @xmath0 or even the barycentric refinement @xmath74 : for a triangle @xmath76 already , @xmath75 a graph which contains the complete graph @xmath77 .
small spheres like the octahedron are examples where @xmath0 and @xmath75 are not homotopic because @xmath75 has euler characteristic @xmath21 while the octahedron , as a 2-sphere has euler characteristic @xmath22 .
however this only is the case because the sphere is too small . for the icosahedron @xmath0 already , the connection graph @xmath75 a two sphere again .
in general , the barycentric refinement @xmath74 of a graph @xmath0 has a connection graph @xmath78 which is homotopic to @xmath74 and so to @xmath0 .
this implies then that all cohomology groups of @xmath0 and @xmath78 agree .
while @xmath0 and @xmath74 are not homeomorphic as already their dimension is different in general , they can be useful in geometry , like barycentric refinements .
they can be used for example to * regularize singularities * as they ' ' homotopically fatten " the discrete manifolds " or simplicial complexes and still have the same homotopy type after applying one barycentric refinement . + we got interested in connection graphs in the context of connection calculus `` , a calculus where differential forms are not functions on simplices but on pairs or @xmath41-tuples of connecting simplices in the simplicial complex .
the corresponding cohomology is compatible with calculus in the sense that common theorems like gauss - bonnet @xcite , poincar - hopf @xcite , euler - poincar or kuenneth @xcite or brouwer - lefschetz fixed point theorem @xcite generalize when euler characteristic is replaced by * wu characteristic * but for which the cohomology is finer .
the cohomology already allows to distinguish spaces which classical simplicial cohomology can not , like the mbius strip and the cylinder @xcite .
+ the bowen - lanford * zeta function * @xcite of a graph @xmath0 with adjacency matrix @xmath1 is defined as the complex function @xmath79 , from @xmath80 to @xmath80 , where @xmath1 is the adjacency matrix of @xmath1 .
if @xmath81 is the spectral radius of @xmath1 , the absolute value of the largest eigenvalue of @xmath1 , then the function @xmath82 is analytic in @xmath83 .
the fredholm determinant of @xmath1 is then @xmath84 , which if @xmath48 is an eigenvalue of @xmath1 is defined as @xmath21 .
zeta functions are of interest as they relate with topology .
we have a taylor expansion @xmath85 for small @xmath86 , where @xmath87 is the number of rooted closed paths of length @xmath41 in the graph .
the zeta function is therefore a generating function for a dynamical property of the graph , the dynamical system being the markov chain defined by the graph .
it is in particular an * artin - mazur zeta function * and a special case of the * ruelle zeta function * @xcite .
since @xmath88 is a rational function , the sum can be understood for general @xmath89 only by analytic continuation . while for general graphs
, @xmath90 can be quite arbitrary , we will see that for connection graphs @xmath75 , the analytic continuation of the divergent series @xmath91 at @xmath92 is either @xmath21 or @xmath93 and that the case @xmath21 appears if and only if there is an even number of odd - dimensional complete subgraphs of the original graph @xmath0 .
if @xmath94 is the number of @xmath41-dimensional simplices , then the euler characteristic @xmath95 is the difference of @xmath96 , where @xmath97 and @xmath98 . + while these cardinalities @xmath99 appear naturally when evaluating the zeta function of a connection graph at @xmath92 , concrete examples of zeta function of a connection graph of some of the simplest graphs indicate , that the result @xmath100 is not that obvious : @xmath101 , @xmath102 , @xmath103 .
these functions evaluated at @xmath48 either have the value @xmath50 or @xmath48 . by the way , the topic of fredholm determinants has appeared in the movie ' ' good will hunting " as one of the problems involves the bowen - lanford function : the last blackboard problem in that movie asks for the generating function for walks from a vertex @xmath105 to a vertex @xmath106 in concrete graph .
the answer is @xmath107_{ij}$ ] which by cramer is expressed by the * adjugate matrix * as the rational function @xmath108 , where @xmath109 is the matrix obtained by deleting row @xmath105 and column @xmath106 in @xmath1 . + in search of a prove of the theorem , it can be helpful to see the connection graph as a geometric space and see a permutation @xmath110 of its vertices as a one - dimensional oriented submanifold " , a collection of disjoint cyclic oriented paths or strings " . since a permutation compatible with the graph as a flow " on the geometry , the fredholm determinant sums over all possible measurable continuous dynamical systems @xmath111 " .
they can be considered flows in @xmath75 in the sense that for every vertex @xmath15 , the pair @xmath112 is an edge in @xmath75 .
we call them measurable because @xmath111 is not continuous in the geodesic distance metric of @xmath75 .
+ the signature @xmath113 of a flow is the product of the signatures of the individual connected cyclic components of the flow = permutation @xmath110 .
the fredholm determinant @xmath114 of the adjacency matrix @xmath115 of the connection graph @xmath75 is then a * path integral * @xmath116 , where @xmath110 runs over all possible flows in @xmath75 .
the unimodularity theorem tells then that the fredholm determinant @xmath99 is equal to the * fermi characteristic * @xmath117 , where @xmath15 runs over all complete subgraphs of @xmath0 , showing so that @xmath19 is a * multiplicative valuation * @xmath118 .
we can compare @xmath99 with the additive valuation @xmath119 on graphs which is the * euler characteristic * @xmath11 or with the * wu characteristic * @xmath120 , summing over all edges @xmath121 of the connection graph @xmath75 @xcite .
but unlike euler characteristic @xmath122 or wu characteristic @xmath123 , the functional @xmath19 is not a combinatorial invariant , as @xmath19 is constant @xmath50 on barycentric refinements
. + functionals like the range of the unimodular green function @xmath124_{ij}$ ] values appear to be combinatorial invariants - at least in experiments .
we have not proven this observation yet .
the closest analogy which comes to mind is an invariant found by bott @xcite who coined the term * combinatorial invariant * as a quantity which is invariant under barycentric subdivision . by cramer
, @xmath125 is the fredholm characteristic of the geometry in which cell @xmath105 is removed and @xmath126 a fredholm characteristic of a geometry , where outgoing connections from cell @xmath105 and incoming connections to cell @xmath106 are snapped .
if @xmath127 is a finite simple graph , we denote by @xmath128 the set of all complete subgraphs of @xmath0 . also named * simplices * or * cliques * , these subgraphs are points of the * barycentric refinement * @xmath74 of @xmath0 , which has as a vertex set @xmath128 the set simplices and where two such simplices are connected if one is contained in the other . the larger * connection graph * @xmath75 has the same vertex set like @xmath74 . in that graph ,
two simplices are connected , if they have a non - empty intersection . unlike for @xmath74 , for which the maximal dimension of @xmath0 and @xmath74 are the same , the graph @xmath75 is in general
fatter " : for a one - dimensional circular graph for example , the graph @xmath75 has triangles attached to each edge .
more generally , @xmath75 contains complete subgraphs @xmath129 if there is a vertex of @xmath15 which is contained in @xmath49 simplices : the unit ball of a simplex @xmath130 belonging to an original vertex is a complete graph . + also if we primarily want to analyze the graph case , it is convenient to look at more general simplicial complex structures on the graph
. assume @xmath0 is an * abstract finite simplicial complex * , a finite set @xmath131 equipped with a collection @xmath5 of finite subsets of @xmath131 such that for every @xmath132 and every subset @xmath133 of @xmath1 also @xmath134 , then the connection graph @xmath135 is the finite simple graph for which two elements in @xmath5 are connected , if they intersect . an abstract simplicial complex is not only a generalization of a graph , we can see it as a * structure * imposed on a graph similarly as a * topology * , an * order structure * or * @xmath136-algebra * is imposed on a set . much of the graph theory literature sees a graph @xmath127 by default equipped with the one - dimensional * skeleton complex * @xmath137 .
the largest complex is the * whitney complex * on @xmath0 , which is the set of all complete subgraphs .
graphs can handle many simplicial complexes as given an abstract finite simplicial complex @xmath0 , the barycentric refinement @xmath138 is a graph which has as vertex set @xmath128 the set of elements in @xmath0 and has @xmath139 .
+ given an abstract finite simplicial complex @xmath0 given by a finite set @xmath131 equipped with a collection @xmath5 of finite subsets , the * connection graph * of @xmath0 is the graph with vertex set @xmath5 , where two vertices @xmath140 are connected , if they intersect as subsets of @xmath131 . + more general than simplicial complexes are discrete * cw complexes*. this structure is built up inductively .
it recursively defines also the notion of contractibility and a notion of sphere in this structure .
this generalizes the evako setup in the graph case ( see @xcite ) start with the empty set , which is declared to be the @xmath141-sphere .
it does not contain any cells .
a * cw - complex * is declared to be a @xmath142-sphere if when punctured becomes contractible and which has the property that every unit sphere @xmath16 of a @xmath143 sphere .
the * unit sphere * of a cell @xmath15 is the cw - sub complex of @xmath0 containing all cells which are either part of @xmath15 or which contain @xmath15 .
also inductively , a cw complex @xmath0 is * contractible * if there exists a cell @xmath15 such that both @xmath16 and @xmath0 without @xmath15 are contractible .
inductively , if @xmath0 is a cw - complex one can build a larger complex by choosing a sphere @xmath144 in @xmath0 , then do an extension over @xmath144 with a new cell @xmath15 , producing so a larger cw complex .
the unit ball of @xmath15 has @xmath145 as a boundary and the new cell is declared to be @xmath146 .
starting with the empty set , one can build up like this structures which are more general than finite simplicial complexes but which still do not ( unlike the classical definition of cw complexes ) invoke the infinity axiom in zermelo - frenkel .
the * connection graph * of a cw complex is the finite simple graph @xmath147 , where @xmath131 is the set of cells and where two cells are connected if they intersect .
the fermi characteristic and fredholm characteristic of a cw complex are defined in the same way as before .
+ * remarks . *
+ * 1 ) * if one looks at the unit balls of cells as a
cover `` of the cw complex , then the connection graph plays the role of the * nerve graph * in ech setups .
the elements in the cover have however more structure because their boundaries are always graph theoretical spheres .
every simplicial subcomplex of the whitney graph is a cw complex in the just given sense and any finite classical cw complex can be described combinatorially as such .
there is a more general notion of * abstract polytope * given as a poset satisfying some axioms but the unimodularity theorem wo nt generalize to that .
the * barycentric refinement * of a cw complex is a graph containing the cells as vertices and where two cells are connected if one is contained in the other .
this graph is equipped with the whitney complex structure which is again a cw complex , but which is much larger . but
again , like for the notion of simplicial complex , we see that it can be implemented as a graph and that we can see therefore a cw complex as a * structure imposed on a graph*. there is still a reason to keep the notion of cw complexes : the connection graph of a cw complex has the same number of vertices than cells and the unimodularity theorem applies to it
. the connection graph of the graph attached to the cw complex would be much larger .
this will be relevant when we look at prime graphs .
it is also relevant in general : if we look at a barycentric refinement of a cw complex , then it is a graph which has the same number of vertices than the connection graph of this cw complex .
the prime graph and prime connection graph considered below are then a morse filtration of the barycentric refinement and connection graphs of the simplest simplicial complex one can imagine : the set of primes equipped with the set of all subsets as complex . +
* 2 ) * as the barycentric refinement of a simplicial complex or cw complex encodes most essential topological features in @xmath0 , there is not much loss of generality by looking at graphs rather than simplicial complexes .
still , the slightly increased generality can make the result more transparent . but
applying it to graphs is more intuitive .
the structure of a cw complex is very natural and practical : look at the cube graph for example .
since the unit spheres are graphs without edges , it is a one - dimensional graph when equipped with the whitney complex .
if we stellate the 6 faces , then we have a larger graph with 14 edges , the stellated cube .
its whitney complex is large as there are already 24 triangles present . but
adding 6 cells with @xmath148 boundaries , we get the familiar picture of a cube with f=6 two dimensional faces , e=12 one dimensional edges and v=8 vertices .
this is how already descartes counted the euler characteristic @xmath149 @xcite .
history shows @xcite how difficult it has been to get to a good notion of ' ' polyhedron " ( see also @xcite ) and one usually refers to euclidean embeddings , using through notions like convexity to define it properly @xcite . + * 3 ) * the notion of cw complex in the discrete allows ( without using any euclidean notions ) to give a decent definition of polyhedron as a cw complex which is a sphere in the sense that removing one cell renders the cw complex contractible and that every unit sphere of any cell is a sphere .
the essential foundation to that is in whitehead @xcite already in the 1930ies .
what was new in the 90ies is the realization that one can do all this on graphs without using the continuum .
there are three reasons why the language of finite graphs is more convenient : it is an intuitive structure which small kids can grasp already ; it is a data structure which exists in many higher level programming languages .
all results discussed here can be explored with a few lines of code ( provided below ) . finally , it is a finite structure ; finite mathematics works also in a framework of finitists like brouwer or strict finitist , which many computer scientists are , wanting to implement the complete structure faithfully .
+ * 4 ) * whitehead * homotopy * has been ported to discrete structures by first defining * contractibility * inductively : it is either the @xmath50 vertex graph or a graph for which there exists a vertex @xmath15 such that both the unit sphere @xmath16 is contractible and such that the graph without @xmath15 is contractible .
contractible graphs have euler characteristic @xmath50 .
a * homotopy step * is the process of an addition or removal of a vertex @xmath15 , for which @xmath16 is contractible .
two graphs are homotopic , if one can get one from the other by applying a sequence of homotopy steps
. a graph can be homotopic to a 1-point graph without being contractible .
an example is the * dunce hat * which shows that one first has to enlarge the graph before it becomes contractible .
homotopic graphs have the same cohomology and euler characteristic however .
+ while the barycentric refinement @xmath74 is homotopic to @xmath0 , @xmath75 is not homotopic to @xmath0 in general but this happens only if * very small * homotopically non - trivial spheres are present . for the octahedron @xmath0 for example , the graph @xmath75 has euler characteristic @xmath21 while @xmath0 has euler characteristic @xmath22 so that @xmath0 and @xmath75 can not be homotopic
this only happened because the geometry was too small and the @xmath150 cohomology of @xmath0 collapsed in @xmath75 .
for smooth enough graphs , @xmath75 is homotopic to @xmath0 it is homotopic and has the same cohomology .
the connection graph of the barycentric refinement @xmath151 of @xmath74 for example is always homotopic to @xmath74 and so to @xmath0 as loosing the additional connection bonds does allows to morph from @xmath78 to @xmath151 without changing the topology . + * 5 ) * the connection graphs emerged for us in the context of the * wu characteristic * @xmath152 of a graph @xmath0 which is a second order " euler characteristic .
the wu characteristic shares all important properties of euler characteristic : it is multiplicative and additive with respect to multiplication or addition of the geometric structures ; there is a compatible calculus , cohomology and theorems like gauss - bonnet , poincar - hopf , euler - poincar , kuenneth or lefschetz generalize .
a relation of the fredholm adjacency matrix and the wu characteristic is given by @xmath153 where @xmath154 is the * checkerboard matrix * @xmath155 .
an integer matrix @xmath156 is called * unimodular * if its determinant is either @xmath50 or @xmath48 . by the explicit * cramer - laplace inversion formula * , unimodularity is equivalent to the fact that its inverse @xmath157 is an integer - valued matrix .
because an unimodular matrix @xmath156 is non - singular in the ring @xmath158 of integer matrices , a unimodular matrix @xmath156 is an element in @xmath159 .
+ let @xmath0 be a graph equipped with a simplicial complex or a cw - complex . with the fermi number @xmath8 giving the number of odd dimensional simplices of cells in @xmath0 , the fermi characteristic is defined as @xmath160 .
the * fredholm characteristic * @xmath161 is the fredholm determinant of the adjacency matrix of the connection graph @xmath75 of the complex . for any graph @xmath0 equipped with a simplicial or cw structure : @xmath162 the following corollary of the theorem is actually equivalent to the theorem , at least on simplicial complexes . as usual , if @xmath127 and @xmath163 are two finite simple graphs , the * intersection graph * @xmath164 and * union graph * @xmath165 are both finite simple graphs . also in the slightly more general case ,
if @xmath166 are finite abstract simplicial complexes or cw complexes , one can look at the intersection @xmath167 and union complex @xmath168 . for any finite simple graphs @xmath166 , simplicial complexes or cw complexes , the determinant formula @xmath169 holds and @xmath170 for @xmath51 and @xmath171 .
the first statement follows directly from the theorem and uses the fact that the explicit knowledge of @xmath8 outs it is a * valuation * , an integer - valued functional @xmath172 on the set of graphs or simplicial complexes satisfying @xmath173 .
the second statement follows from the fact that the number of odd - dimensional simplices in @xmath174 is odd if @xmath51 .
the reason is that the @xmath172-vector of a complete graph @xmath175 is explicitly given in terms of binomial coefficients @xmath176 so that @xmath177 for @xmath178 . in the case
@xmath179 , we have @xmath180 , otherwise @xmath8 is odd .
let @xmath181 denote the number of even - dimensional complete subgraphs of @xmath0 .
the * join * of two graphs @xmath182 is defined as the graph @xmath183 .
the join operation has the same properties as in the continuum : the join of two discrete spheres for example is again a discrete sphere .
the join of a graph @xmath0 with a @xmath21-sphere @xmath184 is the * suspension * of @xmath0 .
the octahedron from example is the suspension of the cyclic graph @xmath148 and repeating the suspension construction on the octahedron and beyond produces all cross polytopes .
an other consequences of the unimodularity theorem is : if @xmath0 is the topological join of @xmath185 with a graph @xmath144 , then @xmath186 especially , if @xmath0 is the suspension of @xmath144 , then @xmath187 .
each newly added odd - dimensional simplex in @xmath0 corresponds to an even - dimensional simplex in @xmath144 . to see the second part , note that the suspension is obtained by performing an second join operation over @xmath144 so that we get again @xmath188 .
* remarks . *
+ * 1 ) * the unimodularity theorem is clear for disjoint graphs @xmath29 or if @xmath189 is a subgraph of @xmath0 .
a special case is if @xmath29 intersect in a single vertex .
then the fredholm characteristics of @xmath189 and @xmath0 multiply .
this special case is related to the known formula @xcite ( corollary 8.7 ) @xmath190 because in the case of two subgraphs of a large complete host graph , intersecting in a point , the adjacency matrix of @xmath191 is @xmath192 .
this is no more true for the connection graph : if we join @xmath29 at a point , there are many simplices @xmath140 from different graphs which join , so that @xmath193 is no more true .
still , the unimodularity theorem implies in that case that @xmath194 for the adjacency matrices of the two connection graphs @xmath195 if @xmath196 .
we see also that except for @xmath197 , removing an edge from @xmath174 removes an even number of odd - dimensional complete subgraphs .
the reason is that @xmath198 is even for positive @xmath41 and odd for @xmath199 .
for example , when removing an edge from @xmath77 , we remove a tetrahedron and an edge to get a kite graph .
the @xmath172-vector encoding the cardinalities of the complete subgraphs changes from @xmath200 to @xmath201 .
+ * 2 ) * the unimodularity theorem implies for example that the path integral sum can be replaced with one single permutation : for example , we can enumerate the odd dimensional simplices as @xmath202 , then define a map @xmath203 from odd to even dimensional simplices by just dropping the first coordinate . define @xmath204 and @xmath205 for any other @xmath206 .
the transformation @xmath207 has the signature @xmath99 .
while @xmath208 is generated by transpositions , it is not a transposition itself in general .
it is not a transposition for a kite graph for example .
but it can be , like in the case @xmath148 .
while some permutations are continuous , the just mentioned one is not .
the closed set @xmath209 of a vertex @xmath15 is mapped into an open set @xmath210 .
the map is not continuous . while on @xmath211 , there are @xmath212 different transformations , only two of them are continuous and only the identity is a continuous flow . the other two transpositions @xmath213 and @xmath214 are two signature @xmath48 transformations but they are not continuous .
the sum @xmath215 is @xmath99 .
+ the next corollary expresses the fredholm characteristic of a ball by the euler characteristic of its boundary : for any graph @xmath127 and every vertex @xmath216 , we have @xmath217 .
the number of odd - dimensional simplices in @xmath218 not in @xmath16 is equal to the number of even - dimensional simplices in @xmath16 .
therefore , @xmath219 .
now exponentiate @xmath220 ( we could write this as @xmath221 , where @xmath14 is a poincar - hopf index at @xmath15 . )
* examples : * +
* 1 ) * if @xmath0 is a graph for which every unit sphere is a discrete sphere of euler characteristic @xmath21 or @xmath22 , then the unit ball @xmath218 has @xmath222 . for an icosahedron @xmath0 for example , where every unit ball is a wheel graph @xmath223 with @xmath224 spikes , there are @xmath225 edges in each unit ball . + * 2 ) * while the just mentioned corollary shows that
for discrete spheres @xmath16 with @xmath226 , the unit balls always have @xmath227 , there are spheres of arbitrary large dimension which can come both with @xmath228 or @xmath227 . here
are examples : start with one dimensional sphere @xmath52 with odd @xmath49 .
it has @xmath228 .
every suspension is again a sphere and @xmath19 does not change under the suspension operation .
besides enlarging the base set @xmath131 of a graph @xmath127 , there is an other possibility to deform the geometry of a graph : we can * refine the simplicial complex structure * imposed on it .
the simplest simplicial complex on a finite set is the * 0 * -skeleton , where the set of subsets of @xmath131 is the set @xmath229 . since there are no paths except the trivial path with signature @xmath50 , the fredholm characteristic of such a @xmath21- dimensional space is @xmath50 and the unimodularity result is obvious in that case as no odd dimenional simplices exist then .
+ the next possibility is to take the * 1 * -skeleton complex @xmath137 .
this renders the graph one - dimensional ; it is the structure which is often associated with a graph , when defining a graph as a one - dimensional simplicial complex . also in this case
, one can see the unimodularity theorem .
but it is less obvious already .
first of all , we have @xmath230 as no higher dimensional simplices are present . in the case
when the graph is a tree , then there are no closed paths beside @xmath197 paths . in order to analyze this
, we have to see that for each edge there are two loops exchanging the edge with the attached vertices .
what happens if we merge two edges is that the two transpositions merge to two circular loop depending on the order and that there is a new transposition exchanging edges .
the functional @xmath19 is multiplicative . + if we start including two - dimensional simplices by
switching on " the triangle , the number of paths increases considerably .
lets call @xmath231 the triangular graph @xmath232 equipped with the one - dimensional simplicial complex .
we have @xmath233 as there are three odd dimensional simplices present , the 3 edges .
we also have @xmath234 as there are no new odd - dimensional simplices .
the reason why the functional does not change when moving from @xmath231 to @xmath232 is because @xmath235 .
but in the three dimensional case , because @xmath236 , the functional @xmath19 changes by @xmath237 , becoming @xmath238 .
then again , the euler characteristic of @xmath239 and there is no change in @xmath119 .
+ the unimodularity theorem holds more generally when the graph is equipped with a * finite cw complex structure*. if a finite simple graph @xmath127 is equipped with such a simplicial or cw structure , it still defines a * connection graph * and so a fredholm determinant .
the definition of a cw structure requires for a good notion of a sphere " in graph theory . traditionally , a cw complex is a hausdorff space equipped with a collection of * structure maps * from @xmath41-balls to @xmath240 .
as we have seen , a * finite cw complex structure * works the same way , but cells do nt need to be simplices any more .
this is of practical value as we can work in general with smaller complexes when dealing with homotopy invariant notions like cohomology .
it allows us to see the deformation of the simplicial structure as the process of adding or removing cells to the cw complex .
the following key result will allow to see how the fredholm determinant changes if we add a cell to a cw complex .
the poincar - hopf indices are not necessarily @xmath20-valued any more in general , because @xmath119 can take values different from @xmath21 or @xmath22 .
the fredholm determinant extension formula needs a slightly more general extension process : + let @xmath0 be a finite simple graph and let @xmath144 be a subgraph @xmath144 of @xmath0 .
define the * pyramid connection graph * @xmath241 as the pyramid extension over the subgraph of @xmath75 generated by the simplices which intersect a vertex in @xmath144 .
this new graph @xmath242 is equipped not with the whitney complex of the pyramid extension but with the union of the complex of @xmath75 together with all simplices @xmath243 , where @xmath244 is a simplex in @xmath144 .
one can build up any simplicial complex as such .
given for example the triangle graph @xmath127 equipped with the 1-skeleton complex @xmath245 .
the pyramid connection complex @xmath242 has now as vertices the elements in @xmath246 together with a new element @xmath15 .
now @xmath15 is connected to any element in @xmath246 which intersects @xmath144 additionally , any two elements in @xmath246 are connected if they intersect . in this case , the graph @xmath242 agrees with the connection graph of @xmath232 equipped with the whitney complex .
+ here is the result : for any graph @xmath0 and any subgraph @xmath144 of @xmath0 we have @xmath247
the same holds for simplicial complexes or cw complexes @xmath0 if @xmath144 is a sub complex . *
( i ) * the map @xmath248 is an additive valuation .
the functional super counts the number of new paths which appear by adding the cell @xmath15 and if a path passes through @xmath15 , then no other path can pass through it .
for any two sub complexes @xmath249 we therefore have @xmath250 * ( ii ) * it follows that also @xmath251 is a valuation : the value @xmath99 is constant as we take @xmath144 as an argument of the functional .
+ * ( iii ) * the claim of the proposition holds if @xmath144 is a complete graph @xmath175 inside @xmath0 .
we have to show @xmath252 . proof . in the fredholm matrix of @xmath242 ,
the row belonging to the @xmath142-simplex belonging to @xmath144 in @xmath0 and the row belonging to the new vertex @xmath15 all have entries @xmath50 .
the matrix is singular .
+ * ( iv ) * since by ( iii ) , @xmath253 if @xmath144 is a complete graph , the valuation @xmath240 takes the value @xmath50 for every complete subgraph @xmath254 . claim : a valuation with this property must be the euler characteristic : @xmath255 proof . by discrete hadwiger
, every valuation has the form @xmath256 .
we know @xmath257 implying @xmath258 .
inductively this shows that the vector @xmath259 is @xmath260 .
an other , more elaborate argument is to see that the assumption implies that the euler characteristic of a barycentric refinement of a simplex is @xmath50 too , showing that @xmath240 is invariant under barycentric refinement . since one knows that the eigenvalues and eigenvectors of the barycentric refinement operator and especially that euler characteristic is the only fixed point of the barycentric operator , we are done .
* examples . *
+ * 1 ) * if @xmath261 , then there is one empty path and one involution in the leibnitz - fredholm determinant , which is a path integral sum .
the fredholm characteristic is @xmath262 .
+ * 2 ) * if @xmath263 , then there are 2 graphs of length @xmath55 , the empty graph and three involutions .
again we have @xmath264 . + * 3 ) * if @xmath265 , then the fredholm matrix @xmath266 \ ; .\ ] ] the proposition makes the path counting problem additive as @xmath267 is a valuation .
it shows that path integral of the set of new paths which appear when adding @xmath15 .
+ * examples : * + * 1 ) * assume that @xmath144 has @xmath41 vertices and no edges . for every of the vertices
@xmath244 we can look at the paths @xmath268 of length @xmath22 . since we can not have simultaneously two paths hitting @xmath15 , the number of additional paths is @xmath269 .
they are counted negatively as they have even length . + * 2 ) * if @xmath144 is @xmath197 then we have three cycles of length @xmath22 , three cycles of length @xmath55 , and one path of length @xmath270 paths passing through @xmath15 .
the last one cancels the empty path .
we get zero .
+ * 3 ) * if @xmath144 is a complete graph of @xmath41 vertices , then we have a unit ball in the connection graph which is a complete graph .
the extended graph has fredholm characteristic @xmath21 . + * 4 ) * if @xmath144 is a cyclic graph of @xmath271 vertices .
now the sum over new path contributions is zero : while @xmath272 is not true any more in general , the proposition shows why the unimodularity theorem works .
we can interpret the pyramid extensions over @xmath41-spheres as adding a @xmath273-dimensional cell or simplex .
the proposition allows to relate fredholm characteristic to euler characteristics of extensions . in the case of simplicial complex extensions , the @xmath274
are always @xmath50 or @xmath48 .
+ * examples : * + * 1 ) * let @xmath0 be a one point graph .
the graph @xmath242 is @xmath197 and @xmath275 .
+ * 2 ) * let @xmath0 be the two point graph without edges .
then @xmath242 is the line graph @xmath276 and @xmath277 . indeed
+ * 3 ) * if @xmath144 is contractible , then @xmath279 . for example , if @xmath144 is a one point graph @xmath280 we just assign a new cell for which every simplex connected to @xmath244 is belonging .
we have now a complete subgraph @xmath218 . we know already that if a connection graph has a unit ball which is a complete graph , then the fredholm characteristic is zero .
+ if @xmath0 is an even dimensional @xmath41-sphere , then @xmath281 and if @xmath0 is an odd dimensional @xmath41-sphere , then @xmath282 .
this implies that if we add an odd dimensional simplex to a simplicial complex , the fredholm characteristic changes sign and if we add an even dimensional simplex , then the fredholm characteristic stays the same .
for ever integer @xmath283 let @xmath131 be the set of natural numbers in @xmath284 which are square free . connect two integers if one is a factor of the other @xcite .
the corresponding graph @xmath285 is the * prime graph*. one can see it as the part @xmath286 of the barycentric refinement of the complete graph on the spectrum @xmath31 of the integers @xmath287 , where @xmath288 is the counting function .
the * prime connection graph * @xmath289 has the same vertex set @xmath131 but now , two integers are connected if they have a common factor larger than @xmath50 @xcite .
the two graphs allow us to illustrate some of the theorems .
for the graph @xmath285 , the poincar - hopf theorem and the euler - poincar relation to cohomology is interesting as it allows to express the mertens function interms of betti numbers . in the case of @xmath289
, we can now formulate a consequence of the proposition which is the analogue of poincar - hopf and the unimodularity result relating the fermi characteristic with a fredholm determinant of the adjacency matrix .
we see @xmath285 as part of the barycentric refinement of the complete graph on the spectrum of the integers @xmath290 and see @xmath289 as the connection graph of that complete graph .
this point of view comes into play when seeing * square free integers * as simplices in a simplicial complex .
+ despite the fact @xmath289 is not directly the connection graph of an other graph ( it is part of the connection graph of an finite graph , the complete graph with vectex set @xmath291 ) , lets still denote by @xmath292 the fredholm characteristic , the fredholm determinant of the adjacency matrix of @xmath289 .
let @xmath293 denote the poincar - hopf index of the counting function @xmath172 at @xmath15 .
it is @xmath294 , the * mbius function*. we have seen that poincar - hopf implies that @xmath295 , where @xmath296 is the * mertens function*. + we have now directly from the unimodularity theorem for @xmath297 complexes a multiplicative euler characteristic formula analogous to @xmath298 @xmath299 . the left hand side is the fredholm characteristic of the cw complex @xmath285 .
the right hand side is the fermi characteristic of @xmath285 .
here is a comparison between the graph @xmath285 and @xmath289 . lets define @xmath300 as the value of the poincar polynomial evaluated at @xmath48 , where @xmath301 is the @xmath41th betti number .
the cohomological data of course correspond exactly to the corresponding notions in the continuum .
an evako @xmath142-sphere for example has the poincar polynomial @xmath302 and a contractible space has poincar polynomial @xmath50 . [ cols="<,<",options="header " , ] * example . *
the prime connection graph @xmath303 is a graph with 18 vertices @xmath304 and @xmath305 edges .
we can look at @xmath289 as a cw complex , where the individual nodes represent the cells .
the fredholm matrix of @xmath289 is @xmath306 \ ; .\ ] ] its determinant is @xmath48 .
[ [ section ] ] the unimodularity theorem does not hold for an arbitrary * generalized matrix function * as defined by marcus and minc @xcite
. it does not work for * permanents * for example as we can check that @xmath19 is not a multiplicative valuation : the fredholm permanent of the kite graph is @xmath307 .
the fredholm permanent of the intersection is @xmath22 .
the fredholm permanent of the two triangular pieces is @xmath53 .
as mentioned in the introduction , the fredholm permanent of a complete graph is @xmath308 , while the determinant is the number of derangements .
the fact that the product property or the unimodularity result can not hold for permanents is similar to the fact is that the euler characteristic @xmath309 is constant @xmath50 for a contractible graph and invariant under homotopy deformations , while the bosonic analogue , the * kalai number * @xmath310 counting the number of simplices in a graph is a valuation but not invariant under homotopy deformations .
+ [ [ section-1 ] ] why does the proof only work for the connection graphs and not for the graphs themselves ? already for small graphs like @xmath197 or @xmath232 ,
the fredholm matrices are not unimodular any more .
indeed , for complete graphs , the fredholm adjacency matrix is the matrix which has @xmath50 everywhere .
it seems that what is needed is the high connectivity of the simplices in the connection graph : paths in the unit ball in @xmath75 of every original vertex @xmath15 are in 1 - 1 correspondence to permutations of the vertices in the ball @xmath218 of @xmath75 .
this is important as the sum of the signatures over all non - identity transformations in @xmath218 is @xmath48 .
+ [ [ section-2 ] ] we initially tried to prove the result by gluing two graphs @xmath311 to a larger graph @xmath312 and intersection @xmath313 and take a permutation @xmath15 of @xmath314 and @xmath244 of @xmath315 and combine them to a permutation @xmath316 in @xmath75 .
the most natural choice is a composition of the permutations .
while @xmath317 holds , we are not able to match the composed transformations @xmath316 with a transformation @xmath318 , where @xmath89 is a permutation in @xmath75 and @xmath319 a permutation in @xmath320 .
the product @xmath316 is not necessarily a permutation any more in @xmath75 and we need @xmath319 in @xmath320 to achieve that . now
if @xmath321 , then @xmath322 is a permutation in @xmath144 . in one of the simplest cases , where @xmath323 and @xmath324 and @xmath325 . now
the number of permutations in @xmath314 or @xmath315 are 11 .
there are @xmath326 product transformations .
there are @xmath305 transformations in @xmath75 and @xmath55 transformations in @xmath320 .
there are only @xmath327 products .
we see , we can not get all the pairs @xmath268 with pairs @xmath318 .
cooking up a proof strategy which pairs the permutations up and then proves that that the signatures of the remaining sum up to zero has not yet worked .
+ [ [ section-3 ] ] let @xmath127 be a finite simple graph .
the determinant of the adjacency matrix @xmath1 is a super count of cyclic permutations of @xmath131 .
the determinant of the fredholm matrix @xmath328 is a super count of the permutations of @xmath131 .
the pseudo determinant of the laplacian @xmath329 counts the number of rooted trees in @xmath0 and the determinant of the fredholm laplacian @xmath330 counts the number of rooted forests in @xmath0 . in each case
we see that the determinant counts some zero or one - dimensional directed subgraphs . what the significance of the unimodularity of the fredholm matrix @xmath45 is remains to be seen .
it will depend on properties of the green functions , the entries of the matrix @xmath331 which is the frchet derivative of the map @xmath332 from @xmath333 matrices to the real axes @xcite ( corollary 5.2 ) .
+ [ [ section-4 ] ] it follows from the explicit formula for the @xmath172-vector that the barycentric refinement @xmath74 of any graph @xmath0 is always a positive graph .
there is an upper triangular matrix @xmath1 such that @xmath334 for all simplicial complexes .
the matrix is @xmath335 , where where @xmath336 are the * stirling numbers * @xmath336 of the second kind . since @xmath337 is even for @xmath338 all rows of @xmath1 beyond the first one are even .
the barycentric refinement matrix for simplicial complexes with maximal dimension @xmath270 for example is @xmath339 \;.\ ] ] having seen that @xmath340 for any graph @xmath0 or simplicial complex if @xmath74 is the barycentric refinement , it follows by the unimodularity theorem that @xmath19 is * not * a combinatorial invariant .
+ [ [ section-5 ] ] having discovered unimodularity experimentally in february 2016 and announced in october @xcite it took us several attempts to do the proof presented here .
induction by looking algebraically at the matrices appeared difficult .
an other impediment is the lack of an additive structure on the category of graphs
. it would be nice to see @xmath341 as a group homomorphism for some larger group and then prove it for a basis only .
this has not yet worked : in the larger category of @xmath342-chains we have also connection graphs but @xmath19 does not extend .
one can also look at the ring of graphs obtained by taking the connection graphs defined by a set of complete subgraphs ( which form a boolean ring ) .
but @xmath19 does not remain multiplicative on that larger structure .
+ [ [ section-6 ] ] a finite simple graph @xmath0 with unimodular adjacency matrix @xmath1 is called an * unimodular graph*. an example is the linear graph @xmath343 for which the adjacency matrix has determinant @xmath344 .
as this is not the same than having the fredholm matrix @xmath45 unimodular , we here did not use the terminology unimodular graph " .
experiments show that the fredholm matrices @xmath345 of connection graphs often also have unimodular or zero determinant submatrices obtained by deleting one row and one column which is equivalent that @xmath346 is a @xmath347 matrix . in general
, for a connection graph @xmath0 , the inverse @xmath348 has only few elements different from @xmath347 .
these appear to be interesting * divisors * to consider .
+ [ [ section-7 ] ] an * abstract finite simplicial complex * @xmath349 on a finite set @xmath131 is a set of subsets of @xmath131 such that if @xmath350 then also any subset of @xmath1 is in @xmath349 .
special simplicial complexes are * matroids * , simplicial complexes with the * augmention property * telling that if @xmath351 then one can enlarge @xmath15 with and element in @xmath244 to get an element in the simplex .
our point of view was to see simplicial complexes on a finite set @xmath131 as a * geometric structure * imposed on @xmath131 similarly as one imposes a * set theoretical topology * @xmath352 in topology or a * @xmath136-algebra * @xmath353 in measure theory . like
any topology , @xmath136-algebra or simplicial complex @xmath349 on @xmath131 defines a * connection graph * @xmath354 of the structure : the vertices of @xmath354 are the sets in @xmath349 and two such vertices are connected , if they intersect as subsets of @xmath131 .
the deformation of the simplicial structure gave more freedom than the deformation of the graph .
besides complete graphs , one can define other simplicial complexes based on graphs .
one can for example look at the matroid on the set of edges in which the faces are the forests .
this simplicial complex defines what one calls a * graphic matroid*. the fermi characteristic of this matroid is the number of forests with an odd number of edges .
the connection graph of the matroid has as vertices the forests and connects two if they intersect in at least one edge .
the universality theorem applies here too .
for example , if @xmath76 , there are 6 forests in @xmath0 .
the fredholm matrix of the connection graph @xmath355\ ] ] has determinant @xmath48 so that the graphic matroid has fredholm characteristic @xmath228 .
there are @xmath55 one - dimensional forests in @xmath0 so that also the fermi characteristic is @xmath48 .
[ [ section-8 ] ] a finite simple graph @xmath127 carries a natural associated simplicial complex , the * whitney complex * @xmath349 .
it consists of all subsets of @xmath131 which are complete subgraphs of @xmath0 .
the * one skeleton complex * @xmath356 is an other example of a simplicial complex .
simplicial complexes do not form a boolean ring : take the boolean addition of @xmath357 with @xmath358 , then this is the complement of @xmath359 in @xmath357 which is no more a simplicial complex in general as @xmath360 and @xmath361 shows .
the unimodularity theorem has no simple algebraic proof because the category of graphs or the category of simplicial complexes only forms a * boolean lattice * and not a * boolean ring*. an algebraic statement @xmath362 does therefore not work for the simple reason that the boolean sum @xmath363 is not a graph .
also , since the result does not extend to the * ring of chains * , ( the free abelian group generated by the simplices ) , we had to proceed differently .
the boolean algebra of the set of simplices in @xmath0 is a ring however so that for every element , we can define a connection graph .
however , the functional @xmath19 does not extend as one can build any graph like that .
[ [ section-9 ] ] the * line graph * of a graph @xmath0 has the edges as vertices and two edges are connected , if they intersect .
it is a subgraph of the connection graph @xmath75 of the @xmath50-skeleton complex defined by the graph @xmath0 .
we mention this structure , as it is also sometimes called the * intersection graph * of @xmath0 .
it plays a role in topological graph theory .
if @xmath364 is the incidence matrix of @xmath0 which has as a kernel the cycle space , then @xmath365 is kirchhoff laplacian , where @xmath133 is the vertex degree diagonal matrix .
now @xmath366 is the adjacency matrix of the line graph if @xmath367 .
we see that the fredholm matrix of @xmath368 is an adjacency matrix of some graph .
[ [ section-10 ] ] the formula @xmath369 which converges for small @xmath86 gives an interpretation of the fredholm determinant in terms of closed paths , which can self intersect .
note that @xmath370 is a polynomial of degree @xmath49 , where @xmath49 is the number of vertices in @xmath0 . since @xmath371 , where @xmath87 is the number of closed paths in the graph of length @xmath41 .
there is a relation with ihera zeta functions , which satisfies @xmath372 , where @xmath1 is the * hashimoto edge adjacency matrix * of @xmath0 .
+ [ [ section-11 ] ] connection graphs could be defined for infinite , and even uncountable graphs .
its not clear yet which generalizations lead to fredholm operators and so to fredholm determinants .
take the points of a circle @xmath373 for example as the vertex set , fix two irrational numbers @xmath374 like @xmath375 and @xmath376 .
connect two points if there exists @xmath377 with @xmath378 such that @xmath379 .
the points of the connection graph is then the set of orbit pieces of the commutative @xmath380 action as these are the simplices " in the original graph @xmath381 .
[ [ section-12 ] ] connection graphs have a high local connectivity : let @xmath0 be a connected graph which is not a complete graph and let @xmath75 is its connection graph .
for any vertex @xmath382 the fredholm determinant of @xmath383 is zero .
assume first that @xmath15 is not the maximal central simplex @xmath244 containing @xmath15 .
then , there are parallel columns @xmath140 in the extended adjacency matrix .
let @xmath89 be in @xmath218 .
then @xmath89 intersects @xmath15 and @xmath244 .
the only way , the unit ball @xmath218 in a connection graph can have nonzero fredholm determinant is if @xmath0 is a complete graph and @xmath15 is the maximal central simplex .
for example , for @xmath384 , the connection graph is @xmath276 .
the unit sphere of @xmath385 is @xmath386 which has fredholm determinant @xmath50 .
for all other @xmath15 , the unit ball is @xmath197 , which has zero fredholm determinant .
benjamin landon and ziliang che , members of the harvard random matrix group tell me that current technology of matrix theory like @xcite could be close in being able to prove a central limit theorem for determinants of random graphs but that the existence of a central limit in that graph case still remains to be done .
* 1 ) * let @xmath387 $ ] , where @xmath388 is the counting probability measure on the erds rnyi space @xmath389 .
we believe the limsup is actually a limit for every @xmath390 $ ] and that the limit is @xmath391 for every @xmath392 .
we know @xmath393 for @xmath394 because a graph @xmath0 without edges has @xmath227 and for @xmath395 , because a complete graph @xmath0 has @xmath228 .
why do we think that the limit in general is @xmath391 ?
because of the central limit theorem for @xmath342-valued random variables .
what happens if we go from @xmath389 to @xmath396 is that we are given a distribution of a @xmath342-valued random variable @xmath397 on subgraphs of @xmath0 get the new graph by adding a pyramid over @xmath144 . since we expect the probability of picking a graph @xmath144 with @xmath398 or @xmath399 is more and more @xmath391 , we essentially multiply with some independent @xmath342-valued random variable when adding a new vertex . by the central limit theorem of random variables taking values in compact groups with a haar measure
, the limiting distribution has to be the uniform distribution which means equal distribution on finite groups . the case
is not yet settled because we convolute each time with a distribution which we want to show to converge .
this requires some estimates .
of course , a situation with some nonzero thresholds with a probability switching from @xmath21 to @xmath391 would have been more exciting , but it seems not to happen here ; there is no reason in sight why the number of odd - dimensional subgraphs should pick a particular parity . + * 2 ) * it appears that both the function @xmath400 as well as @xmath401 have a continuous limiting distribution on random graphs when the functions are normalized so that the mean is @xmath21 and the standard deviation is @xmath50 . establishing this would settle a central limit theorem for determinants .
one usually looks also at the random variable @xmath402 when looking at limits but that requires first establishing that the determinant @xmath21 has asymptotically zero probability . +
* 3 ) * what is the structure of the inverse matrices of @xmath23 , where @xmath403 is the connection graph of the connection graph @xmath75 of @xmath0 ? here are examples : for many graphs , including the complete graph , the wheel graphs , cross polytopes or cycle graphs , the inverse of the matrix @xmath23 only takes the values @xmath404 . for * star graphs * with @xmath49 rays
, we see that the values of the inverse of @xmath405 is @xmath406 . for the utility graph , or the petersen graph ,
the values are @xmath407 . on the @xmath408 archimedian solids ,
we see 6 for which the value takes values in @xmath409 and @xmath410 , where the value takes value in @xmath411 . on the 13 catalan solids , the value cases are @xmath411 ,
@xmath412 , @xmath413 and @xmath414 occur .
+ * 4 ) * we measure that the * green function values * given by the inverse matrix of @xmath415_{ij}\ ] ] is the same for @xmath0 and the barycentric refinement @xmath74 of @xmath0 .
we also measure that maximal and minimal value among the diagonals , the * green function values * are independent of barycentric refinement .
the diagonal entries @xmath416 are up to a sign the fredholm determinants of the graph @xmath417 , where entry @xmath418 has been removed . if the entry @xmath416 is @xmath50 or @xmath48 , there is a chance that the modified graph is the connection graph of a cw complex .
the green function value @xmath419 is the fredholm determinant of the structure obtained by snapping the @xmath420 connection between cell @xmath105 and @xmath106 .
as the so lobotomized graph is no more a connection graph in general , the @xmath419 can be different .
we have no explanation yet for these measurements .
this would actually indicate that the minimal values taken for example are a * combinatorial invariants * of @xmath0 . and that one could assign the invariant to the limiting continuum object . following a suggestion of noam elkies , we looked also whether an edge refinement does not change the range of the green functions @xmath421_{ij}$ ]
. indeed also to be the case and a first thing which should be checked theoretically .
the invariance under edge refinements would imply topological invariance under barycentric refinements for graphs without triangles .
+ * 5 ) * because unimodular matrices form a group , we can define for all @xmath49 , the group of all fredholm connection matrices of subgraphs of a given graph @xmath0 .
assume @xmath49 is the number of simplices in @xmath0 .
if @xmath144 is a subgraph of @xmath0 , then the simplices are part of the simplices of @xmath0 .
all matrices therefore can be made the same size @xmath333 .
the set of all these fredholm connection matrices now forms a subgroup of the unimodular group in @xmath159 .
can we characterize these groups somehow ?
+ * 6 ) * we know that unit balls of connection graphs have fredholm characteristic @xmath21 and euler characteristic @xmath50 .
the structure of the unit spheres is less clear .
we measure so far that unit spheres of connection graphs either have euler characteristic @xmath50 or @xmath22 .
the later case is more rare but appears already for the octahedron graph @xmath0 .
lets see : given a simplex @xmath15 , then the unit sphere consists of all simplices which intersect @xmath15 .
now , there are three possibilities : either @xmath244 is a subgraph .
the set of subgraphs is are all connected with each other and we can homotopically reduce them to a point .
then there are all graphs which contain @xmath15 . also all these
are connected with each other and we can homotopically reduce them to a point . finally , there are all graphs which intersect .
we have to show that they are always homotopic to an even dimensional sphere . + * 7 ) * we tried to correlate @xmath99 with cohomological data , both for simplicial cohomology as well as connection cohomology .
simlilarly as homotopy deformations do change @xmath19 but not cohomology , this also appears to happen for connection cohomology even so the later is not a homotopy invariant : the cylinder and the mbius strip are homotopic but have different connection cohomology . + * 8) * what is the relation between the f - vector of @xmath0 and the f - vector of @xmath75 ?
unless for barycentric refinement , the @xmath172-vector of @xmath0 does not determine the @xmath172-vector of @xmath75 .
the euler characteristic is not an invariant as the octahedron graph @xmath0 for which the connection graph @xmath75 is contractible .
the following code is written in the mathematica language which knows graphs as a fundamental data structure .
we first compute the fredholm determinant @xmath422 of a graph @xmath0 with adjacency matrix @xmath1 . the procedure whitney " produces the whitney complex of a graph , the set of all complete subgraphs which is a finite abstract simplicial complex .
this procedure is then is used to get the connection graph of a graph .
we then compute @xmath99 and @xmath423 for 100 random graphs .
cliquenumber[s_]:=length[first[findclique[s ] ] ] ; listcliques[s_,k_]:=module[{n , t , m , u , r , v , w , u , l={},l},l = length ; vl = vertexlist;el = edgelist;v = vl[s];w = el[s ] ; m = l[w ] ; n = l[v ] ; r = subsets[v,{k , k}];u = table[{w[[j,1]],w[[j,2]]},{j , l[w ] } ] ; if[k==1,l = v , if[k==2,l = u , do[t = subgraph[s , r[[j ] ] ] ; if[l[el[t]]==k(k-1)/2,l = append[l , vl[t]]],{j , l[r]}]]];l ] ; whitney[s_]:=module[{f , a , u , v , d , v , lc , l = length},v = vertexlist[s ] ; d = if[l[v]==0,-1,cliquenumber[s]];lc = listcliques ; if[d>=0,a[x_]:=table[{x[[k]]},{k , l[x ] } ] ; f[t_,l_]:=if[l==1,a[lc[t,1]],if[l==0,{},lc[t , l ] ] ] ; u = delete[union[table[f[s , l],{l,0,d}]],1 ] ; v= { } ; do[do[v = append[v , u[[m , l]]],{l , l[u[[m]]]}],{m , l[u]}],v={}];v ] ; connectiongraph[s _ ] : = module[{c = whitney[s],n , a},n = length[c ] ; a = table[1,{n},{n}];do[if[disjointq[c[[k]],c[[l]]]|| c[[k]]==c[[l]],a[[k , l]]=0],{k , n},{l , n}];adjacencygraph[a ] ] ; fvector[s _ ] : = delete[bincounts[length /@ whitney[s ] ] , 1 ] ; fermicharacteristic[s_]:=module[{f = fvector[s ] } , ( -1)^sum[f[[2k]],{k , floor[length[f]/2 ] } ] ] ; fredholmcharacteristic[s_]:=fredholmdet[connectiongraph[s ] ] ; .... connectiongraph[s_,v_]:=module[{c = whitney[s],n , a},n = length[c ] ; a = table[1,{n+1},{n+1}];do[if[disjointq[c[[k]],c[[l ] ] ] || c[[k]]==c[[l]],a[[k , l]]=0],{k , n},{l , n } ] ; a[[n+1,n+1]]=0 ; do[if[disjointq[v , c[[k]]],a[[k , n+1]]=0 ; a[[n+1,k]]=0],{k , n } ] ; adjacencygraph[a ] ] ; eulerchi[s_]:=module[{c = whitney[s ] } , sum[(-1)^(length[c[[k]]]-1),{k , length[c ] } ] ] ; experiment3 = do [ s = randomgraph[{30,50 } ] ; v = vertexlist[s ] ; v = randomchoice[v , random[integer , length[v ] ] ] ; h = subgraph[s , v ] ; ss = connectiongraph[s ] ; sss = connectiongraph[s , v ] ; u={fredholmdet[ss],fredholmdet[sss],1-eulerchi[h ] } ; print[u , " " , u[[1 ] ] u[[3]]==u[[2]]],{100 } ] ; .... and here is an other experiment .
what is the probability of a random erds rnyi graph in @xmath389 to have connection fredholm determinant @xmath50 ? in this example , we experiment with @xmath424 and @xmath425 : .... erdoesrenyi[n_,p_]:=randomgraph[{n , floor[p n(n-1)/2 ] } ] ; g[n_,p_]:=fredholmdet[connectiongraph[erdoesrenyi[n , p ] ] ] ; n=12;p=0.6;k=0.;m=0 ; experiment4 = do[m++;if[g[n , p]==1,k++];print[k / m],{infinity } ] ; .... and here are the procedures producing the prime graphs @xmath285 and prime connection graph @xmath289 . in the first case
we verify numerically @xmath426 , in the second case we verify numerically @xmath427 , where @xmath428 is the mbius function of the integer @xmath41 . ....
primegraph[m_]:=module[{v={},e , s } , do[if[moebiusmu[k]!=0,v = append[v , k ] ] , { k,2,m}];e= { } ; do[if[(divisible[v[[k]],v[[l ] ] ] || divisible[v[[l]],v[[k ] ] ] ) , e = append[e , v[[k]]->v[[l]]]],{k , length[v]},{l , k+1,length[v ] } ] ; undirectedgraph[graph[v , e ] ] ] ; primeconnectiongraph[m_]:=module[{v={},e , s } , do[if[moebiusmu[k]!=0,v = append[v , k ] ] , { k,2,m}];e= { } ; do[if[gcd[v[[k]],v[[l]]]>1 || gcd[v[[l]],v[[k]]]>1 , e = append[e , v[[k]]->v[[l]]]],{k , length[v]},{l , k+1,length[v ] } ] ; undirectedgraph[graph[v , e ] ] ] ; test5[m_]:=module[{v , g , h , j } , h = primeconnectiongraph[m ] ; g = primegraph[m ] ; v = vertexlist[g ] ; j = table[-moebiusmu[v[[k]]],{k , length[v ] } ] ; print[{total[j ] , eulerchi[g ] , product[j[[k]],{k , length[j]}],fredholmdet[h ] } ] ] ; experiment5 = do[test5[k],{k,10,200 } ] ; .... r. bowen and o. e. lanford , iii .
zeta functions of restrictions of the shift transformation . in _ global analysis ( proc
pure math .
xiv , berkeley , calif . , 1968 )
_ , pages 4349 .
soc . , 1970 . | for a finite simple graph @xmath0 with adjacency matrix @xmath1 , the fredholm determinant @xmath2 is @xmath3 for the bowen - lanford zeta function @xmath4 of the graph .
the connection graph @xmath0 of @xmath0 is a new graph which has as vertices the set @xmath5 of all complete subgraphs of @xmath0 and where two such complete subgraphs are connected , if they have a non - empty intersection .
more generally , the connection graph of an abstract finite simplicial complex or even cw complex @xmath0 has as vertices the simplices or cells in @xmath0 , where two are connected if they intersect .
we prove that for any @xmath0 , the fredholm characteristic @xmath6 is equal to the fermi characteristic @xmath7 , where @xmath8 is the number of odd dimensional cells in @xmath0 ; the functional @xmath8 is a valuation for which poincar - hopf and gauss - bonnet formulas hold . given @xmath9 , we can see the fermi characteristic @xmath10 as a cousin of the euler characteristic @xmath11 which sums the signatures of simplices .
the main result is the unimodularity theorem @xmath12 which relates an algebraic and a combinatorial quantity .
we illustrate this with prime graphs , where @xmath13 is the mbius function of an integer . a key proposition for the proof of the theorem
is that if @xmath14 is the poincar - hopf index of @xmath15 , where @xmath16 is the unit sphere of @xmath15 , then @xmath17 .
if @xmath16 is a graph theoretical sphere , then @xmath18 proving inductively that @xmath19 is @xmath20-valued .
the unimodularity theorem follows then by induction by building up the simplicial complex cell by cell , using that spheres have euler characteristic @xmath21 or @xmath22 .
having established that the fredholm matrix @xmath23 of a simplicial complex @xmath0 is unimodular , the entries of the green function @xmath24_{ij}$ ] are always integers .
they appear of interest as experiments indicate that the range of @xmath25 is a combinatorial invariant for @xmath0 : we conjecture that global or local barycentric refinements of simplicial complexes @xmath0 do not change the range of @xmath25 . |
in a preceding paper,@xcite we performed a statistical mechanical investigation on rayleigh - bnard convection of a dilute classical gas based on the boltzmann equation .
we specifically calculated entropy change through the convective transition for the case of the stress - free boundaries as a function of macroscopic mechanical variables .
we thereby tested the validity of the principle of maximum entropy proposed for nonequilibrium steady states.@xcite the present paper extends the consideration to a more realistic case of the rigid boundaries.@xcite
our starting point is eq . (
46 ) of ref . : [ boussinesq ] @xmath0
@xmath1 where @xmath2 and the units are described in 3.1 . we adopt the condition of the rigid boundaries along @xmath3 , i.e. , @xmath4 at @xmath5 . combined with @xmath6 , it yields the boundary conditions : [ bc ] @xmath7 @xmath8 with @xmath9 denoting the @xmath10 components . as for the horizontal directions ,
we consider the region @xmath11 with @xmath12 and impose the periodic boundary conditions .
we also fix macroscopic mechanical variables of the system , i.e. , the total particle number , energy , and energy flux along @xmath3 .
these conditions lead to eq .
( 42 ) of ref .
, i.e. , [ boundary - t ] @xmath13
@xmath14 where @xmath15 denotes the temperature difference between @xmath5 that would be realized in the heat - conducting state . we solve the above equations with the method developed by pesch.@xcite first , the boundary conditions of eq .
( [ bc ] ) is treated with the galerkin method,@xcite i.e. , by expanding every @xmath3 dependence in terms of some basis functions satisfying the boundary conditions . specifically , the basis functions for eq .
( [ bc1 ] ) are obtained from the second - order differential equation @xmath16 with @xmath17 as [ bases ] @xmath18 they satisfy @xmath19 . on the other hand , those for eq .
( [ bc2 ] ) are constructed from the fourth - order differential equation @xmath20 with @xmath21 as @xmath22 where @xmath23 is determined by @xmath24 note @xmath25 for @xmath26 .
the quantity @xmath27 is the normalization constant : @xmath28 so that @xmath29 . the functions @xmath30 may be called chandrasekhar functions.@xcite now that appropriate basis functions are obtained , we expand @xmath31 and @xmath32 in eq .
( [ boussinesq ] ) as [ expansions ] @xmath33 @xmath34 @xmath35 here @xmath36 component is excluded in the expansion of @xmath32 to seek only periodic current distributions in the @xmath10 plane .
we have also incorporated into eq .
( [ expansionst ] ) the fact that the temperature is uniform at @xmath5 .
now , @xmath2 is transformed into @xmath37 .
it hence follows that @xmath38 can be expressed generally as @xmath39 on the other hand , eq .
( [ boundary - t ] ) is transformed into [ t1dt ]
@xmath40 @xmath41 let us substitute eq .
( [ expansions ] ) with eq .
( [ j_perp ] ) into eq .
( [ boussinesq ] ) and perform space integrations using the orthonormality of the basis functions .
we thereby obtain algebraic equations for the expansion coefficients as [ boussinesq - alg ] @xmath42\tilde{t } ( { \bm k}_{\perp},n ) \nonumber \\ & & \hspace{-10 mm } -\delta t \sum_{n ' } \langle s_{n}|c_{n'}\rangle \tilde{j}_{z}({\bm k}_{\perp},n ' ) \nonumber \\ & & \hspace{-10 mm } = -\frac{1}{l^{2}}\langle { \rm e}^{i{\bm k}_{\perp}\cdot{\bm r}}s_{n } | \hat{\bm j}^{(1.5)}\!\cdot{\bm\nabla}\tilde{t}^{(1)}\rangle , \label{boussinesq - alg1}\end{aligned}\ ] ] @xmath43 \tilde{j}_{z } ( { \bm k}_{\perp},n ' ) \nonumber \\ & & \hspace{-10 mm } -u_{g}^{(2)}k_{\perp}^{2}\sum_{n ' } \langle c_{n } |s_{n'}\rangle \tilde{t}({\bm k}_{\perp},n ' ) \nonumber \\ & & \hspace{-10 mm } = -\frac{1}{l^{2}}{\bm e}_{z}\cdot \langle { \rm e}^{i{\bm k}_{\perp}\cdot{\bm r}}c_{n } | { \bm\nabla}\!\times\!{\bm\nabla}\!\times ( \hat{\bm j}^{(1.5)}\!\cdot{\bm\nabla}\hat{\bm j}^{(1.5)})\rangle , \label{boussinesq - alg2}\end{aligned}\ ] ] @xmath44\frac{\partial \tilde{j}_{p } ( { \bm k}_{\perp},n)}{\partial t}+\nu^{(2 ) } [ k_{\perp}^{2}\!+\!(n\pi)^{2}]^{2}\tilde{j}_{p } ( { \bm k}_{\perp},n ) \nonumber \\ & & \hspace{-10mm}= -\frac{1}{l^{2}}\frac { { \bm e}_{z}\times { \bm k}_{\perp}}{k_{\perp}}\cdot \langle { \rm e}^{i{\bm k}_{\perp}\cdot{\bm r}}s_{n } | { \bm\nabla}\!\times\!{\bm\nabla}\!\times ( \hat{\bm j}^{(1.5)}\!\cdot{\bm\nabla}\hat{\bm j}^{(1.5)})\rangle , \nonumber \\ \label{boussinesq - alg3}\end{aligned}\ ] ] with @xmath45 . finally , entropy characteristic of convection is obtained by substituting eq .
( [ expansionst ] ) into eq .
( 50 ) of ref .
as @xmath46 with @xmath47 and @xmath48 given by eq .
( [ t1dt ] ) .
it follows from the stability analysis for the heat - conducting state@xcite that the critical rayleigh number @xmath49 is determined from eq .
( [ boussinesq - alg ] ) by setting the nonlinear terms and time derivatives equal to zero .
the relevant instability originates from the linear coupled equations for @xmath50 and @xmath51 . eliminating @xmath50 in favor of @xmath51
, we obtain the equation for @xmath49 as @xmath52 where matrix @xmath53 is defined by @xmath54 with @xmath55 the critical rayleigh number @xmath49 corresponds to the minimum value of @xmath56 in eq .
( [ deta ] ) as a function of @xmath57 .
equation ( [ deta ] ) is solved by approximating @xmath53 by a finite dimension of @xmath58 , and the convergence is checked by increasing @xmath59 . choosing @xmath60 already yields an excellent result of @xmath61 with @xmath62.@xcite
the nonlinear terms become relevant in eq .
( [ boussinesq - alg ] ) for @xmath63 .
they are evaluated for given expansion coefficients @xmath64 and @xmath65 as follows .
we first construct @xmath66 and @xmath67 by eqs .
( [ expansions])-([t1dt ] ) .
the fast fourier transform ( fft)@xcite is used in this procedure to obtain the @xmath10 dependence .
we then perform the space differentiations numerically in the @xmath10 plane and analytically along the @xmath3 direction .
we specifically use the following formulas of @xmath68 in the @xmath10 plane : [ nd ] @xmath69 , \label{nd1}\end{aligned}\ ] ] @xmath70 - 490f(x , y)\biggr\}. \label{nd2}\end{aligned}\ ] ] the quantity @xmath71 is obtained with eq .
( [ nd1 ] ) by averaging the derivatives performed in different order .
the nonlinear overlap integrals are evaluated finally , where we again use the fft in the @xmath10 plane . on the other hand ,
all the calculations along @xmath3 are performed by preparing the relevant overlap integrals in advance , e.g. , @xmath72 and @xmath73 for eq .
( [ boussinesq - alg1 ] ) , and performing the summations over @xmath74 and @xmath75 .
time evolutions of eq .
( [ boussinesq - alg ] ) are calculated as follows .
we first multiply eqs .
( [ boussinesq - alg2 ] ) and ( [ boussinesq - alg3 ] ) by @xmath76 and @xmath77^{-1}$ ] with @xmath78 , respectively , and perform summation over @xmath79 for eq .
( [ boussinesq - alg2 ] ) .
time integrations are then carried out numerically by treating @xmath80 as @xmath81 .
a disadvantage of this simple method is that we have to make @xmath82 small enough to avoid an explosion in the numerical time integration .
one may alternatively use the split - step integration scheme developed by pesch which approximates @xmath83 as @xmath84\delta t/2 $ ] , thereby treating the linear part @xmath85 exactly .
this latter scheme removes the explosion at the expense of larger numerical errors to make a rapid time integration possible ; the extra computational time for diagonalizing @xmath86 in the calculation of @xmath87 is negligible in the whole numerical procedures .
we here focus on periodic solutions of eq .
( [ boussinesq ] ) in the @xmath10 plane and express @xmath88 , where @xmath89 and @xmath90 denote the basic vectors .
accordingly , we adopt the periodic boundary condition for the region spanned by @xmath91 and @xmath92 with @xmath93 ( @xmath94 ) a large integer .
the above theoretical framework can also be used in this case with a minor modification .
indeed , we only have to perform the change of variables @xmath95 in the @xmath10 integrations of the nonlinear terms .
those integrations have to be carried out now only over the unit cell of @xmath96 .
the corresponding wave vector @xmath97 is given by @xmath98 , where @xmath99 $ ] , @xmath100 $ ] , and @xmath101 denotes an integer .
the linear stability analysis for the heat - conducting state suggests that the stable solution satisfies @xmath102 . the parameters in eq .
( [ boussinesq - alg ] ) are chosen the same as those used for the free boundaries , i.e. , eq .
( 37a ) of ref . , which correspond to ar at 273k under atmospheric pressure .
we also fix the heat - flux density @xmath103 at @xmath104 so that the temperature difference @xmath105k is realized between @xmath5 in the heat - conducting state .
the rayleigh number @xmath56 is controlled by changing the thickness @xmath106 .
practical calculations of eq .
( [ boussinesq - alg ] ) are performed as follows : we first multiply eqs .
( [ boussinesq - alg1])-([boussinesq - alg3 ] ) by @xmath107 , @xmath108 and @xmath108 , respectively , and rewrite them in terms of @xmath109 and @xmath110 to obtain equations of @xmath111 .
the summations over @xmath79 are truncated at a finite value @xmath59 , whereas @xmath112 discrete points are used to perform fft for each direction in the @xmath10 plane . as for periodic structures
, we investigate the three candidates : the roll , the square lattice and the hexagonal lattice with @xmath113 .
we then trace time evolutions of the expansion coefficients until they all acquire constant values . choosing @xmath114 , @xmath115 and @xmath116 yields excellent convergence for the calculations presented below even with the simplest time - integration scheme .
the initial state is chosen as the conducting state with small fluctuations @xmath117 for the basic harmonics @xmath97 .
the constants @xmath48 and @xmath118 are updated at each time step by using eq .
( [ t1dt ] ) . also evaluated at each time step is entropy measured with respect to the heat - conducting state : @xmath119 where @xmath120 is given by eq .
( [ s^2 ] ) and @xmath121 .
we thereby trace time evolution of @xmath122 simultaneously .
the above procedure is carried out for each fixed periodic structure .
one of the advantages of the present approach is that we only have to change the basis functions to study other boundary conditions .
for example , the case of the stress - free boundaries can also be treated within the present framework by simply changing @xmath123 and @xmath124 in the expansions of eqs .
( [ expansionsz ] ) and ( [ expansionsp ] ) , respectively .
this replacement has been checked to reproduce the results obtained in ref . appropriately . .
the letters r , s and h denote roll , square and hexagonal , respectively , distinguishing initial fluctuations around the heat - conducting solution ; see text for details .
the final state of @xmath125 is the roll convection , whereas the intermediate plateaus of s and h correspond to the square and hexagonal convections , respectively . ] .
the four curves correspond to the different rayleigh numbers : @xmath126 , @xmath127 , @xmath128 and @xmath129 .
the initial state is the heat - conducting state with the fluctuation @xmath130\!=\ !
1.0\times 10^{-2}/\sqrt{8}$ ] and @xmath131 , whereas all the final states are the roll convection .
the broken line near the top indicates the upper bound of @xmath122 . ]
we now present numerical results on the rigid boundaries , which turn out to be qualitatively the same as those of the stress - free boundaries.@xcite figure [ fig:1 ] shows time evolution of @xmath122 for the rayleigh number @xmath132 .
the letters r , s and h denote ( r ) roll , ( s ) square and ( h ) hexagonal , respectively , distinguishing initial conditions ; they are exactly the same as those for the stress - free boundaries.@xcite writing @xmath133 $ ] and introducing the angle @xmath134 by @xmath135 , those initial conditions are given explicitly as follows : ( r ) @xmath130\!=\!1.00\times 10^{-2}/\sqrt{8}$ ] with @xmath136 ; ( s ) @xmath130\!=\!1.01\times 10^{-2}/\sqrt{8}$ ] and @xmath137\!=\!0.99\times 10^{-2}/\sqrt{8}$ ] with @xmath138 and @xmath139 ; ( h ) @xmath130\!=\!\tilde{t}'[0,\pm 1,1]\!=\!1.00\times 10^{-2}/\sqrt{8}$ ] and @xmath140\!=\!\tilde{t}'[-1,-1,1]\!=\!1.01\times 10^{-2}/\sqrt{8}$ ] with latexmath:[$|{\bm b}_{1}|\!=\ ! observe clearly that entropy increases monotonically in all the three cases to reach a common final value of the roll convection . the intermediate plateaus seen in s and
h correspond to the metastable square and hexagonal lattices , respectively , with ( s ) @xmath130\!\sim\!\tilde{t}'[0,\pm 1,1]$ ] and ( h ) @xmath130\!\sim\!\tilde{t}'[0,\pm 1,1]\!\sim\!\tilde{t}'[\pm 1,\pm 1,1]$ ] .
figure [ fig:2 ] displays time evolution of @xmath122 for four different rayleigh numbers , all developing from the initial fluctuation @xmath130\!=\!1.00\times 10^{-2}/\sqrt{8}$ ] with @xmath136 .
each final state is the roll convection .
compared with the case of the stress - free boundaries,@xcite we observe an enhanced oscillatory behavior for @xmath142 and @xmath129 after the first rapid increase of @xmath122 . in the roll convection normalized by the temperature difference @xmath143 in the heat - conducting state .
the rayleigh numbers are @xmath144 , @xmath145 , @xmath127 , @xmath128 and @xmath129 from top to bottom on the left part . ]
figure [ fig:3 ] shows profile of the average temperature variation @xmath147 along @xmath3 in the roll convection for five different rayleigh numbers .
figure [ fig:4 ] plots @xmath148 corresponding to the maximum of @xmath122 as a function of @xmath146 . again
the basic features are qualitatively the same as those of the stress - free boundaries.@xcite thus , we have seen that the principle of maximum entropy proposed in ref .
is satisfied through the rayleigh - bnard convective transition of a dilute classical gas even in the realistic case of the rigid boundaries .
i would like to thank w. pesch for explaining his method to solve the boussinesq equations .
this work is supported in part by the 21st century coe program `` topological science and technology , '' hokkaido university . | the previous investigation on rayleigh - bnard convection of a dilute classical gas [ t. kita : j. phys . soc .
jpn .
* 75 * ( 2006 ) 124005 ] is extended to calculate entropy change of the convective transition with the rigid boundaries .
we obtain results qualitatively similar to those of the stress - free boundaries . above the critical rayleigh number ,
the roll convection is realized among possible steady states with periodic structures , carrying the highest entropy as a function of macroscopic mechanical variables . |
454 is a powerful flat - spectrum radio quasar located at a redshift @xmath3 .
this source is very bright in the gev energy range ; during strong flares , its apparent ( isotropic ) luminosity can reach @xmath4 erg s@xmath5 ( e.g. * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the mass of the central black hole ( bh ) in 454 is estimated in the range @xmath6 @xcite .
this implies an eddington luminosity @xmath7 , which is several orders of magnitude below @xmath8 .
although the large gap within @xmath9 and @xmath8 is naturally explained by relativistic doppler boosting , the estimates of the jet power during these flares appear , in any realistic scenario , close to or even larger than the eddington luminosity @xcite .
the gev emission from 454 is highly erratic , with variability timescales as short as 3 hr , as reported , in particular , for the december 2009 flare @xcite .
the most spectacular flare regarding both variability and gamma - ray luminosity was observed in november 2010 by and @xcite telescopes . during this high state , with the most active phase lasting for 5 days ,
the apparent luminosity in gev achieved @xmath10 . around the flare maximum ,
the rising time was @xmath11 hr , and the decay time , @xmath12 hr .
the detection of photons with energies up to @xmath13 gev , the short variability , and the contemporaneous x - ray flux constrain the doppler boosting of the emitter to @xmath14 to avoid severe internal @xmath15 absorption in the x - ray radiation field @xcite .
a remarkable feature of the gamma - ray emission from 454 is the so - called plateau phase revealed during the bright flare in 2010 .
it is characterized by a long - term brightening of the source , a few weeks before the appearance of the main flare .
such plateau states have been observed by for three flares ( e.g. * ? ? ?
* ; * ? ? ?
* ) , with the plateau emission being about an order of magnitude fainter than that of the main flare . remarkably , the rapid gamma - ray variability of 454 is accompanied by an activity at lower energies .
the simultaneous multiwavelength observations of the source during flares have revealed a strong correlation with optical and x - rays .
it has been interpreted as evidence that the gamma - ray source is located upstream from the core of the 43 ghz radio source , which is at a distance @xmath16 few pc from the central bh ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . over the recent years
, several works have attempted to explain the flaring gamma - ray activity of 454 within the framework of the standard synchrotron self - compton ( ssc ) or external inverse - compton ( eic ) models @xcite . in the ssc scenario , it is possible to reproduce the spectral energy distribution ( sed ) from optical wavelengths to gamma - rays . in these models
most of the jet power is ( unavoidably ) carried by protons , and only a small fraction is contained in relativistic electrons and the magnetic field .
the required proton - to - poynting flux ratio of @xmath17 is quite large .
such a configuration would be hard to reconcile , at least in the gamma - ray emitting region close to the central bh , with an undisturbed jet which is launched by the blandford - znajek @xcite type ( bz ) process , in which the luminosity of the jet is dominated by poynting flux and the jet consists of @xmath18-pairs . in this regard
we should mention that recent relativistic magnetohydrodynamical ( mhd ) simulations of jet acceleration yield much less efficiency of conversion of the magnetic energy into bulk motion kinetic energy ; these calculations @xcite predict a quite modest ratio @xmath19 .
the jet - rg interaction ( jrgi ) scenario has been invoked to explain the day - scale flares in the nearby non - blazar type agn m87 @xcite , it has been applied also to the tev blazar @xcite to demonstrate that the jet - driven acceleration of debris from the rg atmosphere can explain ultra - fast variability of very high - energy gamma - ray emission on timescales as short as @xmath20 s. a distinct feature of the jrgi scenario is the high magnetization ( @xmath21 ) of the relativistic flows located at sub - parsec distances , where the gamma - ray production supposedly takes place . although the strong magnetic field ,
@xmath22 g , dramatically reduces the efficiency of the inverse compton scattering of electrons , it opens an alternative channel of gamma - ray production through synchrotron radiation of protons @xcite .
the latter can be effectively realized only in the case of acceleration of protons to the highest possible energies , up to @xmath23ev .
thus the second ( somewhat hidden " ) requirement of this model is a very effective acceleration of protons with a rate close to the theoretical limit dictated by the classical electrodynamics @xcite .
it is interesting to note that also inverse compton models can be accommodated , at least in principle , in the jrgi scenario .
moreover , unlike most of the leptonic models of powerful blazars , in which the requirement of a very low magnetic field , implying a deviation from the equipartition condition by orders of magnitude , generally is not addressed and explained , the jrgi scenario can offer a natural way for leptonic models to be effective assuming that the gamma - ray emission is produced through the inverse compton scattering in shocked clouds originally weakly magnetized ( see * ? ? ? * ) . in this work , we show that the jrgi scenario gives a viable mechanism for the explanation of the flares seen in 454 .
we also argue that within this model the plateau state can form due to the interaction of the jet with a stellar wind excited by nonthermal ( accelerated ) particles that penetrate into the red giant atmosphere .
in the _ fast cooling regime _ , the proper intensity of the nonthermal emission , i.e. the intensity in the blob co - moving reference frame , is proportional to the energy released at the jet - blob interface .
this energy release can be characterized by a simple dynamical model , which describes the acceleration of the blob by the jet ram pressure . in this model
there are just a few relevant parameters that describe the basic properties of the jet and the blob : the jet ram pressure ( @xmath24 ) and bulk lorentz factor ( @xmath25 ) , and the blob mass ( @xmath26 ) and radius ( @xmath27 ; or , equivalently , its cross - section : @xmath28 ) ( for details , see * ? ? ?
the time dependence of the intensity of the jet / blob interaction corrected for the doppler boosting can be treated as a first - order approximation for the radiation lightcurve .
the mass of the cloud @xmath29 formed at the initial stage can be estimated by comparing the jet ram pressure , @xmath30 with the gravitational force . here , @xmath31 is the jet cross - section at the star crossing height .
this gives the following estimate : @xmath32 where @xmath33 and @xmath34 are the rg mass and radius , respectively . since the initial size of the expelled cloud should be comparable to the size of the star , it is possible to estimate the cloud expansion time as @xmath35 , where @xmath36 is the sound speed of the shocked material : @xmath37^{1/2}$ ] .
the cloud expansion time is @xmath38 where @xmath39 is the plasma adiabatic coefficient , and according to the rhd simulation by @xcite , a value of @xmath40 can be adopted for @xmath41 .
the blob acceleration occurs on a timescale of @xcite @xmath42 the @xmath43parameter that will be often used in the paper has a simple meaning .
it is a dimensionless inverse mass of the blob : @xmath44 the above timescale corresponds to the blob acceleration in the laboratory reference frame .
however , since the blob gets accelerated towards the observer , the emission delay , as seen by the observer , should be approximately corrected by a factor of @xmath45 .
thus , the observed peak of the emission should be delayed by a time interval of @xmath46 the emission produced by lighter clouds allows an estimation of the time required for the star to cross the jet .
once the star enters into the jet , the process of jet - star interaction should proceed steadily , with the production of these lighter clouds being roughly constant on average .
thus , the whole duration of the light cloud - associated emission , if observed , can be taken as a direct measurement of the jet crossing time @xmath47 , where @xmath48 is the star velocity . adopting the paradigm of magnetically - accelerated jets ,
it is possible to derive a very simple expression for this timescale : @xmath49 in this way , the duration of the jet - star interaction is determined by the interaction distance from the central bh . assuming a fixed efficiency @xmath50 in the blob reference frame for the transfer of jet power to nonthermal gamma rays ( where @xmath51 ) , and correcting for doppler boosting
, one can estimate the luminosity of a blob : @xmath52 where the correction function @xmath53 depends on time ; or , equivalently , on the blob location in the jet .
we note that the structure of the jet , i.e. , the dependence of the jet lorentz factor on @xmath54 , determines the actual dependence of @xmath53 on @xmath54 .
the maximum value of @xmath53 monotonically depends on the @xmath43 parameter , approaching a value of @xmath55 if @xmath56 and being @xmath57 for @xmath58 .
this relatively weak dependence allows us to derive the maximum blob luminosity .
also it is possible to obtain an estimate of the total energy emitted by a blob or an ensemble of sub - blobs as a result of the fragmentation of the original cloud ( @xmath59 ) , @xmath60 which accounts for the total energy transferred by the jet to a blob during the acceleration process , @xmath61 , and for the anisotropy of the emission due to relativistic effects represented by the factor @xmath62 .
the total apparent energy of the gev gamma - ray radiation detected during the flare observed from 454 in november 2010 was about @xmath63 .
the exceptionally high flux during this period allows the derivation of a very detailed lightcurve , as seen from figure 1 in @xcite .
the nonthermal activity lasted for @xmath64 days .
the onset of the activity period was characterized by a plateau stage . during the first @xmath65 days ,
a rather steady flux was detected , with an apparent luminosity @xmath66 erg s@xmath5 .
the plateau stage was followed by an exceptionally bright flare , the total duration which was @xmath67 days , with a rise time of @xmath68 h. the maximum flux reached was @xmath69 , which corresponds to a luminosity of @xmath70 .
the final stage of the flare phase was characterized by variable emission with a flux approximately a factor of @xmath71 weaker than the main flare , but still a factor of @xmath72 above the plateau level .
the observed luminosity of the plateau phase allow us to determine a lower limit on the star mass - loss rate , which can be derived by differentiating equation ( [ eq_toten ] ) : @xmath73 where @xmath74 . to evaluate the feasibility of the jrgi scenario for the 454 main flare , it is necessary to check whether the flux , the total energy release , and the flare delay with respect to the onset of the plateau , are well described by equations [ eq_delay ] , [ eq_lum ] , and [ eq_toten ] for a reasonable choice of jet / star properties .
a total energy budget of the flare of @xmath75 is feasible , according to equation ( [ eq_toten ] ) , if @xmath76 where @xmath77 g is the mass of the blown up rg envelop ( i.e. the initially formed cloud ) .
this requirement appears to be very close to the one provided by equation [ dms ] : @xmath78 where @xmath79 and @xmath80 , respectively .
the second term in equation [ eq_delay ] is expected to be short compared to the duration of the plateau phase , even for @xmath81 , and thus the duration of the initial plateau phase constrains the expansion time ( see equation ( [ eq_expansion ] ) ) : @xmath82 @xmath83 @xcite found that the emission of the main flare consisted of 5 components ( see figure 2 in that work ) : a nearly steady contribution , like a smooth continuation of the plateau emission , and 4 sub - flares of similar duration and energetics . in the framework of the jrgi scenario ,
such a description is very natural .
the steady component would be attributed to light clouds , continuously ejected by the star , and the four sub - flares would correspond to much heavier blobs formed out of the blown - up stellar envelop during the initial stage . on the other hand , the decomposition of the main flare in four sub - flares implies a strict limitation on the variability timescale .
the flare rise / decay timescales should be longer than the blob light crossing time corrected for the doppler boosting .
since the shortest variability scale was @xmath84 s , the maximum possible size of the emitting blobs can be estimated as : @xmath85 if the jet is magnetically driven , this size constraint can be expressed through the mass of the central bh : @xmath86 which is restrictive only in the case of @xmath87 . for @xmath88 , the blobs can cover the entire jet without violating the causality constraint . in summary ,
if the flare detected with was produced by an rg entering into the jet , the jet properties should satisfy to the restrictions imposed by the flux level , total energy release , and the duration of the plateau stage , respectively .
interestingly , this set of equations allows the derivation of a _
unique _ solution , which can constrain all the key parameters through the value of the @xmath43 parameter : @xmath89 @xmath90 @xmath91 and @xmath92 the lower limit on the jet luminosity is @xmath93 which exceeds the eddington limit for the mass of the central bh @xmath94 . to assess the feasibility of such a strongly super - eddington jet remains out of the scope of this paper , although we note that @xcite have presented observational evidence indicating that such jets may not be uncommon .
the coherent picture emanating from the jet properties derived above suggests that the jrgi scenario can be responsible for the flare detected from 454 for a solution of the problem with a reasonable set of model parameters .
this solution is designed to satisfy the requirements for ( i ) the total energy ; ( ii ) the peak luminosity ; and ( iii ) the duration of the plateau phase .
therefore , some additional observational tests are required to prove the feasibility of the suggested scenario .
finally , the flare raise time , which is related to the blob acceleration timescale ( see equation [ dtpeak ] ) , can be calculated for the obtained solution .
interestingly , in the limit of small @xmath43-values , this timescale appears to be independent on @xmath43 , the only remaining free parameter , and matches closely the detected raise time of @xmath95 : @xmath96 this agreement can be treated as a cross - check that shows the feasibility of the proposed scenario .
to check whether jrgi plus synchrotron radiation can explain the observations in the case of magnetically dominated jets ( i.e. , @xmath97 ) , we have computed the lightcurve of the november 2010 flare and the sed for one of its subflares .
the radiation output was assumed to be dominated by proton synchrotron , being external or synchrotron self - compton neglected due to the strong magnetic field . to derive the lightcurve ,
equation [ eq_lum ] has been used . in figure
[ lc ] , a computed lightcurve that approximately mimics the november 2010 flare is presented .
the lightcurve has been obtained assuming four subflares of total ( apparent ) energy of @xmath98 erg each , plus a plateau component with luminosity of @xmath99 erg s@xmath5 . for each subflare ,
we have adopted @xmath58 .
the normalization of the lightcurve has been determined adopting the following values : the lorentz factor @xmath100 , the ram pressure @xmath101 erg @xmath102 , blob radius @xmath103 cm and @xmath104 .
these parameters imply a minimum jet luminosity of @xmath105 erg s@xmath5 .
the remaining parameters for the emitter are @xmath106 cm and @xmath107 g. the corresponding mass of the matter lost by the rg to explain the four subflares is @xmath108 g , not far from the upper - limit given in equation [ dms ] . to calculate the sed ,
we have adopted a spectrum for the injected protons @xmath109 , and an homogeneous ( one - zone ) emitter moving towards the observer with lorentz factor @xmath110 .
the minimum proton lorentz factor has been taken equal to the shock lorentz factor in the blob frame , i.e. @xmath111 . the cutoff energy , @xmath112 , has been obtained fixing @xmath113 , i.e. , a relatively modest acceleration efficiency . for the maximum proton energy , i.e.
how far beyond the cutoff the proton energy is considered , we adopted two values : @xmath114 and @xmath115 . regarding the latter case , we note that assuming a sharp high - energy cut is very natural .
the injection spectrum was selected to be hard , @xmath116 , to optimize the required energetics . in figure
[ sed ] , the sed of a subflare is shown .
the impact of the internal absorption on the gamma - ray spectrum is negligible , although the emission of the secondary pairs appears in the energy band constrained by optical measurements @xcite . for the chosen model parameters , the synchrotron secondary component goes right through the optical observational constraints , and for slightly higher @xmath117-values
, the secondary emission will be well below the optical points . also , we note that the obtained spectrum does not violate the x - ray upper - limits obtained by _
swift_. to illustrate the impact of external @xmath15 absorption , we have introduced a photon field peaking at 40 ev with a luminosity @xmath118 erg s@xmath5 , produced in a ring with radius @xmath119 cm at @xmath120 around the jet base .
two photon fields have been adopted , a black body and one represented by a @xmath121-function , to simulate the impact of a dominant spectroscopic line . as seen in figure [
sed ] , the impact of such an external field can be important .
the treatment of the secondary emission of the produced pairs is beyond the scope of this work .
in addition to optical photons , radio emission was also detected at the flare epoch and thought to be linked to the gamma - ray activity @xcite .
this radiation is strongly sensitive to the details of the flow dynamics , and at this stage we will not try to interpret radio observations .
however , we note that the energetics involved in gamma - ray production is very large , and jrgi comprehends complex magnetohydrodynamical and radiative processes , so it could easily accommodate the presence of a population of radio - emitting electrons at @xmath122 .
_ swift _ x - rays could be also linked to the jrgi activity .
x - rays may come from secondary pairs produced via pair creation , or from a primary population of electrons(/positrons ) .
however , as with radio data , given the complexity of the problem we have not tried at this stage to explain the x - ray emission contemporaneous to the gev flare .
erg s@xmath5 , @xmath123 cm @xmath124 , @xmath125 g , @xmath126 cm , and @xmath104 .
we show 4 subflares ( dashed lines ) , plateau background ( dot - dashed line ) , and the sum of all of them ( solid line ) .
the observational data points and error bars are obtained from the 3h binned count rates and photon index using luminosity distance of @xmath127 and assuming a pure powerlaw spectrum between 0.1 and 5 gev.,scaledwidth=43.0% ] . dotted and dot - dot - dashed line shows gamma - ray spectra corrected for internal absorption only for @xmath114 and @xmath115 , respectively .
the thin solid and the dot - dashed lines correspond to the cases when absorption is dominated by a black body and a monoenergetic photon field , respectively .
the computed synchrotron sed of the secondary pairs produced via internal pair creation is also shown ( dotted line ) .
the parameters of the flare are the same as in figure [ lc ] .
the shown observational data are from , swift @xcite , and the flux in the r band @xcite.,scaledwidth=43.0% ]
the observations of 454 with _ fermi
_ revealed several quite puzzling features , in particular the peculiar lightcurve , with a nearly steady plateau phase that was interrupted by an exceptionally bright flare .
the detected flux corresponds to an apparent luminosity of @xmath128 , which almost unavoidably implies a presence of a very powerful jet ( see e.g. * ? ? ?
* ; * ? ? ?
* ) . in the case of powerful jets , the jrgi scenario should proceed in a quite specific way as compared to other cases already considered in the literature @xcite . in particular , the mass of the material initially removed from the star might be very large , resulting in rather long cloud expansion and acceleration timescales , the main flare being significantly delayed with respect to the moment of the star entrance into the jet .
the plateau emission would otherwise start just after the jet penetration , and come from the jet crushing of lighter clouds ejected from the stellar surface while the star travels through the jet .
the duration of the plateau phase would be determined by the time required by the main cloud to expand and accelerate .
we have studied the lightcurve obtained with _
fermi _ in the context of the jrgi scenario aiming to satisfy three main properties of the flare : total energy , maximum luminosity and duration of the plateau stage .
it was shown that the key properties of the jet , i.e. the jet ram pressure ( linked to its luminosity ) and lorentz factor , as well as the cloud / blob characteristics , i.e. mass and cross section , can be reconstructed as functions of the dimensionless parameter @xmath43 .
it was also shown that in the limit of small @xmath43-values , the parameter space is less demanding concerning the jet luminosity , and the key characteristics of the model saturate at values independent of @xmath43 , which allows conclusive cross - checks of the scenario .
in particular , the flare raise time appeared to be an independent parameter , with its value of @xmath129h closely matching the rising time of of @xmath130h obtained observationally .
furthermore , it was shown ( see @xcite ) that for the inferred jet properties the jet - induced stellar wind can provide a mass - loss rate large enough to generate a steady emission component with a luminosity comparable to that of the plateau .
although the analysis of different radiation channels involves additional assumptions regarding the spectrum of the nonthermal particles and density of the target fields , it was possible to show that for magnetic fields not far below equipartition ( as expected in a magnetically launched jet ) all the conventional radiation channels can be discarded , and the emission detected with _
fermi _ can be produced through proton synchrotron emission ( unless @xmath131 , making electron synchrotron also feasible ) .
we note that in this case the emission from pairs created within the blob may also explain the reported optical enhancement at the flare epoch . since the duration of the expansion phase determines the delay between the onset of the plateau phase and the flare itself , it is important to check whether the suggested scenario is consistent with other flares registered with _
fermi _ from the source , e.g. , in december 2009 and april 2010 @xcite .
this issue can be addressed through a simple scaling that relates the duration of the plateau phase to the total energy released during the active phase : @xmath132 .
therefore , for the previous events , with energy releases 1 - 2 orders of magnitude smaller than that of the november 2010 flare , a rough estimation of the plateau duration gives plateau durations between @xmath133 and @xmath134 days , consistent with observations .
the authors are thankful to s. kelner for useful discussions and benoit lott for kindly provided observational data .
bmv acknowledge partial support by jsps ( japan society for the promotion of science ) : no.2503786 , 25610056 , 26287056 , 26800159 .
bmv also mext ( ministry of education , culture , sports , science and technology ) : no.26105521 and rfbr grant 12 - 02 - 01336-a . v.b .-
r . acknowledges support by the spanish ministerio de ciencia e innovacin ( micinn ) under grants aya2010 - 21782-c03 - 01 and fpa2010 - 22056-c06 - 02 . v.b .-
r . acknowledges financial support from mineco through a ramn y cajal fellowship . this research has been supported by the marie curie career integration grant 321520 | we propose a model to explain the ultra - bright gev gamma - ray flares observed from the blazar 454 .
the model is based on the concept of a relativistic jet interacting with compact gas condensations produced when a star ( red giant ) crosses the jet close to the central black hole .
the study includes an analytical treatment of the evolution of the envelop lost by the star within the jet , and calculations of the related high - energy radiation @xcite .
the model readily explains the day - long , variable on timescales of hours , gev gamma - ray flare from 454 , observed during november 2010 on top of a weeks - long plateau . in the proposed scenario ,
the plateau state is caused by a strong wind generated by the heating of the star atmosphere by nonthermal particles accelerated at the jet - star interaction region .
the flare itself could be produced by a few clouds of matter lost by the red giant after the initial impact of the jet . in the framework of the proposed scenario ,
the observations constrain the key model parameters of the source , including the mass of the central black hole : @xmath0 , the total jet power : @xmath1 , and the doppler factor of the gamma - ray emitting clouds , @xmath2 .
whereas we do not specify the particle acceleration mechanisms , the potential gamma - ray production processes are discussed and compared in the context of the proposed model .
we argue that synchrotron radiation of protons has certain advantages compared to other radiation channels of directly accelerated electrons .
4543c454.3 501mkn 501 |
for an r.v . @xmath0 with density @xmath2 , mean @xmath3 and finite variance @xmath4 , goldstein and reinert @xcite showed the identity ( see also @xcite ) @xmath5,\ ] ] which holds for any absolutely continuous function @xmath6 with a.s .
derivative @xmath7 such that the right - hand side is finite . in ( [ eq1.1 ] )
, @xmath8 is defined to be the r.v . with density @xmath9 , @xmath10
. identity ( [ eq1.1 ] ) extends the well - known stein identity for the standard normal @xcite ; a discrete version of ( [ eq1.1 ] ) can be found in , for example , @xcite , where the derivative has been replaced by the forward difference of @xmath11 . in particular
, identities of the form ( [ eq1.1 ] ) have many applications to variance bounds and characterizations @xcite , and to approximation procedures @xcite .
several extensions and applications can be found in @xcite . in @xcite
, the ( continuous ) pearson family is parametrized by the fact that there exists a quadratic @xmath12 such that @xmath13 typically , the usual definition of a pearson r.v . is related to the differential equation @xmath14 , with @xmath15 being a polynomial of degree at most 2 .
in fact , the set - up of ( [ eq1.2 ] ) ( including , e.g. , the standard uniform distribution with @xmath16 ) will be the framework of the present work and will hereafter be called `` the pearson family of continuous distributions '' .
it is easily seen that under ( [ eq1.2 ] ) , the support of @xmath0 , @xmath17 , must be an interval , say @xmath18 with @xmath19 , and @xmath20 remains strictly positive for @xmath21 . clearly , under ( [ eq1.2 ] ) , the covariance identity ( [ eq1.1 ] ) can be rewritten as @xmath22={\mathbb{e}}[q(x)g'(x)].\ ] ] it is known that , under appropriate moment conditions , the functions @xmath23,\qquad x\in(r , s),\ k=0,1,\ldots , m\ ] ] ( where @xmath24 can be finite or infinite ) are orthogonal polynomials with respect to the density @xmath2 so that the quadratic @xmath20 in ( [ eq1.2 ] ) generates a sequence of orthogonal polynomials by the rodrigues - type formula ( [ eq1.4 ] ) .
in fact , this approach is related to the sturm
liouville theory , @xcite , section 5.2 ; see also @xcite . in the present paper
, we provide an extended stein - type identity of order @xmath1 for the pearson family .
this identity takes the form @xmath25= { \mathbb{e}}\bigl [ q^k(x)g^{(k)}(x)\bigr],\ ] ] where @xmath26 is the @xmath1th derivative of @xmath11 ( since @xmath27 , ( [ eq1.5 ] ) for @xmath28 reduces to ( [ eq1.3 ] ) ) . identity ( [ eq1.5 ] ) provides a convenient formula for the @xmath1th fourier coefficient of @xmath11 , corresponding to the orthogonal polynomial @xmath29 in ( [ eq1.4 ] ) . for its proof
, we make use of a novel `` rodrigues inversion '' formula that may be of some interest in itself .
an identity similar to ( [ eq1.5 ] ) holds for the discrete pearson ( ord ) family .
application of ( [ eq1.5 ] ) and its discrete analog yields the corresponding lower variance bounds , obtained in section [ sec4 ] . the lower bound for the @xmath30 distribution , namely , @xmath31\ ] ] ( cf .
@xcite ) and the corresponding one for the normal@xmath32 distribution @xcite , @xmath33,\ ] ] are particular examples ( examples [ ex4.1 ] and [ ex4.5 ] ) of theorems [ th4.1 ] and [ th4.2 ] , respectively .
both ( [ eq1.6 ] ) and ( [ eq1.7 ] ) are particular cases of the finite form of bessel s inequality and , under completeness , they can be extended to the corresponding parseval identity . in section [ sec5 ] , we show that this can be done for a fairly large family of r.v.s , including , of course , the normal , the poisson and , in general , all the r.v.s of the pearson system which have finite moments of any order . for instance , when @xmath0 is @xmath34 , inequality ( [ eq1.7 ] ) ( and identity ( [ eq1.5 ] ) ) can be strengthened to the covariance identity @xmath35= \sum_{k=1}^{\infty } \frac{(\sigma^2)^{k}}{k ! } { \mathbb{e}}\bigl[g_1^{(k)}(x)\bigr ] { \mathbb{e}}\bigl[g_2^{(k)}(x)\bigr],\ ] ] provided that for @xmath36 , @xmath37 , @xmath38 , @xmath39 and that @xmath40 ^ 2<\infty$ ] .
similar identities hold for poisson , negative binomial , beta and gamma distributions . these kinds of variance / covariance expressions
may sometimes be useful in inference problems see , e.g. , the applications [ ap5.1 ] and [ ap5.2 ] at the end of the paper .
in order to simplify notation , we assume that @xmath0 is a non - negative integer - valued r.v . with mean @xmath41 .
we also assume that there exists a quadratic @xmath42 such that @xmath43 where @xmath44 is the probability function of @xmath0 .
relation ( [ eq2.1 ] ) describes the discrete pearson system ( ord family ) @xcite .
let @xmath45 be the forward difference operator defined by @xmath46=g(x+1)-g(x)$ ] and @xmath47=\delta[\delta^{k-1}[g(x)]]$ ] ( @xmath48\equiv g$ ] , @xmath49 ) .
we also set @xmath50}(x)=q(x)q(x+1)\cdots q(x+k-1)$ ] ( with @xmath51}\equiv1 $ ] , @xmath52}\equiv q$ ] ) .
we first show some useful lemmas .
[ le2.1 ] if @xmath53 for @xmath54 and @xmath55 \delta^{k - j}[g(x)]|&<&\infty \qquad\mbox{for } j=0,1,\ldots , k , \\
\label{eq2.3}\lim_{x\to\infty } \delta^j [ h(x - j ) ] \delta^{k - j-1}[g(x)]&=&0\qquad \mbox{for } j=0,1,\ldots , k-1,\end{aligned}\ ] ] then @xmath56g(x ) \nonumber \\[-8pt ] \\[-8pt ] \nonumber & & \quad= \sum_{x=0}^{\infty } h(x)\delta^k[g(x)].\end{aligned}\ ] ] we have @xmath57 & = & \lim_{n\to\infty } \sum_{x=0}^n h(x ) \bigl(\delta^{k-1}[g(x+1)]-\delta ^{k-1}[g(x)]\bigr ) \\ & = & \lim_{n\to\infty } \biggl [ h(n+1)\delta^{k-1}[g(n+1)]-\sum_{x=0}^{n+1 } \delta[h(x-1 ) ] \delta^{k-1}[g(x ) ] \biggr ] \\ & = & \lim_{n\to\infty } h(n+1)\delta^{k-1}[g(n+1)]-\sum_{x=0}^{\infty } \delta[h(x-1 ) ] \delta^{k-1}[g(x ) ] \\ & = & -\sum_{x=0}^{\infty } \delta[h(x-1 ) ] \delta^{k-1}[g(x)].\end{aligned}\ ] ] by the same calculation , it follows that @xmath58 \delta^{k - j}[g(x)]=(-1)^{j+1 } \sum_{x=0}^{\infty } \delta^{j+1 } [ h(x - j-1 ) ] \delta^{k - j-1}[g(x)]\ ] ] for any @xmath59 .
[ le2.2 ] for each @xmath60 , there exist polynomials @xmath61 , @xmath62 , such that the degree of each @xmath63 is at most @xmath64 and @xmath65}(x - n)p(x - n)\bigr ] = q^{[n - i]}(x - n+i)p(x - n+i)q_{i , n}(x),\qquad i=0,1,\ldots , n.\hspace*{26pt}\end{aligned}\ ] ] moreover , the leading coefficient ( i.e. , the coefficient of @xmath66 ) of @xmath67 is given by @xmath68 , where an empty product should be treated as @xmath69 . for @xmath70 , the assertion is obvious and @xmath71 . for @xmath72 ,
the assertion follows from the assumption ( [ eq2.1 ] ) with @xmath73 , @xmath74 . for the case
@xmath75 , the assertion will be proven using ( finite ) induction on @xmath64 .
indeed , for @xmath76 , ( [ eq2.5 ] ) holds with @xmath77 . assuming that the assertion holds for some @xmath78 and setting @xmath79}(x - n)p(x - n)$ ] , it follows that @xmath80 \\ & = & \delta[h_{n - i}(x ) q_{i , n}(x ) ] \\ & = & \delta\bigl [ q(x - n+i)p(x - n+i)\bigl(q^{[n - i-1]}(x - n+i+1)q_{i , n}(x)\bigr)\bigr ] \\ & = & q(x - n+i+1)p(x - n+i+1 ) \delta\bigl[q^{[n - i-1]}(x - n+i+1)q_{i , n}(x)\bigr ] \\ & & { } + \bigl(\mu-(x - n+i+1)\bigr)p(x - n+i+1)q^{[n - i-1]}(x - n+i+1)q_{i , n}(x),\end{aligned}\ ] ] where the obvious relation @xmath81=(\mu-(x+1))p(x+1)$ ] , equivalent to ( [ eq2.1 ] ) , has been used with @xmath82 in place of @xmath83 .
moreover , @xmath84}(x - n+i+1)q_{i , n}(x)\bigr]\\ & & \quad= q^{[n - i-1]}(x - n+i+2)\delta[q_{i , n}(x ) ] + q_{i , n}(x)\delta\bigl[q^{[n - i-1]}(x - n+i+1)\bigr]\end{aligned}\ ] ] and @xmath85}(x - n+i+1)\bigr ] = q^{[n - i-2]}(x - n+i+2 ) \bigl(q(x)-q(x - n+i+1)\bigr)\ ] ] ( where , for @xmath86 , the right - hand side of the above should be treated as @xmath87 ) . observing that @xmath88}(x - n+i+2)\bigl(q(x)-q(x - n+i+1)\bigr )
\\ & & \quad = q^{[n - i-1]}(x - n+i+1)\bigl(q(x)-q(x - n+i+1)\bigr)\end{aligned}\ ] ] ( which is also true in the case where @xmath86 ) , the above calculations show that ( [ eq2.5 ] ) holds with @xmath89,\ ] ] where @xmath90 is a linear polynomial or a constant . from the above recurrence , it follows immediately that @xmath91 , @xmath92 ; this , combined with the fact that @xmath93 , yields the desired result .
it is easy to see that under ( [ eq2.1 ] ) , the support of @xmath0 , @xmath94 , is a finite or infinite integer interval .
this integer interval will be denoted by @xmath95 . here
, the term `` integer interval '' means that `` if @xmath96 and @xmath97 are integers belonging to @xmath95 , then all integers between @xmath96 and @xmath97 also belong to @xmath95 '' .
[ le2.3 ] for each @xmath98 define the functions @xmath99 , @xmath100 , by the rodrigues - type formula @xmath101}(x - k)p(x - k)\bigr ] \nonumber \\[-8pt ] \\[-8pt ] \nonumber & = & \frac{1}{p(x)}\sum_{j=0}^k ( -1)^{k - j}{k\choose j}q^{[k]}(x - j)p(x - j).\end{aligned}\ ] ] we then have the following : each @xmath29 is a polynomial of degree at most @xmath1 , with @xmath102 .
let @xmath11 be an arbitrary function defined on @xmath95 and , if @xmath95 is infinite , assume in addition that the functions @xmath11 and @xmath103}(x)p(x)$ ] satisfy the requirements and of lemma [ le2.1 ] .
then , @xmath104}(x)|\delta^{k}[g(x)]|\bigr]<\infty\ ] ] and the following identity holds : @xmath105={\mathbb{e}}\bigl[q^{[k]}(x)\delta^{k}[g(x)]\bigr].\ ] ] part ( a ) follows from lemma [ le2.2 ] since @xmath106 . for part ( b ) , assume first that @xmath107 , that is , the support @xmath95 is either @xmath108 or @xmath109 for a positive natural number @xmath110 . in the unbounded case , ( [ eq2.7 ] ) follows by an application of lemma [ le2.1 ] to the functions @xmath111}(x)p(x)$ ] and @xmath11 since they satisfy the conditions ( [ eq2.2 ] ) and ( [ eq2.3 ] ) .
for the bounded case , it follows by ( [ eq2.1 ] ) that @xmath112 is a zero of @xmath20 so that @xmath113 , where , necessarily , @xmath114
. thus , in the case where @xmath115 , @xmath50}(x)=0 $ ] for all @xmath100 so that the right - hand side of ( [ eq2.7 ] ) vanishes .
on the other hand , the left - hand side of ( [ eq2.7 ] ) ( with @xmath29 given by ( [ eq2.6 ] ) ) is @xmath116}(x - j)p(x - j)$ ] and it is easy to verify that for all @xmath117 and all @xmath118 , the quantity @xmath50}(x - j)p(x - j)$ ] vanishes .
thus , when @xmath119 , ( [ eq2.7 ] ) holds in the trivial sense @xmath120 . for @xmath121 , the left - hand side of ( [ eq2.7 ] ) equals @xmath122}(x ) p(x)g(x+j)\\ & & \quad= \sum_{j=0}^k(-1)^{k - j } \pmatrix{k\cr j } \sum_{x=0}^{n - j } q^{[k]}(x)p(x)g(x+j ) \\ & & \quad= \sum_{x=0}^{n } q^{[k]}(x ) p(x)\sum_{j=0}^{\min\{k , n - x\}}(-1)^{k - j } \pmatrix{k\cr j}g(x+j ) \\ & & \quad= \sum_{x=0}^{n - k } q^{[k]}(x ) p(x ) \sum_{j=0}^k ( -1)^{k - j } \pmatrix{k\cr j}g(x+j ) , \ ] ] where we have made use of the facts that @xmath123 for @xmath54 and @xmath50}(x)=0 $ ] for @xmath124 .
thus , ( [ eq2.7 ] ) is proved under the assumption @xmath107 . for the general case , where the support of @xmath0 is either @xmath125 or @xmath126
, it suffices to apply the same arguments to the r.v . @xmath127 .
[ th2.1 ] suppose that @xmath0 satisfies and has @xmath128 finite moments for some @xmath129 .
the polynomials @xmath29 , @xmath130 , defined by , then satisfy the orthogonality condition @xmath131=\delta_{k , m } k!{\mathbb{e}}\bigl[q^{[k]}(x)\bigr]\prod_{j = k-1}^{2k-2}(1-j\delta),\qquad k , m=0,1,\ldots , n , \ ] ] where @xmath132 is kronecker s delta .
let @xmath133 .
first , observe that @xmath50}$ ] is a polynomial of degree at most @xmath134 and @xmath135 a polynomial of degree at most @xmath136 .
note that the desired result will be deduced if we can apply ( [ eq2.7 ] ) to the function @xmath137 ; this can be trivially applied when @xmath95 is finite . in the remaining case where @xmath95 is infinite
, we have to verify conditions ( [ eq2.2 ] ) and ( [ eq2.3 ] ) of lemma [ le2.1 ] for the functions @xmath138}p$ ] and @xmath137 , or , more generally , for any polynomial @xmath11 of degree less than or equal to @xmath1 . since the case @xmath139 is obvious ( @xmath140 ) , we assume that @xmath141 .
observe that ( [ eq2.2 ] ) is satisfied in this case because of the assumption @xmath142 . indeed , from lemma [ le2.2 ] , @xmath143}(x - j)p(x - j)]=p(x ) q^{[k - j]}(x)q_{j , k}(x - j+k)$ ] , where @xmath144 is of degree at most @xmath145 and @xmath146}$ ] is of degree at most @xmath147 so that their product is a polynomial of degree at most @xmath148 , while @xmath149 $ ] is a polynomial of degree at most @xmath145 . on the other hand ,
since @xmath150 is eventually decreasing and @xmath151 , we have , as @xmath152 , that @xmath153 hence , @xmath154 . for @xmath155
, we have , from lemma [ le2.2 ] , that @xmath156 \delta^{k - j-1}[g(x)]=q(x)p(x)r(x ) , \ ] ] where @xmath157}(x+1)q_{j , k}(x - j+k)\delta^{k - j-1}[g(x ) ] \ ] ] is a polynomial of degree at most @xmath158 .
this , combined with the above limit , verifies ( [ eq2.3 ] ) and a final application of ( [ eq2.7 ] ) completes the proof . the covariance identity ( [ eq2.7 ] ) enables the calculation of the fourier coefficients of any function @xmath11 in terms of its differences @xmath159 $ ] , provided that conditions ( [ eq2.2 ] ) and ( [ eq2.3 ] ) are fulfilled for @xmath11 and @xmath111}(x)p(x)$ ] . since this identity is important for the applications , we state and prove a more general result that relaxes these conditions .
the proof of this result contains a novel inversion formula for the polynomials obtained by the rodrigues - type formula ( [ eq2.6 ] ) .
[ th2.2 ] assume that @xmath0 satisfies and has @xmath134 finite moments , and suppose that for some function @xmath11 defined on @xmath95 , @xmath160}(x ) |\delta^{k}[g(x)]|\bigr]<\infty .
\ ] ] then , @xmath161 and the covariance identity holds .
moreover , provided that @xmath162 and @xmath0 has @xmath158 finite moments , the following `` rodrigues inversion '' formula holds : @xmath163}(x)p(x ) = \frac{1}{(k-1)!}\sum_{y = x+1}^{\infty } ( y - x-1)_{k-1}p_k(y ) p(y),\qquad x=0 , 1,\ldots , \ ] ] where , for @xmath10 , @xmath164 , @xmath165 and @xmath166 .
relation ( [ eq2.7 ] ) is obvious when @xmath167 .
also , the assertion follows trivially by lemma [ le2.3](b ) if @xmath95 is finite .
assume , now , that @xmath162 and that @xmath95 is unbounded , fix @xmath168 and consider the function @xmath169 it is easily seen that the pair of functions @xmath170 and @xmath111}(x)p(x)$ ] satisfy ( [ eq2.2 ] ) and ( [ eq2.3 ] ) and also that @xmath171=i(x = s)$ ]
. thus , by ( [ eq2.7 ] ) , we get ( [ eq2.9 ] ) ( cf . formulae ( [ eq3.6 ] ) and ( [ eq3.7 ] ) , below , for the continuous case ) , where @xmath29 is defined by ( [ eq2.6 ] ) and the series converges from lemma [ le2.3](a ) and the fact that @xmath0 has @xmath158 finite moments ( note that @xmath158 finite moments suffice for this inversion formula ) . since @xmath29 is a polynomial and the left - hand side of ( [ eq2.9 ] ) is strictly positive for all large enough @xmath83 , it follows that @xmath172 for all large enough @xmath173 . using ( [ eq2.9 ] ) , the assumption on @xmath159 $ ] is equivalent to the fact that @xmath174| \sum_{y = x+1}^{\infty } ( y - x-1)_{k-1}p_k(y ) p(y)<\infty \ ] ] and arguments similar to those used in the proof of theorem [ th3.1](b ) below show that @xmath174| \sum_{y = x+1}^{\infty } ( y - x-1)_{k-1}|p_k(y)| p(y)<\infty . \ ] ] therefore , we can interchange the order of summation , obtaining @xmath175}(x)\delta^{k}[g(x)]\bigr ] & = & \frac{1}{(k-1 ) ! } \sum_{x=0}^\infty \delta^{k}[g(x ) ] \sum_{y = x+1}^{\infty } ( y - x-1)_{k-1}p_k(y ) p(y ) \\ & = & \frac{1}{(k-1)!}\sum_{y=1}^\infty p_k(y ) p(y ) \sum_{x=0}^{y-1 } ( y - x-1)_{k-1 } \delta^{k}[g(x ) ] \\ & = & { \mathbb{e}}[p_k(x ) g(x ) ] , \ ] ] where @xmath176,\qquad x=0,1,\ldots , \ ] ] and where an empty sum should be treated as @xmath87 .
taking forward differences , it follows that @xmath177=\delta^{k}[g(x)]$ ] so that @xmath178 , where @xmath179 is a polynomial of degree at most @xmath180 , and the desired result follows from the orthogonality of @xmath29 to polynomials of degree lower than @xmath1 .
this completes the proof .
the orthogonality of polynomials ( [ eq1.4 ] ) has been shown , for example , in @xcite ; see also @xcite . for our purposes ,
we review some details .
the induction formula , as in ( [ eq2.5 ] ) , here takes the form @xmath181=q^{k - i}(x)f(x)q_{i , k}(x),\qquad i=0,1,\ldots , k , \ ] ] with @xmath182 , where @xmath183 is a polynomial of degree at most @xmath64 and the explicit recurrence for @xmath183 is @xmath184 this immediately implies that @xmath185 , as in the discrete case ; see also @xcite .
the covariance identity , as in ( [ eq2.7 ] ) , here takes the form ( after repeated integration by parts ; cf .
@xcite ) @xmath186 = { \mathbb{e}}\bigl[q^k(x)g^{(k)}(x)\bigr ] , \ ] ] provided that the expectations are finite and @xmath187 \\[-8pt ] \nonumber & & \quad=\lim_{x\to s-}q^{i+1}(x)f(x)q_{k - i-1,k}(x ) g^{(i)}(x)=0,\qquad i=0,1,\ldots , k-1 \ ] ] ( here , @xmath188 ; that the support is an interval follows from ( [ eq1.2 ] ) ) , where @xmath183 are the polynomials defined by the recurrence above .
obviously , an alternative condition , sufficient for ( [ eq3.3 ] ) ( and , hence , also for ( [ eq3.2 ] ) ) , is @xmath189 \\[-8pt ] \nonumber & & \quad i=0,1,\ldots , k-1 , j=0,1,\ldots , k - i-1 .
\ ] ] assuming that @xmath0 has @xmath134 finite moments , @xmath162 , it is seen that ( [ eq3.4 ] ) , and hence ( [ eq3.2 ] ) , is fulfilled for any polynomial @xmath11 of degree at most @xmath1 .
for example , for the upper limit , we have @xmath190 so that ( [ eq3.4 ] ) trivially holds if @xmath191 ; also , if @xmath192 , then ( [ eq3.4 ] ) follows from @xmath193 as @xmath194 , which can be shown by observing that @xmath195 is eventually decreasing , positive and , by the assumption of finite @xmath134th moment , satisfies @xmath196 .
therefore , the explicit orthogonality relation is @xmath197=\delta_{k , m } k!{\mathbb{e}}[q^{k}(x)]\prod_{j = k-1}^{2k-2}(1-j\delta),\qquad k , m=0,1,\ldots , n , \ ] ] where @xmath132 is kronecker s delta , provided that @xmath0 has @xmath128 finite moments .
the proof follows by a trivial application of ( [ eq3.2 ] ) to @xmath198 , for @xmath133 ( cf .
. it should be noted , however , that the condition ( [ eq3.4 ] ) or ( [ eq3.3 ] ) imposes some unnecessary restrictions on @xmath11 .
in fact , the covariance identity ( [ eq3.2 ] ) ( which enables a general form of the fourier coefficients of @xmath11 to be constructed in terms of its derivatives ) holds , in our case , in its full generality ; the proof requires the novel inversion formula ( [ eq3.6 ] ) or ( [ eq3.7 ] ) below , which may be of some interest in itself .
[ th3.1 ] assume that @xmath0 satisfies and consider the polynomial @xmath99 defined by , where @xmath188 . if @xmath0 has @xmath158 finite moments , then the following `` rodrigues inversion '' formula holds : @xmath199 if @xmath0 has @xmath134 finite moments and @xmath200 , then @xmath201 and the covariance identity holds .
\(a ) let @xmath202 , @xmath203 be the left - hand side and right - hand side , respectively , of ( [ eq3.6 ] ) .
it is easy to see that the integral @xmath203 is finite ( this requires only @xmath158 finite moments ) .
moreover , expanding @xmath204 in the integrand of @xmath203 according to newton s formula , it follows that @xmath205 , @xmath206 , and , thus , by the definition ( [ eq1.4 ] ) , @xmath207 vanishes identically in @xmath18 .
therefore , @xmath208 is a polynomial of degree at most @xmath180 . by ( [ eq3.1 ] ) , @xmath209 for all @xmath210 because @xmath211 as @xmath212 and , for the case @xmath213 , @xmath214 as @xmath215
since the @xmath216th moment is finite , @xmath217 is of degree at most @xmath218 and @xmath195 is increasing and positive in a neighborhood of @xmath219 .
similarly , using the fact that @xmath29 is a polynomial of degree at most @xmath1 and observing that @xmath220 for all @xmath210 , @xmath221 ( again , @xmath158 finite moments suffice for this conclusion ) , it follows that @xmath222 for all @xmath210 .
this proves that @xmath208 vanishes identically in @xmath18 and ( [ eq3.6 ] ) follows .
finally , ( [ eq3.7 ] ) follows from ( [ eq3.6 ] ) and ( [ eq3.2 ] ) with @xmath223 ( the validity of ( [ eq3.2 ] ) for polynomials @xmath11 of degree at most @xmath180 can be shown directly , using repeated integration by parts , as above ) .
\(b ) suppose that @xmath162 ; otherwise , since @xmath140 , we have nothing to show .
since @xmath224={\mathbb{e}}[p_k(x)p_0(x)]=0 $ ] from ( [ eq3.5 ] ) , either @xmath99 vanishes identically for @xmath21 ( in which case , ( [ eq3.2 ] ) trivially holds ) or , otherwise , it must change sign at least once in @xmath18 .
assume that @xmath29 is not identically zero and consider the change - sign points of @xmath29 in @xmath18 , say , @xmath225 ( of course , @xmath226 because @xmath29 is a polynomial of degree at most @xmath1 ) .
fix a point @xmath227 in the finite interval @xmath228 $ ] and write , with the help of ( [ eq3.6 ] ) and ( [ eq3.7 ] ) , @xmath229\nonumber\\ & & \quad= \frac{(-1)^k}{(k-1)!}\int_r^{\rho } \bigl|g^{(k)}(x)\bigr| \int_{r}^x ( x - y)^{k-1}p_k(y ) f(y)\,\mathrm{d}y \,\mathrm{d}x \\ & & \qquad { } + \frac{1}{(k-1)!}\int_{\rho}^s \bigl|g^{(k)}(x)\bigr| \int_{x}^s ( y - x)^{k-1}p_k(y ) f(y)\,\mathrm{d}y \,\mathrm{d}x.\nonumber \ ] ] because of the assumption on @xmath11 , both integrals on the right - hand side of ( [ eq3.8 ] ) are finite .
we wish to show that we can change the order of integration in both integrals in the right - hand side of ( [ eq3.8 ] )
. this will follow from fubini s theorem if it can be shown that @xmath230 and similarly for the other integral in ( [ eq3.8 ] ) .
since @xmath231 for all @xmath21 ( see ( [ eq1.2 ] ) ) , it follows that @xmath232 for all @xmath21 as well .
thus , from ( [ eq3.7 ] ) , we get @xmath233 on the other hand , @xmath99 does not change sign in the interval @xmath234 and , hence , @xmath235 for all @xmath236 , showing that @xmath237 \\[-8pt ] \nonumber & & \quad = \int_{\rho_m}^s \bigl|g^{(k)}(x)\bigr| \int_{x}^s ( y - x)^{k-1 } p_k(y ) f(y)\,\mathrm{d}y \,\mathrm{d}x . \ ] ] by the above considerations , it follows that @xmath238 } h(x):=\inf_{x\in[\rho,\rho_m ] } \int_{x}^s ( y - x)^{k-1 } p_k(y ) f(y)\,\mathrm{d}y > 0 , \\
\label{eq3.13 } s(\rho)&:=&\sup_{x\in[\rho,\rho_m ] } s(x ) : = \sup_{x\in[\rho,\rho_m]}\int_x^{\rho_m } ( y - x)^{k-1 } \\ \label{eq3.14}d(\rho)&:=&\sup_{x\in[\rho,\rho_m]}{d}(x):=\sup_{x\in[\rho,\rho_m ] } \int_{\rho_m}^s ( y - x)^{k-1 } p_k(y ) f(y)\,\mathrm{d}y < \infty\end{aligned}\ ] ] because the three positive functions @xmath239 , @xmath240 and @xmath241 defined above are obviously continuous and @xmath83 lies in the compact interval @xmath242 $ ] ( note that @xmath243 by ( [ eq3.10 ] ) ) .
now , from the inequalities @xmath244 and @xmath245 , @xmath246 $ ] , we conclude that @xmath247 combining ( [ eq3.11 ] ) and ( [ eq3.15 ] ) , we see that there exists a finite constant @xmath248 ( take , for example , @xmath249 ) such that @xmath250 and , thus , by ( [ eq3.9 ] ) , we can indeed interchange the order of integration in the second integral in the right - hand side of ( [ eq3.8 ] ) .
similar arguments apply to the first integral . by the above arguments and by interchanging the order of integration in both integrals in the right - hand side of ( [ eq3.8 ] ) ( with @xmath26 in place of @xmath251 ) , we obtain @xmath252 = { \mathbb{e}}[p_k(x ) g(x ) ] , \ ] ] where @xmath253 differentiating ( [ eq3.17 ] ) or ( [ eq3.18 ] ) @xmath1 times , it is easily seen that @xmath254 so that @xmath255 is a polynomial of degree at most @xmath180 and the desired result follows by ( [ eq3.16 ] ) and the orthogonality of @xmath29 to polynomials of degree lower than @xmath1 .
this completes the proof of the theorem .
a simple application of theorem [ th2.2 ] leads to the following lower variance bound .
[ th4.1 ] fix @xmath256 and assume that @xmath0 satisfies and has @xmath128 finite moments .
then , for any function @xmath11 satisfying @xmath257}(x)|\delta^{k}[g(x)]| \bigr]<\infty \qquad\mbox{for } k=0,1,\ldots , n , \ ] ] the bound @xmath258}(x)\delta ^{k}[g(x)]]}{k!{\mathbb{e}}[q^{[k]}(x)]\prod_{j = k-1}^{2k-2}(1-j\delta ) } \ ] ] holds with equality if and only if @xmath11 is a polynomial of degree at most @xmath259 .
assume that @xmath260<\infty$ ] ( otherwise , we have nothing to show ) . by theorem [ th2.1 ] , the polynomials @xmath261 } ; k=0,1,\ldots,\min\{n , n\}\}$ ]
form an orthonormal basis of all polynomials with degree up to @xmath259 , where @xmath262 is the cardinality of @xmath95 . observing that the @xmath1th fourier coefficient for @xmath11 is , by ( [ eq2.7 ] ) , ( [ eq2.8 ] ) and theorem [ th2.2 ] , @xmath263}{{\mathbb{e}}^{1/2}[p_k^2(x ) ] } = \frac{{\mathbb{e}}[q^{[k]}(x)\delta^{k}[g(x)]]}{(k!{\mathbb{e}}[q^{[k]}(x ) ] \prod_{j = k-1}^{2k-2}(1-j\delta))^{1/2}},\qquad k\leq \min\{n , n\ } , \ ] ] the desired result follows by an application of the finite form of bessel s inequality .
it is worth mentioning here the similarity of the lower variance bound ( [ eq4.2 ] ) with the poincar - type ( upper / lower ) bound for the discrete pearson family , obtained recently in @xcite , namely @xmath264}(x ) ( \delta^{k}[g(x)])^2\bigr]\biggr)\geq0 . \ ] ] the following examples can be verified immediately .
[ ex4.1]if @xmath0 is @xmath30 , then @xmath265 so that @xmath266 and ( [ eq1.6 ] ) follows from ( [ eq4.2 ] ) ( see also @xcite ) .
moreover , the equality in ( [ eq1.6 ] ) holds if and only if @xmath11 is a polynomial of degree at most @xmath259 .
[ ex4.2 ] for the @xmath267 distribution , @xmath268 so that @xmath266 and @xmath269}(x)]=(n)_k p^k(1-p)^k$ ]
. thus , ( [ eq4.2 ] ) yields the bound @xmath270\bigr ] \ ] ] with equality only for polynomials of degree at most @xmath259 .
note that there is equality if @xmath271 .
[ ex4.3 ] for the negative @xmath272 with @xmath273 , @xmath274 ( [ eq2.1 ] ) is satisfied with @xmath275 .
thus , @xmath266 , @xmath269}(x)]=(1-p)^k [ r]_k /p^{2k}$ ] and ( [ eq4.2 ] ) produces the bound @xmath276_k } { \mathbb{e}}^2\bigl [ [ r+x]_k \delta^{k}[g(x)]\bigr ] \ ] ] ( in the above formulae and elsewhere in the paper , @xmath277_n = x(x+1)\cdots(x+n-1)$ ] if @xmath129 and @xmath277_0=1 $ ] ) .
[ ex4.4 ] for the discrete uniform@xmath278 , ( [ eq2.1 ] ) is satisfied with @xmath279 ; thus , @xmath280 and ( [ eq4.2 ] ) entails the bound ( which is an identity if @xmath281 ) @xmath282_k(n - x)_k \delta^{k}[g(x)]\bigr].\ ] ] the lower variance bound for the continuous pearson system is stated in the following theorem ; its proof , being an immediate consequence of ( [ eq3.2 ] ) , ( [ eq3.5 ] ) , theorem [ th3.1 ] and a straightforward application of the finite form of bessel s inequality ( cf .
the proof of theorem [ th4.1 ] above ) , is omitted .
[ th4.2 ] assume that @xmath0 satisfies and has finite moment of order @xmath128 for some fixed @xmath129 .
then , for any function @xmath11 satisfying @xmath283<\infty$ ] , @xmath130 , we have the inequality @xmath284}{k!{\mathbb{e}}[q^{k}(x)]\prod_{j = k-1}^{2k-2}(1-j\delta ) } \ ] ] with equality if and only if @xmath11 is a polynomial of degree at most @xmath259 . some examples now follow . [ ex4.5 ] if @xmath0 is @xmath34 , then @xmath285 and ( [ eq4.4 ] ) yields the houdr kagan variance bound ( [ eq1.7 ] ) ( see @xcite ) , under the weaker assumptions @xmath286 , @xmath130 ; equality holds if and only if @xmath11 is a polynomial of degree at most @xmath259 .
[ ex4.6 ] if @xmath0 is @xmath287 with density @xmath288 , @xmath289 , then @xmath290 , @xmath291=[a]_k/\lambda^{2k}$ ] and ( 4.4 ) entails the bound @xmath292_k } { \mathbb{e}}^2\bigl [ x^k g^{(k)}(x)\bigr ] \ ] ] with equality only for polynomials of degree at most @xmath259 .
[ ex4.7 ] for the standard uniform density , @xmath16 , @xmath280 , @xmath293=(k!)^2/(2^k ( 2k+1)!)$ ] and we get the bound @xmath294 \ ] ] with equality only for polynomials of degree at most @xmath259 .
similar inequalities hold for all beta densities . of course , in the above three examples , the corresponding orthogonal polynomials ( [ eq1.4 ] ) are the well - known hermite , laguerre and jacobi ( legendre ) , respectively , so that one can alternatively obtain the results using explicit expressions for the variance and generating functions for the polynomials ( see , e.g. , @xcite ) ; this is also the case for the discrete polynomials corresponding to examples [ ex4.1][ex4.4 ] ( namely , charlier , krawtchouck , meixner and hahn , respectively ) . however , as will become clear from the following example , the considerations given here do not only simplify and unify the calculations , but also go beyond the classical polynomials . [ ex4.8 ]
if @xmath0 follows the @xmath295 distribution ( student s @xmath296 with @xmath110 degrees of freedom ) with density @xmath297 then it is well known that @xmath0 has only @xmath298 finite integral moments . however , for @xmath299 , @xmath0 satisfies ( [ eq1.2 ] ) and its quadratic @xmath300 has @xmath301 .
thus , ( [ eq4.4 ] ) applies for sufficiently large @xmath110 ( see also @xcite for a poincar - type bound corresponding to ( [ eq4.3 ] ) ) . to this end , it suffices to calculate @xmath302= \biggl(\frac{n}{n-1}\biggr)^k \prod_{j=1}^{k}\biggl(1+\frac{1}{n-2j}\biggr),\qquad k\leq ( n-1)/2 , \ ] ] and @xmath303 .
theorem [ th4.2 ] yields the ( non - classic ) bound @xmath304}{k!n^k ( n - k)_k\prod_{j=1}^{k}(1 + 1/(n-2j))},\qquad n\leq\frac{n-1}{2 } , \ ] ] with equality only for polynomials of degree at most @xmath259 .
it seems that it would be difficult to work with the explicit forms of the corresponding orthogonal polynomials , obtained by ( [ eq1.4 ] ) .
note that a similar bound can be easily obtained for the fisher
snedecor @xmath305 distribution and that schoutens @xcite has obtained the corresponding stein s equation , useful in approximating the @xmath295-distribution .
the main application of the present article ( section [ sec4 ] ) presents a convenient procedure for approximating / bounding the variance of @xmath306 when @xmath11 is smooth enough and when @xmath0 is `` nice '' enough ( pearson ) .
although the procedure is based on the corresponding orthonormal polynomials , @xmath307 $ ] , the main point is that we do not need explicit forms for @xmath29 .
all we need are the fourier coefficients of @xmath11 , @xmath308 $ ] , but , due to the identities ( [ eq3.2 ] ) and ( [ eq2.7 ] ) , the fourier coefficients can be simply expressed in terms of the quadratic @xmath217 and the derivatives / differences of @xmath11 when @xmath0 belongs to the pearson / ord family , that is , when ( [ eq1.2 ] ) or ( [ eq2.1 ] ) is satisfied .
a natural question thus arises : _ is it true that the @xmath259th partial sums , given by the right - hand sides of and , converge to the variance of @xmath306 as @xmath309 _ ? more generally , assume that we want to calculate @xmath310 , when @xmath0 is an r.v . with finite moments of any order .
let @xmath3 be the probability measure of @xmath0 , that is , @xmath3 satisfies @xmath311=\pr(x\leq x)$ ] , @xmath10 , and assume that the support of @xmath3 , @xmath312 , is not concentrated on a finite number of points ( otherwise , the following considerations become trivial ) .
it is well known that there exists an orthonormal polynomial system ( ops ) @xmath313 , which can be obtained by an application of the gram
schmidt orthonormalization process to the real system @xmath314 .
each ( real ) polynomial @xmath315 is of degree ( exactly ) @xmath1 and satisfies the orthonormality condition @xmath316=\int_{{\mathbb{r } } } \phi_i(x)\phi_j(x ) \,\mathrm{d}\mu(x)=\delta_{ij } $ ] .
the members of @xmath317 are uniquely determined by the moments of @xmath0 ( of course , any element of @xmath317 can be multiplied by @xmath318 ) . for any function @xmath319 ( i.e. , with finite variance ) , we first calculate the fourier coefficients @xmath320=\int_{{\mathbb{r}}}\phi_k(x)g(x)\,\mathrm{d}\mu(x),\qquad k=0,1,2,\ldots , \ ] ] and then use the well - known bessel inequality , @xmath321 to obtain the desirable lower variance bound . clearly , theorems [ th4.1 ] and [ th4.2 ] just provide convenient forms of the @xmath259th partial sum in ( [ eq5.2 ] ) for some particularly interesting cases ( pearson / ord system ) .
it is well known that the sum in the right - hand side of ( [ eq5.2 ] ) is equal to the variance ( for all @xmath319 ) if and only if the ops @xmath317 is complete in @xmath322 ( and , thus , it is an orthonormal basis of @xmath322 ) .
this means that the set of real polynomials , @xmath323 $ ] , is dense in @xmath322 . if this is the case , then bessel s inequality , ( [ eq5.2 ] ) , is strengthened to parseval s identity , @xmath324 the following theorem summarizes and unifies the above observations .
[ th5.1 ] assume that @xmath0 has probability measure @xmath3 and finite moments of any order . if @xmath0 satisfies and if @xmath325 , then can be written as @xmath326 } { k!{\mathbb{e}}[q^{k}(x)]\prod_{j = k-1}^{2k-2}(1-j\delta)},\ ] ] provided that the polynomials are dense in @xmath322 and that @xmath327<\infty,\qquad k=0,1,2,\ldots.\ ] ] similarly , if @xmath0 satisfies , if @xmath3 is not concentrated on a finite integer interval and if @xmath319 , then yields the identity @xmath328}(x)\delta ^k[g(x)]]}{k!{\mathbb{e}}[q^{[k]}(x)]\prod_{j = k-1}^{2k-2}(1-j\delta)},\ ] ] provided that the polynomials are dense in @xmath322 and that @xmath329}(x)|\delta^k[g(x)]|\bigr]<\infty,\qquad k=0,1,2,\ldots.\ ] ] it should be noted that ( [ eq5.6 ] ) is always true ( and reduces to a finite sum ) if @xmath0 belongs to the discrete pearson ( ord ) family and the support of @xmath3 is finite .
in this case , the sum adds zero terms whenever @xmath269}(x)]=0 $ ] . on the other hand
, it is well known ( due to m. riesz ) that the polynomials are dense in @xmath322 whenever @xmath3 is determined by its moments ; see @xcite or @xcite , page 45 .
an even simpler sufficient condition is when @xmath3 has a finite moment generating function at a neighborhood of zero , that is , when there exists @xmath330 such that @xmath331 a proof can be found in , for example , @xcite . since condition ( [ eq5.8 ] )
can evidently be checked for all pearson distributions , we include an alternative proof in the .
taking into account the above , we have the following covariance representation .
[ th5.2 ] assume that @xmath0 has a finite moment generating function at a neighborhood of zero .
if @xmath0 satisfies and if @xmath332 , @xmath36 , then @xmath333= \sum_{k=1}^{\infty } \frac{{\mathbb{e}}[q^{k}(x)g_1^{(k)}(x ) ] \cdot{\mathbb{e}}[q^{k}(x)g_2^{(k)}(x ) ] } { k!{\mathbb{e}}[q^{k}(x)]\prod_{j = k-1}^{2k-2}(1-j\delta ) } , \ ] ] provided that for @xmath36 , @xmath334<\infty,\qquad k=0,1,2,\ldots . \ ] ] similarly , if @xmath0 satisfies , then @xmath335= \sum_{k=1}^{\infty } \frac{{\mathbb{e}}[q^{[k]}(x)\delta ^k[g_1(x ) ] ] \cdot{\mathbb{e}}[q^{[k]}(x)\delta^k[g_2(x)]]}{k!{\mathbb{e}}[q^{[k]}(x)]\prod_{j = k-1}^{2k-2}(1-j\delta ) } , \ ] ] where each term with @xmath269}(x)]=0 $ ] should be treated as zero , provided that for @xmath36 , @xmath336}(x)|\delta^k[g_i(x)]|\bigr]<\infty,\qquad k=0,1,2,\ldots .
\ ] ] let @xmath337 $ ] and @xmath338 $ ] be the fourier coefficients of @xmath339 and @xmath340 .
it is then a standard inner product property in the hilbert space that @xmath341=\sum_{k=1}^{\infty } \alpha_k\beta_k$ ] .
substituting , for example , @xmath342=\frac{{\mathbb{e}}[p_k(x)g_1(x)]}{{\mathbb{e}}^{1/2}[p_k^2(x ) ] } = \frac{{\mathbb{e}}[q^{[k]}(x)\delta^k[g_1(x ) ] ] } { ( k!{\mathbb{e}}[q^{[k]}(x ) \prod_{j = k-1}^{2k-2}(1-j\delta)])^{1/2}}\ ] ] ( and similarly for the continuous case and for @xmath343 ) , we obtain ( [ eq5.9 ] ) and ( [ eq5.11 ] ) .
[ ap5.1 ] assume that @xmath344 is a random sample from @xmath345 , , with probability function @xmath346 and let @xmath347 be the complete sufficient statistic .
the uniformly minimum variance unbiased estimator , umvue , of @xmath348 is then given by ( see @xcite ) @xmath349 since no simple form exists for the variance of @xmath350 , the inequalities ( [ eq4.3 ] ) have been used in @xcite in order to prove asymptotic efficiency .
however , @xmath0 is negative @xmath351 and @xmath352=\frac{(-1)^{k-1 } ( k-1)!}{[\nu+x]_k},\qquad k=1,2,\ldots , \ ] ] so that one finds from ( [ eq5.6 ] ) the exact expression ( cf .
example [ ex4.3 ] ) @xmath353 observe that the first term in the series , @xmath354 , is the cramr
rao lower bound .
the above series expansions are in accordance with the corresponding results on bhattacharyya bounds given in @xcite ; these results are also based on orthogonality and completeness properties of bhattacharyya functions , obtained by seth @xcite .
similar series expansions for the variance can be found in @xcite .
next , we present a similar application for the exponential distribution .
[ ap5.2 ] assume that @xmath344 is a random sample from @xmath365 , @xmath366 , with density @xmath367 , @xmath289 , and let @xmath347 be the complete sufficient statistic .
we wish to obtain the umvue of @xmath368 and its variance .
setting @xmath369 , we find that @xmath370 where @xmath371 is euler s constant .
therefore , @xmath372 is unbiased and it follows that the umvue of @xmath368 is of the form @xmath373=-\log(x)-\gamma+\sum _ { j=1}^{\nu-1 } \frac{1}{j}. \ ] ] since @xmath0 follows a @xmath374 distribution and @xmath375 , we obtain from ( [ eq5.4 ] ) ( cf .
example [ ex4.6 ] ) the formula @xmath376 which is quite similar to the formula for @xmath377 in application [ ap5.1 ] .
( once again , the first term in the series , @xmath378 , is the cramr rao bound . )
moreover , the series @xmath379 can be simplified in a closed form . indeed , observing that @xmath380 and taking into account the identity @xmath381_{\nu}}-\frac{1}{[k+1]_{\nu}}\biggr ) , \qquad \nu , k=1,2,\ldots , \ ] ] we have @xmath382_{\nu}}-\frac{1}{[k+1]_{\nu } } \biggr)=\frac{1}{\nu^2 } \ ] ] so that @xmath383 finally , using the last expression and the fact that @xmath384 , we get the inequalities @xmath385 , that is , @xmath386 which shows that @xmath387 is asymptotically efficient .
_ a completeness proof under ( _ .
it is well known that , under ( [ eq5.8 ] ) , @xmath0 has finite moments of any order so that the ops exists ( and is unique ) . from the general theory of hilbert spaces , it is known that @xmath317 is a basis ( i.e. , it is complete ) if and only if it is total , that is , if and only if there does not exist a non - zero function @xmath319 such that @xmath11 is orthogonal to each @xmath315 .
therefore , it suffices to show that if @xmath319 and if @xmath388=0\qquad \mbox{for all } k=0,1,\ldots , \ ] ] then @xmath389 .
since each @xmath315 is a polynomial with non - zero leading coefficient , ( [ eqa.1 ] ) is equivalent to @xmath390=\int_{{\mathbb{r}}}x^n g(x ) \,\mathrm{d}\mu(x)=0 , \qquad n=0,1,\ldots , \ ] ] and it thus suffices to prove that if ( [ eqa.2 ] ) holds , then @xmath391 .
since , by assumption , the functions @xmath392 and @xmath393 ( @xmath394 ) belong to @xmath322 , it follows from the cauchy
schwarz inequality that @xmath395<\infty$ ] for @xmath394 and , thus , that @xmath396\leq{\mathbb{e}}[\mathrm{e}^{-tx}|g(x)|]+{\mathbb{e}}[\mathrm{e}^{tx}|g(x)|]<\infty,\qquad |t|<t_0/2 . \ ] ] from beppo levi s theorem and ( [ eqa.3 ] ) it follows that @xmath397<\infty , \qquad |t|<t_0/2 , \ ] ] and , therefore , from fubini s theorem and ( [ eqa.2 ] ) , we get @xmath398= { \mathbb{e}}\biggl [ \sum_{n=0}^{\infty } \frac{t^n x^n}{n ! } g(x)\biggr]= \sum_{n=0}^{\infty } \frac{t^n}{n ! } { \mathbb{e}}[{x^n}g(x)]=0 , \qquad |t|<t_0/2 . \ ] ] write @xmath399 , @xmath400 and observe that from ( [ eqa.2 ] ) with @xmath70 , @xmath401 .
it follows that @xmath402 , say , where @xmath403 .
clearly , @xmath404 because both @xmath405 and @xmath406 are dominated by @xmath407 and , by assumption , @xmath408 .
we assume now that @xmath409 and we shall obtain a contradiction . under @xmath409 , we can define two borel probability measures , @xmath410 and @xmath411 , as follows : @xmath412 by definition , both @xmath413 and @xmath411 are absolutely continuous with respect to @xmath3 , with radon nikodym derivatives @xmath414 since , by ( [ eqa.4 ] ) , @xmath415 it follows that the moment generating functions of @xmath413 and @xmath411 are finite ( for @xmath394 ) and identical because , from ( [ eqa.5 ] ) , we have that , for any @xmath416 , @xmath417 thus , @xmath418 and choosing @xmath419 , we are led to the contradiction @xmath420 .
this work was partially supported by the university of athens research fund under grant no .
we would like to thank professor roger w. johnson for providing us with a copy of his paper .
thanks are also due to an anonymous referee for the careful reading of the manuscript and for some helpful comments .
n. papadatos is devoting this work to his six years old little daughter , dionyssia , who is suffering from a serious health disease .
cacoullos , t. , papadatos , n. and papathanasiou , v. ( 2001 ) .
an application of a density transform and the local limit theorem .
i primenen . _
* 46 * 803810 ; translation in _ theory probab .
* 46 * ( 2003 ) 699707 .
cacoullos , t. , papathanasiou , v. and utev , s.a .
variational inequalities with examples and an application to the central limit theorem .
* 22 * 16071618 .
chen , l.h.y .
poisson approximation for dependent trials .
probab . _ * 3 * 534545 .
papadatos , n. and papathanasiou , v. ( 2001 ) .
unified variance bounds and a stein - type identity . in _ probability and statistical models with applications _ ( c.a .
charalambides , m.v .
koutras and n. balakrishnan , eds . ) 87100 .
new york : chapman & hall / crc .
stein , c.m .
a bound for the error in the normal approximation to the distribution of a sum of dependent random variables . in _ proc .
sixth berkeley symp .
* 2 * 583602 .
berkeley , ca : univ .
california press . | for an absolutely continuous ( integer - valued ) r.v .
@xmath0 of the pearson ( ord ) family , we show that , under natural moment conditions , a stein - type covariance identity of order @xmath1 holds ( cf .
[ goldstein and reinert , _ j. theoret .
probab .
_ * 18 * ( 2005 ) 237260 ] ) .
this identity is closely related to the corresponding sequence of orthogonal polynomials , obtained by a rodrigues - type formula , and provides convenient expressions for the fourier coefficients of an arbitrary function .
application of the covariance identity yields some novel expressions for the corresponding lower variance bounds for a function of the r.v .
@xmath0 , expressions that seem to be known only in particular cases ( for the normal , see [ houdr and kagan , _ j. theoret .
probab . _
* 8 * ( 1995 ) 2330 ] ; see also [ houdr and prez - abreu , _ ann .
probab . _
* 23 * ( 1995 ) 400419 ] for corresponding results related to the wiener and poisson processes ) .
some applications are also given . |
with xml @xcite being the de facto standard for business and web data representation and exchange , storage and querying of large xml data collections is recognized as an important and challenging research problem .
a number of xml databases @xcite have been developed to serve as a solution to this problem .
while xml databases can employ various storage models , such as relational model or native xml tree model , they support standard xml query languages , called xpath and xquery . in general , an xml query specifies which nodes in an xml tree need to be retrieved .
once an xml tree is stored into an xml database , a query over this tree usually requires two steps : ( 1 ) finding the specified nodes , if any , in the xml tree and ( 2 ) reconstructing and returning xml subtrees rooted at found nodes as a query result .
the second step is called xml subtree reconstruction @xcite and may have a significant impact on query response time .
one approach to minimize xml subtree reconstruction time is to cache xml subtrees rooted at frequently accessed nodes as illustrated in the following example .
consider an xml tree in figure [ fig : xmltree ] that describes a sample bookstore inventory .
the tree nodes correspond to xml elements , e.g. , _ bookstore _ and _ book _ , and data values , e.g. , _
`` arthur '' _ and _ `` bernstein '' _ , and the edges represent parent - child relationships among nodes , e.g. , all the _ book _ elements are children of _ bookstore_. in addition , each element node is assigned a unique identifier that is shown next to the node in the figure . as an example , in figure [ fig : relation ] , we show how this xml tree can be stored into a single table in an rdbms using the edge approach @xcite .
the edge table @xmath8 stores each xml element as a separate tuple that includes the element i d , i d of its parent , element name , and element data content .
a sample query over this xml tree that retrieves books with title `` database systems '' can be expressed in xpath as : = = = = = = + + this query can be translated into relational algebra or sql over the edge table to retrieve ids of the _ book _ elements that satisfy the condition : = = = = = = + @xmath9 + @xmath10 + @xmath11 + @xmath12 + @xmath13 + @xmath14 + where @xmath15 , @xmath16 , and @xmath17 are aliases of table @xmath8 .
for the edge table in figure [ fig : relation ] , the relational algebra query returns i d `` 2 '' , that uniquely identifies the first _ book _ element in the tree .
however , to retrieve useful information about the book , the query evaluator must further retrieve all the descendants of the _ book _ node and reconstruct their parent - child relationships into an xml subtree rooted at this node ; this requires additional self - joins of the edge table and a reconstruction algorithm , such as the one proposed in @xcite . instead , to avoid expensive xml subtree reconstruction
, the subtree can be explicitly stored in the database as an xml reconstruction view ( see figure [ fig : view ] ) .
this materialized view can be used for the above xpath query or any other query that needs to reconstruct and return the _ book _ node ( with i d `` 2 '' ) or its descendant . in this work
, we study the problem of selecting xml reconstruction views to materialize : given a set of xml elements @xmath18 from an xml database , their access frequencies @xmath19 ( aka workload ) , a set of ancestor - descendant relationships @xmath20 among these elements , and a storage capacity @xmath21 , find a set of elements @xmath22 from @xmath18 , whose xml subtrees should be materialized as reconstruction views , such that their combined size is no larger than @xmath21 . to our best knowledge
, our solution to this problem is the first one proposed in the literature .
our main contributions and the paper organization are as follows . in section [ sec : rw ] , we discuss related work . in section [ sec : problem ] , we formally define the xml reconstruction view selection problem . in sections [ sec :
complexity ] and [ sec : fptas ] , we prove that the problem is np - hard and describe a fully polynomial - time approximation scheme ( fptas ) for the problem .
we conclude the paper and list future work directions in section [ sec : conclude ] .
we studied the xml subtree reconstruction problem in the context of a relational storage of xml documents in @xcite , where several algorithms have been proposed . given an xml element returned by an xml query , our algorithms retrieve all its descendants from a database and reconstruct their relationships into an xml subtree that is returned as the query result . to our best knowledge , there have been no previous work on materializing reconstruction views or xml reconstruction view selection .
materialized views @xcite have been successfully used for query optimization in xml databases .
these research works rewrite an xml query , such that it can be answered either using only available materialized views , if possible , or accessing both the database and materialized views .
view maintenance in xml databases has been studied in @xcite .
there have been only one recent work @xcite on materialized view selection in the context of xml databases . in @xcite
, the problem is defined as : find views over xml data , given xml databases , storage space , and a set of queries , such that the combined view size does not exceed the storage space .
the proposed solution produces minimal xml views as candidates for the given query workload , organizes them into a graph , and uses two view selection strategies to choose views to materialize .
this approach makes an assumption that views are used to answer xml queries completely ( not partially ) without accessing an underlying xml database .
the xml reconstruction view problem studied in our work focuses on a different aspect of xml query processing : it finds views to materialize based on how frequently an xml element needs to be reconstructed .
however , xml reconstruction views can be complimentarily used for query answering , if desired .
finally , the materialized view selection problem have been extensively studied in data warehouses @xcite and distributed databases @xcite .
these research results are hardly applicable to xml tree structures and in particular to subtree reconstruction , which is not required for data warehouses or relational databases .
in this section , we formally define the xml reconstruction view selection problem addressed in our work
. * problem formulation . * given @xmath23 xml elements , @xmath24 , and an ancestor - descendant relationship @xmath20 over @xmath18 such that if @xmath25 , then @xmath26 is an ancestor of @xmath27 , let @xmath28 be the access cost of accessing unmaterialized @xmath27 , and let @xmath29 be the access cost of accessing materialized @xmath27 .
we have @xmath30 since reconstruction of @xmath27 takes time .
we use @xmath31 to denote the memory capacity required to store a materialized xml element , @xmath32 and @xmath33 for any @xmath34 . given a workload that is characterized by @xmath35 representing the access frequency of @xmath27 .
the _ xml reconstruction view selection problem _ is to select a set of elements @xmath22 from @xmath18 to be materialized to minimize the total access cost @xmath36 under the disk capacity constraint @xmath37 where @xmath38 if @xmath39 or for some ancestor @xmath26 of @xmath27 , @xmath40 , otherwise @xmath41 .
@xmath21 denotes the available memory capacity , @xmath42 .
next , let @xmath43 means the cost saving by materialization , then one can show that function @xmath44 is minimized if and only if the following function @xmath45 is maximized @xmath46 under the disk capacity constraint @xmath37 where @xmath47 represents all the materialized xml elements and their descendant elements in @xmath18 , it is defined as @xmath48 .
in this section , we prove that the xml reconstruction view selection problem is np - hard .
first , the maximization problem is changed into the equivalent decision problem .
* equivalent decision problem . * given @xmath18 , @xmath20 , @xmath31 , @xmath19 , @xmath49 and @xmath21 as defined in section [ sec : problem ] , let @xmath50 denotes the cost saving goal , @xmath51 .
is there a subset @xmath52 such that @xmath53 and @xmath54 @xmath47 represents all the materialized xml elements and their descendant elements in @xmath18 , it is defined as @xmath48 . in order to study this problem in a convenient model
, we have the following simplified version .
the input is a tree @xmath0 , in which every node @xmath1 has a size @xmath2 and profit @xmath3 , and the size limitation @xmath4 .
the target is to find a subset of subtrees rooted at nodes @xmath55 such that @xmath6 , and @xmath7 is maximal .
furthermore , there is no overlap between any two subtrees selected in the solution .
we prove that the decision problem of the xml reconstruction view selection is an np - hard . a polynomial time reduction from _ knapsack _
@xcite to it is constructed .
[ theoremnpcomplete ] the decision problem of the xml reconstruction view selection is np - complete . it is straightforward to verify that the problem is in np . restrict the problem to the well - known np - complete problem _ knapsack
_ @xcite by allowing only problem instances in which : assume that a knapsack problem has input @xmath56 , and parameters @xmath50 and @xmath4 .
we need to determine a subset @xmath57 such that @xmath58 and @xmath59 .
build a binary tree @xmath0 with exactly leaves .
let leaf @xmath1 have profit @xmath3 and size @xmath2 .
furthermore , each internal node , which is not leaf , has size @xmath60 and profit @xmath60 .
clearly , any solution can not contain any internal due to the size limitation .
we can only select a subset of leaves .
this is equivalent to the knapsack problem .
finally , we state the np - hardness of the xml reconstruction view selection problem .
the xml reconstruction view selection problem is np - hard .
it follows from theorem [ theoremnpcomplete ] , since the equivalent decision problem is np - complete .
we assume that each parameter is an integer .
the input is @xmath23 xml elements , @xmath61 which will be represented by an @xmath20 tree @xmath62 , where each edge in @xmath62 shows a relationship between a pair of parent and child nodes .
we have a divide and conquer approach to develop a fully approximation scheme .
given an @xmath20 tree @xmath62 with root @xmath63 , it has subtrees @xmath64 derived from the children @xmath65 of @xmath63 .
we find a set of approximate solutions among @xmath66 and another set of approximate solutions among @xmath67 .
we merge the two sets of approximate solutions to obtain the solution for the union of subtrees @xmath64 .
add one more solution that is derived by selecting the root @xmath63 of @xmath62 .
group those solutions into parts such that each part contains all solutions @xmath68 with similar @xmath69 .
prune those solution by selecting the one from each part with the least size .
this can reduce the number of solution to be bounded by a polynomial .
we will use a list @xmath68 to represent the selection of elements from @xmath18 . for a list of elements @xmath68 , define @xmath70 , and @xmath71 .
define @xmath72 be the largest product of the node degrees along a path from root to a leaf in the @xmath20 tree @xmath62 .
assume that @xmath73 is a small constant with @xmath74 .
we need an @xmath75 approximation .
we maintain a list of solutions @xmath76 , where @xmath77 is a list of elements in @xmath18 .
let @xmath78 with @xmath79 .
let @xmath80 and @xmath81 .
partition the interval @xmath82 $ ] into @xmath83 such that @xmath84 $ ] and @xmath85 $ ] with @xmath86 for @xmath87 , and @xmath88 $ ] , where @xmath89 . two lists @xmath77 and
@xmath90 , are in the same region if there exist @xmath91 such that both @xmath92 and @xmath93 are @xmath94 . for two lists of partial solutions
@xmath95 and @xmath96 , their link @xmath97 . * prune * ( @xmath98 ) input : @xmath98 is a list of partial solutions @xmath99 ; partition @xmath98 into parts @xmath100 such that two lists @xmath77 and @xmath90 are in the same part if @xmath77 and @xmath90 are in the same region .
for each @xmath101 , select @xmath90 such that @xmath102 is the least among all @xmath90 in @xmath101 ; * end of prune * * merge * ( @xmath103 ) input : @xmath104 and @xmath105 are two lists of solutions .
let @xmath106 ; for each @xmath107 and each @xmath108 append their link @xmath109 to @xmath98 ; return @xmath98 ; * end of merge * * union * ( @xmath110 ) input : @xmath111 are lists of solutions . if @xmath112 then return @xmath104 ; return prune(merge(union@xmath113 , union@xmath114 ; * end of union * * sketch * ( @xmath62 ) input : @xmath62 is a set of @xmath23 elements according to their @xmath20 . if @xmath62 only contains one element @xmath27 , return the list @xmath115 with two solutions .
partition the list @xmath62 into subtrees @xmath64 according to its @xmath116 children .
let @xmath117 be the list that only contains solution @xmath62 . for @xmath118 to @xmath116 let @xmath119=sketch(@xmath120 )
; return union@xmath121 ; * end of sketch * for a list of elements @xmath68 and an @xmath20 tree @xmath62 , @xmath122 $ ] is the list of elements in both @xmath68 and @xmath62 .
if @xmath64 are disjoint subtrees of an @xmath20 tree , @xmath123 $ ] is @xmath124\circ \cdots \circ p[j_k]$ ] . in order to make it convenient
, we make the tree @xmath62 normalized by adding some useless nodes @xmath27 with @xmath125 .
the size of tree is at most doubled after normalization . in the rest of the section , we always assume @xmath62 is normalized .
[ lemma - union ] assume that @xmath119 is a list of solutions for the problem with @xmath20 tree @xmath120 for @xmath126 .
let @xmath127 for @xmath126
. then there exists @xmath128union(@xmath111 ) such that @xmath129 and @xmath130 .
we prove by induction .
it is trivial when @xmath112 .
assume that the lemma is true for cases less than @xmath116 .
let @xmath131 @xmath132 union@xmath113 and @xmath133 @xmath132 union@xmath134 . assume that @xmath131 contains @xmath135 such that @xmath136 and @xmath137 .
assume that @xmath133 contains @xmath138 such that @xmath139)$ ] and @xmath140 .
let @xmath141 .
let @xmath142 be the solution in the same region with @xmath143 and has the least @xmath144 .
therefore , @xmath145 since @xmath146 and @xmath147 , we also have @xmath148 [ lemma-1 ] assume that @xmath68 is an arbitrary solution for the problem with @xmath20 tree @xmath62 .
for @xmath98@xmath132sketch(@xmath62 ) , there exists a solution @xmath149 in the list @xmath98 such that @xmath150 and @xmath151 .
we prove by induction .
the basis at @xmath152 is trivial .
we assume that the claim is true for all @xmath153 .
now assume that @xmath154 and @xmath62 has @xmath116 children which induce @xmath116 subtrees @xmath64 .
let @xmath155 $ ] for @xmath126 . by our hypothesis , for each @xmath1 with @xmath156
, there exists @xmath157 such that @xmath158)$ ] and @xmath159))$ ] .
let @xmath22@xmath132union@xmath160 .
by lemma [ lemma - union ] , there exists @xmath161 such that @xmath162)\\ & \le & f^{\log \chi(j)}\lambda(p[j_1,\cdots , j_{k}])\\ & = & f^{\log \chi(j)}\lambda(p).\end{aligned}\ ] ] and @xmath163)\\ & = & \mu(p ) .
\end{aligned}\ ] ] [ time - lemma ] assume that @xmath164 .
then the computational time for sketch(@xmath62 ) is @xmath165 , where @xmath166 is the number of nodes in @xmath62 .
the number of intervals is @xmath167 .
therefore the list of each @xmath119@xmath132prune(@xmath120 ) is of length @xmath168 .
let @xmath169 be the time for union@xmath170 .
it satisfies the recursion @xmath171 .
this brings solution @xmath172 .
let @xmath173 be the computational time for prune(@xmath62 )
. denote @xmath174 to be the number of edges in @xmath62 .
we prove by induction that @xmath175 for some constant @xmath176 .
we select constant @xmath177 enough so that merging two lists takes @xmath178 steps .
we have that @xmath179 we now complete the common procedure . before we give the fptas for our problems , we first give the following lemma which facilitates our proof for fptas .
one can refer the algorithm book @xcite for its proof .
[ basic - lemma ] ( 1)for @xmath180 , @xmath181 .
( 2)for real @xmath182 , @xmath183 .
* algorithm * approximate(@xmath62 , @xmath73 ) input : @xmath62 is an @xmath20 tree with elements @xmath184 and @xmath73 is a small constant with @xmath185 ; let @xmath186sketch@xmath187 ; select @xmath77 from the list @xmath98 that @xmath77 has the optimal cost ; * end of the algorithm * [ main - theorem ] for any instance of @xmath62 of an @xmath20 tree with @xmath23 elements , there exists an @xmath188 time approximation scheme , where @xmath189 .
assume that @xmath68 is the optimal solution for input @xmath62 .
let @xmath98@xmath132prune(@xmath62 ) . by lemma [ lemma-1 ]
, we have @xmath190 that satisfies the condition of lemma [ lemma-1 ] .
@xmath191 furthermore , @xmath192 .
the computational time follows from lemma [ time - lemma ] .
it is easy to see that @xmath193 .
we have the following corollary .
for any instance of @xmath62 of an @xmath20 tree with @xmath23 elements , there exists an @xmath194 time approximation scheme , where @xmath195 .
in this section , we show an approximation scheme for the problem with an input of multiple trees .
the input is a series of trees @xmath64 .
for any instance of @xmath62 of an @xmath20 tree with @xmath23 elements , there exists an @xmath196 time approximation scheme , where @xmath197 and @xmath198 is a tree via connecting all @xmath64 into a single tree under a common root @xmath199 .
build a new tree with a new node @xmath199 such that @xmath64 are the subtrees under @xmath199 .
apply the algorithm in in theorem [ main - theorem ] .
in this work , we studied the problem of xml reconstruction view selection that promises to improve query evaluation in xml databases .
we were first to formally define this problem : given a set of xml elements @xmath18 from an xml database , their access frequencies @xmath19 ( aka workload ) , a set of ancestor - descendant relationships @xmath20 among these elements , and a storage capacity @xmath21 , find a set of elements @xmath22 from @xmath18 , whose xml subtrees should be materialized as reconstruction views , such that their combined size is no larger than @xmath21 .
next , we showed that the xml reconstruction view selection problem is np - hard . finally , we proposed a fully polynomial - time approximation scheme ( fptas ) that can be used to solve the problem in practice .
future work for our research includes two main directions : ( 1 ) an extension of the proposed solution to support multiple xml trees and ( 2 ) an implementation and performance study of our framework in an existing xml database . | query evaluation in an xml database requires reconstructing xml subtrees rooted at nodes found by an xml query .
since xml subtree reconstruction can be expensive , one approach to improve query response time is to use reconstruction views - materialized xml subtrees of an xml document , whose nodes are frequently accessed by xml queries . for this approach to be efficient , the principal requirement is a framework for view selection . in this work ,
we are the first to formalize and study the problem of xml reconstruction view selection .
the input is a tree @xmath0 , in which every node @xmath1 has a size @xmath2 and profit @xmath3 , and the size limitation @xmath4 .
the target is to find a subset of subtrees rooted at nodes @xmath5 respectively such that @xmath6 , and @xmath7 is maximal .
furthermore , there is no overlap between any two subtrees selected in the solution .
we prove that this problem is np - hard and present a fully polynomial - time approximation scheme ( fptas ) as a solution . |
advances in detector technology led to the construction of time - of - flight ( tof ) positron emission tomography ( pet ) scanners , resulting in enhanced signal - to - noise ratio .
goal of this work is the improvement of time resolution of scintillation detectors for pet , by a better understanding of photon propagation inside scintillators and the influence of the cherenkov effect .
the time resolution of scintillation detectors depends on several factors .
it can be distinguished between the time resolution of the scintillator , the photon detector and the readout electronics .
the magnitude of time resolution of the scintillator itself has two major origins : statistical processes of the scintillation ( @xmath0 ) and the photon propagation from the point of interaction to the photon detector ( @xmath1 ) .
the statistical processes are in principle influenced by scintillation rise- and decay - times and the light yield .
the photon propagation process is influenced by factors , such as refractive index , scintillator geometry and surface finishing .
cherenkov photons in scintillators are emitted by electrons , ionised by incident 511kev photons and propagate faster than the speed of light @xmath2 in the scintillator , with _
n _ being the speed of light in vacuum and the refractive index , respectively . making use of the cherenkov effect is very promising for tof - pet detectors , as the time spread of this process is smaller than for scintillation in inorganic materials @xcite .
the direction of the cherenkov photons can be described as a cone relative to the electron motion .
the opening angle is determined by electron velocity and the refractive index of the crystal .
since the electron can be scattered to any direction , the direction of the photons is quasi - random .
the number of emitted cherenkov photons for lso : ce is about 15 per 511kev photon @xcite .
monte carlo simulations were performed to investigate the impact of photon propagation and the cherenkov effect inside scintillators on the time resolution .
these simulations will be described in the following sections .
for the simulations geant4 @xcite was used .
lso : ce was chosen as scintillator material , since this is a common scintillating material for pet .
the optical properties , such as , refractive index and transmission spectrum were taken from @xcite . for the light yield , decay and rise time , typical values of @xmath3 photons / mev , 40ns and 100ps were chosen , respectively . in order to evaluate only the effect of photon production and photon propagation , the detection efficiency of the photon detectors was set to 100% over all wavelengths . instead of simulating a positron source , photons with energies of 511kev
were generated in a point source and emitted into a defined direction .
the photon detectors recorded arrival time , wavelength and creation process of the photons .
compton scattered events were discriminated .
besides photon statistics , the coincidence time resolution ( ctr ) of pet - like detector systems is influenced by variations of the depth of interaction ( doi ) in the opposing scintillators , resulting in differences of the photon propagation lengths to the photon detectors .
this effect can be reduced by shortening the crystals , however , leading to decreasing detection efficiency for 511kev photons . to evaluate the coincidence time resolution , a basic coincidence setup with two finger - like scintillators , each connected to one photon detector ,
was simulated , see figure [ fig : coincidencesetup ] ( a ) .
the crystal sizes were @xmath4mm@xmath5 with z ranging from 1 mm to 30 mm . in figure
[ fig : ctr_lso ] , the simulated ctr is shown for various crystal lengths .
two curves are plotted , one representing the detection of scintillation photons only , the other one showing the improvement of time resolution if the detection of cherenkov photons is included . for both lines
the dependency on the crystal lengths is clearly visible . for crystal lengths from 1 mm to 30 mm
the ctr ranges from 32ps to 144ps fwhm for scintillation and from 12ps to 125ps fwhm by including the cherenkov effect , respectively .
the reason for this behaviour is the decrease in localisation of the 511kev photon interaction inside both of the crystals , simply due to the increasing size of the crystals .
this uncertainty of localisation causes increasing time spread as a function of the crystal length .
the impact of the cherenkov effect on the ctr is clearly visible in figure [ fig : ctr_lso ] .
although time resolution improves with decreasing crystal lengths , longer crystals , providing reasonable sensitivity to the 511kev photons , are used for real pet systems .
the simulated detection efficiency of coincidences ranges from 2% , for @xmath6 mm to above 50% for @xmath7 mm . for real tof - pet scanners
a trade - off between time resolution and sensitivity has to be made .
in the following , the impact of the doi on the photon arrival rates and the photon output of the scintillator at the photon detector will be discussed . for the simulations , the setup of figure [ fig : coincidencesetup ] ( b )
the size of the simulated crystal was @xmath8mm@xmath5 , and was connected to a photon detector .
a source of 511kev photons was placed at the side of the scintillator and the distance _
d _ of the source relative to the photon detector was varied over the whole crystal length from 0 mm to 30 mm . by knowing the distance _
d _ , the doi is determined .
the simulated photon arrival rates for three distances _ d _ at the photon detector are shown in figure [ fig : arrivaltimes ] .
for the photon arrival rates coming from the scintillation process , a fast rise can be seen at early times , followed by an intermediate plateau until a second , smaller and slower increase of the photon rate is visible . the width of the plateau is directly related to the doi of the penetrating 511kev photons and vanishes for dois reaching the length of the scintillator .
the reason is originated in the isotropic emission of scintillation photons .
the photons , emitted towards the photon detector form the first rise of the number of scintillation photons and the consecutive plateau .
the second rise is caused by photons emitted away from the photon detector , getting reflected at the end of the scintillator and reaching the photon detector with a delay , depending on their travel path .
for the cherenkov photons this effect is more obvious , since the duration of the cherenkov process is shorter , compared to the scintillation process .
the cherenkov photons form two subsequent peaks with a distance and width proportional to the effective travel path inside the scintillator . compared to the scintillation photons ,
the rates of cherenkov photons form sharp peaks , providing accurate time information of the interactions of the 511kev photons . from the simulations ,
the mean trigger times at single photon level were determined for various dois , see figure [ fig : meanarrivaltime ] . until large dois , the trigger times are depending linearly on the doi , but deviate at dois @xmath925 mm .
the reason is that for low dois , photons emitted towards the photon detector are triggering the detector , the photons emitted into the opposite direction arrive at the detector with a significant delay .
with increasing doi , the time delay of these two consecutive peaks decreases until they merge at high dois , resulting in a higher photon rate at the photon detector and , therefore , in earlier trigger probability @xcite .
furthermore , the number of detected photons is dependent on the doi , see figure [ fig : nbofphotons ] .
it is clearly visible that the light output of the crystal decreases with increasing doi .
since cherenkov photons provide very fast response to the photon interaction , for good time resolution it is beneficial to detect as many cherenkov photons as possible .
unfortunately , many of them are lost in real detector systems due to low quantum efficiencies of photon detectors in the blue and uv - range and the cut - off frequencies of photon transmission in scintillators . analyzing scintillation pulse shapes and detection of the first and second rise of photon arrivals , can provide information about doi , and help reducing parallax errors of pet systems . due to limited time resolution of state - of - the - art photon detectors
it is difficult to discriminate the first rise , the width of the plateau and the second rise of the photon arrival rate .
nevertheless , this effect should be observable as variation of the rise time of scintillation pulses .
therefore , measuring the rise time or the number of detected photons not only parallax errors can be reduced by estimation of the doi but also improved time resolution by determination of a corrected time stamp of interactions of the 511kev photons inside the scintillator can be achieved . a dependency of scintillation rise times and the number of detected photons on the doi has been measured by @xcite .
for investigations on the time resolution of a scintillator , a setup similar to that in figure [ fig : coincidencesetup ] ( b ) was used , but with the only difference , that the photons were emitted from top , towards the photon detector .
the time stamps were determined triggering on the first arriving photon ( scintillation and cherenkov effect ) .
contrary to the simulation in section [ sec : photontransport ] , the doi is unknown .
the resulting photon arrival times can be seen in figure [ fig : scatter ] ( a ) .
figure [ fig : scatter ] ( b ) shows a scatter plot of the number of optical photons per 511kev photon and the time when the first photon is arriving at the detector .
compton scattered events are discriminated with a discriminator threshold of 1600 photons per event .
92% of the 511kev photons were detected , and 48% of them were compton events . in figure
[ fig : scatter ] ( b ) the time walk is visible . by calculating the mean photon arrival times for increasing amplitudes ,
time walk correction was applied .
the corrected arrival time spectra can be seen in figure [ fig : scatter ] ( c ) and ( d ) .
the total time resolution @xmath10 of the scintillator is @xmath1139ps .
the corrected data of figure [ fig : scatter ] ( c ) gives @xmath12ps .
using @xmath13 results in @xmath14ps .
the corrected time resolution @xmath15 still includes the standard deviation of the scintillation process , the contribution of the cherenkov process and a contribution of photon propagation .
note , that the name of the variables @xmath16 do not imply normal distributions , but are measures for the standard variation .
simulations for determination of the time resolution of lso : ce crystals were performed .
the influence of crystal sizes on the ctr for pet - like detector systems was shown and ranged from 32ps to 144ps fwhm for crystal lengths of 1 mm to 30 mm . including the detection of cherenkov photons showed a significant improvement of the ctr ( 12 to 125ps fwhm ) .
it was shown that the light output and the photon rate distribution at the photon detector are dependent on the doi and that applying time walk correction significantly improves the time resolution from @xmath17ps to @xmath12ps .
the contribution of the time walk was @xmath14ps . by measuring the pulse amplitude or the rise time of a scintillation pulse
, the doi can be estimated .
information on the doi allows to reduce parallax errors for pet and determination of the accurate time stamps of photon interactions results in improvement of tof for pet .
as the development of readout electronics proceeds quickly , extracting amplitude and rise - time information is realistic also for full tof - pet systems @xcite . for real pet - systems , photon detectors with high quantum efficiency in the blue- and uv - range and scintillators with
increased transmission in these wavelength - bands would help to improve tof for pet by making use of the cherenkov effect .
however , for time resolutions of state - of - the - art photon detectors the benefit from the cherenkov effect is small , but becomes increasingly important if the time resolution of photon detectors approximates the time resolution of the scintillators .
this work was partly funded by eu - project hadronphysics3 ( project 283286 ) . | we present results of simulations on the influence of photon propagation and the cherenkov effect on the time resolution of lso : ce scintillators .
the influence of the scintillator length on the coincidence time resolution is shown .
furthermore , the impact of the depth of interaction on the time resolution , the light output and the arrival time distribution at the photon detector is simulated and it is shown how these information can be used for time walk correction .
positron emission tomography ( pet ) , time - of - flight ( tof ) , time resolution , cherenkov effect , depth of interaction ( doi ) , time walk |
clustering is a commonly encountered problem in many areas such as marketing , engineering , and biology , among others . in a typical clustering problem ,
the goal is to group entities together according to a certain similarity measure .
such a measure can be defined in many different ways , and it determines the complexity of solving the relevant clustering problem . clustering problems with the similarity measure defined by regression errors is especially challenging because it is coupled with regression .
consider a retailer that needs to forecast sales at the stock keeping unit ( sku ) level for different promotional plans and mechanisms ( e.g. , 30% off the selling price ) using a linear regression model .
a sku is a unique identifying number that refers to a specific item in inventory .
each sku is often used to identify product , product size , product type , and the manufacturer .
seasonality is an important predictor and is modeled using an indicator dummy input variable for each season , with the length of one season being one week .
the usable data for each sku is limited compared to the possible number of parameters to estimate , among which the seasonality dummies compose a large proportion .
more significant and useful statistical results can be obtained by clustering skus with similar seasonal effects from promotions together , and estimating seasonality dummies for a cluster instead of a single sku .
however , the seasonal effects of skus correspond to regression coefficients , which can only be obtained after grouping skus with similar seasonality .
a two - stage method can be used to solve such difficult clustering problems that are intertwined with regression . in the first stage ,
entities are clustered based on certain approximate measures of their regression coefficients . in the second stage , regressions are performed over the resultant clusters to obtain estimates for the regression coefficients for each cluster . however , good approximate measures are difficult to obtain a priori before carrying out the regressions .
a better alternative is to perform clustering and regression simultaneously , which can be achieved through cluster - wise linear regression ( clr ) , which is also referred to as `` regression clustering '' in the literature .
other application areas of clr include marketing , pavement condition prediction , and spatial modeling and analysis .
more details about these other application areas can be found in openshaw @xcite , desarbo and cron @xcite , desarbo @xcite , and luo and chou @xcite .
the clr problem bears connection to the minimum sum - of - squares clustering ( mssc ) problem , the objective of which is to find clusters that minimize the sum of squared distances from each entity to the centroid of the cluster which it belongs to . contrary to clustering entities directly based on distances , clr generates clusters according to the effects that some independent variables have on the response variable of a preset regression model . each entity
is represented by a set of observations of a response variable and the associated predictors .
clr is to group entities with similar regression effects into a given number of clusters such that the overall sum of squared residuals within clusters is minimal .
although the mssc problem has been extensively studied by researchers from various fields ( e.g. , statistics , optimization , and data mining ) , the work for the clr problem is limited , most of which concerns adapting the lloyd s algorithm based heuristic algorithms of the mssc problem to the clr problem .
the lloyd s algorithm starts randomly from some initial partition of clusters , then calculates the centroids of clusters , and assigns entities to their closest centroids until converging to a local minimum .
recently , several exact approaches have been proposed by carbonneau et al @xcite , which are discussed in detail in sections 1.1 and 2 .
we tackle the problem of clustering entities based on their regression coefficients by modeling it as a generalized clr problem , in which we allow each entity to have more than one observation .
we propose both a mixed integer quadratic program formulation and a set partitioning formulation for generalized clr .
our mixed integer quadratic program formulation is more general than the one proposed by bertsimas and shioda @xcite , which can not be directly applied to the sku clustering problem since they assume each clustering entity to have only one observation and this assumption does not hold for the sku clustering problem .
we identify a connection between the generalized clr and mssc problems , through which we prove np - hardness of the generalized clr problem .
column generation is an algorithmic framework for solving large - scale linear and integer programs .
vanderbeck and wolsey @xcite and barnhart _ et al . _
@xcite overview column generation for solving large integer program .
we design a column generation ( cg ) algorithm for the generalized clr problem using its set partitioning formulation .
the corresponding pricing problem is a mixed integer quadratic program , which we show to be np - hard . to handle larger instances in the column generation framework
, we also propose a heuristic algorithm , referred to as the cg heuristic algorithm .
this heuristic algorithm , inspired by bertsimas and shioda @xcite , first clusters entities to a small number of groups and then performs our column generation algorithm on these groups of entities .
in addition , we propose a metaheuristic algorithm , named the ga - lloyd algorithm , which uses an adapted lloyd s clustering algorithm to find locally optimal partitions and relies on the genetic algorithm ( ga ) to escape local optimums .
furthermore , we introduce a two - stage approach , used frequently in practice due to its simplicity , which performs clustering first and regression second .
we test our algorithms using real - world data from a large retail chain .
we compare the performance of the ga - lloyd , the cg heuristic , and the two - stage algorithms on two larger instances with 66 and 337 skus , corresponding to two representative subcategories under the retailer s product hierarchy .
we observe that the ga - lloyd algorithm performs much better than the two - stage algorithm .
the cg heuristic algorithm is able to produce slightly better results than the ga - lloyd algorithm for smaller instances , but at the cost of much longer running time .
the ga - lloyd algorithm performs the best and identifies distinctive and meaningful seasonal patterns for the tested subcategories .
in addition , we find that the column generation algorithm is able to solve the sku clustering problem with at most 20 skus to optimality within reasonable computation time .
we benchmark the performance of the ga - lloyd and cg heuristic algorithms against the optimal solutions obtained by the column generation algorithm to find that both algorithms obtain close to optimal solutions .
the contributions of our work are as follows . 1 . we are the first to model and solve the sku clustering problem , commonly encountered in retail predictive modeling , through generalized clr .
2 . we propose four heuristic algorithms for the generalized clr problem , including the cg heuristic algorithm , the ga - lloyd algorithm , the two - stage approach , and a variant of spth algorithm .
3 . we propose an exact column generation algorithm that enables us to evaluate the performance of the heuristic algorithms .
we prove np - hardness of the generalized clr problem and np - completeness of the pricing problem of the column generation algorithm .
note that the number of clusters is a parameter in the generalized clr problem that needs to be decided by user beforehand or by enumeration .
although we provide comparison of models with different number of clusters for real - world data in section [ subsec_compare_seasonality ] , it is not straightforward to develop a universal rule for deciding the number of clusters .
this is also a hard task for mssc and clr .
the aic or bic criteria did not give a reasonable number of clusters for the data set we used for the experiment .
they gave more than three times the number of hand - picked number of clusters that work in practice .
hence , in this paper , we assume that the target number of clusters is given in advance . in figure [
fig : compare ] , we summarize and compare the terms in the clr , generalized clr and the sku clustering problem . while clr only has entities , generalized clr allows multiple observations per entity .
the clr problem can be thought of as the generalized clr with one observation per entity .
note that entity and observation in generalized clr are the sku and transactions in the sku clustering problem , respectively .
+ the rest of the paper is organized as follows . in section [ sec : formulations ] , we introduce both the mixed integer quadratic program and the set partitioning formulations of the generalized clr problem . we draw the connection between the generalized clr and mssc problems , and prove np - hardness of the former through this connection . in section
[ sec : algorithms ] , we present the exact column generation algorithm , the cg heuristic algorithm , the ga - lloyd heuristic algorithm , the two - stage algorithm , and a variant of the spth algorithm . the pricing problem of the column generation algorithm is shown to be np - complete . in section [ sec : experiments ] , we present numerical experiments to test the performance of all proposed algorithms .
the literature review is discussed next . to the best of authors knowledge
, no previous work has been conducted that comprehensibly tackles the generalized clr problem .
however , an extensive collection has been proposed for the typical clr problem , which can potentially be adapted to tackle the generalized clr problem .
the algorithms proposed for the typical clr problem are mainly heuristics bearing close similarity to the algorithms for the mssc problem .
for example , spth @xcite proposes an exchange algorithm which , starting from some initial clusters , exchanges two items between two clusters if a cost reduction is observed in the objective function .
desarbo @xcite presents a simulated annealing method to escape local minimums .
@xcite used a self organizing map to perform clusterwise regression . on mathematical programming - based heuristics , lau _ et al .
_ @xcite propose a nonlinear programming formulation that it is solved approximately using commercial solvers with no guarantee to find a global optimum .
their algorithm s performance depends heavily on the initial clusters .
this initial - cluster dependency is overcome by the k - harmonic means clustering algorithm proposed by zhang @xcite .
moreover , bertsimas and shioda @xcite introduce a compact mixed - integer linear formulation for a slight variation of the clr problem with the sum of the absolute error as the objective .
their algorithm first divides entities into a small number of clusters , and then feeds these clusters into their mixed integer program . for exact approaches to clr ,
carbonneau _ et al . _
@xcite proposed a mixed logical - quadratic programming formulation by replacing big m constraints with the logical implication of the constraints .
@xcite proposed an iterative algorithm based on sequencing the data and repetitive use of a branch and bound algorithm .
@xcite proposed a column generation based algorithm based on @xcite and @xcite .
there are two key differences between these works and the one we propose in this paper .
first , we provide both a quadratic mixed - integer program formulation and a set partition formulation of the generalized clr problem .
the former is a generalization of the formulation in @xcite , and the latter is the set partitioning formulation for generalized clr ( recently , carbonneau _ et al .
_ @xcite have proposed a set partitioning formulation for clr ) .
second , we propose two new heuristics , namely the cg heuristic algorithm and the ga - lloyd algorithm for the generalized clr problem . there is another stream of research for the clr problem that assumes a distribution function for regression errors where each entity is assigned to each cluster with a certain probability , i.e. , using `` soft '' assignments .
for example , desarbo and cron @xcite propose a finite conditional mixture maximum likelihood methodology , which assumes normal distribution for regression errors and is solved through the expectation maximization algorithm . since then , a large number of mixture regression models have been developed , including probit and logit mixture regression models as examples .
@xcite compare the performance of the expectation maximization algorithms with their nonlinear programming - based algorithm .
hennig @xcite investigate idenfiability of model - based clusterwise linear regression for consistent estimate of parameters .
durso _ et al . _
@xcite proposed to integrate fuzzy clustering and fuzzy regression . a recent work of ingrassia _
et al_. @xcite uses linear @xmath0 cluster - weighted models for clustering regression .
these model - based approaches allow residual variances to differ between clusters , which the least squares approaches do not allow . in the soft assignment setting , an entity can be assigned to the cluster of highest probability .
we restrict the scope of our review and comparison to least squares approaches because the objective functions are different .
the reader is referred to wedel and desarbo @xcite and hennig @xcite for reviews .
the algorithms for the mssc problem are instructive to solving the clr problem .
there are abundant papers for solving the mssc problem .
hansen and jaumard @xcite survey various forms of clustering problems and their solution methods , including mssc , from a mathematical programming point of view . in their survey ,
solution methods for the mssc problem include dynamic programming , branch - and - bound , cutting planes , and column generation methods .
all these algorithms do not scale well to large size instances or in higher dimensional spaces .
heuristics are also considered , including lloyd s like algorithms ( e.g. , k - means and h - means ) and metaheuristics such as simulated annealing , tabu search , genetic algorithms and variable neighborhood search . with respect to mathematical programming approaches , du merle _ et al .
_ @xcite propose an interior point algorithm to exactly solve the mssc problem .
@xcite improve the algorithm of du merle _ et al . _
@xcite by exploiting the geometric characteristics of clusters , which enables them to solve much larger instances .
we first provide a mixed integer quadratic formulation for the generalized clr problem .
this formulation reveals a close connection between the generalized clr and mssc problems , which enables us to show that the generalized clr problem is np - hard .
consider set @xmath1 of @xmath2 entities .
each entity @xmath3 has @xmath4 observations of dependent variable @xmath5 , and @xmath6 independent variables @xmath7 with @xmath8 for any @xmath9 $ ] . in practice
the number of entities @xmath4 depends on @xmath10 , but we do not show this dependency for improved readability .
( for each integer @xmath11 we introduce @xmath12 = \{1, ... ,g\}$ ] . )
observation @xmath13 is associated with independent variables @xmath14 .
note that vectors are represented in bold symbols .
we want to divide these @xmath2 entities into a partition @xmath15 of @xmath16 clusters where @xmath17 , @xmath18 for any @xmath19 , and @xmath20}c_k = [ i]$ ] .
the minimum size of a cluster is @xmath21 , which is set by the user .
this implies @xmath22 for any @xmath23 $ ] where @xmath24 denotes the cardinality of cluster @xmath25 .
note that the number of observations pertaining to a cluster is at least @xmath26 .
the minimum size constraints are imposed to ensure that there are enough observations for each cluster .
further , in order to avoid regression models with zero error , we require @xmath27 .
we also require @xmath28 such that there is always a feasible solution .
the generalized clr problem is formulated as follows : @xmath29\text { , } k \in [ k]\text { , } l \in [ l]\label{constraint : quadratic : regression1 } \\
t_{il } + ( y_{il } - \sum_{j=1}^j\beta_{kj}x_{ijl } ) + m(1 - z_{ik } ) & \geq 0 \quad & & i \in [ i]\text { , } k \in [ k]\text { , } l \in [ l]\label{constraint : quadratic : regression2 } \\
\sum_{k=1}^k z_{ik } & = 1 \quad & & i \in [ i ] \label{constraint : quadratic : assignment } \\ \sum_{i=1}^i z_{ik } & \geq n \quad & & k \in [ k ] \label{constraint : quadratic : minimumsize }
\\ z_{ik } & \in \{0,1\ } \quad & & i \in [ i]\text { , } k \in [ k ] \nonumber \\ t_{il } & \geq 0 \quad & & i \in [ i]\text { , } l \in [ l ] \nonumber\\ \beta_{kj}&\text{\ unconstrained } \quad & & k \in [ k]\text { , } j \in [ j ] , \nonumber\end{aligned}\ ] ] where @xmath30 is a binary variable , which is equal to one if and only if entity @xmath10 is assigned to cluster @xmath25 .
value @xmath31 , referred to as big @xmath31 in the optimization literature , is a large positive constant . due to constraints and , @xmath32 is equal to the absolute error for the corresponding observation @xmath13 in the optimal solution , and @xmath33 are the regression coefficients for cluster @xmath25 , which are decision variables . the role of @xmath31 is to enforce constraints and only when they are needed ( entity @xmath10 is assigned to cluster @xmath34 ) . in detail , if @xmath35 , then we have @xmath36 , and @xmath37 , which implies @xmath38 because we are minimizing the sum of @xmath39 . if @xmath40 , constraints and require @xmath32 to be greater than a negative number , which holds trivially due to the existence of the nonnegativity constraint on @xmath32 .
constraint requires that every entity is assigned to one cluster , and imposes the limit on the cardinality of each cluster . unlike the clr problem
, the generalized clr allows each entity to have more than one observation , which implies that @xmath4 can be greater than one .
the mixed integer linear program formulation for the clr problem in bertsimas and shioda @xcite has @xmath4 equal to one , and does not have the cluster cardinality constraint . besides , their objective function is the sum of the absolute errors while ours is the sum of squared errors
. our sku clustering problem based on the seasonal effects can be modeled as the generalized clr problem .
the entities to cluster are skus .
the response variable @xmath41 corresponds to a vector of weekly sales for sku @xmath10 .
the independent variables @xmath42 s include promotional predictors such as promotion mechanisms , percentage discount , and seasonal dummies for sku @xmath10 .
aloise _ et al . _
@xcite showed np - hardness of the mssc problem in a general dimension when the number of clusters is two .
general dimension means that the size of the vectors to be clustered is not a constant but part of the input data .
a similar statement can be made for the generalized clr problem with the proof available in appendix [ appendix_proof_theorems ] .
[ theorem : clr ] the generalized clr problem with two clusters in a general dimension is np - hard . with the formulation presented by , we can solve the generalized clr problem using any commercial optimization software that can handle quadratic mixed integer programs .
however , this formulation suffers from two drawbacks , which makes it intractable for large instances .
the first one relates to big @xmath31 .
optimality of the solution and efficiency of integer programming solvers depend on a tight value of @xmath31 . unlike multiple linear regression ,
where obtaining a valid value of @xmath31 is possible @xcite , it is not trivial to calculate a valid value of @xmath31 in and for the generalized clr or clr .
when @xmath40 , @xmath43 s are not from the cluster that entity @xmath10 belongs to , and the residual @xmath32 can be arbitrarily large .
@xcite provide an empirical result that a big m based mip formulation for clr sometimes fails to guarantee optimality of clr for the data sets they consider .
the second one involves the symmetry of feasible solutions .
any permutation of clusters yields the same solution , yet it corresponds to different decision variables .
symmetry unnecessarily increases the search space , and renders the solution process inefficient . to overcome the symmetry problem
, we propose a set partitioning formulation , which has already been used for the clr problem in @xcite .
let @xmath44 denote the set of all clusters of entities with the cardinality equal to or greater than @xmath21 , i.e. , @xmath45 , |s| \geq n\}$ ] .
let @xmath46 equal to one if entity @xmath10 belongs to cluster @xmath47 , and equal to zero otherwise .
let @xmath48 denote the cost of cluster @xmath47 , which is equal to the sum of squared errors when performing the regression over cluster @xmath47 .
introducing binary variables @xmath49 the generalized clr problem can be formulated as : @xmath50 \label{constraints : partition : assignment}\\ z_s & \in \{0,1\}\quad & & s \in \mathscr{s}.\nonumber\end{aligned}\ ] ] constraint ensures that the number of clusters in the partition is @xmath16 and constraint guarantees that each entity occurs in only one cluster within the partition .
the set partitioning formulation has an exponential number of binary variables .
it is very challenging to solve even its linear programming relaxation because there are so many decision variables . to solve large - scale linear and integer programs ,
column generation algorithms have been used in the literature .
the reader is referred to vanderbeck and wolsey @xcite and barnhart _ et al . _
@xcite for reviews of column generation for solving large - scale integer programs . in our work , we employ column generation to handle its linear programming relaxation . at the high level , column generation can be understood as iteratively expanding set @xmath51 ( a subset of @xmath44 ) in - by adding attractive candidate cluster @xmath47 to @xmath51 .
the key challenge is how to select @xmath47 .
the word column is used because adding cluster @xmath47 to @xmath51 is equivalent to adding a column in the matrix form of - .
the column generation algorithm , referred to as the cg algorithm , starts by solving the restricted master problem which has the same formulation as the master problem - , but with set @xmath44 replaced by @xmath51 , a smaller subset of columns .
recall that a column represents a cluster ( subset of entities @xmath52 $ ] ) .
we start the algorithm with small candidate clusters rather than @xmath44 , the set of all possible subsets of @xmath52 $ ] .
the algorithmic framework is presented in algorithm [ alg : exactscg1 ] , which follows the general column generation scheme . in line 1 ,
the initial subset of columns in @xmath51 are randomly generated . in detail , we start from @xmath16 empty clusters
. then , we randomly assign each entity to one of the @xmath16 clusters using a uniform random number .
hence , after line 1 , we have @xmath16 clusters and @xmath53 for the generation procedure . in line 3 ,
optimal dual variables are obtained by solving the restricted master problem and then serve as input to the pricing problem , which will be introduced hereafter , to calculate the smallest reduced cost column . in line 4 ,
the pricing problem returns a column with the smallest reduced cost . in lines 5 - 10 ,
if the reduced cost is nonnegative , then we conclude that the master problem is solved optimally .
otherwise , we add the column with the smallest reduced cost to the restricted master problem and repeat the process .
randomly generate @xmath51 ( a small subset of @xmath44 ) solve master problem and obtain dual solution get a new cluster by solving pricing problem with input of dual solution from line 3 the algorithm is complete with the optimal partition of clusters add the cluster from line 4 to the master problem the pricing problem can be stated as follows .
let @xmath54 be the dual variable for constraint , and @xmath55 s be the dual variables for constraint .
the reduced cost for cluster @xmath47 is @xmath56 , and thus the pricing problem reads : @xmath57 note that we omit the subtraction of @xmath54 in the formulation because it is a constant which does not change the optimal solution .
[ th : pricingnpcomplete ] the pricing problem as stated in is np - complete .
the proof is available in appendix [ appendix_proof_theorems ] . introducing binary variables @xmath58 the pricing problem
can be formulated as a mixed integer quadratic program : @xmath59 \text { , } l \in [ l ] \label{constraints : pricing : residual1}\\
t_{il}+(y_{il}-\sum_{j=1}^j\beta_j x_{ijl})+m(1-z_i)&\geq 0 \text{\ \ } & & i \in [ i ] \text { , } l \in [ l ] \label{constraints : pricing : residual2}\\ \sum_{i=1}^i z_i
& \geq n \label{constraint : pricing : minsize } \\ t_{il}&\geq 0 \text{\ \ } & & i \in [ i ] \text { , } l \in [ l]\nonumber\\ z_i&\in \{0,1\}\text{\ \ } & & i \in [ i],\nonumber\end{aligned}\ ] ] where @xmath31 is a large positive constant and
is assumed to be valid ( does not cut an optimal solution ) .
we can use the same approach from section 2.1 to set up a valid value for @xmath31 .
a feasible solution s sse can be a valid value . by using similar arguments as those for constraints and , @xmath32 is the absolute error for the corresponding observation @xmath13 in the optimal solution if @xmath60 , and it is zero otherwise .
the difference from the pricing problem in @xcite is that is based on big m constraints and is for the generalized clr , while carbonneau _
@xcite used logical implications of the constraints for the clr problem . in the column generation algorithm ,
are solved .
recall that reduced cost for cluster @xmath47 is @xmath56 .
it is easy to see that value @xmath61 is equivalent to to the value of with @xmath62 for @xmath60 and 0 otherwise .
this follows from the fact that @xmath63 and @xmath48 is modeled by variables @xmath0 .
* example * let us consider a data set with 4 entities and suppose @xmath64 .
then , we have @xmath65 , where @xmath66 . in algorithm
[ alg : exactscg1 ] , suppose we start with subset @xmath67 of @xmath44 , which is a set of initial candidate clusters .
the master problem in line 3 picks the best combination of the candidate clusters that has minimum total sse .
suppose we obtain @xmath68 in line 3 together with the associated dual solution .
here we assume that the solution is integral , albeit this might not always be the case . in line 4 ,
the pricing problem is solved to search if there exists a candidate cluster not in @xmath51 that can improve the current best solution @xmath68 .
suppose the pricing problem returns @xmath69 with a negative reduced cost . in line 8 , @xmath51 is updated to @xmath70 .
if the optimal solution obtained by cg is not integral , branching would have to be performed , i.e. , a fractional variable @xmath71 needs to be selected and two new problems created , the first one would impose @xmath72 and the other one @xmath73 .
however , the extensive evaluation conducted on algorithm [ alg : exactscg ] revealed that no fractional solutions were provided by algorithm [ alg : exactscg ] .
for this reason in the remainder we focus on column generation for solving the lp relaxation and not branching .
column generation is known to exhibit the tailing - off effect and for this reason we employ stabilized column generation of du merle _ et al .
_ @xcite . the stabilized column generation algorithm for solving the clr problem
is illustrated in algorithm [ alg : exactscg ] .
the algorithm takes input of stabilization parameters @xmath74 and @xmath75 , and maximum allowed iterations @xmath76 . in line 1 , we start with a set @xmath77 of initial clusters of entities .
the generation procedure is identical to the one in line 1 of algorithm 1 .
for iteration @xmath34 , in line 3 , we solve the stabilized master problem and get the optimal solution @xmath78 and its corresponding dual solution @xmath79 , which provides input parameters for the pricing problem .
the stabilized master problem additionally includes parameters @xmath80 , @xmath81 and variables @xmath82 , @xmath83 but is very similar to - .
see appendix [ appendix_restricted_master_problem ] for the actual formulation . by solving the pricing problem
, we get a new cluster @xmath84 in line 4 .
the reduced cost corresponding to this new cluster is equal to @xmath85 , where @xmath48 is the sum of squared residuals when performing regression over this cluster , and @xmath86 if and only if @xmath60 . in lines 5 - 6 , if the reduced cost is nonnegative and @xmath82 and @xmath83 are equal to zero , then the algorithm is complete with the optimal partition of clusters defined by @xmath87
. otherwise , in lines 8 - 12 , we update @xmath51 and then if the reduced cost @xmath85 is nonnegative , we update the stabilization parameters @xmath80 , and @xmath81 .
@xmath88 , randomly generate @xmath77 ( a small subset of @xmath44 ) @xmath89 solve the stabilized master problem - given in appendix [ appendix_restricted_master_problem ] @xmath90 solve the pricing problem with @xmath91 @xmath92 , and * stop * @xmath93 @xmath94 @xmath95
@xmath96 the numerical experiments introduced later reveal that the column generation algorithm does not scale well to problems with a large number of entities . to overcome the scalability problem
, we propose a heuristic method called the cg heuristic algorithm that relies on column generation .
the cg heuristic algorithm first finds a partition with a large number of clusters by neglecting the cardinality constraint . in the second step ,
we combine the clusters by considering unions to obtain exactly @xmath16 clusters while obeying the cardinality constraint , which is a slight variant of column generation .
we refer to the intermediate clusters from the first part , which are the input to the column generation algorithm in the second part , as groups .
the algorithmic framework is presented in algorithm [ alg : cgheur ] .
we require that @xmath97 since the second step is to combine @xmath98 groups into @xmath16 clusters .
lines 1 - 5 represent the first step to create @xmath98 groups and line 6 represents the second step to find a solution to the original problem .
line 1 follows the same procedure as line 1 of algorithm 1 , except that we have r groups instead of k. lines 2 - 5 are basic and do not need further explanations .
it yields @xmath98 `` low cost '' groups . since @xmath99 ,
in line 6 we combine some groups so that we end up with exactly @xmath16 clusters , each one with cardinality at least @xmath21 .
this regrouping of groups is performed in an optimal way by using the column generation framework .
@xmath100 randomly generate @xmath98 groups perform regression over each group @xmath101 $ ] to obtain regression coefficients @xmath102 for @xmath103 $ ] , reassign entity @xmath10 to group @xmath104 , where @xmath105 execute cg by treating each group as entity for line 6 , we need to revise the master and pricing problems in the following way when we cluster a group of entities instead of single entities .
suppose at the end of line 5 we clustered @xmath2 entities into @xmath98 groups @xmath106 , and then apply the column generation algorithm to the @xmath98 groups of entities .
let @xmath107 be the set of all subsets @xmath47 of @xmath108 $ ] such that @xmath109 , and let @xmath110 if @xmath111 , and @xmath112 otherwise . to obtain the new master problem
, we need to replace @xmath44 with @xmath107 and @xmath46 with @xmath113 in the master problem . in addition , the range of constraints changes to @xmath101 $ ] .
we denote the dual variables of constraints in the master problem by @xmath114 , and introduce the binary decision variables @xmath115 for @xmath101 $ ] to indicate whether group @xmath116 is selected in the cluster with the minimum reduced cost . to obtain the new pricing problem
, we need to replace @xmath117 s with @xmath115 s in the pricing problem .
constraint is changed to @xmath118 , and the range in constraints and now becomes @xmath101 \text { , } i \in g_r \text { , } l \in [ l]$ ] .
the new pricing problem has the same number of constraints as the pricing problem
, however , it has only @xmath98 binary variables , comparing to @xmath2 such variables in the pricing problem refeq : pricing. scientific works as those presented by maulik and bandyopadhyay @xcite and chang _ et al . _
@xcite , effectively suggest to embed the concept of the lloyd s algorithm , into a genetic search metaheuristic framework to find proper clusters for the mssc clustering problem . here
we discuss our proposed adaptation of the ga - based lloyd s clustering algorithms for solving the generalized clr problem . for the lloyd s algorithm part
, a vector of regression coefficients @xmath119 is used to represent cluster @xmath34 , and an entity is recursively assigned to the cluster that gives the smallest sum of squared errors for this entity .
the ga part helps escape local optimal solutions .
the overall algorithmic framework is outlined in algorithm [ alg : galloyd ] .
k , _ maxiter _
, h , @xmath120 for each @xmath121 in @xmath122 $ ] , create @xmath123 by randomly generating @xmath16 clusters randomly select parent chromosomes @xmath124 and @xmath125 using roulette wheel selection create child chromosomes @xmath126 and @xmath127 by performing crossover on @xmath124 and @xmath125 mutation on @xmath128 obtain @xmath129 and @xmath130 based on lloyd s algorithm , calculate @xmath131 and @xmath132 replace @xmath133 , @xmath134 } \gamma(h)$ ] , with the chromosome with larger fitness among @xmath126 and @xmath127 in line 1 , we start by randomly generating @xmath135 partitions @xmath136 , each of which corresponds to @xmath16 clusters of entities with @xmath137 for @xmath138 $ ] .
the generation procedure is based on a uniform random number and is the same as the one in line 1 of algorithm 1 .
any randomly generated partition @xmath123 has to satisfy the constraint that @xmath139 for @xmath23 $ ] .
a population @xmath140 consists of @xmath135 chromosomes , and chromosome @xmath121 is encoded as a vector @xmath141 of size @xmath142 . in any chromosome ,
the first @xmath6 genes represent the regression coefficients @xmath143 for the first cluster , and the next @xmath6 genes represent the regression coefficients @xmath144 for the second cluster , and so on .
the encoding of chromosome @xmath121 is illustrated in figure [ figure : encoding ] .
+ the regression coefficient @xmath145 is obtained by running regression over cluster @xmath146 .
the fitness @xmath147 of the chromosome @xmath121 is defined to be @xmath148 we continue by performing the following genetic operations on the population of chromosomes iteratively until the number of iterations without improvement reaches a specified maximum number _ maxiter_. first , in line 3 , we randomly select two parent chromosomes @xmath124 and @xmath125 from population @xmath140 using roulette wheel selection .
chromosome @xmath121 is chosen with probability @xmath149 .
second , in line 4 , we perform crossover on chromosomes @xmath124 and @xmath125 .
we select a gene position as a random integer in the range of @xmath150 $ ] .
we require this random integer to be no more than @xmath151 so that there is at least one gene positioned to the right of it .
the portions of the chromosome lying to the right of this gene position are exchanged to produce two child chromosomes @xmath126 and @xmath127 encoded by @xmath152 and @xmath153 .
the crossover operation is illustrated in figure [ figure : crossover ] .
+ third , in line 5 , we perform mutation on these two child chromosomes .
the mutation is performed on a child chromosome with a fixed probability @xmath154 , where @xmath154 is a parameter . a gene position with value @xmath54
is randomly picked from the child chromosome using a uniform random number .
after mutation , it is changed to @xmath155 with equal probability if @xmath54 is not zero . here
@xmath156 is a random number with uniform distribution between zero and one .
otherwise , when @xmath54 is zero , it is changed to @xmath157 with equal probability . in this way
, the regression coefficients can take any real values after sufficient number of iterations .
next , in line 6 , we need to decode these two mutated child chromosomes to get the partitions @xmath129 and @xmath130 of clusters they represent . to decode the child chromosome @xmath126
, we assign entity @xmath10 to cluster @xmath158 for @xmath159 then , we perform regression over each cluster of @xmath129 and @xmath130 , and update the encoding of these two child chromosomes @xmath152 and @xmath153 with the resultant regression coefficients .
fitness @xmath160 and @xmath161 are calculated for the child chromosomes using . in lines 6 - 7
, we replace the chromosome in population @xmath140 with the smallest fitness with the child chromosome with the smaller fitness if @xmath162 } \gamma(h)$ ] . during the decoding step of the ga - lloyd algorithm
, we may need to adjust the clusters generated in order to satisfy the minimum size constraints . if cluster @xmath163 has size smaller than @xmath21 , then we sort the entities not in @xmath163 in the increasing order of the sum of squared regression errors , and then reassign these entities in the sorted order to cluster @xmath163 until the size of @xmath163 reaches @xmath21 .
we also skip each entity that would reduce the size of its original cluster below @xmath21 .
when there is more than one cluster with size smaller than @xmath21 , we perform this adjustment for the smallest cluster first . due to its simplicity , two - stage
heuristic algorithms are frequently employed in practice for solving the clr problem . in the first stage ,
entities are partitioned according to certain approximate measures of the regression coefficients . in the second stage
, regression models are built over the resultant clusters .
the clustering method for the first stage is usually problem specific . * * stage 1 * partition @xmath2 entities ( skus ) into @xmath16 clusters .
let @xmath164 $ ] be the index set of entities in cluster @xmath34 , @xmath23 $ ] . * *
stage 2 * for each @xmath23 $ ] , build a regression model using entities in @xmath164 $ ] . in this section
, we describe our two stage heuristic algorithm for the sku clustering problem .
recall that we are given [ cols= " < , < " , ] + this gives a justification to evaluate algorithms for larger instances based on the gap from the target solution .
however , in our experiment for larger instances , we observed that most of the proposed algorithms give a better objective function value than the target solution .
hence , we did not use the target solution to evaluate the algorithms in the paper .
however , the target solutions are available on the website stated in section [ sec : experiments ] .
in this section , we present a stabilized version of the master problem - , referred to as restricted master problem , by applying the technique of du merle _ et al .
_ @xcite . for iteration @xmath34 ,
the restricted master problem is written as @xmath165 \label{constraints : restrictedpartition : assignment}\\ & 0 \leq \boldsymbol{q}^- \leq \boldsymbol{\xi}^{(k)}\\ & 0 \leq \boldsymbol{q}^+ \leq \boldsymbol{\xi}^{(k ) } \label{restr222}\\ & z_s \in \{0,1\}\quad & s \in \mathscr{s},\nonumber\end{aligned}\ ] ] | cluster - wise linear regression ( clr ) , a clustering problem intertwined with regression , is to find clusters of entities such that the overall sum of squared errors from regressions performed over these clusters is minimized , where each cluster may have different variances .
we generalize the clr problem by allowing each entity to have more than one observation , and refer to it as generalized clr .
we propose an exact mathematical programming based approach relying on column generation , a column generation based heuristic algorithm that clusters predefined groups of entities , a metaheuristic genetic algorithm with adapted lloyd s algorithm for k - means clustering , a two - stage approach , and a modified algorithm of spth @xcite for solving generalized clr .
we examine the performance of our algorithms on a stock keeping unit ( sku ) clustering problem employed in forecasting halo and cannibalization effects in promotions using real - world retail data from a large supermarket chain . in the sku clustering problem
, the retailer needs to cluster skus based on their seasonal effects in response to promotions .
the seasonal effects are the results of regressions with predictors being promotion mechanisms and seasonal dummies performed over clusters generated .
we compare the performance of all proposed algorithms for the sku problem with real - world and synthetic data . |
the manuscript deals with the evolution of fields in background space - times with certain symmetry , mainly the higher - dimensional schwarzschild black holes ( or general spherically symmetric curved ) geometries in a linearized approach .
we solve the klein - gordon equation and the equation for metric perturbation introducing a phase space .
a number of results are derived .
since it is an important problem , some motivations are described below .
the reader should probably like to read the original works by reggae and wheeler and others , for instance : general theory of perturbation analysis and motivation @xcite @xcite @xcite @xcite , quasi - normal modes @xcite @xcite , and late - time tails in black hole backgrounds @xcite @xcite @xcite @xcite @xcite .
space - time is described by @xmath5 where @xmath6 , @xmath7 and @xmath8 are the metric , ric scalar and ric tensor of empty universe . the einstein equations ( [ 1 ] ) can be derived from the einstein - hilbert action @xmath9 where @xmath10 . a consistent theory of quantum gravity , such as string theory , seems to require the existence of higher dimensions , which in order to not contradicting observational evidence must be compact on small scales .
string theory has made some important progress in explaining the entropy of certain black holes by counting microscopic degrees of freedom , and thus scenarios with extra , compact dimensions must be taken seriously . in turn , some of the most interesting objects to be studied within string theory are those that possess event horizons , possibly extended in the compact extra dimensions .
higher - dimensional theories of gravitation have recently attracted much attention , and static symmetrical homogeneous solutions have played an important role in revealing various aspects of gravitation peculiar to higher dimensional systems @xcite @xcite @xcite @xcite .
a noteworthy fact is that , compared to the severely constrained situation in four space - time dimensions @xcite
@xcite @xcite , higher dimensional black holes can have many very different varieties @xcite @xcite @xcite @xcite @xcite . for this reason ,
it is important to study the dynamics of fields @xcite @xcite @xcite @xcite @xcite and space - time perturbations @xcite @xcite @xcite in such black hole backgrounds and to find stable solutions . amongst the various types of static symmetrical homogeneous solutions ,
a natural higher dimensional generalization of a black hole also known as a schwarzschild - tangherlini black hole @xcite has been assumed to be ( or believed to be ) stable , like its four - dimensional counterpart , and for this reason it has been conjectured that this black hole describes a final equilibrium state of evolution of an isolated system , or a certain class of unstable static solutions in higher dimensions .
however , the question of the asymptote of a perturbation of a black hole itself is not a trivial matter , and , to our knowledge , it has not yet been fully analyzed in higher dimensions .
our manifold will be the mathematical manifold @xmath11 , the first layer of the would - be physical space - time of general relativity .
the space - time is assumed to be @xmath12 dimensional .
it has been common practice in recent years to denote @xmath13 as the number of extra dimensions , besides the usual four .
we are consistent with ref .
@xcite . in the mathematical literature about topological spaces
, it is always implicitly assumed that the entities of the set can be distinguished and considered separately ( provided the hausdorff conditions are satisfied ) , otherwise one could not even talk about point mappings or homeomorphisms .
it is well known , however , that the points of a homogeneous space can not have any intrinsic individuality . for concreteness
, we consider as background the cartesian product of a two - dimensional orbit space @xmath14 and a horizon @xmath15 with @xmath161 compact dimensions , where 1+@xmath0 is the total space - time dimension and @xmath4 4 ( in order to avoid space - times with one or less compact spatial dimensions where the presence of a massive source is inconsistent with asymptotic flatness ) .
tangherlini s idea was to consider pure gravity in 1+@xmath0 dimensions . the ( 1+@xmath0)-dimensional action is simply @xmath17 where now @xmath6 is the 1+@xmath0 dimensional metric , and @xmath7 is the corresponding ric scalar .
we have the original first two dimensions labeled with coordinates @xmath18 where @xmath19 .
the @xmath161 remaining dimensions are compacted on a hyper sphere and are labeled by the coordinate @xmath20 .
now we can expand the metric as a spherical harmonics series of the form @xmath21 we find that we get an infinite number of fields in the first two dimensions .
modes with @xmath22 correspond to massive fields with mass @xmath23 .
the zero mode corresponds to a mass - less field .
as we take @xmath24 to be smaller and smaller we see that the mass of the first massive field becomes very large .
this means that if we compact on a small enough hyper sphere we can truncate to mass - less modes in the two - dimensional theory .
we can only see the extra dimensions by exciting massive modes which are at energies beyond our reach .
let us now focus on the zero mode @xmath25 , the metric of @xmath14 .
we could define @xmath6 so that g@xmath26 is the two - dimensional field in the schwarzschild- tangherlini metric . in order that our results are more transparent we introduce some notation .
we denote by @xmath27 the radial " coordinate , this is the coordinate away from the horizon . in the vicinity of the static homogeneous horizon
it is useful to introduce
spherical " coordinates @xmath28 as well .
we will actually introduce the components of the metric in the following way : @xmath29 where @xmath30 and @xmath31 are two derivable functions of the variable @xmath27 , and @xmath32 denotes the metric of the horizon ) as a _ maximally symmetric static homogeneous space - time _ @xcite @xcite @xcite . ishibashi and kodama briefly comment on such a generalized static homogeneous space - time in section 4 of @xcite .
] , i. e. , it is the metric of the einstein space ( @xmath15 , h ) whose ric curvature is @xmath33 @xmath34 being the ric scalar of the horizon .
we have truncated to the mass - less fields .
tangherlini type compactifications can be more complicated than simply compacting on a hyper sphere .
the important thing is that the extra dimensions are small so that we do not excite massive modes .
we can truncate to mass - less modes and read off the effective theory in the first two dimensions .
we need not restrict ourselves to just two dimension either .
in fact , higher dimensions have become very fashionable in the last twenty years , mainly due to the success of _ string theory _ as a possible quantum theory of gravity . at the quantum level ,
boson string theory is only consistent in twenty - six ( ! ) dimensions , although this figure is reduced to ten when we introduce super symmetry . furthermore , there are five distinct string theories which can be viewed as different elements of an embracing new theory , m - theory @xcite @xcite @xcite .
m - theory lives in eleven dimensions and has eleven - dimensional super gravity as its low energy limit .
traditionally we achieve the reduction down to the first two dimensions using tangherlini techniques .
if we start with a @xmath35-dimensional theory , we compact on a small ( @xmath36)-dimensional manifold .
different manifolds generally give different effective theories in the first two dimensions .
the one thing all of these manifolds have in common is that they are very small , and compact .
there is , however , an alternative to tangherlini compactification .
this is the idea that we live on something called a _
braneworld _ , where the extra dimensions can be infinite .
in 1+@xmath37 space - time dimensions , an event horizon can be topologically a sphere @xmath38 , but when extended to @xmath39 extra dimensions , it could naturally have topologies either @xmath40 in which case it is a higher dimensional black hole , or @xmath41 being then a black @xmath39-brana or a black string in the @xmath39=1 case @xcite @xcite @xcite .
the @xmath42 topology of the transverse dimensions can be compacted giving flat toroidal topological spaces @xmath43 .
one of the first steps towards understanding these extended higher dimensional solutions is to investigate their classical stability against small perturbations .
if a solution is unstable , then it most certainly will not be found in nature ( unless the instability is secular ) and the solution looses most of its power .
of course the next question one must ask is what is the final stable result of such instability .
now , in ( 1+@xmath37)-space - time dimensions , the schwarzschild - tangherlini geometry is stable against all kinds of perturbations , massive or mass - less @xcite @xcite @xcite @xcite @xcite @xcite @xcite @xcite .
on the other hand , quite surprisingly , gregory and laflamme @xcite showed that this is not the case for higher dimensional black branes and black strings .
these objects are unstable .
the kind of black @xmath39-branes originally studied in @xcite were solutions of ten dimensional low energy string theory with metric of the form @xmath44 where @xmath45 stands for the ( 1+@xmath37)-dimensional schwarzschild- tangherlini line element , the @xmath46 are the coordinates of the compact dimensions , and @xmath47 runs from @xmath48 to @xmath39 .
the total dimension of the space - time is 1+@xmath0 . in string or super gravity theories
one takes 1+@xmath49 or 1+@xmath50 , but for generality one can leave it as a free natural parameter .
the specific metric we shall be interested in is given by a time - independent symmetric metric that can be written in schwarzschild - tangherlini coordinates @xcite as ( [ 2 ] ) .
after this general explanation of our background space - times , we give in the next section our formalism for gravitational perturbations , and the basic strategy of our analysis of the asymptote of a perturbation . at this stage , our arguments
are not limited to the case of black holes , but are equally applicable to any case of maximally symmetric static homogeneous space - times with a non - vanishing cosmological constant .
the study of symmetric perturbations of the schwarzschild - like solution is reduced to solving a single partial differential equation ( [ 3 ] ) .
we attempt to solve this perturbation equation with initial data in a general static symmetrical homogeneous space - time .
deriving the analytic formula for the solutions is important and useful for further study .
the computation of gravitational radiation was done only numerically in a preceding study , e.g. ref @xcite .
the first decision to be made is to choose our coordinates and ansatz for the analysis .
we choose to work with one coordinate patch for the whole space - time . in our coordinates ( where the metric in the @xmath27 line is in the form @xmath51 ) the coordinate size of the horizon is an invariant and hence the coordinate patch does not change .
consider a function @xmath52 of the variable @xmath27 .
we denote by @xmath53 the laplace operator in the metric h. consider a base @xmath54 of eigen - vectors of @xmath53 .
let us consider a field with configuration space @xmath55 with @xmath0 even and global coordinates @xmath56(@xmath27 , @xmath57 , @xmath28 ) described by a singular euler - lagrange equation , @xmath58 @xmath59 the symmetry of the background guarantees that time and spatial variables will separate .
in other words , if we rewrite equation ( [ 3 ] ) in @xmath27 and @xmath57 coordinates we find that the coefficients of the differential equation do not depend on time @xmath57 , and so the problem can be reduced to one dimension by applying a formal transform with respect to the time variable @xmath57 .
however , the perturbation equation ( [ 3 ] ) is not invariant under simultaneous rescaling of time and space variables @xmath60 , @xmath61 .
this lack of invariance reflects the presence of mass / length scale in the model .
the solution is most conveniently analyzed in terms of the phase space coordinates @xmath62 , @xmath63 , @xmath64 . in absence of explicit solutions of the perturbation equation ,
the best we can do is to construct a solution by performing a _ quasi-_transformation from coordinate variables ( @xmath57 , @xmath27 , @xmath28 ) to phase space variables ( @xmath63 , @xmath62 , @xmath64 ) . on the other hand
, such transformation must have the remarkable property that , in the _ special regimes _ with @xmath65 , the variables @xmath66 and @xmath67 form a basis of off - shell observables for the gravitational field _ even if _ the solution of the perturbation equation is not known .
[ [ basic - set - up . ] ] basic set - up .
+ + + + + + + + + + + + + we denote by @xmath68 the function of bessel of order @xmath69 , where @xmath70 if @xmath54 is a discreet set ( continuous set ) , then @xmath71 denotes a sum ( an integral ) . for the transform of @xmath56 we have @xmath72 with the inverse transformation being @xmath73 that is to say ,
we propose a solution of the equation ( [ 3 ] ) of the form ( [ 4 ] ) so the derivatives of the original problem will enter as multiplicative terms .
a vector @xmath74 satisfies @xmath75 the functions of bessel satisfy @xmath76 we substitute ( [ 4 ] ) into ( [ 3 ] ) .
we use ( [ 5 ] ) , ( [ 6 ] ) and ( [ 7 ] ) . ( [ 3 ] ) becomes @xmath77 @xmath78 @xmath79 we suppose that for each @xmath64 and each @xmath80 , @xmath81 we integrate by parts @xmath82 .
( [ 8 ] ) becomes @xmath83 @xmath84 @xmath85 therefore , @xmath84 @xmath86 we define @xmath87 then equation [ 9 ] is as @xmath88 assume the support of the function @xmath89(@xmath90 , @xmath91 , @xmath92 ) lies in a narrow neighborhood of the curve @xmath93 then equation [ 11 ] is approximately equal to @xmath94 therefore , there exists a function @xmath95 such that @xmath96 therefore , ( [ 4 ] ) is as @xmath97 set @xmath98 then ( [ 12 ] ) is as @xmath99 then ( [ 12 ] ) solves equation [ 3 ] with initial condition ( [ 13 ] ) .
now , the usefulness of this approach with respect to previous ones is that we can derive the late - time behavior of fields in this background ( see ref .
@xcite ) with our formalism .
the reader must be aware of the relevant literature on the subject , and of what has been done in recent years .
suppose that @xmath100 then @xmath101 we calculate @xmath102 the previous discussion applies to a class of globally hyperbolic , topologically trivial , non - compact ( asymptotically flat at spatial infinity ) space - times of the type of christodoulou - klainermann ones @xcite . in them
we have @xcite @xcite @xcite @xcite : \1 ) the imposition of suitable boundary conditions on the fields and the gauge transformations of metric gravity reduces the asymptotic symmetries at spatial infinity to the asymptotic poincare group .
the asymptotic implementation of poincare group makes the general - relativistic definition of angular momentum and the matching of general relativity with particle physics .
\2 ) the boundary conditions of point 1 ) require that the leaves of the foliations associated with the admissible 1+@xmath0 splittings of space - time must tend to minkowski space - like hyper - planes asymptotically orthogonal to the ( 1+@xmath0)-momentum in a direction - independent way .
this property is concretely enforced by using a technique introduced by @xcite @xcite for the selection of space - times admitting asymptotically flat ( 1+@xmath0)-coordinates at spatial infinity .
these facts entail that _ there is an effective evolution in the mathematical time _
@xmath57 which parametrizes the leaves of the foliation associated with any 1+@xmath0 splitting . the idea behind a new quantization scheme is to arrive directly to the physical hilbert space by quantizing only the observables of the system and treating the configuration variables as _ c
- numbers _ ( like _ time _ in the time - dependent schrodinger equation ; the momenta become derivatives with respect to the configuration variables , like the energy is replaced by the time derivative ) . in gravity
this scheme would make sense only if the configuration variables are coordinate - independent .
there will be as many coupled schrodinger equations as configuration variables ( gauge invariant ) and the wave function will depend on as many _ times _ ( besides the standard one ) as configuration variables .
every line in this parameter space will correspond to a configuration of the classical theory .
now consider the evolution of a mass - less scalar field @xmath56 in the background described by ( [ 2 ] ) .
the evolution is governed by the curved space klein - gordon equation @xmath103 the metric appearing in ( [ 14 ] ) should describe the geometry referring to both the black brana and the scalar field , but if we consider that the amplitude of @xmath56 is so small that its contribution to the energy content can be neglected , then the metric ( [ 2 ] ) should be a good approximation to @xmath104 in ( [ 14 ] ) .
we shall thus work in this perturbation approach .
it turns out that it is possible to simplify considerably equation ( [ 14 ] ) if we separate the horizon variables from the radial and the time variables , as is done in four dimensions @xcite @xcite . for higher dimensions we follow @xcite .
the evolution of a minimally coupled scalar field @xmath56 is described by the mass - less klein - gordon equation ( [ 8 ] ) @xcite @xcite @xcite @xcite @xcite , with @xmath105 where ( [ 8 ] ) is a partial differential operator that contains information about the initial shape of the wave packet at @xmath57=0 .
the explicit form of the operator ( [ 8 ] ) is the simplest in this general setting . in section [ seccion 2 ] ,
the linear perturbation equation is derived and solved . setting @xmath106
we see that far from the horizon a mass - less field decays as @xmath1 in time , where @xmath2 is the harmonic mode of the sphere .
this indicates that it behaves like @xmath107 for fixed @xmath27 .
@xcite clarify that the field behaves like @xmath107 for even @xmath0 , and @xmath108 for odd @xmath0 .
the situation for the tangherlini hyper sphere is clear . near the hyper sphere
the fluctuations in the field behave in the same way as for flat hyper spheres .
however , the field does vanish at the point at infinity , and close to it the field perturbations start to decrease .
since this point lies far from the hyper sphere we might yet believe that gravity is localized at low enough energies . at finite temperature
we could hide the point at infinity behind another black hole horizon .
why a fluctuation about a black hole is regular at infinity ? this is an important problem in gravitation and has been studied for long time . although we do not have a definite answer to this question yet ,
we have rather well established answers to the questions of stable ( unstable ) gaussian fluctuations about our black hole , which are summarized as the instabilities of gravitation . in local particle physics problems in the present universe , the assumption of the asymptotic flatness will be a good approximation .
however , the success of the inflationary universe model implies that the cosmological constant can not be neglected even locally in the early universe .
further , recent developments in unifying theories suggest the possibility that the universe has dimensions higher than four on microscopic scales @xcite @xcite and that mini black holes might be produced in elementary particle processes in colliders as well as in high energy cosmic shower events . therefore , it is a quite important problem whether the regular character of a fluctuation about a black hole also holds in the non - vanishing cosmological constant case and/or in higher dimensions .
here we explain the derivation of the master equation for the perturbation of a higher - dimensional schwarzschild black hole .
our purpose is to study and present an analytic procedure to obtain solutions for the master equation of a small perturbation in a static symmetrical homogeneous space - time .
we investigate the asymptote of a linear perturbation of higher dimensional space - times in the framework of a gauge - invariant formalism for gravitational perturbations of maximally symmetric static homogeneous space - times .
this formalism was recently developed by ishibashi and kodama @xcite . in the four - dimensional case ,
the perturbation analysis of a black hole was first carried out by reggae and wheeler @xcite , and its stability was essentially confirmed by vishveshwara @xcite , followed by further detailed studies @xcite @xcite @xcite .
the basic observation of those works is that unstable perturbations , whose growth in time is unbounded , are initially divergent at the event horizon ( provided that they are vanishing at large distances ) and therefore are physically unacceptable .
a key step in the analysis is the reduction of the perturbation equations to a simple , tractable form . for this purpose , in the case of a four - dimensional black hole , the symmetry of the background space - time is utilized to classify the black hole perturbations into two types , axial ( odd ) modes and polar ( even ) modes , according to their behavior under the parity transformation on the two - sphere .
to linear order , the perturbation equations for these two types of perturbations are independent , and therefore they reduce to a set of two schrodinger - type second - order ordinary differential equations , the reggae - wheeler equation for axial perturbations and the zerilli equation @xcite @xcite for polar perturbations . in the higher dimensional case ,
the role of the two - sphere is played by an ( @xmath161)-variety : according to their tensorial behavior on the ( @xmath161)-variety , perturbations are classified into three types , those of _ tensor , vector _ , and _
modes , according to their tensorial behavior on the section of the background space - time .
the first of these is a novel type of perturbation that exists only in the higher dimensional case , while the vector- and scalar - type modes correspond , respectively , to the axial and polar modes in the four - dimensional case . in order to investigate the asymptote of a perturbation of higher dimensional space - times
, one must study the asymptote of all the types of perturbations mentioned above .
however , besides from ishibashi and kodama s work , all analysis made to this time has been carried out for only tensor perturbations @xcite , and the situation for vector and scalar perturbations has not yet been investigated : the asymptote of a perturbation of space - times in higher dimensions is not yet established .
the treatment of scalar perturbations is the most difficult part of this analysis , because the linearized einstein equations for a scalar perturbation are much more complicated than those for a tensor perturbation and do not reduce to the form of a single second - order schrodinger - type in the higher dimensional case .
recently , this reduction was carried out by ishibashi and kodama @xcite in the background of maximally symmetric black holes with a non - vanishing cosmological constant in an arbitrary number of space - time dimensions . from the result of that work , along with
the previously obtained ordinary differential equations for tensor and vector perturbations @xcite , we have the full set of equations for the perturbations of higher dimensional maximally symmetric black holes in the form of a single self - ad - joint second - order ordinary differential equation which we refer to as the _ master equation _ for each tensorial type of perturbation .
these master equations of course , can be applied to the perturbation analysis of higher dimensional space - times .
let us mention , furthermore , that these master equations have been derived in the framework of the gauge - invariant formalism for the perturbations , and for this reason , they do not involve the problem of the choice of gauge . in the gauge invariant
approach we can isolate the _ configuration variables _ , which carry the descriptive arbitrariness of the theory , from the _ observables _ , which are gauge invariant quantities providing a coordination of the reduced phase space of general relativity , and are subjected to hyperbolic ( and therefore _ causal _ in the customary sense ) evolution equations . in physics
the _ hole argument _ is an aspect of the fact that also einstein s theory is interpreted as a gauge theory .
recall that we consider as background the cartesian product of a two - dimensional orbit space @xmath14 and a horizon with @xmath161 compact dimensions , where 1+@xmath0 is the total space - time dimension and @xmath4 4 ( in order to avoid space - times with one or less compact spatial dimensions where the presence of a massive source is inconsistent with asymptotic flatness )
. in this background one expects two linearly independent master solutions .
when the size of the radial variable is small ( the order of the size of the horizon ) one expects that the solutions closely resemble a ( 1+@xmath0)-dimensional static homogeneous space - time , while as one increases the distance one expects that at some point the perturbed solution will be flat . as mentioned above , we have already obtained the master equation for each tensorial type of perturbation in the background space - time .
our main focus in this section is therefore to prove that for each harmonic mode of the horizon there are two solutions that behave similarly at large . in the basic mode ,
this behavior is @xmath3 . in particular ,
an eigentensor of the lichnerowicz operator in a ( possibly higher dimensional ) black hole with zero eigenvalue decays at infinity in this way .
these solutions occur in an integral form .
in particular , there exist no unstable solutions to the master equation with physically acceptable initial data .
ishibashi and kodama showed that there exist no well - behaved perturbations that are regular everywhere outside the event horizon .
this is consistent with the uniqueness @xcite of a certain class of black holes in an asymptotically flat background .
although our main interest is in the standard higher dimensional black holes , that is , asymptotically flat , spherically symmetric static vacuum solutions in higher dimensions , we also investigate the asymptote of a perturbation of other maximally symmetric static homogeneous space - times with a non - vanishing cosmological constant with respect to tensor- and vector - type perturbations .
there is a reason to expect a good analytic control of static homogeneous space - time perturbations even if we do not have a complete analytic solution since we have a good approximation : the space - time is expected to resemble closely a ( 1+@xmath0)-dimensional static homogeneous black hole @xcite .
the motivations for this research are first to obtain a theoretical description of a small perturbation which is important on its own right , and second to gain understanding of perturbation physics through combination with analytical work .
the symbiosis with analytical work comes close to serve as a partial substitute of experiments ( which are sorely absent in this field ) : the analytics are essential for understanding static solutions close to perturbation where the approximation is expected to hold , and serves to formulate the aims and methods of the theory . as it turns out
, the black holes obtained numerically show only a single multi - pole mode perturbation to their horizon @xcite @xcite , and that lends some hope that the analytic approximation would retain some validity for static homogeneous space - times as well .
our objective is to describe the perturbation process ( to first order ) . writing down the equations of motion and separating the horizontal variables we find that a ric flatness condition in the horizontal directions yields a relation among the radial functions which is similar to a trace condition and allows us to simplify the fields .
after substitution one can express the fields in terms of first and second derivatives .
we are left with a second order tensor differential equation which reduces to a single second order ordinary differential equation in the radial direction , for each metric function ( and for each harmonic mode ) from which the whole metric may be recovered .
this is the master equation , which after a change of variables simplifies further
. it would be nice to have a deeper understanding why these reductions were to be expected .
the master equation belongs to the class of _ fuchsias equations _ , where fuchsias means that the equation has only regular singular points on the complex sphere which includes infinity , and means that there are exactly three such points .
as the _ hyper - geometric _ case of 3 regular singularities there is a general solution to the equation , and several methods are available .
our case is a rather special case of the hyper - geometric equation . in this case , it turns out that the solutions can be written in terms of an _ integral transformation _ , and it would be nice to understand why that had to be the case .
interestingly , we observe that in some of the relevant cases these integral transformations simplify further to _ explicit functions _ , and in particular , all relevant solutions are explicit functions at infinity ( solutions which are of even multi - pole number and are regular at the horizon ) .
we started working out this problem before we were aware of the results of kodama and ishibashi @xcite @xcite @xcite . even though their papers were published more than three years ago , we continued independently after learning about them the formalism of gauge invariant perturbation theory , and the various changes of variables which are employed there .
we were able to do so and actually found a somewhat similar master equation .
yet the final reduction of our master equation to a hyper - geometric one was motivated by those papers . in subsection [ subseccion 4.1 ]
we briefly overview the present status of the investigations of this problem and discuss what kind of new information can be obtained by the linear perturbation theory . in subsection [ subseccion 4.2 ] we discuss properties of the eigentensors of the lichnerowicz laplacian for a space - time with applications to the asymptote of a linear perturbation of higher dimensional black holes . in subsection
[ subseccion 4.3 ] we determine the linearized perturbations of the space - time . we have actually solved the perturbation equation with initial data of a static symmetrical homogeneous space - time in section [ seccion 2 ] .
we apply this to perform the analysis of the asymptote of a perturbation for tensor , vector and scalar perturbations separately . in this subsection
we write the static einstein equations in a form that will be convenient for expansion around a background space - time .
this type of expansion is known as _ gauge invariant perturbation _ and we follow here the usual conventions for this type of expansion .
we denote in this subsection the ( 1+@xmath0)-dimensional space - time metric by @xmath109 . for the gauge invariant
perturbation it is convenient to write the ric tensor in the following form @xcite @xmath110 where @xmath111 and @xmath112 are the christoffer symbols of the first and the second kind , respectively , and in addition one defines @xmath113 the first step in this procedure is to look at the linearized equations valid for weakly gravitating space - times .
the metric is taken to be @xmath114 where we denote the perturbation metric by @xmath115 .
we have ( 1+@xmath0)(2+@xmath0)/2 metric functions @xmath115 which are functions of ( @xmath27 , @xmath57 , @xmath28 ) .
the linearized field equations become @xmath116 where @xmath117 is the laplace operator in the metric g. we open this subsection with one comment . at higher orders in the perturbation procedure the form of the equations
is dominated by the linearized equations and is given by @xmath118 where ( @xmath119 ) is the order under study and @xmath120 are source terms which are quadratic , at least , in lower order metric components and their derivatives .
it is quite easy to find the higher dimensional counterparts of 4d static solutions , such as the tangherlini solution @xcite . in reality , it is also possible to find a slightly more general family of solutions for the einstein - maxwell system with cosmological constant by requiring that the space - time metric and the electromagnetic tensor @xmath121 can be expressed in the form ( [ 2 ] ) and @xmath122 the einstein equations and maxwell equations determine the two dimensional metric and the electric field @xmath123 as @xcite @xcite @xmath124 where @xmath125 this formula includes the solution with cosmological constant , which is mentioned above . in the special case in which @xmath15 is the sphere @xmath126 , the space - time becomes static and spherically symmetric . in particular , for @xmath127 this solution coincides with the tangherlini solution , and for @xmath128 it gives a higher dimensional counter - part of the reissner - nostrum solution .
more generally , when @xmath15 is a constant curvature space , the solution is invariant under so(@xmath0 ) , iso(@xmath161 ) and so(@xmath161 , 1 ) for @xmath129 respectively .
in general , this solution has a naked singularity or does not have a horizon . as shown in ref .
@xcite , it gives a regular black hole only when the parameters @xmath130 , @xmath7 and @xmath34 are in special regions . in this case
, the einstein space @xmath15 describes a spatial section of the static homogeneous space - time horizon and at the same time the spatial infinity . in the static case with @xmath7=0 , the uniqueness about asymptotically flat regular static solutions
are now established in higher dimensions as well : all regular solutions are exhausted by the tangherlini solution for the vacuum system @xcite , reissner - nostrum solution given by ( [ 2 ] ) , ( [ 15 ] ) and ( [ 16 ] ) with @xmath131=1 and @xmath7=0 in the non - degenerate case @xcite and the higher dimensional majumdar - papapetrou solutions in the degenerate case for the einstein - maxwell system @xcite .
it is also proved that the gibbons - mada solution and the tangherlini solution are the only asymptotically flat regular static symmetrical homogeneous solutions for the einstein - maxwell - dilator system and for the einstein - harmonic - scalar system , respectively @xcite @xcite . however , for @xmath132 0 , nothing is known about the uniqueness , although it was pointed out in ref .
@xcite that the approach developed by anderson may also apply to higher dimensional cases with @xmath133 0 .
in the rotating case , the higher dimensional counterpart of the kerr solution is also known .
it is given by the myers - perry solution @xcite , which has a horizon and is asymptotically flat .
in contrast to the static case , however , this is not the unique asymptotically flat rotating regular solution in higher dimensions , although the uniqueness holds for super - symmetric static homogeneous black holes in the 5-dimensional minimal super - gravity model @xcite .
this is because the horizon topology need not be given by the sphere in higher dimensions as discussed in ref .
in fact , emparan and reall @xcite found an asymptotically flat and rotating regular static symmetrical homogeneous solution in five dimensions whose black hole surface is homeomorphic to @xmath134 .
thus , at present , we have two different families of asymptotically flat rotating regular solutions with different horizon topologies . since a more complicated horizon topology is allowed in dimensions greater than 5 , it is highly probable that other families of regular solutions exist .
however , near the static limit , it is likely that a uniqueness theorem holds .
fluctuations about static homogeneous space - times relate to their stability .
in particular , if the cosmic censorship hypothesis does not hold , a regular stationary static homogeneous space - time solution will be unstable against a generic perturbation , because the formation of a naked singularity must be generic @xcite . in reality
, many of the standard static homogeneous space - time solutions are known to be perturbatively unstable @xcite .
in contrast to static homogeneous space - times , the fluctuations about rotating black holes are established only for the kerr solutions @xcite .
we do not discuss the fluctuations about these rotating static solutions in this article . in the proofs of the fluctuations about these static solutions ,
a key role was played by the fact that the basic equations for the perturbations can be reduced to a single second - order ordinary differential equation of the schrodinger type , which is often called a master equation .
now , we explain how such a master equation is derived and how the asymptote of a perturbation is calculated in terms of it . in space - times of dimension 1+@xmath0 ,
the linearizing of the einstein equations yields the following equations for the space - time perturbation @xmath115=@xmath135 : @xmath136 @xmath137 where lichnerowicz@xmath138 is the lichnerowicz operator defined by @xmath139 in order to analyze the behavior of the perturbations on the basis of these equations , the following two technical problems have to be resolved .
firstly , the perturbation variables @xmath115 contain unphysical gauge degrees of freedom that should be eliminated .
second , these perturbation equations are a coupled system of a large number of equations and are quite hard to solve directly . in the case of our interest , these problems can be resolved with the help of the tensorial decomposition and the gauge - invariant formulation @xcite @xcite @xcite .
first , the metric of a static homogeneous space - time solution to the vacuum or electron - vacuum system in 1+@xmath0 dimensions can be written in terms of the two - dimensional orbit space @xmath14 as ( [ 2 ] ) . on this background
, the basic perturbation variables @xmath115 can be grouped into three sets , * 0@xmath140 , @xmath1411 , * 0@xmath1421 and 2@xmath143 + 1 , and * 2@xmath140 , @xmath144 + 1 , according to their tensorial transformation behavior on @xmath15 . among these ,
@xmath115 for 0@xmath1421 and 2@xmath143 + 1 , can be further decomposed into the gradient part @xmath145 and the divergence - free part @xmath146 as @xmath147 where @xmath148 is the co - variant derivative with respect to the metric @xmath149 on @xmath15 .
similarly , @xmath115 for 2@xmath140 , @xmath144 + 1 , can be decomposed into the trace part and trace - free part as @xmath150 and the trace - free part @xmath151 can be further decomposed into the scalar part @xmath152 , the divergence - free vector part @xmath153 and the divergence - free and trace - free part @xmath154 as @xmath155 @xmath156 by this decomposition , the linearized einstein equations are decomposed into three decoupled sets of equations , each of which contains only variables belonging to one of the three sets of variables , the scalar - type variables ( @xmath157 , @xmath158 , @xmath159 , @xmath145 , @xmath160 , @xmath152 ) , vector - type variables ( @xmath146 , @xmath153 ) , and tensor - type variables ( @xmath154 ) .
further , since the co - variant derivative @xmath148 always appears in the combination @xmath53=-@xmath161 in the perturbation equations , the expansion coefficients of these variables in terms of harmonic tensors on @xmath15 , which are denoted as ( @xmath162 , @xmath163 , @xmath164 , @xmath165 , @xmath166 , @xmath167 ) , ( @xmath168 , @xmath169 ) and @xmath170 for the scalar - type , vector - type and tensor - type variables , respectively , are mutually coupled among only those corresponding to the same harmonic tensor ( see ref .
@xcite for details of this harmonic expansion ) .
thus , the perturbation equations can be reduced to sets of equations on the two - dimensional orbit space @xmath14 with small number of entries independent of the space - time dimension . in addition , after this harmonic expansion , we can easily construct a basis for gauge - invariant variables by taking appropriate linear combinations of these two - dimensional variables and their derivatives with respect to our coordinates of @xmath14 , and the linearized einstein equations can be written as differential equations for these gauge - invariant variables @xcite . in this subsection
we find the linear perturbations for the ( 1+@xmath0)-dimensional static symmetrical homogeneous solution and show that an eigentensor of the lichnerowicz operator is regular at infinity .
reggae and wheeler @xcite derived the linear equations that describe small perturbations of the four dimensional black hole , and here we generalize their method for the static case .
these perturbations can be interpreted as deviations of the static homogeneous black hole from symmetry due to remote masses . in the case of a sphere , we are interested in the influence of the static homogeneous black hole ( images ) on itself . as explained in the introduction the zeroth order correction in the horizon metric is the ( 1+@xmath0)-dimensional static symmetrical homogeneous black hole @xcite . more generally , as our background , we consider the ( 1+@xmath0)-dimensional metric g of the form ( [ 2 ] ) on the cartesian product of the two - dimensional orbit space @xmath14 and the ( @xmath161)-dimensional maximally symmetric space @xmath15 , with curvature scalar satisfying @xmath171=0 , @xmath1721 . the constant scalar of curvature @xmath7 of the space - time is related to the ( 1+@xmath0)-dimensional cosmological constant @xmath173 as @xmath173=2@xmath7 , and the parameter @xmath174 defines the static solution mass . for example , in the case @xmath171=1 and @xmath7=0 , this mass is given by @xmath175 where @xmath176=2@xmath177/@xmath178[@xmath0/2 ] is the area of a unit ( @xmath161)-sphere , and @xmath179 denotes the ( 1+@xmath0)-dimensional newton constant .
we assumed that @xmath180 0 to avoid the appearance of a naked singularity in the background .
we are interested in the static region called the _ black hole wedge _ in which @xmath181 so that @xmath57 is a space - like coordinate .
thus , when @xmath182 0 , the curvature scalar @xmath34 must be positive , while when @xmath133 0 , @xmath171 can be 0 or @xmath1721 .
the higher dimensional black hole corresponds to the case @xmath171=1 , @xmath183 0 and @xmath7=0 . to study perturbations in this background
, it is convenient to introduce harmonic tensors in the ( @xmath161)-dimensional symmetric space ( @xmath15 , @xmath184 ) and expand perturbations in terms of them , as we see below .
the type of the tensor defines the type of perturbation . combining the expansion coefficients
, one can construct the gauge - invariant variables in the two - dimensional orbit space metric spanned by our coordinates @xmath185=(@xmath57 , @xmath27 ) .
the linearized einstein equations then reduce to a set of equations for the gauge - invariant quantities .
it is found from the einstein equations that there is a master scalar variable for each tensorial type of perturbation .
consequently , the linearized einstein equations reduce to a set of three master equations for the master variable . thorough studies of the gauge - invariant formalism for perturbations and the derivation of the master equation are given in refs . @xcite and @xcite . to determine the eigenvalues of the lichnerowicz operator we study the solutions of @xmath186 where @xmath187 is transverse , normalizable and @xmath188 @xmath189
a variant of this problem has been treated by reggae and wheeler @xcite .
they investigated the lorentz version of this problem with @xmath190=0 .
their methods were refined by subsequent workers , vishveshwara @xcite , zerilli @xcite , press and teukolsky @xcite @xcite , stewart @xcite , and chandrasekhar @xcite .
@xmath190=0 corresponds to a small perturbation of a black hole that remains a classical solution .
these authors searched for runaway solutions of the form exp(-@xmath191)@xmath192(function of spatial variables with @xmath63 complex ) .
they demonstrated that imaginary@xmath63=0 for all solutions of ( [ 17 ] ) and concluded that black holes were classically stable objects .
we , on the other hand , are interested in solutions to ( [ 17 ] ) , where @xmath109 is a static symmetrical homogeneous solution and @xmath190 is not necessarily zero .
positive ( negative ) values of @xmath190 correspond to stable ( unstable ) gaussian fluctuations about our space - time . by applying the above arguments , we obtain three decoupled sets of equations for the scalar - type , vector - type and tensor - type perturbations .
the first two are often called the even modes ( or the polar perturbation ) and the odd modes ( or the axial perturbation ) , respectively .
this equation can be separated in @xmath57 , @xmath27 and @xmath15 coordinates , and is exhibited in equations ( [ 19 ] ) , ( [ 20 ] ) and ( [ 21 ] ) .
we then follow the approach of reggae and wheeler @xcite and divide the space of eigenfunctions .
since the background is static , by the fourier transformation with respect to the time coordinate @xmath57 , @xmath193 where * @xmath194 * @xmath195 * @xmath196 + and * @xmath197 + for 2@xmath140 , @xmath144 + 1 .
it turns out that the eigenvalues equation of the lichnerowicz operator simplify .
the linearized vacuum einstein equations can be brought to a simplified form @xcite . because our background space - time is static , with a killing parameter @xmath57 in the black hole wedge , substitution of these forms into ( [ 17 ] ) , together with the conditions of trace - free and transversality applied to @xmath115 , leads to sets of decoupled ordinary differential equations for the scalar - type , vector - type and tensor - type perturbations in the background space - time .
the corresponding set of equations can be easily casted into a single second - order ordinary differential master equation of the schrodinger type , as first shown by reggae and wheeler @xcite .
it was shown by zerilli that such reduction exists for the scalar - type perturbation of a black hole @xcite .
although these authors derived these master equations by the gauge - fixing method , the master variable can be written in terms of gauge - invariant variables=@xmath198 . ]
thus , the linear fluctuations about a black hole are determined by the behavior of the functions @xmath170 s for each mode . in the subsequent subsections , we examine tensor , vector and scalar perturbations separately , briefly reviewing the space - time perturbations and the master equation .
the parts of the riemann tensor are @xmath199 @xmath200\times\ ] ] @xmath201 @xmath202-\frac{s}{n-1}\right)\ ] ] @xmath203 @xmath204
@xmath205 we have @xmath206 @xmath207 if @xmath208 set @xmath209 we begin by considering tensor perturbations , which are given by @xmath210 where the tensor @xmath211 is defined as solution to the eigenvalue problem on the ( @xmath161)-variety , i.e. , ref .
( [ 18 ] ) with @xmath212 the eigenvalues @xmath213 are given by @xmath2(@xmath2+@xmath162)@xmath2142 + 2(@xmath161 ) for @xmath215 , ( the hyper sphere ) and form a continuous set of non - negative values for @xmath216 0 .
note that , in the four - dimensional black hole ( @xmath0=3 ) , there are no corresponding modes , as there are no tensors of this type on a 2-sphere .
it can immediately be seen that the expansion coefficient @xmath52 is itself gauge independent , and it can be taken as the master variable for a tensor perturbation .
the master equation follows from the vacuum einstein equations @xmath217\xi=0.\ ] ] next , consider vector perturbations , which are given by @xmath218 where the vector harmonics are introduced as the solutions of @xmath219 with eigenvalues @xmath220 that are given by @xmath221 with @xmath222 0 for @xmath223 and form a continuous set of non - negative values for @xmath216 0 .
a gauge - invariant variable @xmath224 can be constructed . substituting this into the einstein equations ,
we can obtain the master variable @xmath224 .
the radial part of the einstein equations can then be reduced to @xmath225 we now examine scalar - type perturbations in the background space - time .
these perturbations are given by @xmath226 where the scalar harmonics @xmath227 are defined by @xmath219 with the eigenvalues @xmath220 given by @xmath221 for @xmath171=1 and forming a continuous set of non - negative values for @xmath228 0 .
we find that the linearized einstein equations can be reduced to the form : @xmath229 we have now 2(1+@xmath0 ) radial fields @xmath170(@xmath27 ) , @xmath162(@xmath27 ) ... , @xmath230(@xmath27 ) , @xmath164(@xmath27 ) ... , + @xmath231(@xmath27 ) defined in ( [ 19 ] ) , ( [ 20 ] ) , ( [ 21 ] ) for each mode , @xmath232 recently , kodama and ishibashi have shown that for a static charged homogeneous black hole in higher dimensions represented by ( [ 2 ] ) , ( [ 15 ] ) and ( [ 16 ] ) , the perturbation equations can be reduced to decoupled second - order master equations of the schrodinger type , as in four dimensions @xcite @xcite .
these master equations in higher dimensions , however , have some new features .
first , there exists no simple relation between vector and scalar perturbations like the scalar - vector correspondence in four dimensions , for @xmath233=1+@xmath2344 .
this implies that stabilities for scalar and vector perturbations should be studied separately .
second , there exist tensor perturbations for @xmath233=1+@xmath2344 . now , let us see how these new features affect the fluctuations about static homogeneous space - times in higher dimensions . when @xmath15 is a generic einstein space , the @xmath15-coordinates appear in the perturbation equations only through the lichnerowicz operator @xmath235 hence , by expanding tensor perturbations in terms of the eigentensors @xmath236 of @xmath237 , @xmath238 we obtain a single decoupled equation for @xmath170 , which can be easily put into the form ( [ 19 ] ) .
we have solved the master equation with initial data of a static symmetrical homogeneous space - time ( [ 13 ] ) .
consider the function @xmath239 in ( [ 3 ] ) . if @xmath240 then our universal decay law ( -@xmath241 of the field behavior at infinity in space - time ( [ 13 ] ) again agrees with ref .
@xcite because of equations ( [ 9 ] ) and ( [ 10 ] ) .
the present section deals essentially with the field theory in curved space - time by studying a class of exact " ( integral transforms ) solutions to the massive klein - gordon field equation in the background of a schwarzschild - tangherlini black hole ( the higher - dimensional schwarzschild solution ) in an arbitrary number @xmath12 of space - time dimensions .
we again use the relations equations ( [ 9 ] ) and ( [ 10 ] ) in this study .
they are not exact solutions since it is assumed that the amplitude is so small that its contribution to the energy content can be neglected . some justifications for this study are presented .
we also compute and discuss the behavior of the green s function of the klein - gordon equation for a free scalar field in the background of a curved higher dimensional spherically symmetric schwarzschild black hole with coordinates coupled to the curvature of the background space - time @xcite .
the green s function is a sum on the harmonic modes of the sphere .
we consider the first term in subsection [ subseccion 5.6 ] .
this term is a double integration on the energy spectrum and the momentum of the particle .
far from the horizon , we can approximate the double integration by an integration on a line defined by the relation of energy and momentum of a free particle . from here
, we derive the yukawa potential ( [ 29 ] ) in our formalism . consider the evolution of a massive scalar field @xmath56 in the background described by ( [ 2 ] )
the evolution is governed by the curved space klein - gordon equation @xmath242\phi= 0,\ ] ] where we denote by @xmath243 the determinant of the metric .
the metric appearing in ( [ 22 ] ) should describe the geometry referring to both the black brana and the scalar field , but if we consider that the amplitude of @xmath56 is so small that its contribution to the energy content can be neglected , then the metric ( [ 2 ] ) should be a good approximation to @xmath104 in ( [ 22 ] ) .
we shall thus work in this perturbation approach .
it turns out that it is possible to simplify considerably equation ( [ 22 ] ) if we separate the horizon variables from the radial and time variables , as is done in four dimensions @xcite @xcite . for higher dimensions we follow @xcite .
consider a number @xmath170 .
the evolution of a minimally coupled scalar field @xmath56 is described by the massive klein - gordon equation ( [ 3 ] ) @xcite @xcite @xcite @xcite @xcite , with @xmath244 where ( [ 3 ] ) is a partial differential operator that contains information about the initial shape of the wave packet at @xmath57=0 .
this formula also includes the cosmological constant .
the explicit form of the operator ( [ 3 ] ) is the simplest in this general setting .
the whole section is based on an integral solution of the klein - gordon equation . in section [ seccion 2 ] ,
the linear perturbation equation is derived and solved .
let s find a relation of energy and momentum of a particle with mass in a hyper black hole .
we use ( [ 10 ] ) and calculate @xmath245 the first term is the energy of a free particle .
+ the second term is the gravitational potential energy between the masses @xmath119 and @xmath246 . to solve the inhomogeneous version of the differential equation ( [ 3 ] ) , we must construct a green s function out of the fundamental system of solutions of the homogeneous equation @xmath247 where parameters @xmath248 of the differential equation depend on @xmath249 as given by equations ( [ 10 ] ) and @xmath250 with these functions , the green s function is constructed as @xmath251 @xmath252 where the plus - minus sign of @xmath63 is taken depending on the arguments .
once the green s function has been determined , it is simple to construct the later - time evolution of the wave - packet from the initial data using equation ( [ 25 ] ) . as the eigen - value problem ( [ 3 ] )
is not self - ad - joint , the green s function ( [ 25 ] ) need not be symmetric in its arguments @xmath255 , @xmath256 , @xmath257 and @xmath258 , @xmath259 , @xmath260 .
note that the green s function is calculated from the @xmath190-modes and so @xmath261 and @xmath262 depend on @xmath190 , but we omitted the third argument in @xmath263(@xmath27 , @xmath57 , @xmath28 ) for brevity .
we claim that the expression @xmath265 @xmath266 is the green s function for the klein - gordon field in the schwarzschild space - time and utilize it in the subsequent arguments in essential ways .
actually , this expression does not have the global time translation invariance , in spite of the fact that the background space - time is static .
inverse transform integrals ( [ 26 ] ) are precisely of the above type , with phase @xmath34(@xmath63)=@xmath63 and the large parameter @xmath57 being the time coordinate .
so the stationary phase method tells us that the asymptotic @xmath269 behavior of the solution @xmath56(@xmath27 , @xmath57 , @xmath28 ) is given by singular points of @xmath270 as a function of @xmath63 , i. e. the singular point ( [ 10 ] ) .
therefore the study of analytic properties of the green s function ( [ 25 ] ) play a key role in understanding the late - time evolution of the wave - packet .
the possible sources of non - analyticity in the green s function are listed below , the problem with branches of the argument of the square root in @xmath63(@xmath62 , @xmath27 , @xmath190 ) is absent in the green s function @xmath271 when we approximate the double integration by an integration on a line defined by the relation of energy and momentum of a free particle , that is to say , by an integration on @xmath272+@xmath273 .
since the @xmath258 , @xmath259 , @xmath260 space - time is important for the late - time evolution of the wave - packet , it is instructive to take a closer look at the approximation to the green s function ( [ 25 ] ) there . we proceed to apply this into the green s function ( [ 25 ] ) , noticing that the behavior of @xmath263 is fundamentally different depending on the sign of @xmath190 .
i @xmath274 0 , the imaginary exponent @xmath275(@xmath62 , @xmath27 , @xmath190 ) in ( [ 23 ] ) and ( [ 24 ] ) is large and @xmath263 goes to zero for all @xmath27 , @xmath57 , @xmath28 . if @xmath276 0 , the real part of @xmath275(@xmath62 , @xmath27 , @xmath190 ) is negative and hence @xmath263 is finite for @xmath27 , @xmath57 , @xmath28 .
applying this into the green s function and using the asymptotic behavior of @xmath261 , @xmath262 near infinity , we obtain , @xmath277 @xmath278 the case @xmath276 0 is precisely where the inverse transformation should be in . in field theory we define @xmath279 where @xmath280(@xmath281 ) is the so - called source function .
we can choose any @xmath280(@xmath281 ) we want and by exploiting this freedom of choice , we can extract a remarkable amount of physics .
we can go on to consider some possibilities for @xmath280(@xmath281 ) ( which we will refer to generically as sources ) , for example , @xmath282 in other words , @xmath280(@xmath281 ) is a sum of sources that are time - dependent infinitely sharp spikes located at @xmath283 and @xmath284 in space - time .
( if the reader likes more mathematical rigor than is offered here , he is welcome to replace the delta function by lumpy functions peaking at @xmath285 .
the reader would simply clutter up the formulas without gaining much . )
more picturesquely , we are describing two massive lumps sitting at @xmath283 and @xmath284 on the space - time , which are moving [ time dependence in @xmath280(@xmath281 ) ] .
@xmath286(@xmath280 ) contains four terms .
we neglect the self - interaction " term @xmath287 since this contribution would be present in @xmath286 regardless of whether @xmath288 is present or not .
we want to study the interaction between the two massive lumps " represented by @xmath289 and @xmath288 .
similarly we neglect @xmath290 .
recall that in path integration formalism @xmath297=@xmath298 represents @xmath299 , where @xmath300 is the energy due to the presence of the two sources acting on each other . setting @xmath301=@xmath302
we obtain from ( [ 28 ] ) @xmath303 @xmath304 we will see that this energy is negative . the presence of two delta function sources , at @xmath283 and @xmath284 , has lowered the energy . in other words ,
the two sources attract each other by virtue of their coupling to the field @xmath56 .
we have derived a first physical result in quantum field theory !
we identify @xmath300 as the potential energy between two sources . even without doing the integrations we see that as the separation @xmath283 , @xmath284 between the two sources becomes large , the oscillatory exponential cuts off the integration .
the characteristic distance is the inverse of the characteristic value of @xmath62 , which is @xmath119 .
thus , we expect the attraction between the two sources to decrease rapidly to zero over the distance 1/@xmath119 .
far from horizon @xmath306 0 , we can approximate the double integration by an integration on a line defined by the relation of energy and momentum of a free particle , that is to say , by an integration on @xmath272+@xmath273 , @xmath307 we calculate @xmath308 @xmath309 the integral gives @xmath310 the result is as we expected : the potential drops off exponentially over the distance scale 1/@xmath119 .
obviously , @xmath311/@xmath312 0 : the two massive lumps sitting on the space - time can lower the energy by getting closer to each other .
what we have derived is one of the most celebrated results in twentieth - century physics .
yukawa proposed that the attraction between nucleons in the atomic nucleus is due to their coupling to a field like the @xmath56 field described here .
the known range of the nuclear force enabled him to predict not only the existence of the particle associated with this field , now called the @xmath313 meson or the peon , but its mass as well .
as the reader probably knows , the peon was eventually discovered with essentially the properties predicted by yukawa . in deriving ( [ 29 ] )
we direct our attention to the region far from the horizon . in this case , @xmath63 does not depend on @xmath27 , i.e. @xmath63 is reduced to the ordinary relation in the case of a flat space - time .
this means that our result ( [ 29 ] ) is correct even if gravity is taken into account .
t. reggae and j. a. wheeler , stability of a schwarzschild singularity , _ physics rev .
_ * 108 * ( 1957 ) 1063 . c. v. vishveshwara , _ physics rev .
* 1 * ( 1970 ) 2870 .
f. zerilli , _ physics rev .
d _ * 2 * ( 1970 ) 2141 .
m. davis , r. ruffing , w. h. press and r. h. price , _ physics rev .
_ * 27 * ( 1971 ) 1466 .
k. d. kokkotas and b. g. schmidt , _ living rev .
* 2 * ( 1999 ) 2 .
v. cardozo , j. p. s. limos and s. yoshi , _ physics rev .
* 69 * ( 2004 ) 044004 .
r. price , _ physics rev .
d _ * 5 * ( 1972 ) 2419 .
e. leaver , _ physics rev .
d _ * 34 * ( 1986 ) 384 .
e. s. c. ching , p. t. lung , w. m. sue and k. young , _ physics rev .
_ * 74 * ( 1995 ) 2414 .
e. s. c. ching , p. t. lung , w. m. sue and k. young , _ physics rev .
* 52 * ( 1995 ) 2118 . v. cardozo , s. yoshi , o. j. c. dias and j. p. s. limos , _ physics rev .
* 68 * ( 2003 ) 061503 .
r. c. myers and m. j. perry , black holes in higher dimensional space - times , _ ann . physics _
* 172 * ( 1986 ) 304 .
e. witten , _ adv .
math . physics _
* 2 * ( 1998 ) 505 .
s. dimopoulos and g. lands - berg , _ physics rev .
_ * 87 * ( 2001 ) 161602 .
s. b. giddiness and s. thomas , _ physics rev .
* 65 * ( 2002 ) 056010 .
w. israel , _ physics rev .
_ * 164 * ( 1967 ) 1776 .
b. carter , _ physics rev .
_ * 26 * ( 1971 ) 331 .
d. c. robinson , _ physics rev .
_ * 34 * ( 1975 ) 905 . g. w. gibbons , d. ida and t. shiromizu , _ physics rev .
_ * 89 * ( 2002 ) 041101 .
g. w. gibbons , d. ida and t. shiromizu , _ physics rev .
d _ * 66 * ( 2002 ) 044010 .
r. emparan and h. s. reall , a rotating black ring in five dimensions , _ physics rev .
_ * 88 * ( 2002 ) 101101 hep - t/0110260 .
g. w. gibbons and s. a. hartwell , _ physics rev .
* 66 * ( 2002 ) 064024 .
g. w. gibbons , s. a. hartwell and c. n. pope , _ physics rev .
* 67 * ( 2003 ) 084024 . v. cardozo , o. j. c. dias and j. p. s. limos , _ physics rev .
* 67 * ( 2003 ) 064026 .
d. ida , y. uchida and y. morissa , _ physics rev .
* 67 * ( 2003 ) 084019 .
g. horowitz and v. hubey , _ physics rev .
* 62 * ( 2000 ) 024027 .
r. konoplya , gr - q/0303052 .
c. molina , gr - c/0304053 .
s. a. hartwell , hep - t/0305001 .
f. r. tangherlini , schwarzschild field in @xmath0 dimensions and the dimensionality of space problem , _ nov cm . _
* 27 * ( 1963 ) 636 .
h. weyl , _ groups , klein s enlarging program .
quantities _ ch .
i , sec . 4 of _ the classical groups , their invariants and representations _ second ed .
( princeton university , princeton , 1946 ) , pp .
d. birmingham , _ class
. quantum gravity _ * 16 * ( 1999 ) 1197 .
a. ishibashi and h. kodama , stability of higher - dimensional schwarzschild black holes , _ prog .
theo . physics _
* 110 * ( 2003 ) 901 hep - t/0305185 .
e. witten , string theory dynamics in various dimensions , _ nuclear physics _
* b443 * ( 1995 ) 85 http://xxx.lanl.gov/abs/hep-th/9503124[hep-t/9503124 ] .
j. h. schwartz , the power of m theory , _ physics let . _ * b367 * ( 1996 ) 97 http://xxx.lanl.gov/abs/hep-th/9510086[hep-t/9510086 ] .
m. j. duff , m theory ( the theory formerly known as strings ) , _ int .
j. mod . physics _
* a11 * ( 1996 ) 5623 http://xxx.lanl.gov/abs/hep-th/9608117[hep-t/9608117 ] .
g. t. horowitz and a. storming , _ nuclear physics _ * b360 * ( 1991 ) 197 .
h. kodama and a. ishibashi , a master equation for gravitational perturbations of maximally symmetric black holes in higher dimensions , _ prog . theoretical physics _
* 110 * ( 2003 ) 701 hep - t/0305147 .
h. kodama and a. ishibashi , _ prog . theoretical physics _
* 111 * ( 2004 ) 29 .
r. a. konoplya , _ physics rev .
* 68 * ( 2003 ) 124017 .
r. a. konoplya , _ physics rev .
* 68 * ( 2003 ) 024018 .
e. berti , m. cavalier and l. gualterio , _ physics rev .
* 69 * ( 2004 ) 124011 .
r. gregory and r. laflamme , _ physics rev .
_ * 70 * ( 1993 ) 2837 .
yong chen wang , imaginary weyl coordinates and imaginary schwarzschild coordinates , _ sci .
seneca ser .
* 26 * no .
10 , ( 1983 ) 1062 - 1072 .
d. christodoulou and s. klainerman , _ the global nonlinear stability of the minkowski space _ ( princeton , princeton , 1993 ) .
l. lusaka , the rest - frame instant form of metric gravity , _ gen .
rel . gravity _
* 33 * ( 2001 ) 1579 gr - q/0101048 .
l. lusaka and s. russo , a new parametrization for tetra gravity , _ gen .
rel . gravity _
* 34 * ( 2002 ) 189 gr - q/0102074 .
r. de pietro , l. lusaka , l. martucci and s. russo , dirac s observables for the rest - frame instant form of tetra gravity in a completely fixed 3-orthogonal gauge , _ gen .
rel . gravity _
* 34 * ( 2002 ) 877 gr - q/0105084 .
l. lusaka and m. pauli , general covariance and the objectivity of space - time point - events : the physical role of gravitational and gauge degrees of freedom in general relativity , gr - q/0301040 .
s. shanmugadhasan , canonical formalism for degenerate lagrangians , _ j. math
. physics _
* 14 * ( 1973 ) 677 .
l. lusaka , the shanmugadhasan canonical transformation , function groups and the second nether theorem , _ int . j. mod
. physics _
* a8 * ( 1993 ) 4193 .
d. r. brill , p. l. chrzanowski , c. m. peria , e. d. farrell and j. r. ipser , _ physics rev .
d _ * 5 * ( 1972 ) 1913 .
s. a. teukolsky , _ physics rev .
_ * 29 * ( 1972 ) 1114 .
zia hui gan and jinan sang , standing waves of the inhomogeneous klein - gordon equations with critical exponent , _ act math .
( eng . ser . )
* 22 * no .
2 , ( 2006 ) 357 - 366 . ruin cato and jinan sang , periodic wave solutions for the coupled nonlinear klein - gordon equations , ( chinese ) _ siusan shina dixie xuebao zairian kexue ban _ * 29 * no .
2 , ( 2006 ) 158 - 160 .
abdul - maj - id wazwaz , compactions , solutions and periodic solutions for some forms of nonlinear klein - gordon equations , _ chaos solutions fractals _ * 28 * no .
4 , ( 2006 ) 1005 - 1013 .
lix - in ti - an and shuimeng you , non symmetrical compaction and multi - compaction of nonlinear intensity klein - gordon equation , _ chaos solutions fractals _ * 29 * no .
2 , ( 2006 ) 282 - 293 .
jing - bo chen , simplex and multisymplectic fourier pseudo spectral discretizations for the klein - gordon equation , _ let .
math . physics _
* 75 * no .
3 , ( 2006 ) 293 - 305 .
l. randall and r. sun - drum , _ nuclear physics b _ * 557 * ( 1999 ) 79 .
i. antonia - dis , n. armani - hamed , s. dimopoulos and g. dali , _ physics let .
b _ * 436 * ( 1998 ) 257 .
r. m. wald , _
. physics _ * 20 * ( 1979 ) 1056 .
r. m. wald , _
. physics _
* 21 * ( 1980 ) 218 .
f. zerilli , _ physics rev .
_ * 24 * ( 1970 ) 737 .
h. kodama , a. ishibashi and o. soto , _ physics rev .
* 62 * ( 2000 ) 064022 .
s. hollands and a. ishibashi , gr - q/0304054 e. soring , b. kohl and t. pi - ran , caged black holes , black holes in compacted space - times , ii .
5d numerical implementation , _ physics rev .
* 69 * ( 2004 ) 064032 hep - t/0310096 .
h. kudos and t. wise - man , properties of kaluza - klein black holes , _ prog . theoretical physics _
* 111 * ( 2004 ) 475 hep - t/0310104 .
h. kodama and a. ishibashi , stability of generalized static black holes in higher dimensions , gr - q/0312012 .
v. foch , _ the theory of space - time and gravitation _
( pergamon , oxford , 1964 ) , appendix b. s. huang , _ geometry dedicated _ * 71 * ( 1998 ) 5 . m. rogatko , _ physics rev . d _
* 67 * ( 2003 ) 084025 .
m. rogatko , _ class . quantum gravity _ * 19 * ( 2002 ) l151 .
m. t. anderson , p. t. chrusciel and e. delay , _ jeep _ * 0210 * ( 2002 ) 063 .
h. reall , _ physics rev .
* 68 * ( 2003 ) 024024 .
m. cain and g. galloway , _ class .
quantum gravity _ * 18 * ( 2001 ) 2707 .
r. m. wald , gr - q/9710068 .
d. j. gross , m. j. perry and l. g. yaffe , instability of flat space at finite temperature , _ physics rev .
( 3 ) * 25 * no 2 , ( 1982 ) 330 b. whiting ,
. physics _ * 30 * ( 1989 ) 1301 .
j. bardeen , _ physics rev .
* 22 * ( 1980 ) 1882 .
h. kodama and m. saki , cosmological perturbation theory , _ prog . theoretical physics suppl . _
* 78 * ( 1984 ) 1
. w. press and s. a. teukolsky , _ astrophysics j. _ * 185 * ( 1973 ) 649 . w. press and s. a. teukolsky , _ astrophysics j. _ * 193 * ( 1974 ) 443 . j. m. stewart , _ proc .
london a _ * 334 * ( 1975 ) 51 .
s. chandrasekhar , in _ general relativity , an einstein centennial survey _ eds
. s. w. hawking and w. israel ( cambridge university press , london , 1979 ) .
r. m. wald , _ general relativity _
( the university of chicago , chicago 1984 ) , pp .
schoenberg , mario , the green functions of the klein - gordon equation , _ revista union mat .
argentina _ * 12 * 238 - 264 , ( 1947 ) 265 - 296 . | in absence of explicit solutions of the perturbation equation of a static symmetrical homogeneous space - time , the best we can do is to construct a _ quasi-_transformation . in this framework
, we solve the perturbation equation with initial data and a number of results are derived .
far from the horizon of a black hole of even space dimension @xmath0 , a mass - less field decays as @xmath1 in space - time , where @xmath2 is a harmonic number of the sphere .
a relation of energy and momentum of a particle with mass in a hyper black hole is discovered and a solution to the equation of klein - gordon in the metric of schwarzschild - tangherlini with initial data on the hypersphere is proposed .
also , the green s function of the klein - gordon equation in schwarzschild coordinates is calculated .
this function is a sum on the harmonic modes of the sphere .
the first term is a double integration on the spectrum of energy and the momentum of the particle .
far from the horizon , the double integration is approximated by an integration on a line defined by the relation of energy and momentum of a free particle . from here , the potential of yukawa is derived .
finally , the linear perturbation equations are derived and solved exactly .
the master equation with initial data of a small perturbation in a static symmetrical homogeneous space - time , like a ( possibly higher dimensional ) schwarzschild black hole , is studied .
a main statement of the article is that for each harmonic mode of the horizon there are two solutions that behave similarly at large . in the basic mode ,
the asymptote of a field ( an eigentensor of the lichnerowicz operator , for example ) decays at infinity according to the universal law @xmath3 .
these solutions occur in an integral form .
the analysis we present is of a small perturbation to the full static symmetrical solution .
the higher order perturbations will appear in a sequel .
we determine independently perturbations of the space - time in dimension 1+@xmath4 4 , where the system of equations can be reduced to a master equation a tensor differential equation .
the solutions are integral transformations which in some cases reduce to explicit functions . |
the behaviour of a scalar field driven by a fluctuating force field depends strongly on the correlations of the driving field in space and time .
a well known example is the passive scalar problem in fluid mechanics , where one asks for the behavior of a passive field as it is advected by a turbulent fluid flow @xcite . even in situations where the driving force field has simpler correlations ,
the passive scalar field can show interesting , and sometimes unexpected , behaviour .
in particular , while in the fluid context an initial local concentration of passive particles typically spreads out in space , in other types of situations an initially randomly distributed set of particles may be driven into a state with large - scale clustering . in this paper , we study one such example .
we consider a force field which is derived from a fluctuating potential , and ask for its effect on a system of particles which do not interact with each other except through hard - core exclusion .
the problem is then tantamount to the dynamics of hard - core particles which reside on a fluctuating surface and are driven downwards by gravity along local slopes . an especially interesting case arises when the surface is self - affine with a power - law divergence in the height correlation function as a function of the separation . in such cases ,
surface roughening strongly affects the clustering of particles and can lead to new types of states@xcite .
we study a stochastically evolving two - dimensional surface governed by kardar - parisi - zhang ( kpz ) dynamics @xcite and show that surface fluctuations bring about large - scale clustering of particles , akin to phase separation .
the particles are taken to be random walkers with an excluded volume constraint , diffusing in the dynamic potential landscape defined by the local height @xmath0 at a base position @xmath1 at time @xmath2 .
in addition , the particles are assumed to be sufficiently massive that the dynamics operates effectively at zero temperature ; this means that particles only move locally downwards at a fixed rate , subject to the conditions that the local slope is favourable and the target site is unoccupied .
this model is a generalization to two dimensions of the 1-d model studied in @xcite where it was found that the particles reach a phase separated state with unusual properties arising from strong surface fluctuations .
since the nature and strength of fluctuations depends strongly on the dimensionality , it is important to see to what extent these features survive in higher dimensions .
this is one of the principal aims of this paper .
let us summarize the main results .
our numerical simulations support the argument that the particle density exhibits phase ordering over a characteristic time - dependent coarsening length scale @xmath3 set by surface fluctuations @xcite . here
@xmath4 is the dynamical exponent for surface fluctuations ; for the kpz surface with @xmath5 , it is known that @xmath6 @xcite .
the pair - correlation function is found to depend on the separation scaled by @xmath7 , as is characteristic of a system undergoing phase ordering , but the scaling function has a cusp near the origin , unlike usual phase separating systems @xcite .
the cluster size distribution in the steady state has a power law decay @xmath8 for @xmath9 , where the exponent @xmath10 is different for particles and holes .
in addition , unlike the 1-d case , there is a distinct contribution to @xmath11 from an ` infinite ' cluster which contains a finite fraction of the total number of particles .
an understanding of these results can be gained from a study of the surface itself . to this end , it proves useful to distinguish between regions where the height is less than or more than a fixed reference level .
this is incorporated in a coarse - grained depth ( cd ) model of the surface @xcite , which itself exhibits phase separation .
many of its properties reflect the underlying asymmetry of the kpz surface .
finally , in both the sliding particle and depth models , the order parameter has a distribution that remains broad in the thermodynamic limit .
we introduce the model in section ii . in section iii
, we present results for two - point correlation functions , cluster - size distributions and the order - parameter distribution , while section iv is the conclusion .
we study an autonomously evolving two - dimensional surface , whose behaviour over large length and time scales is described by the kardar - parisi - zhang equation@xcite @xmath12 with short range correlations for the noise , @xmath13 . the kpz surface in @xmath5
is known to be self - affine in steady state : @xmath14 where @xmath15 is the roughness exponent . for the kpz problem with @xmath5 , @xmath16@xcite .
the nonlinear term in eq.(1 ) breaks @xmath17 symmetry , and this has important consequences for the models we study .
we simulate the surface through a discrete solid - on - solid ( sos ) algorithm@xcite , where the height difference between nearest neighbour ( nn ) points on a square lattice is maintained at @xmath18 .
a point is selected at random and its height is increased by 2 units with probability @xmath19 if all four of its nn points are at greater height , and decreased with probability @xmath20 if all four are at a lower height . otherwise , the site is not updated .
it is believed that the asymptotic properties of this single step model are the same as those of the ( 2 + 1)-dimensional kpz equation , though this has not been proved . in our simulations we chose @xmath21 and @xmath22 , so that we have an ` evaporating ' , rather than a growing surface whose average height decreases in time .
the sliding particle ( sp ) model is defined as follows .
particles are initially distributed at random on surface sites with no more than one particle per site , the overall particle density being @xmath23 . the external force field which drives the particles acts downwards , so that particles tend to move in the same direction as the average surface height , while holes tend to move upwards , opposite to the direction of surface motion . in a microstep ,
a randomly chosen particle attempts a move to a randomly chosen neighbouring site .
if the local slope is favourable and the target site is unoccupied , the move is made , otherwise the particle stays at the same site .
one mc step is counted when , on average , all surface sites and all particles have been updated in a random sequence .
we verified that a faster rate of updating for particles with respect to surface updates does not change the qualitative results .
we did our simulations on square lattices with @xmath24 and sizes ranging from @xmath25 to @xmath26 .
we used periodic boundary conditions for both the surface and the particles which reside on it .
it is helpful to introduce two sets of discrete ising - like spin variables to characterise the particle and surface configurations . for the particle configuration ,
we define @xmath27 where @xmath28 is the local occupation index , i.e. , @xmath29 if there is a particle at @xmath30 and zero otherwise .
further , to characterise the height fluctuations of the surface configuration , we define a coarse - grained depth model in which we categorize sites according to whether they are above or below a certain fixed height . to this end , we define @xmath31 where @xmath32 is a chosen reference level .
we call the set @xmath33 as sp spins and the set @xmath34 as cd spins . in order to have a rough correspondence with the half - filled case @xmath24 of the sp model ,
we take @xmath32 to be a spatial average over the configuration of surface heights at that instant .
we expect a correlation between sp and cd spins as local slopes in the surface guide the particles towards the local minima , so that over sufficiently large time scales , particles are expected to preferentially cluster in low - depth regions with predominantly positive cd spins . if particles were to occupy the lowest available positions so as to minimise the total potential energy , we would have a close match of the @xmath35 and @xmath36 .
however , this ` ground state ' is never reached as the surface reconfigures itself before particle rearrangements can occur .
the sp - cd correspondence gives an insight into the extent of clustering of the particles .
for example , a hill ( a region with negative @xmath37 ) of linear extension @xmath38 is expected to overturn in typical time @xmath39 , causing clustering of particles as they fall into the valleys that form . since the largest hill that overturns within time
@xmath2 has typical size @xmath3 , this defines the characteristic length scale for particle clustering @xcite .
this is verified in a numerical study of the pair spin correlation function @xmath40 in the sp model . as we see in fig.1 ,
this function can be fitted into a scaling form @xmath41 with @xmath3 and @xmath42 , which is close to the value of the dynamical exponent for the ( 2 + 1 ) dimensional kpz surface @xcite .
this dynamic scaling form is characteristic of a phase ordering system .
however , the scaling function has a cusp singularity at small argument , unlike usual coarsening systems where the function is linear for small arguments @xcite .
= 1.6 in in steady state , the system size @xmath43 replaces @xmath44 as the relevant length scale , and a similar scaling form is expected for the correlation function . to quantify the deviation from linearity at small arguments
, we studied the structure factor in the steady state , @xmath45 as shown in fig .
2 , the direction - averaged structure factor is well described by the form @xmath46 , where @xmath47 and @xmath48 at large @xmath49 .
the first term arises from a short - distance analytic contribution @xmath50 which adds on to the scaling part @xmath51 of @xmath52 .
we find @xmath53 , a pronounced difference from typical coarsening systems where @xmath54 ( porod law ) @xcite .
the non - porod form implies a cusp in the real space scaling function at small argument , @xmath55 with @xmath56 .
we have also studied the two - point correlation function in the cd model , and find a similar behaviour , with a cusp exponent @xmath57 in the next subsection we will show that within the independent interval approximation , the cusp exponent @xmath58 of the cd model is equal to the roughness exponent @xmath15 .
= 1.6 in let us define a cluster as a set of like sp spins ( particles or holes ) , each of which is a nearest neighbor of at least one other like spin in that cluster .
a study of the size distribution of clusters shows ( fig .
3 ) that the number of connected clusters @xmath59 with @xmath60 particles ( holes ) has a power - law decay for @xmath61 : @xmath62 where @xmath63 and @xmath64 .
this power - law distribution of cluster sizes , reminiscent of critical systems , is another characteristic feature of the unusual phase separated state under study it occurs in one dimension as well @xcite .
there are , however , two important differences from the 1-d case .
first , there is a marked difference in the powers @xmath65 and @xmath66 for particle @xmath67 and hole @xmath68 clusters , which reflects the asymmetry of the @xmath69 and @xmath70 clusters in the cd model as discussed below .
second , for the particle cluster distribution we find that there is an additional large @xmath60 contribution of the form @xmath71 peaked at @xmath72 . for finite system sizes , @xmath73 is somewhat broad , but with increasing @xmath43 , the width of the peak narrows down .
we find that the area under the peak is unity , implying that there is a single very large cluster in every configuration .
also , the fraction of particles contained in this cluster is @xmath74 .
the power - law distribution of cluster sizes in the sp model has its counterpart in the cd model too .
it is known that when a rough surface is intersected by a horizontal plane , the distribution of areas enclosed by the closed contours of intersection has the power - law form @xmath75 for @xmath9 , where @xmath76 @xcite .
a typical contour encloses several other contours , and the overall structure is scale invariant@xcite . for the kpz surface one should moreover distinguish between contours whose inner perimeter sites lie above or below the cut , as there is no _ a priori _ reason for both the distributions to be identical . in fact , numerical simulations show that cluster size distributions for cd spins ( which are closely related to area distributions ) , indeed has an asymmetry between positive and negative spins ( fig .
4 ) . both the distributions follow power - law decays , but with different exponents .
the exponent for negative cd spins is numerically close to @xmath77 , while that for positive spins is significantly larger .
morover , there is a contribution from an infinite cluster of positive spins , as for the particle distribution in the sp problem .
= 1.8 in within the cd model , we argue that power laws in cluster distributions of a different sort are related to the occurrence of a cusp in the scaled correlation function . for a self - affine surface with roughness exponent @xmath15
, it is known that the probability @xmath78 that the surface first returns to its starting height @xmath79 after moving a distance @xmath80 along an arbitrary linear direction has a power - law decay at small @xmath80 : @xmath81 for @xmath82@xcite .
each such segment defines a linear cluster of cd spins @xmath83 of the same sign , along the linear cut .
now let us make the independent interval approximation ( iia ) in which the lengths of such segments are taken to be independent random variables .
this enables the laplace transforms of the cluster size distribution and pair correlation function for cd spins ( defined relative to @xmath32 ) to be related along any linear cut .
the nonstandard feature is that the mean cluster size diverges as @xmath84 , but the correlation function can be calculated as in @xcite .
it has the scaling form @xmath85 comparing with eq.[eq : eq3 ] , we see that the iia predicts that the cusp exponent @xmath58 in the cd model is equal to the roughness exponent @xmath15 .
the kpz value @xmath15 is quite close to the measured value of @xmath58 for the cd model , and also to that for the sp model .
the behaviours of both models on large scales of distance and time appear to be similar in 2-d , even though the microscopic configurations of the two match only roughly .
to get an idea of the latter , we monitored the overlap index @xmath86 , and found that @xmath87 varies between @xmath88 when the frequency of updates of particles and surface is equal , and a saturation value @xmath89 as the ratio between the two is increased . however , there was no corresponding significant change in the numerical values of the exponents @xmath58 and @xmath10 .
= 1.8 in for a system with conserved magnetisation like the sp model , an appropriate quantity to characterise the ordered state is the steady state average of the magnitude of the fourier components of the density @xcite , defined as @xmath90 where @xmath91 and @xmath92 . taking the magnitude
guarantees that @xmath93 receives the same contribution from all configurations that can be obtained from each other by translational shifts @xcite . in fig . 5
, we plot @xmath94 along the ( 1,0 ) direction for four different lattice sizes .
the sequence of curves suggests that for any fixed , finite @xmath95 , @xmath96 approaches zero as @xmath43 increases .
however , if @xmath97 is held fixed so that @xmath98 approaches zero as @xmath99 , the corresponding @xmath100 approaches a finite limit . of this set of @xmath100 s ,
the largest is @xmath101 , and provides the simplest characterization of the order .
= 1.8 in = 1.8 in fluctuations in the sp model are unusually strong and are reflected in the probability distribution of @xmath102 ( fig .
the distribution seems to remain broad and approach a limit as @xmath43 increases , with @xmath103 and variance @xmath104 .
similar studies of the corresponding cd model shows that there is a broad distribution there as well , and we find @xmath105 .
the fact that the rms fluctuation of the order parameter does not vanish in the large - size limit is an unusual and characteristic feature of the ordered state of this model .
note that fluctuations which drive @xmath102 towards zero need not take the system into a disordered state .
a study of the temporal fluctuations of @xmath102 in one dimension showed that a low value of @xmath102 occurs simultaneosly with an increase of the value of @xmath100 for another close - by @xmath97 , so that the state retains macroscopic order , but with a few more coexisting macroscopic domains @xcite .
a more complete characterization of the order in two dimensions would thus involve finding the joint probability @xmath106 $ ] as @xmath84 , but this has not been attempted here .
it is interesting to compare the distribution @xmath107 for this model with the corresponding quantity for a more familiar system such as the 2-d ising model evolving under conserved ( kawasaki ) dynamics . for the ferromagnetic ising model with equal numbers of up and down spins ,
the ordered state in a finite system consists of strips of width @xmath108 parallel to either @xmath109 or @xmath110 axes , with equal probability .
thus the distribution @xmath107 ( say , with @xmath101 ) in that case would have two peaks , one at a non - zero value of @xmath102 corresponding to strips forming along the @xmath109-direction , and another peak at @xmath111 which corresponds to strips forming in the @xmath110-direction . in our case , however , the peak at the origin is absent as the interface between phases is much more diffuse and there is no strip formation . in this sense ,
the ordering observed in this model is very different from that in traditional equilibrium models .
to conclude , we have studied a model where a fluctuating self - affine surface drives a set of downward - drifting particles to a phase separated state .
the phase separation has several unusual characteristics arising from the presence of strong fluctuations which survive even in the thermodynamic limit .
these include a power - law distribution of cluster sizes , a cusp in the scaled pair correlation function and a finite width of the order parameter distribution .
these features mirror the properties of a coarse - grained height model of surface fluctuations , where they follow directly from the self - affine nature of surface fluctuations .
in particular , the asymmetry in the cluster size distribution for particles and holes is a consequence of the fact that kpz growth breaks up - down symmetry
. it would be interesting to see if a similar asymmetry exists in other related quantities such as the distribution of lengths of contours and the areas enclosed by them
. it would also be interesting to study the sliding particle problem on a fluctuating edwards - wilkinson surface , where particle - hole symmetry should be restored in the half - filled case .
this surface is logarithmically rough in 2-d , suggesting that the scaled two - point correlation functions for the sliding particle and depth problems should show an even sharper cusp than for the kpz surface studied in this paper .
finally , we remark that it would be interesting to explore the effects of removing the hard core interaction between particles .
a study of noninteracting particles sliding on a kpz surface in one dimension indicates a power law decay of the two - point correlation function @xcite .
a study of other characteristics of the resulting state should prove quite revealing .
we would like to thank d. das , d. dhar , s.n .
majumdar , b. schmittmann and r.k.p zia for helpful discussions .
g. m. acknowledges the hospitality of the tata institute of fundamental research where this work was initiated . this research is supported in part by a grant ( dmr-0088451 ) from the us national science foundation .
it is a pleasure and a privilege to contribute this article to the festschrift volume for the 70th birthday of prof .
michael e. fisher , whose work on various aspects of co - operative ordering has inspired several generations of statistical physicists .
mb would like to acknowledge many stimulating interactions over the years with professor fisher , both professional and personal .
r. h. kraichnan , phys .
rev . lett.*72 * , 1016 ( 1994 ) ; b. l. shraiman and e. d. siggia , nature * 405 * , 639 ( 2000 ) ; g. falkovich , k. gawedzki and m. vergassola , rev .
phys . * 73 * , 913 ( 2001 ) .
b. drossel and m. kardar , phys .
* 85 * , 614 ( 2000 ) .
d. das and m. barma , phys .
* 85 * , 1602 ( 2000 ) ; d. das , m. barma and s. n. majumdar , phys .
e * 64 * , 046126 ( 2001 ) .
m. kardar , g. parisi and y. -c .
zhang , phys .
* 56 * , 889 ( 1986 ) .
j. m. kim , m. a. moore and a. j. bray , phys .
a * 44 * , 2345(1991 ) .
a. j. bray , adv
. phys . * 43 * , 357 ( 1994 ) .
j. m. kim , a. j. bray and m. a. moore , phys .
e * 45 * , 8546 ( 1992 ) .
tang , b. m. forrest and d. e. wolf , phys .
a * 45 * , 7162 ( 1992 ) ; m. lassig , phys .
lett . * 80 * , 2366 ( 1998 ) .
p. meakin , _ fractals , scaling and growth far from equilibrium _ , cambridge university press .
g. porod , in _ small angle x - ray scattering _ , edited by o. glatter and l. kratky ( academic press , new york , 1983 ) .
m. matsushita , s. ouchi and k. honda , j. phys .
japan * 60 * , 2109 ( 1991 ) .
z. olami and r. zeitak , phys .
lett . * 76 * , 247 ( 1996 ) .
j. kondev , c. l. henley and d. g. salinas , phys .
e * 61 * , 104(2000 ) .
j. schmittbuhl , j -p .
vilotte and s. roux , phys .
e * 51 * , 131 ( 1995 ) .
g. korniss , b. schmittmann and r.k.p .
zia , europhys .
* 45 * , 431 ( 1999 ) . | we study phase separation in a system of hard - core particles driven by a fluctuating two - dimensional self - affine potential landscape which evolves through kardar - parisi - zhang ( kpz ) dynamics .
we find that particles tend to cluster together on a length scale which grows in time .
the final phase - separated steady state is characterized by an unusual cusp singularity in the scaled correlation function and a broad distribution for the order parameter . unlike the one - dimensional case studied earlier ,
the cluster - size distribution is asymmetric between particles and holes , reflecting the broken reflection symmetry of the kpz dynamics , and has a contribution from an infinite cluster in addition to a power law part .
a study of the surface in terms of coarse - grained depth variables helps understand many of these features . 2 |
in many theoretical treatments of condensed matter systems , the hamiltonian consists of a sum of kinetic energy and coulomb potential energy . in reality , the transverse electromagnetic ( radiation ) fields are also important .
if the transverse electromagnetic field is included in the description of condensed matter , then phenomena such as the casimir effect@xcite , the lamb shift@xcite and the long ranged inter - molecular interaction forces@xcite can be explained employing standard low order quantum electrodynamic perturbation theory .
if the coupling of the transverse electromagnetic field with matter is strong enough to produce an instability in the photon oscillators , then a transition occurs from a normal radiative phase to a super - radiant phase .
the dicke model@xcite has been studied as a prime example of this phase transition . in the dicke model , many two level molecules interact with a single photon mode of the radiation field .
the dicke model has been extensively studied by many workers@xcite , each adopting a somewhat different method to explain the physical properties of the system .
hepp and lieb@xcite , analyzed the thermodynamic properties of the dicke maser model and elucidated the mathematical structure of the super - radiant phase transition . in subsequent works of wang and hoe@xcite and hepp and lieb@xcite ,
the coherent states of glauber @xcite were employed for multi - photon oscillators in the dicke model and the resulting phase transition has been investigated in great detail in these papers . the existence of such phase transitions in condensed matter systems has been investigated in a series of papers by preparata and coworkers@xcite .
preparata has proposed several examples of super - radiant phases in a recent review@xcite .
water@xcite is a particularly interesting example .
the purpose of this work is to present a very direct method for computing the thermodynamic region wherein super - radiant phase exists .
when a condensed matter system interacts with the electromagnetic field , the field degrees of freedom may be described by a renormalized thermodynamic potential . for small oscillator displacements ( in the normal radiation phase )
the effective potential is parabolic .
as the thermodynamic parameters change , the effective oscillator frequency may turn imaginary .
the resulting oscillator instability is the signature of the phase transition into a super - radiant state . in sec .
ii , the statistical thermodynamic method for calculating the effective potential of a photon oscillator mode will be discussed .
the conditions for the stability of the oscillator will be exhibited . in sec .
iii , the dynamics in real time of the oscillator will be discussed in terms of the photon mode propagator .
the conditions for dynamic real time stability are shown to be the same as for thermodynamic stability . in sec .
iv , the example of a mesoscopic object placed in an optical cavity is considered .
if the mesoscopic object has a sufficiently high polarizability , then a cavity mode can become unstable . in the sec .
v , the nature of the non - linear forces which stabilize the oscillator mode in the super - radiant phase are explored .
the shift in the equilibrium position required to stabilize the oscillator in the low temperature super - radiant phase is exhibited . in the concluding sec .
vi , the notion of super - radiance with electric dipole coupling is related to the notion of coherent ferro - electricity . a material grain with a high electric polarizability , but otherwise without a net dipole moment , can be _ induced _ into a ferro - electric phase by an interaction with transverse electric fields . the transverse fields may induce a coherence in the electronic states then yielding a net electric dipole moment .
let us consider a simple harmonic oscillator with momentum @xmath0 , displacement @xmath1 and frequency @xmath2 .
the oscillator is coupled to a thermal bath described by the hamiltonian @xmath3 .
the free energy @xcite of the thermal bath @xmath4 is then determined by @xmath5 the coupling between the thermal bath and the oscillator is described by a force @xmath6 .
the total hamiltonian is then @xmath7 which determines the total free energy @xmath8 via @xmath9 if @xmath10 and @xmath11 are classical to a sufficient degree of accuracy , then the the _ total trace _ may be approximated by @xmath12 from eqs.(1 - 4 ) one obtains @xmath13
@xmath14 i.e. @xmath15 where @xmath16 @xmath17 finally , if the force @xmath6 is _ also classical _ to a sufficient degree of accuracy , then @xmath18 yielding @xmath19 the `` effective '' thermally induced potential of the oscillator is given by @xmath20 where ( without loss of generality ) one may assume ( for @xmath21 ) that @xmath22 . for small values of the oscillator displacement , eq.(10 ) implies @xmath23 where , for a _ classical force _
@xmath6 , the shifted frequency @xmath24 of the oscillator is given by @xmath25 the frequency shift ( from @xmath26 downward to @xmath27 ) of the oscillator due to the interaction of the oscillator with the thermal bath is given by @xmath28 the oscillator is `` stable '' if the shifted frequency is real ; i.e. @xmath29 or equivalently @xmath30 the oscillator is `` unstable '' if the shifted frequency is imaginary ; i.e. @xmath31 or equivalently @xmath32 one may appreciate the dynamical significance of the shifted frequency by considering the spectral density @xmath33 of the force fluctuations @xcite defined by @xmath34 for equal times @xmath35 the classical fluctuation response theorem asserts that @xmath36 where @xmath37 for @xmath38 is the dynamic damping coefficient for the oscillator .
@xmath37 will be discussed in detail in sec .
iii below .
here we note that eqs.(13 ) , ( 17 ) and ( 18 ) imply that @xmath39 if the integrated ( over frequency ) dissipative damping on the right hand side of eq.(19 ) is sufficiently large , the the oscillator becomes unstable .
the oscillator will be stable for @xmath40 eq.(20 ) has been here deduced by statistical thermodynamic reasoning .
however , it is perhaps more simple to deduce eq.(20 ) directly from the simple dynamical equations of motion for the oscillator .
let us now consider a simple harmonic oscillator with a canonical momentum @xmath0 , displacement @xmath1 and frequency @xmath41 .
the oscillator is driven by a small time varying external force @xmath42 .
the total hamiltonian , including the external force is then @xmath43 where @xmath44 is defined in eq.(2 ) . in a phenomenological linear response theory ,
the small mean deviation from thermal equilibrium in the oscillator coordinate obeys the equation of motion @xmath45 where @xmath46 represents the induced change in the frictional damping force exerted on the coordinate by the thermal bath . for simple engineering purposes
, one often considers that the frictional damping force has a simple local ( in time ) form @xmath47 .
more generally , the damping force will be non - local in time ; i.e. @xmath48 for complex frequency @xmath49 , with @xmath50 , if @xmath51 then @xmath52 defines the dynamical oscillator propagator @xmath53 . from eqs.(23 ) and ( 25 ) @xmath54 where @xmath55 from eqs.(22 ) , ( 25 ) and ( 26 ) , one finds that @xmath56 i.e. @xmath57 is the `` self energy '' of the harmonic oscillator propagator .
from the retarded ( causal ) nature @xcite of eqs.(23 ) and ( 27 ) follows the dispersion relation @xmath58 equivalently , in terms of the frequency dependent damping coefficient @xmath59 , @xmath60 eq.(29 ) reads @xmath61 as @xmath62 , eq.(25 ) reads as the thermal ( static ) response function @xmath63 where @xmath64 equivalently , employing eqs.(31 ) and ( 33 ) @xmath65 as previously derived in eq.(19 ) from purely thermodynamic reasoning .
if the dynamical oscillator propagator is presumed to be retarded , it too must satisfy a dispersion relation @xmath66 if the oscillator is stable , then causality is obeyed with the dissipative condition @xmath67 .
the response function @xmath68 is then analytic in the upper half @xmath69 of the complex frequency plane . on the other hand
if the oscillator is unstable , then at least one pole at @xmath70 appears in the upper - half plane and the dispersion relation eq.(35 ) for the dynamical oscillator propagator is no longer valid .
the occurrence of an instability in the system can be viewed as inducing the non - analyticity of the dynamical oscillator propagator given in eq.(28 ) . to locate an unstable oscillator pole in the upper half plane at @xmath71 ,
we look for a zero in the denominator of eq.(28 ) ; i.e. @xmath72 from eqs.(31 ) and ( 36 ) we seek a solution for @xmath73 of the equation @xmath74 as a function of imaginary frequency in the range @xmath75 , it is evident that the right hand side of eq.(36 ) is smoothly and monotonically decreasing from @xmath76 to zero , while the left hand side of eq.(36 ) is smoothly and monotonically increasing from @xmath77 to infinity . if @xmath78 , then the two smooth curves intersect at precisely one value of @xmath79 .
it is easily proved that @xmath80 is an isolated zero in the denominator of eq.(28 ) .
thus @xmath81 has an isolated pole in the upper - half plane .
the condition for this pole to exist , i.e. @xmath82 is the same as the existence of the oscillator instability .
consider an electromagnetic mode in an empty optical cavity @xcite .
the electric field of the mode @xmath83 may be related to the oscillator displacement @xmath10 of the mode by equating the mean energy stored in the electric field to the potential energy of the oscillator ; i.e. @xmath84 in order to implement eq.(39 ) , one may choose @xmath85 where @xmath86 is normalized according to @xmath87 let us further suppose that cavity has been designed so that the mode electric field is localized in a spatial region of length scale @xmath88 about the origin .
then @xmath89 where @xmath90 is a unit vector ; @xmath91 .
let us finally suppose that a mesoscopic object is placed within the mode neighborhood .
the mesoscopic object interacts with the electric field via the electric dipole moment @xmath92 .
the interaction hamiltonian between the mesoscopic object and the optical mode oscillator then takes the form @xmath93 so that the force on the oscillator coordinate is due to the electric dipole moment .
the force is @xmath94 eq.(44 ) follows from eqs.(40 ) , ( 42 ) and ( 43 ) . if the oscillator were driven by an external force at frequency @xmath95 , then from eqs.(25 ) and (
26 ) the damping force and displacement at frequency @xmath96 obey @xmath97 the electric dipole moment response to the electric field defines the polarizability @xmath98 of the mesoscopic object via @xmath99 equivalently , @xmath100 where eqs.(40 ) , ( 42 ) and ( 46 ) have been employed . from eqs.(44 ) , ( 45 ) and ( 47 ) , one can relate the self energy part of the mode propagator to the polarizability of the mesoscopic object ; the relationship is @xmath101 the condition for the oscillator stability @xmath102 may be written in terms of the static polarizability @xmath103 for a very highly polarizable mesoscopic object , say a single domain ferroelectric grain @xcite , the photon oscillator can exhibit an instability .
the general nature of the non - linear forces which eventually stabilize the oscillator must come into play .
the problem of how an oscillator reaches a new thermodynamic equilibrium position in the `` unstable '' regime can be discussed only in the context of a specific microscopic environmental hamiltonian @xmath104 which enters into eq.(7 ) .
an often discussed model for a photon oscillator employs @xmath105 identical `` two level atoms''@xcite in the thermal environment .
the hamiltonian is @xmath106 where @xmath107 is the energy difference between the single atom ground state and the excited state , @xmath108 are the pauli matrices for the @xmath109 two level atom and @xmath110 is formally the `` total spin '' of the mesoscopic object . the object is placed in the region of the cavity where the electric fields of the photon oscillator are concentrated . the electric field is presumed to interact with the atomic system via the total electric dipole moment component in the @xmath111 direction ; i.e. @xmath112 where @xmath113 is the dipole moment matrix element . using eqs.(44 ) and ( 52 ) the force exerted on the photon oscillators by the dipole moment of the atomic system is @xmath114 one is now in a position to compute the free energy @xmath115 for _ all _ domains ( stable or unstable ) of the ( now non - linear ) oscillator .
substituting eqs ( 51 ) and ( 53 ) into eq.(7 ) and taking the trace over the formal spin degrees of freedom yields @xmath116}-\ ] ] @xmath117.\ ] ] using eqs ( 11 ) and ( 54 ) , the shift in the frequency of the photon oscillators from @xmath118 down to the temperature dependent @xmath119 is found to be @xmath120 the stability of the oscillator requires @xmath121 , which in virtue of eq.(55 ) reads @xmath122 where the critical temperature for stability obeys @xmath123 the shift in the equilibrium position required to stabilize the oscillator when @xmath124 is found from the free energy minimum condition @xmath125 .
the resulting equilibrium position @xmath126 obeys the transcendental equation @xmath127 @xmath128 in fig . 1 , with @xmath129 , we show the following : ( i ) above the critical temperature @xmath130 the renormalized frequency of the stable oscillator @xmath131
is lowered as the temperature is lowered .
( ii ) below the critical temperature @xmath132 , the oscillator position shift @xmath133 ( required to restore stability ) increases as the temperature is lowered .
thus as the temperature is lowered from above the critical temperature , the oscillator frequency decreases until the `` soft mode '' has vanishing frequency at the critical temperature .
the transition to an ordered phase begins as the temperature is lowered below the critical temperature .
the oscillator seeks a new equilibrium position @xmath134 displaced from the old equilibrium position of @xmath135 .
the further the temperature is lowered , the more growth is exhibited in @xmath126 . in fig .
2 the free energy is plotted as a function of the displacement of the oscillator for two values of the temperature . for @xmath136 ,
the potential has a single minimum at @xmath135 . for @xmath137 ,
the potential has a double minimum at @xmath138 .
the oscillator will choose one of these minima in the super - radiant phase . if a highly polarizable mesoscopic grain is coupled strongly to the transverse electromagnetic field via the electric dipole moment , then there is a transition from a normal phase ( @xmath139 ) to a super radiant phase ( @xmath140 ) .
it has been shown that any simple harmonic oscillator subject to a random force ( producing damping ) can be made unstable if the induced damping of the oscillator is sufficiently strong .
if @xmath141 is the frequency of the unperturbed oscillator and if @xmath57 in eq.(31 ) is the self energy induced by the damping coefficient @xmath142 , then the precise condition for the instability is that @xmath143 it is possible to observe this instability for a mesoscopic grain which is highly polarizable . the grain must be placed in an optical cavity with a mode of spatial extent @xmath144 large on the scale of the grain size .
the condition for an optical cavity mode to go unstable is that @xmath145 where @xmath146 is the dynamic polarizability of the grain . the inequality in eq.(59 ) can surely be reached with a single domain highly polarizable ferroelectric grain .
the manner in which the oscillator is eventually stabilized ( when eq.(59 ) holds true ) is not a trivial problem .
the example of a grain which consists of `` two - level atoms '' was worked out in detail .
the normal and super - radiant free energies were exhibited in fig .
2 . when the frequency of the oscillator mode softens towards zero as @xmath147 , the equilibrium position of the oscillator shifts to a nonzero value @xmath134 . for two level atoms ,
the details of the super - radiant transition can be computed .
casimir and d. polder , _ phys .
rev . _ * 106 * , 1117 ( 1957 ) .
w.e . lamb and r.c .
retherford , _ phys . rev . _
* 72 * , 241 ( 1947 ) .
bethe , _ phys .
rev . _ * 72 * , 339 ( 1947 ) .
n.m . kroll and w.e .
lamb , _ phys .
rev . _ * 75 * , 388 ( 1949 ) .
feynman , _ the quantum theory of fields : proceedings of the twelfth solvay conference on physics _
, page 61 , interscience , new york ( 1961 ) .
i.e. dzyaloshinski and l.p .
pitayevski , _ sov .
. jetp _ * 9 * , 1282 ( 1959 ) .
i.e. dzyaloshinski , e.m . lifshitz and l.p .
pitayevski , _ sov .
. jetp _ * 10 * , 161 ( 1960 ) .
i.e. dzyaloshinski , e.m . lifshitz and l.p .
pitayevski , _ advances in physics _ * 10 * , 165 ( 1961 ) .
dicke , _ phys .
rev . _ * 93 * , 99 ( 1954 ) .
k. hepp and e.h .
lieb , _ ann .
phys . _ * 76 * , 360 ( 1973 ) .
wang and f.t .
hoe , _ phys .
_ * a7 * , 831 ( 1973 ) .
k. hepp and e.h .
lieb , _ phys .
_ * a8 * , 2517 ( 1973 ) .
v.i . emelanov and yu.l .
klimontovich , _ phys .
lett . _ * a59 * , ( 1976 ) .
r. glauber , _ phys .
rev . _ * 131 * , 2766 ( 1963 ) .
e. del giudice , g. preparata and g. vitiello , _ phys .
_ * 61 * , 1085 ( 1988 ) .
g. preparata , _ phys .
rev . _ * a38 * , 233 ( 1988 ) .
g. preparata , `` quantum field theory of super radiance '' in _ problems of fundamental modern physics _ ,
r. cherubini , p. dalpiaz and b. minetti , world scientific , singapore ( 1990 ) .
g. preparata , _
`` qed coherence in matter '' _ , world scientific , singapore ( 1995 ) .
e. del giudice and g. preparata , _ `` a new qed picture of water '' _ in _ `` macroscopic quantum coherence '' _ edited by e. sassaroli , y. srivastava , j. swain and a. widom , world scientific , singapore ( 1998 ) .
e. del guidice , g. preparata and m. fleischmann , _
_ * 482 * , 110 ( 2000 ) .
landau and e.m .
lifshitz , _ statistical physics _ , pregamon press , london ( 1958 ) .
r. kubo , _ j. phys
( japan ) _ * 12 * , 570 ( 1957 ) . p.c .
martin and j. schwinger , _ phys .
rev . _ * 115 * , 1342 ( 1959 ) .
l.p . kadanoff and p.c .
martin , _ ann .
* 24 * , 419 ( 1963 ) .
martin , _ measurements and correlation functions _ , gordan and breach science publishers , new york ( 1968 ) .
a.a . abrikosov , l.p .
gorkov and i.e. dzyaloshinski , _ methods of quantum field theory in statistical physics _ , dover publications , new york ( 1975 ) .
p. goy , j.m .
raimond , m. gross and s. haroche , _ phys .
lett . _ * 50 * , 1903 ( 1983 ) .
heinzen , j.j .
childs , j.e .
thomas , and m.s .
feld , _ phys .
lett . _ * 58 * , 1320 ( 1987 ) .
d.j . heinzen and m.s .
feld , _ phys .
lett . _ * 59 * , 2623 ( 1987 ) .
a. anderson , s. haroche , e.a .
hinds , w. jhe and d. meschede , _ phys .
* a9 * , 3594 ( 1988 ) .
thompson , g. rempe and h.j .
kimble , _ phys .
lett . _ * 68 * , 1132 ( 1992 ) .
v. sandoghdar , c.i .
sukenik , e.a .
hinds and s. haroche , _ phys .
lett . _ * 68 * , 3432 ( 1992 ) .
wang , z.k .
qin and d.l .
lin , _ phys .
* b40 * , 680 ( 1989 ) .
a. gordon and s. dorfman , _ phys .
rev . _ * b50 * , 13132 ( 1994 ) . v. weisskopf and e. wigner , _ z. phys . _ * 63 * , 54 ( 1930 ) .
arecchi and e. courtens , _ phys .
rev . _ * a2 * , 1730 ( 1970 ) .
eberly , _ am .
_ * 40 * , 1374 ( 1972 )
. s. stenholm , _ phys .
reports _ * 6 * , 1 ( 1973 ) | if the damping of a simple harmonic oscillator from a thermally random force is sufficiently strong , then the oscillator may become unstable . for a photon oscillator ( radiatively damped by electric dipole moments )
, the instability leads to a low temperature hepp - lieb - preparata super - radiant phase transition .
the stable oscillator regime is described by the free energy of the conventional casimir effect .
the unstable ( strongly damped ) oscillator has a free energy corresponding to dicke super - radiance . |
cosmic rays ( crs ) travel to us through the chaotic magnetic field of the galaxy . therefore , the only way to connect them to their accelerator is through the on - site radiation .
accelerated electrons have already been detected in supernova remnant ( snr ) shocks @xcite .
however , electrons comprise only 1 - 2% of the cr intensity above 4 - 5 gev .
therefore , distinguishing between leptonic and hadronic origin of the observed tev emission from snrs is a key to the proof of the supernova origin of crs .
simultaneous monitoring of x- and @xmath1 energy bands is an indispensable tool for that purpose .
the same tev _
electrons _ radiate in x- and @xmath0-rays via synchrotron and inverse compton ( ic ) mechanisms , respectively ( _ e.g. , _ * ? ? ?
* ; * ? ? ?
accelerated _ protons _ , can be detected _ only _ near their own ( most interestingly tev ) energy band through their interaction with an ambient gas .
therefore , the `` smoking gun '' for _ proton acceleration _ should be the @xmath0-ray emission , without an x - ray emission that can be identified as the electron synchrotron radiation .
molecular clouds ( mc ) adjacent to the snr will dramatically enhance proton visibility ( _ e.g. , _ * ? ? ?
* ; * ? ? ?
however , to conclusively detect acceleration of super - tev nuclei turned out to be a very difficult task . despite extensive search campaigns ( _ e.g. , _ * ? ? ?
* ; * ? ? ?
* ) , only three snrs have shown detectable tev emission so far @xcite .
nevertheless , a signature of protons accelerated to super - tev energies was reportedly discovered in the snr rx j1713.7 - 3946 by the cangaroo team @xcite .
recently , this remnant has been confirmed as a tev source , with significantly reduced systematic and statistical errors in the range 1 - 10 tev , by the hess collaboration @xcite .
this remnant is of a shell type , typical for the major acceleration models , and rather abundant in the galaxy .
therefore , after some verification of the data and analysis , it could serve as a direct evidence of the cr - snr connection . a number of groups ( _ e.g. , _ * ? ? ?
* ; * ? ? ?
* ) reanalyzed the data of @xcite and claimed that the nucleonic interpretation is `` highly unlikely '' because of an alleged inconsistency with the commonly assumed first order fermi acceleration theory .
we argue below that if ( as it is assumed for this snr by * ? ? ?
* ; * ? ? ?
* ) , some of the accelerated particles begin to interact with a partially ionized dense gas such as a mc in the _ northwestern _ rim of the remnant , the standard acceleration model is not applicable . in particular , the nonlinearity of the acceleration process as well as alfven wave evanescence in the mc are essential .
therefore , the analyses by both sides of the controversy are oversimplified and we believe that a conclusive case for or against nucleonic origin of the emission has not been made .
moreover , a recent analysis by @xcite indicates that the standard acceleration model applied to tev electrons does not adequately fit the tev data , either , unless the x - ray emitting filaments have an unreasonably low filling factor of @xmath2 .
@xcite have increased it to @xmath3 by allowing a smoother spectrum cut - off .
these analyses indicate that modifications to the accelerated particle spectra are required .
namely , a low - energy cut - off above @xmath4 and/or a _ spectral break _ would support the nucleonic scenario @xcite while models with an exponential cut - off rather than a break provide only a bad fit .
such modifications are suggested below .
the controversy about the tev observations of the snr rx j1713.7 - 3946 is fundamental and will clearly impair understanding of future observations of snrs nearby mcs .
therefore , in this letter we revisit shock acceleration theory and include the following physical phenomena in the model : ( i ) the nonlinearity of the acceleration process , i.e. , the modification of the flow by accelerated particles ( ii ) a position dependent _ low - energy cut - off _ of accelerated particles ahead of the shock , which is _ well known _ to be present in analytic solutions for the shock acceleration problem .
note that this cut - off is usually ignored in shock acceleration models since only the downstream solution is considered . when the dense gas is present upstream of the shock this cut - off
should be included in the calculation of @xmath0-radiation .
( iii ) impairment of cr confinement in a dense gas ( mc ) due to non - propagation of some alfven waves .
the latter phenomenon produces a break ( curvature ) in the particle and radiation spectra in the tev energy range and , in combination with ( i ) , results in a spectral slope significantly different from the usual predictions of linear theory supplemented with the high energy cut - off , which was used in the analyses of @xcite and @xcite , for example .
the diffusive shock acceleration ( dsa ) mechanism operates by scattering charged particles off magnetic irregularities ( alfven waves ) across a shock , such as a sn blast wave ( _ e.g. , _ * ? ? ?
* ) . in the linear acceleration regime ,
the momentum distribution is a power - law , @xmath6 where @xmath7 is a normalization constant and the index @xmath8 depends only on the shock compression ratio @xmath9 .
as soon as the flow is disturbed by the pressure of accelerated particles , the spectrum becomes more complex .
it flattens toward higher energies and may become steeper at lower energies .
the spectrum @xmath10 is the _ downstream _ spectrum , which is coordinate independent in both the linear and nonlinear regimes . here
the upstream spectrum is more important since the interaction with the adjacent mcs starts _ upstream _ where a low energy _ cut - off _ occurs .
indeed , the solution upstream , valid for both the linear and nonlinear regimes , reads ( _ e.g. , _ * ? ? ?
* ) @xmath11.\label{spec}\ ] ] here @xmath12 is the spectrum downstream , @xmath13 .
the @xmath5 coordinate points upstream ( @xmath14 ) from the shock ( @xmath15 ) and @xmath16 is the speed of the plasma flowing into the shock ( @xmath17 ) .
now , @xmath18 grows with @xmath19 , most likely linearly , @xmath20 . according to eq.([spec ] ) , there exists a cr precursor , which is of the length @xmath21 .
furthermore , high energy particles diffuse ahead of low energy particles so the spectrum ( [ spec ] ) has an exponential _ low - energy cut - off _ at @xmath22 for the bohm diffusion we have @xmath23 where @xmath24 is the magnetic field in @xmath25 .
this low - energy decay can explain why the same mc can be visible in the tev energy range and invisible in the gev range .
let the near - most edge of an adjacent mc be upstream of the shock at a distance @xmath26 , such that @xmath27 .
this low - energy cut - off @xmath28 is shown in fig.[fig : sp ] ) : @xmath29 , @xmath30 , and the distance from the sub - shock to the mc , @xmath31 .
note that for the visibility of the tev and higher energy protons , it is sufficient that @xmath32 . ]
that will be discussed below .
the last condition means that the leading edge of the cr precursor ( filled with tev protons ) has already penetrated into the mc , but the sub - shock itself ( with the gev particles ahead of it ) still has not .
obviously , the cangaroo and hess could detect the proton tev emission while the egret could not , because of the low density inside the wind bubble the sub - shock is located in .
the strong x - ray emission expected when the sub - shock crashes into the dense mc will not yet be visible , either .
in the presence of weakly ionized dense gas the particle and emission spectra undergo significant modifications in the tev energy range , where a spectral break can form . according to eq .
( [ fig : sp ] ) the accelerated particles occupy an extended precursor of the size @xmath33 ahead of the shock . here
@xmath34 is the particle gyroradius .
one can estimate @xmath35 as + @xmath36 thus , the shock precursor may be as long as @xmath37 for @xmath38 . therefore
, the accelerated particles start to interact with a mc before it becomes significantly ionized by the shock wave or the ionizing precursor @xcite .
hence , they propagate in the mc under conditions of strong ion - neutral collisional damping of the self - generated alfven waves that are needed to confine accelerated particles .
more importantly , there is a gap in wavenumber space at @xmath39 where the waves do not propagate @xcite . here
, it is critical to realize that waves _ literally do not exist _ , as opposed to simply being damped by collisions .
the above wave evanescence range is bounded by @xmath40 and @xmath41 , where @xmath42 is the ratio of the ion to neutral mass density , @xmath43 is the ion - neutral collision frequency and @xmath44 is the alfven speed .
the resonance condition for the wave generation and particle scattering off them is @xmath45 , where @xmath46 and @xmath47 is the parallel ( to the magnetic field ) component of particle momentum .
therefore , particles having @xmath48 , where @xmath49 are not scattered by any waves , so that the confinement of all particles with @xmath50 is dramatically degraded .
we assume that the gap in @xmath51 is sufficiently broad , so that resonance broadening ( _ e.g. , _ * ? ? ?
* ) does not bridge it .
we also assume that there is no significant background turbulence at the scales @xmath52 , that could reduce particle diffusivity to a limit which is not significantly higher than the bohm value .
for the parameters used in our calculations below , @xmath53 .
then , particles with @xmath50 will escape from the mc upstream of the shock along the magnetic field at a speed comparable to @xmath54 or , at least , their confinement time to the shock will be much shorter than those with @xmath55 . only the latter particles will generate @xmath56 mesons and @xmath0 emission efficiently in the mc .
their pitch angle distribution for @xmath57 is , however , limited to the interval @xmath58 .
therefore , the contribution to the @xmath0 emission of particles with momentum @xmath57 is reduced by the phase volume filling factor @xmath59 .
the resulting emission spectrum is thus one power steeper as compared to the standard calculations ( _ e.g. , _ * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) based on the isotropic particle distribution . the latter yields the @xmath0 spectrum which reproduces that of the particles up to about @xmath60 where it declines . hence , the energy spectrum of @xmath0 emission must be the same as that of the particles , @xmath61 , for the photon energy @xmath62 ( particle momenta @xmath63 ) , and it must scale as @xmath64 for @xmath65 up to about @xmath60 ( particle momenta @xmath66 ) .
the break energy @xmath67 can be estimated using eq .
( [ eq : p12 ] ) for @xmath68 by substituting @xmath69 , where @xmath70 is the cross - section of the ion - neutral collisions and @xmath71 is the collision velocity averaged over a thermal distribution . using an approximation of @xcite for @xmath72
, @xmath68 can be estimated as @xmath73 where @xmath74 is measured in the units of @xmath75 , @xmath76 and @xmath77 ( number densities corresponding to the mass densities @xmath78 and @xmath79 ) -in @xmath80 . to illustrate these results we consider three different acceleration regimes , corresponding to the three substantially different shock compression ratios @xmath9 , shown by the points 1 - 3 in the inset to fig.[fig : sp ] .
( here @xmath9 means a total shock compression , that is adiabatic compression across the precursor times the subshock jump . ) as it can be seen from this plot , the shock compression is difficult to calculate accurately due to its sharp dependence on the injection rate ( usually not well known ) and other parameters , such as the maximum momentum .
we discuss this spectrum in more detail in the next section .
in this paper we have considered the _ upstream _ particle spectrum since this is necessary for interpretation of cases where a shock is expanding into a low - density pre - supernova wind bubble and is approaching a denser material such as a swept - up shell or a mc .
perhaps the most discussed snr of this kind is the rx j1713.7 - 3946 claimed by the cangaroo team to be a long anticipated proton super - tev accelerator @xcite .
significant part of the tev emission detected by the cangaroo and hess @xcite instruments comes from the northwestern rim of the remnant where the interaction with a molecular cloud is believed to take place ( _ e.g. , _ * ? ? ?
* ; * ? ? ?
therefore these observations are pertinent to the subject of this letter . to demonstrate that the observed tev spectra are consistent with the mechanisms suggested in this paper , we put the calculated upstream _ proton _ spectra on one plot with the cangaroo / hess and egret _ photon _ fluxes that would be radiated by these protons .
since the emission generated by protons has typically an order of magnitude higher energy , we up - shifted the _ observed _ emission energy accordingly in fig .
[ fig : sp ] .
as the spectrum between @xmath81 is considerably flatter than in the case of the linear theory , its magnitude at @xmath82 is an order of magnitude higher for the same injection rate .
this lowers the ambient gas density required to produce the same tev emission as computed by @xcite using the linear spectrum .
in addition , this allows to accommodate the egret points , should they belong to the same object @xcite .
however , because of the controversy about the distance to the remnant ( 6 vs 1 kpc ) @xcite and other uncertainties , it is impossible to present detailed normalized fits of the photon flux in this short letter .
the form of the spectrum is marked by increasing the spectral index by one at the break momentum @xmath82 , fig .
[ fig : sp ] .
this is crucial to fitting the observed spectra without imposing a specific form and value of the energy cut - off .
the latter procedure would give only a poor fit ( _ e.g. , _ * ? ? ? * ) .
note that there are two cangaroo points that are not in a good agreement with both the hess and with the theoretical spectrum .
it should be noted , however , that any disagreement between the three sets of spectral points is substantially enhanced in this particular spectrum format ( the particle phase density is multiplied by @xmath83 ) . except for these two points
the agreement is remarkably good .
interestingly enough the agreement is excellent for a subset of hess points with the lowest and highest energy points excluded . unlike the power - law index , it is more difficult to constrain the parameters determining the position of the break on the spectrum given by eq .
( [ eq : p1 ] ) due to the poor information about the target gas for @xmath84 reactions ( mc ) . for the particular case shown in fig.[fig : sp ] , we have chosen such combination of parameters in eq .
( [ eq : p1 ] ) that @xmath68 amounts to @xmath85 .
mcs in general are known to be `` clumpy '' with the inter - clump gas density of 5 - 25 @xmath80 and a less than 10% filling factor ( see ,
_ e.g. , _ * ? ? ?
* ; * ? ? ?
* and references therein ) .
the results shown in fig . [ fig : sp ] are obtained for the following values of parameters in eq.([eq : p1 ] ) : @xmath86 , @xmath87 , @xmath88 , and @xmath89 which yields @xmath90 . by contrast , in the model of @xcite much lower target density is assumed since its authors do not consider the possibility that the cr precursor reach the dense gas while the shock itself is still in the wind bubble .
therefore they concluded that the @xmath84 reactions do not contribute significantly to the tev emission because of the lack of target protons .
in contrast to the conclusion of @xcite and to that of @xcite the most recent hess observations allow their authors @xcite to assume that both protons and electrons contribute to the tev emission significantly .
note that a complementary mechanism of suppression of the low - energy ( gev ) emission has been suggested earlier by @xcite and discussed recently in the context of snr rx j1713.7 - 3946 by @xcite .
this mechanism requires an impulsive release of accelerated particles at some distance from a mc ( target ) and subsequent ( energy dependent ) diffusive propagation of crs to the target .
by contrast , the mechanism discussed in this paper is based on a quasi - stationary solution of the acceleration problem ( eq . [ [ spec ] ] ) which implies that the accelerated particles are bound to the propagating shock front via self - generated alfven waves .
it is clear , however , that in both cases the suppression of the low - energy emission is based on a slower diffusion of low - energy particles .
note that @xcite also discussed qualitatively the same effect .
it is also important to mention that gev emission may not be suppressed all over the remnant , and was demonstrated by @xcite to be likely of proton origin .
obviously , the spectral modifications considered in this letter are equally applicable to electrons of similar rigidity .
it should be pointed out that our model in its current version does not provide a complete fit to the data , particularly in the x- and radio energy bands .
the tev - gev fluxes are given in arbitrary units , so only the _ form _ of the spectrum ( including possible gev signal ) is provided here .
a complete fit would require the convolution of the proton spectrum with appropriate emissivities given the densities and the distance .
the thermal peak in fig.[fig : sp ] is a maxwellian , approximately normalized relative to the high - energy part of the spectrum , in accordance with the calculated downstream temperature without turbulent heating .
more accurate calculation of the thermal peak would require adequate injection model and turbulent gas heating in the precursor ( _ e.g. , _ * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
such calculations of the thermal x - ray emission should be tested specifically against very low observed upper limits .
one _ potentially _ observable prediction of our model is the break in the pitch angle averaged particle spectrum at @xmath82 that would be produced by almost instant losses along the field lines of all particles with @xmath50 once they enter a partially ionized gas with significant ion - neutral collisions .
this break may very well be in the currently `` obscure '' energy range between the egret and the ground based cherenkov telescope energy bands .
the next generation of space gamma telescopes , such as glast , will be able to explore this energy range .
the principal results of this letter , which are the spectrum softening by one degree above the spectral break , and the possibility of a low energy cut - off clearly show that these physical phenomena need to be included in the models to conclusively differentiate between the nucleonic and leptonic sources of the tev emission from the remnants nearby molecular clouds such as the rx j1713.7 - 3946 .
previous claims to the contrary have ignored necessary refinements in the dsa theory beyond the level of test particle theory .
berezinsky and ptuskin(1989 ) achterberg , a. 1981 , , * 98 * , 161 aharonian , f.a . , drury , l.oc . and
vlk , h.j .
1994 , , 285 , 645 aharonian , f.a .
, atoyan , a.m. 1996 , , 309 , 917 aharonian , f. a. _ et al .
_ 2001 , , 370 , 112 aharonian , f. a. _ et al .
_ 2004 , nature , 432 , 75 berezinsky , v.s .
& ptuskin , v.s .
1989 , 340 , 340 berezhko , e. g. , yelshin , v. , & ksenofontov , l. 1996 , sov .
jetp , 82 , 1 blandford , r.d . & eichler , d. , 1987 , , 154 , 1 buckley , j. h. _ et al . _ , 1998 , , 329 , 639 butt , y.m .
_ et al .
_ 2002 , nature , 418 , 499 butt , y.m . _ et al .
_ 2001 , , 562 , l167 bykov , a.m. _ et al .
_ , 2000 , , 538 , 203 chevalier , r.a . , 1999 , , 511 , 798 draine , b.t . , mckee c.f .
, 1993 , annu .
, 31 , 373 drury , l. oc , aharonian , f. a. & vlk , h. j. , 1994 , , 287 , 959 drury , l. oc , duffy , p. kirk , j. k. , 1996 , , 309 , 1002 ellison , d.c . , slane , p. & gaensler , b.m .
2001 , , 563 , 191 enomoto , r. _ et al .
_ 2002 , nature , 416 , 823 fukui , y. _ et al .
_ 2003 , pasj , 55 , l61 hartman , r.c .
_ et al .
_ 1999 , , 123 , 79 hiraga , j.s . _
et al . _ , 2003 , astro - ph/0407401 kang , h. , jones , t.w . &
gieseler , u.d.j .
2002 , 579 , 337 koyama , k. _ et al .
_ 1995 , nature , 378 , 255 kulsrud , r. & pearce , w.p .
1969 , , 156 , 445 kulsrud , r.m .
& cesarsky , c.j .
1971 , , 8 , 189 lazendic , j.s .
_ et al . _ , 2004 , , 602 , 271 malkov , m. a. & drury l.oc . , 2001 , _
_ , 64 , 429 muraishi , h. _ et al .
_ 2000 , , 354 , l57 naito , t. & takahara , f. 1994 , j. phys .
g20 , 477 pannuti , t.g . ,
allen , g.e . ,
houck , j.c .
and sturner , s.j .
2003 , , 593 , 377 reynolds , s.p .
1998 , 493 , 375 reimer , o. & pohl , m. 2002 , , 390 , l43 slane , p. _ et al .
_ 1999 , 525 , 357 sturner , s.j . , skibo , j.g .
, dermer , c.d . &
mattox j.r .
1997 , 490 , 619 tanimori , t. _ et al .
_ 1998 , , 497 , l25 uchiyama , y. , aharonian , f.a . , and takahashi , t. , 2003 , , 400 , 567 vlk , h.j .
2000 , in aip conference proceedings , 515 , 197 zweibel , e. , and shull , j.m .
1982 , , 259 , 859 | cosmic rays ( crs ) are thought to be accelerated in snrs .
the most favorable situation for proving that the main , hadronic cr component is accelerated there is when crs interact with dense gases , such as molecular clouds ( mc ) which surround the sn shock . here , a new mechanism of spectrum formation in partially ionized gases near snrs is proposed . using an analytic model of nonlinear diffusive shock acceleration
, we calculate the spectra of protons and estimate the resulting @xmath0-ray emission occurring when the snr shock approaches a mc .
we show that the spectrum develops a break in the tev range and that its gev component is suppressed .
these modifications to the standard theory occur because of the proximity of the partially ionized mc - gas and because of the physics of particle and alfven wave propagation inside the gas .
possible applications of the new spectra to the recent cangaroo and hess observations of the snr rx j1713.7 - 3946 are discussed . |
on approaching a critical point , some quantity diverges in the thermodynamic limit with a characteristic critical exponent . the coherent anomaly method ( cam )
has proven quite successful in determining critical exponents from certain sequences of approximations @xcite .
the cam requires a systematic , or _ canonical _ , sequences of approximations , all of which yield identical , known critical exponents .
the prototypical example is a sequence of mean - field approximations in which successively larger clusters allow more and more fluctuations to be properly taken into account @xcite ; all of the critical exponents in this example assume their `` classical '' values .
the cam uses the known critical exponents of the approximate systems and the behaviour of the critical _ amplitudes _ as the quality of approximation is improved to determine the true critical exponents of the original system being approximated .
the purpose of this paper is to show how sequences of approximations in which the critical points are ill defined and divergences do not occur can be used to construct canonical sequences through extrapolation .
the basic idea is as follows .
suppose that in the thermodynamic limit a nonnegative quantity @xmath4 diverges as a system parameter @xmath5 approaches a critical value @xmath6 as @xmath7 where @xmath1 is a critical exponent of unknown value .
the reciprocal of @xmath4 then converges to zero as @xmath8^{-1 } \sim ( \beta_{\rm c}^ * - \beta)^\nu \ ; .\ ] ] suppose further that we have a sequence of monotonically decreasing approximations @xmath9 such that @xmath10 by taking two or more consecutive values of @xmath11 , we make an extrapolation @xmath12 which _ intentionally underestimates _ the rate of convergence in @xmath13 .
clearly , these extrapolations will be negative at @xmath6 , and they generically cross zero as @xmath14 for some @xmath15 . for any reasonable extrapolation procedure ( e.g. power - law extrapolation or exponential extrapolation ) ,
the cam hypothesis , which can be justified on the basis of an envelope argument @xcite , is that @xmath17 so that , in analogy with , @xmath18 this provides us with a convenient means of measuring @xmath1 . according to ,
a plot of @xmath19 versus @xmath20 should ( in the limit @xmath21 ) be a straight line with slope @xmath22 .
furthermore , since all that is required of the extrapolation is that it must underestimate the convergence , we are free to choose an extrapolation which leads to a particularly clear coherent anomaly .
the organization of the remainder of this paper is as follows . in
we take as @xmath23 the mass gap , i.e. , the reciprocal of the correlation length , as a function of inverse temperature in the square - lattice ising ferromagnet .
this model has the advantage that the correlation length is known analytically @xcite . in
we analytically study the asymptotic behaviour of the estimated critical exponent if @xmath24 is given by a finite - size scaling function . in we study the critical behaviour of the mass gap in the cubic - lattice ising ferromagnet , and demonstrate that the method works even when @xmath25 . in we study the critical behaviour of the energy gap in an antiferromagnetic spin-1 heisenberg chain with bilinear and biquadratic interactions at the phase transition between the haldane and dimerized phases . in we summarize and discuss possible extensions of this work
. note that although all of the initial approximations @xmath24 used in this study come from systems that are finite in at least one dimension , this is only to provide convenient examples .
the method itself does not restrict us to such systems .
as the first example , we use the extrapolation - cam method to determine @xmath1 for the classical ising ferromagnet with hamiltonian @xmath26 where @xmath27 .
the sum @xmath28 runs over all nearest - neighbour pairs on a periodic square lattice which is of length @xmath29 in the @xmath30-direction and of infinite length in the @xmath31-direction .
the unit of length is the lattice constant .
this model has some very advantageous properties : the mass gaps @xmath32 can be calculated analytically for systems of arbitrary finite width @xmath29 , and the mass gap @xmath23 for the thermodynamic limit of the model can also be calculated analytically @xcite .
consequently , we can make a detailed comparison of the extrapolation - cam estimates for the critical exponent @xmath1 with its rigourously known value , @xmath0 . for this model
the variable @xmath5 is the reciprocal of the dimensionless temperature . since it is known that @xmath33 for sufficiently large systems at the critical point @xcite , we extrapolate by solving @xmath34 for @xmath35 at fixed @xmath5 for @xmath36 .
the result of such an extrapolation is shown in
. shows @xmath37 vs. @xmath38 , defined by and , to be a curve with a small slope .
the slope tends to zero as @xmath39 becomes large that is , when @xmath40 becomes a very good approximation to @xmath6 .
we can estimate @xmath1 from the slope of the line connecting two adjacent points @xmath41 and @xmath42 .
this estimate will depend not only on the quality of the initial approximations @xmath43 , but also on the parameter @xmath44 .
we can constrain @xmath44 by taking a third point @xmath45 and demanding that the three points be colinear , so that the value of @xmath46 will be unambiguous .
shows this for values of @xmath1 based on systems with @xmath47 , 9 , 16 , and 25 .
the resulting estimate is @xmath48 , in good agreement with the exact value @xmath0 .
as may be expected , increasing the sizes of the four systems needed to form the estimate of @xmath1 increases the accuracy of @xmath46 . shows the estimated value of @xmath1 vs. @xmath44 for @xmath49 , @xmath50 , @xmath51 , and @xmath52 , where @xmath53 .
the estimate for @xmath54 is @xmath55
. further increase of @xmath56 actually causes the accuracy of the estimate to become worse due to the increasingly large sums required to calculate @xmath57 and @xmath58 and the finite ( eight - byte ) numerical precision of our programs .
the convergence of @xmath46 depends on how the four system sizes are chosen , and can be both complicated and nonmonotonic .
some insight into the dependence of @xmath46 on the system sizes can be gained by assuming that @xmath59 satisfies the scaling equation @xcite @xmath60 where @xmath61 and @xmath62 this assumption is well justified for the systems actually analyzed in this paper , since they all are finite in at least one dimension . in order to be consistent with
, we must have @xmath63 , and @xmath64 @xcite .
the variable @xmath65 is the correction - to - scaling amplitude and @xmath66 is the correction - to - scaling exponent , where the correction to scaling is assumed to arise from the leading irrelevant field @xcite . for large @xmath67 , yields @xmath68 \ , .
\end{aligned}\ ] ] for the remainder of this section , we consider @xmath29 to be a single continuous variable and drop the second index @xmath69 .
the @xmath70 term in comes from the truncation error in a difference formula approximation of a derivative .
this term can not be neglected if @xmath71 , unless we generalize the extrapolation procedure to allow for a sufficiently high - order difference formula @xcite ; for the remainder of this section we assume that this is done if necessary and neglect the truncation error . for the moment
, we set @xmath72 and perform the extrapolation - cam procedure in the absence of corrections to scaling .
it is convenient at this point to use a slightly different definition for @xmath37 , @xmath73 using , we find @xmath74 \nonumber \\ & = & \left(1 + \frac{\omega}{\theta}\right ) \left[\ln x - x \right ] - \ln
b + \ln \frac{\partial f}{\partial x}(x , b ) \ , .\end{aligned}\ ] ] we find @xmath75 from the condition @xmath76 , which implies @xmath77 the estimate for @xmath1 is given by calculating @xmath78 while holding both @xmath79 and @xmath44 constant , which is accomplished by holding @xmath31 constant : @xmath80 finally , the `` local straightness '' constraint @xmath81 is simply the identity @xmath82 .
thus without corrections to scaling , the @xmath83 but @xmath31 and @xmath44 are undetermined .
note , however , that if @xmath84 , @xmath85 . in order to study the effects of corrections to scaling , we could expand and the left - hand side of ( which is proportional to @xmath65 ) in @xmath31 , @xmath30 , and @xmath86 ; then the requirements @xmath87 and @xmath88 yield a pair of equations of the form @xmath89 where @xmath90 and @xmath91 are constants .
the solution of , @xmath92 shows that both @xmath31 and @xmath93 are proportional to @xmath30 . in the same way
, we can expand the left - hand side of in terms of @xmath31 , @xmath30 , and @xmath93 .
the result then is that @xmath94 and @xmath95 privman and fisher have shown that the asymptotic convergence for finite - size scaling renormalisation techniques is also of the form @xcite ; their discussion of the difficulties in actually observing this asymptotic behaviour should be relevant in the present case as well .
for the two - dimensional ising model , @xmath96 , so we should expect @xmath97 , as indeed seems plausible from .
in three dimensions , the ising model defined by has not been solved analytically , but it has been the subject of a large amount of numerical study .
we use a monte carlo renormalisation group estimate for the critical point of the cubic - lattice ising model , @xmath98 @xcite .
recent estimates of @xmath1 include @xmath99 @xcite , @xmath100 @xcite , and @xmath101 @xcite . as a result of some earlier studies @xcite ,
we have transfer - matrix results already available for cubic - lattice ising ferromagnets with periodic boundary conditions and square @xmath102 cross - sections , where @xmath103 . these system sizes are obviously quite small , and are not really competitive with some current methods , such as the transfer - matrix monte carlo method @xcite .
the method here is basically the same as in the previous section , although we are much more restricted both in the system sizes available and , for @xmath104 , the number of data available . since references @xcite dealt with phenomena for @xmath105 , we have only a few points @xmath106 with @xmath107 [ see ] .
we use cubic splines to evaluate both @xmath43 and the first @xmath58 as continuous functions .
except at the low-@xmath5 end of the @xmath104 data , we always use `` natural '' boundary conditions on our cubic splines , i.e. specifying that @xmath108 at the end points of our data .
for the low-@xmath5 end of the @xmath109 spline , we use both natural boundary conditions and `` clamped '' boundary conditions by extrapolating @xmath110 from the smaller system sizes .
there is little difference in @xmath109 itself for these two splines .
we extrapolate @xmath111 using and use , , and to estimate @xmath1 . as shows , the splines yield unreliable estimates near the end of the @xmath109 data . however ,
as long as @xmath112 is reasonably large , the estimates of @xmath1 do not depend on the boundary conditions used for the spline and seem more reliable . choosing five such points and using aitken s @xmath113 method @xcite is a part of the name of the numerical method and should not be confused with either @xmath23 or @xmath114 . ] to accelerate the convergence , we extrapolate to @xmath115^{-1 } \rightarrow \infty$ ] , yielding @xmath116 . given the very small systems used in this estimate , this is in good agreement with other recent estimates of @xmath1 .
the heisenberg chain we study is defined by the quantum hamiltonian @xmath117 \ ; , \ ] ] where @xmath118 is the quantum spin-1 operator for the spin at site @xmath119 and @xmath120 ( periodic boundary conditions ) . in the limit @xmath121 ,
the spectrum of this hamiltonian has been exactly solved for at @xmath122 using bethe - ansatz techniques , and the resulting energy spectrum is gapless @xcite .
the ground state of the hamiltonian has also been found @xcite for even values of @xmath29 at @xmath123 , and the gap above it has been proven to be nonvanishing .
it has been argued @xcite that @xmath122 marks a phase transition between the gapped haldane phase @xcite and the gapped dimerized phase , and that this phase transition is in the same universality class as the two - dimensional ising model .
these arguments have been supported numerically @xcite .
shows the estimates @xmath43 obtained by exact diagonalization of small systems .
the finite - size effects are quite strong , and we were unable to make an extrapolation of the form , since we require @xmath124 for all extrapolations in order for @xmath38 and @xmath37 to be real and finite .
instead we solve @xmath125 for @xmath126 at fixed @xmath5 , and then solve @xmath127 for @xmath128 using the value of @xmath126 found from . here
the parameter @xmath130 allows us to tune the extrapolation as in the preceding sections .
shows an example of extrapolations which in fact yield the best estimate of @xmath1 .
even with the extrapolation procedure outlined above , we are not able to eliminate the curvature from the cam plot as we did in sections [ sec-2di ] and [ sec-3di ] .
instead , we perform a fit to the form @xmath131 while varying @xmath130 to find @xmath1 , assigning an arbitrary fixed weight equally to all of the cam points .
the best estimate for @xmath1 is the one that minimizes @xmath132 [ ] .
the best fit , shown in , is for @xmath3 , in good agreement with theoretical predictions .
in this paper we propose a new and very general method for constructing canonical sequences for use in the coherent anomaly method . a critical point @xmath133 is always marked by the vanishing of some quantity @xmath23 , though approximations of @xmath23 often remain nonzero for all values of @xmath5 . by intentionally underestimating the rate of convergence of an initial sequence of approximations @xmath9 , we form extrapolations @xmath134 that cross zero at some value of @xmath5 , which serves as the approximate critical point @xmath135 . furthermore
, @xmath114 can very generally be expected to cross zero linearly with @xmath5 .
because there are many ways in which the extrapolations can be made from an initial sequence @xmath136 , we have a great deal of freedom to choose an extrapolation which shows a clear coherent anomaly .
we apply this method to the square lattice ising model , using the exact values of the mass gaps for semi - infinite systems of finite width @xmath137 at temperature @xmath138 as @xmath43 .
we find that the method yields @xmath139 for moderate values of @xmath29 and that @xmath140 , as should be expected from the exact solution of the two - dimensional ising model @xcite . the convergence may be complicated and nonmonotonic , however , depending on which sets of @xmath29 are chosen to form the extrapolations . in order to investigate the convergence of @xmath141 to @xmath1 further
, we assume that @xmath59 follows the scaling equation .
this allows us to show that @xmath142 , which is the same convergence rate as has been found for finite - size scaling @xcite .
we also apply this method to the cubic - lattice ising model , using numerical transfer - matrix values of the mass gaps for semi - infinite systems with @xmath102 cross sections and periodic boundary conditions .
even though we are limited to systems with @xmath143 and have a limited amount of data for @xmath104 , we are able to estimate @xmath2 , which is within a few percent of the best current estimates of @xmath1 @xcite . finally , we apply this method to a one - dimensional spin-1 quantum heisenberg antiferromagnet , using numerical exact - diagonalization estimates of the energy gaps for systems of width @xmath29 and periodic boundary conditions . in spite of large finite - size effects , we are able to estimate @xmath3 , in good agreement with theoretical and numerical studies indicating @xmath0 @xcite . * although we use data from finite systems of successively larger size , we are not performing finite - size scaling @xcite .
the initial sequence @xmath136 could be derived from other techniques , such as the density - matrix renormalisation group algorithm @xcite , in which the system size is not the most important parameter affecting the quality of the approximation . *
although we use extrapolations , we are not seeking the _ best _ extrapolations in the sense of extrapolations which are nearest to the thermodynamic limit of the model in question .
this is because the extrapolation is just one step in the process .
instead , we seek a sequence of extrapolations for which the coherent anomaly is clear .
* taylor expansions have been used previously to form sequences of approximations with @xmath144 from series expansions @xcite , but in those studies the linear behaviour of the cam plot could not be improved , since the physics of the series expansion left no room for change .
the extrapolation - cam method has flexibility to choose the `` straightest '' cam plot . *
this method requires good numerical precision .
this is clear from .
care should therefore be exercised when applying this method to monte carlo data , where statistical uncertainty may be significant . * although we here are estimating only @xmath1 , other critical exponents could be found in the same way .
for instance , instead of extrapolating the gap , one could extrapolate the reciprocal of the specific heat . * some initial sequences
@xmath9 may already cross zero as @xmath5 is varied .
for instance , the variational method of references @xcite produces gap estimates for the spin-1 heisenberg chain that have this property .
in such a case , extrapolations could still be used to look for a clearer coherent anomaly .
the extrapolations should then be faster than the convergence of @xmath9 , rather than slower , so that the sequence @xmath145 still crosses zero .
this work was supported by the inoue foundation for science .
the authors wish to thank professors m kaburagi and h nishimori for useful comments and for the use of their programs kobepack and titpack , respectively .
the data for the three - dimensional ising model were taken at florida state university ; consult reference @xcite for details .
the authors thank professor per arne rikvold for comments on the manuscript . | we propose and test a new method for generating canonical sequences for analysis by the coherent anomaly method ( cam ) from non - mean - field approximations . by intentionally underestimating the rate of convergence of exact - diagonalization values for the mass or energy gaps of finite systems , we form families of sequences of gap estimates .
the gap estimates cross zero with generically nonzero linear terms in their taylor expansions , so that @xmath0 for each member of these sequences of estimates .
thus , the cam can be used to determine @xmath1 . our freedom in deciding exactly how to underestimate
the convergence allows us to choose the sequence that displays the clearest coherent anomaly .
we demonstrate this approach on the two - dimensional ferromagnetic ising model , for which @xmath0 .
we also use it on the three - dimensional ferromagnetic ising model , finding @xmath2 , in good agreement with other estimates .
finally , we apply it to an antiferromagnetic spin-1 heisenberg chain , finding @xmath3 at the phase transition between the haldane phase and the dimerized phase , in agreement with the field - theoretic prediction @xmath0 .
although the specific systems used to test the extrapolation - cam procedure involve finite system sizes , the method could be applied to other finite approximations , such as systematic variational approximations .
[ extrapolation - cam theory ] |
the monitoring of galactic sources by the _ wide - field camera _ on bepposax and the _ all - sky monitor _ on the rossi x - ray timing explorer have detected large type i bursts ( hereafter referred to as `` superbursts '' ) from six accreting neutron stars ( cornelisse et al .
2000 ; strohmayer 2000 ; heise et al .
2000 ; wijnands 2001 ; kuulkers 2001 ) .
these bursts share a number of characteristics ; burst energies @xmath7 ( roughly 1000 times larger than a normal type i burst ) ; durations of a few hours and accretion rates @xmath8 ( wijnands 2001 ) .
the superburst recurrence time is not well known , though one recurrent burst has been seen 4.7 years later ( wijnands 2001 ) .
the light curves are similar to a type i burst in spectral evolution , though the immediate decay timescale is about two hours ; 1000 times longer than a typical type i x - ray burst .
the superbursts appear to be thermonuclear flashes from fuel at much larger depths than a typical type i burst .
for example , if the nuclear energy release is 1 mev per accreted nucleon ( @xmath9 ) , the accumulated mass would be @xmath10 , implying a recurrence time of @xmath11 months for an accretion rate of @xmath12 , much longer than a typical type i burst .
flashes from pure @xmath1layers have been discussed previously ( woosley & taam 1976 ; taam & picklum 1978 ; brown & bildsten 1998 ) and might well apply to the pure helium accretor 4u 1820 - 30 ( strohmayer 2000 ) . however , there are two difficulties with such a model for the superbursts from hydrogen / helium accretors : ( 1 ) the recurrence times are very long , and the burst energies correspondingly large , unless a large heat flux from the crust of the star is present deep in the ocean ; and , ( 2 ) our current understanding of the ashes from hydrogen / helium burning does not point to accumulation of pure @xmath1 . instead , schatz et al .
( 1999 , 2001 ) have shown that only a small amount of @xmath1remains after all of the hydrogen and helium has burned via the rapid proton ( rp ) process ( wallace & woosley 1981 ) ; either from steady state burning or unstable type i x - ray bursting .
even though a small fraction by mass ( typically @xmath13 ) , we show here that it is enough to trigger a thermonuclear runaway with energy comparable to the superbursts .
most important , however , is the role played by the heavy ashes from the rp process .
these reduce the thermal conductivity enough to force a large temperature gradient in the ocean .
this ignites the @xmath1at much lower masses than previously expected and reduces the energy and recurrence times to levels consistent with the superbursts .
in addition , the conductive cooling times from these flashes are consistent with the long decay times observed .
we presume that the ashes of hydrogen / helium burning in the upper atmosphere primarily consist ( by mass ) of a single dominant nucleus of mass @xmath14 and charge @xmath15 and a small amount of @xmath1with mass fraction @xmath0 .
schatz et al . ( 2001 ) recently showed that the onset of a closed cycle ( called the snsbte cycle ) in the rp - process naturally stops the increasing @xmath16 trajectory once nuclei of mass @xmath17 are reached .
though a range of nuclei are made , those with masses in this region dominate the mix . for simplicity
, we take @xmath18as our fiducial nucleus ( @xmath19 and @xmath20 ) with mass fraction @xmath21 . the energy released when @xmath1fuses ( for simplicity , we presume that the reaction makes @xmath22 ) is @xmath23 , so that if @xmath24 , an accumulated mass of @xmath25 is needed for a @xmath26 burst . at an accretion rate of @xmath27 onto a @xmath28 neutron star , the recurrence time would be @xmath29 years ( we define @xmath30 , the local eddington accretion rate for solar composition ) .
the pressure at the base of such a pile is @xmath31 , where the surface gravity is @xmath32 . at such high pressures ,
the equation of state is that of degenerate , relativistic electrons with fermi energy @xmath33 , where @xmath34 . supporting this pressure requires a density @xmath35 and @xmath36 .
though they exert little pressure , the nuclei set the thermal conductivity via electron - ion scattering . for a classical one component plasma with ion separation , @xmath37 , defined by @xmath38 , where @xmath39 , the importance of coulomb physics for the ions is measured by @xmath40 where @xmath41 . for this initial calculation
we use the microphysics of the liquid state ; @xmath42 ( farouki & hamaguchi 1993 and references therein ) , even when , in some instances ( mostly when @xmath43 ) @xmath44 .
electron conduction dominates the heat transport in the liquid ocean .
the heat flux is given by @xmath45 , where @xmath46 is the thermal conductivity ( see yakovlev & urpin 1980 ; itoh et al .
1983 , potekhin et al . 1999 ) .
for @xmath47 , we find @xmath48 where we assume the electrons are relativistic , and take the coulomb logarithm to be unity ( a good approximation in the deep ocean ; bildsten & cutler 1995 ) . integrating the heat equation and hydrostatic balance gives the temperature contrast required between two points ( at pressure @xmath49 and @xmath50 ) to transport a heat flux @xmath51 ( bildsten & cutler 1995 ; brown & bildsten 1998 ) @xmath52 where @xmath53 is the eddington flux for a @xmath54 , @xmath28 star .
the flux in the ocean is the sum of compressional heating and heat flowing out from the crust , deposited there by electron captures and pycnonuclear reactions ( haensel & zdunik 1990 ; brown 2000 ) .
since both contributions are proportional to the local accretion rate per unit area , we write @xmath55 , with @xmath56 . equation ( [ eq : trel ] ) then gives @xmath57 in a careful study of the overall thermal balance of the neutron star , brown ( 2000 ) found @xmath58 for accretion rates @xmath59 ( i.e. , of the @xmath60 per nucleon deposited in the crust , @xmath29% escapes through the surface ; the rest is emitted as neutrinos from the inner crust and core ) .
the compressional heating term contributes @xmath61 per nucleon , small compared to the heat flux from the crust . for @xmath62 , the second term in equation ( [ eq : trel2 ] )
is close to unity for a pure @xmath1ocean ( @xmath63 ) , but @xmath64 times larger for a @xmath18ocean ( @xmath65 ) ( @xmath66 ) .
thus , an ocean of heavy ashes is hotter than that of pure @xmath1at the same depth .
this is critical , as it allows for ignition at lower column densities , @xmath67 .
figure [ fig : settle ] shows the temperature as a function of column depth for a pure @xmath1ocean , and for a @xmath1@xmath18mixture with @xmath1mass fraction @xmath68 . in each case
, we take @xmath69 , and show two accretion rates , @xmath70 , and @xmath71 .
the dashed line is the crystallization condition for @xmath18 .
the lower thermal conductivity makes the @xmath1@xmath18oceans much hotter than the pure @xmath1oceans .
we terminate the thermal profiles in figure [ fig : settle ] at the depth at which the @xmath1ignites .
brown & bildsten ( 1998 ) showed that the condition for an unstable @xmath1ignition in a neutron star ocean is set by competition with conductive cooling , so that a thermal runaway occurs when the local nuclear burning energy release , @xmath72 , satisfies @xmath73 , where @xmath74 is an approximation of the cooling rate ( e.g. , fujimoto , hanawa , & miyaji 1981 ) and the nuclear reaction rates are as described in brown & bildsten ( 1998 ) .
for the thermal conductivity given by equation ( [ eq : cond ] ) , @xmath75 , unstable ignition requires @xmath76 where @xmath77 is the temperature sensitivity of the fusion reaction .
ignition curves are shown in figure [ fig : settle ] for two cases .
the lower dotted line is for a pure @xmath1ocean and matches that in brown & bildsten ( 1998 ) .
the upper dotted line shows the ignition curve for a @xmath1@xmath18mixture with @xmath68 . because of the lower abundance
, a larger column density is needed to ignite @xmath1 in a @xmath18ocean for a given temperature
. however , ignition occurs at smaller column depths in this case , because the @xmath1@xmath18ocean is hotter . for the pure @xmath1models of figure [
fig : settle ] , ignition occurs at @xmath78 ( @xmath79 ) for @xmath80 ( @xmath81 ) . writing the burst energy @xmath82 , or @xmath83 , and recurrence time @xmath84
, we find that pure @xmath1ignitions have recurrence times @xmath85 ( @xmath86 ) and burst energies @xmath87 ( @xmath88 ) .
these burst energies exceed the energy of the superbursts by at least a factor of ten . for the @xmath1@xmath18models
, ignition occurs at @xmath89 ( @xmath90 ) for @xmath71 ( @xmath70 ) , with recurrence time @xmath91 ( @xmath92 ) , and burst energy @xmath93 ( @xmath94 ) .
not only must the thermal settling solution reach the ignition curve , but the @xmath1must also have survived to that depth .
thus , in order to obtain a flash , the @xmath1lifetime to the fusion reaction , @xmath95 must exceed the accumulation time , @xmath96 , at the ignition point . combining this condition with equation ( [ eq : ignite ] )
implies that @xmath97 must exceed the critical value @xmath98 when we evaluate this using @xmath46 and the @xmath99 , @xmath100 relations for degenerate relativistic electrons , the density cancels out , giving @xmath101 this relation is robust and shows that large accretion rates are required for unstable @xmath1ignition . in order to eliminate the temperature
, we neglect the outer temperature in equation ( [ eq : trel2 ] ) and substitute for the temperature at the base in equation ( [ eq : mdotc ] ) .
we then find that the explicit accretion rate dependences cancel and the requirement for an instability is just @xmath102 at smaller @xmath0 , the @xmath1depletes and burns steadily before reaching ignition .
figure [ fig : mdotx ] shows the @xmath1mass fraction required to achieve an explosion as a function of @xmath97 and @xmath103 .
the accretion rate dependence is mostly from the ignition pressure dependence on @xmath97 .
the top two panels in figure [ fig : gom ] show the burst energies and recurrence times for unstable models .
the curves are for @xmath68 and ( top to bottom ) , @xmath104 and @xmath105 .
having derived the conditions for an unstable ignition of a small mass fraction of @xmath1 , we now calculate the time dependent nature of the flash and estimate the time it takes for the energy to escape from the burning layer .
the `` one - zone '' version of the time dependent heat equation is @xmath106 in which we write the specific heat at constant volume @xmath107 rather than constant pressure ( an excellent approximation since @xmath108 throughout the flash ) . in the case of predominantly high @xmath109 nuclei ,
the electron specific heat dominates that of the ions , so @xmath110 .
once ignited , the @xmath1burning is extremely rapid and there is no time for cooling to occur while the @xmath1burns .
the fluid then evolves to a final temperature set by the total energy released @xmath111 .
using @xmath107 for the electrons and taking @xmath112 gives @xmath113 thus , @xmath114 is adequate to substantially change the temperature from that at ignition and trigger a thermal instability .
if @xmath115 , the peak temperatures would exceed @xmath116 and neutrino cooling would reduce the observed energetics of the event .
after all the @xmath1has burned and the fluid has reached @xmath117 , the layer cools . integrating equation ( [ eq : dtdt ] ) starting from @xmath117 with only the @xmath118 term , we find @xmath119 , where @xmath120 depends only on the depth in the star and the composition , and is temperature independent .
the luminosity during the conductive cooling phase is thus set by the temperature at the base of the layer ; @xmath121 .
thus , the observer should see a timescale of luminosity decay equal to equation ( [ eq : tcool ] ) , as shown in figure [ fig : gom ] .
if we use @xmath117 from equation ( [ eq : tfinal ] ) for an upper limit , we find the maximum conductive cooling luminosity @xmath122 due to our neglect of convective heat transport , this @xmath123 will be less than the peak luminosity of the burst .
however , since convection can not last much longer than the burning time , we expect something similar to this simple luminosity evolution to occur during the burst decay , over times relevant to detection by flux monitoring on hourly timescales .
detection of these flashes should be easiest at low @xmath2 s , as @xmath123 increases due to ignition at higher densities , whereas the accretion luminosity , @xmath124 , is dropping .
hence , the contrast of the burst cooling tail is largest for low @xmath2 sources , as shown in the bottom panel of figure [ fig : gom ] .
detecting one of these flares from a rapidly accreting @xmath109 source would require flux sensitivity at the @xmath125 level on a timescale of a few hours and spectral sensitivity to distinguish that the flux rise is from extra thermal emission .
since these bursts will occur roughly once a month , it is an archival project to find the rare event in the accumulated observations .
we have shown that the ignition of a small mass fraction of @xmath1 in a neutron star ocean consisting mostly of heavier elements is thermally unstable and leads to flashes whose energy , recurrence times and durations are similar to the superbursts .
the energy from these flashes takes a long time to escape the star , possibly explaining the persistent `` offset '' in the flux nearly a day after the superburst in 4u 1735 - 44 ( see figure 1 of cornellise et al .
we also predict flashes at high local accretion rates which include the z sources and accreting x - ray pulsars . in the pulsar case , some of the difficulties associated with explaining flares from lmc x-4 via pure @xmath1burning ( see brown & bildsten 1998 )
might well be alleviated in the mixed case we present here .
cornellise et al .
( 2000 ) reported an absence of regular type i bursts for seven days after the superburst .
this might be due to thermal stabilization of the hydrogen / helium burning layers by the large heat flux from the cooling ashes of the @xmath1burning .
paczynski ( 1983 ) and bildsten ( 1995 ) have shown that luminosities in excess of the helium burning flux ( or @xmath126 ) will stabilize the burning , as then the thermal state of the burning layer is independent of the local burning rates .
this appears most likely at low accretion rates , where the conductive cooling is slowest , thus halting type i bursts for @xmath127 days after the superburst .
there is still much to be done , including a time dependent study of the onset of the instability .
our initial integrations of the time - dependent code from bildsten ( 1995 ) demonstrate a transition from steady - state burning at low @xmath97 to unstable flashes at higher rates . in future work , this code will be used to more accurately calculate the transition accretion rate , @xmath128 with realistic mixes of @xmath1and rp - process ashes .
only then can we make the comparisons to observations that will allow us to constrain the rp - process ashes .
we thank erik kuulkers and marten van kerkwijk for conversations about the properties of the superbursts and hendrik schatz for insights on the rp - process .
we also thank deepto chakrabarty and bob rutledge for comments on the manuscript .
this research was supported by nasa via grant nag 5 - 8658 and by the national science foundation under grants phy99 - 07949 and ay97 - 31632 .
l. b. is a cottrell scholar of the research corporation . | we show that the burning of a small mass fraction @xmath0 of @xmath1 in a neutron star ocean is thermally unstable at low accumulated masses when the ocean contains heavy ashes from the hydrogen burning rapid proton ( rp ) process .
the key to early unstable ignition is the decreased thermal conductivity of a heavy element ocean .
the instability requires accretion rates , @xmath2 , in excess of one - tenth the eddington limit when @xmath3 .
lower @xmath2 s will stably burn a small mass fraction of @xmath1 .
the unstable flashes release @xmath4 over hours to days and are likely the cause of the recently discovered large type i bursts ( so - called `` superbursts '' ) from six galactic low - mass x - ray binaries .
in addition to explaining the energetics , recurrence times and durations of the superbursts , these mixed @xmath1flashes also have an @xmath2 dependence of unstable burning similar to that observed . though the instability is present at accretion rates
@xmath5 , the flashes provide less of a contrast with the accretion luminosity there , thus explaining why most detections are made at @xmath6 . future comparisons of time dependent theoretical calculations with observations will provide new insights on the rp - process . |
standard accretion disc theory ( * ? ? ?
* ; * ? ? ?
* hereafter ss73 and nt73 , respectively ) predicts a disc s effective temperature as a function of radius , from which its half - light radius as a function of wavelength can be calculated .
the predicted disc temperature profile is @xmath1 at @xmath2 , where @xmath3 with @xmath4 the black hole mass and @xmath5 the quasar s bolometric luminosity relative to its eddington luminosity .
the corresponding prediction for the variation of the half - light radius with wavelength is @xmath6 for a broad range of wavelengths .
quasar accretion disc half - light radii can be inferred using observations of gravitationally lensed quasars ( see , e.g. , * ? ? ?
* ) . microlensing by stars in the lensing galaxy causes the flux ratios of the lensed images to deviate from the predictions of models with smooth mass distributions .
the amplitude of these deviations depends on the size of the emission region at the wavelength being observed relative to the typical einstein radius of the lensing stars .
observationally , it has been found that quasar accretion discs have _ microlensing sizes _ ( inferred from microlensing observations ) which are larger than the _ theory sizes _
@xcite predicted from theory given each quasar s estimated black hole mass and assuming accretion at the eddington rate , 10% accretion efficiency and an average disc inclination to the line of sight of 60@xmath7 .
in the three studies to date with sample sizes of ten or more objects , the microlensing sizes range from a factor of @xmath8 higher on average at mean rest - frame wavelength 2660 ( * ? ? ?
* hereafter m10 ) to a factor of @xmath9 @xcite to @xmath10 ( * ? ? ?
* hereafter b11 ) higher on average at mean rest - frame wavelength 1736 .
the above results show that microlensing sizes are larger than theory sizes at only marginal significance ( assuming accretion at the eddington limit , which may be an overestimate ; see @xcite ) .
however , microlensing sizes are significantly larger than quasar _ flux sizes _ ( m10 ) .
flux sizes are found by determining how large a standard disc would need to be to generate the specific luminosity corresponding to the observed magnification - corrected flux of the quasar at a given wavelength .
m10 find that microlensing sizes are @xmath11 times larger than flux sizes .
similar results have been obtained in studies of individual objects ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
as for the wavelength dependence of quasar half - light radii , the microlensing sizes in the study of b11 are very weakly dependent on wavelength in ten of eleven cases studied , implying in the simplest case a temperature profile steeply decreasing with radius at temperatures that generate ultraviolet emission . on average , b11 finds @xmath12 with @xmath13 ( quoting random and systematic errors ) , where @xmath14 is predicted by theory .
the results of b11 contrast with studies of individual objects which have generally found results consistent with the theoretical prediction : @xmath15 for he 1104@xmath161805 @xcite ; @xmath17 for mg 0414 + 0534 @xcite , as compared to @xmath18 for that object in b11 ; @xmath19 for he 0435@xmath161223 @xcite , as compared to @xmath20 for that object in b11 ; @xmath21 for q 2237 + 0305 ; @xmath22 for sbs 0909 + 532 @xcite ; and @xmath23 at 94% confidence for sdss j0924 + 0219 @xcite , as compared to @xmath24 for that object in b11 .
further studies of microlensed quasars are clearly needed @xcite .
meanwhile , investigation of possible explanations for unexpectedly large sizes and nonstandard temperature profiles in quasar accretion discs is warranted .
to summarize : there is marginal evidence that quasar accretion discs are larger than their theory sizes ( though the evidence strengthens if quasars are not typically accreting at the eddington limit ) , considerable evidence that they are larger than their flux sizes , and possible evidence that their half - light radii have a different wavelength dependence than that predicted by theory . as discussed in m10 , a complication to the latter issue is that reconciling the three size measurements may require _ flatter _ temperature profiles ( @xmath25 ) and thus a steeper size - wavelength relation ( @xmath26 ) .
the two constraints are not necessarily contradictory ; a nonmonotonic temperature profile can be locally steep yet effectively flat overall .
if the correct temperature profile and physical parameters are used , theory sizes and flux sizes should agree with each other and with the microlensing sizes and the observed wavelength dependence of the sizes should match that predicted by the temperature profile .
other options discussed in m10 to help explain the three discrepant size measurements include low accretion efficiency for unobscured quasars or increasing the apparent disc size through contamination by line emission arising on larger physical scales or through scattering ( * ? ? ?
* see also ) .
for example , a large fraction of the disc continuum could be intercepted by a strongly warped disc ( e.g. , * ? ? ? * ; * ? ? ? * ) . in that case ,
the half - light radius will increase with a wavelength dependence related to the disc albedo , as radiation not scattered by the disc will be absorbed and reradiated at the local equilibrium temperature ( which will be lower than the characteristic temperature of the continuum emission from smaller radii ) .
disc sizes also increase with eddington ratio , so super - eddington accretion may also help explain discrepant size measurements .
furthermore , @xcite have suggested that the steep temperature profile found by b11 arises from the formation of a scattering photosphere in gas outflowing from the inner regions of an accretion disc undergoing super - eddington accretion ( see also * ? ? ?
* ; * ? ? ?
the small size of the x - ray emitting regions in quasars ( e.g. , * ? ? ? * ) could be explained in their model if the outflowing gas forms a compton - thick funnel - shaped wind and photosphere , so that quasars are only visible within the cone of the funnel because the funnel obscures both the x - ray and uv continuum source regions when viewed from other angles . another possible explanation for the temperature profiles found by b11 may arise from changes to an accretion disc resulting from misalignment of its initial angular momentum vector and the angular momentum vector of a spinning black hole .
any material in the vicinity of a massive spinning object experiences a precession induced by the frame dragging due to the object @xcite .
the result of this precession , known as the lense - thirring effect , on a misaligned disc around the object is that each individual ring in the disc will precess at different rate , warping the disc .
@xcite showed that the discs around spinning black holes evolve in such a way that in the final configuration the disc is warped , with the orbital angular momentum of the inner region either aligned or anti - aligned with the spin axis of the black hole while the disc s outer region remains unperturbed .
warped - disc dynamics has been the subject of extensive research ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in several recent papers , nixon and colleagues have suggested that the inner regions of inclined discs could evolve by tearing apart instead of smoothly warping .
* hereafter nkpf ) showed that for realistic parameters , the inner region of randomly oriented accretion discs around a spinning black hole will tear , due to the lense - thirring effect . if the angle between the outer disc and the black hole spin is in the range @xmath27 , the misalignment between the inner edge of the outer disc and the annulus ( of width @xmath28 ) torn off it will be @xmath29 after half a precession period .
the rotational velocities of those gas parcels will then be partially opposed , leading to cancellation of angular momentum and rapid infall ( see * ? ? ?
nkpf suggested that , given the smallness of the misalignment angle upper limit for avoiding disc tearing ( of order a few degrees ) , the effect should be common in accretion around the supermassive black holes that power quasars . when gas in a precessing annulus torn from a disc intersects with gas in an annulus still connected to the disc , the intersecting gas will be collisionally heated .
high - temperature gas in intersection regions at relatively large distances from the central black hole can in principle help explain the large , relatively wavelength - independent half - light radii inferred for quasar accretion discs . in addition , the wide range of behaviors possible in the torn disc model as a function of the initial disc - bh spin misalignment angle could in principle help explain the wide range of wavelength dependences in quasar accretion disc half - light radii @xcite . in this paper
we examine whether the temperature profiles of torn and other non - standard discs can simultaneously reproduce the observed spectral energy distributions and inferred large half - light radii of quasar accretion discs . in
[ toy ] we outline the initial torn - disc model used in our study . in [ tprof ]
we discuss the temperature profiles and seds for selected torn discs . in [ models ] we explore a small set of additional temperature profiles .
we discuss our results in
[ discuss ] .
where necessary , we assume a flat universe with @xmath30 , @xmath31 , and @xmath32=70kms@xmath33mpc@xmath33 .
our initial model begins with accreting material forming a keplerian disc at large radii from the black hole .
the disc s spin axis is oriented at a random angle @xmath34 relative to the black hole spin axis
. the disc spreads inwards and different annuli in it precess at different rates due to the lense - thirring effect , so the disc warps as it spreads inwards .
the disc eventually reaches a radius where viscous dissipation can not overcome the shear between adjacent annuli imposed by the lense - thirring effect .
the innermost annulus of the disc then tears away from the rest of the disc and precesses independently .
this annulus has width @xmath28 , where @xmath35 is the local disc height .
there will be two contact points between the precessing annulus and the next innermost annulus ( which forms the inner edge of the outer disc ) . at these points gas from each annulus on intersecting orbits is likely to shock , converting some orbital motion to heat
. the post - shock gas will be unable to remain in circular orbits at the intersection radius .
the original nkpf picture consists of gas intersecting at angles up to @xmath36 .
the disc tears at a break radius @xmath37 given by eq . 7 and eq .
8 of nkpf : @xmath38 @xmath39 where @xmath40 instead of @xmath35 appears in the above equations because we use @xmath35 to refer to the disc scale height , whereas nkpf use it to refer to the disc thickness .
gas which precesses and intersects at @xmath37 loses angular momentum , meaning that it can recircularize its orbit only at some circularization radius @xmath41 .
@xcite give an approximate expression for @xmath42 ( their eq .
3 ) , but the exact expression for the circularization radius of an annulus of width @xmath35 centered at @xmath37@xmath16@xmath43 and an annulus of width @xmath35 centered at @xmath37@xmath44@xmath43 , intersecting at an angle @xmath36 is @xmath45 note that the above is the minimum possible @xmath42 ; if gas intersects at angle @xmath46 , it will circularize at an intermediate radius @xmath47 . when an annulus torn off a disc oriented at angle @xmath48 to the black hole spin precesses , its intersection angle with the adjacent annulus that forms the inner edge of the outer disc will cycle sinusoidally between 0 and 2@xmath48 .
the intersection angle between these annuli as a function of time , @xmath49 , is given by @xmath50 where the precession period is @xmath51 , with @xmath52 the lense - thirring precession frequency in degrees per unit time .
at each intersection , each annulus approaches at an angle @xmath53 to their mutual bisector .
we assume that the kinetic energy from motion transverse to the bisector is converted into thermal energy .
this kinetic energy is @xmath54 ^ 2 $ ] , where @xmath55 is the number of particles ( of mean mass @xmath56 ) in each annulus and @xmath57 is the keplerian velocity at the intersection radius .
if all this transverse kinetic energy is converted to heat , the resulting thermal energy is distributed among the same @xmath58 particles : @xmath59 , where @xmath60 is the temperature increase of the gas .
thus , for gas intersecting at angle @xmath61 at radius @xmath62 , @xmath63 ^ 2 \sin^2\frac{\eta}{2}.\ ] ] this @xmath60 forms an additional term for the surface temperature @xmath64 in a region of width @xmath65 wherever a precessing annulus is present .
if gas is shock - heated to a temperature higher than the compton temperature of the ambient radiation field , the gas will cool by transferring its thermal energy to the radiation field through inverse compton scattering .
the compton temperature of a quasar radiation field is @xmath66 k ( eq .
25 and 26 of * ? ? ?
we therefore cap the value of @xmath64 at @xmath67 k.
the temperature profile and sed of a torn disc ultimately depend on the disc / bh spin misalignment angle @xmath48 , the dimensionless bh spin parameter @xmath68 ( @xmath69 ) , the dimensionless viscosity parameter @xmath70 , the disc height to radial distance ratio @xmath71 , and the black hole mass @xmath4 and normalized mass accretion rate @xmath72 .
b11 write this latter quantity as @xmath73 , where the eddington ratio is @xmath74 and the accretion efficiency is @xmath75 .
b11 assumed @xmath76 and @xmath77 , whereas in all cases we assume a schwarzschild black hole ( @xmath78 ) . to match the value of @xmath73 used by b11 , we assume @xmath79
. we also assume @xmath80 ( so that the disc height at any radius is @xmath81 ) , @xmath82 , and @xmath83 . in our illustrative calculations below , we adopt a ss73 temperature profile ( valid only for @xmath84 ) even for the case of @xmath85 , as our arguments are insensitive to the differences between ss73 and nt73 seds .
equations [ rb ] , [ rcirc ] and [ deltat ] generate a family of temperature profiles as a function of @xmath48 which consist of : @xmath86 a normal ss73 disc ( @xmath87 ) at @xmath88 , @xmath86 a gap at @xmath89 , @xmath86 a shock - heated region ( @xmath90 ) at radii @xmath91 , @xmath86 and a normal ss73 disc again at @xmath92 . to find the temperature profile for the scenario put forth by nkpf , in which the precessing disc intersects with its neighboring annulus over half to one precession periods , for a given @xmath48 we use eq .
[ deltat ] to calculate @xmath93 for @xmath94 ( @xmath95 ) or @xmath96 ( @xmath97 ) , as the average of @xmath61 over one precession period is about half the maximum @xmath61 .
we then calculate spectral energy distributions ( seds ) and half - light radii using the @xmath98 values from the regions @xmath99 , @xmath100 and @xmath101 .
figure [ fig:1 ] shows an ss73 disc sed and the seds resulting from these shock - heated disc temperature profiles .
the seds are much bluer and more x - ray luminous than observed quasar seds .
( note that the ss73 disc model is known to be a poor fit to observations at @xmath102 , and that at x - ray wavelengths an additional power - law component is needed in addition to thermal disc emission .
those wavelengths are plotted here to examine if an sed overpredicts the short - wavelength emission , not in expectation of a good fit at short wavelengths from any sed , even a baseline ss73 one . )
we have assumed that an entire annulus at @xmath37 is at the post - shock temperature , but it may be more accurate to assume that only the two intersection regions of surface area @xmath103 on each face of the disc are at that temperature .
figure [ fig:2 ] shows the seds in that case ; they are still much bluer and more x - ray luminous than observed quasar seds .
we discuss the half - light radii results for this and the next model in [ toyresults ] .
[ cols="<,<,^,^,^",options="header " , ] + see appendix [ sizes ] for a description of @xmath104 , and
[ results ] for details on other quantities presented herein .
[ tab : chi2min ] for each quasar we calculate a temperature profile for all allowable combinations of the specified parameter values given for each model above . for each model
ss73 , forming disc , and magnetically restrained for values of @xmath105 from 0 to 0.81 we find the minimum @xmath106 value over all tested values of @xmath5 , @xmath107 , @xmath108 and @xmath109 .
we plot the resulting half - light radii vs. wavelength , seds , and @xmath110 values for each quasar individually in figure [ fig : newer ] ; see the figure caption for details .
we present the @xmath110 values and best - fit parameter values for the ss73 , forming disc , and best - fit magnetically restrained models in table [ tab : chi2min ] .
for ease of comparison with other size values , table [ tab:2 ] presents @xmath111-band flux sizes @xmath112 for both the @xmath113 case and the best - fit magnetically restrained models , calculated following appendix [ sizes ] .
the first point to notice about the best fits is that the best - fit ss73 and forming disc models usually have large @xmath5 ( 0.7 @xmath114 @xmath5@xmath114 1 in nine out of eleven cases ) .
larger @xmath5 results in a larger accretion disc , making it easier to match the large inferred half - light radii .
this effect is less pronounced in the best - fit magnetically restrained models ; they have @xmath5=1 in six of eleven cases , but three of those are poor fits .
( however , we have only considered face - on discs in our modelling .
an inclined disc will have half - light radii smaller than it does when face - on , and given that quasar discs will have a nonzero average inclination angle , the true best - fit @xmath5 of our objects in all models will , statistically , be somewhat larger than the values quoted . )
the second point to notice is that the forming disc model is never a significantly better fit than the ss73 model .
the third point to notice is that in eight of eleven cases , the magnetically restrained models can provide a fit which is better than the ss73 or forming disc fit and is also a statistically acceptable fit at the 99.73% confidence level if the full errors @xmath115 from b11 are used in the @xmath106 calculations .
( if the @xmath116 values are used instead , only in five of eleven cases are statistically acceptable fits found . )
two of the cases without acceptable fits using @xmath115 , he 0230 and wfi j2033 , have unusual wavelength dependences of their half - light radii and are discussed below .
the third case , sdss j1138 , has the lowest @xmath4 and the largest ratio of @xmath117 to @xmath4 in our sample .
the ss73 and forming disc models can match the luminosity of sdss j1138 , but not its large half - light radii , even with @xmath5=1 .
the magnetically restrained models with @xmath5=1 can better match the shape of its half - light radius vs. wavelength curve , but still can not match their large values , and is a worse fit to the observed luminosity .
the large @xmath110 in this object could be reduced if future observations reveal that its @xmath4 is larger than the value we adopt , which is quite plausible given the considerable uncertainties on @xmath4 estimates ( see , e.g. , the comprehensive review of * ? ? ?
the same might be true of he 0230 and wfi j2033 , which in addition to the unusual wavelength dependences of their half - light radii have the next largest ratios of @xmath117 to @xmath4 in our sample .
however , table [ tab:2 ] shows that the flux sizes of the best - fit magnetically restrained models are still about an order of magnitude smaller , on average , than the measured microlensing sizes .
( they are larger than the flux sizes of a @xmath113 disc by @xmath118 on average . )
the one exception is he 0230@xmath162130 , which has a best - fit magnetically restrained model with @xmath119 , corresponding to negligible thermal emission within the radius of the temperature spike . except in that case ,
our best - fit magnetically restrained models produce more flux than observed in these objects . incorporating flux sizes as a constraint in the fitting might shift the best - fit @xmath105 values to @xmath119 in many cases , as the @xmath106 minimum as a function of @xmath105 is quite broad . the best - fit ss73 and forming disc models tend to overestimate @xmath117 , while the best - fit magnetically restrained model tends to slightly underestimate it .
better constraints on the models might be obtainable by including multiwavelength seds for these quasars in the @xmath106 fits , rather than just the total @xmath117 . in the cases of he 0230 , pg 1115 and wfi j2033 , such constraints would considerably increase the @xmath106 of the best - fit magnetically restrained models ; those best - fit models have seds that peak closer to 10@xmath120 than 10@xmath121 .
it is not clear if a magnetically restrained model would still yield a low @xmath110 for pg 1115 if measurements of its multiwavelength
sed were available for consideration in the fitting .
there are some similarities among the eight acceptable magnetically restrained model fits .
those best fits have @xmath122 , corresponding to temperatures betweeen 20% and 60% of the ss73 value .
the best - fit @xmath109 is 12 or 18 , except for mg j0414 ( @xmath123 ) .
( note that mg j0414 is also consistent with an ss73 or forming disc . )
the best - fit @xmath108 is between 12000 k and 16000 k , again except for mg j0414 ( @xmath108=28000 k ) .
the best - fit @xmath107 is between 25000 k and 35000 k , except for mg j0414 ( @xmath107=60000 k ) and pg 1115 ( @xmath107=15000
the best - fit values of @xmath5 , however , range from 0.03 to 1 .
we can therefore characterize the best - fit temperature profile in seven of eleven cases as a magnetically restrained model with a temperature ( 40@xmath12420)% of the ss73 value within the radius where the ss73 disc reaches ( 14000@xmath1242000 ) k , at which radius it has a temperature spike reaching ( 30000@xmath1245000 ) k , which falls off as a power law @xmath125 .
two quasars in b11 ( he 0230 and wfi j2033 ) appear to have half - light radii that decrease with increasing wavelength , albeit with large uncertainties .
a temperature profile which increases with radius and then abruptly decreases can in principle reproduce the sign of this trend .
( conceptually , such a temperature profile might arise if the accretion luminosity is dominated by a shock at some radius @xmath37 , with cooling gas undergoing radial infall within @xmath37 . )
the two quasars above have maximum @xmath126 and @xmath127 , respectively , from @xmath128 to @xmath129 . however , a radially inverted temperature profile like that described above can not explain changes in @xmath130 greater than about @xmath131 . that @xmath132 corresponds to the case of a narrow temperature spike at @xmath133 which generates all the luminosity at @xmath128 , and a flat temperature profile within that radius which dominates the luminosity observed at @xmath129 , yielding @xmath134 , so that @xmath135 . a temperature profile increasing with radius within @xmath133
will produce an even smaller @xmath132 value .
a temperature profile decreasing with radius within @xmath133 before spiking to a high temperature at that radius will not match the observed smooth change in @xmath130 with wavelength ; such a temperature profile would yield @xmath130 values decreasing and then increasing with wavelength .
b11 note that the anomalous wavelength dependence of these two objects size estimates might also be produced by unusual microlensing caustic patterns such as those found at the center of an astroid caustic .
( the magnification increases away from the center of an astroid caustic , leading to larger flux anomalies for larger sources . )
that explanation for these objects half - light radii seems more likely after consideration of the difficulty in reproducing the results with actual disc temperature profiles .
we predict that further observations of these two objects will not show the anomalous size estimates , as long as the time baseline between old and new observations is sufficient for the unusual microlensing caustic pattern to have moved out of our line of sight to the continuum source .
+ + + + + + + + + + + +
we have tested several energetically plausible quasar accretion disc temperature profile models against the large quasar accretion disc half - light radii inferred by b11 from microlensing observations .
we have shown that the temperature profiles from our toy model of the torn accretion discs proposed by nkpf are not able to simultaneously match observed quasar seds and the microlensing sizes of b11 . in our toy model ,
torn discs yield precessing rings of gas orbiting at supersonic speeds which shock when they intersect and yield seds which are much bluer than observed quasar seds .
we have demonstrated that acceptable fits ( at the 3@xmath115 level ) to the half - light radii of b11 in other words , theory sizes matching the microlensing sizes can be found in eight of eleven cases for model discs with a temperature spike at some radius and lowered temperatures within that radius ( our magnetically restrained models ) .
acceptable fits can be found in only four of eleven cases for discs with normal ss73 temperature profiles and in only three of eleven cases for discs with normal temperature profiles inside a temperature spike ( our forming disc models ) .
the poor @xmath110 for both forming disc models and magnetically restrained models with large @xmath105 indicate that the shape of the temperature profile outside the temperature spike is not a major factor in obtaining an acceptable fit .
we find that the best - fit temperature profile in seven of eleven cases is a magnetically restrained model with a temperature ( 40@xmath12420)% of the ss73 value within the radius where the ss73 disc reaches ( 14000@xmath1242000 ) k , at which radius it has a temperature spike reaching ( 30000@xmath1245000 ) k , which falls off as a power law @xmath125 .
however , such profiles do not produce flux sizes as large as their theory sizes , by about an order of magnitude . constraining the magnetically restrained models to explicitly match the flux sizes
may produce better fits by reducing the best - fit @xmath105 values .
we note that the two quasars in b11 which have half - light radii that appear to decrease with increasing wavelength can not be fully explained by a radially increasing temperature profile with an abrupt cutoff .
they must instead be examples of unusual microlensing patterns ; e.g. , astroid caustics ( [ astroid ] ) .
future progress on the issue of quasar accretion disc temperature profiles will require additional and improved observations and modelling .
measurements of the wavelength dependence of half - light radii in lensed quasars could be expanded to larger samples and extended to shorter wavelengths using the hubble space telescope .
improved @xmath4 estimates for gravitationally lensed quasars would yield better constraints on models providing acceptable fits to half - light radii , as would multiwavelength sed measurements ( such as photometrically calibrated flux measurements of lens images at observed optical and infrared wavelengths ) .
one worthwhile improvement to the modelling would be to self - consistently use the luminosity profiles which have here been found to provide good fits to the half - light radii as input to simulations of the microlensing flux deviations .
currently the simulations use elliptical gaussians , whereas at high temperatures the best - fit profiles found here will resemble elliptical rings on the sky .
further exploration of model temperature profiles consistent with improved observations could then follow .
models with @xmath5@xmath1361 and super - eddington luminosities ( e.g. , * ? ? ?
* ) may be worth considering , as the best - fit models we find here tend to cluster at our imposed limit of @xmath5=1 .
also , each of our magnetically restrained models has an angular velocity which is a fixed fraction @xmath105 of the keplerian value , but a more general extension of the models of @xcite would involve a power - law dependence of the angular velocity on radius ( @xmath137 ) , arising from an assumed radial power - law dependence of the magnetic field strength .
eventually , such parametric modelling should be superceded by the use of temperature profiles from theoretical simulations of accretion discs .
if the results of b11 are borne out by further observations , a strong test of such simulations will be how naturally they can generate the temperature profiles allowed by observations . alternatively , if future observations rule out the possibility that quasar accretion discs have flat temperature profiles , inhomogeneous accretion discs @xcite offer a potential solution to the discrepancy between microlensing and flux sizes .
inhomogeneous discs simultaneously increase the radius within which emission at a given wavelength occurs ( increasing the microlensing size ) and decrease the area within that radius which emits at that wavelength ( increasing the flux size ) .
we thank c. kochanek for valuable comments .
pbh thanks nserc for research support for lsc and ew .
esn thanks nserc for an undergraduate student research award .
cjn acknowledges support provided by nasa through the einstein fellowship program , grant pf2 - 130098 .
44 natexlab#1#1 p. , shakura n. i. , 2012 , mnras , 427 , 1867 m. a. , czerny b. , lasota j. p. , szuszkiewicz e. , 1988 , astrophys . j. , 332 , 646 j. m. , petterson j. a. , 1975 , astrophys . j. lett . , 195 , l65 n. f. , floyd d. j. e. , webster r. l. , wyithe j. s. b. , 2008 , mnras , 391 , 1955 j. a. , pooley d. , rappaport s. , schechter p. l. , 2011 , astrophys . j. , 729 , 34 e. w. , shields g. a. , stevens a. c. , salviander s. , 2013 , astrophys . j. , 770 , 30 x. , kochanek c. s. , chartas g. , kozowski s. , morgan c. w. , garmire g. , agol e. , 2010 , astrophys . j. , 709 , 278 j. , agol e. , 2011 , apjl , 727 , l24 a. , courbin f. , meylan g. , agol e. , anguita t. , schmidt r. w. , wambsganss j. , 2008 , astron .
astrophys . , 490 , 933 d. j. e. , bate n. f. , webster r. l. , 2009 , mnras , 398 , 233 j. , mediavilla e. , muoz j. a. , kochanek c. s. , 2012 , astrophys .
j. , 751 , 106 a. r. , pringle j. e. , 2006 , mon . not .
r. astron .
soc . , 373 , l90 , 2007 , mon . not .
r. astron .
soc . , 377 , l25 j. a. , onken c. a. , kochanek c. s. , gould a. , weinberg d. h. , dietrich m. , cool r. , dey a. , eisenstein d. j. , jannuzi b. t. , le floch e. , stern d. , 2006 , astrophys .
j. , 648 , 128 s. , pringle j. e. , 1985 , mnras , 213 , 435 j. , thirring h. , 1918 , physikalische zeitschrift , 19 , 156 s. h. , ogilvie g. i. , pringle j. e. , 2002 , mnras , 337 , 706 e. , muoz j. a. , kochanek c. s. , guerras e. , acosta - pulido j. , falco e. , motta v. , arribas s. , manchado a. , mosquera a. , 2011 , astrophys .
j. , 730 , 16 c. w. , hainline l. j. , chen b. , tewes m. , kochanek c. s. , dai x. , kozlowski s. , blackburne j. a. , mosquera a. m. , chartas g. , courbin f. , meylan g. , 2012 , astrophys .
j. , 756 , 52 c. w. , kochanek c. s. , dai x. , morgan n. d. , falco e. e. , 2008 , astrophys .
j. , 689 , 755 c. w. , kochanek c. s. , morgan n. d. , falco e. e. , 2010 , astrophys .
j. , 712 , 1129 a. m. , kochanek c. s. , 2011 , astrophys .
j. , 738 , 96 a. m. , muoz j. a. , mediavilla e. , kochanek c. s. , 2011 , astrophys .
j. , 728 , 145 r. , igumenshchev i. v. , abramowicz m. a. , 2003 , publ .
jpn . , 55 , l69 s. , 2005 , mon . not .
r. astron .
soc . , 359 , 545 c. , king a. , price d. , frank j. , 2012 , astrophys . j. lett . , 757 , l24 c. j. , king a. r. , price d. j. , 2012 , mnras , 422 , 2547 i. d. , thorne k. s. , 1973 , in black holes ( les astres occlus ) , dewitt c. , dewitt b. s. , eds . , pp . 343450 g. i. , 1997 , mnras , 288 , 63 , 1999 ,
mnras , 304 , 557 j. c. b. , lin d. n. c. , 1995 , astrophys .
j. , 438 , 841 j. c. b. , pringle j. e. , 1983 , mnras , 202 , 1181 s. , kochanek c. s. , 2010 , astrophys .
j. , 712 , 668 s. , morgan n. , kochanek c. s. , 2008 , astrophys .
j. , 673 , 34 d. , blackburne j. a. , rappaport s. , schechter p. l. , 2007 , astrophys .
j. , 661 , 19 j. e. , 1992 , mnras , 258 , 811 g. t. , lacy m. , storrie - lombardi l. j. , hall p. b. , gallagher s. c. , hines d. c. , fan x. , papovich c. , vanden berk d. e. , trammell g. b. , schneider d. p. , vestergaard m. , york d. g. , jester s. , anderson s. f. , budavri t. , szalay a. s. , 2006 , apjs , 166 , 470 s. y. , ostriker j. p. , sunyaev r. a. , 2004 , mnras , 347 , 144 a. e. , stewart g. c. , mateos s. , 2012 , mon . not .
r. astron .
, 423 , 2633 n. i. , sunyaev r. a. , 1973 , a&a , 24 , 337 y. , 2013 , bulletin of the astronomical society of india , 41 , 61 a. d. , roberts t. p. , middleton m. j. , 2013 , arxiv e - prints s. , davis s. w. , 2013 , arxiv e - prints j. , 2006 , in saas - fee advanced course 33 : gravitational lensing : strong , weak and micro , meylan g. , jetzer p. , north p. , schneider p. , kochanek c. s. , wambsganss j. , eds . , pp .
the energy per second emitted from the disc in its forming region ( @xmath138 ) must satisfy @xmath139 ( see [ fd ] ) . considering both sides of the disc , @xmath140 ^ 4 da \leq \frac{gm_{bh}\dot{m}}{2r_b}\ ] ] writing @xmath141 at @xmath142 , with @xmath143
, we have @xmath144 @xmath145 making the substitution @xmath146 , @xmath147 @xmath148 @xmath149 \frac{gm_{bh}\dot{m}}{2\pi r_b^3}\ ] ]
in our magnetically restrained disc models , gas accreting through a keplerian disc at large radii is presumed to reduce its orbital velocity to a fraction @xmath150 of its keplerian value by the time it reaches radius @xmath37 .
the disc thus has a fraction @xmath151 of the kinetic energy of the accreting gas at @xmath37 available to emit as heat at @xmath138 , in addition to the normal disc emission ( @xmath152 ) at @xmath138 .
consider a disc obeying @xmath141 from @xmath153 and a normal ss73 @xmath154 profile at @xmath155 , where @xmath156 so that @xmath157 and @xmath158 .
we have a limit on the excess energy emitted by the disc , which is equal to the total energy emitted by the disc minus the energy that would have been emitted by an ss73 disc between those radii . again considering both sides of the disc ,
@xmath159 expanding @xmath160 and making the substitution @xmath146 , @xmath161 \nonumber\\ \leq ( 1-f^2)\frac{gm_{bh}\dot{m}}{2r_b } \end{aligned}\ ] ] which evaluates to @xmath162 \nonumber\\ \leq ( 1-f^2)\frac{gm_{bh}\dot{m}}{2r_b } \end{aligned}\ ] ] defining @xmath163 , we have : @xmath164\nonumber\\ \leq ( 1-f^2)\frac{gm_{bh}\dot{m}}{2r_b } \end{aligned}\ ] ] examining the above equation , we see that for it to be satisfied in the limit @xmath165 , where no excess energy is available , then we must have @xmath166 , or @xmath167 , as expected .
thus , the constraint on our magnetically restrained disc models is given by : @xmath168 \end{aligned}\ ] ] where @xmath163 .
the _ flux size _ of a quasar accretion disc ( m10 ) is found by using the quasar s observed flux to calculate its luminosity and determining the size a standard ss73 disc would need to be to generate that luminosity . given a lensed quasar s demagnified apparent magnitude @xmath111 , the flux size in the @xmath111 band ( @xmath169 ) for the accretion disc of that quasar , located at angular size distance @xmath170 ( in units of the hubble radius @xmath171 ) with inclination @xmath172 and with a temperature profile @xmath173 , is given by : @xmath174 where @xmath175 kms@xmath33mpc@xmath33 and where @xmath176^{-1 } \end{aligned}\ ] ] is the integral over the disc surface brightness , normalized to unity for @xmath113 . the @xmath177 integral
is obtained as follows . for observations at rest wavelength @xmath178 ,
we can write @xmath179 , with @xmath180 and @xmath181 where @xmath182 . we then have @xmath183^{-1 } \\
\propto \int_{0}^{\infty } r ~dr ~[\exp(r^\beta / r_\lambda^\beta)-1]^{-1 } \\
\propto r_\lambda^2 \int_{0}^{\infty } u ~du ~[\exp(u^\beta)-1]^{-1}\end{aligned}\ ] ] where we have defined @xmath184 . in the case of our magnetically restrained models ( [ mr ] ) , where the disc temperature profile has three parts
, we can generalize @xmath177 to @xmath185 : @xmath186 \nonumber\end{aligned}\ ] ] where @xmath184 as before . in the equation
above , we define @xmath187 and @xmath188 , with @xmath37 and @xmath189 as defined in
[ mr ] , and we define @xmath190 . the constant of proportionality is the same for @xmath177 and @xmath185 , so the ratio of those two quantities gives the ratio of the flux sizes for an ss73 model disc and a magnetically arrested one . | microlensing observations indicate that quasar accretion discs have half - light radii larger than expected from standard theoretical predictions based on quasar fluxes or black hole masses .
blackburne and colleagues have also found a very weak wavelength dependence of these half - light radii .
we consider disc temperature profile models that might match these observations .
nixon and colleagues have suggested that misaligned accretion discs around spinning black holes will be disrupted at radii small enough for the lense - thirring torque to overcome the disc s viscous torque .
gas in precessing annuli torn off a disc will spread radially and intersect with the remaining disc , heating the disc at potentially large radii .
however , if the intersection occurs at an angle of more than a degree or so , highly supersonic collisions will shock - heat the gas to a compton temperature of @xmath0 k , and the spectral energy distributions ( seds ) of discs with such shock - heated regions are poor fits to observations of quasar seds . torn discs where heating occurs in intermittent weak shocks that occur whenever the intersection angle reaches a tenth of a degree pose less of a conflict with observations , but do not have significantly larger half - light radii than standard discs . our toy model for torn accretion discs
is therefore unable to simultaneously match observed seds and large half - light radii in quasars .
we also study two phenomenological disc temperature profile models .
we find that discs with a temperature spike at relatively large radii and lowered temperatures at radii inside the spike yield improved and acceptable fits to microlensing sizes in most cases . such temperature profiles could in principle occur in sub - keplerian discs partially supported by magnetic pressure .
however , such discs overpredict the fluxes from quasars studied with microlensing except in the limit of negligible continuum emission from radii inside the temperature spike .
[ firstpage ] galaxies : nuclei - quasars : general - quasars : absorption lines |
in order to keep track of associativity data for our weakly monoidal categories , we will need a categorical equivalent of the associahedron @xmath19 . to begin with we formalize the notion of a parenthesized word : a _ parenthesized word _ @xmath20is a finite linear order @xmath21 together with a ( possibly empty ) collection of closed intervals @xmath22\}$ ] subject to the following requirements . *
the cardinality of each @xmath23 is at least 2 and is strictly smaller than the cardinality of @xmath21 .
* for any @xmath24 , either @xmath25 , @xmath26 or @xmath27 a parenthesized word @xmath20 can be converted into a parenthesized string of characters by putting as many left parentheses in front of an element @xmath28 as @xmath29 is an initial element of some @xmath30 and as many right parentheses after an element @xmath31 as @xmath32 is a final element in some @xmath30 , and concatenating the resulting characters .
for instance @xmath33,[x_2,x_4],[x_5,x_6]\}\right\}\mapsto x_1((x_2x_3x_4)(x_5x_6)).\ ] ] it is clear that @xmath20 can be recovered from the parenthesized string and we will often find it convenient to represent @xmath20 in this way . in most cases we will use the standard linear orders @xmath34 . in some induction arguments however we will need to consider subintervals of the @xmath35 .
[ kmdefi ] we define @xmath36 to be the poset of parenthesized words on the linear order @xmath35 , where @xmath37 iff @xmath38 .
the minimal elements in this order are called the _ fully parenthesized _ words of length @xmath39 . in the degenerate cases @xmath40 and @xmath41 , the poset @xmath42 consists of the single parenthesized word @xmath43 , and @xmath44 consists of the single parenthesized word @xmath45 .
the string @xmath46 is the terminal object in @xmath36 .
as noted above , sometimes it will be convenient to use some other linear order @xmath47 of the same cardinality @xmath39 . in that case
the unique order isomorphism between @xmath47 and @xmath35 specifies a canonical isomorphism between @xmath36 and the corresponding poset of parenthesized words on @xmath47 .
the poset @xmath48 : @xmath49\ar[dr]\ar[ddd]\\ & ( x_1x_2)x_3x_4\ar[ddr ] & & x_1x_2(x_3x_4)\ar[ddl]\\ { ( ( x_1x_2)x_3)x_4}\save[]+<0pt,-8pt>*{}="a"\restore\ar[ur]\ar[drr ] & & & & x_1(x_2(x_3x_4))\save[]+<0pt,-8pt>*{}="f"\restore\ar[ul]\ar[dll]\\ & & x_1x_2x_3x_4\\ { } \save[]+<38pt,10pt>*{(x_1x_2x_3)x_4}="b"\ar[urr]\restore & & & & { } \save[]+<-38pt,10pt>*{x_1(x_2x_3x_4)}="d"\ar[ull]\restore\\ \\ & \save[]+<0pt,8pt>*{}="c"\restore(x_1(x_2x_3))x_4\ar[r]\ar[uuur ] & x_1(x_2x_3)x_4\ar[uuu ] & \save[]+<0pt,8pt>*{}="e"\restore x_1((x_2x_3)x_4)\ar[l]\ar[uuul ] \ar " a";"b " \ar " c";"b " \ar " e";"d " \ar " f";"d " } \ ] ] the @xmath39-th associahedron is defined to be the polytope which has one vertex for every fully parenthesized word of length @xmath39 .
two vertices @xmath50 and @xmath51 are on the same @xmath52-dimensional face if they share at least @xmath53 parentheses , i.e. @xmath54 has cardinality at least @xmath53 .
thus our poset @xmath36 is exactly the face poset of the @xmath39-th associahedron , and so the geometric realization of the nerve of @xmath36 is simply the barycentric subdivision of the @xmath39-th associahedron .
note that we are using a fraktur font to distinguish the poset @xmath36 from the associahedron @xmath19 which is the geometric realization of its nerve ( as a topological space ) .
the following lemma will prove to be surprisingly useful : [ surprise ] let @xmath55 in @xmath56 .
then the subposet @xmath57=\{(w_k , p'')\in\sck_k | ( w_k , p)\le(w_k , p'')\le(w_k , p')\}\ ] ] is isomorphic to the poset @xmath58 where @xmath59 is the poset @xmath60 and @xmath39 is the number of parentheses in @xmath61 which are not in @xmath62 . in other words , the factorizations of a fixed morphism in @xmath56 form a commutative cubical diagram .
we can uniquely associate to each element @xmath63 in @xmath64 $ ] a characteristic function on the set of parentheses in @xmath61 which are not in @xmath62 by giving the value 1 to each parenthesis which occurs in @xmath63 and 0 to any which do not so occur .
but such a characteristic function is evidently the same thing as an object of @xmath58 and it is clear that order relations match .
if we take @xmath65 , then geometrically this gives a decomposition of the associahedra into cubes .
the decomposition of @xmath11 into 5 squares looks like this : @xmath49\ar[dr]\\ & ( x_1x_2)x_3x_4\ar[ddr ] & & x_1x_2(x_3x_4)\ar[ddl]\\ { ( ( x_1x_2)x_3)x_4}\save[]+<0pt,-8pt>*{}="a"\restore\ar[ur ] & & & & x_1(x_2(x_3x_4))\save[]+<0pt,-8pt>*{}="f"\restore\ar[ul]\\ & & x_1x_2x_3x_4\\ { } \save[]+<38pt,10pt>*{(x_1x_2x_3)x_4}="b"\ar[urr]\restore & & & & { } \save[]+<-38pt,10pt>*{x_1(x_2x_3x_4)}="d"\ar[ull]\restore\\ \\ & \save[]+<0pt,8pt>*{}="c"\restore(x_1(x_2x_3))x_4\ar[r ] & x_1(x_2x_3)x_4\ar[uuu ] & \save[]+<0pt,8pt>*{}="e"\restore x_1((x_2x_3)x_4)\ar[l ] \ar " a";"b " \ar " c";"b " \ar " e";"d " \ar " f";"d " } \ ] ] the decomposition of @xmath66 into 14 cubes can be found here : http://arxiv.org / src/1005.3979v4/anc / cubical.flv .
there is a one - to - one correspondence between parenthesized words and _ stable rooted trees_. briefly these are planar rooted trees where each node has at least two input edges .
we refer to @xcite for a formal definition .
the correspondence is given by labelling the leaves of such a tree with the labels @xmath67 in left to right order .
[ in the degenerate cases @xmath68 and @xmath69 , the identity @xmath70 corresponds to the tree with a single edge and no nodes and @xmath71 corresponds to the empty tree with no edges and no nodes .
] then for each node of the tree , except for the bottom root node , one takes the set of labels sitting over that node as one of the intervals @xmath30 in the collection @xmath72 , thus giving us a parenthesized word @xmath73 .
for example , here are all of the parenthesized words on the linear order @xmath74 and their corresponding stable rooted trees : the following trees@xmath75 & & \ar@{-}[dl]&&\ar@{-}[dr ] & & \ar@{-}[dl]\\ & { \scriptscriptstyle\bullet}\ar@{-}[ddrr]&&&&{\scriptscriptstyle\bullet}\ar@{-}[ddll]\\ \\ & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & & & { } } $ ] @xmath75 & & \ar@{-}[dl]&&&\ar@{-}[ddll ] & & & \ar@{-}[dddlll]\\ & { \scriptscriptstyle\bullet}\ar@{-}[drr]\\ & & & { \scriptscriptstyle\bullet}\ar@{-}[drr]\\ & & & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & & & } $ ] @xmath76 & & & \ar@{-}[dr]&&\ar@{-}[dl ] & & & \ar@{-}[dddllll]\\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[dl]\\ & & & { \scriptscriptstyle\bullet}\ar@{-}[dr]\\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & & } $ ] @xmath77 & & & \ar@{-}[dr]&&\ar@{-}[dl ] & & & \ar@{-}[ddlll]\\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[dr]\\ & & & & & { \scriptscriptstyle\bullet}\ar@{-}[dl]\\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & & } $ ] @xmath77 & & & \ar@{-}[ddrr]&&\ar@{-}[dr ] & & & \ar@{-}[dll]\\ & & & & & & { \scriptscriptstyle\bullet}\ar@{-}[dl]\\ & & & & & { \scriptscriptstyle\bullet}\ar@{-}[dl]\\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & & } $ ] @xmath78&\ar@{-}[dd]&\ar@{-}[ddl ] & & & \ar@{-}[dddlll]\\ \\ & { \scriptscriptstyle\bullet}\ar@{-}[dr]\\ & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & } $ ] @xmath79&&&\ar@{-}[ddr]&\ar@{-}[dd]&\ar@{-}[ddl]\\ \\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[dl]\\ & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & } $ ] @xmath75&&\ar@{-}[dl]&\ar@{-}[dddl]&&&\ar@{-}[dddllll]\\ & { \scriptscriptstyle\bullet}\ar@{-}[ddr]\\ \\ & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & } $ ] @xmath79&&\ar@{-}[dr]&&\ar@{-}[dl]&&\ar@{-}[dddlll]\\ & & & { \scriptscriptstyle\bullet}\ar@{-}[dd]\\ \\ & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & } $ ] @xmath77&&&\ar@{-}[dddr]&&\ar@{-}[dr]&&\ar@{-}[dl]\\ & & & & & & { \scriptscriptstyle\bullet}\ar@{-}[ddll]\\ \\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & & } $ ] @xmath79&&\ar@{-}[dddr]&&\ar@{-}[dddl]&&\ar@{-}[dddlll]\\ \\ \\ & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & } $ ] represent @xmath80 , @xmath81 , @xmath82 , @xmath83 , @xmath84 , @xmath85 , @xmath86 , @xmath87 , @xmath88 , @xmath89 , @xmath90 respectively . the poset structure on @xmath91 of definition [ kmdefi ] can be described in terms of trees as follows : @xmath92 if @xmath93 can be obtained from @xmath94 by shrinking some of the edges of @xmath94 .
it is of course more convenient to use parenthesized words when describing @xmath91 as a poset .
however the language of trees is more convenient to describe the operad structure on the @xmath91 .
the @xmath95 form an operad in @xmath18 , the category of small categories . given stable rooted trees @xmath96 and
@xmath97 for @xmath98 , we obtain a new tree by grafting the root of @xmath99 to the @xmath100-th leaf of @xmath101 . in terms of parenthesized words
we are substituting the word for @xmath99 in place of the @xmath100-th character of the word for @xmath101 , and reindexing to insure that all characters in the resulting word are distinct .
this only makes sense if @xmath102 and all @xmath103 .
if @xmath104 , we define the composed tree to be @xmath105 .
if @xmath106 , then we leave the @xmath100-th leaf of @xmath101 unchanged .
if @xmath107 , then we delete the @xmath100-th leaf of @xmath101 .
if this leaves only one input edge for the node below , we delete that node as well .
if it leaves no input edges for the node below , we delete both that node and the edge below .
we apply this algorithm recursively : if the next node below receives only one input edge or no input edges , we delete that node or that node together with the edge below , and so on . in the special case
when @xmath108 or @xmath109 , the resulting degenerate trees are defined to be @xmath110 or @xmath111 respectively .
this process is is clearly functorial in each of @xmath112 , and so we obtain a functor @xmath113 these functors define a categorical operad @xmath114 .
the associahedral operad @xmath115 has an operadic filtration @xmath116 where @xmath117 is the subposet of @xmath118 consisting of trees where each node has input valence @xmath119 ( i.e has at most @xmath4 incoming edges ) .
we define @xmath120 .
we note for future reference that if an element @xmath20 of @xmath56 lies in filtration @xmath4 and @xmath121 then @xmath122 also lies in filtration @xmath4 .
this follows from our description above of the poset structure in terms of trees . [ connectivity ]
the poset @xmath117 is the face poset of a subcomplex of the ( unsubdivided ) associahedron @xmath123 .
this subcomplex contains all cells of @xmath123 of dimension @xmath124 .
consequently the nerve of @xmath117 is @xmath125-connected . in particular
if @xmath126 the nerve of @xmath117 is simply connected .
we note that if an element @xmath20 of @xmath118 is in @xmath117 and @xmath121 then @xmath122 is also contained in @xmath117 , since the tree representing @xmath20 is obtained from the tree representing @xmath122 by shrinking internal edges .
it follows that @xmath117 is the face poset of a subcomplex of @xmath123 .
now the vertices of @xmath123 are parametrized by the elements of @xmath127 , which are represented by binary trees .
it follows that the cells of @xmath123 of dimension @xmath128 are obtained by shrinking @xmath128 internal edges of a binary tree .
it easily follows that the dimension of the cell parametrized by a given tree is the sum over all nodes of the incoming valence of that node minus 2 .
thus the maximal possible incoming valence of a node in a tree parametrizing a cell of dimension @xmath128 is @xmath129 .
hence the subcomplex of @xmath123 parametrized by @xmath117 contains all cells of @xmath123 of dimension @xmath124 .
now the nerve of @xmath117 is the barycentric subdivision of this subcomplex of @xmath123 .
moreover @xmath123 is obtained from this subcomplex by adding cells of dimensions @xmath130 . since @xmath123 is contractible
, it follows that the complex and hence the nerve of @xmath117 is @xmath125-connected .
@xmath117 is generally larger than the face poset of the @xmath131-skeleton of @xmath123 .
for instance @xmath132 is the face poset of the subcomplex of the 3-dimensional associahedron @xmath66 consisting of all the edges together with the three square faces .
our categorical operad @xmath133 is almost the same as leinster s @xmath134 ( @xcite ) .
the only difference is that he has @xmath135 , whereas we have @xmath136 .
so our approach encodes the notion of a unit for algebras over @xmath133 .
leinster expected that the nerve of @xmath137 is homeomorphic to the associahedron , which we prove .
thus leinster s topological operad is precisely the same as stasheff s .
+ the tree description of a @xmath18-operad containing @xmath133 appears in @xcite .
since the nerve of a product in @xmath18 is a product in @xmath138 , it follows that the nerve of a @xmath139 algebra is an @xmath0-space in the sense of stasheff .
for @xmath140 , an @xmath0-monoidal category is a category @xmath141 together with multiplications @xmath142 for @xmath143 such that 1 .
@xmath144 is the identity functor .
@xmath145 is an object @xmath146 that acts as a strict unit in the sense that @xmath147 for any @xmath24 such that @xmath148 .
@xmath141 is also equipped with natural transformations ( associators ) @xmath149 for @xmath150 , satisfying @xmath151 and the coherence conditions specified by the following commutative diagrams @xmath152\ar[d]^{\alpha^{a+b+1,c , d}_{(\overline{a},0,\overline{b}),\overline{c},\overline{d } } } & \mu_{a+b+d+1}(\overline{a},\overline{b},\mu_c(\overline{c}),\overline{d } ) \ar[d]^{\alpha^{a+b , c , d}_{(\overline{a},\overline{b}),\overline{c},\overline{d}}}\\ \mu_{a+b+c+d+1}(\overline{a},0,\overline{b},\overline{c},\overline{d})\ar@{=}[r ] & \mu_{a+b+c+d}(\overline{a},\overline{b},\overline{c},\overline{d } ) } \\
\lefteqn{\raisebox{-20pt}{\mbox{(iii ) } } } & \qquad&\qquad\qquad\xymatrix { \mu_{a+c+d+2}(\overline{a},\mu_b(\overline{b}),\overline{c},0,\overline{d } ) \ar@{=}[r]\ar[d]^{\alpha^{a , b , c+d+1}_{\overline{a},\overline{b},(\overline{c},0,\overline{d } ) } } & \mu_{a+c+d+1}(\overline{a},\mu_b(\overline{b}),\overline{c},\overline{d } ) \ar[d]^{\alpha^{a , b , c+d}_{\overline{a},\overline{b},(\overline{c},\overline{d})}}\\ \mu_{a+b+c+d+1}(\overline{a},\overline{b},\overline{c},0,\overline{d})\ar@{=}[r ] & \mu_{a+b+c+d}(\overline{a},\overline{b},\overline{c},\overline{d } ) } \\
\lefteqn{\raisebox{-20pt}{\mbox{(iv ) } } } & \qquad&\qquad\qquad\xymatrix { \mu_{a+d+1}(\overline{a},\mu_{b+c+1}(\overline{b},0,\overline{c}),\overline{d } ) \ar@{=}[r]\ar[d]^{\alpha^{a , b+c+1,d}_{\overline{a},(\overline{b},0,\overline{c}),\overline{d } } } & \mu_{a+d+1}(\overline{a},\mu_{b+c}(\overline{b},\overline{c}),\overline{d } ) \ar[d]^{\alpha^{a , b+c , d}_{\overline{a},(\overline{b},\overline{c}),\overline{d}}}\\ \mu_{a+b+c+d+1}(\overline{a},\overline{b},0,\overline{c},\overline{d})\ar@{=}[r ] & \mu_{a+b+c+d}(\overline{a},\overline{b},\overline{c},\overline{d } ) } \\\end{aligned}\ ] ] @xmath153^{\alpha^{a+c+1,d , e}_{(\overline{a},\mu_b(\overline{b}),\overline{c}),\overline{d},\overline{e } } } \ar[rr]^{\alpha^{a , b , c+e+1}_{\overline{a},\overline{b},(\overline{c},\mu_d(\overline{d}),\overline{e } ) } } & & \mu_{a+b+c+e+1}(\overline{a},\overline{b},\overline{c},\mu_d(\overline{d}),\overline{e } ) \ar[d]^{\alpha^{a+c+1,d , e}_{(\overline{a},\overline{b},\overline{c}),\overline{d},\overline{e}}}\\ \mu_{a+c+d+e+1}(\overline{a},\mu_b(\overline{b}),\overline{c},\overline{d},\overline{e } ) \ar[rr]^{\alpha^{a , b , c+d+e}_{\overline{a},\overline{b},(\overline{c},\overline{d},\overline{e } ) } } & & \mu_{a+b+c+d+e}(\overline{a},\overline{b},\overline{c},\overline{d},\overline{e } ) } \\
\lefteqn{\raisebox{-20pt}{\mbox{(vi ) } } } & \qquad&\xymatrix { \mu_{a+e+1}(\overline{a},\mu_{b+d+1}(\overline{b},\mu_c(\overline{c}),\overline{d}),\overline{e } ) \ar[rrr]^{\mu_{a+e+1}(id_{\overline{a}},\alpha^{b , c , d}_{\overline{b},\overline{c},\overline{d}},id_{\overline{e } } ) } \ar[d]^{\alpha^{a , b+d+1,e}_{\overline{a},(\overline{b},\mu_c(\overline{c}),\overline{d}),\overline{e } } } & & & \mu_{a+e+1}(\overline{a},\mu_{b+c+d}(\overline{b},\overline{c},\overline{d}),\overline{e } ) \ar[d]^{\alpha^{a , b+c+d , e}_{\overline{a},(\overline{b},\overline{c},\overline{d}),\overline{e}}}\\ \mu_{a+b+d+e+1}(\overline{a},\overline{b},\mu_c(\overline{c}),\overline{d},\overline{e } ) \ar[rrr]^{\alpha^{a+b , c , d+e}_{(\overline{a},\overline{b}),\overline{c},(\overline{d},\overline{e } ) } } & & & \mu_{a+b+c+d+e}(\overline{a},\overline{b},\overline{c},\overline{d},\overline{e } ) } \end{aligned}\ ] ] here @xmath154 , @xmath155 , @xmath156 , @xmath157 , @xmath158 are taken to be objects of @xmath159 , @xmath160 , @xmath161 , @xmath162 , @xmath163 , respectively .
essentially coherence conditions ( i)-(iv ) require the associators to be compatible with the strict unit 0 , while ( v ) and ( vi ) just say that if we are removing two pairs of matching parentheses in a multiplication , it does nt matter which we remove first . in @xcite , leinster defines the notion of a lax monoidal category , which is similar in spirit to the above definition , but there are some crucial differences . a lax monoidal category in his sense , has multiplications @xmath164 for all @xmath165 together with natural transformations @xmath166 and a natural transformation @xmath167 the natural transformations @xmath168 satisfy a coherence condition which is essentially our coherence conditions ( v ) and ( vi ) combined into a single diagram .
there is no unit condition for @xmath169 ( so one might as well require the existence of @xmath170 for @xmath171 only ) .
moreover his natural transformation @xmath172 is not the identity .
thus a lax monoidal category in his sense possesses arbitrarily long nondegenerate strings of composable natural transformations between unary multiplications @xmath173 it follows that the operad controlling such a structure has an infinite dimensional nerve .
the main result of this paper is : [ main ] a category @xmath141 is a @xmath139-algebra iff it is an @xmath0-monoidal category . given an action , @xmath174 , define @xmath175 to be the restriction of this action to @xmath176 , where @xmath177 is the terminal object of @xmath118 .
this makes sense for @xmath178 , since in those cases @xmath179 is contained in the @xmath4-th filtration @xmath139 .
we then define @xmath180 to be the restriction of @xmath181 to @xmath182\}\right ) \longrightarrow(w_{i+j+k},\emptyset)\right\}\times\scc^{i+j+k}$ ] , for @xmath150 .
conditions ( 1 ) , ( 2 ) and ( i ) follow from the fact that @xmath183 is the identity of the operad and composing the constant @xmath184 into any input of @xmath185 gives @xmath186 .
conditions ( ii)-(iv ) also follow from the latter fact . finally conditions ( v ) and
( vi ) follow from the restriction of @xmath187 to @xmath188 and @xmath189 , where @xmath190 and @xmath191 are the following commutative diagrams in @xmath192 : @xmath193,[x_{a+b+c+1},x_{a+b+c+d}]\right\}\right)\ar[r]\ar[d ] & \left(w_{a+b+c+d+e},\left\{[x_{a+1},x_{a+b}]\right\}\right)\ar[d]\\ \left(w_{a+b+c+d+e},\left\{[x_{a+b+c+1},x_{a+b+c+d}]\right\}\right)\ar[r ] & ( w_{a+b+c+d+e},\emptyset ) } \ ] ] @xmath194,[x_{a+b+1},x_{a+b+c}]\right\}\right)\ar[r]\ar[d ] & \left(w_{a+b+c+d+e},\left\{[x_{a+b+1},x_{a+b+c}]\right\}\right)\ar[d]\\ \left(w_{a+b+c+d+e},\left\{[x_{a+1},x_{a+b+c+d}]\right\}\right)\ar[r ] & ( w_{a+b+c+d+e},\emptyset ) } \ ] ] respectively .
conversely suppose that @xmath141 is an @xmath0-monoidal category
. then we define @xmath195 by induction on @xmath100 as follows .
we define @xmath196 to be @xmath169 and @xmath197 to be @xmath198 .
having defined @xmath199 for @xmath200 , consider an object @xmath94 in @xmath117 represented by a tree @xmath201&\ar@{-}[ddr ] & & \ar@{-}[ddl]\\ \\ & & { \scriptscriptstyle\bullet}\ar@{-}[dd]\\ \\ & & } \ ] ] with @xmath202 and where @xmath203 has @xmath204 input edges , so that @xmath205 .
let @xmath206 represent an object in @xmath207 , with @xmath208 , @xmath209 . by induction @xmath210
are already defined for @xmath209 .
we then define @xmath211 [ here we use implicitly the canonical isomorphisms between the associahedral posets based on subintervals of @xmath212 with the associahedral posets based on the standard linear orders @xmath213 of the same cardinality , c.f .
[ kmdefi ] . ]
we define @xmath214 for morphisms in @xmath207 similarly .
this completes the induction .
next we extend the definition of @xmath214 to define natural transformations @xmath215 where @xmath216 are the indecomposable morphisms in @xmath117 , i.e. morphisms which ca nt be factored nontrivially ( or equivalently morphisms given by dropping a single pair of matching parentheses in a parenthesized word ) .
again we proceed by induction on @xmath100 , starting with @xmath217 and @xmath218 where these are vacuously defined .
now consider an indecomposable morphism @xmath219 in @xmath117 , where @xmath94 has the form @xmath201&\ar@{-}[ddr ] & & \ar@{-}[ddl]\\ \\ & & { \scriptscriptstyle\bullet}\ar@{-}[dd]\\ \\ & & } \ ] ] with @xmath202 and where @xmath203 has @xmath204 input edges , so that @xmath205
. then @xmath220 is obtained by shrinking a single interior edge in @xmath94 .
there are two possibilities : ( 1 ) an interior edge of some tree @xmath203 is shrunk or ( 2 ) an edge below some @xmath203 is shrunk .
now let @xmath206 represent an object in @xmath207 , with @xmath208 , @xmath209 . in the first case
we define @xmath221 + @xmath222 where @xmath223 is the indecomposable morphism in @xmath224 given by shrinking that particular edge . in the second case
we define @xmath225 this completes the induction .
finally to extend @xmath214 to all morphisms in @xmath118 , we must show that for any factorization of a morphism in @xmath117 into indecomposable morphisms , the corresponding composition of natural transformations defines the same morphism in @xmath141 . but
according to lemma [ surprise ] , the factorizations of any morphism in @xmath118 give rise to a cubical diagram in @xmath141 .
according to coherence conditions ( v ) and ( vi ) of an @xmath0-monoidal category , all the 2-dimensional faces of this cubical diagram commute .
it is an elementary consequence that the entire cubical diagram in @xmath141 commutes , c.f .
lemma [ cubicallemma ] below .
it follows that there are well defined functors : @xmath226 for all @xmath227 .
the fact that @xmath214 are compatible with the operadic compositions @xmath228 follows from the inductive construction of @xmath214 if all the @xmath229 . if @xmath230 or @xmath40 , the compatibility follows from conditions ( 1 ) , ( 2 ) and ( i)-(iv ) of the definition of an @xmath0-monoidal category . [ cubicallemma ] a cubical diagram in any category commutes iff each of its 2-dimensional faces commutes .
we proceed by induction on the dimension of the cube .
the statement is vacuously true if the dimension is @xmath231 .
suppose it is true for all cubical diagrams of dimension @xmath232 , and suppose we are given an @xmath39-dimensional cubical diagram .
consider two edge paths from the initial object @xmath233 of the diagram to @xmath234 , the terminal object .
let these edge paths factor as @xmath235 respectively , where @xmath236 and @xmath237 are edges of the diagram and @xmath238 and @xmath239 are composites of the remainders of these edge paths .
if @xmath240 , then by induction @xmath241 and we are done . otherwise let @xmath242^\alpha\ar[d]^\beta & b\ar[d]^\gamma\\ c\ar[r]^\delta & d } \ ] ] be the 2-dimensional face spanned by @xmath236 and @xmath237 .
pick any edge path @xmath243 , and consider the diagram @xmath244_\gamma\ar[drrr]^f\\ a\ar[ur]^\alpha\ar[dr]_\beta & & d\ar[rr]^{h\quad } & & z\\ & c\ar[ur]^\delta\ar[urrr]_g } \ ] ] by hypothesis the square commutes and by induction the two triangles commute .
hence @xmath245 .
this completes the induction and proof .
we say that an @xmath0-monoidal category is _ undirected _ if all the associativity natural transformations @xmath180 are isomorphisms .
an undirected @xmath0-monoidal category is a monoidal category if @xmath126 .
if @xmath141 is an undirected @xmath0-monoidal category , then the corresponding action functors @xmath246 extend to @xmath247 , where @xmath248 is obtained from @xmath117 by formally inverting all the morphisms . by proposition [ connectivity ] the nerve of @xmath117
is simply connected . now recalling that inverting all the morphisms in a connected category has the effect of killing off the higher homotopy groups of its nerve ( c.f .
* proposition 1 ) ) , we see that the nerve of @xmath248 is contractible , and it follows that the objects of @xmath127 are connected to each other by uniquely defined isomorphisms in @xmath248 .
the images of these isomorphisms under @xmath214 specify uniquely defined natural isomorphisms connecting all possible different ways of associating the binary product @xmath249 into an @xmath100-fold product @xmath250 so that all diagrams involving them commute
. thus @xmath141 is a monoidal category , in the classical sense of mac lane .
next we derive laplaza s coherence theorem @xcite , which generalizes mac lane s coherence theorem to the case where the associativity natural transformation for a monoidal structure on a category is not required to be an isomorphism .
we begin with a preliminary version of this result .
[ cohthm0 ] let @xmath251 be a directed monoidal category with a strict unit . that is
, @xmath252 is a bifunctor and 0 is an object of @xmath141 which serves as a strict unit for @xmath253 , i.e. the restrictions of @xmath253 to @xmath254 and @xmath255 are the identity .
finally @xmath256 is a natural transformation ( not necessarily an isomorphism ) such that @xmath257 is the identity whenever one of @xmath233 , @xmath258 , @xmath259 is 0 and such that the pentagonal diagram @xmath260^{\eta_{a , b , c\box d}}\\ \\ ( ( a\box b)\box c)\box d\ar[uurr]^{\eta_{a\box b , c , d}}\ar[ddddr]^{\eta_{a , b , c}\box id_d } & & & & a\box(b\box(c\box d))\\ \\ \\ \\ & ( a\box(b\box c))\box d\ar[rr]^{\eta_{a , b\box c , d } } & & a\box((b\box c)\box d)\ar[uuuur]^{id_a\box\eta_{b , c , d } } } \ ] ] commutes . then @xmath141 can be endowed with the structure of an @xmath7-monoidal category .
we define @xmath261 , @xmath262 to be the identity , @xmath263 and then we inductively define @xmath264 to be the composite @xmath265^{id_\scc\times\mu_{i-1}}&&\scc\times\scc\stackrel{\box}{\longrightarrow}\scc } \ ] ] thus @xmath266 and we have @xmath267 for any objects @xmath268 , @xmath269 .
now let @xmath269 and @xmath270 .
we define the associativity @xmath271 inductively on @xmath32 .
we assume @xmath272 , since @xmath273 is required by definition to be the identity . for @xmath274 , we also require @xmath275 to be the identity .
so suppose @xmath276 and @xmath277 .
then we define @xmath278 to be the composite@xmath279 @xmath280^{\eta_{b_1,\mu_{b-1}(\overline{b}'),\mu_c(\overline{c } ) } } & & & & b_1\box(\mu_{b-1}(\overline{b}')\box\mu_c(\overline{c } ) ) } \ ] ] followed by the composite @xmath281 @xmath282^{id_{b_1}\box\alpha^{0,b-1,c}_{\overline{b}',\overline{c}}\qquad\qquad}&&&b_1\box\mu_{b+c-1}(\overline{b } ' , \overline{c})=\mu_{b+c}(\overline{b},\overline{c } ) . } \ ] ] this completes the inductive definition of @xmath271 .
we then define @xmath283 to be the composite@xmath284 @xmath285^{\mu_{a+1}(id_{\overline{a}},\alpha^{0,b , c}_{\mu_b(\overline{b}),\overline{c } } ) \qquad\qquad\qquad\qquad}&&&\mu_{a+1}(\overline{a},\mu_{b+c}(\overline{b},\overline{c } ) ) = \mu_{a+b+c}(\overline{a},\overline{b},\overline{c } ) . } \ ] ] note that this implies that @xmath283 is the identity if @xmath286 , and that @xmath287 conditions ( 1 ) , ( 2 ) , ( i)-(iv ) for an @xmath7-monoidal category are either true by construction or follow by a straight forward induction argument using the hypotheses that 0 is a strict unit for @xmath253 and that @xmath257 is the identity whenever one of @xmath233 , @xmath258 or @xmath259 is 0 . by ( * * ) the verification of condition ( v ) reduces to the special case of the diagram @xmath288^{\alpha^{c+1,d , e}_{(\mu_b(\overline{b}),\overline{c}),\overline{d},\overline{e } } } \ar[rr]^{\alpha^{0,b , c+e+1}_{\overline{b},(\overline{c},\mu_d(\overline{d}),\overline{e } ) } } & & \mu_{b+c+e+1}(\overline{b},\overline{c},\mu_d(\overline{d}),\overline{e } ) \ar[d]^{\alpha^{c+1,d , e}_{(\overline{b},\overline{c}),\overline{d},\overline{e}}}\\ \mu_{c+d+e+1}(\mu_b(\overline{b}),\overline{c},\overline{d},\overline{e } ) \ar[rr]^{\alpha^{0,b , c+d+e}_{\overline{b},(\overline{c},\overline{d},\overline{e } ) } } & & \mu_{b+c+d+e}(\overline{b},\overline{c},\overline{d},\overline{e } ) } \ ] ] since the general diagram for ( v ) can be obtained from this one by applying the functor @xmath289 to it . by ( * ) and ( * * ) , this diagram in turn is the same as the diagram @xmath290^{\mu_2(id_{\mu_b(\overline{b})},\alpha^{c , d , e}_{\overline{c},\overline{d},\overline{e } } ) } \ar[rrr]^{\alpha^{0,b,1}_{\overline{b},\mu_{c+e+1}(\overline{c},\mu_d(\overline{d}),\overline{e } ) } } & & & \mu_{b+1}(\overline{b},\mu_{c+e+1}(\overline{c},\mu_d(\overline{d}),\overline{e } ) ) \ar[d]^{\mu_{b+1}(id_{\overline{b}},\alpha^{c , d , e}_{\overline{c},\overline{d},\overline{e}})}\\ \mu_{2}(\mu_b(\overline{b}),\mu_{c+d+e}(\overline{c},\overline{d},\overline{e } ) ) \ar[rrr]^{\alpha^{0,b,1}_{\overline{b},\mu_{c+d+e}(\overline{c},\overline{d},\overline{e } ) } } & & & \mu_{b+1}(\overline{b},\mu_{c+d+e}(\overline{c},\overline{d},\overline{e } ) ) } \ ] ] this last diagram in turn commutes because @xmath291 is a natural transformation . by similar reasoning
, the verification of condition ( vi ) reduces to the special case of the diagram @xmath292^{\mu_{e+1}(\alpha^{b , c , d}_{\overline{b},\overline{c},\overline{d}},id_{\overline{e } } ) } \ar[d]^{\alpha^{0,b+d+1,e}_{(\overline{b},\mu_c(\overline{c}),\overline{d}),\overline{e } } } & & & \mu_{e+1}(\mu_{b+c+d}(\overline{b},\overline{c},\overline{d}),\overline{e } ) \ar[d]^{\alpha^{0,b+c+d , e}_{(\overline{b},\overline{c},\overline{d}),\overline{e}}}\\ \mu_{b+d+e+1}(\overline{b},\mu_c(\overline{c}),\overline{d},\overline{e } ) \ar[rrr]^{\alpha^{b , c , d+e}_{\overline{b},\overline{c},(\overline{d},\overline{e } ) } } & & & \mu_{b+c+d+e}(\overline{b},\overline{c},\overline{d},\overline{e } ) } \ ] ] by ( * ) it follows that this diagram is unchanged if we replace @xmath293 throughout by @xmath294 .
hence we may as well suppose that @xmath295 and @xmath296 is an object of @xmath141 .
then by the inductive definition of @xmath297 and ( * * ) , we can factor the diagram as follows : @xmath298^{(id_{b_1}\box\alpha^{b-1,c , d}_{\overline{b}',\overline{c},\overline{d}})\box id_{e } } \ar[d]^{\eta_{b_1,\mu_{b+d}(\overline{b}',\mu_c(\overline{c}),\overline{d}),e } } & & & ( b_1\box\mu_{b+c+d-1}(\overline{b}',\overline{c},\overline{d}))\box e \ar[d]^{\eta_{b_1,\mu_{b+c+d-1}(\overline{b}',\overline{c},\overline{d}),e}}\\ b_1\box(\mu_{b+d}(\overline{b}',\mu_c(\overline{c}),\overline{d})\box e ) \ar[rrr]^{id_{b_1}\box(\alpha^{b-1,c , d}_{\overline{b}',\overline{c},\overline{d}}\box id_{e } ) } \ar[d]^{id_{b_1}\box\alpha^{0,b+d,1}_{(\overline{b}',\mu_c(\overline{c}),\overline{d}),\overline{e } } } & & & b_1\box(\mu_{b+c+d-1}(\overline{b}',\overline{c},\overline{d})\box e ) \ar[d]^{id_{b_1}\box\alpha^{0,b+c+d-1,1}_{(\overline{b},\overline{c},\overline{d}),e } } \\
b_1\box\mu_{b+d+1}(\overline{b}',\mu_c(\overline{c}),\overline{d},e ) \ar[rrr]^{id_{b_1}\box\alpha^{b-1,c , d+1}_{\overline{b},\overline{c},(\overline{d},e ) } } & & & b_1\box\mu_{b+c+d}(\overline{b}',\overline{c},\overline{d},e ) } \ ] ] the upper square commutes by naturality of @xmath299 .
the commutativity of the lower square corresponds to a reduction of the problem from @xmath32 to @xmath300 .
recursing on this reduction we reduce to the case @xmath301 , i.e. showing that the diagram @xmath302^{\alpha^{0,c , d}_{\overline{c},\overline{d}}\box id_{e } } \ar[d]^{\alpha^{0,d+1,1}_{(\mu_c(\overline{c}),\overline{d}),e } } & & & \mu_{c+d}(\overline{c},\overline{d})\box{e } \ar[d]^{\alpha^{0,c+d,1}_{(\overline{c},\overline{d}),e}}\\ \mu_{d+2}(\mu_c(\overline{c}),\overline{d},e ) \ar[rrr]^{\alpha^{0,c , d+1}_{\overline{c},(\overline{d},e ) } } & & & \mu_{c+d+1}(\overline{c},\overline{d},e ) } \ ] ] commutes . by ( * ) and the inductive definition of @xmath297 this diagram can be replaced and expanded into the following diagram @xmath303^{\alpha^{0,c,1}_{\overline{c},\mu_d(\overline{d})}\box id_{e } } \ar[d]^{\alpha^{0,2,1}_{(\mu_c(\overline{c}),\mu_d(\overline{d})),e } } & & & \mu_{c+1}(\overline{c},\mu_d(\overline{d}))\box{e } \ar[d]^{\alpha^{0,c+1,1}_{(\overline{c},\mu_d(\overline{d})),e}}\\ \mu_{3}(\mu_c(\overline{c}),\mu_d(\overline{d}),e ) \ar[rrr]^{\alpha^{0,c,2}_{\overline{c},(\mu_d(\overline{d}),e ) } } \ar[d]^{\mu_2(id_{\mu_c(\overline{c})},\alpha^{0,d,1}_{\overline{d},e } ) } & & & \mu_{c+2}(\overline{c},\mu_d(\overline{d}),e ) \ar[d]^{\mu_{c+1}(id_{\overline{c}},\alpha^{0,d,1}_{\overline{d},e } ) } \\
\mu_2(\mu_c(\overline{c}),\mu_{d+1}(\overline{d},e ) ) \ar[rrr]^{\alpha^{0,c,1}_{\overline{c},\mu_{d+1}(\overline{d},e ) } } & & & \mu_{c+1}(\overline{c},\mu_{d+1}(\overline{d},e ) ) } \ ] ] the lower square commutes by naturality of @xmath236 .
so it suffices to show the upper square commutes .
this is just the previous diagram with @xmath157 replaced by @xmath304 .
thus we have reduced to the case @xmath305
. we will find it convenient to display this diagram in reflected form : @xmath306^{\alpha^{0,c,1}_{\overline{c},d}\box id_{e } } \ar[rrr]^{\alpha^{0,2,1}_{(\mu_c(\overline{c}),d),e } } & & & \mu_{3}(\mu_c(\overline{c}),d , e ) \ar[d]^{\alpha^{0,c,2}_{\overline{c},(d , e)}}\\ \mu_{c+1}(\overline{c},d)\box{e } \ar[rrr]^{\alpha^{0,c+1,1}_{(\overline{c},d),e } } & & & \mu_{c+2}(\overline{c},d , e ) } \ ] ] we have to show this diagram commutes , where @xmath307 and @xmath293 are objects of @xmath141 and @xmath156 is an object of @xmath161 .
this diagram commutes trivially if @xmath308 .
so assume @xmath309 and @xmath310 . again using ( * ) and the inductive definition of @xmath297
, we can expand this diagram into @xmath311^{\alpha^{0,2,1}_{\mu_2(c_1,\mu_{c-1}(\overline{c}')),d}\box id_{e } } \ar[rrr]^{\alpha^{0,2,1}_{(\mu_2(c_1,\mu_{c-1}(\overline{c}'),d),e } } & & & \mu_{3}(\mu_2(c_1,\mu_{c-1}(\overline{c}')),d , e ) \ar[d]^{\alpha^{0,2,2}_{(c_1,\mu_{c-1}(\overline{c}')),(d , e)}}\\ \mu_{3}(c_1,\mu_{c-1}(\overline{c}'),d)\box{e}\ar@{=}[d ] \ar[rrr]^{\quad\alpha^{0,3,1}_{(c_1,\mu_{c-1}(\overline{c}'),d),e } } & & & \mu_{4}(c_1,\mu_{c-1}(\overline{c}'),d , e)\ar@{=}[d ] \\
\mu_{2}(c_1,\mu_2(\mu_{c-1}(\overline{c}'),d))\box{e } \ar[d]^{\mu_2(id_{c_1},\alpha^{0,1,1}_{\mu_{c-1}(\overline{c}'),d})\box id_e } & & & \mu_{2}(c_1,\mu_3(\mu_{c-1}(\overline{c}'),d , e ) ) \ar[d]^{\mu_2(id_{c_1},\alpha^{0,1,2}_{\mu_{c-1}(\overline{c}'),(d , e ) } ) } \\ \mu_{2}(c_1,\mu_c(\overline{c}',d))\box e \ar@{=}[d ] & & & \mu_{2}(c_1,\mu_{c+1}(\overline{c}',d , e ) ) \ar@{=}[d ] \\
\mu_{c+1}(\overline{c},d)\box{e } \ar[rrr]^{\alpha^{0,c+1,1}_{(\overline{c},d),e } } & & & \mu_{c+2}(\overline{c},d , e ) } \ ] ] the top square in this diagram is the original diagram with @xmath156 replaced by @xmath312 , thus reducing it to the case @xmath313 .
this top square can be expanded into the pentagonal diagram of the hypothesis of the theorem and thus commutes .
it remains to show that the bottom square commutes .
after rewriting the bottom square in reflected form and applying the inductive definition of @xmath297 we obtain the following expanded diagram @xmath314^{\quad(id_{c_1}\box\alpha^{0,1,1}_{\mu_{c-1}(\overline{c}'),d})\box id_e } \ar[d]^{\eta_{c_1,\mu_{c-1}(\overline{c}')\box d , e } } & & & ( c_1\box\mu_c(\overline{c}',d))\box e \ar[d]^{\eta_{c_1,\mu_c(\overline{c}',d),e } } \\
c_1\box((\mu_{c-1}(\overline{c}')\box d)\box e ) \ar[rrr]^{\quad id_{c_1}\box(\alpha^{0,1,1}_{\mu_{c-1}(\overline{c}'),d}\box
id_e ) } \ar[d]^{id_{c_1}\box\alpha^{0,2,1}_{(\mu_{c-1}(\overline{c}'),d),e } } & & & c_1\box(\mu_c(\overline{c}',d)\box e ) \ar[d]^{id_{c_1}\box\alpha^{0,c,1}_{(\overline{c}',d),e } } \\ c_1\box\mu_3(\mu_{c-1}(\overline{c}'),d , e ) \ar[rrr]^{\qquad id_{c_1}\box\alpha^{0,c-1,2}_{\overline{c}',(d , e ) } } & & & c_1\box\mu_{c+1}(\overline{c}',d , e ) } \ ] ] the top square commutes by naturality of @xmath299 and the bottom square by naturality of @xmath236 .
this completes the verification that we have constructed an @xmath7-monoidal structure on @xmath141 . to obtain the full version of laplaza s coherence theorem
, we start with an operadic reformulation of theorem [ cohthm0 ] .
the laplaza operad @xmath315 is the operad in @xmath18 which acts on directed monoidal categories as in the hypothesis of theorem [ cohthm0 ] .
specifically @xmath316 can be described as a full subcategory of the free directed monoidal category on @xmath39 generating objects @xmath317 , whose objects look like @xmath318 after removing all parentheses
. thus @xmath319 , @xmath320 , and for @xmath321 the objects of @xmath316 are in bijective correspondence with planar binary trees with @xmath39 input edges .
@xmath322 is the trivial poset @xmath323 , @xmath324 is the poset @xmath325 isomorphic to @xmath59 , while @xmath326 is the pentagonal poset generated by the labelled arrows shown below .
@xmath327^{\eta_{x_1,x_2,x_3\box x_4}}\\ ( ( x_1\box x_2)\box x_3)\box x_4 \ar[ur]^{\eta_{x_1\box x_2,x_3,x_4}\quad } \ar[d]_{\eta_{x_1,x_2,x_3}\box id_{x_4 } } \ar[rr ] \ar[drr]|(.5){\quad\hole } & & x_1\box(x_2\box(x_3\box x_4))\\ ( x_1\box(x_2\box x_3))\box x_4\ar[rr]_{\eta_{x_1,x_2\box x_3,x_4 } } \ar[urr ] & & x_1\box((x_2\box x_3)\box x_4)\ar[u]_{\ id_{x_1}\box\eta_{x_2,x_3,x_4 } } } \ ] ] laplaza s coherence theorem states that @xmath316 is a poset for all @xmath39 .
laplaza works with natural transformations @xmath328 .
moreover , he does not consider units .
so , if @xmath329 is the suboperad obtained from @xmath330 by dropping the unit , the operad @xmath331 is dual to laplaza s original one .
note also that @xmath332 for @xmath333 .
our poset @xmath316 , @xmath334 is precisely the poset considered by tamari @xcite , and is now commonly called the _ tamari lattice _ @xcite . with this notation , we can reformulate theorem [ cohthm0 ] as follows .
[ cohthm1 ] there is a map of @xmath18-operads @xmath335 which is a surjection .
the existence of @xmath336 is clear from the statement of theorem [ cohthm0 ] .
surjectivity follows from the proof of theorem [ cohthm0 ] , where it is shown that @xmath337 maps via @xmath336 to @xmath325 and the fact that @xmath338 generates @xmath330 as a @xmath18-operad .
laplaza s coherence theorem is not immediately apparent from theorem [ cohthm1 ] , since a quotient category of a poset need not be a poset .
we need the following additional observation .
[ inverse_images ] for any object @xmath339 , the inverse image under @xmath336 of the subcategory @xmath340 is a subposet of @xmath36 containing both a minimal and a maximal object . for @xmath341 ,
the functor @xmath336 is an isomorphism and there is nothing to prove . for @xmath334
, we may regard @xmath94 as a planar binary tree .
clearly the minimal object of @xmath342 is @xmath94 regarded as an object of @xmath36 .
the maximal object of @xmath342 is obtained from @xmath94 by successively shrinking the rightmost incoming edge to every node of @xmath94 , with the exception of those edges which are leaves , till the rightmost edge of each node is a leaf .
the inverse images in lemma [ inverse_images ] for @xmath343 are as follows : @xmath344 & & \ar@{-}[dl]&&&\ar@{-}[ddll ] & & & \ar@{-}[dddlll]\\ & { \scriptscriptstyle\bullet}\ar@{-}[drr]\\ & & & { \scriptscriptstyle\bullet}\ar@{-}[drr]\\ & & & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & & & } $ } \right\}\\ \lambda^{-1}\{(x_1\box x_2)\box ( x_3\box x_4)\ } & = & \left\ { \begin{picture}(250,20 ) \put(0,0){$(x_1x_2)(x_3x_4)=$ \raisebox{10pt}{$\xymatrix@=5pt@m=-1pt@w=-1pt { \ar@{-}[dr ] & & \ar@{-}[dl]&&\ar@{-}[dr ] & & \ar@{-}[dl]\\ & { \scriptscriptstyle\bullet}\ar@{-}[ddrr]&&&&{\scriptscriptstyle\bullet}\ar@{-}[ddll]\\ \\ & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & & & { } } $ } } \put(120,4){\vector(1,0){20 } } \put(145,0){$(x_1x_2)x_3x_4=$ \raisebox{10pt}{$\xymatrix@=5pt@m=-1pt@w=-1pt { \ar@{-}[dr]&&\ar@{-}[dl]&\ar@{-}[dddl]&&&\ar@{-}[dddllll]\\ & { \scriptscriptstyle\bullet}\ar@{-}[ddr]\\ \\ & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & } $ } } \end{picture } \right\}\\ \lambda^{-1}\{(x_1\box ( x_2\box x_3))\box x_4\ } & = & \left\ { \begin{picture}(250,20 ) \put(0,0){$(x_1(x_2x_3))x_4=$ \raisebox{10pt}{$\xymatrix@=5pt@m=-1pt@w=-1pt { \ar@{-}[ddrrr ] & & & \ar@{-}[dr]&&\ar@{-}[dl ] & & & \ar@{-}[dddllll]\\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[dl]\\ & & & { \scriptscriptstyle\bullet}\ar@{-}[dr]\\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & & } $ } } \put(120,4){\vector(1,0){20 } } \put(145,0){$(x_1x_2x_3)x_4=$ \raisebox{10pt}{$\xymatrix@=5pt@m=-1pt@w=-1pt { \ar@{-}[ddr]&\ar@{-}[dd]&\ar@{-}[ddl ] & & & \ar@{-}[dddlll]\\ \\ & { \scriptscriptstyle\bullet}\ar@{-}[dr]\\ & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & } $ } } \end{picture } \right\}\\ \lambda^{-1}\{x_1\box ( ( x_2\box x_3)\box x_4)\ } & = & \left\ { \begin{picture}(250,20 ) \put(0,0){$x_1((x_2x_3)x_4)=$ \raisebox{10pt}{$\xymatrix@=5pt@m=-1pt@w=-1pt { \ar@{-}[dddrrrr ] & & & \ar@{-}[dr]&&\ar@{-}[dl ] & & & \ar@{-}[ddlll]\\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[dr]\\ & & & & & { \scriptscriptstyle\bullet}\ar@{-}[dl]\\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & & } $ } } \put(120,4){\vector(1,0){20 } } \put(145,0){$x_1(x_2x_3)x_4=$ \raisebox{10pt}{$\xymatrix@=5pt@m=-1pt@w=-1pt { \ar@{-}[dddrrr]&&\ar@{-}[dr]&&\ar@{-}[dl]&&\ar@{-}[dddlll]\\ & & & { \scriptscriptstyle\bullet}\ar@{-}[dd]\\ \\ & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & } $ } } \end{picture } \right\}\\\end{aligned}\ ] ] @xmath345 @xmath346 & & & \ar@{-}[ddrr]&&\ar@{-}[dr ] & & & \ar@{-}[dll]\\ & & & & & & { \scriptscriptstyle\bullet}\ar@{-}[dl]\\ & & & & & { \scriptscriptstyle\bullet}\ar@{-}[dl]\\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & & } $ } } \put(120,0){\vector(1,0){170 } } \put(120,5){\vector(2,1){40 } } \put(150,30){$x_1(x_2x_3x_4)=$ \raisebox{10pt}{$\xymatrix@=5pt@m=-1pt@w=-1pt { \ar@{-}[dddrrr]&&&\ar@{-}[ddr]&\ar@{-}[dd]&\ar@{-}[ddl]\\ \\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[dl]\\ & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & } $ } } \put(120,-5){\vector(2,-1){40 } } \put(150,-35){$x_1x_2(x_3x_4)=$ \raisebox{10pt}{$\xymatrix@=5pt@m=-1pt@w=-1pt { \ar@{-}[dddrrrr]&&&\ar@{-}[dddr]&&\ar@{-}[dr]&&\ar@{-}[dl]\\ & & & & & & { \scriptscriptstyle\bullet}\ar@{-}[ddll]\\ \\ & & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & & } $ } } \put(250,25){\vector(2,-1){40 } } \put(250,-30){\vector(2,1){40 } } \put(300,0){$x_1x_2x_3x_4=$ \raisebox{10pt}{$\xymatrix@=5pt@m=-1pt@w=-1pt { \ar@{-}[dddrrr]&&\ar@{-}[dddr]&&\ar@{-}[dddl]&&\ar@{-}[dddlll]\\ \\ \\ & & & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & & } $ } } \end{picture } \right\}\ ] ] ( laplaza coherence theorem ) for all @xmath39 , the category @xmath316 is a poset ( known as the tamari lattice for @xmath347 ) .
let @xmath348^f\ar@<-2pt>[rr]_g & & t}\ ] ] be morphisms in @xmath316 .
since @xmath336 is surjective , we may find preimages @xmath349 under @xmath336 of @xmath238 and @xmath239 respectively .
now let @xmath350 be the minimal element of @xmath351 and let @xmath352 be the maximal element of @xmath342 .
then since @xmath36 is a poset , we have a commutative diagram @xmath353^{f ' } & t'\ar[dr]\\ s'''\ar[ur]\ar[dr ] & & & t'''\\ & s''\ar[r]^{g ' } & t''\ar[ur ] } \ ] ] applying @xmath336 to this diagram and noting that @xmath336 sends the unlabelled arrows to identities , we obtain @xmath241 .
we will now give an explicit description of the posets @xmath316 for @xmath354 .
similar considerations apply to @xmath36 and give an alternative description of those posets .
let @xmath355 be a @xmath18-operad with a single nullary operation @xmath356 ( such as @xmath133 or @xmath330 ) .
suppose @xmath357 and let @xmath358 .
we define the functor @xmath359 to be the composite @xmath360 here @xmath361 if @xmath362 , @xmath363 , the first map takes @xmath364 to @xmath365 , where @xmath366 with all other @xmath367 , and the second map is composition in @xmath355 .
there is a commutative diagram @xmath368^(.3){\{\pi_{a , b , c}\}}\ar[d]^\lambda & & \prod_{\{1\le a < b < c\le m\}}\sck_3\ar[d]^{\prod\lambda}\\ \scl_m\ar@{^(->}[rr]^(.3){\{\pi_{a , b , c}\ } } & & \prod_{\{1\le a < b < c\le m\}}\scl_3\ar[r]^\cong&\prod_{\{1\le a <
b < c\le m\}}\sci } \ ] ] with the horizontal arrows being full imbeddings of posets .
the proof is straight forward and left as an exercise for the reader .
it is well known that a monoidal category is equivalent to a strictly monoidal category , c.f . @xcite .
[ recall that a monoidal category is strict if the associativity natural transformations @xmath257 of theorem [ cohthm0 ] are the identities .
] we establish an analogous result for @xmath7-monoidal categories . for @xmath369
we define the poset @xmath370 to have as objects combinatorial trees as defined in ( * ? ? ?
* appendix e ) with @xmath52 input edges .
all nodes except the root node , i.e. the node at the output of the tree , are required to have more than one incoming edge .
the root node may have zero , one , or more incoming edges .
we define @xmath92 if @xmath93 can be obtained from @xmath94 by shrinking some internal edges .
we define @xmath371 to consist of the single tree : @xmath372\\ \\ { \scriptscriptstyle\bullet}\ar@{-}[dd]\\ \\ \\ } \ ] ] and @xmath373 to consist of the single tree @xmath374\\ \\ \\ \\ } \ ] ] @xmath375 & & \ar@{-}[dl]\\ & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & { } } \ ] ] is allowed in @xmath376 , while @xmath377\\ \ar@{-}[dr ] & & { \scriptscriptstyle\bullet}\ar@{-}[dl]\\ & { \scriptscriptstyle\bullet}\ar@{-}[d]\\ & & { } } \ ] ] is not allowed .
the collection @xmath378 is a right module over the associahedral operad , that is there are maps of posets @xmath379 satisfying the usual associativity and unit conditions .
this right action is defined in exactly the same way as we defined the operad structure on @xmath133 , with the single exception that when we compose with @xmath380 , we never delete the root node .
moreover , @xmath381 is also a left module over @xmath382 , the trivial operad parametrizing strictly monoidal structures .
the left action @xmath383 is given by @xmath384&\ar@{-}[ddr ] & & \ar@{-}[ddl]\\ \\ & & { \scriptscriptstyle\bullet}\ar@{-}[dd]\\ \\ & & } $ ] @xmath385&&\ar@{-}[ddr ] & & \ar@{-}[ddl]&&\ar@{-}[ddlll]\\ \\ & & & { \scriptscriptstyle\bullet}\ar@{-}[dd]\\ \\ & & & } $ ] there is a functorial construction @xmath386 together with functors @xmath387 and @xmath388 which associates to each @xmath7-monoidal category @xmath141 a strictly monoidal category @xmath389 such that 1 .
the induced maps by @xmath390 and @xmath293 on the nerves of the categories are mutually inverse homotopy equivalences , 2 .
the functor @xmath390 induces a lax homomorphism of @xmath7-spaces in the sense of @xcite , 3 .
if @xmath141 is strictly monoidal then @xmath293 is a strictly monoidal functor .
let @xmath391 where the equivalence relation is given by @xmath392 where @xmath393 , @xmath394 , @xmath395 , for @xmath396 .
the left action of @xmath382 on @xmath381 then induces a strict monoidal structure on @xmath389 .
there are functors @xmath397 and @xmath398 .
the functor @xmath390 takes a tree @xmath399 to the tree @xmath400\\ \\
{ \scriptscriptstyle\bullet}\ar@{-}[dd]\\ \\ \\ } \ ] ] the functor @xmath293 takes a tree in @xmath370 , deletes the root vertex if it has only one or no incoming edges , and regards it as a tree in @xmath133 .
the composite @xmath401 is the identity of @xmath133 .
there is a natural transformation from the composite @xmath402 to the identity of @xmath381 , given by shrinking the edge above the root vertex .
the functors @xmath390 , @xmath293 and this natural transformation are compatible with the right actions of @xmath133 on @xmath381 and on itself .
hence there are induced functors @xmath403 such that @xmath401 is the identity of @xmath141 and we have an induced natural transformation from @xmath402 to the identity of @xmath389 .
it follows that the maps induced by @xmath390 and @xmath293 on the nerves of these categories are mutually inverse equivalences .
moreover the nerve of @xmath389 is homeomorphic to the topological construction described in ( * ? ? ?
* theorem 1.26 ) applied to the nerve of @xmath141 .
there it is shown that the map induced by @xmath390 is a lax homomorphism of @xmath7-spaces .
the fact that @xmath293 is strictly monoidal , when @xmath141 is strictly monoidal , is straight forward . | this paper answers the following question : what algebraic structure on a category corresponds to an @xmath0 structure ( in the sense of stasheff ) on the geometric realization of its nerve ? in his trailblazing paper @xcite ,
stasheff constructed an infinite hierarchy of higher homotopy associativity conditions for an h - space @xmath1 .
these conditions are parametrized by a family @xmath2 of polyhedra , which came to be known as associahedra .
the vertices of @xmath3 are in 1 - 1 correspondence with all possible ways of associating an @xmath4-fold product @xmath5 , and an h - space @xmath1 is said to be an @xmath0-space if there is a map @xmath6 whose restriction to the vertices enumerates all possible ways of associating the binary multiplication on @xmath1 into an @xmath4-fold multiplication .
an @xmath7-space is known to be equivalent to a strict monoid @xmath8 and hence , up to group completion , to a loop space .
at the same time mac lane @xcite analyzed higher associativity conditions for monoidal structures on categories .
he formulated analogs of stasheff s @xmath0 conditions for categories . for @xmath9
the analogy is perfect . in particular mac lane
s @xmath10 condition is that a pentagonal diagram commute , whereas stasheff s @xmath11 is a pentagon .
however for @xmath12 the analogy breaks down .
mac lane s coherence theorem states that the @xmath10 condition implies all the higher @xmath0 conditions for @xmath12 .
by contrast for any @xmath13 one can construct h - spaces @xmath1 which satisfy the @xmath0 condition but not the @xmath14 ( or any higher ) condition . in this paper
we show how mac lane s notion of a monoidal structure on a category can be weakened so as to obtain a full hierarchy of @xmath0 conditions .
the paper is similar in spirit to @xcite where an @xmath15 hierarchy of commutativity conditions on categories was considered , analogous to those on @xmath4-fold loop spaces .
similarly to the case of associativity for categories , joyal and street ( * ? ? ?
* proposition 5.4 ) showed that if these commutativity conditions are required to hold up to natural isomorphisms , then the @xmath16 condition implies all higher @xmath15 conditions .
in @xcite we demonstrated that we could recover the entire @xmath15 hierarchy for categories by weakening these commutativity conditions to hold up to natural transformations instead .
this strategy does not work for associativity , since laplaza ( * ? ? ?
* theorem 5 ) showed that even if the associativity conditions are weakened to hold up to natural transformations , instead of isomorphisms , this laxened form of mac lane s @xmath10 condition still implies all higher @xmath0 conditions .
thus a different strategy for weakening mac lane s @xmath0 conditions for categories is required . in sections 1 and 2
we develop this strategy : we define the category theoretical analogues of stasheff s associahedra in section 1 and @xmath0-monoidal categories in section 2 . in section 3
we relate our work to that of laplaza ( and implicitly to that of tamari ) and give a simpler proof of his coherence result . in section 4
we prove a rectification result for @xmath7-monoidal categories , similar in spirit to mac lane s rectification of a monoidal category to a strictly monoidal one , by translating the rectification of an @xmath7-space to a monoid into category theory .
this paper presupposes some familiarity with the notion of operad and related concepts .
a prcis of the relevant definitions may be found in @xcite and some historical context in @xcite .
since we will be dealing exclusively with noncommutative operations , we will be using the non-@xmath17 forms of operads throughout . to forestall any possible misunderstanding ,
it should be pointed out that this paper is not related in any significant way to the notion of @xmath7-category as developed by fukaya , kontsevich , soibelman and others ( c.f .
@xcite for an overview ) . in this paper
we are discussing ordinary categories with weak monoidal structures , not some notion of a weak higher category .
we would like to take this opportunity to thank jim stasheff and stefan forcey for some helpful suggestions and references to previous work in this area . |
after analyzing the results from approximately 5 fb@xmath7 of integrated luminosity at both 7 tev and 8 tev centre - of - mass collision energies , both atlas and cms revealed the discovery of a new resonance in the @xmath8 , @xmath9 and @xmath10 decay channels , consistent with a higgs boson at a mass of approximately 125 gev , with a combined significance of more than @xmath11 @xcite .
now that the remaining 8 tev collision data , corresponding to approximately 20 fb@xmath7 of integrated luminosity , has been analyzed , this new state continues to be consistent with a higgs boson . however , there are indications that its branching ratios might deviate from those of the standard model ( sm ) higgs boson , particularly in the diphoton decay rate that is sensitive to the presence of new physics @xcite .
more data will be needed for precise determination of the branching ratios .
prior to moriond 2013 , both the cms and the atlas experiments found an enhancement in the diphoton signal strength , without significant deviations from sm values in the @xmath12 and @xmath13 signal strengths @xcite . following moriond 2013 ,
cms updated their diphoton analysis with results that were in better agreement with the sm predictions @xcite .
of interest , however , is that atlas @xcite still observes an excess in the diphoton rate at a significance of approximately @xmath14 ; and the atlas diphoton resonance provides a best fit invariant mass for the higgs boson that is larger than the measured resonance mass for the @xmath12 final state ( 126.8 gev versus 124.3 gev ) . beyond the standard model physics may be significantly constrained by comparing its predictions to the measured mass and the various measured signal strengths ( @xmath15 ) of the higgs - like state . in general ,
vector boson fusion ( vbf ) , vector boson associated production ( vh ) , and the @xmath12/@xmath13 decay modes are sensitive to modifications of the higgs boson couplings to gauge bosons ; top quark associated production ( tth ) and the fermion decay modes are sensitive to modifications of the higgs boson couplings to fermions ; and gluon fusion ( ggf ) and the diphoton ( @xmath8 ) and @xmath16 loop - induced decay modes are sensitive to the presence of new coloured and electrically charged states , respectively , that couple to the higgs boson , as well as modifications to the @xmath17 and @xmath18 couplings . in general , little higgs models without @xmath19-parity are more highly constrained by precision electroweak measurements than by the lhc higgs results @xcite , while @xmath19-parity models are primarily constrained by relic abundance considerations @xcite and lhc search results @xcite .
the recent non-@xmath19-parity little higgs model of schmaltz , stolarski and thaler , which we will refer to as the blh model @xcite , is not as constrained by precision measurements due to the presence of a custodial symmetry and a disassociation of the masses of the top partner and heavy gauge boson states .
the blh model features a global @xmath20 symmetry that is broken to a diagonal @xmath21 at a scale @xmath22(tev ) when a non - linear sigma field , @xmath23 , develops a vev ( @xmath24 ) .
the resulting 15 pseudo - nambu goldstone bosons are parameterized as two real @xmath25 triplets , @xmath26 and @xmath27 ( @xmath28 ) , two complex higgs doublets , @xmath29 and @xmath30 , and a real singlet @xmath31 . a general two higgs doublet potential
is generated in part explicitly and in part radiatively , where the quartic coupling for the higgs arises when integrating out the heavy scalar singlet , @xmath31 .
a second global symmetry of the form @xmath32 is also present , and is broken to a diagonal @xmath33 at a scale @xmath34 when a second non - linear sigma field , @xmath35 , develops a vev ( @xmath36 ) . to connect these two non - linear sigma models , the @xmath37 and @xmath38 symmetries are gauged with the same @xmath39 gauge bosons , while the @xmath40 and @xmath41 symmetries are gauged with the same @xmath42 gauge bosons .
the diagonal subgroup of @xmath43 is then identified as the standard model @xmath25 . meanwhile ,
the diagonal combination of @xmath44 is gauged by the hypercharge @xmath45 , while leaving the @xmath35 sector unchanged .
this symmetry breaking leads to an extra heavy gauge boson triplet ( @xmath46 , @xmath47 ) , with large squared masses proportional to @xmath48 , which reduces their contribution to precision electroweak observables .
fermions in the blh model , including the newly introduced top partners ( @xmath19 , @xmath49 , @xmath50 , @xmath51 , @xmath52 and @xmath53 ) , only transform under the global @xmath20 .
this leads to top partners with masses proportional only to the scale @xmath54 , lighter than the heavy gauge bosons , and results in a lesser degree of fine tuning than in other little higgs models @xcite .
the blh model has a large parameter space , allowing for a wide range of experimental signatures that could potentially reproduce either the cms or the atlas results .
since the blh model is a type i two higgs doublet model ( 2hdm ) , it also presents the possibility for a near - degeneracy between two or three physical scalar fields ( @xmath0 , @xmath2 and @xmath3 ) , which would have a large effect on the measured signal rates of the observed scalar resonance .
in particular , since a cp - odd scalar ( @xmath3 ) boson does not couple directly to pairs of gauge bosons ( @xmath13 and @xmath12 ) , it is possible for a ( nearly ) degenerate cp - odd scalar to contribute to the diphoton rate without affecting these signal strengths .
type i 2hdm are not as strongly affected by meson factory constraints as type ii 2hdm @xcite , and so the near - degenerate case presents a very interesting possibility that we explore in this paper .
although this scenario can lead to an enhancement in @xmath55 , it also leads to a large enhancement in @xmath56 . as we will show
, this effectively rules out the entirety of the near - degenerate scenario .
alternatively , the large number of new vector boson , fermion and scalar fields in the blh model can contribute to the loop - induced production and decay modes of a light higgs boson .
these additional states can also reproduce the observed enhancement of the diphoton rate without significantly affecting the non - loop - induced couplings @xcite . furthermore , mixing between flavour eigenstates in the blh model further leads to modifications of the couplings from the normal sm expressions that can also result in changes to the higgs boson signal strengths .
these three features ( extra higgs states , new gauge and fermion states , modified couplings ) combined lead to the possibility of large variances in the higgs boson signal strength rates , and the potential to reproduce either the atlas or cms measurements . in this paper
, we explore the higgs results in the blh model , accounting for all production and decay modes explored separately by atlas and cms , in two scenarios : a general case , which assumes the @xmath0 state is light ( @xmath57 gev ) and the masses of the other states ( @xmath2 and @xmath3 ) are allowed to vary , and a second case with a near - degeneracy between the masses of the @xmath0 and @xmath3 fields .
the masses of the @xmath0 and @xmath3 fields are input parameters for the model , while the @xmath2 mass is calculated from these input parameters and from the values of @xmath58 , the ratio of the vacuum expectation values of the two higgs doublets , and @xmath59 ( see eq .
[ derived_params ] ) .
therefore , the @xmath2 state may or may not be similarly near - degenerate in the latter scenario , depending on the values of the input parameters ( @xmath60 ) .
these two regions are not orthogonal , as the general scenario does allow for the possibility of near degeneracy in the masses of the @xmath3 and @xmath2 states ; this will be discussed further in sec .
[ sec : general_scenario ] . in sec .
[ sec : calc ] , we describe the formalism we use in our calculations of the higgs results , while in sec . [
sec : model ] we describe the details of the blh model that are relevant to our calculations . in sec .
[ sec : results ] , we compare the blh model predictions to the measured results from atlas and cms in both scenarios .
we summarize our results in sec .
[ sec : summary ] .
as shown in @xcite , the scalar interactions of a little higgs model lagrangian can be normalized to the form of the sm expressions by introducing scaling factors @xmath61 , such that @xmath62 for fermion species @xmath54 , vector bosons @xmath63 and scalars @xmath64 .
the @xmath65 label denotes the @xmath0 , @xmath2 , and @xmath3 ( with an appropriate @xmath66 factor ) for the fermion interactions , and the @xmath0 and @xmath2 for the vector boson and scalar boson interaction terms .
the parameters ( @xmath59 , @xmath67 , @xmath68 , @xmath61 ) are model dependent .
the expression for the vev , @xmath59 , will be discussed further in sec .
[ sec : model ] .
a higgs boson of approximately 125 gev decays predominantly to pair produced , kinematically accessible states , such as @xmath69 and @xmath70 , and to one on - shell and one off - shell vector boson ( @xmath12 and @xmath13 ) .
decays to diphotons , digluons and @xmath16 also occur through loop interactions @xcite . in the blh model , all three physical higgs states ( @xmath0 , @xmath2 and @xmath3 )
can decay to light fermions and to loop - induced final states , but only the cp - even states ( @xmath0 and @xmath2 ) can decay to pairs of weak gauge bosons , as shown in figures [ fig : partialdirect ] , [ fig : partialyy ] and [ fig : partialgg ] . in our calculations , we account for the contributions from all three higgs states , and include all new particle states of the model that contribute to the loop diagrams . when the scalar couplings are expressed in the form of eq .
[ eq : blhlag ] , the partial decay widths of the @xmath71 can be written in terms of the sm calculated values ( we use the values calculated in @xcite ) multiplied by some combination of scaling factors . for direct decays ,
this is straightforward . for
loop - induced decays the scaling factors must also include the loop factors .
the direct decays are given by @xmath72 where @xmath73 the factor @xmath74 accounts for the differences between the vev in the blh model and the sm vev , such that @xmath75 , and is discussed in further detail in the following section .
the factor @xmath76 is the ratio of the @xmath77 boson mass calculated in the blh model and in the sm , and appears when making the replacement @xmath78 in the expression for the @xmath79 partial width @xcite .
the factors of @xmath80 are given by @xmath81 and @xmath82 , and account for differences in the expressions of the couplings of the light vector bosons to fermions in the blh model .
for our chosen parameter values , we found that these factors amount to at most a 1% correction to the @xmath83 and @xmath84 couplings . although these corrections are numerically small , it is appropriate to include them when the experimental measurements rely on leptonic decay modes of the vector bosons .
thus , we account for @xmath85 and @xmath86 decays ( @xmath87 ) in the calculation of the @xmath12 and @xmath13 decay widths , corresponding to the experimental results from atlas and cms @xcite .
likewise , the @xmath85 decay ( @xmath87 ) is taken into account in the @xmath88 channel as measured by atlas and cms @xcite , by including a factor of @xmath89 in the expression for @xmath90 in equations [ eq : loopdec ] and [ eq : loopdec_r ] below .
, @xmath2 and @xmath3 states.,scaledwidth=50.0% ] the expressions for the partial widths of the loop - induced decay modes are given by @xmath91 where @xmath92 refers to all fermions with a colour charge , including heavy , vector - like quark states .
this is also the process for the gluon - fusion production mode.,scaledwidth=30.0% ] in this form , the @xmath93 terms in each channel account for the contributions from loops involving only the new ( non - sm ) particles of the blh model .
thus , we find the expressions @xmath94 the expressions for the @xmath95 terms in the blh model are given by @xcite : @xmath96 q_f n_{c}^f f_{f}(\tau_f,\lambda_f)\cr\cr a_{v , z\gamma}^{blh } & \equiv & y_{s_0 vv } \
, g_{zvv}^{blh } \
, q_v \ , f_{v}(\tau_v,\lambda_v)\cr\cr a_{s , z\gamma}^{blh } & \equiv & y_{s_0 ss } \ , g_{zss}^{blh } \ , q_s \ , f_{s}(\tau_s,\lambda_s)\cr\cr a_{f , gg}^{blh } & \equiv & y_{s_0\bar{f}f } \ ,
f_{f}(\tau_f ) , \label{eq : loopdec3}\end{aligned}\ ] ] where the sm expressions from the denominators in eq .
[ eq : loopdec_r ] can be found by setting the @xmath61 scaling factors in eq .
[ eq : loopdec3 ] to unity . in these expressions
, @xmath97 is the electric charge of the particle , and @xmath98 is the number of colours ( 3 for quarks , 1 for leptons ) .
the form factors @xmath99 , @xmath100 and @xmath101 ( and similarly for @xmath102 , @xmath103 and @xmath104 ) are found by integrating over the fermion ( @xmath54 ) , gauge boson ( @xmath63 ) and scalar ( @xmath64 ) loops , respectively .
these are given by @xcite : @xmath105 where @xmath106 ^ 2 , & x \ge 1\\ -\displaystyle\frac{1}{4}\left[\ln\left(\frac{1+\sqrt{1-x}}{1-\sqrt{1-x}}\right ) -i \pi\right]^2 , & x < 1\end{array}\right .
\cr\cr\cr g(x ) & = & \displaystyle\left\{\begin{array}{l l}\sqrt{x-1}\arcsin(1/\sqrt{x } ) , & x \ge 1\\ \displaystyle\frac{\sqrt{1-x}}{2}\left[\ln\left(\frac{1+\sqrt{1-x}}{1-\sqrt{1-x}}\right ) -i \pi\right ] , & x < 1\end{array}\right .
\nonumber \\\end{aligned}\ ] ] in the above , @xmath107 and @xmath108 , for field @xmath54 , @xmath63 or @xmath64 , with corresponding mass @xmath109 . in the limit that @xmath110 , these factors approach the values of @xmath111 , @xmath112 , and @xmath113 . in the blh model
, it is useful to consider the contributions of each of three sectors - gauge , fermion and scalar - to the diphoton rate .
this can be performed by summing up all contributing states @xmath114 , @xmath115 and @xmath116 .
the relative value and sign of these terms , @xmath117 , @xmath118 , and @xmath119 , represent the size of the contribution of that sector of particle states to the diphoton effective coupling .
the diphoton effective coupling is dominated by gauge boson loops in the sm , with a sub - dominant contribution from fermions ( predominantly the top quark ) .
this is not expected to change significantly in the blh model in regions of parameter space that produce sm - like signal strength ratios .
the reason for this is that , for example , a decrease in the gauge contribution to the diphoton rate would likely correspond to a smaller @xmath120 rate , assuming a negligible contribution from the heavy @xmath121 . for new fields
whose masses are proportional to @xmath54 or @xmath122 , such as the new vector - like quarks and heavy gauge bosons in the blh model , the scaling factors behave as @xmath123 , where @xmath124 .
thus , while the factor of @xmath125 may increase in magnitude with increasing mass of the new state , the scaling factor , @xmath61 , decreases at a faster rate , and very high mass states typically have only a small contribution to the loop factors .
it is also possible that loops involving both @xmath126 and @xmath47 or @xmath127 and @xmath128 contribute to the @xmath129 and @xmath130 decay widths due to vertices of the form of @xmath131 and @xmath132 .
however , since the @xmath133 and @xmath134 couplings of the blh model are suppressed at @xmath135 , these mixed loops can be safely neglected in this model .
in addition to the decays discussed above , the heavy cp - even scalar , @xmath2 , and the cp - odd scalar , @xmath3 , of the blh model , when kinematically allowed , may also decay to top quarks ( @xmath136 ) , pairs of scalars ( @xmath137 , @xmath138 ) , or a scalar and a gauge boson ( @xmath139 and @xmath140 , where in the case of off - shell decays , @xmath141 and @xmath142 ) @xcite .
these decays have the effect of lowering the @xmath2 and @xmath3 branching ratios to the standard modes considered above .
hence , their contributions to the higgs signal strengths under consideration in this study are lowered , particularly for @xmath2 and @xmath3 masses above roughly 200 - 300 gev @xcite .
thus , it is important to include their effects in our analysis .
the partial widths for the top quark decays of @xmath143 were calculated using the data from @xcite , scaled in a similar manner as the direct decays of equations [ directdecays ] and [ directdecays2 ] : @xmath144 where @xmath145 and @xmath146 for @xmath147 is an additional factor that accounts for the difference in kinematics of the cp - even @xmath2 and cp - odd @xmath3 decays .
the scalar - scalar and scalar - gauge partial widths were calculated using hdecay @xcite in a generic 2hdm using the mixing angle parameters @xmath148 and @xmath149 for a range of scalar masses from 125 gev to 1200 gev in 5 gev increments . to calculate the widths for a particular set of blh model parameters ,
interpolation was used on the discrete set of 2hdm widths and the results were scaled according to @xmath150 where @xmath151 where the expressions for the 2hdm couplings , @xmath152 , used in hdecay are written in appendix e of @xcite .
we have verified that these couplings agree with those listed in appendix a of @xcite .
the couplings , @xmath153 , of the blh model are directly related to the scaling factors , @xmath61 , and are listed in eq .
[ y_sss ] and [ g_ssv ] of section [ sec : model ] .
expressions similar to those for the decay modes exist for the production modes . at the lhc ,
the primary production mode is through gluon - gluon fusion ( ggf ) , which occurs via the reverse of the diagram in figure [ fig : partialgg ] .
the other production mechanisms at the lhc include vector boson fusion ( vbf ) , associated production with a vector boson ( vh ) , and associated production with top quarks ( tth ) .
the diagrams for these subdominant processes are given in fig .
[ fig : production ] . in the blh model
, the cross sections are modified from the sm expressions in a similar manner as the decay widths , and can be expressed as a multiplicative factor times the sm values , such as : @xmath154 the factor modifying the gluon fusion production mode , @xmath155 , is the same as for the @xmath156 decay mode , defined in eq .
[ eq : loopdec_r ] . since @xmath157 and @xmath158 in the blh model due to custodial symmetry ,
the vbf and vh expressions are essentially independent of whether the gauge boson is a @xmath159 or a @xmath77 .
the following subsections deal with the determination of the scaling factors , @xmath61 , for all neutral higgs states ( @xmath160 ) coupling to the charged scalars ( @xmath161 ) , the gauge bosons ( @xmath162 ) , the sm fermions , and the additional fermions ( @xmath163 ) of the blh model .
the blh model higgs fields , @xmath29 and @xmath30 , form a two higgs doublet potential that undergoes spontaneous symmetry breaking .
the most basic form of this potential is given below as in @xcite , and is sufficient to understand ewsb in the model : @xmath164 in this form , each of the higgs doublets " are written as * 4 * s of @xmath165 , which have the same degrees of freedom and hypercharge values as the standard two higgs doublets from @xcite .
additionally , while there are other quartic terms present in the blh model ( namely @xmath166 , @xmath167 , @xmath168 , and @xmath169 ) , their coefficients are generated at loop level and do not significantly affect the details of electroweak symmetry breaking . as these terms have a small effect on the diagonalization , they are ignored in the determination of the mass eigenstates .
they are not ignored in the determination of the couplings , however , as they comprise the dominant contribution to the interactions of the scalar triplets with the higgs fields .
the parameters in this potential are generated in part explicitly and in part radiatively , as with other little higgs models , and spontaneous symmetry breaking occurs for parameter values that satisfy @xmath170 .
the minimization condition is achieved by shifting the first component of each of @xmath29 and @xmath30 by a respective vacuum expectation value ( vev ) , @xmath171 and @xmath172 .
these vevs can be parameterized by a mixing angle , @xmath173 , where @xmath174 is related to the standard model vev via the relation @xmath175 ( the expression for @xmath74 will be addressed in the next subsection ) . diagonalizing the mass matrix for the scalar sector results in three physical neutral scalar fields ( @xmath0 , @xmath2 and @xmath3 , along with the unphysical @xmath176 ) , and two physical charged scalar fields ( @xmath177 , along with the unphysical @xmath178 ) , parameterized by an angle @xmath179 such that @xmath180 & = & \cos\alpha \ , h_0 - \sin\alpha \ , h_0 + v \sin\beta \nonumber\\ h_{2}[1 ] & = & \sin\alpha \ , h_0 + \cos\alpha \ , h_0 + v \cos\beta \nonumber\\ h_{1}[2 ] & = & -\cos\beta \ , a_0 + \sin\beta \ , g_0 \nonumber\\ h_{2}[2 ] & = & + \sin\beta \ ,
a_0 + \cos\beta \ ,
g_0 \nonumber\\ h_{1}^\pm & = & -\cos\beta \ , h^\pm + \sin\beta \ , g^\pm \nonumber\\ h_{2}^\pm & = & + \sin\beta \ , h^\pm + \cos\beta \ , g^\pm .\end{aligned}\ ] ] the four parameters in the higgs potential ( @xmath181 , @xmath182 , @xmath183 and @xmath184 ) can be replaced by a more phenomenologically accessible set consisting of the masses of the @xmath0 and @xmath3 states , along with the mixing angle @xmath185 and the vev , @xmath59 . in this parameterization
, we find the following expressions : @xmath186 since the quartic interaction , with coupling @xmath184 , is produced when the heavy singlet state ( @xmath31 ) is integrated out , it is related to the fundamental parameters @xmath187 and @xmath188 via @xmath189 .
the parameters @xmath187 and @xmath188 are the coefficients of the quartic potential , defined in eq .
9 of @xcite , and must both be non - zero to achieve collective symmetry breaking and generate a higgs quartic coupling . rather than expressing @xmath190 in terms of these two free parameters ,
we instead choose to parameterize it in terms of @xmath184 and a single free parameter , @xmath191 , as shown in the last line of eq .
[ derived_params ] .
there exist a number of theoretical constraints that can be placed on these parameters , primarily due to perturbativity requirements .
the value of the mixing angle @xmath185 is limited by two constraints , the first of which is the requirement that @xmath192 , leading to an upper bound of @xmath193 a lower bound also exists , and is set by examining the radiatively - induced contributions to @xmath181 and @xmath182 in the model , which suggest that @xmath194 @xcite .
furthermore , there are limits on @xmath191 from requiring that @xmath195 are real valued and @xmath196 , such that @xmath197 which leads to the bounds @xmath198 the blh model also contains two physical , real triplets , with charged fields @xmath127 and @xmath128 that are relevant to the diphoton and @xmath16 decay modes .
these scalar triplet fields obtain a contribution to their mass from the explicit symmetry breaking terms in the model , as defined in eq .
38 of @xcite , that depend on the parameter @xmath199 gev .
however , their masses are dominated by contributions from the coleman - weinberg potential involving gauge boson loops for @xmath127 and hypercharge boson loops for @xmath128 . the dominant contributions to their masses are given by @xmath200 where @xmath201 is the compositeness scale , and @xmath202 and @xmath203 are taken as @xmath204 factors that account for the details of the cancellation of the gauge logarithmic , and hypercharge quadratic divergent loops , respectively , at the scale @xmath205 . in our calculations , we take into account all contributions up to @xmath135 , including the subdominant explicit symmetry breaking mass terms @xcite , but do not show them here for brevity . the factors
@xmath206 and @xmath207 are the gauge couplings for @xmath39 and @xmath42 , respectively , which will be discussed in further detail in section [ sec : gauge ] .
the scaling factors for the charged scalar interactions are given by @xmath208 ( c_\beta s_\alpha + c_\alpha s_\beta ) - 32 \kappa_s \lambda_0 m_\sigma^2 \log\left[\frac{\lambda^2}{m_\sigma^2}\right ] ( 3 c_\beta s_\alpha + 2 c_\alpha s_\beta ) \cr & & + 128 \pi^2 c_\beta s_\beta ( -6 f^2 \lambda_0 ( c_\alpha c_\beta + s_\alpha s_\beta ) + m_{a_0}^2 ( c_\alpha c_\beta - s_\alpha s_\beta ) ( c_\beta^2 - s_\beta^2))\biggr)\cr\cr y_{h_0h^+h^- } & \approx & \frac{v^2}{(768 f^2 m_{a_0}^2 \pi^2 ) } \biggl(9 \kappa_g f^2 g_a^2 g_b^2 \log\left[\frac{\lambda^2}{m_{w^\prime}^2}\right ] ( c_\beta c_\alpha - s_\alpha s_\beta ) + 32 \kappa_s \lambda_0 m_\sigma^2 \log\left[\frac{\lambda^2}{m_\sigma^2}\right ] ( 3 c_\beta s_\alpha + 2 c_\alpha s_\beta)(c_\beta^2 - s_\beta^2 ) \cr & & + 128 \pi^2 c_\beta s_\beta ( -6 f^2 \lambda_0 ( c_\beta s_\alpha - c_\alpha s_\beta ) + m_{a_0}^2 ( c_\beta s_\alpha + c_\alpha s_\beta ) ( c_\beta^2 - s_\beta^2))\biggr)\cr\cr y_{h_0 a_0 a_0 } & \approx & \frac{v^2}{(768 f^2 m_{a_0}^2 \pi^2 ) } \biggl(9 \kappa_g f^2 g_a^2 g_b^2 \log\left[\frac{\lambda^2}{m_{w^\prime}^2}\right ] ( c_\beta c_\alpha - s_\alpha s_\beta ) + 32 \kappa_s \lambda_0 m_\sigma^2 \log\left[\frac{\lambda^2}{m_\sigma^2}\right ] ( 3 c_\beta s_\alpha + 2 c_\alpha s_\beta)(c_\beta^2 - s_\beta^2 ) \cr & & + 36 \kappa_y g^2 \tan^2\theta_w \lambda^2 ( c_\beta c_\alpha - s_\beta s_\alpha ) + 128 \pi^2 c_\beta s_\beta ( -6 f^2 \lambda_0 ( c_\beta s_\alpha - c_\alpha s_\beta ) + m_{a_0}^2 ( c_\beta s_\alpha + c_\alpha s_\beta ) ( c_\beta^2 - s_\beta^2))\biggr)\cr\cr y_{h_0 h_0 h_0 } & \approx & \frac{v^2}{(768 f^2 m_{h}^2 \pi^2 ) } \biggl(9 \kappa_g f^2 g^4 \cot^2\theta_g \log\left[\frac{\lambda^2}{m_{w^\prime}^2}\right ] ( c_\beta c_\alpha - s_\alpha s_\beta)\cr & & + 96 \kappa_s \lambda_0 m_\sigma^2 \log\left[\frac{\lambda^2}{m_\sigma^2}\right ] ( c_\beta c_\alpha -s_\beta s_\alpha + s_\alpha c_\alpha ( s_\beta c_\alpha -c_\beta s_\alpha ) ) + 12 \kappa_y g^2 \tan^2\theta_w \lambda^2 ( c_\beta c_\alpha - s_\beta s_\alpha ) \cr & & -3072 \pi^2 c_\beta s_\beta
( f^2 \lambda_0 ( c_\beta c_\alpha -s_\beta s_\alpha + 3s_\alpha c_\alpha ( s_\beta c_\alpha -c_\beta s_\alpha ) ) + m_{a_0}^2 ( ( c_\alpha^2-s_\alpha^2)(c_\beta s_\alpha + s_\beta c_\alpha ) ) ) \biggr ) \end{aligned}\ ] ] where @xmath209 , @xmath210 , and @xmath211 , for @xmath212 and @xmath213 .
the factor of @xmath214 is also taken as an @xmath204 factor that accounts for the details of the cancellation of the logarithmic divergence involving scalar loops .
here we have also neglected contributions from higher orders in the expansion of @xmath215 , and terms proportional to the small explicit symmetry breaking parameters in the blh model @xcite .
the gauge couplings @xmath206 and @xmath207 , associated with @xmath216 , can be parameterized in a more phenomenological fashion in terms of a mixing angle @xmath217 ( @xmath218 ) and the @xmath25 gauge coupling , @xmath219 . furthermore ,
@xmath220 and @xmath221 , the gauge coupling associated with the @xmath45 symmetry , can be parameterized as in the sm in terms of the fundamental charge , @xmath222 , and the weak mixing angle , @xmath223 , such that @xmath224 , and @xmath225 .
the masses of the gauge bosons in the blh model are given by : @xmath226 where @xmath227 and @xmath228 and @xmath229 . while both @xmath230 and @xmath231 obtain corrections of @xmath232 , which do not vanish as @xmath233 , the model retains custodial symmetry and such factors cancel out in their contribution to the @xmath234 parameter .
additionally , corrections to the @xmath235 couplings are proportional to @xmath236 , and do vanish for @xmath233 @xcite .
we use the experimental value of the fine structure constant , @xmath237 , to determine the value of @xmath222 ( @xmath238 ) , and calculate @xmath59 and @xmath239 from the experimental values of @xmath240 and the pole mass of the @xmath159 boson , by exploiting the custodial symmetry relation , @xmath241 .
accounting for the contributions from the @xmath77 and @xmath121 in the process @xmath242 , gives an expression for the vev , @xmath59 , in terms of the sm vev , such that @xmath243 where @xmath244 combining this with the expression for the @xmath159 pole mass ( @xmath245 ) , the weak mixing angle in the blh model can be expressed as @xmath246 the scaling factors important to the vbf and vh production modes , and to the diboson decay modes ( @xmath247 ) are given by @xmath248 we see that @xmath249 is the most important parameter controlling these scaling factors and note that the coupling to the @xmath121 gauge boson is suppressed at @xmath250 . the heavier higgs states , @xmath3 and @xmath2 , may also decay to combinations of scalar and gauge states , such as @xmath140 and @xmath139 .
these couplings do not have an equivalent expression in the sm , so we provide the explicit blh coupling expressions below : @xmath251 the yukawa terms for the light fermions in the blh model lagrangian take the form @xmath252 where @xmath253 is the second component of the triplet of @xmath254 generators corresponding to the @xmath255 subgroup of @xmath256 , and performs the charge conjugation of the higgs fields for interactions with the down - type quarks .
this identifies the blh model as a type i 2hdm , with light fermion masses given by @xmath257 and scaling factors of the form @xmath258 the following expressions give the leading order contributions in terms of @xmath249 and @xmath259 : @xmath260 the blh model also includes a number of heavy , vector - like quarks that act to protect the higgs boson mass from developing quadratic divergences from the top quark loops .
the yukawa interactions for the heavy fermions , which includes the top quark , take the form given in eq .
46 of @xcite : @xmath261 these couplings can be parameterized in terms of @xmath262 and two mixing angles , defined such that @xmath263 and @xmath264 . the value of @xmath262 is then fixed through the measured top quark mass . as we showed in @xcite , analytic methods for determining
the mass eigenstates for the heavy quarks fail in the region where @xmath265 , due to the degeneracy between the @xmath19 and @xmath52 states .
this can clearly be seen by observing a slice of the @xmath266 parameter space , where @xmath267 , as in fig .
[ fig : fmass ] . as a result ,
we use numerical diagonalization to determine the mass eigenstates of the heavy quarks , and thus their coupling to the @xmath0 , @xmath2 and @xmath3 states , for all values of @xmath268 and @xmath269 .
therefore , we do not provide analytic solutions for the relevant scaling factors .
however , we can provide some insight into their contributions to the higgs production and decay . since the vector - like quarks ( @xmath163 ) obtain contributions to their masses proportional to @xmath54 , the lowest order contribution to their scaling factor is @xmath270 , and their contribution to the loops in the @xmath271 and @xmath272 effective vertices drop off rapidly with increasing @xmath54 .
in particular , the scaling factors for the @xmath49 and @xmath51 are suppressed at @xmath273 , making these two states effectively decouple from the process .
the scaling factors for the @xmath52 and @xmath19 are the largest for the set , but with values typically smaller than 0.02 , while for the @xmath53 and @xmath50 the scaling factor is suppressed by an additional factor of about 10 .
since the top quark is generated from the heavy quark yukawa terms , it gets a mass proportional to @xmath59 , and thus the dependence of the scaling factor for the top quark on @xmath185 and @xmath179 behaves at lowest order like those of the light up - type quarks .
this dependence on @xmath185 and @xmath179 , characteristic of a type i 2hdm , results in the most significant deviations from the sm in the fermion contributions to the loop interactions .
we use the following fixed values in our calculations @xcite : @xmath274 of note , the parameters @xmath275 are explicit symmetry breaking mass terms in the full blh model scalar potential @xcite that are small and provide a negligible contribution to the interactions examined .
as discussed in @xcite , these parameters are introduced to break all the axial symmetries in the higgs potential , giving positive masses to all scalars . in particular
, the @xmath276 state receives a mass equal to @xmath199 gev .
since this state couples to top quarks and would decay predominantly to @xmath277 quarks , it may be visible in @xmath278 final states .
the value of @xmath279 would strongly affect the rate for this process , but this is not relevant for the signal strengths considered here . to be thorough ,
we randomize all remaining parameters in the blh model over the ranges : @xmath280 where @xmath281 we determine min(@xmath122 ) as follows .
electroweak precision observables ( ewpo ) place constraints on the mass of the heavy gauge bosons , as a function of the heavy gauge boson mixing angle , @xmath282 , as shown in @xcite for a light higgs boson mass of 125 gev .
this lower limit on the @xmath121 mass in turn determines min(@xmath122 ) for each value of @xmath282 via the third line in equation [ gauge_masses ] .
so the parameter sets we generate satisfy this ewpo constraint . since the mass of the observed resonance is not precisely known , and especially since the @xmath12 and @xmath8 mass peak values in the atlas measurements are distinct , we allow the mass of the light higgs boson to vary over a small range .
all other values are calculated from this base set , as described in the preceding sections of this paper .
additionally , we separate our analysis into two scenarios - a general case " , where @xmath283 varies up to 700 gev , and a near - degenerate " case , where @xmath283 can take a maximum value of @xmath284 gev .
both cms and atlas have now published updated signal strengths for each given production mode and final state using the 7 tev and 8 tev data .
we have calculated the expected value for these signal strength ratios in the blh model for 60k parameter sets for each of the general and the near - degenerate cases .
each set corresponds to a point on the plots which follow .
specifically , we calculate : @xmath285 where @xmath286 the sum over the index @xmath287 accounts for all contributing production modes , including the @xmath3 and @xmath2 mediated production . since the signal strengths from cms and atlas are calculated based on the experimentally determined best fit mass , our @xmath288 values are normalized to the expected results for a sm higgs boson with the mass given in the experimental study , and are weighted by the 7 tev and 8 tev integrated luminosities , @xmath289 and @xmath290 respectively .
summaries of the published results from cms and atlas can be found in tables [ tab : cmsresults ] and [ tab : atlasresults ] , respectively .
while we include contributions from the @xmath3 and @xmath2 in our determination of the signal strength ratios , some constraints currently exist on heavy scalars .
in particular , both cms @xcite and atlas @xcite have placed 95% c.l .
constraints on the decays of a heavy higgs boson to @xmath120 , and cms provides 95% c.l .
exclusions on @xmath291 for the @xmath292 final state for masses up to 150 gev @xcite .
we have incorporated these results and show parameter points that violate these constraints in light red / pink to distinguish them from the unconstrained results .
we present our results using two different @xmath293 measures in figures [ fig : rdeg : muyyvsmuww ] through [ fig : norm : flvsmt6 ] . in order to show the blh parameter sets that produce results in better agreement with the measured data than the sm
, we calculate @xmath294 , where @xmath295 where the sum over the index @xmath287 includes all channels listed in tables [ tab : cmsresults ] and [ tab : atlasresults ] and @xmath296 .
this measure is employed in figures [ fig : deg : tbvssapb ] , [ fig : norm : tbvssapb]-[fig : norm : muwwvsmuzz ] , and [ fig : norm : flvsmt6 ] . for figures [ fig : rdeg : muyyvsmuww]-[fig : rdeg : tbvssapb ] , [ fig : rdeg : muyyvsmutata ] , and [ fig : rnorm : muyyvsmuww]-[fig : rnorm : hmuyy1vsamuyy1 ] , we calculate an alternate measure @xmath297 , defined as : @xmath298 where the sum in this case is only over the @xmath8 and @xmath13 channels from tables [ tab : cmsresults ] and [ tab : atlasresults ] and @xmath299 this measure focuses only on the most precisely measured signal strength values in order to identify regions of the blh parameter space which favour either the distinctive cms diphoton and @xmath13 results or those of atlas .
the values of @xmath300 in the near - degenerate and general scenarios are listed in table [ tab : chisq ] . in all figures , both
the atlas ( right panel ) and cms ( left panel ) results are included , in order to show the differences between the atlas and cms preferred regions . [
cols="^,^,^,^,^,^",options="header " , ] we define a near - degenerate scenario such that the mass of the @xmath3 is constrained to be within a few gev of the mass of the lighter scalar higgs state ( @xmath0 ) , which allows it to contribute to any signal strength that involves fermionic couplings in both the production mode ( ggf , tth ) and the decay mode ( @xmath69 , @xmath70 , @xmath8 and @xmath16 ) . depending on the other parameters
, the mass of the heavier scalar higgs state ( @xmath2 ) can also be nearly degenerate with the @xmath0 , and thus also contribute significantly to the measured signal strengths , including @xmath13 and @xmath12 .
ultimately , this scenario can reproduce an enhanced diphoton signal strength , as measured by atlas , due to multiple scalar states contributing .
however , at the same time , the @xmath70 mode is also enhanced , beyond what is experimentally observed .
so the scenario is largely ruled out .
however , it remains useful as a tool to study the contribution of the different scalar states to the various processes . since several scalar fields may be contributing significantly to the diphoton production rate
, it is useful to consider the individual contribution to @xmath55 from each of the @xmath0 , @xmath3 and @xmath2 . for this purpose ,
we define @xmath301 where @xmath302 includes only the cross section from the production and decay of the stated scalar , @xmath64 .
we first focus on the issue of the cms results versus those of atlas .
figure [ fig : rdeg : muyyvsmuww ] shows that both the more sm - like @xmath55 from cms and the excess @xmath55 from atlas can be accommodated in the near - degenerate scenario .
together , figures [ fig : rdeg : hmuyyvshmuww ] , [ fig : rdeg : hmuyyvsamuww ] and [ fig : rdeg : tbvssapb ] provide insight into the reproduction of the data of each experiment within the blh parameter space , in terms of the particular scalar states that are being produced , and in terms of the parameters @xmath303 and @xmath58 . those parameter sets that yield results consistent with the cms data involve a strongly suppressed @xmath0 contribution to the diphoton signal strength ratio , @xmath304 , as seen in the left panels of figures [ fig : rdeg : hmuyyvshmuww ] and [ fig : rdeg : hmuyyvsamuww ] .
the reduced @xmath0 contribution results from a suppression of the coupling between the @xmath0 and the @xmath77 boson that occurs for small to moderate values of @xmath303 ( see equation [ eq : ygauge ] for @xmath179 and @xmath185 dependence ) , as seen in the left panel of figure [ fig : rdeg : tbvssapb ] .
the @xmath0 coupling to the top quark is slightly enhanced for this range of @xmath303 and @xmath58 values , which results in destructive interference with the gauge boson contribution to the diphoton effective coupling .
the diphoton rate as measured by cms , which is near that of the sm , is instead understood as predominantly a result of contributions from both the @xmath3 and @xmath2 decays .
the @xmath13 signal strength ratio , as measured by cms , is produced from a combination of the decays of the @xmath0 and @xmath2 .
the right panels of figures [ fig : rdeg : hmuyyvshmuww ] , [ fig : rdeg : hmuyyvsamuww ] and [ fig : rdeg : tbvssapb ] give equivalent information for the reproduction of the atlas data . here
it is primarily blh parameter sets with large @xmath303 and small @xmath58 that are favoured by the data .
this range of the parameters yield a diphoton signal ratio approximately 40% due to production and decay of the cp - odd state , @xmath3 , with an admixture of decays of the @xmath2 and @xmath0 providing the rest of the contribution .
the @xmath13 production arises primarily from the @xmath0 , with a smaller contribution from the @xmath2 , which is proportional to @xmath305 .
the results presented in figures [ fig : rdeg : muyyvsmuww ] through [ fig : rdeg : tbvssapb ] only include the diphoton and @xmath13 signal rates in the @xmath293 measure .
the picture is drastically different when accounting for the full data set , as shown in figure [ fig : deg : tbvssapb ] . comparing the full @xmath306 to that of @xmath307 ,
it is clear that there are almost no parameter values in which the near - degenerate blh model is a better fit to the data than is the sm , nor is it even close .
this is due entirely to the @xmath56 signal strength ratio , as shown in figure [ fig : rdeg : muyyvsmutata ] .
when the cp - odd scalar field is nearly degenerate with the light higgs boson , it contributes a significant amount to the production of @xmath70 pairs . for most parameter regions producing a signal consistent with the diphoton rates observed by either cms or atlas , the @xmath70 signal strength ratio would need to be three to five times larger than the sm rate , and several sigma larger than the respective measured values .
these calculations do not include the effect of interference .
there are no interference effects between the cp - even ( @xmath0 and @xmath2 ) states and the cp - odd ( @xmath3 ) state due to cp invariance . in the case
where the heavier cp - even state , @xmath2 , is nearly degenerate with the @xmath0 , the mass difference between the two is typically large enough compared to the corresponding boson widths that the interference effects can be neglected .
this was the approach in @xcite wherein the authors point out that the experimental mass resolutions are significantly larger than the higgs widths and , so , the assumption does not significantly constrain their analysis . since we are democratic in our scan
, we do include some parameter points in which the mass separation is small enough such that interference effects should be included , however we find that the value of @xmath308 , the contribution of the cp - odd scalar to the di - tau signal strength ratio , increases for decreasing @xmath309 .
the region with small @xmath309 is thus ruled out due to considerations of the cp - odd scalar contributions alone , suggesting that the inclusion of interference effects will not change our conclusions . in the general scenario ,
we allow the mass of the cp - odd scalar to vary between the mass of the light higgs boson and 700 gev .
thus , there is some overlap between the regions of parameter space explored in this and the near - degenerate scenario .
however , since the @xmath70 results exclude the entirety of the near - degenerate scenario , the points of overlap will be greyed out in figures [ fig : norm : tbvssapb ] through [ fig : norm : muwwvsmuzz ] , which are coloured based on @xmath310 .
this is discussed in further detail below .
figures [ fig : norm : tbvssapb ] through [ fig : norm : muwwvsmuzz ] distinctly show that the blh model is a better fit to the cms data than is the sm for a significant portion of the parameter space , and is a better fit to the atlas data than is the sm for a smaller set of parameter points .
the region of better fit occurs for values of @xmath311 , @xmath312 gev and @xmath313 gev for cms , as shown on the left side of figures [ fig : norm : tbvssapb ] and [ fig : norm : flvsma0 ] , respectively .
the right side of figure [ fig : norm : flvsma0 ] indicates that agreement between the blh model and atlas results occurs predominantly for @xmath312 gev and for @xmath314 gev , with an extended region of agreement between @xmath315 gev for parameter sets on the lower boundary of @xmath316 gev .
the favoured large values of @xmath303 for both experimental result sets are understandable in order to achieve @xmath317 , and larger values of @xmath54 reduce the contribution from higher order terms in the expansion of @xmath215 in the couplings .
both the atlas and cms results can be realized in the general scenario , as is evident in figure [ fig : norm : muyyvsmuww ] , where we show @xmath318 versus @xmath319 .
figure [ fig : norm : tbvsma0 ] shows that the parameter sets which produce a value of @xmath320 , as in the atlas side of figure [ fig : norm : muyyvsmuww ] , occur more frequently near the boundaries of excluded regions of @xmath58 and @xmath283 . in both the cms and atlas cases ,
the region with @xmath321 gev is mostly ruled out .
this means the @xmath3 contribution to @xmath56 is small for parameter points that are allowed and makes that signal strength more consistent with the sm , as shown in figure [ fig : norm : muyyvsmutata ] .
these results are therefore effectively orthogonal to the near - degenerate scenario , even though there is a small amount of overlap between the parameter spaces generated . to enhance the excluded ( pink ) points in figure [ fig : norm : tbvsma0 ] ,
a dashed line of the approximate region enclosing the excluded points has been included . for low values of @xmath322 ( less than 250 gev )
, parameter sets are excluded primarily due to the constraint on @xmath323 production . for larger values of @xmath322 ,
the exclusion comes entirely from the @xmath5 search . in the areas with overlap between excluded and non - excluded points ,
there is large variation in the @xmath324 and @xmath325 values due to influences from the other fundamental parameters ( such as @xmath54 and the scalar sector parameters ) , and large variations in the total width of the @xmath2 .
this allows for many parameter sets to avoid exclusion , with sufficiently low branching ratio or production rate , or both .
the small region of pure exclusion at approximately @xmath326 and @xmath327 gev occurs due to all parameter points having a sufficiently large production rate and branching ratio for exclusion .
this results in the appearance of an apparently isolated region with a high density of excluded parameter sets .
however , this appearance is simply a result of the two dimensional display of values that depend on multiple degrees of freedom in which the excluded points are displayed as the lowest layer .
as is clear from figures [ fig : norm : muyyvsmuww ] ( @xmath55 , @xmath318 ) , [ fig : norm : muyyvsmutata ] ( @xmath55 , @xmath56 ) , [ fig : norm : muyyvsmubb ] ( @xmath55 , @xmath328 ) , and [ fig : norm : muyyvsmuzy ] ( @xmath55 , @xmath329 ) , the general scenario of the blh model includes parameter sets that can produce signal strength ratios in better agreement with either cms or atlas than is the sm .
the blh model also allows for the possibility of @xmath330 , as supported by the atlas measurement set and visible in the right side of figure [ fig : norm : muwwvsmuzz ] .
this occurs due to our calculation of an inclusive cross section that includes contributions from the @xmath2 , which has larger suppressions to the @xmath120 branching ratio at larger masses than the @xmath331 .
an invariant mass windowing of @xmath332 events would likely exclude any excess @xmath333 contributions , and result in a measurement of @xmath334 .
it remains to understand the physics underlying the enhancement of the diphoton rate , in agreement with the atlas results , in the general scenario .
it is not a result of significant contributions from the production and decay of the other neutral higgs bosons in the 2hdm .
while figures [ fig : norm : tbvssapb ] through [ fig : norm : muwwvsmuzz ] show the parameter sets that are in better agreement with the experimental results than is the sm , they do not focus on the parameter sets which best agree with the experimental values . to examine the physics underlying the enhancement in the atlas diphoton rate , we consider the reduced @xmath293 ( @xmath297 ) that includes only the @xmath318 and @xmath55 values .
this will focus on the results which agree with the diphoton rate while constraining the results to also agree with the precisely measured @xmath335 rate . in figure
[ fig : rnorm : muyyvsmuww ] , we plot @xmath318 versus @xmath319 .
figure [ fig : rnorm : muyyvsmuww ] shows that the points in agreement using the reduced @xmath293 ( green points with @xmath336 ) are a subset of those that are a better fit to the data than is the sm , as previously shown in figure [ fig : norm : muyyvsmuww ] ( green points with @xmath337 ) .
we now investigate the contributions of different particles to the diphoton loop process . by separately examining the real component of the fermion , scalar and gauge contribution to the diphoton effective coupling , @xmath338 , @xmath339 , and @xmath340 respectively , as in figures [ fig : rnorm :
muyyvsreshyyf ] , [ fig : rnorm : muyyvsreshyyg ] and [ fig : rnorm : muyyvsreshyys ] , it becomes clear that there are two possibilities that result in a significant enhancement of the diphoton rate .
as shown in figure [ fig : rnorm : muyyvsreshyyf ] ( right panel ) , the contribution from fermions is approximately sm - like .
this is as expected since significant deviation would also result in significant alterations of the gluon fusion effective coupling to the higgs .
similarly , as shown in figure [ fig : rnorm : muyyvsreshyyg ] ( right panel ) , the contribution from gauge bosons is also sm - like .
the additional heavy gauge bosons of the blh model do not affect the results due to their large mass and suppressed couplings .
any significant alteration of the @xmath341 coupling from its sm value would result in disagreement with the @xmath318 results .
the right panel of figure [ fig : rnorm : muyyvsreshyys ] shows the most obvious source of enhancement of the diphoton rate , as measured by atlas .
the enhancement occurs when the scalar contribution to the diphoton effective coupling becomes significant , dominated by the contribution from the charged higgs field , @xmath177 .
the dominant component of the diphoton effective coupling comes from the @xmath77 boson loop , which has a negative relative value , as do the contributions from the charged scalars ( @xmath177 , @xmath128 and @xmath127 ) , while the sum of all of the fermion loops contribute positively .
thus , enhancement of the scalar loop contribution increases the effective coupling strength of @xmath342 ( reductions of the contribution from fermion loops would have a similar effect , but are not important here ) .
we have determined that the diphoton enhancement occurs where the @xmath343 coupling becomes large for large values of @xmath184 , corresponding to where @xmath58 is near the the upper boundary of its allowed range as seen in the right panel of figure [ fig : rnorm : tbvsma0 ] .
this region lies close to the border of perturbativity constraints , as discussed in eq .
[ eq : tbmax ] .
figure [ fig : rnorm : hmuyy1vsamuyy1 ] also shows several parameter points which fit the data well but do not have a significant scalar loop contribution to the diphoton effective coupling .
these points correspond to a moderate contribution to the diphoton rate from the production of the cp - odd scalar at masses above 300 gev .
this is an artifact of the method we use for calculating , which determines an inclusive ( @xmath344 ) cross section , but excludes parameter sets only for which the heavier resonances would be distinguishable at 95% c.l .
in other words , if the heavier resonances do not result in an exclusion , they contribute to the total cross section calculated .
little information was given by the experiments regarding any invariant mass windowing incorporated into the determination of the diphoton excess , and so we chose an inclusive cross section calculation . with regards to the other parameters in the model , particularly the heavy quark mixing angles , @xmath345 and @xmath346 , and heavy gauge boson mixing angle , @xmath217
, no constraints can be placed with the existing data .
this is because these states do not get a large component of their mass from the higgs vacuum expectation value , and so they do not contribute significantly to the loop factors . in figure [ fig : norm : flvsmt6 ] , we show , as an example , the @xmath53 mass versus @xmath54 .
the only constraint that can be determined is an overall mass constraint arising directly from the constraint on @xmath54 . in particular , the minimum heavy quark mass as determined from the higgs data is approximately 300 gev - the precise value of which is unimportant , as direct constraints from pair production searches for heavy vector - like quarks rule out much heavier states @xcite .
the blh model provides , in principle , a rich source of phenomenology .
apart from incorporating a two higgs doublet model , it includes additional scalar triplets , heavy gauge bosons , and a set of six new heavy quarks . in this paper , we have investigated whether the model includes parameter sets that are consistent with the higgs boson signal strength ratios recently measured by the cms and atlas experiments .
we have used a couple of @xmath293 measures , one to compare the fit of the blh model to the data relative to that of a sm fit and another to identify regions of blh model parameter space favoured by the results of the two experiments .
as described above , we have found that the blh model can reproduce the results of either experiment but primarily via modifications to the couplings of top quarks to the higgs states and through the contributions of additional scalar states in the context of a 2hdm . as a 2hdm ,
the blh model provides two possible scenarios : the general case in which a single higgs state dominates contributions to the signal strength ratios , and the near - degenerate case in which multiple higgs states contribute to the observed results . at this time , the experimental data remains statistically limited and there remain discrepancies between the central values of the cms and atlas results .
however , using the full set of measured signal strength ratios , we find that the blh model in the near - degenerate scenario is a worse fit to the data of each experiment than is the sm for almost all parameter sets . this is a consequence of a large enhancement of the @xmath56 signal strength by the contribution of the cp - odd scalar state nearly degenerate with the light higgs boson . in this scenario , for parameter sets
consistent with the observed diphoton rates , @xmath56 is predicted to be three to five times larger than its sm value , and several sigma larger than the value measured by cms or atlas .
consequently , a precise measurement of @xmath56 will be sufficient to exclude the near - degenerate scenario in the blh model .
on the other hand , large regions of the general blh parameter space provide a better fit to the experimental results than does the sm .
this corresponds to @xmath311 in order to achieve @xmath317 , while @xmath313 gev is necessary such that higher order corrections in the expansion in @xmath215 do not reduce couplings between the @xmath347 quark and @xmath0 , and the @xmath77 boson and @xmath0 .
a cp - odd scalar mass of @xmath348 gev is favoured , resulting in a value of @xmath56 that is consistent with the sm .
these rather general constraints provide good agreement with the cms diphoton results , allowing a large range of @xmath58 and @xmath322 values .
the signal strength ratios are produced primarily through the light higgs boson with approximately sm - like couplings . due to the enhancement of the overall scalar and cp - odd scalar production for @xmath349 gev , much of this region of parameter space
is already directly ruled out at 95% c.l . by heavy higgs searches in the @xmath335 and @xmath8 channels .
additionally , all parameter regions of the blh model predict the values of @xmath318 and @xmath350 to be similar , such that the entirety of the blh model would be excluded if the difference in the @xmath350 and @xmath318 results becomes statistically significant to the degree currently measured by atlas .
it is possible to reproduce the atlas @xmath55 measurement with a sm - like @xmath56 in the blh model , but for a restricted space of parameter sets .
the physical origin of the enhancement of the diphoton rate is a significant enhancement of the charged higgs field ( @xmath177 ) contribution to the diphoton loop .
the diphoton enhancement occurs for maximal values of @xmath58 , where the @xmath343 coupling becomes large .
more accurate measurements , including the @xmath69 , @xmath70 and @xmath16 final states , with higher luminosity will be crucial to determining the status of the blh model , and for determining the values of the scalar sector parameters .
the 2hdm sector of the blh model is likely its most accessible aspect , with fairly light cp - odd and , consequently , charged higgs states allowed .
discovery and measurement of heavy quark partners can lead to further constraints on the value of the scale @xmath54 , as discussed in @xcite .
however , in the blh model , measurements of the mass and branching ratios of the lightest heavy quark partner is insufficient to significantly constrain the value of @xmath54 , due to the presence of degenerate states and the involvement of the two mixing angles , @xmath268 and @xmath269 . in @xcite
, it was shown that measurement of several of the heavy quark masses is needed to isolate the value of the scale @xmath54 from the values of the heavy quark mixing angles .
the authors would like to thank heather logan , thomas grgoire , stephen godfrey , michael spira , and peter winslow for guidance and assistance in this project .
the authors would also like to apologize to m. schmaltz , d. stolarski and j. thaler for not using the full name of their model from @xcite in our work .
this research was supported in part by the natural sciences and engineering research council of canada .
g. aad _ et al .
_ [ atlas collaboration ] , `` observation of a new particle in the search for the standard model higgs boson with the atlas detector at the lhc , '' phys .
b * 716 * , 1 ( 2012 ) [ arxiv:1207.7214 [ hep - ex ] ] .
s. chatrchyan _ et al . _
[ cms collaboration ] , `` observation of a new boson at a mass of 125 gev with the cms experiment at the lhc , '' phys .
b * 716 * , 30 ( 2012 ) [ arxiv:1207.7235 [ hep - ex ] ] .
[ atlas collaboration ] , `` an update of combined measurements of the new higgs - like boson with high mass resolution channels , '' atlas - conf-2012 - 170 .
[ cms collaboration ] , `` combination of standard model higgs boson searches and measurements of the properties of the new boson with a mass near 125 gev , '' cms - pas - hig-13 - 005 .
k. cheung , j. s. lee and p. -y .
tseng , `` higgs precision ( higgcision ) era begins , '' jhep * 1305 * , 134 ( 2013 ) [ arxiv:1302.3794 [ hep - ph ] ] .
[ atlas collaboration ] , `` observation and study of the higgs boson candidate in the two photon decay channel with the atlas detector at the lhc , '' atlas - conf-2012 - 168 .
[ atlas collaboration ] , `` search for the standard model higgs boson in @xmath351 decays in proton - proton collisions with the atlas detector , '' atlas - conf-2012 - 160 .
[ atlas collaboration ] , `` observation of an excess of events in the search for the standard model higgs boson in the @xmath352 channel with the atlas detector , '' atlas - conf-2012 - 169 .
[ cms collaboration ] , `` combination of standard model higgs boson searches and measurements of the properties of the new boson with a mass near 125 gev , '' cms - pas - hig-12 - 045 .
[ cms collaboration ] , `` updated results on the new boson discovered in the search for the standard model higgs boson in the @xmath331 to 4 leptons channel in @xmath353 collisions at @xmath354 = 7 and 8 tev , '' cms - pas - hig-12 - 041 .
[ cms collaboration ] , `` higgs to tau tau ( sm ) ( hcp ) , '' cms - pas - hig-12 - 043 .
[ cms collaboration ] , `` updated measurements of the higgs boson at 125 gev in the two photon decay channel , '' cms - pas - hig-13 - 001 .
[ atlas collaboration ] , `` combined coupling measurements of the higgs - like boson with the atlas detector using up to 25 fb@xmath7 of proton - proton collision data , '' atlas - conf-2013 - 034 .
[ atlas collaboration ] , `` measurements of the properties of the higgs - like boson in the two photon decay channel with the atlas detector using 25 @xmath355 of proton - proton collision data , '' atlas - conf-2013 - 012 . j. reuter and m. tonini , `` can the 125 gev higgs be the little higgs ?
, '' jhep * 1302 * , 077 ( 2013 ) [ arxiv:1212.5930 [ hep - ph ] ] .
x. -f .
han , l. wang , j. m. yang and j. zhu , `` little higgs theory confronted with the lhc higgs data , '' arxiv:1301.0090 [ hep - ph ] .
l. wang , j. m. yang and j. zhu , `` dark matter in little higgs model under current experimental constraints from lhc , planck and xenon , '' arxiv:1307.7780 [ hep - ph ] . j. reuter , m. tonini and m. de vries , `` littlest higgs with t - parity : status and prospects , '' arxiv:1310.2918 [ hep - ph ] . m. schmaltz , d. stolarski and j. thaler , `` the bestest little higgs , '' jhep * 1009 * , 018 ( 2010 ) [ arxiv:1006.1356 [ hep - ph ] ] .
we apologize to these authors for not using the full name of their model in our work .
p. m. ferreira , r. santos , h. e. haber and j. p. silva , `` mass - degenerate higgs bosons at 125 gev in the two - higgs - doublet model , '' phys .
d * 87 * , 055009 ( 2013 ) [ arxiv:1211.3131 [ hep - ph ] ] .
n. bonne and g. moreau , `` reproducing the higgs boson data with vector - like quarks , '' phys .
b * 717 * , 409 ( 2012 ) [ arxiv:1206.3360 [ hep - ph ] ] .
j. f. gunion , h. e. haber , g. l. kane and s. dawson , `` the higgs hunter s guide , '' front .
* 80 * , 1 ( 2000 ) .
r. martinez , m. a. perez and j. j. toscano , `` the two photon decay width of the higgs boson in left - right symmetric theories , '' phys .
d * 40 * , 1722 ( 1989 ) .
t. han , h. e. logan , b. mcelrath and l. -t .
wang , `` loop induced decays of the little higgs : @xmath356 , '' phys .
b * 563 * , 191 ( 2003 ) [ erratum - ibid .
b * 603 * , 257 ( 2004 ) ] [ hep - ph/0302188 ] .
s. dittmaier _ et al . _
[ lhc higgs cross section working group collaboration ] , `` handbook of lhc higgs cross sections : 1 .
inclusive observables , '' arxiv:1101.0593 [ hep - ph ] .
d. eriksson , j. rathsman and o. stal , `` 2hdmc : two - higgs - doublet model calculator physics and manual , '' comput .
commun . * 181 * , 189 ( 2010 ) [ arxiv:0902.0851 [ hep - ph ] ] .
s. godfrey , t. gregoire , p. kalyniak , t. a. w. martin and k. moats , `` exploring the heavy quark sector of the bestest little higgs model at the lhc , '' jhep * 1204 * , 032 ( 2012 ) [ arxiv:1201.1951 [ hep - ph ] ]
. j. beringer _
( particle data group ) , review of particle physics , " phys .
d * 86 * , 010001 ( 2012 ) .
s. chatrchyan _ et al .
_ [ cms collaboration ] , `` search for a standard - model - like higgs boson with a mass in the range 145 to 1000 gev at the lhc , '' eur .
j. c * 73 * , 2469 ( 2013 ) [ arxiv:1304.0213 [ hep - ex ] ] .
the atlas collaboration , atlas - conf-2013 - 067 .
[ cms collaboration ] , `` properties of the higgs - like boson in the decay @xmath360 in @xmath353 collisions at @xmath354 = 7 and 8 tev , '' cms - pas - hig-13 - 002 .
[ cms collaboration ] , `` evidence for a particle decaying to @xmath120 in the fully leptonic final state in a standard model higgs boson search in @xmath353 collisions at the lhc , '' cms - pas - hig-13 - 003 .
[ cms collaboration ] , `` search for the standard model higgs boson produced in association with @xmath77 or @xmath159 bosons , and decaying to bottom quarks for hcp 2012 , '' cms - pas - hig-12 - 044 .
[ cms collaboration ] , `` search for the standard - model higgs boson decaying to tau pairs in proton - proton collisions at @xmath354 = 7 and 8 tev , '' cms - pas - hig-13 - 004
. [ atlas collaboration ] , `` measurements of the properties of the higgs - like boson in the four lepton decay channel with the atlas detector using 25 fb@xmath7 of proton - proton collision data , '' atlas - conf-2013 - 013 .
[ atlas collaboration ] , `` measurements of the properties of the higgs - like boson in the @xmath361 decay channel with the atlas detector using 25 fb@xmath7 of proton - proton collision data , '' atlas - conf-2013 - 030 .
[ atlas collaboration ] , `` coupling properties of the new higgs - like boson observed with the atlas detector at the lhc , '' atlas - conf-2012 - 127 .
s. chatrchyan _ et al . _
[ cms collaboration ] , `` inclusive search for a vector - like t quark with charge 2/3 in pp collisions at @xmath354=8 tev , '' arxiv:1311.7667 [ hep - ex ] .
the atlas collaboration , `` search for pair production of heavy top - like quarks decaying to a high-@xmath362 @xmath77 boson and a @xmath277 quark in the lepton plus jets final state in @xmath353 collisions at @xmath363 tev with the atlas detector , '' atlas - conf-2013 - 060 .
t. a. w. martin , `` examining extra neutral gauge bosons in non - universal models and exploring the phenomenology of the bestest little higgs model at the lhc '' , ph.d .
thesis , carleton university ( 2012 ) . | in this paper , we use the latest higgs measurements from atlas and cms to constrain the parameter space of the model of schmaltz , stolarski and thaler , a little higgs model with two higgs doublets , which we will refer to as the blh model .
we account for all production and decay modes explored at atlas and cms in two scenarios : a general case , which assumes the @xmath0 state is light ( @xmath1 gev ) and the masses of the other neutral scalars ( @xmath2 and @xmath3 ) are allowed to vary , and a case with a near - degeneracy between the masses of the @xmath0 and @xmath3 and , for some choices of parameters , the @xmath2 states .
the near - degeneracy scenario can result in an enhanced diphoton rate , as measured by atlas , but is largely ruled out by a combination of the @xmath4 and the heavy @xmath5 measurements . in the general case , we find large regions of parameter space that are in better agreement with either the atlas or cms results than is the sm .
however , a significantly enhanced diphoton rate is only possible through large contributions to the @xmath6 effective coupling from charged higgs bosons in a region of parameter space that borders on violation of perturbativity in the scalar sector . |
from the approximate chiral symmetry of qcd it is known that the pion cloud of the nucleon plays a vital role in understanding the nucleon s long - range structure .
it provides important corrections to static nucleon properties , such as the mass , magnetic moment and axial charge , and significantly influences its electric and magnetic charge distributions ( for a review see ref .
@xcite ) . at the quark level
, the preferential coupling of a proton to a @xmath2 and a neutron provides a natural explanation of the excess of @xmath3 quarks over @xmath4 in the proton sea @xcite , which has now been unambiguously established experimentally @xcite .
since it is pseudoscalar , the emission of a pion from a nucleon also leads to a nontrivial redistribution of the spin and angular momentum of its quark constituents @xcite , which partially resolves the proton spin problem @xcite .
more recently , it has been established from the chiral expansion in qcd that pion cloud contributions to moments of twist - two parton distributions of the nucleon have a leading nonanalytic ( lna ) behavior characteristic of goldstone boson loops in chiral perturbation theory @xcite .
since it is determined by the infrared properties of chiral loops , the lna behavior is model - independent @xcite , and places the physics of the pion cloud on a firm footing in qcd .
in addition to the traditional studies of elastic form factors and parton distributions , there is a great deal of interest in the more recently defined generalized parton distributions ( gpds ) @xcite . as with the ordinary parton distribution functions ,
the physical interpretation of gpds in terms of probability distributions is most natural on the light - front , or in the infinite momentum frame ( imf ) of time - ordered perturbation theory ( topt ) @xcite , and one knows that here too chiral corrections can be very important .
it is timely , therefore , to address the question of how to provide a consistent derivation of the chiral corrections to all of these observables on the light - front , and to explain in some detail the technical differences between these calculations on the light - front and either in equal - time or covariant formulations .
historically the realization of chiral symmetry on the light - front has posed a serious theoretical challenge , and extreme care must usually be taken to avoid pathologies associated with so - called zero modes or spurious end - point singularities in light - front calculations . indeed , the lna behavior of twist - two matrix elements calculated in the meson cloud model on the light - front ( or in the imf in equal - time quantization ) with a pseudoscalar @xmath5 interaction @xcite appeared to be in conflict with the results from covariant chiral perturbation theory @xcite , which uses a pseudovector coupling .
this led to questions being raised @xcite about the suitability of computing chiral corrections to hadronic matrix elements in meson cloud models on the light - front @xcite . in other applications ,
the equivalence of light - front and manifestly covariant formalisms was demonstrated by bakker _
@xcite for the vector two - point function and pseudoscalar charge form factor in @xmath6 dimensions .
( for an analogous discussion in qed see ref .
using several different methods , it was shown @xcite that spurious divergences can be avoided when performing loop integrals by properly taking into account contributions from the arc used to close the contour of integration at infinity .
furthermore , sawicki @xcite demonstrated for a scalar @xmath7 theory that a smooth transition from equal - time perturbation theory to the light - front can be made without reference to the imf limit . in some cases
, however , extreme care must be taken when computing the arc contributions , _ viz .
_ , when one encounters so - called moving poles where the pole in the @xmath8 momentum variable depends on the @xmath9 integration variable .
in particular , the contributions coming from the end points @xmath10 and @xmath11 in the @xmath9 range @xmath12 must be taken into account as discussed in ref .
@xcite . without these end point contributions
the complete equivalence between the light - front and manifestly covariant results can not in general be demonstrated .
it is amusing to see that the corrections we find restore covariance in the same way as the kinetic mass counter - term does in the work of ref .
@xcite . in this paper
we utilize some of these techniques to demonstrate the equivalence between equal - time and light - front dynamics for the interactions of nucleons with pions .
we demonstrate that there is , in fact , no conflict between the results in equal - time , light - front or covariant frameworks , provided care is taken when performing loop integrations and if compares consistently the same theories ( with pseudovector or pseudoscalar @xmath5 interactions ) .
while a detailed analysis of the twist - two matrix elements in the different frameworks will be the subject of an upcoming work @xcite , here for illustration purposes we consider the specific example of the self - energy @xmath0 of a nucleon dressed by a pion loop , and examine in particular the model - independent , nonanalytic behavior of @xmath0 in the chiral limit . in sec .
[ sec : def ] we define the lagrangian for the pseudovector ( pv ) @xmath1 interaction , and introduce the self - energy for the dressing of a nucleon by a pion loop .
although the pseudoscalar ( ps ) @xmath1 interaction does not preserve chiral symmetry ( without the introduction of scalar fields ) , for completeness we also consider the pseudoscalar theory in appendix [ app : ps ] .
we also present a convenient reparametrization of the momentum dependence in the loop integrations which allows the self - energy to be expressed entirely in terms of scalar propagators .
the calculation of the self - energy within a covariant framework is presented in sec .
[ sec : cov ] using dimensional regularization .
we derive results for the total @xmath0 , including finite and divergent parts , but focus in particular on the structure of the model - independent , leading ( and next - to - leading ) nonanalytic contributions , and recover the standard results of chiral perturbation theory @xcite . in sec .
[ sec : et ] we examine the self - energy in detail using time - ordered perturbation theory , both in the familiar rest frame of the nucleon , and in the imf , where a probabilistic interpretation is applicable .
the computation of @xmath0 on the light - front , discussed in sec .
[ sec : lf ] , is closely related to the imf formulation in equal - time dynamics .
we verify that in all cases the correct results are obtained for the nonanalytic contributions . finally , in sec .
[ sec : conc ] we summarize our findings and outline future applications of the results
. a presentation of the results for the self - energy with the ps interaction is given in appendix [ app : ps ] , and the lna behavior of some relevant integrals is listed in appendix [ app : lna ] .
the lowest order @xmath5 interaction with a pseudovector coupling which is relevant for the self - energy is defined by the lagrangian density @xcite @xmath13 where @xmath14 and @xmath15 are the nucleon and pion fields , @xmath16 is the pauli matrix operator in nucleon isospin space , @xmath17 is the pion mass , and @xmath18 is the pseudovector @xmath1 coupling constant with @xmath19 . the analogous pseudoscalar interaction is given in appendix [ app : ps ] . often the pv coupling is expressed in terms of the ps coupling constant @xmath20 , @xmath21 where @xmath22 is the nucleon mass , with @xmath23 . using the goldberger - treiman relation the @xmath1 coupling
can also be expressed in terms of the axial vector charge of the nucleon , @xmath24 , @xmath25 where @xmath26 and @xmath27 mev is the pion decay constant .
the self - energy operator @xmath28 is given by @xcite @xmath29 and can be decomposed into scalar and vector components according to @xmath30 taking the matrix element of @xmath28 between nucleon states , the self - energy ( of mass shift ) of a nucleon with momentum @xmath31 dressed by a pion loop with momentum @xmath32 is given by @xmath33 where the sum is over the nucleon spins @xmath34 , and we have used the spinor normalization convention of bjorken and drell @xcite . the expression in eq .
( [ eq : sigmapv ] ) can be simplified by writing the momentum variables in the numerator of the integrand ( [ eq : sigmapv ] ) in terms of the pion and nucleon propagators @xmath35 and @xmath36 , where [ eq : d ] @xmath37 rearranging eqs .
( [ eq : d ] ) , and dropping the irrelevant @xmath38 terms , one can make the replacements [ eq : momd ] @xmath39 after applying a trace over the nucleon spins and implementing these substitutions , the self - energy can be written as @xmath40 \label{eq : sigmapv1 } \\ & = & - { 3 i g_a^2 \over 4 f_\pi^2 } \int{d^4k \over ( 2\pi)^4 } { 1 \over 2 m } \left [ 4 m^2 \left ( { m_\pi^2 \over d_\pi d_n } + { 1 \over d_n } \right ) + { 2 p \cdot k \over d_\pi } \right ] , \label{eq : sigmapv2 } \ ] ] where in ( [ eq : sigmapv2 ] ) we have used the goldberger - treiman relation ( [ eq : gt ] ) . note that because the term in eq .
( [ eq : sigmapv2 ] ) proportional to @xmath41 is odd in the pion momentum @xmath32 , it will integrate to zero , provided the ultra - violet regulator does not introduce additional dependence on @xmath42 . in any case
, this should not affect the infrared behavior of the integrand , and hence not affect the chiral behavior .
the individual scalar and vector contributions to the self - energy are given by @xmath43 , \label{eq : sigma_v } \\ \sigma_s & = & - { 3 i g_a^2 \over 4 f_\pi^2 } \int{d^4k \over ( 2\pi)^4 } m \left [ { 1 \over d_n } + { m_\pi^2 \over d_\pi d_n } \right ] .
\label{eq : sigma_s } \ ] ] interestingly , provided that the @xmath41 term in the vector self - energy does not contribute , the contributions to the total self - energy from the vector and scalar components are identical , @xmath44 this is not the case , however , for a pseudoscalar theory , as we discuss in appendix [ app : ps ] .
we can interpret the expression in eq .
( [ eq : sigmapv2 ] ) as a reduction of the pseudovector theory to an effective theory of `` scalar nucleons '' @xcite and pions , involving a scalar self - energy and nucleon tadpole diagrams , as in fig .
[ fig : pv ] . in the following sections we compute the self - energy @xmath0 in several ways , including a direct manifestly covariant computation using dimensional regularization , in equal - time dynamics , both in the rest frame and in the infinite momentum frame , and on the light - front .
the self - energy can be evaluated in a manifestly covariant manner using dimensional regularization to regulate the ultra - violet divergence in the integral .
we can compute the self - energy covariantly by either integrating the full expression in eq .
( [ eq : sigmapv1 ] ) or the reduced result in eq .
( [ eq : sigmapv2 ] ) .
since the latter is more straightforward , we present results calculated from the two propagator terms in ( [ eq : sigmapv2 ] ) ( the @xmath45 term integrates to zero ) , although we have verified that identical results are obtained with both expressions . for the @xmath46 term , the product of the pion and nucleon propagators can be reduced using the feynman parametrization [ eq : feynman ] @xmath47 changing variables to @xmath48 , the denominator in the self - energy can be written as @xmath49 where @xmath50 performing a wick rotation to euclidean space , the integral over the pion and nucleon propagators in @xmath51 dimensions , in the limit @xmath52 , can then be written as @xmath53 where @xmath54 is a mass parameter introduced to give the correct mass dimensions in @xmath55 dimensions , @xmath56 , and @xmath57 is euler s constant .
for the integral over the tadpole - like @xmath58 term , we find @xmath59 where we have used the recurrence relation for the @xmath60 function , @xmath61 .
the infinitesimal parameter @xmath62 is set to zero at the end of the calculation , leading to singular results for the integrals , which in principle can be absorbed into counter - terms when computing observables .
however , our concern here is the finite part of the integrals , and in particular the lna behavior of @xmath0 . combining the results in eqs .
( [ eq : dpidn ] ) and ( [ eq : dpi ] ) , the self - energy becomes @xmath63 it is remarkable that a closed form exists for the complete result of the self - energy , even in the relativistic formulation ( see also refs .
expanding @xmath64 in powers of @xmath65 and isolating the nonanalytic terms ( namely , ones which are odd powers or logarithms of @xmath17 ) , the lna behavior as @xmath66 is given by @xmath67 where in addition to the @xmath68 term , which agrees with the established results from chiral perturbation theory @xcite , for completeness we have also kept the next order , @xmath69 , nonanalytic term .
this is not the complete contribution to this order , however , as there exists an @xmath69 term arising from diagrams with a pion loop accompanied by a @xmath70 intermediate state @xcite .
in fact , since this contribution depends on @xmath71 , where @xmath72 is the mass of the @xmath70 , rather than on @xmath73 as in eq .
( [ eq : sigmalnapv ] ) , it will give the next - to - leading contribution in the heavy baryon limit .
note also that because the @xmath58 term in eq .
( [ eq : sigmapv2 ] ) is independent of @xmath17 , it does not contribute to the nonanalytic behavior of @xmath0 , which is solely determined by the @xmath46 term .
while the covariant calculation in the previous section is straightforward , it is instructive to examine the relative contributions to the self - energy from the different time orderings of the intermediate state . this can be realized using time - ordered perturbation theory , in which the pion and nucleon propagators are split up into their positive and negative energy poles .
the relative contributions will naturally depend on the frame of reference , and we consider two commonly used examples , namely , the nucleon rest frame and the infinite momentum frame . here
the 4-dimensional integrals are computed by first performing the integrations over the energy @xmath74 .
the 3-momentum integration over @xmath75 will generally be divergent , so in order to regularize the integrals we employ a 3-momentum cut - off on @xmath76 .
the lna behavior of the self - energy , which is the primary focus of this work , will of course be independent of the details of the ultraviolet regularization .
the computation of the self - energy in the nucleon rest frame is most straightforward in terms of the reduced expression for @xmath0 in eq .
( [ eq : sigmapv2 ] ) .
in contrast to the standard expression in eq .
( [ eq : sigmapv1 ] ) involving pion momenta in the numerator , the self - energy expressed solely through pion and nucleon propagators does not receive contributions from the arc at infinity when performing the @xmath74 integration .
the results are of course identical if one uses the original expression ( [ eq : sigmapv1 ] ) , although as we demonstrate below , in that case one must consider both pole terms and nonzero arc contributions .
consider first the @xmath46 term , which when expanded into positive and negative energy components can be written @xmath77 where in the rest frame the target nucleon has energy @xmath78 , the recoil nucleon energy is @xmath79 and the pion energy is @xmath80 . multiplying out the terms in the parentheses in eq .
( [ eq : dpidn_rf ] ) , the resulting @xmath74 integral has four contributions : * @xmath81 : pion pole in the lower half - plane ( @xmath82 ) and nucleon pole in the upper half - plane ( @xmath83 ) ; * @xmath84 : pion pole in the upper half - plane ( @xmath85 ) and nucleon pole in the lower half - plane ( @xmath86 ) ; * @xmath87 : pion pole in the lower half - plane ( @xmath82 ) and nucleon pole in the lower half - plane ( @xmath86 ) ; * @xmath88 : pion pole in the upper half - plane ( @xmath85 ) and nucleon pole in the upper half - plane ( @xmath83 ) , where the superscripts @xmath89 refer to the signs of @xmath74 in the pion and nucleon parts of the the energy denominators , respectively .
note that the contribution @xmath81 corresponds to the positive energy diagram , while @xmath84 is the so - called `` z - graph '' .
since the terms @xmath87 and @xmath88 have both poles in the same half - plane , one can choose the contour of integration to render their residues zero .
however , in addition to the residues of the poles , the contour integrations also contain contributions from the arc at infinity , which must be subtracted , @xmath90 one can verify that closing the contour @xmath91 in either the upper or lower half - plane gives the same results for @xmath81 and @xmath84 .
for @xmath87 the contour can be chosen in the upper half - plane and for @xmath88 in the lower half - plane to exclude the pole contributions . from the powers of the energy @xmath74 in the numerator and denominators in eq .
( [ eq : dpidn_rf ] ) , however , one sees that the arc contributions will vanish at infinity .
performing the @xmath74 integration , the four contributions to the self - energy then become [ eq : sigmaeti ] @xmath92 while it does not contribute to the lna behavior of @xmath0 , for completeness we consider also the @xmath58 term in eq .
( [ eq : sigmapv2 ] ) . closing the @xmath74 contour integration in either the upper or lower half - planes , the integral can be written @xmath93 combining all the contributions , the total self - energy in the nucleon rest frame is @xmath94 the @xmath95 integration can be performed using spherical polar coordinates , with a high - momentum cut - off @xmath96 on @xmath76 , yielding the final result for the self - energy , @xmath97 we observe that the @xmath17-dependent terms in eq .
( [ eq : sigmaet ] ) are identical to those in the covariant calculation of @xmath0 in eq .
( [ eq : sigmacov ] ) , with the only differences appearing in terms that are analytic in @xmath17 or which depend on the ultraviolet regulator . not surprisingly , therefore ,
expanding @xmath98 in powers of @xmath65 , the lna structure in the chiral limit is @xmath99 which is consistent with the covariant result in eq .
( [ eq : sigmalnapv ] ) .
note that the @xmath68 term in eq .
( [ eq : et_lna ] ) arises entirely from @xmath81 , while the @xmath69 term receives contributions from both the positive energy and z - graphs , @xmath100 so that combined they reproduce the lna behavior in eq .
( [ eq : et_lna ] ) . had we worked from the original expression for the self - energy in eq .
( [ eq : sigmapv1 ] ) , @xmath101 then the contributions from the corresponding four cross products would be [ eq : sigmaetj ] @xmath102 \nonumber\\ & & \hspace*{4 cm } + \ { i r_\infty \over \pi } ( m+e'+{\omega_k})\ + \ { \cal o}(1/r_\infty ) \big\ } , \\ \sigma_{\rm et}^ { ' ( -- ) } & = & -{3 g_a^2 \over 32 \pi^3 f_\pi^2 } \int{d^3\bm{k}\over 2e ' } { 1 \over 2{\omega_k } } \big\ { { 1 \over 2 } \left [ \bm{k}^2 - { \omega_k}^2 + { \omega_k}(m - e ' ) - ( m - e')^2 \right ] \nonumber\\ & & \hspace*{4 cm } + \ { i r_\infty \over \pi } ( m - e'-{\omega_k})\ + \ { \cal o}(1/r_\infty ) \big\ } .\end{aligned}\ ] ] here @xmath103 is the magnitude of the energy @xmath74 parametrized for the arc contribution , @xmath104 , with @xmath105 ranging from 0 to @xmath106 , for contours closed in the upper or lower half - planes , respectively .
in contrast to eqs .
( [ eq : sigmaeti ] ) , the terms @xmath107 and @xmath108 are nonzero , with @xmath109 contributions arising from the arc at infinity .
remarkably , while each individual term in eqs .
( [ eq : sigmaetj ] ) contains a divergent piece from the arc as @xmath110 , the total @xmath109 contribution vanishes once all of the terms are summed .
one can verify that adding the four terms in eqs .
( [ eq : sigmaetj ] ) leads to the same result for the total self - energy @xmath98 as in eq .
( [ eq : sigmaet ] ) . the nonanalytic behavior of the individual components of the self - energy in eqs .
( [ eq : sigmaetj ] ) is given by : [ eq : sigmaetjlna ] @xmath111 interestingly , while lna behavior of the positive - energy @xmath112 term is @xmath68 , the lna behavior of the z - graph @xmath113 actually has a lower order , @xmath114 , which arises partly from the arc contribution .
this cancels , however , with an analogous arc contribution to the @xmath107 component , so that the lna behavior of the total is identical to that in eq .
( [ eq : et_lna ] ) .
it is clear , therefore , that contributions from the arc at infinity are vital if the self - energy in the original formulation ( [ eq : sigmapv1 ] ) is to reproduce the correct behavior of @xmath0 in the chiral limit . as noted above , the reduced form ( [ eq : sigmapv2 ] ) simplifies the computation of @xmath0 considerably by avoiding arc contributions altogether .
the equal - time calculation allows one to track explicitly the origins of the various lna contributions in terms of the respective time - orderings , which are otherwise obscured in a covariant calculation .
the price that one pays , however , is that many more time - ordered diagrams need to be evaluated than in a covariant formulation ; while four graphs for the self - energy is tractable , for other quantities , such as vertex corrections or multi - loop diagrams , the number of time orderings quickly escalates .
it was realized some time ago @xcite that by viewing the system in a lorentz - boosted frame in which the nucleon is moving along the @xmath115 direction with infinite momentum , many time - ordered diagrams which contribute in the rest frame are suppressed by powers of the nucleon momentum , @xmath116 , as @xmath117 . in particular , diagrams involving backward - moving nucleons ( _ i.e. _ , backwards in time ) in intermediate states do not contribute in this limit .
the infinite momentum frame ( imf ) therefore provides a simplifying framework with a much reduced number of diagrams , while at the same time retaining an intuitive , probabilistic interpretation of particle production .
the self - energy of the nucleon due to pion loops , as well as vertex and wave function renormalization , was considered by drell , levy and yan ( dly ) @xcite using a pseudoscalar @xmath1 interaction .
they found , however , that extreme care must be taken to correctly treat certain cases when the nucleon s momentum fraction carried by the pion takes its limiting values of 0 or 1 , and that naive application of topt rules in the imf can lead to important contributions to integrals being omitted .
here we perform an analogous treatment of the self - energy for the pseudovector case , and illustrate how so - called `` treacherous '' points can be avoided .
we begin with the reduced form of the self - energy in eq .
( [ eq : sigmapv2 ] ) . for the @xmath46 term ,
the integral over the energy @xmath74 can be decomposed into four contributions as in eq .
( [ eq : dpidn_rf ] ) . for the positive energy term , @xmath118 , with the pion pole taken in the lower half - plane and nucleon pole in the upper half - plane
, we parametrize the initial nucleon momentum @xmath31 and intermediate nucleon and pion momenta @xmath119 and @xmath32 by @xcite @xmath120 where @xmath121 is the nucleon s longitudinal momentum , and @xmath122 is the longitudinal momentum fraction carried by the intermediate nucleon .
the parametrization ( [ eq : imfkin ] ) respects the momentum conservation condition @xmath123 .
the contribution to the self - energy can then be written @xmath124 where we have changed the variable of integration from @xmath125 to @xmath122 . as we have already shown in the rest frame calculation in sec .
[ ssec : rf ] , the contributions from the arc at infinity to the @xmath74 integration of the @xmath46 term are @xmath126 due to the absence of loop momenta in the numerator and the quadratic dependence on @xmath74 in the denominator . only pole terms contribute , therefore , to the integral over @xmath46 . for the @xmath122 integration we need to consider three regions : @xmath127 , @xmath128 and @xmath129 . for @xmath127 , which corresponds to the intermediate nucleon moving backward ( in space ) , from the momenta parametrizations in eq .
( [ eq : imfkin ] ) the energy denominator in eq .
( [ eq : imf+- ] ) is @xmath130 counting powers of the large momentum @xmath131 , one sees that the integral vanishes as @xmath132 . in the region @xmath128 the energy denominator becomes @xmath133 which yields a nonzero contribution to ( [ eq : imf+- ] ) in the limit @xmath132 . finally , when @xmath129 , which corresponds to the pion moving backward , one has @xmath134 which again gives a suppressed contribution as @xmath132 .
thus the only nonzero contribution is from the diagram where both the nucleon and pion are forward moving , _
i.e. _ , @xmath135 and @xmath136 , in which case the self - energy for the positive energy diagram is @xmath137 where an ultraviolet cut - off @xmath138 is introduced to regulate the @xmath139 integration . for the z - graph
, the intermediate nucleon and pion momenta are parametrized according to @xcite @xmath140 satisfying the momentum conservation condition @xmath141 .
the contribution to the self - energy is then given by @xmath142 where again arc contributions at infinity are suppressed . unlike for the @xmath118 case above ,
however , since each of the @xmath143 terms in the energy denominator add , the sum @xmath144 for all @xmath122 . counting the large momentum factors in eq .
( [ eq : imf-+ ] ) reveals that the z - graph contribution to the self - energy is @xmath145 and thus vanishes in the limit @xmath132 .
moreover , since there are no arc contributions for the @xmath46 term in the equal - time dynamics , the terms with both poles in the upper or lower half - plane will be zero , @xmath146 , as for the rest frame calculation in eq .
( [ eq : sigmaeti ] ) . for the @xmath58 nucleon tadpole term in eq .
( [ eq : sigmapv2 ] ) , the integration can be performed in an analogous way , using the parametrization of the momenta as in eq .
( [ eq : imfkin ] ) .
although it does not contain nonanalytic structure in @xmath17 , it is still instructive to examine its computation in the imf , and to demonstrate the equivalence of the results with those of the formalisms .
using the fact that the integrand is symmetric in @xmath147 , the integral of @xmath58 can be written @xmath148 where @xmath138 cuts off the large transverse momenta in the @xmath139 integration .
after integrating over @xmath139 , we note that because of the simpler structure of the denominator , we need only consider a single region of @xmath122 integration .
since the @xmath122 integration is also divergent , we introduce a large-@xmath122 cut - off parameter , @xmath149 , @xmath150 with @xmath151 .
one could also perform the @xmath122 integration first , as in dly @xcite for the pseudoscalar model , with yields the same results . combining the @xmath5 loop diagram with the nucleon tadpole
, the total self - energy in the imf can be written @xmath152 which is identical to the rest frame expression ( [ eq : sigmaet ] ) in its nonanalytic structure .
expanding in @xmath17 then gives the lna behavior of the self - energy in the imf , which arises solely from the positive energy @xmath5 loop contribution , @xmath153 the results differ , not surprisingly , from those of dly @xcite for the pseudoscalar @xmath5 interaction . in that case the calculation is performed using the original expression for @xmath0 in eq .
( [ eq : sigmaps1 ] ) below , where the presence of loop momenta in the numerator means that the main contribution to the self - energy arises from the infinitesimal end - point regions @xmath154 and @xmath155 @xcite ( see appendix [ app : ps ] ) . for the more consistent pseudovector theory , working with the reduced form ( [ eq : sigmapv2 ] ) allows one to avoid this problem .
a formalism which is similar to topt in the imf is light - front , where the fields are quantized along the light - cone , @xmath156 .
the advantage of this formulation is that , as with the imf in equal - time , many diagrams are suppressed , but the suppression is realized in any reference frame .
however , as also in the imf calculation , care needs to be taken to correctly include contributions from end - point regions corresponding to pion light - front momentum fractions equal to 0 or 1 . we define four - momenta on the light - front as @xmath157 , with the plus / minus components @xmath158 .
note that particles on the light - front are on their mass shells , while the minus components of momenta are not conserved at vertices . in terms of light - front variables , the self - energy in eq .
( [ eq : sigmapv2 ] ) can be written as @xmath159 where we have dropped the term odd in @xmath32 .
choosing the @xmath160 frame , and performing the @xmath8 integration , the integral of the @xmath46 term can be written @xmath161 where @xmath162 is the plus momentum fraction of the nucleon carried by the pion .
there is no arc contribution here due to the quadratic @xmath8 dependence in the denominator .
thus , for the region @xmath163 or @xmath164 , where both poles of @xmath8 are located either in the lower or upper half plane , respectively , the integral of @xmath46 vanishes .
consequently , only the region @xmath165 contributes to the integration .
note that the result in ( [ eq : sigmalf_dpidn ] ) identical to the imf result for @xmath118 in eq .
( [ eq : sigmaimf+- ] ) , provided the @xmath139 integration is regulated by the same high - momentum cut - off @xmath138 as in ( [ eq : sigmaimf+- ] ) . for the @xmath58 term care
needs to be taken in computing the arc contribution because of the dependence on @xmath9 in the pole of @xmath8 .
namely , the @xmath8 pole is moving as @xmath9 changes , which leads to a treacherous point as has been discussed previously in the literature @xcite . to see this point we can change the integration variable @xmath166 and rewrite the @xmath58 term as @xmath167 note here that the @xmath8 pole ( _ i.e. _ , @xmath168 depends on @xmath9 .
not only does the position of the pole depend on the sign of @xmath9 , but also the pole moves to infinity as @xmath169 , changing the degree of divergence of the @xmath8 integral from logarithmic to linear . at the point
@xmath10 the @xmath8 integral is linearly divergent , not logarithmically divergent as one would naively expect from eq .
( [ eq : dn_lf ] ) for the @xmath170 and @xmath171 regions .
the computation of the @xmath58 term is thus highly nontrivial in light - front dynamics .
the details of this treacherous point have been discussed by bakker _
following ref .
@xcite , we use the light - front cylindrical coordinates ( @xmath172 , @xmath173 ) to perform the @xmath9 and @xmath8 integration as follows : @xmath174 \nonumber\\ & = & \lim_{r\to\infty } \left ( -\pi^2 + 2\pi i\ \log{r_0 ^ 2 \over r^2 } + { \cal o}(1/r^4 ) \right ) , \end{aligned}\ ] ] where @xmath175 .
the result contains the same @xmath176 term in the integrand as in the imf calculation above .
the @xmath139 integration is then straightforward , and adding the two terms in eq .
( [ eq : sigmalfdef ] ) gives the total light - front self - energy @xmath177 the nonanalytic part of @xmath178 is identical to the results of the covariant and equal - time ( rest frame and imf ) calculations , @xmath179 demonstrating the equivalence of the light - front formalisms to the equal time and covariant formulations .
we have demonstrated the equivalence of the covariant , equal - time ( both in the rest frame and the imf ) , and light - front formalisms in computing the renormalization of the bare nucleon by pion loops . as a specific example
, we have focussed on the self - energy of the nucleon , which finds applications in computations of pion cloud corrections to deep inelastic structure functions , form factors , and other observables .
while this equivalence is expected , it has not to our knowledge been demonstrated explicitly in the literature .
the calculations involve some non - trivial aspects , especially in the imf and lf formulations , which require special care when dealing with end - point singularities , corresponding to momentum fractions approaching 0 or 1 .
we focused on the chirally preferred case of pseudovector coupling , although the comparison with pseudoscalar coupling ( discussed in appendix [ app : ps ] ) , which was employed in the classical work of drell , levy and yan ( dly ) , reveals several interesting features .
the leading nonanalytic behavior of the self - energy in the latter case is incorrect , with a term behaving as @xmath180 .
in addition , whereas in the pseudovector case the scalar and vector pieces are equal and have the same lna behavior , in the pseudoscalar theory they have even lower order ( incorrect ) nonanalytic terms of order @xmath17 , which cancel in the full combination @xmath181 . finally , we note that the technical problems encountered by dly , with critical contributions from the regions @xmath182 and @xmath183 , are peculiar to the pseudoscalar case in the light - front formalism .
there are a number of applications of the methodology developed here which will be important to pursue in future . as well as the lna behavior of the vertex renormalization , @xmath184 , it is important to compute the lna behavior of the moments of the twist - two parton distribution functions .
in particular , one will be able to explore within the lf formalism the results obtained by dly @xcite in the imf , as well as the application to this problem of effective field theory @xcite .
this should permit a satisfactory resolution of the discrepancy between the results of imf and rest frame equal - time calculations cited by chen and ji @xcite .
it will also pave the way for a consistent interpretation of the physics of the pion cloud at the partonic level and for the development of realistic chiral models of hadron structure on the lf suitable for discussing form factors , parton distribution functions and gpds .
this work was supported by the doe contract no .
de - ac05 - 06or23177 , under which jefferson science associates , llc operates jefferson lab . we acknowledge partial support from the triangle nuclear theory fund and the ncsu theory group fund ( doe contract no .
de - fg02 - 03er41260 )
although the pseudoscalar lagrangian is not invariant under chiral transformations without the introduction of a scalar field ( as in the linear sigma model , for example @xcite ) , since this theory is often discussed in the literature , for completeness we discuss here the results for the ps theory .
this also allows us to contrast a number of important features pertinent to the pseudovector and pseudoscalar calculations .
the lowest order lagrangian density for a ps @xmath5 interaction relevant for the self - energy is given by @xmath185 where @xmath186 is the ps coupling constant . for on - shell nucleons obeying the free dirac equation the ps and pv lagrangians ( see eq .
( [ eq : lpv ] ) ) give identical results for matrix elements , provided the couplings are related by eq .
( [ eq : cc ] ) . for bound or off - shell nucleons ,
the ps and pv interactions lead to different results . however , the couplings @xmath18 and @xmath20 are still related through eq .
( [ eq : cc ] ) since these are defined at the nucleon ( and pion ) poles .
the self - energy operator for the ps coupling is given by @xmath187 taking the spin trace and using eq .
( [ eq : d ] ) to replace the momentum dependence in the numerator by the nucleon and pion propagators gives for the self - energy @xmath188\ .
\label{eq : sigmaps2}\end{aligned}\ ] ] the transformed effective ps theory is represented in fig .
[ fig : ps ] , where the self - energy now contains contributions from a scalar nucleon self - energy and a nucleon tadpole diagram , as in the pv theory , as well as a pion tadpole term .
the individual vector and scalar contributions to @xmath0 are given by @xmath189\ , \end{aligned}\ ] ] which when combined recover the full expression in eq .
( [ eq : sigmaps2 ] ) .
the result of the covariant calculation of the full self - energy @xmath190 using dimensional regularization is @xmath191 expanding in powers of @xmath17 , the leading nonanalytic behavior of the self - energy in the ps theory is given by @xmath192 here the @xmath68 term is identical to the leading term in the pv self - energy in eq .
( [ eq : sigmalnapv ] ) .
this is not surprising since the @xmath193 term in the integrand , which produces the @xmath194 behavior , is common to both the pv and ps self - energies .
however , the presence of the pion tadpole in @xmath190 ( arising from the @xmath195 term in the integrand ) leads to an additional contribution proportional to @xmath196 , which is of lower order than the @xmath194 term . moreover , if we consider the leading nonanalytic behavior of the vector and scalar parts separately , @xmath197\ , \\ \sigma^{\rm ps}_{s , \rm lna } & = & -{3 g_a^2 \over 32 \pi f_\pi^2 } \left [ 2m^2 m_\pi + { m \over \pi } m_\pi^2 \log
m_\pi^2 + \cdots \right]\ , \end{aligned}\ ] ] we find that , in contrast to the pv case , the leading order terms in both @xmath198 and @xmath199 are @xmath200 !
while these terms cancel in the total @xmath201 , this demonstrates that the lna behavior of @xmath190 is very different to that for the pv theory . using the original formulation as in eq .
( [ eq : sigmaps1 ] ) , drell , levy and yan @xcite computed the mass shift in the imf formulation , finding that the entire contribution arises from the infinitesimal end - point regions @xmath154 and @xmath155 for the positive energy and @xmath202-graph contributions , respectively .
moreover , the two terms separately diverge as @xmath132 or @xmath203 , but the sum of the two becomes independent of @xmath131 and @xmath204 . while the results are ultimately the same , this illustrates the fact that the original formulation ( [ eq : sigmaps1 ] ) leads to a somewhat more tortuous path , at least in the imf ( and light - front ) treatments than with the reduced form ( [ eq : sigmaps2 ] ) .
in this appendix we summarize the leading nonanalytic behavior of various integrals which enter into calculations of the self - energy in the pv and ps theories .
the relevant integrands are ones which involve a product of pion and nucleon propagators , one pion propagator ( for the pion tadpole diagram in fig .
[ fig : pv ] ) , or one nucleon propagator ( for the nucleon tadpole diagrams in figs .
[ fig : pv ] and [ fig : ps ] ) . for the product of pion and nucleon propagators ,
one finds leading terms of order @xmath17 and @xmath196 , @xmath205 the pion tadpole diagram involves the integral over a single pion propagator , and contributes to the nonanalytic behavior only at a higher order , @xmath206 note that there are no higher order contributions in @xmath17 for this term .
finally , since it is independent of the pion mass , integrals involving nucleon propagators only will give zero nonanalytic contributions , @xmath207 for any power @xmath208 .
f. myhrer and a. w. thomas , phys .
b * 663 * , 302 ( 2008 ) .
a. w. thomas , w. melnitchouk and f. m. steffens , phys .
lett . * 85 * , 2892 ( 2000 ) .
w. detmold , w. melnitchouk , j. w. negele , d. b. renner and a. w. thomas , phys . rev
. lett . * 87 * , 172001 ( 2001 ) .
j .- w . chen and x. ji , phys .
b * 523 * , 107 ( 2001 ) .
d. arndt and m. j. savage , nucl .
a * 697 * , 429 ( 2002 ) . for review see x. ji , ann .
sci . * 54 * , 413 ( 2004 ) ; a. v. belitsky and a. v. radyushkin , phys .
rept . * 418 * , 1 ( 2005 ) .
s. d. drell , d. j. levy and t. m. yan , phys .
d * 1 * , 1035 ( 1970 ) .
j .- w . chen and x. ji , phys .
lett . * 87 * , 152002 ( 2001 ) [ erratum - ibid . *
88 * , 249901 ( 2002 ) ] .
chen and x. ji , phys .
b * 523 * , 73 ( 2001 ) .
w. melnitchouk and a. w. thomas , phys .
d * 47 * , 3794 ( 1993 ) .
j. speth and a. w. thomas , adv .
* 24 * , 83 ( 1997 ) .
b. l. g. bakker , m. a. dewitt , c .- r .
ji and y. mishchenko , phys .
d * 72 * , 076005 ( 2005 ) .
a. misra and s. warawdekar , phys .
d * 71 * , 125011 ( 2005 ) .
m. b. hecht , m. oettel , c. d. roberts , s. m. schmidt , p. c. tandy and a. w. thomas , phys .
c * 65 * , 055204 ( 2002 ) .
j. d. bjorken and s. drell , _ relativistic quantum fields _
( mcgraw hill , new york , 1965 ) .
s. scherer and m. r. schindler , arxiv : hep - ph/0505265 .
m. procura , t. r. hemmert and w. weise , phys . rev .
d * 69 * , 034505 ( 2004 ) .
d. b. leinweber , a. w. thomas , k. tsushima and s. v. wright , phys .
d * 61 * , 074502 ( 2000 ) .
s. weinberg , phys .
rev . * 66 * , 1313 ( 1966 ) . | we demonstrate the equivalence of the light - front and equal - time formulations of pionic corrections to nucleon properties . as a specific example , we consider the self - energy @xmath0 of a nucleon dressed by pion loops , for both pseudovector and pseudoscalar @xmath1 couplings .
we derive the leading and next - to - leading nonanalytic behavior of @xmath0 on the light - front , and show explicitly their equivalence in the rest frame and infinite momentum frame in equal - time quantization , as well as in a manifestly covariant formulation . |
the diffusion of particles is a subject that has been studied extensively in many different contexts ( see @xcite and cites within ) . one of this contexts is the diffusion of particles inside narrow channels @xcite .
however , the diffusion of interacting particles inside a confined environment has received much less attention , even when there are numerous examples in nature @xcite . in this paper
we studied the transport of brownian particles in a channel of variable transverse width . on the one hand
, we studied how the current of particles that cross an asymmetric cavity is modified by increasing the concentration , considering particles that interact through a hard - core potential . on the other hand , we introduced a modified fick
jacobs equation to describe this process .
this equation has been solved using numerical methods for the case of asymmetric cavities .
we mainly studied how the particles arrange inside the cavity , when a small force is applied .
numerical simulations with the method of monte carlo where carried out .
the results obtained with the equation are in complete agreement with the numerical simulations .
this paper is organized as follows . in sec .
[ sec : triangle ] , we analyze the current of particles in a triangular cavity as a function of the concentration and we expose some of the characteristics of the numerical simulations . in sec .
[ sec : equation ] , we explain the most important steps to derive the non linear fick jacobs equation . in sec . [
sec : results ] , we show how the theoretical solution matches with the results obtained with the numerical simulations .
some final remarks are stated in sec .
[ sec : conclusion ] .
0.45 . @xmath0.,title="fig : " ] 0.45 . @xmath0.,title="fig : " ] in the first part of this work we analyzed the diffusion of hard
core interacting particles inside a triangular cavity . to perform the numerical simulations we depicted the triangular cavity inside a regular square lattice of @xmath1 sites , as seen in fig .
[ fig : cavidadtriangular ] .
the distance between neighboring sites is @xmath2 ( a value @xmath3 was considered ) .
there is a small exit of size @xmath4 on each side .
periodic boundary conditions were set in those exits to simulate a long chain of identical cavities .
each site inside the cavity can be either occupied by a particle or empty .
the hard core interaction is realized by allowing only one particle on each site of the lattice .
we define the mean concentration @xmath5 , where @xmath6 is the total number of sites inside the cavity and @xmath7 is the number of particles . at the beginning of the simulation ,
the particles are distributed randomly inside the cavity .
the size of a particle is exactly one site , and they move by jumping to one nearest neighbor site . they are allowed to jump freely inside the cavity , except when the new site is occupied with another particle . in this case
the particle remains on its site .
we say that one monte carlo step occurs when , in average , every particle has had the opportunity to jump once .
the particles are driven by a dimensionless external force , @xmath8 where . at any given time
, a particle has four possible directions to move .
one of the them has a a higher jump rate , due to the effect of the external force applied .
the other three have the same ( lower ) jump rate .
thus , if the force is applied to the right , the jump rates in different directions are @xmath9 where @xmath10 , and @xmath11 is the diffusion coefficient . in the same way , if the force is applied to the left , and @xmath12 . in the first part of this paper we will analyze the difference in the current of particles of a system when the force is applied first to the right , and then to the left .
we let the system evolve to the stationary state , and then we start to measure the relevant quantities , like the distribution of the particles inside the cavity , the distance covered by each one or the mean velocity , among others . here
we are interested in the total current of particles that cross the cavity , @xmath13 .
if there were no interaction between particles , the current will be higher as the concentration or the force becomes higher .
but that is not necessarily true for interacting particles . also ,
because of the particular shape of the cavity , the current will not be the same when the external force is applied to the right , @xmath14 , or to the left , @xmath15 .
since this asymmetric effect depends on the size of the particles , it has been proposed by reguera _ _ et al.__@xcite as a possible way to separate a mixture of different size particles . a a function of the mean concentration @xmath16 in the cavity of fig .
[ fig : cavidadtriangular ] .
the values of the force are : @xmath17 .
average over @xmath18 samples.,scaledwidth=70.0% ] in fig .
[ fig : deltaj ] we show the difference of currents of particles , @xmath19 as a function of the concentration for a wide range of forces .
there are a few things that we can notice from this plot .
around @xmath20 we can see there is a maximum value in the difference of the current @xmath21 .
there is a value of the mean concentration , @xmath16 , for each value of the force , @xmath22 , for which the difference in the current is maximum .
increasing the concentration after this point , will only decrease the asymmetric effect of the cavity . particularly , if the mean concentration is @xmath23 the asymmetric effect vanishes , and the current of particles is the same when the force is applied to the left or to the right , i.e. @xmath24 , independently of its magnitude .
finally , if the concentration is higher than @xmath25 , we found that the current of particles when the force is applied to the left is always higher than the current of particles when the force is applied to the right .
this is specially interesting because the asymmetry in the cavity is designed to favor the diffusion to the right .
after this we can conclude that the hard
core interaction , whose influence increases as @xmath16 is increased , is the cause of an inversion of the asymmetry effect .
the fick - jacobs equation describes the diffusion of non interacting particles inside a channel of variable width .
it has been used to describe the diffusion of not
interacting particles or in the low concentrations regime @xcite . in this work
we have extended this equation to include the case of diffusion of particles with hard core interaction . with this
we mean that two particles can not occupy the same space .
the excluded volume effect of the other particles has a considerable influence on the diffusion of a tagged particle .
this equation allows us to describe a rather complex system , of many interacting particles inside a two or three dimensional environment , with a simple uni - dimensional equation . to derive the modified fick
jacobs equation , let us consider a set of hard core interacting particles in a one dimensional discretized system , the so called single file diffusion @xcite . in a given configuration ,
the current of particles between the sites @xmath26 and @xmath27 is given by , @xmath28 where @xmath29 is the occupation number of site @xmath26 , @xmath30 is the variation of the external potential and @xmath31 .
@xmath32 is the jumping rate , @xmath10 where @xmath11 is the diffusion constant and @xmath3 is the length of the jump .
we are assuming that @xmath33 .
the terms @xmath34 and @xmath35 take into account the hard core interaction . averaging over configurations one has @xmath36 and @xmath37 . in the continuous limit @xmath38 , @xmath39 and @xmath40 , where @xmath41 . assuming the decorrelation @xmath42 and small variation of @xmath43 we obtain @xmath44 .
\label{eq : j}\ ] ] let us now consider a two dimensional channel composed by a chain of cavities ( as the one shown in fig .
[ fig : color ] ) where the external potential @xmath45 is independent of the transverse direction , @xmath46 . assuming local equilibrium ,
the site concentration @xmath47 does not depend on the transverse direction either .
that is , @xmath48 .
then , the total current is @xmath49 , where @xmath50 is the transverse width of the channel and from eq .
( [ eq : j ] ) one obtains @xmath51 , \label{eq : final}\ ] ] where @xmath52 is the external force . now from the continuity equation @xmath53 we get @xmath54 , \ ] ] which corresponds to our extension of the fick - jacobs equation where the hard - core interaction is taken into account . for non interacting particles @xmath55 and the well known fick
jacobs equation is recovered .
0.45 axis , @xmath43 .
comparison between analytic solution to eq .
( [ eq : final ] ) and simulations for @xmath23 . in each case
, the force is applied to the right and the following values were used : @xmath56 ( blue triangles ) , @xmath57 ( black squares ) and @xmath58 ( red circles ) . dots were obtained with numerical simulations ( @xmath59 samples between @xmath60 and @xmath61 monte carlo time - steps ) .
( b ) site concentration inside the cavity
. corresponds to the case @xmath62 and @xmath58 ( red circles in ( a ) ) .
, title="fig : " ] 0.53 axis , @xmath43 .
comparison between analytic solution to eq .
( [ eq : final ] ) and simulations for @xmath23 . in each case
, the force is applied to the right and the following values were used : @xmath56 ( blue triangles ) , @xmath57 ( black squares ) and @xmath58 ( red circles )
. dots were obtained with numerical simulations ( @xmath59 samples between @xmath60 and @xmath61 monte carlo time - steps ) .
( b ) site concentration inside the cavity . corresponds to the case @xmath62 and @xmath58 ( red circles in ( a ) ) .
, title="fig : " ]
to test how accurate eq . ( [ eq : final ] ) is , we solved it for a particular case using a smoother asymmetric cavity than the one shown in fig .
[ fig : cavidadtriangular ] , see fig .
[ fig : color ] . in fig .
[ fig : fj ] we show the stationary state of site concentration along the direction of the channel in an asymmetric cavity , for different values of the force . in the three curves
the force is applied to the right . however , as a consequence of the analogy in the definition of a particle and a vacancy , they also represent the system behavior when the force is applied to the left . for every particle that move to the right
there is a vacancy that moves to the left .
all we have to do is to change @xmath63 ( then @xmath64 ) , @xmath65 and @xmath66 .
note that in eq .
( [ eq : final ] ) this symmetry holds , see also fig .
[ fig : deltaj ] .
further details on this matter may be found in @xcite .
the results show that the agreement between theory and simulations is very good almost everywhere .
however , it can be noticed that , in some regions the agreement is better than others .
this is due to the fact that we assumed @xmath67 to be constant along the @xmath46 axis .
when this condition is not fulfilled , the theoretical curves slightly differs from simulations ( see fig . [
fig : comparison ] ) .
in this paper we studied how the diffusion of interacting particles is modified by increasing the concentration .
we showed that the interaction between particles plays a fundamental role in this kind of systems .
one reason is the formation of clogs where the available space is limited .
apart from that , we were able to obtain a non linear fick - jacobs equation to describe the transport of hard core particles inside a cavity with variable cross section .
this equation correctly predicts the steady state of the concentration of particles along the direction of the channel .
the agreement of theory and simulations confirms that the modified equation of fick jacobs here introduced is a useful tool to describe the transport of particles when the interactions between them can not be neglected .
this work was partially supported by consejo nacional de investigaciones cientficas y tcnicas ( conicet , argentina , pip 0041 2010 - 2012 ) .
gps wishes to thank the organizing committee of the xxvi iupap conference on computational physics for financial support to assist to the conference . | in this paper we investigate the diffusion of particles inside a chain of asymmetric cavities .
we are considering particles that interact through a hard core potential and are driven by an external force .
we show that the difference in the current when the force is applied to the left and to the right strongly depends on the concentration inside the cavity .
we found that , when the concentration is high enough , the hard
core interaction vanishes and inverts the asymmetric effect of the cavity .
we also introduce a new equation , a modification to the fick
jacobs equation , to describe this system analytically .
finally , we used numerical simulations to verify the analytic results , finding a good agreement between theory and simulations . |
we thank a. notari and a. strumia for useful discussions .
n. cabibbo , l. maiani , g. parisi and r. petronzio , nucl .
b * 158 * ( 1979 ) 295 .
p. q. hung , phys .
* 42 * ( 1979 ) 873 .
m. lindner , z. phys .
c * 31 * ( 1986 ) 295 .
m. lindner , m. sher and h. w. zaglauer , phys .
b * 228 * ( 1989 ) 139 . m. sher , phys .
* 179 * ( 1989 ) 273 .
b. schrempp and m. wimmer , prog .
* 37 * ( 1996 ) 1 [ hep - ph/9606386 ] .
j. a. casas , j. r. espinosa and m. quiros , phys .
b * 382 * ( 1996 ) 374 [ hep - ph/9603227 ] .
g. isidori , g. ridolfi and a. strumia , nucl . phys .
b * 609 * ( 2001 ) 387 [ hep - ph/0104016 ] . j. r. espinosa , g. f. giudice and a. riotto , jcap * 0805 * ( 2008 ) 002 [ arxiv:0710.2484 [ hep - ph ] ] .
n. arkani - hamed , s. dubovsky , l. senatore and g. villadoro , jhep * 0803 * ( 2008 ) 075 [ arxiv:0801.2399 [ hep - ph ] ] . g. isidori , v. s. rychkov , a. strumia and n. tetradis , phys .
d * 77 * ( 2008 ) 025034 [ arxiv:0712.0242 [ hep - ph ] ] .
j. elias - miro , j. r. espinosa , g. f. giudice , g. isidori , a. riotto and a. strumia , arxiv:1112.3022 s. r. coleman , phys .
d * 15 * , 2929 ( 1977 ) [ erratum - ibid .
d * 16 * , 1248 ( 1977 ) ] .
t. biswas and a. notari , phys .
d * 74 * ( 2006 ) 043508 [ hep - ph/0511207 ] .
f. di marco and a. notari , phys .
d * 73 * ( 2006 ) 063514 [ astro - ph/0511396 ] .
i. masina and a. notari , phys .
d * 85 * ( 2012 ) 123506 [ arxiv:1112.2659 [ hep - ph ] ] .
g. aad _ et al .
_ [ atlas collaboration ] , phys .
b [ arxiv:1207.7214 [ hep - ex ] ] .
s. chatrchyan _ et al . _
[ cms collaboration ] , phys .
b [ arxiv:1207.7235 [ hep - ex ] ] .
i. masina and a. notari , phys .
* 108 * ( 2012 ) 191302 [ arxiv:1112.5430 [ hep - ph ] ] .
e. komatsu _ et al .
_ [ wmap collaboration ] , astrophys . j. suppl .
* 192 * ( 2011 ) 18 [ arxiv:1001.4538 [ astro-ph.co ] ] .
p. a. r. ade _ et al .
_ [ planck collaboration ] , arxiv:1303.5082 [ astro-ph.co ] .
i. masina , phys .
d * 87 * ( 2013 ) 053001 f. bezrukov , m. y. .kalmykov , b. a. kniehl and m. shaposhnikov , jhep * 1210 * ( 2012 ) 140 .
g. degrassi , s. di vita , j. elias - miro , j. r. espinosa , g. f. giudice , g. isidori and a. strumia , jhep * 1208 * ( 2012 ) 098 . see also : d. buttazzo , g. degrassi , p. p. giardino , g. f. giudice , f. sala , a. salvio and a. strumia , jhep * 1312 * ( 2013 ) 089 [ arxiv:1307.3536 ] . | for a narrow band of values of the top quark and higgs boson masses , the standard model higgs potential develops a shallow local minimum at energies of about @xmath0 gev , where primordial inflation could have started in a cold metastable state . for each point of that band , the highness of the higgs potential at the false minimum is calculable , and there is an associated prediction for the inflationary gravitational wave background , namely for the tensor to scalar ratio @xmath1 .
we show that the recent measurement of @xmath1 by the bicep2 collaboration , @xmath2 at @xmath3 , combined with the most up - to - date measurements of the top quark and higgs boson masses , reveals that the hypothesis that a standard model shallow false minimum was the source of inflation in the early universe is viable .
the fact that , for a narrow band of values of the top quark and higgs boson masses , the standard model ( sm ) higgs potential develops a local minimum @xcite is very interesting , as this happens at energy scales of about @xmath0 gev which are suitable for inflation in the early universe . inflation from a local minimum @xcite is a viable scenario , provided a graceful exit to a radiation - dominated era can be obtained via some mechanism beyond the sm . developing a model with graceful exit in the framework of a scalar - tensor theory of gravity @xcite , in ref .
@xcite we pointed out that the hypothesis that inflation took place in a sm shallow false vacuum was consistent only with a narrow range of values of the higgs boson mass , which subsequently turned out to be compatible with the experimental range indicated by atlas and cms @xcite .
these very suggestive results provide a strong motivation to further investigate the scenario of sm false vacuum inflation , by looking for complementary experimental tests .
inflation can generate tensor ( gravity wave ) modes as well as scalar ( density perturbation ) modes .
it is common to define the tensor contribution through @xmath1 , the ratio of tensor to scalar perturbation spectra at large scales . if inflation happened at a very high scale , as is the case for the sm false vacuum scenario , quantum fluctuations during inflation produced a background of gravitational waves with a relatively large amplitude .
as argued in @xcite , the tensor to scalar ratio , combined with the top quark and higgs boson mass measurements , does represent a test of the hypothesis that inflation started from the sm false vacuum .
the upper bounds on @xmath1 provided by the wmap @xcite and planck experiments @xcite were too weak for the sake of such test .
it was anyway possible to conclude @xcite that , for the sm false vacuum to be a realistic inflationary scenario , an experimental detection of @xmath1 would have been possible in the case that the top quark mass turned out to be close to its lower allowed value at @xmath4 .
the recent measurement of @xmath1 by the bicep2 collaboration @xcite , @xmath5 at @xmath3 , on the one hand represents a hint in favor of the hypothesis that inflation took place in the sm shallow false vacuum and , on the other hand , allows to perform a sensible test of the allowed parameter space for such scenario .
that is the goal of the present letter .
let us consider the higgs potential in the sm of particle physics .
for very large values of the higgs field @xmath6 , the quadratic term @xmath7 can be neglected and we are left with the quartic term , whose dimensionless coupling @xmath8 depends on the energy scale , which can be identified with the field @xmath6 itself : @xmath9 it is well known that , for some narrow band of the higgs and top masses , the higgs potential develops a new local minimum @xcite .
if the higgs field is trapped in a cold coherent state in the false minimum @xmath10 and dominates the energy density of the universe , the standard friedmann equation leads to a stage of inflationary expansion h^2 h_i^2 , a(t)e^h_i t , [ eq - m ] where @xmath11 is the scale factor , @xmath12 is the hubble rate and @xmath13 gev is the reduced planck mass .
a nontrivial model - dependent ingredient is how to achieve a graceful exit from inflation . in order to end inflation
the higgs field has to tunnel to the other side of the potential barrier by nucleating bubbles @xcite that eventually collide and percolate .
subsequently the higgs field could roll down the potential , reheat the universe and relax in the electroweak vacuum .
a graceful exit can be generically realized only if at the end of inflation there is a very shallow false minimum , otherwise the tunneling rate becomes negligibly small , the probability being exponentially sensitive to the barrier @xcite .
the shape of the potential is thus very close to an inflection point configuration .
this leads to a powerful generic prediction for the scale of inflation , and therefore for @xmath1 @xcite .
so , if the false vacuum is very shallow , the specific model only affects the prediction for the spectral index of cosmological density perturbations @xmath14 , see e.g. the models of refs .
@xcite . using the renormalization group equations ( rge ) at next - to - next - to - leading order ( nnlo ) , _ i.e. _ 3-loops for beta functions @xcite and the 2-loops for matching conditions @xcite
, we investigated @xcite the values of the top quark and higgs masses allowing for the presence of a second degenerate minimum or a shallow local minimum at high scale . for the technical details on the rge procedure at nnlo we refer the interested reader to ref .
@xcite , while here we summarize only the issues relevant for the present work .
the are two main sources of uncertainty in the rge calculation at nnlo .
the first is of experimental nature and is associated to the uncertainty in the determination of the value of @xmath15 , for which the pdg @xcite gives the range @xmath16 at @xmath3 .
the second is of theoretical nature and is associated to the matching conditions .
as suggested in ref .
@xcite , using the value of the top mass in the @xmath17 scheme , @xmath18 , one has to consider only the theoretical uncertainty associated to the matching of @xmath8 ; this kind of top mass is experimentally known with a quite large @xmath3 error , @xmath19 gev according to @xcite , and @xmath20 gev according to the pdg @xcite .
using instead the top pole mass , the theoretical error becomes bigger because one has to include also the uncertainty associated to the matching of the top yukawa coupling ( see e.g. @xcite ) ; the top pole mass @xmath21 is however known quite accurately , @xmath22 gev at @xmath3 according to the pdg @xcite , and @xmath23 gev according to the recent combination of the atlas , cdf , cms , d0 measurements @xcite . in the present analysis , as a crosscheck , we adopt both procedures , finding that the two methods essentially provide similar results ( as they should ) , even though the second method is somewhat more stringent .
we first start with the analysis based on the @xmath17 top mass . in fig.[fig-1 ] , the ( red and orange ) lines spanning the @xmath4 range of @xmath15 , show the points in the @xmath24 plane where the shallow sm false minima exists .
the thickness of the lines is associated to the theoretical uncertainty of the nnlo procedure , due to the matching of @xmath8 .
the latter are extremely close to the line marking the transition from stability to metastability ( two degenerate vacua ) , so that it is not possible to distinguish them by eye , the difference being a few mev for the top mass @xcite .
the ( green ) shaded horizontal region corresponds to the @xmath25 range @xmath19 gev , as calculated in ref .
@xcite .
the ( pink ) shaded vertical regions are the 1 and 2 @xmath26 ranges of @xmath27 according to the pdg @xcite . the shallow false minimum configuration ( and ,
more generally , any configuration close to it , like e.g. two degenerate minima ) is thus compatible with the higgs and top quark mass values , provided the latter is quite light . , taken to vary in its @xmath4 range .
the ( green ) shaded horizontal region corresponds to the @xmath25 range @xmath19 gev @xcite .
the ( pink ) shaded vertical regions are the 1 and 2 @xmath26 ranges of @xmath27 @xcite .
the ( blue ) shaded bands are the @xmath28 and @xmath29 @xmath26 ranges of bicep2 , @xmath30 @xcite ; for reference , also the value @xmath31 is displayed . , width=283 ] .2 cm for each shallow false minimum configuration the highness of the higgs potential at the false minimum , @xmath32 , can be calculated and , thanks to eq .
( [ eq - m ] ) , the same is true for @xmath33 .
let now consider the tensor to scalar ratio of cosmological perturbations .
the amplitude of density fluctuations in the observed universe as seen by the cmb and large - scale structure data is parametrized by the power spectrum in @xmath34-space p_s(k)=_r^2 ( ) ^n_s-1 , where @xmath35 is the amplitude at some pivot point @xmath36 .
we consider the best - fit value from @xcite , @xmath37 at @xmath38 . in any inflationary model that can be analyzed through the slow - roll approximation
, there is a relationship between the scale of inflation , the amplitude of density perturbations , and the amount of gravity waves that can be produced : _
r^2 = .
[ eq - r ] if inflation actually started from a sm shallow false minimum , then each point in the @xmath24 plane has to be associated with a specific value of @xmath1 .
this is done in fig.[fig-1 ] , where the ( blue ) shaded bands represents the @xmath28 and @xmath29 @xmath26 ranges of bicep2 measurement @xmath30 @xcite ; for reference , also the value @xmath31 is displayed . taking into account the theoretical error in the determination of the position of the false minimum ( the thickness of the red and orange lines ) , the position of the ( blue ) @xmath1 bands is also uncertain by about @xmath39 gev along the vertical axis .
one can see that , given the results from bicep2 , the shallow false minimum is a viable framework for models of higgs inflation provided that : i ) the higgs mass is close to its upper 1 - 2 @xmath26 range , more precisely between @xmath40 and @xmath41 gev , or ii ) both @xmath15 and the top mass @xmath18 are quite small , say respectively close to the lower 1 - 2 @xmath26 range ( between @xmath42 and @xmath43 ) and between @xmath44 and @xmath45 gev . clearly , the smaller is @xmath1 , the more the three previous parameters can go in the direction of their central values .
it will thus be important to further improve the @xmath1 measurement in the future .
show the points where the sm false vacuum configuration is realized , their thickness being the theoretical error in the nnlo rge procedure .
the ( green ) shaded horizontal region corresponds to the @xmath28 and @xmath28 @xmath26 ranges of the combined measurement by the atlas , cdf , cms , d0 collaborations , @xmath23 gev @xcite .
the ( pink ) shaded vertical regions are the 1 and 2 @xmath26 ranges of @xmath27 @xcite .
the ( blue ) shaded bands are the @xmath28 and @xmath29 @xmath26 ranges of bicep2 , @xmath30 @xcite ; for reference , also the value @xmath31 is displayed .
, width=283 ] .2 cm the same conclusions can be drawn by making the analysis using the top pole mass , see fig .
[ fig-2 ] . as the theoretical error has to include also the uncertainty due to the matching of the top yukawa coupling , the thickness of the lines is bigger than with the previous method , and turns out to be about @xmath46 gev in the vertical axis ( see e.g. @xcite ) . for the sake of clearness
, we thus display only three lines , corresponding the the central and @xmath4 values of @xmath15 .
taking into account the theoretical error in the determination of the position of the false minimum ( the thickness of the red and orange lines ) , the position of the ( blue ) bands representing the bicep2 result is also uncertain by a shift along the vertical axis of about @xmath46 gev .
the ( green ) shaded horizontal regions show the 1 and 2 @xmath26 ranges of the combined measurement by the atlas , cdf , cms , d0 collaborations , @xmath23 gev @xcite .
the ( pink ) shaded vertical regions are the 1 and 2 @xmath26 ranges of @xmath27 according to the pdg @xcite . also from fig .
[ fig-2 ] one can see again that , for the shallow false minimum configuration , the bicep2 results are compatible with the present ranges for @xmath15 , @xmath27 , @xmath21 .
in particular the overlapping requires : @xmath21 close to its lower @xmath4 value , @xmath27 close to its upper @xmath4 value , @xmath15 close to its central one . notice that the bicep2 result at @xmath3 , @xmath30 @xcite , corresponds to the range @xmath47 gev , and to @xmath48 gev . summarizing , we have argued that the present status of the measurements of @xmath1 , @xmath21 , @xmath27 , @xmath15 , is consistent with of the hypothesis @xcite that inflation occurred in a sm shallow false vacuum at about @xmath49 gev . to account for the bicep2 result at @xmath3 , @xmath30 @xcite ,
in particular , the top quark mass should be close to its lower @xmath4 range , while the higgs mass should be close to its @xmath50 upper one , see figs .
[ fig-1 ] and [ fig-2 ] .
it is intriguing that for such values of the top and higgs masses , the quadratic divergences of the higgs mass ( corresponding to the running veltman condition ) cancel at the planck scale , see fig . 2 of ref .
@xcite .
future precision measurements of the top quark and higgs masses will thus be crucial to further test the sm shallow false minimum inflationary scenario . |
in the last few decades corrections to newton s law of gravitation and constraints on them have become the subject of considerable study ( see the monograph @xcite and reviews @xcite ) . from the experimental standpoint , it is of most importance that at separations between the test bodies below 0.1 mm newton s law is not confirmed by measurements with sufficient precision .
theoretically , many extensions of the standard model of elementary particles and their interactions predict corrections to the newton law of power- and yukawa - type due to exchange of light and massless elementary particles @xcite . on the other hand , similar corrections
are predicted by the extra - dimensional physics with a low - energy compactification scale @xcite .
this makes the search for such corrections , or at least constraining their parameters , interesting for the problems of dark matter and unification of gravitation with other fundamental interactions .
a lot of successful work has been done on constraining the power- and yukawa - type corrections to newton s law of gravitation from gravitational experiments of etvos- and cavendish - type @xcite .
although the most strong constraints on the power - type corrections were obtained in this way , it was found that the resulting constraints on the yukawa - type corrections become much weaker in the interaction range below a few micrometers .
this is explained by the fact that at sufficiently small separations between the test bodies the van der waals @xcite and casimir force @xcite becomes the dominant background force in place of gravitation .
the two names belong to a single force of quantum origin caused by the zero - point and thermal fluctuations of the electromagnetic field .
they are usually used at short and relatively large separations , respectively , where the effects of relativistic retardation are immaterial or , on the contrary , are influential and should be taken into account . the possibility to constrain corrections to newton s law from the van der waals and casimir force
was proposed long ago @xcite for the cases of yukawa - type and power - type corrections , respectively . at that time , however , reasonably precise measurements of the van der waals and casimir force were not available .
things have changed during the last 15 years when a lot of more precise experiments on measuring the casimir force between metallic , dielectric and semiconductor bodies have been performed ( see reviews @xcite ) .
the measure of agreement between the measurement data of these experiments and theoretical description of the casimir force in the framework of the lifshitz theory resulted in the strengthening of previously known constraints on the parameters of yukawa - type corrections up to a factor of @xmath0 @xcite . using different experiments on measurement of the casimir force
, the strongest constraints on the corrections to newton s law were obtained over a wide interaction region from about 1 nm to a few micrometers .
note that for shorter interaction regions the strongest constraints on the yukawa - type corrections follow from precise atomic physics @xcite , whereas starting from a few micrometers the gravitational experiments @xcite remain the most reliable source of such constraints . in this paper
we obtain constraints on the yukawa - type corrections to newton s gravitational law from two recently performed experiments on measuring the casimir interaction by means of an atomic force microscope ( afm ) .
each of these experiments is highly significant , as compared with all earlier measurements of the casimir interaction . in the first @xcite the dynamic afm
was used to measure the gradient of the casimir force between a plate and a sphere both coated with the ferromagnetic metal ni with no spontaneous magnetization . as a result , the predictions of the lifshitz theory generalized for the case of magnetic materials more than 40 years ago @xcite were experimentally confirmed .
the distinguished feature of the experiment with two magnetic surfaces is also that it allows to shed light on the role of some important systematic effects ( see sec .
ii for details ) and , thus , remove any doubt in the reliability of constraints obtained . in the second experiment of our interest here @xcite the static afm was used to measure the casimir force between a plate and a sphere both with sinusoidally corrugated surfaces coated with nonmagnetic metal au . the unusual feature of this experiment , as compared with earlier performed experiments with corrugated surfaces , is that the casimir force was measured at various angles between the longitudinal corrugations on both bodies .
this introduced into the problem an additional parameter ( the angle between corrugations ) that can be chosen to obtain the most strong constraints from the measure of agreement between the experimental data and theory of the casimir force for corrugated surfaces based on the derivative expansion @xcite .
the constraints on corrections to newton s law obtained by us from the experiment with magnetic surfaces are in agreement with those obtained earlier @xcite for smooth au surfaces by means of dynamic afm @xcite , but slightly weaker due to different densities of au and ni . the advantage of constraints following from the experiment with ni surfaces is that they are not only fully justified on their own , but add substantiation to the constraints obtained from the experiments with nonmagnetic metal surfaces . as to constraints obtained from the experiment with corrugated surfaces ,
they are stronger up to a factor of 4 than the most strong constraints reported so far @xcite in the interaction region from 11.6 to 29.2 nm .
the paper is organized as follows . in sec . ii we present the exact expression for the yukawa - type interaction in the experimental configuration of ref .
@xcite and derive the respective constraints on corrections to newton s gravitational law .
the advantages of using magnetic materials are elucidated .
section iii is devoted to the experiment with corrugated surfaces @xcite . here , we derive an expression for the yukawa - type force in configurations with different angles between corrugations .
the obtained expression is used to find the stronger constraints on corrections to newton s law .
some modifications in the setup are proposed allowing further strengthening of the constraints in configurations with corrugated surfaces . in sec .
iv the reader will find our conclusions and discussion .
we begin with the standard parametrization of the spin - independent yukawa - type correction to newton s gravitational law @xcite ( for spin - dependent corrections see refs .
@xcite ) . the total gravitational potential between the two point - like masses @xmath1 and @xmath2 spaced at a separation @xmath3 takes the form @xmath4 where @xmath5 and @xmath6 are the newtonian part and the yukawa - type correction , respectively . here , @xmath7 is the newtonian gravitational constant , and @xmath8 and @xmath9 are the strength and interaction range of the yukawa - type correction .
similar to ref .
@xcite it can be shown that at separations below a few micrometers the newtonian gravitational force between the test bodies @xmath10 and @xmath11 in experiments under consideration is much smaller than the error in measurements of the casimir force .
because of this , in all subsequent calculations the newtonian potential can be neglected , and the yukawa - type addition to it is considered on the background of the measured casimir force
. then the gravitational force acting between the test bodies at short separations can be obtained by the integration of the yukawa - type correction @xmath6 defined in eq .
( [ eq1 ] ) over the volumes of both bodies @xmath12 here , @xmath13 is the closest separation between the test bodies and @xmath14 and @xmath15 are the respective mass densities ( note that @xmath16 and @xmath17 are not constant because in the experiments used below each test body consists of several homogeneous layers of different densities ) .
the gravitational force due to the yukawa - type correction and its gradient are given by @xmath18 in the experiment @xcite the gradient of the casimir force was measured between a ni - coated plate and a ni - coated hollow microsphere attached to the tip of an afm cantilever operated in the dynamic regime @xcite .
the silicon plate ( @xmath10 ) of a few millimeter diameter and thickness can be considered as having an infinitely large area and an infinitely large thickness when we have to deal with the submicrometer region of @xmath9 .
the density of si is @xmath19 . for technological purposes
the si plate was coated first with a layer of cr having a thickness @xmath20 nm and density @xmath21 and then with a layer of al having a thickness @xmath22 nm and density @xmath23 .
finally the plate was coated with an outer layer of magnetic metal ni with a thickness @xmath24 nm and density @xmath25 .
the hollow microsphere ( @xmath11 ) was made of glass with density @xmath26 .
the thickness of the spherical envelope was @xmath27 m .
the sphere was also coated with successive layers of cr , al and ni having the thicknesses @xmath28 , @xmath29 , and @xmath30 nm . the outer radius of the sphere with all the coatings included is @xmath31 m .
the exact integration over the volumes of a plate and a sphere in eq.([eq2 ] ) with account of their layer structure can be performed like in ref .
@xcite . then , substituting the obtained result in eq .
( [ eq3 ] ) , we arrive at @xmath32 where the following notations are introduced : @xmath33 and the following notation is introduced @xmath34 the constraints on the parameters ( @xmath35 ) , which are often referred to as the parameters of _ non - newtonian gravity _ , can be obtained from the comparison between the measurement data for the gradient of the casimir force @xmath36 and respective theory . in ref . @xcite it was found that within the entire separation region from 223 to 550 nm there is an excellent agreement between the data and theoretical predictions of the lifshitz theory of the van der waals and casimir force @xcite with omitted relaxation properties of conduction electrons ( the so - called _ plasma model approach _ ) .
the predictions of the lifshitz theory with included relaxation properties of free charge carriers ( the so - called _ drude model approach _ ) were excluded by the measurement data at a 95% confidence level within the separation region from 223 to 350 nm ( in the end of this section we provide a brief discussion of different approaches to the lifshitz theory which is essential for obtaining constraints on non - newtonian gravity ) .
the measure of agreement with the adequate theory is characterized by the total experimental error @xmath37 in the measured gradient of the casimir force determined at a 67% confidence level @xcite . keeping in mind that within the limits of this error no additional interaction of yukawa - type was observed , the constraints on the parameters @xmath9 and @xmath8 can be obtained from the inequality @xmath38 we have substituted eqs .
( [ eq4])([eq6 ] ) in eq .
( [ eq7 ] ) and analyzed the resulting inequality at different separations .
it was found that for @xmath39 nm the strongest constraints are determined at the shortest separation @xmath40 nm where @xmath41n / m @xcite . for @xmath42 nm and @xmath43 nm
the strongest constraints follow at @xmath44 and 300 nm , respectively ( with respective @xmath45 and @xmath46n / m ) . finally , at @xmath47 nm
the strongest constraints are obtained at @xmath48 nm ( @xmath49n / m ) .
the resulting constraints are shown by the solid line in fig .
[ fg1 ] . here and
below the region of ( @xmath35 ) plane above each line is prohibited and below is allowed by the results of respective experiment . in the same figure by the dashed line we show the constraints obtained in ref .
@xcite from measurements of the gradient of the casimir force between two au - coated surfaces by means of dynamic afm @xcite .
the dotted line shows the constraints obtained @xcite from the experiment on measuring gradient of the casimir force between an au - coated sphere and a ni - coated plate @xcite using the same setup . as can be seen in fig .
[ fg1 ] , the constraints indicated by the solid line are slightly weaker than those shown by the dashed and dotted lines .
this is caused by the fact that density of ni is smaller than density of au and by different experimental errors .
note also that our constraints shown by the solid line can be obtained in a simpler way by using the proximity force approximation @xcite @xmath50 to calculate the gradient of the yukawa - type force , where @xmath51 is the energy per unit area of yukawa - type interaction between two plane - parallel plates having the same layer structure as our test bodies . according to the results of refs .
@xcite , this is possible under the conditions @xmath52 which are satisfied in our experimental configuration with a wide safety margin . in this case
the function @xmath53 with any argument @xmath3 can be approximately replaced with @xmath54 .
it would be interesting also to compare the constraints on non - newtonian gravity obtained here from the experiment with two ni surfaces ( solid line in fig .
[ fg1 ] ) with the strongest constraints obtained so far using the alternative setups . for this purpose in fig .
[ fg2 ] we reproduce the solid line of fig . [ fg1 ] as the solid line 1 .
the solid line 2 in fig .
[ fg2 ] was obtained @xcite from measurements of the thermal casimir - polder force between @xmath55pb atoms belonging to the bose - einstein condensate and a sio@xmath56 plate @xcite , and the solid line 3 was obtained @xcite from measurements of gradient of the casimir force between an au sphere and a rectangular corrugated semiconductor ( si ) plate by means of a micromachined oscillator @xcite .
next , the solid line 4 in fig .
[ fg2 ] was found from an effective measurement of the casimir pressure between two parallel au plates by means of a micromachined oscillator @xcite , and the dashed line was obtained from the casimir - less experiment @xcite . as can be seen in fig .
[ fg2 ] , various constraints obtained using quite different setups are consistent with the constraints of line 1 obtained from the most recent experiment with two magnetic surfaces . in the end of this section it is pertinent to note that the experiment with two magnetic surfaces @xcite plays the key role in the test of validity of the lifshitz theory . keeping in mind that constraints on non - newtonian gravity are derived from the measure of agreement between the measurement data and theory , this experiment is also important to validate the reliability of constraints obtained . as mentioned above
, the lifshitz theory is in agreement with the plasma model approach to the casimir force , which disregards the relaxation properties of free charge carriers , and excludes the drude model approach taking these properties into account ( see the experiments of refs .
@xcite and earlier experiments reviewed in refs .
this is against expectations of many and gave rise to the search of some systematic effects which could reverse the situation .
after several unsuccessful attempts ( see ref .
@xcite for a review ) the influence of large surface patches was selected as the most probable systematic effect which could bring the data in agreement with the drude model approach @xcite . in two experiments on measuring the casimir force between au surfaces
@xcite hypothetical large patches were described by models with free fitting parameters and used in respective fitting procedures .
in these experiments , which are not independent measurements of the casimir force , an agreement of the data with the drude model approach has been claimed ( see refs .
@xcite for a critical discussion ) .
the crucial point to underline here is that for nonmagnetic metals the drude model approach leads to smaller gradients of the casimir force than the plasma model approach @xcite .
thus , the effect of large patch potentials ( which leads to an attraction similar to the casimir force ) is added to the predictions of the drude model approach and might make the total theoretical force compatible with the measurement data @xcite .
by contrast , for two magnetic metals the drude model approach leads to larger gradients of the casimir force than the plasma model approach @xcite .
thus , if the effect of patches were important in this case , it would further increase the disagreement between the predictions of the drude model approach and the measurement data observed in ref .
this confirms that surface patches do not play an important role in precise experiments on measuring the casimir force in accordance with the model of patches @xcite predicting a negligibly small effect from patches @xcite .
recently the patches on au samples used in measurements of the casimir force were investigated by means of kelvin probe microscopy @xcite .
the force originating from them was found to be too small to affect the conclusions following from precise measurements of the casimir force .
it is the matter of fact that the experimental data of all independent measurements of the casimir interaction between both nonmagnetic and magnetic metals are in excellent agreement with the predictions of the lifshitz theory combined with the plasma model approach and exclude the drude model approach .
although the fundamental reasons behind this fact have not yet been finally understood , the constraints on non - newtonian gravity obtained on this basis can be already considered as reliable enough .
in sec . ii we have used the most recent measurement of the casimir interaction where the material dependence played a major role in theory - experiment comparison .
another recent experiment @xcite is of quite a different nature . in ref .
@xcite the normal casimir force acting perpendicular to the surface was measured between the sinusoidally corrugated surfaces of a sphere and a plate .
the corrugated boundary surfaces have long been used in measurements of the casimir force ( see refs .
@xcite for a review ) .
for example , the normal casimir force between a rectangular corrugated semiconductor ( si ) plate and a smooth au sphere has been measured by means of a micromachined oscillator and compared with theory based on the exact scattering approach @xcite . the obtained constraints on non - newtonian gravity are discussed in sec .
ii ( see solid line 3 in fig . [ fg2 ] ) .
a further example is the lateral casimir force between a sinusoidally corrugated plate and a sinusoidally corrugated sphere , both coated with au , which was measured and compared with exact theory in refs .
this experiment resulted in the maximum strengthening of constraints on non - newtonian gravity from the casimir effect by a factor of @xmath57 discussed in sec .
i. in experiments with corrugated surfaces the nontrivial geometry plays a major role in the theory - experiment comparison whereas different approaches to the description of material properties can not be differentiated due to the lower experimental precision . the specific feature of the experiment of ref .
@xcite is that the normal casimir force between a sinusoidally corrugated au - coated plate and a sinusoidally corrugated au - coated sphere was measured at various angles between corrugations using an afm .
the plate in this experiment is the diffraction grating with uniaxial sinusoidal corrugations of period @xmath58 nm and amplitude @xmath59 nm .
the grating was made of hard epoxy with density @xmath60 and coated with an au layer of thickness @xmath61 nm .
the corrugated plate was used as a template for the pressure imprinting of the corrugations on the bottom surface of a sphere .
the polystyrene sphere has a density @xmath62 .
it was coated with a layer of cr of thickness @xmath63 nm , then with a layer of al of thickness @xmath64 nm and finally with a layer of au of thickness @xmath65 nm .
the outer radius of the coated sphere is @xmath66 m
. the imprinted corrugations on the sphere have the same period as on the plate and the amplitude @xmath67 nm .
the size of an imprint area was measured to be @xmath68 m , i.e. , it is much larger than @xmath69 . in ref .
@xcite the casimir force between the sphere and the plate was measured at the following angles between the axes of corrugations on both bodies : @xmath70 , @xmath71 , @xmath72 , and @xmath73 .
now we calculate the yukawa - type force in the experimental configuration of ref . @xcite . for this purpose
we first consider the yukawa - type energy per unit area in the configuration of two plane - parallel plates spaced at a separation @xmath13 having the same layer structure as a plate and a sphere in the experiment .
the result is @xcite @xmath74 where now @xmath75 next , we introduce corrugations at an angle @xmath76 on the parallel plates and find their effect by means of the geometrical averaging @xcite @xmath77 here , @xmath78 is the energy per unit area defined in eq .
( [ eq9 ] ) calculated at different separations @xmath79 between the corrugated plates which are assumed parallel to the ( @xmath80 ) plane @xmath81 note that there is no phase shift between the corrugations on both plates , so that @xmath82 .
finally , to obtain the yukawa - type force between a corrugated plate and a corrugated sphere , we apply the proximity force approximation ( [ eq7a ] ) taking into account different radii of separate spherical layers . after an easy calculation using eqs .
( [ eq9])([eq12 ] ) , the yukawa - type force between a corrugated plate and a corrugated sphere takes the form @xmath83 where @xmath84 and the function @xmath85 is defined as @xmath86/\lambda}. \label{eq15}\ ] ] for zero angle between corrugations at both surfaces ( @xmath87 ) one arrives to a more simple representation @xmath88/\lambda}. \label{eq16}\ ] ] the integral in eq .
( [ eq16 ] ) can be evaluated analytically using the formula 2.5.10(3 ) in ref .
@xcite if there is an integer @xmath89 such that @xmath90 . in this case @xmath91 where @xmath92 is the bessel function of imaginary argument .
if @xmath93 where @xmath94 , eq . ( [ eq17 ] ) is satisfied only approximately .
if in the interaction region of our interest ( see fig .
[ fg4 ] below ) it occurs @xmath95 , the maximum error arising from the use of eq .
( [ eq17 ] ) achieves 5% . in the case
@xmath96 this error is equal to @xmath97% . in the general case of an arbitrary @xmath76
the quantity @xmath85 can be computed numerically . in fig .
[ fg3 ] the computational results are plotted by the solid lines as functions of @xmath9 at @xmath70 , @xmath71 , @xmath72 , and @xmath73 used in the experiment of ref .
@xcite from bottom to top , respectively , to a logarithmic scale . the measurement data of ref .
@xcite for the normal casimir force between corrugated surfaces were compared with the results of numerical computations based on the derivative expansion approach @xcite and a good agreement was found within the limits of the experimental errors @xmath98 determined at the 67% confidence level ( minor disagreement at the shortest separations in fig . 3 of ref .
@xcite comes from the use of an oscillator model in place of the optical data for the complex index of refraction ) .
then the constraints on the parameters @xmath9 and @xmath8 of the corrections to newton s law were found from the inequality @xmath99 where @xmath100 is given by eq .
( [ eq13 ] ) with the notations in eqs .
( [ eq10 ] ) , ( [ eq14 ] ) and ( [ eq15 ] ) .
we have numerically analyzed eq .
( [ eq18 ] ) at different separations @xmath13 and with different values of the angle @xmath76 between corrugations .
the strongest constraints were obtained at the shortest separation @xmath101 nm where @xmath102pn .
they are shown by the solid lines in figs .
[ fg4](a)[fg4](d ) at the values of @xmath70 , @xmath71 , @xmath72 , and @xmath73 , respectively . for comparison purposes , the dashed lines 1 and 2 in figs .
[ fg4](a)[fg4](d ) show the strongest constraints obtained earlier @xcite within this interaction region from measurements of the lateral casimir force between sinusoidally corrugated surfaces @xcite and from effective measurements of the casimir pressure between metallic plates by means of a micromachined oscillator @xcite . as can be seen in fig .
[ fg4 ] , at any @xmath76 measurements of the normal casimir force between sinusoidally corrugated surfaces result in stronger constraints within some interaction region than were known so far .
thus at @xmath70 the strengthening of previously available constraints up to a factor 1.8 holds within the interaction region @xmath103 nm with the largest strengthening achieved at @xmath104 nm [ see fig .
[ fg4](a ) ] . at @xmath105 and @xmath72
the strengthening up to factors 2.8 and 3.5 occurs for @xmath106 nm and @xmath107 nm , respectively .
the maximum strengthening up to a factor 4 ( achieved at @xmath104 nm ) within the interaction region @xmath108 nm takes place at the angle between corrugations @xmath109 .
the obtained stronger constraints following from measurements of the normal casimir force between sinusoidally corrugated surfaces can be further strengthened at the expense of some modification of the experimental setup .
thus , it would be useful to switch from a static afm mode used in this experiment to the dynamic mode used in refs .
this results in a higher experimental precision though makes it necessary to perform measurements at larger separation distances . as an example
, we calculate the prospective constraints on @xmath35 which can be obtained from dynamic measurements of the gradient of the casimir force between corrugated surfaces at @xmath110 nm . in
so doing we assume that the total experimental error obtainable at this experiment is @xmath111n / m .
for the sake of simplicity we consider the case @xmath70 which does not lead to the maximum strengthening of the respective constraints .
then from eq .
( [ eq13 ] ) one obtains @xmath112 substituting eq .
( [ eq19 ] ) in the left - hand side of eq .
( [ eq7 ] ) adapted for the case of corrugated surfaces , we arrive at the prospective constraints shown by the dotted line in fig .
[ fg5 ] . in the same figure
the strongest constraints obtained @xcite from measurements of the lateral casimir force between sinusoidally corrugated surfaces @xcite , from effective measurements of the casimir pressure between metallic plates by means of a micromachined oscillator @xcite , and from the casimir - less experiment @xcite are indicated by the dashed lines 1 , 2 , and 3 , respectively . as can be seen in fig .
[ fg5 ] , the prospective constraints shown by the dotted line are stronger than the strongest current constraints over a wide interaction region from 12 to 160 nm . at the moment three different experiments are used to constrain the yukawa - type corrections to newton s law within this interaction region .
the maximum strengthening up to a factor of 12.6 occurs at @xmath104 nm .
in the foregoing we have obtained constraints on the parameters of yukawa - type corrections to newton s law of gravitation following from two recent experiments on measuring the casimir interaction .
each of these experiments is of particular interest , as compared with all previous work in the field . the experiment of ref .
@xcite pioneered measuring the gradient of the casimir force between two magnetic surfaces and confirmed the predictions of the lifshitz theory combined with the plasma model approach . in this way it was demonstrated that magnetic properties of the material boundaries influence the casimir force .
the outstanding property of magnetic materials is that the force gradients predicted by the drude model approach are larger than those predicted by the plasma model approach ( just opposite to the case of nonmagnetic metals ) .
thus , it was confirmed that such a widely discussed systematic effect as the patch potentials can not be used for the reconciliation of the measurement data with the drude model approach leading to further support of constraints on non - newtonian gravity obtained from the measure of agreement between experiment and theory . although constraints following from the experiment with magnetic surfaces are slightly weaker than the previously known ones ( this is due to smaller density of ni as compared to au ) , the increased reliability can be considered as an advantage . the experiment of ref .
@xcite pioneered measurements of the normal casimir force between metallized sinusoidally corrugated surfaces at various angles between corrugations .
it was demonstrated that the casimir force depends on these angles in accordance with theory using the derivative expansion .
we have calculated the yukawa - type force in the experimental configuration with corrugated surfaces and obtained the respective constraints on its parameters .
it was shown that the strength of constraints increases with increasing angle between corrugations .
the maximum strengthening up to a factor of 4 , as compared to the strongest previously known constraints , was shown to occur within the interaction range from 11.6 to 29.2 nm .
we have also proposed some modification in the measurement scheme allowing strengthening of the previously known constraints up to a factor of 12.6 within a wide interaction region presently covered using the results of three different experiments .
this means that measurements of the casimir interaction retain considerable potential for further strengthening of constraints on the yukawa - type corrections to newton s gravitational law in submicrometer interaction region .
this work was supported by the nsf grant no .
phy0970161 and doe grant def010204er46131 ( u.m . ) .
99 e. fischbach and c. l. talmadge , _ the search for non - newtonian gravity _ ( springer , new york , 1999 ) .
e. g. adelberger , b. r. heckel , c. w. stubbs , and w. f. rogers , ann . rev .
nucl . part .
sci . * 41 * , 269 ( 1991 ) .
e. g. adelberger , b. r. heckel , and a. e. nelson , ann .
nucl . part .
sci . * 53 * , 77 ( 2003 ) .
e. g. adelberger , j. h. gundlach , b. r. heckel , s. hoedl , and s. schlamminger , progr .
. phys . * 62 * , 102 ( 2009 ) .
v. m. mostepanenko , v. b. bezerra , g. l. klimchitskaya , and c. romero , int . j. mod .
a * 27 * , 1260015 ( 2012 ) .
r. d. peccei and h. r. quinn , phys .
* 38 * , 1440 ( 1977 ) .
s. ferrara , j. scherk , and b. zumino , nucl .
b * 121 * , 393 ( 1977 ) .
y. fujii , int .
phys . a * 6 * , 3505 ( 1991 ) .
s. deser and b. zumino , phys .
* 38 * , 1433 ( 1977 ) .
s. dimopoulos and g. f. giudice , phys .
b * 379 * , 105 ( 1996 ) .
i. antoniadis , n. arkani - hamed , s. dimopoulos , and g. dvali , phys .
b * 436 * , 257 ( 1998 ) .
n. arkani - hamed , s. dimopoulos , and g. dvali , phys .
b * 429 * , 263 ( 1998 ) .
n. arkani - hamed , s. dimopoulos , and g. dvali , phys .
d * 59 * , 086004 ( 1999 ) .
e. g. floratos and g. k. leontaris , phys .
b * 465 * , 95 ( 1999 ) .
a. kehagias and k. sfetsos , phys .
b * 472 * , 39 ( 2000 ) .
s. j. smullin , a. a. geraci , d. m. weld , j. chiaverini , s. holmes , and a. kapitulnik , phys .
d * 72 * , 122001 ( 2005 ) .
a. a. geraci , s. j. smullin , d. m. weld , j. chiaverini , and a. kapitulnik , phys .
d * 78 * , 022002 ( 2008 ) .
v. a. parsegian , _ van der waals forces : a handbook for biologists , chemists , engineers , and physicists _ ( cambridge university press , cambridge , 2005 ) . m. bordag , g. l. klimchitskaya , u. mohideen , and v. m. mostepanenko , _ advances in the casimir effect _ ( oxford university press , oxford , 2009 ) . v. a. kuzmin , i. i. tkachev , and m. e. shaposhnikov , pisma v zhetf * 36 * , 49 ( 1982 ) [ jetp lett . * 36 * , 59 ( 1982 ) ] . v. m. mostepanenko and i. yu .
sokolov , phys .
a * 125 * , 405 ( 1987 ) .
g. l. klimchitskaya , u. mohideen , and v. m. mostepanenko , rev .
phys . * 81 * , 1827 ( 2009 ) .
g. l. klimchitskaya , u. mohideen , and v. m. mostepanenko , int .
b * 25 * , 171 ( 2011 ) .
a. w. rodriguez , f. capasso , and s. g. johnson , nature photonics * 5 * , 211 ( 2011 ) .
r. onofrio , new j. phys .
* 8 * , 237 ( 2006 ) .
v. b. bezerra , g. l. klimchitskaya , v. m. mostepanenko , and c. romero , phys .
d * 81 * , 055003 ( 2010 ) .
s. g. karshenboim , phys .
d * 82 * , 073003 ( 2010 ) .
a. a. banishev , g. l. klimchitskaya , v. m. mostepanenko , and u. mohideen , phys .
* 110 * , 137401 ( 2013 ) .
richmond and b. w. ninham , j. phys .
c : solid st .
* 4 * , 1988 ( 1971 ) .
a. a. banishev , j. wagner , t. emig , r. zandi , and u. mohideen , phys .
lett . * 110 * , 250403 ( 2013 ) . c. d. fosco , f. c. lombardo , and f. d. mazzitelli , phys .
d * 84 * , 105031 ( 2011 ) .
g. bimonte , t. emig , r. l. jaffe , and m. kardar , europhys . lett . * 97 * , 50001 ( 2012 ) .
g. bimonte , t. emig , and m. kardar , appl .
. lett . * 100 * , 074110 ( 2012 ) .
g. l. klimchitskaya , u. mohideen , and v. m. mostepanenko , phys .
d * 86 * , 065025 ( 2012 ) .
chang , a. a. banishev , r. castillo - garza , g. l. klimchitskaya , v. m. mostepanenko , and u. mohideen , phys .
b * 85 * , 165443 ( 2012 )
. r. s. decca , d. lpez , e. fischbach , g. l. klimchitskaya , d. e. krause , and v. m. mostepanenko , phys .
d * 75 * , 077101 ( 2007 ) .
r. s. decca , d. lpez , e. fischbach , g. l. klimchitskaya , d. e. krause , and v. m. mostepanenko , eur .
j. c * 51 * , 963 ( 2007 ) .
b. a. dobrescu and i. mocioiu , jhep * 0611 * , 005 ( 2006 ) .
l. hunter , j. gordon , s. peck , d. ang , and j .- f .
lin , science * 339 * , 928 ( 2013 ) .
m. bordag , b. geyer , g. l. klimchitskaya , and v. m. mostepanenko , phys .
d * 62 * , 011701(r ) ( 2000 ) .
r. s. decca , e. fischbach , g. l. klimchitskaya , d. e. krause , d. lpez , and v. m. mostepanenko , phys .
d * 79 * , 124021 ( 2009 ) .
a. a. banishev , c .- c .
chang , g. l. klimchitskaya , v. m. mostepanenko , and u. mohideen , phys .
b * 85 * , 195422 ( 2012 ) .
e. fischbach , g. l. klimchitskaya , d. e. krause , and v. m. mostepanenko , eur .
j. c * 68 * , 223 ( 2010 ) .
j. m. obrecht , r. j. wild , m. antezza , l. p. pitaevskii , s. stringari , and e. a. cornell , phys . rev
98 * , 063201 ( 2007 ) .
v. b. bezerra , g. l. klimchitskaya , v. m. mostepanenko , and c. romero , phys .
d * 83 * , 075004 ( 2011 ) .
y. bao , r. gurout , j. lussange , a. lambrecht , r. a. cirelli , f. klemens , w. m. mansfield , c. s. pai , and h. b. chan , phys .
105 * , 250402 ( 2010 ) .
r. s. decca , d. lpez , e. fischbach , d. e. krause , and c. r. jamell , phys .
* 94 * , 240401 ( 2005 ) . v. m. mostepanenko and g. l. klimchitskaya , int . j. mod .
a * 25 * , 2302 ( 2010 ) .
r. o. behunin , f. intravaia , d. a. r. dalvit , p. a. maia neto , and s. reynaud , phys .
a * 85 * , 012504 ( 2012 ) .
a. o. sushkov , w. j. kim , d. a. r. dalvit , and s. k. lamoreaux , nature phys .
* 7 * , 230 ( 2011 ) .
d. garcia - sanchez , k. y. fong , h. bhaskaran , s. lamoreaux , and h. x. tang , phys .
* 109 * , 027202 ( 2012 ) .
v. b. bezerra , g. l. klimchitskaya , u. mohideen , v. m. mostepanenko , and c. romero , phys .
b * 83 * , 075417 ( 2011 ) . g. l. klimchitskaya and v. m. mostepanenko , int
phys . a * 26 * , 3944 ( 2011 ) .
g. l. klimchitskaya , m. bordag , e. fischbach , d. e. krause , and v. m. mostepanenko , int . j. mod .
a * 26 * , 3918 ( 2011 ) .
g. l. klimchitskaya , m. bordag , and v. m. mostepanenko , int . j. mod .
a * 27 * , 1260012 ( 2012 ) .
m. bordag , g. l. klimchitskaya , and v. m. mostepanenko , phys . rev
. lett . * 109 * , 199701 ( 2012 ) .
b. geyer , g. l. klimchitskaya , and v. m. mostepanenko , phys .
b * 81 * , 104101 ( 2010 ) .
g. l. klimchitskaya , b. geyer , and v. m. mostepanenko , int .
a * 25 * , 2293 ( 2010 ) . c. c. speake and c. trenkel , phys .
lett . * 90 * , 160403 ( 2003 ) .
r. decca and g. voisin , aps bulletin * 58 * , r41.00008 ( 2013 ) .
chiu , g. l. klimchitskaya , v. n. marachevsky , v. m. mostepanenko , and u. mohideen , phys .
b * 80 * , 121402(r ) ( 2009 ) .
chiu , g. l. klimchitskaya , v. n. marachevsky , v. m. mostepanenko , and u. mohideen , phys .
b * 81 * , 115417 ( 2010 ) . a. p. prudnikov , yu .
a. brychkov , and o. i. marichev , _ integrals and series _ ( gordon and breach , new york , 1986 ) , vol .
1 . constraints on the parameters of yukawa - type corrections to newton s gravitational law obtained in this work from measurement of the gradient of the casimir force between two ni surfaces ( solid line ) , between two au surfaces ( dashed line ) and between an au and a ni surfaces ( dotted line ) . here and in figs .
[ fg2],[fg4 ] the regions of @xmath113 plane below each line are allowed and above each line are prohibited ( see text for further discussion ) . ]
constraints on the parameters of yukawa - type corrections to newton s gravitational law obtained in this work ( solid line 1 ) , in ref .
@xcite from measurements of the thermal casimir - polder force @xcite ( solid line 2 ) , in ref .
@xcite from measurements of the gradient of the casimir force between metallic and corrugated semiconductor surfaces @xcite ( solid line 3 ) , in refs .
@xcite from measurements of the gradient of the casimir force between two metallic surfaces ( solid line 4 ) , and in ref .
@xcite from the casimir - less experiment ( dashed line ) . ]
constraints on the parameters of yukawa - type corrections to newton s gravitational law obtained in this work ( solid line ) , in ref .
@xcite from measurements of the lateral casimir force between sinusoidally corrugated surfaces @xcite ( dashed line 1 ) , and from effective measurements of the casimir pressure between metallic plates by means of a micromachined oscillator @xcite ( dashed line 2 ) .
the angle between the axes of corrugations is equal to ( a ) @xmath70 , ( b ) @xmath105 , ( c ) @xmath114 , and ( d ) @xmath109 .
] prospective constraints on the parameters of yukawa - type corrections to newton s gravitational law which can be obtained from dynamic measurement of the gradient of the casimir force between sinusoidally corrugated surfaces are shown by the dotted line . for comparison purposes
the dashed lines 1 , 2 , and 3 indicate the strongest current constraints obtained in ref .
@xcite from measurements of the lateral casimir force between sinusoidally corrugated surfaces @xcite , from effective measurements of the casimir pressure between metallic plates by means of a micromachined oscillator @xcite , and in ref .
@xcite from the casimir - less experiment , respectively . ] | we obtain constraints on parameters of the yukawa - type corrections to newton s gravitational law from measurements of the gradient of the casimir force between surfaces coated with ferromagnetic metal ni and from measurements of the casimir force between au - coated sinusoidally corrugated surfaces at various angles between corrugations .
it is shown that constraints following from the experiment with magnetic surfaces are slightly weaker than currently available strongest constraints , but benefit from increased reliability and independence of systematic effects .
the constraints derived from the experiment with corrugated surfaces within the interaction region from 11.6 to 29.2 nm are stronger up to a factor of 4 than the strongest constraints derived from other experiments . the possibility of further strengthening of constraints on non - newtonian gravity by using the configurations with corrugated boundaries is proposed . |
few years ago chanfray et al.@xcite showed that the @xmath1 excitation is present also in the longitudinal response of a nucleus to an electro - magnetic ( e.m . ) probe .
the key issue in @xcite is elegant and simple .
let us assume that a nucleon at rest is excited to a @xmath0 through a m1 excitation .
the above is purely transverse but since the longitudinal / transverse separation is frame - dependent , a ( relativistic ) boost is able to give , in a moving frame , a longitudinal component too . since in a nucleus
the nucleons are moving thanks to the fermi motion , then the conversion of a single ( moving ) nucleon to a @xmath0 contains necessarily a ( small ) longitudinal component . here
we wish to improve this analysis by accounting both for the possibility of creating a @xmath0 through a e2 or c2 multipole and ( most relevant ) for many - body effects , that can also lead to a ( broadened ) resonant structure . concerning the former problem ( developed in sect .
2 ) one has to face the limited experimental knowledge of the e2 and c2 excitations .
the e2 current is believed to be small ( the coupling constant being expected , on the ground of the presently available data , to range between the -1% and -2.5% of the m1 one ) .
it is anyway negligible here , since it is transverse too and of higher order than m1 in the non - relativistic reduction .
moreover the e.m . coupling constant is squared in the inclusive response so that a factor of about @xmath2 is expected .
thus we shall not present in the following the effects of this contribution since , as expected , our calculations do not provide any perceptible change in the responses .
the c2 excitation is longitudinal and therefore could play a role even for small coupling constants ( in the peccei lagrangian@xcite , for example , it is about 10% of the m1 ) .
it is thus accounted for in the following leaving the corresponding coupling constant free to range between 0 and a 15% of the m1 coupling constant . the second problem , discussed in sect .
3 , is of course by far more involved and the calculations are consequently less firmly grounded .
we make use in the present paper of the boson loop expansion ( ble ) developed by two of us in some previous papers ( see for instance @xcite ) .
as it is well known , the longitudinal response is given by @xmath3 where @xmath4 is the polarisation propagator ( here written in configuration space ) .
the @xmath5-@xmath0 transition e.m .
current operator , @xmath6 , @xmath7 is expressed in terms of the nucleon and @xmath0 fields , i.e. @xmath8 and @xmath9 respectively .
the operators @xmath10 are usually written in fourier transform , according to @xcite , as @xmath11 @xmath12 @xmath13 here @xmath14 and @xmath15 are the @xmath0 and nucleon masses respectively and @xmath16 in absence of interaction ( relativistic free fermi gas model or simply rffg ) the evaluation of @xmath17 is rather simple : @xmath18 @xmath19 being the usual in - medium relativistic propagator for nucleons , namely @xmath20 ( where @xmath21 ) , and @xmath22 the usual rarita - schwinger propagator in the vacuum .
symbolically @xmath23 takes the form @xmath24 } \label{eq:2}\ ] ] where , using eqs .
( [ m1m],[e2m],[c2 m ] ) and evaluating some complicated traces , @xmath25 turns out to be @xmath26 with @xmath27 \left(m^2+m m_\delta + q_0 e_k-{\bf k}\cdot { \bf q}\right ) } \label{m1 } \\
\lefteqn{\tilde p^e(k , q)= \frac{1}{m_\delta^2(q^-)^2 } \left[{\bf k}^2 { \bf p}^2 - ( { \bf k } \cdot { \bf q})^2\right ] } \label{e2}\\ & & \times\bigl\ { - ( { \bf k } \cdot { \bf q})^3 + ( { \bf k } \cdot { \bf q})^2 \left[m^2-m m_\delta + 3 q_0 e_k\right ] \nonumber\\ & & + 2 { \bf k } \cdot { \bf q } q_0 m ( m_\delta - m ) e_k \nonumber\\ & & - { \bf k } \cdot { \bf q } \left[m_\delta^2 ( q^-)^2 + m^2 { \bf q}^2 + 2 q_0 ^ 2 m^2 + 3 q_0 ^ 2 { \bf k}^2 \right ] \nonumber\\ & & + q_0\left[m_\delta^2 ( q^-)^2 + m^2 { \bf q}^2+q_0 ^ 2 { \bf k}^2 \right]e_k \nonumber\\ & & + m\left[m_\delta^2 ( q^-)^2(m+m_\delta)+(m - m_\delta ) \left(m^2 { \bf q}^2+q_0 ^ 2 { \bf k}^2\right)\right]\bigr\ } \nonumber\\ \lefteqn{\tilde p^c(k , q)= \frac{4 } { m_\delta^4 ( q^-)^2 } \left[q^2 m_\delta^2-\left(q^2+q_0 e_k-{\bf k}\cdot { \bf q}\right)^2\right ] } \label{c2}\\ \nonumber\\ & & \times \left[q_0 e_k-{\bf k}\cdot { \bf q}+m(m - m_\delta)\right ] \left[q^2e_k-(q_0 e_k-{\bf k}\cdot { \bf q})q_0\right]^2 \nonumber\end{aligned}\ ] ] even if eq.([eq:2 ] ) appears to be divergent , it actually becomes finite when its value at @xmath28 is subtracted @xcite .
the removal of the divergences concerns in any case the real part only of the polarisation propagator .
the longitudinal response at the end of the story ( separating the three contribution m1 , e2 and c2 ) is obtaining by taking the imaginary part of the @xmath0-hole term of @xmath23 and is the sum of the three terms , namely @xmath29 ( each of them , of course , multiplied by its corresponding form factor ) where @xmath30 .
the m1 coupling constant has been chosen according to @xcite , i.e. @xmath31 , we have left instead the c2 coupling constant free to range between 0 . and 0.4 ( since no well established experimental data give serious indication about this value , we ha chosen a rather high one : actually , as we shall see , it does not significantly change the results , so that its uncertainty does not really matter in the present case ) . the free response functions for @xmath32 , evaluated at @xmath33 are displayed in fig . 1 .
here the quasi - elastic peak ( qep ) is treated relativistically and without dynamics ( i.e. , in rffg approximation ) .
the @xmath0 peak is evaluated according to the above equations .
[ fig1 ] since the e.m .
transition form factors of the @xmath0 are not so well determined as those of the nucleon @xcite , we have simply used the standard nucleon dipole form factor for them . in the figure
the expected contribution appears as a small bump on the right of the qep .
the m1 contribution here is by far dominant , the c2 term , even at its maximum coupling constant , being hardly distinguishable in the figure .
we now come to deal with the @xmath0-excitations through many - body channels . in the last two years , after a long time of unsatisfactory efforts , some calculations have been able to reproduce fairly well both the transverse and longitudinal responses in the qep region ( see @xcite among others ) . here
we adopt the approach developed by two of us , that naturally embeds the @xmath0-degree of freedom into the formalism .
the main theoretical tool in @xcite is the bosonisation , which amounts to transform the original action containing nucleons and @xmath0 s as true degrees of freedom interacting via a potential into an effective bosonic one ( i.e. containing only bosonic degrees of freedom actually auxiliary fields introduced by means of a hubbard - stratonovitch transformation ) fully equivalent to the former .
then the natural approximation scheme over the bosonic action is the boson loop expansion . as described in @xcite , the semi - classical limit ( mean field )
amounts to collect all diagrams having no boson loops and is easily recognised to be the familiar ( not antisymmetrized ) rpa .
the next - to - leading order collects all the diagrams with one boson loop : since fermion loops are of no matter in determining the order of the ble , then each boson line has to be thought as an rpa - dressed potential . in this way
a boson line embodies an infinity of p - h and @xmath0-h insertions .
thus not only a remarkable part of 2p-2h , but also some 3p-3h and higher are included .
however , not all the 2p-2h contributions are accounted for : for instance , of the diagrams listed in ref .
@xcite the ivb ) and vb ) of fig .
9 are absent .
more generally all diagrams with two ( rpa - dressed ) interaction lines and only one fermion loop are lacking . instead , our approach warrants the preservation of gauge invariance and of general theorems and sum rules . concerning a comparison with other approaches , the only possible
one ( being expressed in terms of feynman diagrams too ) is with ref .
@xcite . in our case both
higher order corrections to @xmath0 self - energy , pion - in - flight and kroll - rudermann term are absent ( and this because the complications and the large amount of work they require to be embedded in the formalism ) , while instead the possibility of having @xmath0-excitations in different fermionic loops ( in practice the so - called box diagrams of the bonn potentials @xcite ) are accounted for , that not only play a relevant role in the qep , as is well known , but are also fundamental in providing the response in the @xmath0-h region .
note that the absence of part of the mec causes an underestimate of the response in the dip region , as will be apparent later in fig .
4 . in any case at present no separate calculations for the longitudinal and transverse responses in the @xmath0 region have been provided by oset and coworkers . still a comparison between our approach and the one of ref .
@xcite seems to be beyond our present possibilities . of central interest to us
is that the one loop order diagrams can contain @xmath0 s even if the incoming and outgoing virtual photons interact only with nucleons .
this originates a resonant structure , broadened by a further p - h pair contemporarily present .
typical diagrams are displayed in fig .
2 . there
it is shown that a @xmath0-h can be excited , with vector - isovector character , provided that another p - h pair again vector - isovector in nature re - establishes the correct quantum numbers .
remarkably the whole class of diagrams considered in @xcite and schematized in fig .
3 contains at least one @xmath0-h pair , since they surely appear in the rpa - dressed potential .
there the dashed lines are the rpa - dressed , and consequently also contain @xmath0-h pairs , while the solid ones can be either a nucleon or a @xmath0 compatibly with the conservation laws .
the vertices are assumed to be pure nucleon - nucleon or @xmath0-@xmath0 currents since this part of calculation is intrinsically non - relativistic .
the results of our calculations extend some partial outcomes hidden in @xcite : there the existence of a tail on the right of qep was partially displayed but not discussed in one of the many figures , and , mainly , its resonant structure was not exhibited . here
the whole @xmath0 peak is displayed and its resonant nature can be recognised
an outcome not warranted _ a priori _ and the balance with the direct m1 excitation established .
furthermore the crucial physical point is that the effect discovered in @xcite is expected to rise with the transferred momentum , as the quoted paper shows , while the many - body corrections are expected to decrease .
the calculations are carried out with the same parameters given in @xcite .
the results are shown in fig .
since the relativistic kinematics can be of great relevance , because it tends to shift both the qep and the @xmath0 one to lower energies , i.e. , at greater distances from the light cone , the electro - magnetic form factor will also decrease somehow the height of the peaks .
as repeatedly said throughout the paper we are not able at present to perform a full relativistic calculation of the one - loop ble : the best thing we can do is to shrink somehow the @xmath34-axis in fig . 4 in such a way to account for the relativistic kinematics , if not for the relativistic dynamics .
what we have done in practice is a rescaling of the @xmath34-axis in such a way that the maximum of the qep , the maximum of the @xmath0 peak and the right corner of the free qep will exactly coincide in energy with the position required by the relativistic kinematics .
in particular it is extremely important to fix the right corner of the free qep because this strongly reduces the oscillations that are still observable in that point and that are due to the failure of the ble nearby a singular point .
this procedure makes our pseudo - relativistic results in the left side of the qep rather unreliable , but on the other hand we are more interested , in the present paper , to the dip and @xmath0-peak region .
let us now comment our main results , namely those displayed in fig .
the solid line denotes a pseudo - relativistic calculation : this means that the mean field is exactly evaluated in rffg approximation both in the qep and in the @xmath0-peak and that the one - loop correction are evaluated non - relativistically but the energy scale is shrinked as described above .
the dashed line instead presents a fully non - relativistic calculation . in this case coherently no free @xmath0-peak is accounted for and the corresponding bump on the right of the qep comes solely from the one - loop corrections ; finally the dotted lines describe the pure rffg model .
as we expected from the very beginning the qep is well reproduced by the non - relativistic calculations , because the parameters of the model ( in practice the effective interactions in the various possible channels ) where adjusted on a non - relativistic dynamics . since relativity reduces the free qep the solid line underestimates it . on the other hand in the @xmath0-region the incoherence between the relativistic kinematics and the non - relativistic treatment of the one - loop corrections
is the price we must pay to give a numerically acceptable description of the response .
looking to this region , we see that the many - body excitations are clearly displayed and in general larger than the pure m1 response ( again a result not warranted _ a priori _ since both effects were expected to be small ) .
another relevant question concerns the behaviour of the many - body correlations at high momentum transfer .
for this reason we have attempted to perform a calculation also for a transferred momentum of 600 mev / c , where we know that the non - relativistic approximation is at least questionable . disregarding in any case the qep we see ( if in the @xmath0-region our approach
still can provide meaningful results ) that the fading out of the many - body correction is not so strong as naively expected , and it opens the serious questions about how fast the nuclear correlations will disappear .
it suggests indeed that nucleon - nucleon correlation will die much before than the @xmath0-nucleon one .
furthermore , the dashed line ( that does not contain the free @xmath0-peak ) shows a clear resonant behaviour at a transferred momentum @xmath35= 400 mev / c , while the resonant structure seems to fully disappear at @xmath35= 600 mev / c , where instead the dip region is significantly filled . since these conclusions depend in a crucial way upon the behaviour of the response at high values of @xmath35 , where our calculation suffer for the lack of relativity , then the present outcome clearly asks for fully relativistic many - body calculations , a task that is at this moment forbidden to us due to some technicalities discussed in @xcite .
another relevant point is that even neglecting correlations some strength is present in the resonant region and can give a significant ( possibly of the order of some percent ) contribution to the coulomb sum rule .
this will require qualitative changes in the derivations of the former to account for the @xmath0 degrees of freedom as well .
of course no serious comparison with the experimental data can be performed until a higher precision will be reached by the experimental data , such to enable the rosenbluth separation even in a region where the longitudinal response is naturally depressed with respect of the transverse one .
nevertheless the indications coming from our calculation are surely not in contrast with the experimental data . | the presence of a resonant structure corresponding to the @xmath0 excitation in the longitudinal response to an e.m .
probe is investigated .
it is shown that many - body effects could significantly increase the relativistic contribution suggested by m.ericson and coworkers |
bl lacertae objects ( bl lacs ) are an extreme form of active galactic nuclei ( agn ) which exhibit rapid flux variability at all frequency , high polarization , and weak or non - existent spectral features . along with their cousins , the flat - spectrum radio quasars ,
they form the blazar class of agn . the model which best describes blazars is one in which a jet of relativistic material is beamed directly at us @xcite . although among the rarest types of agn , bl lacs are particularly important precisely because we can look into their jets . in this way
, we can see the site of energy production which is presumably close to the central engine , or black hole .
thus , by studying these enigmatic objects we can better understand the central black hole in all other agn . one way of studying these objects , which is independent of any assumptions made about the nuclear regions and energy production , is by their environments . there is growing consensus that bl lacs are found on average in poor clusters of galaxies @xcite .
however , some individual objects ( e.g. pks 0548322 ) are found in rich clusters @xcite .
work has progressed on the kilo - parsec - scale environments , or host galaxies of bl lacs as well .
bl lacs are found in giant elliptical galaxies with @xmath2 mag @xcite .
some individual objects were thought to be hosted by spiral galaxies , but high resolution observations have shown them to be ellipticals @xcite .
high resolution images of the host galaxies of six radio - selected bl lac objects observed with the _ hubble space telescope ( hst ) _ have been presented in and .
these are relatively deep wfpc2 images from cycle 5 . in addition
, we are currently analyzing the host galaxies and extended environments of more than 100 bl lac objects at redshifts 0.03 - 1 observed by hst as part of a snapshot project ( pesce et al .
2000 , in preparation ; scarpa et al .
2000 in preparation ; urry et al .
2000 , in preparation ) .
the combined dataset will allow a thorough investigation of bl lac environmental properties over a large redshift range .
in this paper we present the analysis of the extended environments of six radio selected objects from our hst cycle 5 go project ( pi urry ) . to this sample we have added three x - ray selected objects obtained from the hst archive and observed during cycle 5 ( pi jannuzi ) .
we assume h@xmath3 km s@xmath4 mpc@xmath5 and q@xmath6 .
the objects were observed with the wfpc2 and the f814w ( i - band ) filter , the bl lac was centered in pc camera .
exposure times varied and were approximately 10 min to 1.5 hours ( see table 1 ) .
more details about the reduction and analysis of these fields can be found in .
we used the focas software , in iraf , to detect and classify all objects on an image and produce a catalog ( figure 1 ) .
for the central bl lac , the point spread function ( psf ) was created with tiny tim @xcite ; each object is classified as ` star ' or ` galaxy ' based on an automatic comparison of its profile to a standard psf profile . in these images , we detect objects down to a `` completeness limit '' of @xmath7 .
objects fainter than the completeness limit are not considered in the analysis of the images .
the automatic identifications from focas were checked with simulated images consisting of stars , galaxies , and noise .
focas detected and correctly classified all objects above the completeness limit and 90% of objects below the limit .
as a further check , the galaxy density on each image was compared to the average background density for the f814w filter from casertano ( 1997 , priv com ) .
and the pc is @xmath8280 kpc square , while the maximum complete radius sampled ( the straight lines and arc ) is @xmath80.7 mpc.,width=377 ] [ cols="^,^,>,^",options="header " , ] @xmath9 apparent magnitude of a typical galaxy , assuming @xmath10 .
no k - correction has been applied .
to avoid background contamination , we only count galaxies with @xmath11 where @xmath12 .
the pc images subtend 0.3 arcmin@xmath13 while the wf chips subtend 1.5 arcmin@xmath13 giving a maximum radial extent from our bl lacs of 0.4 - 0.8 mpc .
most bl lac clusters are found within 0.5 mpc of object @xcite . however , because of the wfpc2 geometry , we only observe @xmath81/2 of the volume surrounding the bl lac .
we checked our procedures by analyzing an archival image of the rich cluster abell 2390 ( richness class 1 , @xmath14 , exposure = 2100 s ) . on this image
, we detected a significant ( factor 9 - 20 ) overdensity of galaxies .
for our objects , we * detect enhancements ( factor 2 - 4 ) above background in four cases : * 1407 + 599 , 1538 + 149 , 1823 + 568 , 2143 + 070 .
a modest enhancement is found around 2254 + 074 , but this object has the lowest redshift so we are observing only a small part of the potential cluster volume . * no enhancements * are found for 0814 + 425 , 0828 + 493 , 1221 + 249 , 1308 + 326 . given our completeness limit , and assuming correct redshifts , we _ should _ detect any cluster present ( if the cluster is approximately symmetric ) . *
these objects appear to be truly isolated . *
we have analyzed the extended environments of nine bl lac objects observed with the hst wfpc2 in the f814w filter .
the long exposures provide deep ( @xmath15 mag ) images .
we find significant enhancement of galaxy density around four of the nine bl lac objects ( 1407 + 599 , 1538 + 149 , 1823 + 568 , 2143 + 070 ) . in some cases ( 1407 + 599 ,
1823 + 568 ) the enhancement is extreme , with dozens of galaxies within @xmath8150 kpc of the bl lac object .
the remaining five fields show no indication of excess galaxy counts .
in fact , some fields are slightly below the average background galaxy density , although in general the galaxy density is consistent with the background .
these objects are either truly isolated , their clusters are unusually asymmetric ( and strangely all out of the field of view ) , or the redshifts are incorrect . of these three explanations ,
the first is most likely .
thus , some bl lac objects appear to be in regions of significantly enhanced galaxy density , while others seem to be completely isolated .
this result is not surprising given that bl lac hosts appear to be otherwise normal giant elliptical galaxies .
such galaxies are found in all types of environments - from rich clusters to isolated regions of space . nonetheless if galaxy interactions play an important role in the agn activity ,
future work needs to address this point .
our sample is too small to determine differences between x - ray and radio selected objects , a point to be addressed in our analysis of the larger hst snapshot survey .
abraham , r. g. , mchardy , i. m. , & crawford , c. s. 1991 , mnras , 252 , 482 blandford , r. d. , & rees , m. j. 1978 , in `` pittsburgh conference on bl lac objects '' , ed .
a. n. wolfe ( pittsburgh : univ . of pittsburgh press ) , 328 falomo , r. 1996 , mnras , 283 , 241 falomo , r. , pesce , j. e. , & treves , a. 1993 , apj , 411 , l63 falomo , r. , pesce , j. e. , & treves , a. 1995 , apj , 438 , l9 falomo , r. , urry , c. m. , pesce , j. e. , scarpa , r. , treves , a. , & giavalisco , m. 1997 , apj , 476 , 113 hasan , h. , & burrows , c. j. 1995 , pasp , 107 , 289 pesce , j. e. , falomo , r. , & treves , a. 1994 , aj , 107 , 494 pesce , j. e. , falomo , r. , & treves , a. 1995 , aj , 110 , 1554 smith , e. p. , odea , c. p. , & baum , s. a. 1995 , apj , 441 , 113 stickel , m. , fried , j. w. , & khr , h. 1993 , a&as , 98 , 393 urry , c. m. , falomo , r. , scarpa , r. , pesce , j. e. , treves , a. , giavalisco , m. 1999 , apj , 512 , 88 wurtz , r. , stocke , j. t. , & yee , h. k. c. 1996 apjs , 103 , 109 wurtz , r. , stocke , j. t. , ellingson , e. , & yee , h. k. c. 1997 , apj , 480 , 547 | we analyze images of bl lacertae objects obtained with the _ hubble space telescope _ wfpc2 and the f814w filter .
the nine objects cover a redshift range of 0.19 to 0.997 .
the relatively deep images are sufficient to detect galaxies at least one magnitude below m@xmath0 ( 21.4 ) and in most cases to three magnitudes below m@xmath1 .
galaxy enhancement over the average background is found around four out of the nine objects .
results for some cases are confirmed by ground - based imaging . in the other cases ,
the redshifts of the target bl lac objects may be incorrect or they are truly isolated .
these findings reinforce the idea that on average , bl lac objects are found in regions of above average galaxy density
. however , isolated objects apparently can host bl lac nuclei too , a result that has implications for the processes that trigger / fuel the nuclear activity . , , , , , |
m , width=316 ] a major challenge in biology is to determine how proliferating cells in developing and adult tissues behave _ in vivo_. a powerful technique in solving this problem is clonal analysis , the labelling of a sample of cells within the tissue to enable their fate and that of their progeny to be tracked @xcite .
this approach gives access to information on proliferation , migration , differentiation ( into other cell types ) , and cell death ( apoptosis ) of the labelled cell population .
the most reliable method of labelling is through genetic modification leading to the expression of a reporter gene in a random sample of cells .
recently it has become possible to activate genetic labelling at a defined time in transgenic mice , enabling the kinetics of labelled cells to be studied with single - cell resolution _ in vivo _ @xcite . from a theoretical perspective , the analysis of clonal fate data presents a challenging `` inverse problem '' in population dynamics : while it is straightforward to predict the time - evolution of a population distribution according to a set of growth rules , the analysis of the inverse problem is more challenging , open to ambiguity and potential misinterpretation .
these principles are exemplified by the mechanism of murine epidermal homeostasis : mammalian epidermis is organised into hair follicles interspersed with _
interfollicular _ epidermis ( ife ) , which consists of layers of specialised cells known as keratinocytes @xcite ( see fig . [
fig : skincrosssection](a ) ) .
proliferating cells are confined to the basal epidermal layer .
as they differentiate into specialised skin cells , the basal cells withdraw from the cycle of cell proliferation and then leave the basal layer , migrating towards the epidermal surface from which they are ultimately shed . to maintain the integrity of the tissue , new cells must be generated to replace those lost through shedding . for many years , it has been thought that interfollicular epidermis is maintained by two distinct progenitor cell populations in the basal layer .
these comprise long - lived stem cells ( s ) with the capacity to self - renew , and their progeny , known as transit - amplifying cells ( ta ) , which go on to differentiate and exit the basal layer after several rounds of cell division @xcite .
stem cells are also found in the hair follicles , but whilst they have the potential to generate epidermis in circumstances such as wounding , they do not appear to contribute to maintaining normal epidermis @xcite .
the prevailing model of interfollicular homeostasis posits that the tissue is organised into regularly sized `` epidermal proliferative units '' or epus , in which a central stem cell supports a surrounding , clonal , population of transit amplifying cells , which in turn generate a column of overlying differentiated cells @xcite .
several experimental approaches have been used to attempt to demonstrate the existence of epus , but conclusive evidence for their existence is lacking .
the epu model predicts that slowly - cycling stem cells should be found in a patterned array in the ife ; cell labelling studies have failed to demonstrate such a pattern @xcite . in chimaeric mice
the epu model predicts that the boundaries of mosaicism in the ife should run along the boundaries of epus ; instead boundaries were found to be highly irregular @xcite .
genetic labelling studies using viral infection or mutation to activate expression of a reporter gene in epidermal cells have demonstrated the existence of long - lived , cohesive clusters of labelled cells in the epidermis , but these clusters do not conform to the predicted size distribution of the epu @xcite .
thus , until recently the means by which homeostasis of ife was achieved has been unclear .
however , by exploiting inducible genetic labelling , recent studies have allowed the fate of a representative sample of progenitor cells and their progeny to tracked _ in vivo _ @xcite .
as well as undermining the basis of the stem / ta cell hypothesis , the range of clone fate data provide the means to infer the true mechanism of epidermal homeostasis . in particular ,
these investigations indicate that the maintenance of ife in the adult system conforms to a remarkably simple birth - death process involving a single progenitor cell compartment .
expanding upon the preliminary theoretical findings of ref .
@xcite , the aim of this paper is to elucidate in full the evidence for , and the properties of , the model of epidermal maintenance , and to describe the potential of the system as a method to explore early signatures of carcinogenic mutations . to organise our discussion , we begin with an overview of the experimental arrangement , referring to ref .
@xcite for technical details of the experimental system . to generate data on
the fate of individual labelled cells and their progeny , hereafter referred to as clonal fate data , inducible genetic marking was used to label a sample of cells and their progeny in the epidermis of transgenic mice . the enhanced yellow fluorescent protein ( eyfp )
label was then detected by confocal microscopy , which enables 3d imaging of entire sheets of epidermis .
low - frequency labelling of approximately 1 in 600 basal - layer epidermal cells at a defined time was achieved by using two drugs to mediate a genetic event which resulted in expression of the eyfp gene in a cohort of mice .
this low efficiency labelling ensures that clones are unlikely to merge ( see discussion in section [ sec : model ] ) . by analysing samples of mice at different time
points it was possible to analyse the fate of labelled clones at single cell resolution _ in vivo _ for times up to one year post - labelling in the epidermis ( see , for example , fig .
[ fig : skincrosssection](b ) ) @xcite .
weeks post - labelling of ( a ) a detached clone in which all cells have undergone a transition to terminal differentiation by week @xmath0 , and ( b ) a persisting clone in which some of the cells maintain a proliferative capacity , according to model ( [ modelratelaws ] ) .
circles indicate progenitor cells ( p ) , differentiated cells ( d ) , and suprabasal cells ( sb ) .
note that , because the birth - death process ( [ modelratelaws ] ) is markovian , the lifetime of cells is drawn from a poisson distribution with no strict minimum or maximum lifetime . the statistics of such lineage trees do not change significantly when we account for a latency period between divisions that is much shorter than the mean cell lifetime ( see discussion in section [ sec : stochastic ] ) .
_ bottom : _ the total number of proliferating , differentiated and supra - basal cells for the two clones as a function of time.,width=316 ] with the gradual accummulation of eyfp levels , the early time data ( less than two weeks ) reveals a small increase in the number of labelled clones containing one or two cells . at longer times , clones increase in size while cells within clones begin to migrate through the suprabasal layers forming relatively cohesive irregular columns ( see fig .
[ fig : skincrosssection](a ) ) . the loss of nuclei in the cornified layer ( fig .
[ fig : skincrosssection ] ) makes determination of the number of cornified layer cells in larger clones by microscopy unreliable .
therefore , to identify a manageable population , attention was focused on the population of basal cells in `` persisting clones '' , defined as those labelled clones which retain at least one basal layer cell , such as is exemplified in the theoretical lineage maps in fig .
[ fig : tree ] .
after two weeks , the density of persisting clones was seen to decrease monotonically indicating that the entire cell population within such clones had become differentiated and the clone detached from the basal layer ( shown schematically in figs .
[ fig : skincrosssection](a ) and [ fig : tree](a ) ) .
however , the population of persisting clones showed a steady increase in size throughout the entire duration of the experiment .
weeks ) and at long times ( beyond @xmath1 weeks ) follows the two simple analytical approximations described in the main text ( lower and upper dashed curves ) . for times earlier than two weeks ( referring to section [ sec : eyfp ] ) , clones remain approximately one cell in size . the experimental data are consistent with the behaviour predicted by process ( [ modelratelaws ] ) ( black line ) when it is assumed that only a - type cells are labelled at induction .
in contrast , assuming that a and b type cells label in proportion to their steady - state population leads to an underestimate of average clone size between two and six weeks ( lower curve , red online ) , as does the assumption that type b cells label with better efficiency ( not shown ) . _ inset : _ the underlying distribution of basal cells per clone at 2 weeks and 26 weeks post - labelling .
the data is binned by cell count in increasing powers of @xmath2.,width=316 ] to
what extent are the clone fate data consistent with the orthodox stem / ta cell model of epidermal maintenance ? referring to fig .
[ fig : avgsize ] , one observes an inexorable increase in the average size of an ever - diminishing persisting clone population . this result is incompatible with any model in which the ife is supported by a population of long - lived stem cells . with the latter , one would expect the number density of persisting clones to reach a non - zero minimum ( commensurate with the labelling frequency of stem cells ) while the average clone size would asymptote to a constant value characteristic of a single epidermal proliferative unit .
we are therefore lead to abandon , or at least substantially revise , the orthodox stem / ta cell hypothesis and look for a different paradigm for epidermal maintenance . but , to what extent are the clone fate data amenable to theoretical analysis
? indeed , the application of population dynamics to the problem of cell kinetics has a long history ( see , e.g. , refs . ) with studies of epidermal cell proliferation addressed in several papers @xcite .
however , even in the adult system , where cell kinetics may be expected to conform to a `` steady - state '' behaviour , it is far from clear whether the cell dynamics can be modelled as a simple stochastic process .
regulation due to environmental conditions could lead to a highly nonlinear or even non - local dependence of cell division rates .
indeed , _ a priori _ , it is far from clear whether the cell kinetics can be considered as markovian , i.e. that cell division is both random and independent of the past history of the cell . therefore , instead of trying to formulate a complex theory of cell division , taking account of the potential underlying biochemical pathways and regulation networks @xcite
, we will follow a different strategy looking for signatures of steady - state behaviour in the experimental data and evidence for a simple underlying mechanism for cell fate .
intriguingly , such evidence is to be found in the scaling properties of the clone size distribution @xcite . to identify scaling characteristics ,
it is necessary to focus on the basal layer clone size distribution , @xmath3 , which describes the probability that a labelled progenitor cell develops into a clone with a total of @xmath4 basal layer cells at a time @xmath5 after labelling .
( note that , in general , the total number of cells in the supra - basal layers of a clone may greatly exceed the number of basal layer cells . ) with this definition , @xmath6 describes the `` extinction '' probability of a clone , i.e. the probability that _ all _ of the cells within a labelled clone have migrated into the supra - basal layers . to make contact with the experimental data ,
it is necessary to eliminate from the statistical ensemble the extinct clone population ( which are difficult to monitor experimentally ) and single - cell clones ( whose contribution to the total ensemble is compromised by the seemingly unknown relative labelling efficiency of proliferating and post - mitotic cells at induction ) , leading to a reduced distribution for persisting " clones , @xmath7 then , to consolidate the data and minimise fluctuations due to counting statistics , it is further convenient to _ bin _ the distribution in increasing powers of 2 , @xmath8 i.e. @xmath9 describes the probability of having two cells per clone , @xmath10 describes the probability of having 3 - 4 cells per clone , and so on . referring to fig .
[ fig : scaling ] , one may see that , after an initial transient behaviour , the clone size distribution asymptotes in time to the simple scaling form , @xmath11 this striking observation brings with it a number of important consequences : as well as reinforcing the inapplicability of the stem cell / ta cell hypothesis , such behaviour suggests that epidermal maintenance must conform to a simple model of cell division .
the absence of further characteristic time - scales , beyond that of an overall proliferation rate , motivates the consideration of a simple kinetics in which _ only one process dictates the long - time characteristics of clonal evolution_. moreover , from the scaling observation one can also deduce two additional constraints : firstly , in the long - time limit , the average number of basal layer cells within a _ persisting _ clone _ increases linearly with time _ , viz .
@xmath12 secondly , if we assume that labelled progenitor cells are representative of _ all _ progenitor cells in the epidermis , and that the population of clones with only one basal layer cell is not `` extensive '' ( i.e. @xmath13 ) , this means that , in the long - time limit , the clone persistence probability must scale as @xmath14 such that @xmath15 where the constant , @xmath16 , is given by the fraction of proliferating cells in the basal layer . without this condition ,
one is lead to conclude that the labelled population of basal layer cells either grows or diminishes , a behaviour incompatible with the ( observed ) steady - state character of the adult system .
, plotted as a function of the rescaled time coordinate @xmath17 .
the data points show measurements ( extracted from data such as shown in fig .
[ fig : avgsize](inset ) , given fully in ref .
@xcite ) , while the solid curves show the probability distributions associated with the non - equilibrium process ( [ modelratelaws ] ) for the basal - layer clone population as obtained by a numerical solution of the master equation ( [ pmastereqn ] ) .
( error bars refer to standard error of the mean ) . at long times ,
the data converge onto a universal curve ( dashed line ) , which one may identify with the form given in eq .
[ pk_exact_all_cells ] .
the rescaling compresses the time axis for larger clones , so that the large - clone distributions appear to converge much earlier onto the universal curve.,width=316 ] although the manifestation of scaling behaviour in the clone size distributions gives some confidence that the mechanism of cell fate in ife conforms to a simple non - equilibrium process , it is nevertheless possible to conceive of complicated , multi - component , models which could asymptote to the same long - time evolution . to further constrain the possible theories ,
it is helpful to draw on additional experimental observations @xcite : firstly , immunostaining of clones with a total of two cells ( using the proliferation marker ki67 and , separately , the replication licensing factor cdc6 ) reveals that a single cell division may generate either one proliferating and one non - proliferating daughter through asymmetric division , or two proliferating daughters , or two non - proliferating daughters ( cf . ) .
secondly , three - dimensional imaging of the epidermis reveals that only @xmath18 of mitotic spindles lie perpendicular to the basal layer indicating that divisions may be considered to be confined to the basal layer , confirming the results of earlier work that indicates a dividing basal cell generates two basal layer cells @xcite .
this completes our preliminary discussion of the experimental background and phenomenology . in summary ,
the clone fate data reveal a behaviour wholely incompatible with any model based on the concept of long - lived self - renewing stem cells .
the observation of long - time scaling behaviour motivates the consideration of a simple model based on a stochastic non - equilibrium process and is indicative of the labelled cells being both a representative ( i.e. self - sustaining ) population and in steady - state . in the following
, we will develop a theory of epidermal maintenance which encompasses all of these observations .
taken together , the range of clonal fate data and the observation of symmetric and asymmetric division are consistent with a remarkably simple model of epidermal homeostasis involving only one proliferating cell compartment and engaging just three adjustable parameters : the overall cell division rate , @xmath19 ; the proportion of cell divisions that are symmetric , @xmath20 ; and the rate of transfer , @xmath21 , of non - proliferating cells from the basal to the supra - basal layers . to maintain the total proliferating cell population , a constraint imposed by the steady - state assumption , we have used the fact that the division rates associated with the two channels of symmetric cell division must be equal . denoting the proliferating cells as type a , differentiated basal layer cells as type b , and supra - basal layer cells as type c , the model describes the non - equilibrium process , @xmath22 finally , the experimental observation that the total basal layer cell density remains approximately constant over the time course of the experiment leads to the additional constraint that @xmath23 reducing the number of adjustable parameters to just two . by ignoring processes involving the shedding of cells from the surface of the epidermis ,
the applicability of the model to the consideration of the _ total _ clone size distribution is limited to appropriately short time scales ( up to six weeks post - labelling ) . however ,
if we focus only on the clone size distribution associated with those cells which occupy the basal layer , the model can be applied up to arbitrary times . in this case , the transfer process must be replaced by one in which @xmath24 . in either case , if we treat all instances of cell division and cell transfer as independent stochastic events , a point that we shall revisit later , then the time evolution associated with the process ( [ modelratelaws ] ) can be cast in the form of a master equation .
defining @xmath25 as the probability of finding @xmath26 type a cells and @xmath27 type b cells in a given clone after some time @xmath5 , the probability distribution evolves according to the master equation : @xmath28\nonumber\\ & & \qquad\qquad + r\lambda[(n_{\rm a}+1)p_{n_{\rm a}+1,n_{\rm b}-2}-n_{\rm a } p_{n_{\rm a},n_{\rm b}}]\nonumber\\ & & \qquad\qquad + ( 1 - 2r)\lambda[n_{\rm a } p_{n_{\rm a},n_{\rm b}-1}-n_{\rm a } p_{n_{\rm a},n_{\rm b}}]\nonumber\\ & & \qquad\qquad + \gamma[(n_{\rm b}+1)p_{n_{\rm a},n_{\rm b}+1}-n_{\rm b } p_{n_{\rm a},n_{\rm b}}]\,.\end{aligned}\ ] ] if we suppose that the basal layer cells label in proportion to their population , the latter must be solved subject to the boundary condition @xmath29 . later , in section
[ sec : eyfp ] , we will argue that the clone size distribution is compatible with a labelling efficiency which favours a over b type cells . either way , by excluding single cell clones from the distribution , this source of ambiguity may be safely eliminated .
although the master equation ( and its total cell number generalisation ) is not amenable to exact analytic solution , its properties can be inferred from the consideration of the a cell population alone for which an explicit solution may be derived . when considered alone ,
a type cells conform to a simple set of rate laws , @xmath30 an example of a galton - watson process , long known to statisticians ( see , e.g. , ref .
@xcite ) . in this case
, the probability distribution , which is related to that of the two - component model through the relation @xmath31 , can be solved analytically .
( here , we have used a lower case @xmath32 to discriminate the probability distribution from its two - component counterpart . ) for an initial distribution @xmath33 it may be shown that @xcite , @xmath34 from this system and its associated dynamics , one can draw several key implications : starting with a single labelled cell , the galton - watson process predicts that the persistance probability of the resulting clone ( i.e. , in this case , the probability that the clone retains at least one proliferating cell ) , is given by @xmath35 i.e. as with the experiment , the persistance probability of a clone decays monotonically , asymptoting to the form @xmath36 at time scales @xmath37 , the time scale for symmetric division .
applied to the experimental system , this suggests that labelled clones continue to detach from the basal layer indefinitely . at the same time , defining @xmath38 as the size distribution of _ persisting _ clones , the mean number of basal layer cells in a persisting clone grows steadily as @xmath39 such that the overall cell population remains constant , viz .
@xmath40 , i.e. the continual extinction of clones is compensated by the steady growth of persisting clones such that the average number of proliferating cells remains constant : given enough time , all cells would derive from the same common ancestor , the hallmark of the galton - watson process .
this linear increase in clone size may lead one to worry about neighbouring clones coalescing .
fortunately , the continual extinction of clones ensures that the fraction of clones conjoined with their neighbours remains small and of same order as the initial labelling density . ] .
the fact that this fraction is constant is again indicative of the steady - state condition maintained throughout the experiment .
if , at some instant , a clone is seen to have , say , @xmath41 proliferating cells then , after a further time @xmath5 , its size will fluctuate as @xmath42 thus clones ( as defined by the a cell population ) will maintain an approximately stable number of cells providing @xmath43 . for larger clones
this time may exceed the lifetime of the system . at the limit where macroscopic sections of the basal layer are considered ,
the statistical fluctuations are small .
the increased stability of larger clones also explains the surprising prediction that , given enough time , all clones eventually become extinct ( viz .
@xmath44 ) .
calculated explicity , the extinction probability for a clone of size @xmath45 scales as @xmath46 @xcite approaching unity at long times .
however , because this extinction probability is small when @xmath47 , a large enough clone may easily persist beyond the lifetime of the system . at asymptotically long times
, one may show ) , we treat @xmath48 as continuous variables in eq .
( [ pmastereqn ] ) ( a good approximation at large values of @xmath4 ) . then , making the ansatz that the b - cell population remains slave to the a - cell population viz .
@xmath49 , the master equation simplifies to the approximate form @xmath50 which is solved by @xmath51 , leading to eq .
[ p_exact_all_cells ] . ] that the full probability distribution for finding @xmath52 cells within a persisting clone scales in proportion to @xmath53 , viz .
@xmath54\ , , \ ] ] and so @xmath55 - \exp\left[-2^k\frac{\rho}{r\lambda t}\right]\ , , \ ] ] i.e. the probability distribution acquires the scaling form found empirically . referring to eq .
( [ scaling ] ) , we can therefore deduce the form of the scaling function , @xmath56-\exp[-\rho x / r\lambda]\,.\ ] ] as a result , at long times , the average basal layer population of persisting clones becomes proportional to the average number of proliferating cells per clone , @xmath57 , a behaviour consistent with that seen in experiment ( see fig . [
fig : avgsize ] ) . in fitting the model to the data ( see below )
, we will find that the rates @xmath19 and @xmath21 at which differentiated cells are created and then transferred into the super - basal region are significantly larger than the rate of symmetric division @xmath58 , which dictates the long - time behaviour of the clone size distribution . in this case , at early times ( @xmath59 ) , the clone size distributions are dominated by the differentiation and transfer rates , which remain prominent until the population of labelled differentiated cells associated with each proliferating cell reaches its steady - state value of @xmath60 .
one may therefore infer that , at short times , the mean number of basal layer cells in clones arising from proliferating cells is given by @xmath61 and that the early - time clone size distribution is poisson - distributed , viz .
@xmath62 with these insights it is now possible to attempt a fit of the model to the data . referring to fig .
[ fig : fit2data ] , one may infer the rate of cell division @xmath19 from the short - time data , and the symmetric division rate @xmath58 from the long - time scaling data .
in particular , taking the fraction of proliferating cells in the basal layer to be @xmath63 , a figure obtained experimentally by immunostaining using ki67 @xcite , a fit of eq .
( [ eqn : shorttimep ] ) to the short - time data ( fig .
[ fig : fit2data](a ) ) is consistent with a transfer rate of @xmath64/week which , in turn , implies a rate of cell division of @xmath65/week .
furthermore , by plotting the long - time , large-@xmath66 , size - distributions in terms of the `` inverse '' to the scaling function , @xmath67 \right)^{-1}\\ & = & \left ( 2\ln\left[(1-(1-\mathcal{p}^{\rm pers.}_k(t))^{1/2})/2\right ] \right)^{-1},\end{aligned}\ ] ] the data converge onto a linear plot ( fig .
[ fig : fit2data](b ) ) .
the resulting slope takes the value @xmath68 , from which we may infer the symmetric division rate @xmath69/week , and @xmath70 .
these figures compare well with an optimal fit of the _ entire _ basal layer clone size distribution ( fig .
[ fig : scaling ] ) , obtained by numerically integrating the master equation ( [ pmastereqn ] ) .
the fitting procedure is shown in fig .
[ fig : fit2data](c ) ( solid curves ) , where the likelihood of the model is evaluated for a range of values of @xmath19 and @xmath58 , as assessed from a @xmath71 test of the model solution @xcite .
one may see that the likelihood is maximised with an overall division rate of @xmath65/week and a symmetric division rate in the range @xmath72/week , thus confirming the validity of the asymptotic fits . moreover , the corresponding fit of both the basal layer distribution and the _ total _ clone size distribution , including both basal and supra - basal cells , is equally favourable ( fig .
[ fig : fit2data](c ) , dashed ) .
thus , in the following sections we shall use the asymptotically fitted value of @xmath73 , however any choice of the parameter in the range @xmath74 gives similar results .
although the comparison of the experimental data with the model leaves little doubt in its validity , it is important to question how discerning is the fit . by itself ,
the observed increase in the size of persisting clones is sufficient to rule out any model based on long - lived self - renewing stem cells , the basis of the orthodox epu model .
however , could one construct a more complicated model , which would still yield a similar fit ?
certainly , providing the long - time evolution is controlled by a single rate - determining process , the incorporation of further short - lived proliferating cell compartments ( viz .
transit - amplifying cells ) would not affect the observed long - time scaling behaviour .
however , it seems unlikely that such generalisations would provide an equally good fit to the short - time data .
more importantly , it is crucial to emphasize that the current experimental arrangement would be insensitive to the presence of a small , quiescent , long - lived stem cell population .
yet , such a population could play a crucial role in _
non_-steady state dynamics such as that associated with wound healing or development .
we are therefore led to conclude that the range of clone fate data for normal adult ife are consistent with a simple ( indeed , the simplest ) non - equilibrium process involving just a single progenitor cell compartment . at this stage , it is useful to reflect upon the sensitivity of the model to the stochasticity assumption applied to the process of cell division
. clearly , the scaling behaviour ( eq . [ p_exact_all_cells ] ) depends critically on the statistical independence of successive cell divisions ; each cell division results in symmetric / asymmetric cell fate with relative probabilities as detailed in ( [ modelratelaws2 ] ) .
but , to what extent would the findings above be compromised if the cell _ cycle - time _
, i.e. the time between consecutive cell divisions , were not determined by an independent stochastic process ?
this question may have important ramifications , because the assumption of independent cell division , used in formulating the master equation ( [ pmastereqn ] ) , introduces a manifestly unphysical behaviour by allowing cells to have arbitrarily short cycle times .
moreover , although a wide distribution of cell cycle - times has been observed for human keratinocytes _ in vitro _ @xcite , it is possible that that keratinocytes _ in vivo _ may divide in _ synchrony _ , giving a cell cycle - time distribution narrowly centered about the mean ( @xmath75 ) . in the following , we shall address both of these points : firstly , we shall show that , up to some potential latency period ( the time delay before a newly - divided cell is able to divide again ) , consecutive cell divisions occur independently as an asychronous , poisson process . secondly , while the data is insufficient to detect a latency period of @xmath0 hours or less between consecutive cell divisions , the data does discriminates against a period lasting longer than @xmath76 hours . to investigate the degree to which the model is sensitive to the particular cell cycle - time distribution , let us revisit the original model of independent cell division with several variations : firstly , we introduce a latency period of @xmath77 immediately following cell division , in which daughter cells can not divide .
this biologically - motivated constraint renders a more complicated yet more realistic model of cell division than the idealised system studied in the previous section .
motivated by observations of the minimal cycle - time of ( human ) keratinocytes @xcite , where a latency period of @xmath78 hours was observed _ in vitro _ , we shall here consider the a range latency periods of up to @xmath79 hours .
secondly , we compare the empirical clone size distributions with a model where all progenitor cells have a cycle - time of exactly @xmath75 , i.e. where cells within each clone divide in perfect synchrony .
finally , we shall investigate a range of intermediate models with different distributions of progenitor cell cycle - time ( see fig .
[ fig : stochastic](a ) ) .
technically , the resulting clone size distributions may be evaluated through monte carlo simulations of the non - equilibrium process ( [ modelratelaws ] ) with the cycle - time @xmath80 of each proliferating cell selected at random from a gamma distribution of the form @xmath81 where @xmath82 is the average time to division following the initial latency period @xmath83 , and @xmath84 is the `` shape parameter '' of the gamma distribution . in particular
, the choice of shape parameter @xmath85 corresponds to the exponential distribution which characterises the independent cell cycle - time distribution , whereas @xmath86 describes the case in which all a - cells have an exact cycle - time of @xmath75 ( see fig .
[ fig : stochastic](a ) ) .
then , to reflect the assumption that initially - labelled , spatially separated , progenitor cells have uncorrelated cell cycles , the time to the initial division event post - labelling is adjusted by a random time @xmath87 $ ] . finally , for an unbiased comparison of the models
, we optimise the value of @xmath19 for each model separately against the empirical data , whilst keeping @xmath88 to ensure an optimal fit of the long - time data , as discussed below . the resulting clone size distributions are shown in fig .
[ fig : stochastic](b ) , where the case of independent division following a 12-hour latency ( @xmath85 ) and the exact cycle - time case ( @xmath89 ) are compared to the empirical _ total _ clone size distribution , which includes both basal and supra - basal ( type c ) cells , over the first 6 weeks post - labelling .
two intermediate cases are also shown for comparison ( @xmath90 ) .
focusing first on the results for the case @xmath85 , which bears closest resemblance to the markovian model analysed using the master equation ( [ pmastereqn ] ) , one may see by inspection that the quality of the fit to the data remains good even when the effects of a latency period between cell divisions is taken into account .
more rigorously , a likelihood analysis reveals that the two cases are statistically indistinguishable ( see fig .
[ fig : fit2data](c ) , inset ) , which indicates that the duration of a latency period of @xmath91 hours is beyond the current empirical resolution .
however , referring to fig .
[ fig : fit2data]c ( inset ) , a similar analysis of longer latency periods reveals that for periods of @xmath92 hours , the fit to the data is significantly poorer .
turning next to the predicted basal - layer clone size distributions at late times ( @xmath93 ) ( not shown ) , one may see that all of the proposed distributions asymptotically converge : starting with exactly one cell , then the moment - generating function @xmath94 associated with the a cell population distribution @xmath95 after @xmath96 cell cycles satisfies the recursion relation @xcite : @xmath97 which asymptotes to the continuous master equation @xmath98 with the relative magnitude of the leading - order correction dropping off as @xmath99 . but
with @xmath100 , this equation is simply the master equation for the moment - generating function associated with the original model , eq .
[ p_exact_acells ] , and so the two models converge . one may therefore conclude that , beyond the first several weeks of the experiment ( @xmath101 ) , the fit to the data is sensitive only to the average cycle time of progenitor cells . with this in mind , we note that for the case of perfectly synchronous cell division , an optimal ( albeit poor ) numerical fit was obtained when @xmath102/week , a figure that compares well with the fit for the independent case .
it appears therefore that _ the predicted average cell division rate ( @xmath19 ) is insensitive to the shape of the cell cycle distribution_. finally , let us turn to the early time behaviour ( @xmath103 ) , where the predicted distributions are distinct . referring to fig .
[ fig : stochastic](b ) , one may see , at 2 - 4 weeks post - labelling , that relatively large clones ( @xmath104 cells ) appear earlier than expected by a model assuming synchronous division , and that , compared with the same model , a sizeable proportion of small clones ( e.g. , @xmath2 cells ) lingers on for far longer than expected . the same behaviour is observed for the basal layer clone size distribution ( not shown ) .
one may therefore infer that cell division conforms to a model of _ independent _ rather than _ synchronous _ division , allowing for some progenitor cells to divide unusually early , and for others to remain quiescent for an unusually long period of time . in summary
, we have established that , following division , progenitor cells do not divide for a period that is likely to last up to @xmath0 hours , and not more than @xmath76 hours .
after this latency period , the data is consistent with cells switching to a mode of independent , asynchronous , cell division .
these results shed light on why the simple model of independent cell division presented in section [ sec : model ] succeeds in producing such a remarkable fit to the data .
although the integrity of the fit of the model to the data provides some confidence in its applicability to the experimental system , its viability as a model of epidermal homeostasis rests on the labelled clone population being representative of all cells in the ife .
already , we have seen that the model , and by inference , the labelled clone population , has the capacity to self - renew .
however , the slow accumulation of eyfp after induction , together with the question of the relative labelling efficiency of the two basal layer cell types , leaves open the question of the very short - time behaviour . accepting the validity of the model , we are now in a position to address this regime . in doing so , it is particularly useful to refer to the time evolution of clone size as measured by the average number of basal cells in a persisting clone . as expected from the scaling analysis discussed in section [ sec : scaling ]
, a comparison of the experimental data with that predicted by the proposed cell kinetic model shows a good agreement at long times ( fig .
[ fig : avgsize ] ) .
however , comparison of the data at intermediate time - scales provides significant new insight .
in particular , if we assume _ equal _ labelling efficiency of progenitor and differentiated cells , i.e. that both cell types label in proportion to their steady - state population ( shown as the lower ( red ) curve in the fig . [
fig : avgsize ] ) , then there is a substantial departure of the predicted curve from the experimental data for times of between two and six weeks .
intriguingly , if we assume that differentiated cells simply do nt label , then the agreement of the data with theory is excellent from two weeks on ! we are therefore lead to conclude that , at least from two weeks , all labelled clones derive from progenitor cells labelled at induction . with this in mind
, we may now turn to the average clone size as inferred from the data at two days and one week .
here one finds that the model appears to substantially over - estimate the clone size .
indeed , fig .
[ fig : avgsize ] suggests that the average clone size is pinned near unity until beyond the first week post - labelling , i.e. the relative population of single - cell clones is significantly _ larger _ than expected at one week , yet falls dramatically to the theoretical value at two weeks . referring again to the slow accumulation of eyfp , can one explain the over - representation of single - cell clones at one week post - labelling ?
at one week , two - cell clones are observed soon after cell division , and thus express lower concentrations of eyfp compared to single - cell clones . as a result
they may be under - represented . at later times , all labelled clones become visible as eyfp concentration grows , explaining the coincidence of experiment and theory at two weeks .
it follows , of course , that the size distributions at later time points would be unaffected by slow eyfp accumulation
. however , a full explanation of this effect warrants further experimental investigation , and is beyond the scope of this paper .
having elucidated the mechanism of normal skin maintenance , it is interesting to address its potential as a predictive tool in clonal analysis . conceptually
, the action of mutations , drug treatments or other environmental changes to the tissue can effect the non - equilibrium dynamics in a variety of ways : firstly , a revision of cell division rates or `` branching ratios '' ( i.e. symmetric vs. asymmetric ) of _ all _ cells may drive the system towards either a new non - equilibrium steady - state or towards a non - steady state evolution resulting in atrofication or unconstrained growth of the tissue . (
the development of closed non steady - state behaviour in the form of limit cycles seems infeasible in the context of cellular structures . )
secondly , the stochastic revision of cell division rates or branching ratios of _ individual _ cells may lead to cancerous growth or extinction of a sub - population of clones .
the former may be referred to as a `` global perturbation '' of the cell division process while the second can be referred to as `` local '' . in both cases
, one may expect clonal analysis to provide a precise diagnostic tool in accessing cell kinetics . to target our discussion to the current experimental system , in the following we will focus on the action of a local perturbation in the form of a carcinogenic mutation , reserving discussion of a global perturbation , and its ramifications for the study of drug treatment , to a separate publication .
let us then consider the action of a local perturbation involving the activation of a cancer gene in a small number of epidermal cells , which leads to the eventual formation of tumours . in the experimental system , one can envisage the treatment coinciding with label induction , for example by simultaneously activating the eyfp and the cancer gene . in this case , clonal fate data should simply reflect a modified model of cell proliferation leading to the eventual failure of the steady - state model of tissue maintenance . to quantify the process of cancer onset , we start by establishing the simplest possible changes to process ( [ modelratelaws ] ) which may be associated with tumour growth .
cancer is widely held to be a disease caused by genetic instability that is thought to arise when a progenitor cell undergoes a series of mutations @xcite . as a result , cells within the mutant clone prefer to proliferate , on average , over processes leading to terminal differentiation or death . in this investigation
we shall consider a `` simple '' cancer resulting from _ two _ rate - limiting mutations : referring to our proposed labelling experiment , the controlled induction of a cancer - causing mutation during label induction defines the first mutation ; a second , rate - limiting step then occurs with the stochastic occurrence of a second cancer causing mutation .
examples of the first type of mutation may be genes that affect the ability of a cell to respond to genetic changes of the cell , e.g. _ p53 _
, whilst the second mutation may be of a gene that affects clone fate such as the _ ras _ oncogene @xcite .
we may therefore distinguish between `` stage one '' mutated cells , which maintain the steady - state , and `` stage two '' cells , which have the capacity for tumour formation .
the resulting process of cell proliferation is set by three parameters : the overall rate of mutation @xmath105 from a stage one a cell into a cancerous stage two cell ; the division rate @xmath106 of the stage two cells ; and the degree of imbalance @xmath107 between their stochastic rate of proliferation and differentiation . in summary , focusing on the proliferating cell compartment only , and denoting the stage two mutated cells as type a@xmath108 , then the revised cell proliferation model includes the additional non - equilibrium processes @xmath109 the rate @xmath105 may be interpreted as the mean rate with which a stage - one cell acquires an additional mutation necessary to activate a second oncogene .
the mutated cells then give rise , on average , to an exponentially growing cell lineage with growth rate @xmath110 .
this nonequilibrium process was originally addressed by kendall , who predicted the distribution in the number of tumours detected at time @xmath5 after mutation @xcite .
his focus on tumour statistics may reflect the experimental limitations in clonal analysis at the time : until recently it was not possible to reliably detect clones at all , let alone to count the number of cells per clone .
experimentally , however , the clone size distributions are a more efficient measure of cell kinetics than the tumour number distributions , because they result in a far richer data set , and are accessible within weeks rather than months
. we shall therefore extend kendall s approach to predict the clone size distributions at times far earlier than tumour appearance . to familiarise ourselves with the modified model ,
consider the evolution of the average clone size with time . focusing on the proliferating cell compartment with @xmath4 type a cells and @xmath111 type a@xmath108 cells in a clone , the relevant mean - field equations are @xmath112 which give the expected shift from linear growth of clones in normal skin to that of exponential growth , @xmath113 . more interestingly , referring to the master equation below
, one may show that the variance in clone size also changes qualitatively : whereas for normal skin the rms variance in clone size grows as @xmath114 , here the variance in the long - time limit is _ finite _ , @xmath115 that is , the relative broadening of the clone size distribution observed in normal skin is halted by the introduction of an exponentially growing cell population .
these observations may already provide a crude method for identifying carcinogenesis through clonal analysis . to do better ,
it becomes necessary to solve for the full size distribution by extending the master equation ( [ pmastereqn ] ) to include process ( [ processeqn : cancer ] ) .
if we neglect the fate of differentiated cells , then the master equation now describes the evolution of the probability @xmath116 for finding @xmath4 type a cells and @xmath111 type a@xmath108 cells in a clone , @xmath117 + r\lambda\left[(n+1)p_{n+1,n^*}-n p_{n , n^ * } \right]+ \nu\left[(n+1)p_{n+1,n^*-1}-n p_{n , n^ * } \right]\nonumber\\ & & \qquad\qquad + \frac{1+\delta}{2}\mu\left[(n^*-1)p_{n , n^*-1 } - n^ * p_{n , n^ * } \right]+ \frac{1-\delta}{2}\mu\left[(n^*+1)p_{n , n^*+1 } - n^ * p_{n , n^ * } \right]\ , , \nonumber\end{aligned}\ ] ] subject to the experimental boundary condition @xmath118 corresponding to exactly one `` stage one '' cell per clone at @xmath119 . as for the case of normal skin , we shall later be interested in the distribution of persistent clones , defined as , @xmath120 while it is not possible to solve eq .
( [ eqn : cancermaster ] ) analytically , progress may be made when we allow for the widely - accepted view that tumours are _ monoclonal _ , that is they arise from a single `` stage two '' mutated cell @xcite .
this assumption conveniently limits us to the parameter space @xmath121 , for which an approximate long - time solution for the full clone size distribution may be found . referring to the appendix for details
, we find that the binned clone size distribution takes the long - time asymptotic scaling form , @xmath122\,,\end{aligned}\ ] ] where @xmath123 @xmath124 @xmath125 , @xmath126 , @xmath127 , and @xmath128 . despite its apparent complexity
, this distribution is characterised by a simple scaling behaviour : referring to fig .
[ fig : cancer](a ) , the predicted clone size distributions are plotted using the scaling appropriate to the normal ( unperturbed ) system ( cf .
[ fig : scaling ] ) . in this case
, it is apparent that the scaling @xmath129 fails .
by contrast , from the expression for @xmath130 , it is clear that the size distributions should scale according to the time translation , @xmath131 as confirmed by the results shown in fig . [
fig : cancer](b ) . further consideration of the size distribution exposes several additional features , which may provide further access to the new model parameters : + * the long - time distribution decays with a rate @xmath132 : expanding @xmath133 for small @xmath134 gives us the asymptotic form of the universal decay curve . for @xmath135 ,
consistent with the monoclonicity requirement @xmath136 , we find @xmath137 where @xmath138 denotes the gamma function .
this expression allows us to estimate @xmath139 from the rescaled clone size distributions , providing access to the cell division and mutation parameters of the observed cells . *
the probability of tumour formation is finite : this is a well - known feature of the simple non - critical birth - death process ( [ processeqn : cancer ] ) @xcite . referring to the appendix , we find that the probability @xmath140 for any given clone to survive and form a tumour is finite , @xmath141 as a result , the onset of cancer will halt the steady decrease in the density of labelled clones that is a hallmark of the unperturbed system .
these properties , and especially the change in scaling behaviour , allow the onset of early - stage cancer to be identified from observations of clones less than one hundred cells in size .
this may provide a dramatic improvement both in speed and accuracy over current experimental models , which rely on much later observations of tumours ( or hyperplasias ) in order to deduce the cell kinetics at early - stages .
to summarize , we have shown that the range of clone fate data obtained from measurements of murine tail epidermis are consistent with a remarkably simple stochastic model of cell division and differentiation involving just one proliferating cell compartment .
these findings overturn a long - standing paradigm of epidermal fate which places emphasis on a stem cell supported epidermal proliferative unit . as well as providing significant new insight into the mechanism of epidermal homeostasis
, these results suggest the utility of inducible genetic labelleling as a means to resolve the mechanism of cell fate in other tissue types , and as a means to explore quantitatively the effects of drug treatment and mutation . to conclude
, we note that the analysis above has focused on the dynamics of the clonal population without regard to the spatial characteristics .
indeed , we have implicitly assumed that any model capable of describing the cell size distributions will also succeed in maintaining the near - uniform areal cell density observed in the basal layer . however , it is known that , when augmented by spatial diffusion , a simple galton - watson birth - death process leads to `` cluster '' formation in the two - dimensional system whereupon local cell densities diverge logarithmically @xcite .
significantly , these divergences can not be regulated through a density - dependent mobility .
understanding how the galton - watson process emerges from a two - dimensional reaction - diffusion type process represents a significant future challenge . from a practical perspective
, there is also the significant question of how the cell kinetic model might be generalised to describe other forms of epidermis . in particular , it is not feasible to repeat these experiments _ in vivo _ in human epidermis , a system of obvious medical significance . therefore , it may be of great interest to determine , in future studies , the extent to which our results compare with the behaviour found in other systems .
lastly , our analysis of the cancer system referred to the relatively simple case of a two - stage mutation .
it is , of course , well - known that tumour formation is usually the result of multiple mutations .
understanding whether clonal fate data can be used to probe the kinetics of _ multi - stage _ mutation remains an interesting future challenge .
acknowledgements : we are grateful to sam edwards , martin evans , bill harris , marc kirschner , and gunter schtz for useful discussions .
the experimental aspect of this work was funded by the medical research council , association for international cancer research and cancer research uk .
to derive the clone size distribution given in eq .
( [ cancersummarysoln ] ) , we start by quoting the known result for the probability distribution @xmath142 of finding @xmath111 stage - two a@xmath108 cells at time @xmath5 starting from a single a@xmath108 cell at time @xmath80 @xcite , when @xmath144 , this distribution asymptotes to the form @xmath145 from the value of @xmath146 we see that even when a cell has mutated , it is not guaranteed to result in a tumour : this will only occur with a probability of @xmath147 .
the value of @xmath148 plays an important role in determining the statistics of tumour formation , as will be seen below .
we now make two approximations : first , we take the long - time clone size distribution to be dominated by the statistics of a@xmath108 cells . this is a safe assumption at times @xmath149 and @xmath150 , as may be seen by considering the behaviour of the mean - field equations in section [ sec : cancer : soln ] .
this approximation allows us to focus on the size distribution of a@xmath108 cells only , @xmath151 , which is related to the full clone size distribution by the sum @xmath152 .
secondly , we assume that the entire population of type a@xmath108 cells in each clone arises from the first mutated cell that gives rise to a stable , exponentially growing lineage of cells .
this corresponds to the condition @xmath121 , as discussed in the main text . with these two approximations ,
the probability of finding a labelled clone containing @xmath153 mutated cells is given by the population distribution of the _ first surviving cell lineage _ of a@xmath108 cells , @xmath154 where @xmath155 is some normalisation constant , and we have introduced the probability @xmath156 for the @xmath157-th lineage of mutated cells within a given clone to be created during the interval @xmath158 through the mutation process @xmath159 .
the weight factor @xmath160 gives the probability that the first @xmath161 cell lineages of a@xmath108 cells within a clone will become extinct a situation necessary to make the @xmath157-th cell line relevant to the distribution according to the monoclonal approximation .
the rates @xmath162 may be accessed by considering the probability @xmath163 that a clone containing @xmath4 type a cells at time @xmath5 also contains @xmath157 _ independent lineages _ of mutated a@xmath108 cells , each arising from a separate mutation event .
( later we shall treat the evolution of each of these cell lines post - creation ) . to solve for @xmath163
we must introduce its moment - generating function @xmath164 , which evolves ( from eq .
[ eqn : cancermaster ] ) according to the dynamical equation , @xmath165\partial_qg\,.\ ] ] solving this equation subject to the initial condition of one `` stage - one '' ( a ) cell per clone , we find the solution @xmath166 with @xmath167 .
( [ g1soln ] ) describes the evolution of a single a cell as it proliferates and eventually gives rise to a set of internal lines of mutated cells . before we proceed to find @xmath151 , note that setting @xmath168 in eq .
( [ g1soln ] ) gives us the result ( quoted in the main text ) for the asymptotic fraction @xmath140 of clones in which all mutated cell lines become extinct , @xmath169 on the other hand , setting @xmath170 only in eq .
( [ g1soln ] ) gives the moment - generating function for ( yet another ) distribution @xmath171 of a clone containing @xmath157 independent lines of a@xmath108 cells irrespective of the number of normal cells in the clone . finally , noting that @xmath172 then gives :
@xmath173 which we may substitute into eq .
( [ eqn : case2pdefn ] ) to find ( for @xmath153 ) @xmath174 from this expression , simplified by the large-@xmath111 approximation @xmath175 , we obtain the final form of the binned size distribution given in eq .
( [ cancersummarysoln ] ) . | the rules governing cell division and differentiation are central to understanding the mechanisms of development , aging and cancer . by utilising inducible genetic labelling ,
recent studies have shown that the clonal population in transgenic mouse epidermis can be tracked _ in vivo_. drawing on these results , we explain how clonal fate data may be used to infer the rules of cell division and differentiation underlying the maintenance of adult murine tail - skin .
we show that the rates of cell division and differentiation may be evaluated by considering the long - time and short - time clone fate data , and that the data is consistent with cells dividing independently rather than synchronously .
motivated by these findings , we consider a mechanism for cancer onset based closely on the model for normal adult skin . by analysing the expected changes to clonal fate in cancer emerging from a simple two - stage mutation
, we propose that clonal fate data may provide a novel method for studying the earliest stages of the disease . |
a well known formal difficulty of the standard model of particle physics is that the higgs boson mass has quadratically divergent contributions from fermion and gauge boson loops which must be cancelled by a ` fine - tuned ' tree - level mass term .
this fine - tuning is removed in supersymmetric models by introducing boson partners for standard model fermions , and fermion partners for standard model bosons .
cancellations of quadradic divergences then arise from the differing signs of bosonic and fermionic loops .
recently a new class ` little higgs ' models@xcite , has emerged in which cancellations occur between paired particles of the _ same _ statistics : fermions are paired with new , massive fermions , and bosons with new massive bosons . this little higgs mechanism , as realized in various models , requires a set of new massive gauge bosons , some of which mix with gauge bosons of the standard model . for some range of mixing angles the coupling of gauge bosons to scalars can become strong , resulting in a breakdown of perturbative calculation .
in this work we probe the limits of perturbative reliability by examining the decay partial widths of a massive gauge boson .
the prototype little higgs model is the ` littlest higgs'@xcite .
here we study the decays of a new neutral @xmath3 gauge boson in the littlest higgs model , the @xmath0 .
the presence and properties of this particle are central to the divergence - canceling mechanism of the littlest higgs model .
a catalog of feynman rules for the littlest higgs model can be found in reference @xcite , along with a number of phenomenological results .
a few important corrections can be found in reference @xcite .
additional rules for scalar couplings can be found in reference @xcite . here
we will sketch the outlines of the model as needed to establish notation and motivate our work .
the enlarged electroweak space of the littlest higgs model is the symmetric tensor representation of @xmath6 , which is broken at a scale @xmath7 by a tensor vacuum expectation value to the coset @xmath8 .
the result is @xmath9 goldstone bosons , which are parameterized by a @xmath10 non - linear sigma field @xmath11 .
four of these goldstone bosons will become the complex higgs doublet of the standard model , four will become the longitudinal modes of four new massive gauge bosons , and the remaining six will form a massive complex scalar triplet .
the @xmath6 group contains an @xmath12 subgroup , which is gauged , with gauge couplings @xmath13 and @xmath14 , respectively .
the result is eight gauge bosons , which after diagonalization to mass eigenstates are the gauge bosons of the standard model , here labeled @xmath15 , @xmath16 , @xmath17 , and new gauge bosons with masses of order @xmath7 , labeled @xmath18 , @xmath0 , @xmath19 .
the mass diagonalization can be described in terms of the mixing angles : @xmath20 and @xmath21 with @xmath6 represented in this way , quadratic higgs mass divergences from standard model gauge boson loops are cancelled by opposite - sign contributions from the new massive gauge bosons , leaving a logarithmic divergence . at some scale above @xmath22
tev this logarithmic contribution must also be cancelled if fine - tuning is to be avoided , so we do not view the littlest higgs model as complete , but instead expect at sufficiently high energies additional contributions from an unidentified complete theory , the so - called ultraviolet completion " @xcite .
.[lhparticles ] new particles in the littlest higgs model @xcite@xcite , where @xmath23 , @xmath24 .
[ cols="^,^,^",options="header " , ] although standard model final state masses are negligible at @xmath25 tev , much lower symmetry - breaking scales are not excluded by direct searches for some regions of parameter space@xcite . at @xmath26 tev ,
the final state mass effects are on the order of ten percent .
we surveyed the available littlest higgs couplings and particle masses to determine the possible three - body decays of the @xmath0 .
there are eighteen distinct kinematically - allowed three - body tree - level decays for the @xmath0 , which are listed in table [ threes ] .
many of these decay processes are possible only for small or very large values of @xmath27 ( where @xmath0 becomes very massive ) due to the presence of one or more tev - scale particles in the final state .
note that no decays to the @xmath28 are possible because it has the same mass as the @xmath0 to high order in @xmath29 .
= 1.bp ( 388,140)(3,1 ) ( 5.0,10.0)(14.0,10.0)(0.,)/sine0 ( 6.5,11.0)[b]@xmath0 ( 14.0,10.0)0 ( 14.0,10.0)(21.0,15.0)(0.,)/sine0 ( 18.5,15.0)[b]@xmath16 ( 14.0,10.0)(21.0,10.0)(0.,)/scalardash0 ( 23.0,10.0)[b]h ( 14.0,10.0)(21.0,5.0)(0.,)/scalardash0 ( 23.0,4.5)[b]h ( 7.0,10.0)(14.0,10.0)(0.,)/sine0 ( 8.5,11.0)[b]@xmath0 ( 14.0,10.0)0 ( 14.0,10.0)(23.0,12.0)(0.,)/sine0 ( 18.5,13.0)[b]@xmath30 ( 14.0,10.0)(23.0,5.0)(0.,)/scalardash0 ( 25.0,4.5)[b]h ( 23.0,12.0)0 ( 23.0,12.0)(28.0,9.22)(0.,)/scalardash0 ( 30.0,7.0)[b]h ( 23.0,12.0)(28.0,12.0)(0.,)/sine0 ( 30.0,12.0)[b]@xmath16 ( 12.0,10.0)(21.0,10.0)(0.,)/sine0 ( 13.5,11.0)[b]@xmath0 ( 21.0,10.0)0 ( 22.0,7.25)[b]h ( 21.0,10.0)(28.0,15.0)(0.,)/sine0 ( 25.5,15.0)[b]@xmath16 ( 24.5,7.5)0 ( 24.5,7.5)(28.0,10.0)(0.,)/scalardash0 ( 30.0,10.0)[b]h ( 21.0,10.0)(28.0,5.0)(0.,)/scalardash0 ( 30.0,4.5)[b]h these decays will characteristically each involve a set of eight or so feynman diagrams , including one with a quadrilinear vertex factor .
we exclude from our list processes without a quadrilinear coupling and with a related kinematically possible two - body decay , since these would constitute a double - counting of a two - body decay with an added decay step for one of the decay products . the diagrams for the process @xmath1 are displayed in figure [ feynmans ] .
note that a diagram with a @xmath31 particle propagator is not present because the combined effects of a factor of @xmath29 at each vertex places this diagram at higher order in the sigma expansion than we are working .
since the calculation of a three - body decay is analytically difficult , especially because we wish to retain final state masses , we implemented a monte carlo integration to determine the partial width for @xmath1 .
this partial width becomes very large at large and especially at small @xmath27 .
we would expect this behavior because many vertex factors contain both @xmath27 and @xmath32 in the denominator .
in fact this effect is present in two - body decays , but for three - body decays the width divergence is much more severe .
the width - to - mass ratios for the processes @xmath2 and @xmath1 are shown in figure [ threep ] .
the three - body partial width exceeds the two - body partial width for @xmath33 .
this is clearly pathological , since three - body decays should be coupling - suppressed compared to two - body decays .
what is happening ?
the scalar - gauge boson couplings originate in the scalar kinematic lagrangian term@xcite : @xmath34 where @xmath35 is the gauge - covariant derivative in the bifundamental representation of @xmath6 , @xmath36 and @xmath37 is the nonlinear representation : @xmath38 and where @xmath39 is a @xmath40 x @xmath40 matrix which parametrizes the scalars in the theory , including the longitudinal modes of massive standard model gauge bosons as well as the new tev scale gauge bosons . the constant matrix @xmath41 is the choice of vacuum expectation value for this model@xcite . in the gauge - covariant derivative
the index @xmath42 is over the pairs of electroweak gauge groups , and the index over the generators of @xmath3 are suppressed . as @xmath37
is expanded in a power series , an infinite set of vertex factors will result from the scalar kinematic term , each involving two gauge bosons , some number of scalars and a power of @xmath43 , with the factor @xmath44 appearing due to mixing of the two @xmath3 gauge groups . the lowest order terms will be two gauge bosons plus one scalar , then plus two scalars , then three scalars , and so on .
the existence of vertex factors connecting more than four external particles signals the fact that any nonlinear sigma model is a nonrenormalizable effective theory .
the effects of this nonrenormalizability would become evident in loop calculations , but here we are considering a different kind of divergence - that of the expansion of the @xmath37 field itself .
we might expect this divergence to become parametrically significant at about @xmath45 , which at @xmath25 tev occurs at @xmath46@xcite .
instead , we see the pathological effect of the decay to @xmath47 exceeding the decay to @xmath48 at as high as @xmath49 , as we approach the strong coupling regime . we have not presented the partial widths for other three - body processes , but can note that the decay @xmath50 should approximate that of @xmath47 due to the goldstone boson equivalence across these processes . with the exception of @xmath51 , all other processes will be kinematically suppressed due to the presence of one , two , or three tev - scale particles in the final state .
these processes will in fact be kinematically excluded except for small or very large @xmath27 . by combining these contributions we can roughly estimate that the total partial width to three final state particles is less than an order of magnitude larger than that to @xmath47 for all but the smallest and largest values of @xmath27 .
we have found that a characteristic three - body decay can become large at small and very large values of the @xmath3 mixing angle @xmath27 , and exceeds a related two - body decay at @xmath33 , indicating divergence of the sigma field expansion .
the pathology of the littlest higgs model in this region of parameter space is due to a nonlinear sigma scalar field representation , and occurs over a larger regions of parameter space than a simple parametric calculation would suggest .
although perturbative calculations break down in this region , we can note an interesting qualitative feature : the littest higgs model near the nonperturbative region would produce decays with high scalar multiplicities , especially with multiple higgs boson production , and so with large b quark multiplicities .
this feature of large rates of scalar production in decays may be a general feature of nonlinear sigma fields near their perturbative limits , as the regime of strong dynamics is approached .
in particular , we would expect similar results for decays of the littlest higgs @xmath18 and @xmath52 , and for the new gauge bosons of other models in the little higgs class , such as the simplest higgs @xmath53 .
we should note , however , that all resuts presented here are at tree - level only , and could be significantly modified by the effects of loop corrections .
we have presented the various expressions for two - body decays of the littlest higgs @xmath0 , including all final state masses .
the numerical width - to - mass ratios for the most significant of these processes , and for a characteristic three - body decay process , confirm that the @xmath0 is a narrow resonance , as needed for dilepton invariant mass searches @xcite .
this material is based upon work supported by the department of energy under award number de - fg02 - 91er40685 .
thanks to dave rainwater for various helpful discussions . | the little higgs mechanism , as realized in various models , requires a set of new massive gauge bosons , some of which mix with gauge bosons of the standard model . for a range of mixing angles the coupling of gauge bosons to scalars can become strong , ultimately resulting in a breakdown of perturbative calculation .
this phenomenon is studied in the littlest higgs model , where the approach to strong dynamics is characterized by increasing tree - level decay widths of the neutral @xmath0 boson to lighter gauge bosons plus multiple scalars .
these increasing widths suggest a distinctive qualitative collider signature for the approach to the strong coupling regime of large higgs and other scalar multiplicities . in this work we catalog
the kinematically allowed three - body decays of the @xmath0 , and calculate the partial width of the process @xmath1 .
this partial width is found to be larger than the comparable two - body decay @xmath2 for values of the @xmath3 mixing angle cosine @xmath4 @xmath5 , indicating divergence of the littlest higgs sigma field expansion at values of cosine @xmath4 larger than a simple parametric calculation would suggest .
additionally , we present analytical expressions for all two - body decays of the littlest higgs @xmath0 gauge boson , including the effects of all final - state masses . |
new neutral gauge bosons @xmath1 are a feature of many models addressing the physics beyond the standard model ( sm ) .
the phenomenology of possible heavy gauge bosons has been extensively studied in the past .
there are stringent limits on their mass and mixings from precision electro - weak experiments @xcite and from direct searches @xcite .
the existence of a @xmath1 affects precision electro - weak data ( a ) because @xmath5 mixing pushes the @xmath6 mass below the standard model expectation ; ( b ) this expectation is itself modified by mixing , because it depends on the weak angle , the value of which is confused or distorted by the effects of mixing on other observables .
( c ) both the mixing and the heavy particle exchange lead as well to other changes in the predictions for the various observables , implying new terms in the effective interactions relevant to each process and leading to different apparent values of the weak angle determined in different processes .
thus , the limits from precision experiments vary significantly from model to model because of the different chiral couplings to the ordinary fermions . typically , the mass of a heavy @xmath1 must exceed @xmath7 400 gev and the @xmath5 mixing angle must be smaller than a few times @xmath8 in those models in which the @xmath1 couples significantly to charged leptons , including the standard examples from grand unification @xcite
. models with suppressed couplings to charged leptons can tolerate much larger mixings ( several percent ) , with the dominant constraint coming from the shift in the light @xmath6 mass , as is described in the appendix .
such models may be motivated @xcite by the possible enhancement in @xmath9 decays suggested by the lep data @xcite .
the direct production limits @xcite are likewise very sensitive to the @xmath1 couplings as well as to the number of open channels for the decays ( _ e.g. _ , into exotic particles or superpartners ) , but are again typically in the 300 - 600 gev range for grand unification type models .
the identification and diagnostic study of heavy gauge bosons at future colliders has been investigated in detail @xcite .
it should be possible to discover a heavy @xmath1 with mass up to @xmath10 tev through its leptonic decay channels , @xmath11 , @xmath12 or @xmath13 , at the lhc , while @xmath14 interference effects should be observable at a 600 gev @xmath15 collider ( nlc ) for @xmath16 up to @xmath17 tev .
if such a @xmath1 exists , it should be possible to obtain useful diagnostics about its coupling by forward - backward asymmetries , rapidity distributions , rare decays such as @xmath18 , and associated productions with a @xmath6 , @xmath19 , or @xmath20 , for masses up to @xmath21 tev at the lhc , with the lhc and nlc providing complementary information .
( for a review see ref . @xcite and references therein . )
there have also been studies of the present and future constraints on possible exotic particles , such as heavy non - sequential quarks , leptons , or standard model singlets @xcite .
for example , some models predict the existence of a heavy , @xmath22-singlet , charge @xmath23 quark , @xmath24 , which could be produced at a hadron collider by ordinary qcd processes and decay by @xmath25 mixing into , _
@xmath26 , @xmath27 , or @xmath28 , where @xmath29 is a neutral higgs boson .
typically , the @xmath26 , @xmath27 , and @xmath28 decays occur in the ratio @xmath30
. currently , @xmath31 85 gev if it mixes mainly with @xmath32 @xcite .
additionally , precision experiments ( weak charged current , neutral current , and flavor changing constaints ) typically imply that the mixing between @xmath33 and @xmath34 is less than @xmath35 @xcite .
similarly , some models imply the existence of new @xmath22-doublets of leptons , @xmath36 and @xmath37 , which can be produced by ordinary electro - weak processes , with the lightest decaying by mixing with ordinary leptons . in both examples , the exotic particles are vector , _
i.e. _ , both the @xmath38 and @xmath39 fields have the same standard model transformation properties .
vector particles could actually have bare masses as far as the standard model is concerned , but in most theories there are additional symmetries which forbid these and require that the masses be generated by some form of spontaneous symmetry breaking .
new vector multiplets ( or supermultiplets ) could in principle have arbitrarily large masses since they do not require @xmath40 breaking , and they are the most commonly predicted new type of matter in many standard model extensions . in models with extended gauge symmetry ,
but which do not incorporate constraints of such underlying dynamics as grand unification , supersymmetry , supergravity , or string theory , the mass and couplings of additional gauge bosons are free parameters , and thus their masses can be anywhere from about 1 tev to the planck scale @xmath41 .
in addition , the masses and couplings of the additional exotic particles which usually accompany the @xmath1 are also free parameters .
thus , such models lack a unique predictive power for @xmath1 physics at future experiments .
the situation is different for models that incorporate constraints of supersymmetry : supergravity theories and , in particular , those that are predicted from string theory .
string theory models necessarily include gravitational interactions ( as a consistent part of the theory ) , and supersymmetry is restored at sufficiently large energy scales . in addition , for each of the models the light particle spectrum as well as their couplings are calculable , thus providing a major advantage over models in which both the particle content and the couplings are put in the theory by hand .
unfortunately , there are two major hurdles that prevent one from obtaining unique low energy predictions of string theory : ( i ) by now a very large number of string models have been constructed , and one does not have a first principle guidance to single one model over another ; ( ii ) while these models possess supersymmetry at large energy scales , one does not presently know how to break supersymmetry without introducing new parameters .
both problems are believed to have an ultimate resolution in the non - perturbative string dynamics . in spite of the above unsolved problems one can take a less ambitious attitude and consider only those string models which have a potential to be realistic . those are string models with supersymmetry , the standard model ( sm ) gauge group as a part of the gauge structure , and a particle content that includes three sm families @xcite and at least two sm higgs doublets , _
i.e. _ , the string vacua which have at least the ingredients of the mssm
. a number of such string models have been constructed @xcite . also , one may parameterize our ignorance of supersymmetry breaking by introducing supersymmetry breaking mass and trilinear interaction terms .
the mass terms ultimately drive the electro - weak symmetry breaking , and thus are of the order of the electro - weak scale .
the models which have been constructed often contain additional @xmath0 symmetries and additional exotic matter multiplets .
a class of string models with the features mentioned above and an additional @xmath0 symmetry provide a testing ground to address the following phenomenological issues : ( i ) a scenario , which specifies the scale of @xmath1 and ( ii ) the mass and phenomenological implications of the exotic particles accompanying the new gauge boson .
we have identified several distinct scenarios , each of which can be illustrated by a specific string model .
a main conclusion is that a large class of string models not only predict the existence of additional gauge bosons , but often imply the masses of the new gauge bosons and the exotic particles which necessarily accompany them to be in the electro - weak range .
each specific model leads to calculable predictions ( which , however , depend on the assumed supersymmetry breaking mass terms ) for the masses , couplings , and mixing with the @xmath6 of the new boson(s ) , as well as for the masses and quantum numbers of the associated exotic matter .
we would thus like to advocate that , from the string point of view , new gauge bosons and associated exotic matter are perhaps the next best motivated new physics
after the higgs and supersymmetric particles to be searched for at future experiments .
the paper is organized as follows . in section 2
we specify in more detail the features of the string models and their advantage over sm physics . in section 3
we give the specific scenarios for the @xmath1 masses achievable without excessive fine - tuning of the supersymmetry breaking parameters and constraints on exotic particles accompanying the new gauge boson(s ) .
we show examples in which the additional @xmath1 mass is : ( a ) comparable to that of the @xmath6 ( already excluded ) ; ( b ) in the 300 gev to 1 tev range , which may still be barely allowed but easily within the range of future or present colliders ; ( c ) at an intermediate scale ( _ e.g. _ , @xmath42 gev ) . it is argued that in case ( b ) it is difficult though not impossible to satisfy existing constraints on @xmath5 mixing , especially for lower values of @xmath16 , and that @xmath1 masses above 1 tev are not expected ( given our assumptions ) without excessive fine tuning .
conclusions are given in section 4 .
the appendix discusses the limits on @xmath5 mixing .
the results presented in this paper provide a summary of ref .
@xcite , but with a more phenomenological emphasis .
the more detailed version @xcite also provides additional technical details and illustrations of the scenarios within specific string models .
let us first specify the generic features of supersymmetric string models with the standard model ( sm ) gauge group @xmath43 , three ordinary families , and at least two sm doublets , _
i.e. _ , a set of models with at least a particle content of the minimal supersymmetric standard model ( mssm ) .
in addition , from a set of models we select only those with @xmath44 , @xmath22 and @xmath45 all embedded into the @xmath46 gauge group , since for other types of embedding the normalization of the @xmath45 gauge group coupling is different from the one leading to the gauge coupling unification in the mssm model .
in general these models also contain a set of particles which are singlets of the sm gauge group , but which transform non - trivially under an additional non - abelian `` shadow '' sector group , and a number of additional @xmath47 s , one of them usually anomalous .
the shadow sector non - abelian gauge group may play a role in dynamical supersymmetry breaking . in addition , there are typically a large number of additional matter multiplets , which transform non - trivially under @xmath47 s and/or the standard model symmetry . in general such models also lead to fractionally charged color singlets , which may have important phenomenological@xcite consequences .
the anomalous @xmath47 group of such models is broken at @xmath48 @xcite by nonzero vacuum expectation values ( vev s ) of certain multiplets which preserve supersymmetry and consistency of the theory at the loop level of string theory . at the same time
, this mechanism ensures that a number of additional non - anomalous @xmath47 s are broken and a number of additional multiplets become massive .
thus , the enhanced gauge symmetry and the exotic particle content is in general drastically reduced .
nevertheless , there are often one or more non - anomalous @xmath47 s and associated exotic matter that are left unbroken .
the physics associated with these left over non - anomalous @xmath0 symmetries is the subject of this paper .
one should , however , point out that the models discussed in general may not be consistent with all of the phenomenological constraints : ( i ) the models could have additional color triplets in the spectrum which could mediate a too fast proton - decay@xcite , ( ii ) the detailed mass spectrum of the ordinary fermions@xcite may not be realistic , ( iii ) a scenario for the symmetry breaking of additional @xmath0s may not be consistent with phenomenological constraints on the exotic multiplets , such as gauge coupling unification @xcite .
in addition , it is not clear how to implement the supersymmetry breaking scenario ; we parameterize it by introducing supersymmetry breaking mass and cubic terms .
within the models discussed in the previous section the pattern of additional @xmath0 symmetry breaking is very constraining .
first , the models possess supersymmetry , which relates the interactions of integer ( boson ) and half - integer ( fermion ) spin particles , _
i.e. _ , those between the particle and its superpartner ; second , the the type of particles and their couplings are calculable . with the assumptions that supersymmetry
is parameterized by supersymmetry breaking mass parameters ( and cubic scalar interactions ) , and that dynamical effects in the shadow sector do not break the @xmath0 symmetry , the @xmath0 symmetry breaking must take place via the higgs mechanism , _
i.e. _ , by giving non - zero vacuum expectation values ( vev ) to spin - zero particles @xmath49 , which are singlets summetry by the vev s of higgs doublets is discussed in @xcite . in that case , it is difficult to achieve a hierarchy between the @xmath1 mass and the @xmath50 masses .
] under the sm symmetry , but carry non - zero charges under the @xmath0 .
non - zero vev s of @xmath49 therefore preserve the sm gauge symmetry ; however , they yield a mass term for @xmath1 , which is of the order of the vev s of @xmath49 .
the analogous higgs mechanism for spontaneous symmetry breaking ( ssb ) of the electro - weak symmetry takes place when the @xmath22 doublet(s ) acquire nonzero vev ( s ) , giving mass to the @xmath6 and @xmath51 gauge bosons .
we assume that the soft supersymmetry breaking mass - square terms for the scalar fields are positive at @xmath52 gev and of the order of gravitino mass @xmath53 , which is assumed to be of @xmath54 tev .
the appearance of positive mass - square terms at @xmath48 is the case in almost all models that have been explored
. then the only way of achieving ssb is via a radiative mechanism .
namely , since such sm singlets have positive ( supersymmetry breaking ) mass - square terms at large energies , these terms need to be driven negative at lower energies to ensure a global minimum of their potential with nonzero vev s for such fields .
the magnitude of the couplings and mass parameters depends on the energy scale at which they are measured .
the renormalization group equations account for the loop - corrections to the physical couplings and mass parameters , and track the dependence of these couplings on the energy scale . in particular ,
some supersymmetry breaking mass - square parameters of the higgs field(s ) @xmath49 can become negative if @xmath49 have a large yukawa coupling ( of @xmath54 ) to fermions with appropriate representations under the sm gauge group , _
e.g. _ , sm doublets or triplets . in this case
the loop corrections due to such yukawa couplings ( see fig.1 ) contribute to the renormalization group equations for the running of the mass - square terms in such a way that the mass - square term for @xmath49 can be driven negative , while that for the supersymmetric partners of the fermions ( that couple to @xmath49 via the yukawa couplings ) remain positive ( see fig .
2 ) . gev ) , while the mass - square terms for the stops ( supersymmetric partner of the top quark ) remain positive .
] thus , the scale of @xmath0 symmetry breaking depends on both the type of sm singlets @xmath49 responsible for the @xmath0 symmetry breaking and on the yukawa couplings of such multiplet(s ) to other exotic particles .
one of the features which distinguishes string models from the generic mssm is that the yukawa couplings are calculable and thus are not free parameters .
interestingly , for the class of string models discussed in the the previous section the corresponding yukawa couplings are either of @xmath55 ( @xmath56-gauge coupling ) or zero .
thus , if the relevant coupling is non - zero it may be sufficiently large to ensure radiative breaking of the @xmath0 symmetry . for each of these possibilities the pattern of @xmath0 symmetry breaking and the running of the gauge couplings still depend on the specific exotic particle content and their couplings .
we shall now discuss the possible scenarios for observable sector @xmath0 symmetry breaking .
for the sake of simplicity we shall address scenarios in which the electro - weak symmetry is broken due to the non - zero vev of the higgs doublet that couples to the top - quark , _
i.e. _ , a large @xmath57 scenario of the mssm
. a generalization to scenarios that accommodate other ranges of @xmath57 is straightforward .
we will emphasize the general features which hold in each scenario .
however , we emphasize that in each specific model the @xmath1 mass , mixing , and couplings , as well as the properties of the exotic matter , are in principle calculable , though in practice they depend on the details of the soft supersymmetry breaking
. suppose that the radiative breaking of @xmath0 is due to one sm singlet @xmath58 , that is charged under @xmath0 .
namely , only one @xmath58 has its effective mass - square driven to a negative value at low energies , thus allowing for a non - zero vev .
the @xmath5 mass - square matrix is then : @xmath59 where @xmath29 and @xmath58 denote the vev s for the sm higgs doublet and singlet , respectively , and @xmath60 are the corresponding @xmath0 charges . here
@xmath61 , where @xmath62 are the gauge couplings at the ssb scale for @xmath22 , @xmath45 and @xmath0 , respectively .
the exotic matter to which @xmath58 couples acquires a mass of order @xmath63 , where @xmath64 is the relevant yukawa coupling between the particular exotic matter and @xmath58 .
in general , there will be an additional soft supersymmetry breaking mass term contributing to the mass of the exotic matter even in the absence of the relevant yukawa coupling(s ) .
the nature of the @xmath5 hierarchy now crucially depends on the allowed vev s @xmath58 and @xmath29 , which are constrained by the form of the potential . at low energies
the potential can be written for the particular direction with non - zero vev s as : @xmath65 where we have extracted explicit minus signs from the negative mass - square terms .
note that due to the constraints of supersymmetry the quartic couplings in ( [ potential ] ) are proportional to the gauge couplings and thus are _ not _ free parameters .
one encounters the following two scenarios : * ( i ) : the relative signs of @xmath66 and @xmath67 are opposite .
+ in this case the minimum of the potential ( [ potential ] ) is for : @xmath68 and the @xmath5 mass - square matrix is @xmath69 it is difficult to achieve the needed hierarchy between @xmath70 and @xmath16 , unless @xmath71 and @xmath72 , in such a way that @xmath73 .
the first condition is not normally expected to hold , except in the limiting case @xmath74 . *
( ii ) : the relative signs of @xmath66 and @xmath67 are the same
. + the minimum of the potential ( [ potential ] ) now occurs for : @xmath75 and the @xmath5 mass - square matrix becomes : @xmath76 in this case one encounters an interesting possibility for achieving a hierarchy without an unusually small ratio of @xmath77 , provided @xmath78 . in this limit ,
the @xmath5 mixing angle is @xmath79 for small @xmath80 the @xmath5 mixing could be sufficiently suppressed to be consistent with the experimental bounds for @xmath81 tev , as is further discussed in the appendix .
one can illustrate @xcite the above scenarios in a particular class of string models .
the second case is of interest , because there a reasonable hierarchy can be achieved without fine tuning of the soft supersymmetry breaking parameters or choices of models with unusual values of @xmath82 . _
e.g. , _ such models can provide for a hierarchy @xmath83 , say , @xmath84 .
if , in addition , one has _
@xmath85 , one obtains @xmath86 and the mixing angle @xmath87 . in this example
, @xmath16 is barely within the current experimental bounds , while @xmath88 is too large for most choices of @xmath1 couplings @xcite .
somewhat larger values of @xmath16 and smaller values of @xmath89 may be consistent with observations .
thus , without excessive fine tuning of the soft supersymmetry breaking parameters , the prediction of @xmath16 is within experimental reach of present or future colliders .
however , when the experimental bounds on @xmath16 exceed the 1 tev region , this scenario _ can not _ be implemented without excessive fine tuning of the soft supersymmetry breaking parameters or unusual choices of the @xmath0 charge assignments . in this case ,
negative mass - square terms are induced for two ( or more ) @xmath0 charged sm singlets @xmath90 , whose @xmath91 charges have opposite sign .
it turns out that supersymmetry constraints for the potential ensure that the minimum of the potential is along the vev direction @xmath92 ( @xmath93-flat direction ) .
along this direction the potential for @xmath58 has no quartic terms ; however , the mass square term depends on the energy scale @xmath13 at which it is measured .
namely , one now has to include the renormalization group improved potential , which along the flat direction is of the form : @xmath94 thus , the minimum occurs near the energy scale @xmath95 at which @xmath96 turns negative . in the case of radiative breaking with yukawa couplings of @xmath54 , it turns out that @xmath97 is much larger than the soft supersymmetry breaking mass terms . for the models considered it is @xcite typically four to ten orders of magnitude below @xmath48 .
therefore , in the case of flat directions the scale of symmetry breaking , _
i.e. , _ the vev of @xmath58 , is @xmath98 gev . higher order ( non - renormalizable ) terms , which arise from exchanges of massive ( of @xmath99 ) particles of string theory , are also present in the potential for @xmath58 .
they are of the type @xmath100 ( @xmath101 ) and could compete with the radiative corrections included in ( [ flatpot ] ) .
such terms determine the scale of symmetry breaking to be of the order of @xmath102^{1\over{k+1}})$ ] .
for example , for @xmath103 the symmetry breaking scale is of the order @xmath104 gev . in either case
the @xmath1 acquires mass of the order of the intermediate scale . on the other hand ,
it is straightforward to show that the mass of the physical higgs boson associated with @xmath58 is of the order of the soft supersymmetry breaking mass terms .
the exotic matter to which @xmath58 couples via the yukawa couplings of magnitude @xmath105 acquires mass of order @xmath63 , _
i.e. _ , that of the intermediate scale . in the absence of the relevant yukawa couplings ,
the exotic matter would have a mass set by supersymmetry breaking mass terms , and thus would be of the weak scale .
again there are string models which illustrate such scenarios @xcite , with the appropriate sm singlet multiplets and the necessary couplings to the exotic particles to ensure the symmetry breaking scenario discussed above . even in the absence of a @xmath0 gauge symmetry it is possible for sm singlet scalar fields to acquire negative mass - square values at low energies due to the radiative mechanism if they have sufficiently large yukawa couplings to other fields .
such scalars generally will not have quartic terms in the potential , and thus they would acquire intermediate scale vev s .
let us briefly comment on a few related topics .
we have seen that under a certain set of assumptions the vev s of standard model scalars will typically be either at the electro - weak scale , or at an intermediate scale .
intermediate scales are of interest in implementing the seesaw model of neutrino mass .
however , one may still need non - renormalizable terms in the potential to implement realistic neutrino mass scenarios @xcite . also , one promising scenario for baryogenesis @xcite is that a large lepton asymmetry is generated by the decays of the heavy majorana neutrino associated with the seesaw , and then converted to a baryon asymmetry during the electro - weak transition .
such scenarios , while very attractive , can not occur if there is a @xmath0 which is only broken at the electro - weak to tev scale @xcite , unless , of course , the heavy majorana neutrino is a @xmath0 singlet .
we would also like to point out that the string models with an additional @xmath0 yield a natural solution the so called @xmath13 problem @xcite .
namely , the mass parameter @xmath13 , parameterizing the mixing the two sm higgs doublets , should be of the order of the electro - weak scale to ensure a realistic mass spectrum in the mssm . in string theory ,
at large energy scales the @xmath13 parameter is either of @xmath106 or zero , thus causing the phenomenological @xmath13 problem .
if , in the models with an additional @xmath0 , there is a coupling of the the sm singlet(s ) @xmath58 to the two sm higgs doublets and @xmath58 acquires a non - zero vev s of order 1 tev , then such a term would provide an effective @xmath13 term of the order of the electro - weak scale .
we emphasize that the radiative ssb scenarios discussed in this paper require the existence of sufficiently large yukawa couplings to drive the mass square values of the sm singlet @xmath58 negative .
this is most easily implemented if there exists exotic matter which transforms non - trivially under the sm gauge group .
the exotic matter will then acquire mass terms given by the relevant yukawa coupling times the vev of @xmath58 , as well as contributions from supersymmetry breaking mass terms .
such exotic matter typically exists in string models .
it is usually vector , _
i.e. _ , both the left and right chiral fields occur as @xmath107 doublets , or both as @xmath107 singlets .
however , if it carries sm quantum numbers it can destroy the success of the gauge coupling unification @xcite .
such effects largely cancel if the light exotic matter corresponds to complete @xmath46 multiplets , but that is not typically expected in the types of semi - realistic models we are discussing . preserving gauge unification without fine - tuning is a stringent constraint on string model building , with or without an additional @xmath0 .
we have explored the possible scenarios for ( non - anomalous ) @xmath0 symmetry breaking , as is expected for a class of string models with the standard model gauge group and additional @xmath47 gauge factors which survive into the low energy theory below @xmath48 . under the assumptions that the additional @xmath0 symmetry is not broken by shadow sector effects and that the supersymmetry breaking scalar mass - square terms are positive , the breaking is necessarily radiative .
this requires the existence of additional matter with large enough yukawa coupling to the standard model singlets responsible for the @xmath0 symmetry breaking .
then , within a particular model with definite supersymmetry breaking mass terms , the symmetry breaking pattern , the couplings and the masses of the new gauge bosons , and those of the accompanying exotic particles are calculable . in that sense
the string models yield predictions for the new physics associated with the new gauge bosons .
it turns out that for the class of string models considered the mass of @xmath1 is either of @xmath2 or the intermediate scale of order @xmath108 gev .
however , in both cases the mass of the associated physical higgs bosons is in the electro - weak region .
our major conclusion , therefore , is that a large class of string models considered here not only predict the existence of additional gauge bosons and exotic matter , but often imply that their masses should be in the electro - weak range .
many such models are already excluded by indirect or direct constraints on heavy @xmath1 bosons , and the @xmath5 mixing is often too large , especially for lower values of @xmath16 .
the scenario in which @xmath16 is in the electro - weak range allows , without excessive fine tuning of the supersymmetry breaking mass parameters , for predictions of @xmath16 within experimental reach of present or future colliders . on the other hand ,
when the experimental bounds on @xmath16 exceed the 1 tev region , this scenario _ can not _ be implemented without excessive fine tuning of supersymmetry breaking mass parameters , or unusual choices of @xmath0 charge assignments .
string models provide a concrete set of predictions for new gauge boson physics , thus motivating the @xmath1 searches and the associated exotic particle searches at future experiments , as perhaps the best motivated physics , next to that of the higgs and supersymmetric partner searches .
the work was supported in part by u.s .
department of energy grant no .
doe - ey-76 - 02 - 3071 , the national science foundation grant no .
phy94 - 07194 , the institute for advanced study funds and j. seward johnson foundation ( m.c . ) , and the national science foundation career advancement award phy95 - 12732 ( m.c . ) .
we would like to thank g. kane for encouraging us to write a version of the paper on new gauge bosons with a phenomenological emphasis .
we acknowledge hospitality of the institute for theoretical physics , santa barbara , where the work was initiated .
as described in the introduction , the experimental constraints on the @xmath5 mixing angle @xmath88 from precision ( mainly @xmath6-pole ) data are stringent but very model dependent .
it is common to quote limits on the masses and mixings of a heavy @xmath1 with couplings given by a number of reference models .
these include models based on grand unification : the @xmath109 , occurring in @xmath110 models , and the @xmath111 and @xmath112 , which correspond to two patterns of @xmath113 breaking .
other common reference models include the @xmath114 , from left - right symmetric models , and the @xmath115 , which has the same couplings as the @xmath6 .
the latter would not occur in a realistic gauge theory , but is a convenient reference .
the current 95% cl limits on the mass and mixing for each of these models are shown in the table .
.95% cl lower limit on @xmath16 in gev and 95% cl upper and lower limits on the @xmath5 mixing angle @xmath88 for some common reference models , from precision data ( updated from @xcite ) and direct searches @xcite .
recent ( preliminary ) cdf data increase the direct @xmath115 limit to 625 gev . [
cols="<,^,^,^,^,^",options="header " , ] the limits in the table assume that @xmath88 and @xmath16 are independent parameters .
however , in specific models these are related if one knows the @xmath0 charges of the higgs doublets which break @xmath107 and generate the mixing .
the relation is especially simple if only one @xmath107 doublet is involved .
assuming @xmath116 the mixing angle and masses are related by ( [ thetath ] ) .
this relation is more general than the string or grand unified models considered here , and assumes only that there is a mass hierarachy and that the mixing is dominated by a single higgs doublet . in the reference grand unified and left - right models one finds ( assuming the simplest higgs structure and direct breaking of the grand unified theory to the standard model with an additional @xmath47 ) that @xmath117 is in the range 0.2 0.6 , so the mixing limits in the table imply lower limits on @xmath16 around 1 tev .
( in most cases the limits are relaxed when one allows two or more higgs doublets to contribute , since their contributions can cancel in the off - diagonal term in the @xmath5 mass matrix . ) for such couplings , the mass hierarchies that can be generated without excessive fine tuning in the string models considered here would be inconsistent or only barely compatible with current experimental limits . however , the string - derived models are not expected to have precisely the same couplings and @xmath0 charges as the models based on simple grand unification , and one could have smaller predicted mixings and/or less stringent experimental constraints . for example , it is possible that a new @xmath1 has suppressed couplings to charged leptons or to all ordinary fermions
. even if there are no selection rules to make the relevant charges vanish , one expects @xmath118 to be small if the @xmath1 couples to a large numbers of exotic particles . in a unified theory the overall strength of each gauge interaction , as normalized by @xmath119 ,
should be the same or comparable .
this implies that if a @xmath1 couples to many particles , the coupling to each particle individually is smaller , scaling roughly as @xmath120 , where @xmath121 is the number of particles to which it couples .
( this holds for the couplings at the string scale .
it applies to the running couplings as well if the @xmath121 particles are all light . ) in particular , if @xmath122 is small , then the predicted mixing is reduced .
similarly , small @xmath123 leads to weaker experimental constraints on @xmath88 , since , roughly , the combination @xmath124 enters most of the precision observables .
( of course , suppressed couplings would also make the discovery of a @xmath1 in the future more difficult ) . as an extreme model
, one can consider a @xmath1 that does not have any couplings to ordinary fermions .
the @xmath5 mixing still affects the precision observables because it shifts the @xmath6 mass below the standard model expectation ( see below ) .
a global fit to all data gives the best fit at @xmath16 = 130 gev , although the standard model ( @xmath125 ) is allowed at 90% cl , and @xmath126 .
to understand this relaxed constraint better , it is convenient to consider the relation @xmath127 between the mixing angle , the physical @xmath6 and @xmath1 masses , and @xmath128 , which is the the standard model prediction for the @xmath6 mass in the absence of mixing .
this constraint holds in all models , even when the @xmath1 coupling to the ordinary fermions is absent . assuming @xmath129 and @xmath116
, this implies @xmath130 where @xmath131 and @xmath132 .
the experimental error on @xmath70 is negligible .
however , the standard model prediction for @xmath128 is uncertain due to @xmath133 , @xmath134 , @xmath135 , and @xmath136 , which altogether yield an uncertainty of @xmath137 in @xmath138 .
this uncertainty is increased to @xmath139 if one allows modest couplings of the @xmath1 to electrons , where we have assumed that such couplings can distort the apparent @xmath133 obtained from @xmath6-pole asymmetries by around 0.001 . in figure 3
we indicate the range of upper limits on @xmath88 allowed by ( [ thetamz ] ) for @xmath140 and @xmath141 as a function of @xmath142 .
this is the plausible experimental bound for models with suppressed @xmath1 couplings to charged leptons .
the predicted mixing can be smaller than the grand unification values for small @xmath122 .
it can also be smaller when two or more higgs doublets contribute significantly to the mixing , as is expected to be the case in both superstring models and supersymmetric grand unification ( _ i.e. _ , away from the large @xmath57 limit ) . in this case
, their contributions to the mixing can cancel .
a plausible ( but not rigorous ) range for the effective value of @xmath117 is 0.5 to 0.1 .
this is also shown in figure 3 . | we address the mass ranges of new neutral gauge bosons and constraints on the accompanying exotic particles as predicted by a class of superstring models . under certain assumptions about the supersymmetry breaking parameters
we show that breaking of an additional @xmath0 symmetry is radiative when the appropriate yukawa couplings of exotic particles are of order one , analogous to the radiative breaking of the electro - weak symmetry in the supersymmetric standard model due to the large top - quark yukawa coupling .
such large yukawa couplings occur for a large class of string models .
the @xmath1 and exotic masses are either of @xmath2 , or of a scale intermediate between the string and electro - weak scales . in the former case ,
@xmath3 tev@xmath4 may be achievable without excessive fine tuning , and is within future experimental reach . |
the handling of small liquid droplets in contact with a gas phase on top of a solid substrate , or of liquid lenses at the interface of two other fluid phases plays an important role in the context of microfluidics ( see , e.g. , refs .
@xcite ) . for droplets or lenses with linear extensions in the nanometer regime
a proper thermodynamic description requires to account for not only the bulk and surface properties but also for the special properties of matter in the vicinity of three - phase contact . in order to capture the contribution of the three - phase - contact region to the relevant thermodynamic potential , a certain excess contribution to the appropriate free energy that scales with the linear extension of the system
is associated with the three - phase - contact line , defined by a common intersection of the interfaces meeting in the region of three - phase contact .
the liquid lens is an example of an inhomogeneous system in which three thermodynamic phases , say @xmath0 , @xmath1 , and @xmath2 , coexist in a ( constrained ) equilibrium .
the thermodynamic coexistence of bulk phases is provided by specific choices of the thermodynamic state of the system whereas their spatial coexistence follows from appropriate boundary conditions .
we consider the following set - up .
a drop of the non - wetting @xmath1 phase is placed at the interface between two other fluids . a microscopically thin equilibrium wetting film of @xmath1 phase is formed at the interface and for suitably chosen substances and conditions a surplus of @xmath1 phase forms a lens at the @xmath0@xmath2 interface .
( in principle the exploration of the configuration space allows for shifting the lens laterally along the interface ; however , for the purposes of the present paper one may disregard this degree of freedom . ) for such a system two basic scenarios are possible .
the lens ( @xmath1 phase ) can exchange matter with the surrounding phases , so that the chemical potentials in all phases are equal .
there are cases in which the lens is in a stable equilibrium with the surrounding phases ; in other cases the lens is unstable but could be stabilized by imposing suitable constraints .
alternativly , one may consider a nonvolatile liquid ( @xmath1 phase ) , i.e. , one constrains the volume of the liquid while chemical equilibrium is not attained .
+ in addition to the concept of interfaces separating the bulk phases and of interfacial or surface tensions , the systematic thermodynamic description of such systems leads to considering the contact line @xmath3 along which the three interfaces meet .
the line tension @xmath4 is attributed to the contact line @xmath3 .
it is defined as the line contribution to the grand canonical free energy @xmath5 of the inhomogeneous system per unit length of the contact line @xcite which is the leading term left after subtracting volume and surface or interface contributions from @xmath5 , i.e. , @xmath6 with s.l.t .
we denote subleading terms which vanish for @xmath7 .
the symbol @xmath8 denotes the grand canonical free energy density of the homogeneous bulk phase @xmath9 ( @xmath10 ) , @xmath11 is the volume assigned to this phase , @xmath12 is the area of the @xmath9@xmath13 interface , @xmath14 is the corresponding interfacial tension , and @xmath15 is the length of the three - phase - contact line . the definition of @xmath4 in eq .
( [ eta ] ) refers to a reference state in which uniform bulk phases are extrapolated right up to the interfaces , and analogously the interface or surface properties of laterally homogeneous interfaces or surfaces are extrapolated right up to the contact line @xmath3 . in eq .
( [ eta ] ) we do not take into account contributions to @xmath5 related to the presence of walls enclosing the whole system ; the only inhomogeneities of the system which are relevant in the present analysis are those related to the spatial coexistence of the three thermodynamic phases @xmath0 , @xmath1 , and @xmath2 .
the above decomposition allows one to calculate the line tension for a given thermodynamic system provided the quantities on the lhs of eq .
( [ eta ] ) are determined in separate preceding steps involving similar considerations of suitable thermodynamic limits .
a different typ of an inhomogeneous fluid system is a sessile drop on a solid substrate . in this case
only two thermodynamic phases coexist whereas the substrate acts as an inert spectator phase . in such cases
the three - phase - contact line corresponds to the region where the interface between two coexisting thermodynamic phases meets the substrate @xcite . in the literature the term line energy or line elasticity
is also frequently used in connection with an extra surface energy associated with the deformation of the three - phase - contact line as the result , e.g. , of interactions with surface defects ( see , e.g. , refs .
this quantity has to be clearly distinguished from the line tension as introduced via eq .
( [ eta ] ) . in the literature
the notion of line tension is also used to describe the one - dimensional interface of two coexisting , intrinsically two - dimensional phases such as liquid- and vapor - like phases in langmuir blodget films .
this , however , corresponds to the lower - dimensional version of an interfacial tension and not to the coexistence of three phases as considered here . + there is a growing body of literature describing both experimental and theoretical investigations of the line tension and already a number of reviews
have been published on that subject @xcite .
experimental investigations were carried out for drops on solid substrates @xcite , for liquid lenses at the interface of two fluids @xcite , and for spherical particles or bubbles at the interface of two fluids @xcite ; also the role of line tension in epitaxial growth has been discussed @xcite . on the theoretical side
the theory of capillarity has been extended in a phenomenological way by taking into account line contributions and the consequences of these extensions have been explored @xcite .
furthermore there are microscopic calculations of the line tension @xcite and many studies , in most cases based on microscopic theories as well , concentrate on the behavior of the line tension in the vicinity of a wetting transition @xcite .
theoretical studies are also devoted to the examination of the influence of the droplet size on the line tension @xcite and of the electrostatic contributions to the line tension which arise in the presence of surface charges @xcite .
the boundary tension between two coexisting wetting films of different thicknesses was studied as well ( see , e.g. , refs .
a drop - size dependence of the contact angle was studied in molecular dynamics simulations without deducing values for the line tension ( see , e.g. , ref .
finally , the line tension was studied also via molecular dynamics simulations ( see , e.g. , refs .
@xcite ) or via an analysis of probability distributions as obtained from monte carlo simulations at three - phase contact @xcite .
a closer inspection of these results reveals that certain aspects of the line tension either give rise to conflicting statements or remain unaddressed , leading us to set out to clarify the following basic questions : * in which sense is the concept of the line tension well defined ?
* is it sufficient for the determination of equilibrium shapes to characterize the thermodynamics of the contact line by a line tension only ?
these questions arise because it is not obvious that the definition of @xmath4 via eq .
( [ eta ] ) leads to a unique result .
the reason lies in the arbitrariness in the definition of the position of the interfaces between adjacent phases .
this arbitrariness is due to the smooth spatial transition between the adjacent phases .
once the density distributions of the fluid components across the interfacial region are known , for example on the basis of scattering experiments or atomic force microscopy @xcite , or theoretically by simulations or density - functional calculations , a criterion has to be applied which fixes the interface position somewhere in the transition region .
however , there exists a multitude of sensible choices .
the arbitrariness in the definition of the interface positions leads to an arbitrariness in the definitions of the volumes , areas , lengths , and to some extent even interfacial tensions ( see , e.g. , refs .
although it was shown by widom @xcite that @xmath4 is independent of the particular choice of the dividing interfaces in the case of a straight three - phase - contact line in a system with three thermodynamically coexisting fluid phases separated by planar interfaces , it is not clear that the same will be true in other cases as well , e.g. , if two fluid phases are in contact with an inert solid phase , or for systems in which the curvatures of the interfaces and of the contact line plays a role . a first discussion , which naturally raises the issues , concerning the uniqueness of the line tension for the case in which two fluid phases are in contact with and meet at an inert solid phase , is given in sect .
this discussion , however , turns out to be incomplete and it leads to a contradiction with one of the results obtained in sect . 5 , which states that the line tension _ as defined there _ is unique . a thorough discussion and the resolution of the contradiction is given in sect .
6 . in sect .
5 we mainly investigate the three - phase - contact line in systems with curved interfaces . on the other hand , just these systems
are studied experimentally if one attempts to determine the value of the line tension , e.g. , via measuring the dependence of the contact angle on droplet size ( see , e.g. , refs .
@xcite ) .
the curvature dependence of interfacial tensions leads to contributions to the free energy which scale with the linear extension of the system , i.e. , very much like the line contribution . since in many cases interface curvature and the length of the three - phase - contact line
can not be varied independently , these two contributions are not distinguishable per se .
thus , line tension and curvature effects on the interfacial tensions have to be discussed simultaneously .
moreover , the curvature expansion of the surface tension has to be known in advance before the line tension may be determined from eq .
( [ eta ] ) .
+ since in the present context it is unavoidable to consider curvature effects on the interfacial tensions and because it is known that a consistent description of curved ( spherical ) interfaces must take into account for general dividing interfaces derivatives of the surface tensions with respect to their radii of curvature , i.e. , their bending rigidities ( see , e.g. , refs .
@xcite ) , we expect that a complete thermodynamic description of three - phase - contact lines requires the introduction of further material parameters in addition to the line tension .
indeed , a number of theoretical analyses have appeared in the literature in which additional material parameters were introduced in order to characterize the three - phase - contact line @xcite .
recent studies have been carried out by rusanov et al .
@xcite for a drop on a solid substrate , and by widom and coworkers in connection with the line analogue of the gibbs adsorption equation @xcite for a straight contact line at a genuine three - phase contact . in sect .
5 we systematically study these issues guided by the concept of form - invariance of the basic equations under so - called notional shifts of dividing interfaces ( for a definition see , c.f . , sects . 2 and 5 ) .
as a prerequisite to a thorough study of these issues we also have to answer similar questions related to the concept of surface tensions .
however , before discussing them and in particular the main questions concerning the line tension , we point at related issues recurring in the literature which can be judged only after the basic questions raised above have found a satisfactory answer .
these issues are : * what is the typical magnitude of the line tension ?
+ of course , the line tension depends on the thermodynamic state of the system and the number of relevant thermodynamic degrees of freedom varies from system to system .
for example , for certain systems the temperature dependence of the line tension becomes especially pronounced close to wetting transitions @xcite .
on the other hand , for many systems away from such special thermodynamic states there is now widespread agreement that the value of the line tension is of the order of @xmath16 n @xcite . with the experimental accuracy available at present the corresponding prefactor distinguishes between different systems . *
what is the sign of the line tension ?
+ various experimental and theoretical investigations lead to values of the line tension which include both signs @xcite .
since - contrary to the interfacial tension - there exists no thermodynamic argument for a specific sign of the line tension @xcite , so that experimental findings for the line tension with a certain sign can not be discarded from the outset .
( even in this respect a contrary statement may be found in the literature @xcite . ) on the other hand , if one is interested in the temperature dependence of the line tension for a specific system which undergoes a first - order wetting transition , both theoretical and experimental results show that upon increasing the temperature towards the wetting transition temperature the line tension changes sign from negative to positive values @xcite .
* is a negative line tension compatible with the structural stability of drops ? + the mesoscopic analysis of line tensions including small wavelength fluctuations confirms that negative values of the line tension do not lead to instabilities @xcite .
* what does the line tension depend on ?
+ similarly to the interfacial tension it is a function of the thermodynamic state of the system . for straight contact lines , defined by intersecting planar interfaces
, the phases involved must be at stable thermodynamic coexistence .
for example , one - component fluids in contact with an inert substrate must be at liquid
vapor coexistence @xmath17 , where @xmath18 denotes the chemical potential .
thus in this case the line tension is a function of temperature only .
if the fluid in contact with a substrate consists of a binary mixture of a and b particles , the fluid must be at fluid
fluid coexistence , i.e. , @xmath19 which leaves two thermodynamic variables free . for three - phase contact among three fluid phases
the system must be at the triple line ( @xmath20 ) of a - rich liquid , b - rich liquid , and vapor coexistence so that in this case the line tension is again a function of temperature only .
this dependence can be reparametrized in terms of the temperature dependent contact angle @xmath21 .
the situation is somewhat different for curved contact lines defined by intersecting curved interfaces .
droplets of finite size residing on a substrate or liquid lenses formed at planar fluid fluid interfaces are examples in which the contact lines are curved . due to
their curvature droplets or lenses remain in a ( constrained ) equilibrium with their surrounding phases , which takes place at chemical potentials slightly off their values @xmath22 at stable thermodynamic coexistence with planar interfaces . under these conditions
the pressure is different inside and outside the droplet or lens .
this pressure difference and at the same time the size of the droplet or lens are prescribed by the chosen chemical potentials or alternatively by the temperature and the ambient pressure which deviates from that for stable coexistence at the same temperature .
thus in principle the line tension may now depend on a further thermodynamic variable in addition to @xmath23 .
this dependence can be reparametrized in terms of the size of the droplet or lense .
+ in addition , from the theoretical point of view the line tension @xmath4 is a functional of both the interaction potentials between the fluid particles and the substrate potential .
if the microscopic forces are too long - ranged @xmath4 becomes ill defined while the corresponding interfacial and surface tensions retain their validity @xcite , i.e. , in this case the size dependence of @xmath5 can not be described in terms of a bulk , surface , and line contribution with a line tension which is size- and shape - independent in the thermodynamic limit . +
it appears that in extracting line tensions from experimental data so far the curvature dependence of the interfacial tension , characterized by the tolman length @xcite , has been completely neglected .
therefore the question arises to which extent the experimental determination of the line tension is affected by the tolman length . *
can one measure the line tension ?
+ in the last chapter we shall point at possible difficulties to determine line tensions uniquely .
in the following we discuss a possible source of ambiguity in the definition of the line tension ( see eq . ) in the case that one of the phases considered in sec . 1 is an inert substrate , i.e. , we consider two - phase coexistence in the presence of an inert wall rather than coexistence of three genuine thermodynamic phases .
we focus on the issue of non - uniqueness of the liquid substrate or gas substrate interfacial tensions . to this end
we first briefly mention the related question of the thermodynamic determination of the interfacial tension @xmath14 in a system in which two coexisting _ fluid phases _ , say @xmath9 and @xmath13 , meet along a planar interface @xcite . using the formula
@xmath24 the question appears whether the corresponding value of the interfacial tension depends on the arbitrary choice of the position of the fluid
fluid interface and the corresponding volumes @xmath11 and @xmath25 . in the case of a planar interface the value of the interfacial tension
does not depend on the position of the interface , i.e. , eq .
( [ sigma ] ) leads to a unique result .
the reason is that the total volume @xmath26 as well as the surface area @xmath27 are independent of the location of the dividing surface ; in addition due to thermal equilibrium one has @xmath28 .
( for a complete discussion of this issue see , e.g. , refs .
@xcite . ) + however , in situations in which one of the phases , say phase @xmath2 , is an inert substrate with a planar surface , a difference to the case of fluid fluid interfaces arises because the inert substrate is not one of the thermodynamically coexisting phases .
although it is not a priori obvious whether one should treat the substrate as a part of the system or not , it is usually considered as an external object which just provides a steep external potential defining the boundaries of the system .
+ it might appear that the freedom in positioning the dividing interface is less obvious for the solid fluid interface than for the fluid
fluid interface because a solid surface is defined rather sharply , say by the positions of the nuclei of the atoms forming the topmost layer .
however , even then the position at which an actual experimental technique ( e.g. , atomic force microscopy , optical methods , etc . ) will locate the surface of the solid will certainly deviate from the definition given above , because of the smooth decay of the substrate potential and of the finite extension of the electron cloud of the substrate , and because the fluid phase in contact with the solid only gradually attains its bulk properties . depending on the kind of experiment and on the way the data are analyzed , relative shifts in the location of the dividing surface which are of the order of one or even several atomic radii are conceivable . this shift multiplied by the surface tension under consideration
yields a force which is comparable with the magnitude of the line tension .
moreover , a solid substrate in contact with a vapor phase might be covered with a thin liquid - like wetting film which is in thermal equilibrium with the bulk vapor phase . in a thermodynamic description in terms of interfacial and line tensions
as considered here , this thin wetting film is not treated as a separate entity but as part of the solid vapor interface and as such it contributes to the actual solid vapor ( _ g_as ) surface tension @xmath29 .
accordingly the question arises , where to place the solid vapor interface
. it could be somewhere in the transition region from the liquid - like film to the vapor or at the transition region from the solid to the liquid - like film .
it is very likely that different experimental techniques imply different conventions .
( if nonequilibrium situations are considered @xmath29 may be different from its equilibrium value if the aforementioned thin wetting film present at thermal equilibrium has not yet formed . ) in order to explore the consequences of the freedom in choosing the dividing surface , we consider a system in which a planar inert substrate is exposed to the fluid @xmath30 , i.e. , the liquid or gas phase ( see fig .
1 ) . a fluid phase ( f ) in the presence of a planar substrate ( s ) .
two choices for the position of the planar substrate fluid interface are marked as ( 1 ) and ( 2 ) ; they are shifted with respect to one another by a distance @xmath31 . in each case the area @xmath32 of the substrate
fluid interface is the same . ]
if we consider the solid as a part of our system the value of the substrate fluid surface tension @xmath33 follows from the equation @xmath34 where @xmath5 denotes the grand canonical potential of the fluid plus that of the substrate , @xmath35 is the value of the fluid volume compatible with the chosen location of the planar substrate
fluid interface , @xmath36 is the corresponding volume of the solid , and @xmath37 and @xmath38 are the grand canonical free energy densities of the fluid and solid , respectively .
@xmath39 is the area of the planar solid fluid interface .
if the substrate is in a constrained equilibrium ( e.g. , no interdiffusion of solid and fluid particles ) , one has @xmath40 .
+ since the position of the substrate
fluid interface is not unique we can consider another position of this interface which is parallel to the first one and located a distance @xmath31 above it ( see fig.1 ) . in this case
@xmath41 , @xmath42 , whereas the surface area @xmath39 does not change upon the vertical shift of the interface .
it follows from eq .
( [ sf ] ) that the values of the substrate
fluid surface tensions corresponding to these two choices differ by @xmath43 thus a different choice of the location of the substrate
fluid interface corresponds to a redistribution of the free energy between the bulk and the surface terms , and to different values of the substrate
fluid surface tension .
note that in the case of _ thermodynamically coexisting _ liquid and gas phases the grand canonical free energy densities of the two phases ( i.e. , the negative pressures ) are equal ( @xmath44 ) and thus the well known conclusion @xmath45 follows ( see refs .
at the same time we emphasize that the difference @xmath46 ( which , e.g. , enters into young s law for the contact angle ( see , c.f . , eq . ( [ young ] ) ) does not depend on this choice of the dividing surface provided the dividing interfaces between solid and gas on one hand and between solid and liquid on the other hand are chosen to be at the same height above the solid , and provided the solid is in the same state in both cases .
in the case of three genuine thermodynamically coexisting phases the issue analogous to the one raised in sect . 2 is whether the value of the line tension @xmath4 determined from eq .
( [ eta ] ) depends on the location of the dividing interfaces which in turn determine the location of the contact line .
one can show @xcite that in the case that these phases meet along a _ straight line _
parallel shifts of this line do not influence the value of the line tension .
+ in order to study whether the location of the substrate fluid interface affects the value of @xmath4 , we investigate the following configuration of two - phase coexistence in the presence of a planar solid substrate . due to thermal equilibrium of the gas and liquid phase
one can impose lateral boundary conditions such that far to the left the substrate is exposed to the gas phase whereas far to the right the substrate is exposed to the liquid .
this enforces the formation of the liquid gas interface which meets the substrate with a contact angle @xmath47 ( see fig .
2 ) . on a macroscopic scales the liquid gas interface is also planar .
coexisting gas ( g ) and liquid ( l ) phases in the presence of a planar substrate ( s ) .
the planar liquid
gas interface meets the substrate with a contact angle @xmath47 .
far to the left ( right ) there is the substrate
gas ( liquid ) interface .
two choices of the position of the substrate
fluid interfaces are marked by ( 1 ) and ( 2 ) ; their distance is denoted by @xmath31 .
this results in two parallel contact lines the cross sections of which are indicated by the dots . @xmath48 and @xmath49 denote the corresponding changes in the liquid
gas and substrate liquid interfacial areas , respectively . for both choices
the contact angle @xmath47 is the same . ]
+ we again consider two parallel positions of the substrate
fluid interface at a distance @xmath31 from each other .
this results in two corresponding , parallel contact lines . from simple geometrical considerations
it follows that the values of the line tensions corresponding to these choices differ by @xmath50 in the calculation leading to eq .
( [ deltaeta1 ] ) we have used the aforementioned result @xmath51 and the fact that the freedom in the choice of the substrate fluid interface position does not influence the value of the liquid gas interfacial tension @xmath52 , i.e. , @xmath53 . after using young s equation @xcite @xmath54 and identifying @xmath55 one obtains @xmath56 this result reflects the fact that there is no explicit force balance perpendicular to the solid surface . taken at face value the above result would show that the freedom in positioning the substrate
fluid interfaces with the consequential shift of the contact line is reflected in the change of the value of the line tension .
this change is proportional to the distance @xmath31 between the two arbitrarily selected positions of the substrate
fluid interfaces .
numerically , the rhs of eq .
( [ deltaeta2 ] ) is comparable with @xmath4 .
however , eq . ( [ deltaeta2 ] ) is in conflict with a result which will be obtained in sect .
v of the present work , namely the invariance of the line tension with respect to notional changes of the system .
this puzzle will be resolved in sect .
we only note that it turns out that the two conflicting statements are based on two different definitions of a line tension both of which seem to be absolutely compelling from the point of view of how they are introduced . a relation between the two definitions of @xmath4 will be given .
but we also point out that the way how eq .
( [ deltaeta2 ] ) has been obtained above has to be scrutinized carefully , because , while deriving eq .
( [ deltaeta2 ] ) , all contributions to @xmath5 which are generated by separating a subsystem from its surrounding and which are proportional to the linear extension of the system , have been disregarded .
we shall show that in general this is not permissible .
we start this section by considering genuine three - phase coexistence in which a lense consisting of a fluid phase @xmath1 is located at the interface between two fluid phases @xmath0 and @xmath2 ( see fig .
3(a ) ) .
( a ) a liquid lens ( phase @xmath1 ) at the @xmath0@xmath2 interface between two fluid phases @xmath0 and @xmath2 ; @xmath57 is the radius of the circular intersection between the two spherical caps forming the lens .
the various contact angles are denoted also as @xmath0 , @xmath1 , and @xmath2 .
( b ) a sessile liquid drop ( phase @xmath1 ) in contact with its vapor ( phase @xmath0 ) and a planar substrate ( @xmath2 ) forming a contact angle @xmath47 .
the drop is a spherical cap . ]
the lense is taken to be formed by two spherical caps of different radii intersecting along a circle of radius @xmath57 .
the three - phase - contact line of circular shape is accompanied by the line tension @xmath4 . to simplify the notation
also the corresponding contact angles are denoted by the same symbols as the phases , i.e. , @xmath0 , @xmath1 , and @xmath2 , where @xmath58 . in the absence of the line tension the contact angles @xmath59 , @xmath60 , and @xmath61 fulfill the equation ( see , e.g. , ref .
@xcite ) @xmath62 if a line - tension contribution to the constrained grand canonical free energy @xmath63 is included one obtains , from the minimization of @xmath63 at a constant volume of the liquid phase @xmath1 , the modified neumann equation ( see , e.g. , refs .
@xcite ) @xmath64 provided neither bending rigidities of the interfaces nor further properties attributed to the contact line other than the line tension ( such as rigidities against changes of contact angles ) are taken into account .
equation ( [ mye1 ] ) can be equivalently rewritten as @xmath65 in particular , if one of the phases , say phase @xmath2 , is taken to represent an inert substrate with planar surface ( @xmath66 ) ( see fig .
3(b ) ) , the above equation turns into the modified young equation ( see , e.g. , refs .
@xcite ) @xmath67 where @xmath57 denotes the radius of the circular substrate
phase-@xmath0phase-@xmath1 contact line .
equations ( [ mye10 ] ) and ( [ mye2 ] ) represent asymptotic formulae valid in the limiting case of large lenses or drops , i.e. , @xmath4 governs the _ leading _ behavior for @xmath68 .
+ we note that in many experiments involving sessile liquid drops a so - called line tension @xmath4 is deduced from measurements of the contact angle @xmath47 as a function of the radius @xmath57 via fitting the modified young equation ( eq .
( [ mye2 ] ) ) to the data @xcite .
similar experiments have been carried out with lens - like objects @xcite .
+ however , a closer look at the procedure described above of determining @xmath4 reveals problems in using the modified young ( eq . ( [ mye2 ] ) ) or neumann equations ( eqs .
( [ mye1 ] , [ mye10 ] ) ) which are related to the freedom in positioning the dividing interfaces . a shift of the solid liquid dividing interface ( in the case of a sessile drop ) by @xmath31 or a change of the radius of a spherical interface ( for the lens or the drop ) by @xmath69 leads to changes of the contact angles as well . from simple geometrical considerations
one finds that the corresponding changes in @xmath70 or @xmath71 are of the order @xmath72 or @xmath73 , i.e. , they are of the same order as the corrections stemming from the presence of the line tension . therefore , upon applying the modified neumann or young equation to the same physical object but with different choices for the dividing interfaces one would have to introduce two different and suitably chosen values of @xmath4 for different dividing interfaces in order to obtain the correct relations between the two corresponding contact angles .
( in the line - tension related correction terms in eqs .
( [ mye10 ] , [ mye2 ] ) the quantities other than @xmath4 are either independent of the choice of dividing interfaces or their changes with the dividing interfaces give rise to higher order corrections . ) on the other hand , by decomposing @xmath5 in two different ways we had found that for a straight contact line at a genuine three - phase contact the line tension is independent of the choice of dividing interfaces .
this contradicts the result of the previous argument .
finally , we can also look at what happens if we shift the substrate
fluid interface for the sessile drop . if we compute the difference @xmath74 enforced by the geometrical relations between @xmath75 and @xmath76 together with the requirement that both @xmath75 and @xmath76 fulfil the modified young equation we obtain @xmath77 which has the same structure as the value given in eq .
( [ deltaeta2 ] ) ( obtained from a comparison of two different ways of decomposing @xmath5 ) , but it has the opposite sign . + summarizing these two findings it turns out that the relations between two line tension values obtained from two different decompositions of @xmath5 for two different sets of dividing interfaces are at variance with the currently used modified young and neumann equations combined with elementary geometrical considerations .
+ in order to find equations free from the above inconsistencies we shall investigate in detail two representative systems : a liquid lens at the interface between two fluid phases and a sessile drop on a substrate .
we shall include from the outset the effects of curvature on the interfacial tensions and we shall explicitly state all conventions used in defining different sets of dividing interfaces , and list all the properties which are assigned to the interfaces and to the reference bulk phases .
in this section we first study a lens - shaped fluid phase @xmath1 located at the planar interface between two other fluid phases @xmath0 and @xmath2 , and secondly a sessile liquid drop of @xmath1 phase in contact with its vapor @xmath0 on top of an undeformable inert solid substrate @xmath2 . in this context two main questions are addressed .
first , how does the line tension depend on parallel ` displacements of the positions of the dividing interfaces between two phases ? secondly , what can be learned from the fact that thermodynamic potentials of the total system must be independent of arbitrary conventions regarding the choice of dividing interfaces ?
in particular , we are interested in the resulting requirements with respect to the structure of the equations relating the contact angle(s ) to the system size , e.g. , to the radius @xmath57 of the three - phase - contact line .
these questions are motivated by the inconsistencies encountered for the modified young and neumann equations in the previous chapter .
+ first , we specify the systems in more detail and give the rules that we have chosen to define the dividing interfaces .
these rules still admit parallel shifts of the dividing interfaces .
we also introduce the quantities that are used in order to describe the shifted dividing interfaces .
it is further necessary to fix the properties attributed to the ( reference ) bulk and the interfaces present in the lens or drop , and we do this in accordance with what is known about the bulk fluid and the interfacial properties of spherical drops . only after this step
is completed the properties attributed to the line may be extracted .
+ the lens and the drop are completely characterized once the density distributions are known for all constituent species at given thermodynamic conditions . only for small drops
the line tension and curvature effects are expected to become relevant , and in this case gravity can be neglected .
( the validity of young s law in the presence of gravity is discussed in ref .
we also assume that no other external bulk forces act on the systems . under these conditions , provided the volume of the @xmath1 phase is sufficiently large , one can expect that there exist regions of the lens ( at some distance away from the three - phase contact lines ) where the density distributions exhibit to a good approximation radial symmetry relative to one of the two centers of curvature @xmath78 and @xmath79 ; an additional center ' at infinity characterizing the planar interface is present in the case of a lens ( see fig .
4 ) . a liquid lens at the planar @xmath0@xmath2 interface ( horizontal lines ) .
two choices for the interfaces are marked by ( 1 ) and ( 2 ) .
the relative concentric shifts of the spherical interfaces are characterized by @xmath80 $ ] and @xmath81 $ ] .
the planar @xmath0@xmath2 interface is shifted together with the spherical ones , as indicated by the dashed and solid horizontal lines and the change in the amount of @xmath2 phase shown in lighter gray .
the corresponding contact angles are denoted by @xmath82 , @xmath83 , @xmath84 and @xmath85 , @xmath86 , @xmath87 ; the contact - line radii are @xmath88 and @xmath89 . m@xmath90 and m@xmath91 are the centers for the radii of curvature @xmath92 , @xmath93 $ ] and @xmath94 , @xmath95 $ ] , respectively . ] in the case of the drop we correspondingly expect radial symmetries around one center @xmath96 plus an additional center at infinity characterizing the substrate fluid interface ( see fig . 5 ) . for the lens
, the center @xmath78 characterizes the density distributions at the @xmath0@xmath1 interface , the center @xmath79 characterizes those at the @xmath1@xmath2 interface . a sessile liquid drop on a planar substrate .
two choices for the liquid vapor ( @xmath1@xmath0 ) , the substrate
liquid ( @xmath2@xmath1 ) , and substrate vapor ( @xmath2@xmath0 ) interfaces are denoted as ( 1 ) and ( 2 ) ; the corresponding concentric shifts are characterized by @xmath97 $ ] and @xmath98 $ ] .
the corresponding contact angles are denoted by @xmath99 and @xmath100 , and @xmath88 and @xmath89 are the corresponding contact - line radii .
m is the center of the radii of curvature . ] in view of these radial symmetries it makes only sense to consider concentric shifts of the interfaces with respect to the fixed centers .
the two phases @xmath0 and @xmath2 are assumed to be separated by a planar interface .
if the @xmath0@xmath1 and @xmath1@xmath2 interfacial structures do not overlap except near the three - phase region , the interior of the lens is occupied by an almost homogeneous @xmath1 phase .
this homogeneous @xmath1 phase interpolates smoothly between the radially symmetric density distributions around the two centers associated with the two interfaces .
similar considerations apply to the drop except that the @xmath1@xmath2 interface is planar which can be regarded as a limiting case of a spherical interface with infinite radius of curvature .
+ from the previous remarks it follows that to a large extent the isodensity contours of the @xmath0@xmath1 and @xmath1@xmath2 interfaces are segments of spherical surfaces . in order to define gibbs dividing interfaces separating two adjoining phases we use the spherical parts of the isodensity contours in a two - phase region and extrapolate them into the three - phase - contact region where surfaces of constant densities actually are no longer spherical . which of the infinite number of surfaces of constant densities is chosen in order to construct a gibbs dividing interface is of course a matter of convention .
+ for the lens , once the centers @xmath78 , @xmath79 are given and the radii @xmath101 , @xmath102 are chosen according to a certain convention we can define a three - phase - contact line of circular shape by the intersection of the two spheres ( @xmath78,@xmath101 ) and ( @xmath79,@xmath102 ) as indicated in fig .
the third , planar gibbs dividing interface between the phases @xmath0 and @xmath2 is placed then in such a way that it coincides with the plane determined by the previously defined circular three - phase - contact line . again
this choice is a mere convention but deviating from it would create three different lines of intersection between three pairs of interfaces ( @xmath0@xmath1 intersecting with @xmath1@xmath2 , @xmath0@xmath1 with @xmath0@xmath2 , and @xmath0@xmath2 with @xmath1@xmath2 ) . furthermore , for such a deviating choice a ring - shaped volume , with triangular cross section defined by these three lines and the connecting interfaces , could not be assigned to any of the three phases and would have to be treated separately as a new line phase .
this would create an unrewarding complication because the central idea of introducing the mathematical gibbs dividing surface is to allow for a natural extrapolation of the properties of the adjacent bulk - like phases from both sides right to the dividing surface were they are assumed to change discontinously . within this approach
there is no room for additional phases .
a consistent description of the actual system follows by defining appropriate surface excess quantities and their densities , like the surface tension .
we stick to gibbs idea and we use the convention described above within which all three lines of intersections coincide , and no volume filled with a phase of an unassigned character is left over .
+ the procedure for the drop is analogous .
the sphere ( @xmath103 ) defines the @xmath0@xmath1 gibbs dividing surface .
again , the spherical part of the isodensity contour is extrapolated into the three - phase - contact region . in the same spirit as above the @xmath0@xmath2 and the @xmath1@xmath2 gibbs dividing interfaces
are chosen to lie in the same plane .
the three - phase - contact line is defined by the intersection of the @xmath0@xmath1 gibbs dividing surface with the common @xmath0@xmath2 and @xmath1@xmath2 plane ( see fig . 5 ) . using this convention
has the advantage of being in agreement with the one chosen for the lens in the limit of one of the curvature radii becoming infinite .
furthermore , this construction creates only one line of intersection instead of two . + once the dividing interfaces are defined the total volume @xmath104 is subdivided into domains assigned properly to the phases @xmath0 , @xmath1 , and @xmath2 .
the grand canonical potential of the system is decomposed into the corresponding bulk , interfacial , and line contributions . for the lens one has @xmath105 where @xmath106 is the pressure in the bulk - like phase @xmath9 (= @xmath0 ,
@xmath1 , or @xmath2 ) and @xmath11 is the corresponding volume .
@xmath107 is the area of the planar @xmath0@xmath2 interface in the absence of the lens .
the values of @xmath11 depend on the way the dividing interfaces are chosen .
the total volume @xmath108 of the system is independent of this choice , and independent of the physical size of the lens or the drop .
+ for the drop one has @xmath109 where @xmath110 , @xmath111 (= @xmath112 for the drop ) are the areas of the @xmath0@xmath113 and the @xmath1@xmath2 interfaces , respectively .
@xmath107 is the area of the planar @xmath0@xmath2 interface ( in the absence of the drop ) and it is independent of the choice of the dividing interfaces . the area @xmath112 denotes the @xmath0@xmath2 interfacial area replaced by the drop .
the radius of the circular three - phase - contact line is denoted by @xmath57 and @xmath114 , @xmath115 , and @xmath116 denote the surface tensions of the @xmath0@xmath113 , @xmath1@xmath2 , and @xmath0@xmath2 interfaces , respectively .
finally , @xmath4 is the line tension , i.e. , the excess free energy per unit length of the three - phase - contact line we are interested in . strictly speaking the line tension defined via eqs .
( [ lensdrop1 ] ) and ( [ lensdrop1b ] ) is not identical to its definition via eq .
( [ eta ] ) , because in eqs .
( [ lensdrop1 ] ) and ( [ lensdrop1b ] ) we first keep the subleading terms and treat them as subleading contributions to @xmath4 , whereas in eq .
( [ eta ] ) we isolate and drop these terms from the outset .
however , if in the following we speak about the line tension itself , as deduced from the decomposition of @xmath5 via eqs .
( [ lensdrop1 ] ) and ( [ lensdrop1b ] ) , we shall always drop the subleading terms . in this sense the definitions in eqs .
( [ lensdrop1 ] ) and ( [ lensdrop1b ] ) and that via eq .
( [ eta ] ) do agree and we therefore do not introduce a different notation to distinguish between these two expressions of the line tension . for the lens the pressures in the @xmath0 and @xmath2 phases are equal , @xmath117 because the coexisting phases @xmath0 and @xmath2 are separated by a planar interface . for the drop
the role of @xmath118 is played by @xmath119 , i.e. , the grand canonical free energy density of the solid while for the phase @xmath0 we set @xmath120 .
the pressure of the @xmath1 phase deviates from @xmath121 and is written as ( see , c.f . , the discussion following eq .
( [ lensdrop4b ] ) ) @xmath122 + the expression for the grand potential @xmath5 can be regrouped as @xmath123 where @xmath124 for the lens , and @xmath125 for the drop .
consequently one obtains @xmath126 for the lens , and @xmath127 for the drop .
+ the equations given above hold for the two following scenarios . in the first scenario , the lens or drop can exchange matter with the surrounding phase so that the chemical potentials in all phases are equal and the pressure in the @xmath1 phase is determined by the chemical potentials . in the second scenario , nonvolatile lenses or drops are investigated and the volume of the lens or drop is prescribed . in this case the pressure in the @xmath1 phase is not an independent thermodynamic variable but is determined by the amount of @xmath1 phase . the term @xmath128 ( eq . ( [ lensdrop3 ] ) )
is independent of the lens size and of the choice of dividing interfaces .
this holds also for the drop ( eq .
[ lensdrop3b ] ) ) due to the transformation law given in eq .
( [ deltasigmasf ] ) which is valid if the @xmath0@xmath2 and @xmath1@xmath2 interfaces form a common plane , as is the case for the convention we have chosen . since @xmath5 as a physical quantity and @xmath128 , as just argued , are independent of choices of dividing interfaces , also @xmath129 is independent of such choices .
thus in the following we can focus on @xmath129 . in order to extract the value of @xmath4 from a known , e.g. , calculated @xmath129 , or to infer information about possible changes of @xmath4 upon shifts of the dividing interfaces , at first
it is necessary to specify all other quantities appearing in eqs .
( [ lensdrop4 ] ) and ( [ lensdrop4b ] ) .
the geometrical quantities like volumes , areas , and lengths are defined once the dividing interfaces are chosen .
the surface tension @xmath116 is that of a planar interface between phases @xmath0 and @xmath2 .
if both phases @xmath0 and @xmath2 are fluid it is known ( see , e.g. , ref .
@xcite ) that @xmath116 is independent of the position of the dividing surface between these phases .
if one of the phases , e.g. , @xmath2 is an inert solid phase , @xmath116 must be analyzed more closely ; but we postpone this discussion to a later subsection ( see , c.f . , subsec .
5.3.1 ) in which sessile drops are discussed separately . the quantities @xmath130 and @xmath115 are surface tensions of curved surfaces again the case of a drop requires a separate discussion and thus they depend on the corresponding radii of curvature .
we indicate this dependence @xmath131 , @xmath132 explicitly whenever it is necessary in order to avoid confusion .
as already discussed above we postulate that the interfacial tensions appearing in eqs .
( [ lensdrop4 ] ) and ( [ lensdrop4b ] ) have the same properties as their spherically closed counterparts , such as @xmath133 were @xmath134 is the tolmann length for the @xmath0@xmath135 interface ; subleading terms in @xmath136 are not considered here .
an analogous equation holds for @xmath132 .
although in the following these expressions will not be be used explicitly it will be always understood that surface tensions depend on the radius of curvature in such a way ( concerning the physical radius , see the corresponding remarks below ) .
a related property of the pressure difference @xmath137 is described by the generalized laplace equation @xcite @xmath138 \quad , \nonumber \\ & = & \frac{2\sigma_{\beta \gamma}(r_2)}{r_2 } + \left[\frac{\mathrm{d}\sigma_{\beta \gamma}}{\mathrm{d } r_{2 } } \right ] \quad , \label{lensdrop6}\end{aligned}\ ] ] where the terms in square brackets are quantities termed by rowlinson and widom @xcite notional derivatives .
such a term multiplied by a small [ d@xmath139 approximates the change in surface tension upon increasing the radius of the dividing surface by that value [ d@xmath140 without changing the physical system , i.e. , the density distributions and all thermodynamic variables remain fixed whereas the description of the system has been changed by shifting the dividing surface . in the following square brackets will be always used in order to characterize notional changes , e.g. , [ d@xmath140 denotes the notional change of the radius whereas we would write d@xmath141 if we speak about a physical change of the radius at a fixed convention of choosing the dividing surface ( interface ) . the quantity @xmath142 $ ] depends on the choice of the dividing surface .
the so - called surface of tension is defined as that dividing surface for which @xmath143 $ ] is zero .
for the equimolar dividing surface the notional derivative coincides with the derivative of the surface tension with respect to the physical drop size , i.e. , the derivative of eq .
( [ lensdrop5 ] ) . the second term on the rhs of eq . ( [ lensdrop6 ] )
renders the rhs invariant with respect to changes of the dividing surface .
this must be the case because @xmath137 defined as the pressure difference between two bulk phases at given thermodynamic conditions is a measurable and therefore invariant quantity .
below , notional derivatives will be characterized by putting them into square brackets . a property that will also be used extensively in the following is the fact , that @xmath144 ( @xmath145 ) is independent of the particular choice of the dividing surface up to and including the order @xmath146 . at some occasions
we shall use the relation @xmath147^{(2 ) } - \left[\frac{\mathrm{d}\sigma_{\xi \nu } } { \mathrm{d } r_i } \right]^{(1 ) } = \frac{2\sigma_{\xi \nu } \left[\mathrm{d}r_i\right]}{r_i^2 } + s.l.t.\ ] ] between notional derivatives of surface tensions for two differently chosen dividing interfaces .
the two dividing interfaces are denoted by the superscripts ( 1 ) and ( 2 ) and their radii are related via @xmath148^{(2 ) } = \left[r_i\right ] ^{(1 ) } + \left[\mathrm{d}r_i\right]$ ] ; @xmath149 $ ] could be also understood as a differential .
it could be also finite but then we assume that it is small compared with @xmath141 . in eq .
( [ lensdrop7 ] ) we write @xmath141 and do not introduce @xmath148^{(1)}$ ] or @xmath148^{(2)}$ ] because using one or the other convention would lead to expressions which only differ in subleading terms which are neglected anyway .
equation ( [ lensdrop7 ] ) is valid if the values of both @xmath148^{(1)}$ ] and @xmath148^{(2)}$ ] are close to that of the radius corresponding to the surface of tension and it is a direct consequence of well - known results obtained for spherical interfaces ( see , e.g. , refs .
@xcite ) . the neglected s.l.t . contain terms of the order @xmath150 or higher and in addition they contain @xmath149 $ ] or differences between @xmath148^{(i)}$ ] and the radius corresponding to the surface of tension raised to the second or higher power .
most of the following conclusions , however , do not make explicit use of eq .
( [ lensdrop7 ] ) , but only of the fact that @xmath151^{(2 ) } - \left[\frac{\mathrm{d}\sigma_{\xi \nu } } { \mathrm{d } r_i } \right]^{(1)}$ ] is of higher order than @xmath146 .
we also use ( in the second part of our reasoning ) the following relation known for spherical drops @xcite between the stiffness against changes of the radius of curvature at given thermodynamic conditions ( i.e. , fixed temperature and chemical potentials ) , denoted here as @xmath152 , and the notional derivative of the surface tension : @xmath153 \quad.\ ] ] in the following we discuss lenses and drops .
although the arguments proceed similarly in both cases , we discuss them separately because we want to address certain special problems connected exclusively with the solid fluid interfaces .
first , we investigate the consequences of the requirement that the grand canonical potential of the total system is independent of particularly chosen dividing interfaces .
secondly , we derive from a variational procedure at fixed lens volume the equations yielding the contact angles as a function of the radius @xmath57 characterizing the lens size .
the variational problem is set up in a way which guarantees that the contact angles transform correctly upon notional shifts of the gibbs dividing interfaces .
the results following from these two procedures are then compared .
+ notional variations ( indicated by square brackets ) leave the grand canonical potential unchanged which in differential form is written as @xmath154 = 0\ ] ] with @xmath129 defined in eq .
( [ lensdrop4 ] ) . the left - hand side of eq .
( [ lensdrop9 ] ) gives the notional change of @xmath129 upon notional shifts of the dividing interfaces away from some given but essentially arbitrary set of dividing interfaces , in the limit of very small notional shifts , characterized by , e.g. , @xmath155 $ ] ( @xmath156 ) ( see below ) . in the above equation all notional variations of geometrical quantities are then expressed in terms of the notional variations @xmath155 $ ] ( @xmath156 ) of the two radii of curvature @xmath101 and @xmath102 such as , for example , @xmath157 = \left [ \frac { \mathrm{d}v_{\beta } } { \mathrm{d } r_1 } \right ] \left [ \mathrm{d } r_1 \right ] + \left [ \frac { \mathrm{d}v_{\beta } } { \mathrm{d } r_2 } \right ] \left [ \mathrm{d } r_2 \right ] \quad.\ ] ] similar equations hold for the notional variations of interfacial areas @xmath158 $ ] , @xmath159 $ ] , @xmath160 $ ] , and for the length of the three - phase - contact line @xmath161 $ ] as well as for the surface and line tensions .
the notional derivatives @xmath162 $ ] and @xmath163 $ ] of the surface tensions are as already stated taken to be identical to those defined for completely spherical drops ( see eqs .
( [ lensdrop6 ] - [ lensdrop8 ] ) ) .
in addition we demand that @xmath164 = 0 $ ] and @xmath165 = 0 $ ] because a notional change of the radius @xmath101 of the @xmath0@xmath1 interface should not lead to notional changes of the @xmath1@xmath2 interfacial tension , i.e. , of the interface on the opposite side of the lens , and vice versa ( see fig .
we further introduce two notional derivatives of the line tension : @xmath166 $ ] and @xmath167 $ ] .
these are actually defined by eq .
( [ lensdrop9 ] ) because all other quantities in this equation are either given by geometry or fixed via the set of definitions given above .
certain properties of these new quantities will follow from the analysis given below .
+ the pressure difference @xmath137 is related to the surface tensions and their notional derivatives via eq .
( [ lensdrop6 ] ) . since @xmath137 is unique the right hand sides of these two equations in eq .
( [ lensdrop6 ] ) must be equal .
( the validity of the laplace equation for even small lenses has been checked by simulations , e.g. , in ref .
@xcite . ) after using the geometrical relations @xmath168 and @xmath169 we obtain a first equation relating @xmath0 , @xmath2 and @xmath57 : @xmath170 = \sigma_{\beta \gamma } \sin \gamma + \frac{r}{2 } \left [ \frac{\mathrm{d } \sigma_{\beta \gamma } } { \mathrm{d } r_2 } \right ] \quad.\ ] ] because the notional changes @xmath171 $ ] and @xmath172 $ ] are independent , the prefactors of @xmath173 $ ] and @xmath172 $ ] , obtained after expressing eq .
( [ lensdrop9 ] ) in terms of these variables , must both vanish . in this way two further equations
are obtained leading to three equations in total . on the other hand
only two equations are required in order to determine the contact angles for a given physical lens size , because the geometric parameters @xmath0 , @xmath1 , and @xmath57 are already related through a prescribed volume of the @xmath1 phase . therefore the two aforementioned equations following from eq .
( [ lensdrop9 ] ) have to be identical .
this leads to the following consistency condition : @xmath174 - \cos \alpha \left [ \frac{\mathrm{d } \tau } { \mathrm{d } r_1 } \right ] = \frac{r}{2 } \left\ { \left [ \frac{\mathrm{d } \sigma_{\beta \gamma}}{\mathrm{d } r_2 } \right ] - \left [ \frac{\mathrm{d } \sigma_{\alpha \beta } } { \mathrm{d } r_1 } \right ] \right\ } \ , .
\end{split}\ ] ] the remaining third equation can be written in a symmetrized form , @xmath175 - \left [ \frac{\mathrm{d } \sigma_{\alpha \beta } } { \mathrm{d } r_1 } \right ] \right\ } \\
\hspace*{-0.10 cm } + \frac { \left\ { \sin \gamma \cos \alpha + \sin \alpha \cos \gamma \right\ } } { 2 \cos \alpha \cos \gamma } \left\ { \cos \gamma \left [ \frac{\mathrm{d } \tau } { \mathrm{d } r_2 } \right ] + \cos \alpha \left [ \frac{\mathrm{d } \tau } { \mathrm{d } r_1 } \right ] \right\ } \hspace*{-0.10 cm } , \\ \end{gathered}\ ] ] or , using the consistency condition given in eq .
( [ lens2 ] ) , it can be put into the following form : @xmath176 + \sin \alpha \hspace*{-0.10 cm } \left [ \frac{\mathrm{d } \tau } { \mathrm{d } r_1 } \right ] \hspace*{-0.10cm}. \\\end{gathered}\ ] ] since the contact angles @xmath0 and @xmath177 are natural variables characterizing the three - phase - contact region we express notional derivatives of @xmath4 with respect to the radii in terms of notional derivatives with respect to the contact angles : @xmath178 = \left [ \frac{\mathrm{d } \tau } { \mathrm{d } \alpha } \right ] \left [ \frac{\mathrm{d } \alpha } { \mathrm{d } r_i } \right ] + \left [ \frac{\mathrm{d } \tau } { \mathrm{d } \gamma } \right ] \left [ \frac{\mathrm{d } \gamma } { \mathrm{d } r_i } \right]$ ] , where @xmath179 $ ] and @xmath180 $ ] describe notional changes of @xmath0 and @xmath2 , respectively , upon notional changes of @xmath141 . for given centers @xmath78 and @xmath79 , the descriptions in terms of the pairs ( @xmath0 , @xmath2 ) or ( @xmath101,@xmath102 ) are equivalent .
however , we do not transform the notional derivatives of the surface tensions , because the radii of curvature @xmath141 are the natural variables of the curved interfaces . in the new variables eqs .
( [ lens2 ] ) and ( [ lens4 ] ) take the form @xmath181 \sin ^2 \alpha - \left [ \frac{\mathrm{d } \tau } { \mathrm{d } \gamma } \right ] \sin ^2 \gamma = \frac{r^2}{2 } \left\ { \left [ \frac{\mathrm{d } \sigma_{\beta \gamma}}{\mathrm{d } r_2 } \right ] - \left [ \frac{\mathrm{d } \sigma_{\alpha \beta } } { \mathrm{d } r_1 } \right ] \right\ } \end{split}\ ] ] and @xmath182 \nonumber \\ + \frac{\sin \gamma \cos \gamma}{r } \left [ \frac{\mathrm{d } \tau } { \mathrm{d } \gamma } \right ] \quad . & & \label{lens6}\end{aligned}\ ] ] the next question is whether the line tension @xmath4 depends on the choice of the dividing interfaces . in order to answer it we compare @xmath129 ( eq . ( [ lensdrop4 ] ) ) evaluated for two different dividing interfaces and find _
@xmath4 to be independent of the choices of the dividing interfaces within the leading order . _
more precisely this statement says that notional shifts of the dividing interfaces change @xmath4 only by contributions which decrease as @xmath183 with increasing @xmath57 or faster .
( we recall our remarks after eqs .
( [ lensdrop1 ] ) and ( [ lensdrop1b ] ) saying that @xmath4 as defined by eqs .
( [ lensdrop1 ] ) and ( [ lensdrop1b ] ) may contain subleading terms . )
the _ leading _ contribution to @xmath4 which is independent of @xmath57 , is independent of the choices of the dividing interfaces . although the subleading contributions to @xmath4 itself may be neglected in the term @xmath184 in eq .
( [ lens4 ] ) if we want to keep only terms up to the order @xmath183 , it is not permissible to neglect the terms containing the notional derivatives @xmath178 $ ] .
if these derivatives are of the order @xmath183 in fact we shall see below that notional shifts of the dividing interfaces lead to changes in @xmath178 $ ] which are of that order the pertinent terms in eq .
( [ lens4 ] ) are of the order of @xmath183 , i.e. , they are of the same order as @xmath184 with only the leading term in @xmath4 kept . in proving the aforementioned results concerning the behavior of @xmath4 under notional shifts of the dividing interfaces , we have used the fact that surface tensions are independent of the chosen dividing interfaces up to order @xmath185 .
furthermore we have used the generalized laplace equation ( [ lensdrop6 ] ) and also the fact that the difference @xmath151^{(2 ) } - \left[\frac{\mathrm{d}\sigma_{\xi \nu } } { \mathrm{d } r_i } \right]^{(1)}$ ] is a correction of higher order than we are interested in . moreover
, the contact angles and other geometrical quantities for a choice ( 2 ) for the dividing interfaces have to be expressed in terms of the corresponding quantities characterizing choice ( 1 ) .
this has been carried out up to the order giving rise to contributions of the order of the line - tension contribution .
+ next we investigate the transformation behavior of the quantities @xmath186 $ ] or @xmath187 $ ] and @xmath188 $ ] due to notional shifts of the dividing interfaces . for this purpose
we consider the equations for the contact angles ( eqs .
( [ lens1 ] - [ lens6 ] ) ) for two different sets ( 1 ) and ( 2 ) of dividing interfaces .
the geometrical quantities , say @xmath189 , @xmath190 , and @xmath191 , in equations valid for the set ( 2 ) are then expressed in terms of @xmath192 , @xmath193 , and @xmath194 up to the required order . the same is done for the surface tensions and their notional derivatives .
we also use the above result that @xmath4 is independent of the choice of dividing interfaces to leading order . from comparison of the two systems of equations ( eqs .
( [ lens1 ] ) , ( [ lens2 ] ) , and ( [ lens4 ] ) ) and the fact that both systems must yield the same relations between @xmath192 , @xmath193 , and @xmath195 for a fixed physical system , by using eq .
( [ lensdrop7 ] ) we obtain the following transformation rules : @xmath196^{(2 ) } - \left [ \frac{\mathrm{d } \tau } { \mathrm{d } r_1 } \right]^{(1 ) } & = & - \frac{\sigma_{\alpha \beta } \sin \alpha \cos ( \alpha + \gamma)}{r \sin ( \alpha + \gamma ) } \left[\mathrm{d } r_1 \right ] - \frac{\sigma_{\beta \gamma } \sin \gamma } { r \sin ( \alpha + \gamma ) } \left[\mathrm{d } r_2 \right ] \/ , \nonumber \\ \left [ \frac{\mathrm{d } \tau } { \mathrm{d } r_2 } \right]^{(2 ) } - \left [ \frac{\mathrm{d } \tau } { \mathrm{d } r_2 } \right]^{(1 ) } & = & - \frac{\sigma_{\alpha \beta } \sin \alpha } { r \sin ( \alpha + \gamma ) } \left[\mathrm{d } r_1 \right ] - \frac{\sigma_{\beta \gamma } \sin \gamma \cos ( \alpha + \gamma)}{r \sin ( \alpha + \gamma ) } \left[\mathrm{d } r_2 \right ] \/ .
\nonumber \\ & & \label{lens7 } \end{aligned}\ ] ] alternatively , this may be translated into the notional derivatives of @xmath4 with respect to the contact angles : @xmath197^{(2 ) } - \left [ \frac{\mathrm{d } \tau } { \mathrm{d } \alpha } \right]^{(1 ) } & = & - \sigma_{\alpha \beta } \left[\mathrm{d } r_1 \right ] \/ , \nonumber \\
\left [ \frac{\mathrm{d } \tau } { \mathrm{d } \gamma } \right]^{(2 ) } - \left [ \frac{\mathrm{d } \tau } { \mathrm{d } \gamma } \right]^{(1 ) } & = & - \sigma_{\alpha \beta } \left[\mathrm{d } r_2 \right ] \quad .
\label{lens8 } \end{aligned}\ ] ] equations ( [ lens7 ] ) and ( [ lens8 ] ) show that the notional derivatives of @xmath4 depend on where the dividing interfaces are located .
this is completely analogous to what is known for the notional derivative of the surface tension for closed spherical interfaces . a common procedure aimed at obtaining equations for the contact angles consists of minimization of the grand canonical potential @xmath198 with the constraint of fixed volume @xmath199 ( compare eq .
( ( [ lensdrop4 ] ) ) : @xmath200 in eq .
( [ lens9 ] ) no bulk contributions appear because the volume @xmath199 , the bulk pressures @xmath201 and @xmath202 , and thus the pressure difference @xmath203 are kept fixed due to the fixed thermodynamic conditions specified by the temperature and the chemical potentials .
the fixed volume condition is implemented via the last term in eq .
( [ lens9 ] ) containing the lagrange muliplier @xmath204 .
the volume @xmath199 is the one enclosed by the dividing interfaces .
the independent variables of the variational problem are @xmath0 , @xmath2 , and @xmath57 ( @xmath1 is determined via @xmath205 ) . in order to calculate the variation @xmath206 resulting from the variations of @xmath0 , @xmath2 , and
@xmath57 we have to introduce
in addition to the interfacial and the line tensions new material parameters .
these are the stiffness constants @xmath207 of the interfaces describing the cost in interfacial free energy resulting from changes in the radii of curvature as well as the stiffnesses @xmath208 of the three - phase - contact line against changes of the contact angles @xmath0 and @xmath2 , and the radius of curvature @xmath57 , respectively . to be explicit : in the variation we introduce also contributions @xmath209 describing the cost in free energy associated with a variation of the curvature radius @xmath101 , caused by a change of @xmath57 by @xmath210 , without changing the interfacial area , or @xmath211 etc .. the symbol @xmath212 indicates that the stiffnesses are given by the respective derivatives evaluated _ at fixed thermodynamic conditions .
_ + the stiffness constants of the interfaces , e.g. , @xmath213 should not be confused with the derivatives of surface tensions with respect to the physical radius .
for example , @xmath214 does not coincide with the derivative of @xmath215 given in eq .
( [ lensdrop5 ] ) with respect to @xmath101 because drops of different physical sizes in an unstable or constrained equilibrium with their environment do not correspond to the same thermodynamic conditions given by temperature and chemical potentials .
the stiffness constants of the interfaces may be expressed via eq .
( [ lensdrop8 ] ) in terms of the notional derivatives .
this relation shows that the stiffness constant of an interface vanishes if the so - called surface of tension is chosen as the gibbs dividing surface . in the same way the stiffness constants attributed to the contact line
should not be confused with similarly looking derivatives with respect to an implicit dependence of the line tension on contact angles .
these implicit dependences reflect changes in the thermodynamic conditions .
the meaning of the stiffness constants is completely different . in setting up the variational principle
we explore constrained equilibrium configurations in the neighbourhood of _ the _ equilibrium configuration in order to find the equilibrium by minimizing the free energy at fixed thermodynamic conditions .
surface tensions , the line tension , and the various stiffness constants describe the costs in free energy due to virtual displacements of interfaces away from their equilibrium shape .
( in the summary operational procedures will be discussed how to determine stiffness constants theoretically . )
+ the equation @xmath216 leads to three equations because @xmath0 , @xmath2 , and @xmath57 can be varied independently .
they have the following form : @xmath217 @xmath218 and @xmath219 the ensuing equality of the right hand sides of eqs .
( [ lens11 ] ) and ( [ lens12 ] ) leads to one of two equations relating the variables @xmath0 , @xmath2 , and @xmath57 .
the second relation is provided by eq .
( [ lens13 ] ) .
finally , a third equation expresses the fixed volume @xmath199 of the lens , @xmath220 and thus finally determines @xmath0 , @xmath2 , @xmath57 , and the lagrange parameter @xmath204 via eqs .
( [ lens11 ] ) or ( [ lens12 ] ) .
we now compare two such sets of equations , each for a different choice of the dividing interfaces in order to find relations between the stiffnesses of the line for two different choices for the dividing interfaces .
since the physical object is kept fixed , the relations between one set of variables ( @xmath221 , @xmath222 , @xmath89 ) and another set ( @xmath223 , @xmath224 , @xmath88 ) follow from the definition of the lens by two intersecting spheres @xmath225 and @xmath226 as described in sec . 5 and simple geometrical considerations .
the volume @xmath199 attributed to the lens depends on the choice of the dividing interfaces , i.e. , two different volumes @xmath227 are assigned to the same physical object .
the relation between @xmath228 and @xmath229 is known once the notional changes of the radii @xmath101 and @xmath102 , fixing the relative positions of the dividing interfaces , are given .
therefore eq .
( [ lens13-z ] ) does not contain any new information that could be used in order to relate the stiffnesses which have to be attributed to the line for two different choices for the dividing interfaces .
thus in the following , we have to consider only two pairs of equations : first , the equation resulting from the equality of the right hand sides of eqs .
( [ lens11 ] ) and ( [ lens12 ] ) , and secondly eq . ( [ lens13 ] ) , one pair for each choice of the dividing interfaces .
the parameters @xmath221 , @xmath222 , and @xmath89 can be expressed in terms of @xmath223 , @xmath224 , and @xmath88 , and of the notional changes of the radii @xmath230 = r_i^{(2 ) } - r_i^{(1)}$ ] , @xmath156 . after expressing the pair of equations for convention ( 2 ) in terms of @xmath223 , @xmath224 , and
@xmath88 up to the order of the line tension term , the resulting equations may be compared directly with the pair of equations obtained if convention ( 1 ) was chosen from the beginning .
it is clear that the two pairs of equations must be equivalent since they describe the same physical system in terms of the same variables . in that comparison
we further use the relation ( [ lensdrop8 ] ) between the stiffnesses of the interfaces against changes of the radii of curvature and the related notional derivatives .
in addition we use that the mentioned notional derivatives written for two different dividing interfaces are related via eq .
( [ lensdrop7 ] ) which in the chosen variables read ( up to the relevant order it is not necessary to distinguish between @xmath223 and @xmath221 etc . and thus on the right hand side we omit the superscripts distinguishing between these conventions ) : @xmath231^{(2 ) } - \left[\frac{\mathrm{d}\sigma_{\alpha \beta } } { \mathrm{d } r_1 } \right]^{(1 ) } = \frac{2\sigma_{\alpha \beta } \sin^2\alpha\left[\mathrm{d}r_1\right]}{r^2 } + s.l.t.\ ] ] and @xmath232^{(2 ) } - \left[\frac{\mathrm{d}\sigma_{\beta \gamma } } { \mathrm{d } r_2 } \right]^{(1 ) } = \frac{2\sigma_{\beta \gamma } \sin^2\gamma\left[\mathrm{d}r_2\right]}{r^2 } + s.l.t.\hspace{0.3 cm } \quad.\ ] ] as a result of the comparison and the requirement of equivalence of the two pairs of equations discussed above ( rhs of eq .
( [ lens11 ] ) equals rhs of eq .
( [ lens12 ] ) , and eq .
( [ lens13 ] ) ) we obtain two coupled equations for the following three quantities : @xmath233 , @xmath234 , and @xmath235 .
these equations have a manifold of solutions .
if we pick a particular solution with @xmath236 the two remaining quantities are given by @xmath237 & \quad \\ & = & \frac { - r^2}{2\sin^2 \alpha}\left ( \left [ \frac{\mathrm{d } \sigma _ { \alpha \beta } } { \mathrm{d}r_1 } \right]^{(2 ) } - \left [ \frac{\mathrm{d } \sigma _ { \alpha \beta } } { \mathrm{d}r_1 } \right]^{(1 ) } \right ) & \\
\frac{\mathrm{d}\tau}{\mathrm{d } \gamma } \vert^{(2 ) } - \frac{\mathrm{d}\tau}{\mathrm{d } \gamma } \vert^{(1 ) } & = & - \sigma _ { \beta \gamma } \left[\mathrm{d}r_2 \right ] & \\ & = & \frac { - r^2}{2\sin^2 \gamma}\left ( \left [ \frac{\mathrm{d } \sigma _ { \beta \gamma } } { \mathrm{d}r_2 } \right]^{(2 ) } - \left [ \frac{\mathrm{d } \sigma _ { \beta \gamma } } { \mathrm{d}r_2 } \right]^{(1 ) } \right ) & \quad . \end{array}\ ] ] for the special choice in eq .
( [ lens16 ] ) leading to eq .
( [ lens17 ] ) the lagrange multiplier @xmath204 ( see eqs .
( [ lens11 ] ) and ( [ lens12 ] ) ) becomes independent of the chosen dividing interfaces and therefore it allows for a physical interpretation .
the conditions ( [ lens17 ] ) are obviously fulfilled if the following relations between stiffness constants of the line and those of the interfaces hold : @xmath238\ ] ] and @xmath239 \quad .\ ] ] additional terms which are independent of the choice of dividing interfaces could be added on the right hand sides of eqs .
( [ lens18 ] ) and ( [ lens19 ] ) without violating the conditions ( [ lens17 ] ) .
however , keeping such terms would not lead to more general results , but to equations which are more complicated than those presented below . as we shall see , the special choices of the right hand sides of eqs .
( [ lens18 ] ) and ( [ lens19 ] ) lead to equilibrium conditions for the contact angles which agree with eqs .
( [ lens1 ] ) and ( [ lens6 ] ) which have been derived above from the principle of invariance of @xmath5 under notional changes .
this agreement is achieved for the particular choice @xmath240 , but @xmath241 is still left undetermined by consistency requirements alone .
insertion of eqs .
( [ lens18 ] ) and ( [ lens19 ] ) into eqs .
( [ lens11 ] ) and ( [ lens12 ] ) leads to @xmath242\ ] ] and @xmath243 \quad.\ ] ] in other words , @xmath204 equals minus the laplace pressure expressed via the generalized laplace equation ( compare eq .
( [ lensdrop6 ] ) ) which supports the choice made in eqs .
( [ lens18 ] ) and ( [ lens19 ] ) above .
accordingly , the equations relating the contact angles to the radius @xmath57 read ( recall that this holds only for choices which are in accordance with eq .
( [ lens16 ] ) as well as eqs .
( [ lens18 ] ) and ( [ lens19 ] ) ) : @xmath244 = \sigma _ { \beta \gamma } \sin \gamma + \frac{r}{2 } \left [ \frac{\mathrm{d } \sigma _ { \beta \gamma } } { \mathrm{d}r_2 } \right ] \ ] ] and @xmath245 comparison of eq .
( [ lens23 ] ) with eq .
( [ lens6 ] ) ( eqs .
( [ lens22 ] ) and ( [ lens1 ] ) are identical anyway ) renders the relation
@xmath246 + \frac{\sin \gamma \cos \gamma } { r } \left [ \frac{\mathrm{d } \tau}{\mathrm{d } \gamma } \right ] \quad . \ ] ] adopting this relation for the surface of tension @xmath247 and using the relations in eqs .
( [ lens18 ] ) , ( [ lens19 ] ) , and ( [ lens16 ] ) and noting further that @xmath248^{(\mathrm{s } ) } = 0 = \left [ \frac{\mathrm{d } \sigma _ { \beta \gamma } } { \mathrm{d } r_2 } \right ] ^{(\mathrm{s } ) } $ ] and thus @xmath249 we find ( note that because of eq .
( [ lens16 ] ) @xmath250 is independent of the choice of the dividing interfaces ) @xmath251 ^{(\mathrm{s } ) } + \frac{\sin \gamma \cos \gamma } { r } \left [ \frac{\mathrm{d } \tau}{\mathrm{d } \gamma } \right ] ^{(\mathrm{s } ) } \hspace*{-0.20cm}. \end{split}\ ] ] on the rhs of eq .
( [ lens25 ] ) superscripts @xmath252 are omitted for @xmath0 , @xmath2 , and @xmath57 because differences between the values of @xmath0 , @xmath2 , and @xmath57 for different dividing interfaces give rise only to higher order corrections .
reinserting eq .
( [ lens25 ] ) into eq .
( [ lens24 ] ) we further find @xmath253 - \left [ \frac{\mathrm{d } \tau}{\mathrm{d } \alpha } \right ] ^{(\mathrm{s})}\ ] ] and @xmath254 - \left [ \frac{\mathrm{d } \tau}{\mathrm{d } \gamma } \right ] ^{(\mathrm{s } ) } \quad.\ ] ] of course , due to these relations the transformation behavior for the stiffness constants of the line tension ( eqs .
( [ lens16 ] ) and ( [ lens17 ] ) ) is consistent with that for the notional derivatives ( eq .
( [ lens8 ] ) ) .
it should also be noted that @xmath255 and @xmath256 . in the case of a drop
our analysis also starts from the requirement that the grand potential must be invariant with respect to notional changes .
they consist of a change of the radius of the @xmath0@xmath1 interface by @xmath257 $ ] and a common shift of the @xmath0@xmath2 and @xmath1@xmath2 interfaces by @xmath258 $ ] . from the requirement @xmath259 = 0 $ ]
we obtain two equations ( @xmath257 $ ] and @xmath258 $ ] can be chosen independently ) relating the contact angle @xmath47 to the radius @xmath57 . however , these equations have to be identical because already a single equation is sufficient to determine the relation between @xmath47 and @xmath57 .
this leads to the consistency relation @xmath260 + \cos \theta \left [ \frac{\mathrm{d } \tau } { \mathrm{d } r } \right ] = - \frac{r}{2 } \left [ \frac{\mathrm{d } \sigma_{\alpha \beta } } { \mathrm{d } r } \right]\ ] ] and the following equation for the contact angle : @xmath261 \quad.\ ] ] similarly as for the lens we have introduced notional derivatives @xmath262 $ ] and @xmath263 $ ] .
we also make use of a notional derivative @xmath264 $ ] of the spherically shaped @xmath0@xmath1 interface but at the same time we do not take into account a notional derivative @xmath265 $ ] of the difference in the interfacial tensions of the two planar substrate liquid ( @xmath2@xmath1 ) and substrate vapor ( @xmath2@xmath0 ) interfaces .
this calls for a comment . here
the quantity @xmath266 denotes the difference of two interfacial tensions corresponding to a situation in which the same substrate phase @xmath2 remains in contact with either the @xmath0 or the @xmath1 phase . both phases @xmath0 and @xmath1
are at the same pressure @xmath121 .
this quantity together with the @xmath0@xmath1 surface tension at infinite radius defines via the young equation the contact angle @xmath267 of a macroscopicly large liquid drop ( phase @xmath1 ) in contact with its vapor ( phase @xmath0 ) on top of the substrate @xmath2 .
the angle @xmath267 is an observable quantity and does not depend on the choice of the dividing surface . since @xmath215 does not depend on the choice of the dividing interfaces either , it follows that @xmath268 can not depend on the chosen position @xmath269 of the common @xmath0@xmath2 and @xmath1@xmath2 dividing interfaces .
+ one arrives at the same conclusion by calculating the changes of @xmath270 and @xmath116 with respect to changes @xmath271 in the interface positions . in the difference @xmath268 terms depending on
@xmath271 drop out provided the pressure in the phases @xmath0 and @xmath1 is the same , and a common height above the substrate is chosen for the @xmath0@xmath2 and for the @xmath1@xmath2 dividing interfaces ( see , e.g. , the discussion at the end of sect .
( both these assumptions are implicitly contained in the young equation . ) actually , further complications would arise if , in an attempt to mimic the situation in a drop more closely , we would carry out the calculations at different pressures , e.g. , for @xmath272 with the laplace pressure @xmath137 .
a quantity @xmath273 defined in that way would depend on the choice of the dividing interfaces and additional terms of the same order as those introduced by the line tension would appear in the final equations .
furthermore the difference @xmath273 would not be related to @xmath267 via the standard young equation and thus could not be determined via an independent experiment of measuring the shape of a macroscopic drop .
therefore it does not appear to be useful to introduce quantities like @xmath273 ( see , however , the discussion in subsect .
+ equations ( [ drop1 ] ) and ( [ drop2 ] ) may be rewritten in terms of @xmath47 and @xmath57 . in these variables
the consistency equation and the modified young equation take the following form : @xmath274 = \frac{r^2}{2 } \left [ \frac{\mathrm{d } \sigma_{\alpha \beta } } { \mathrm{d } r } \right ] \quad\ ] ] and @xmath275 - \frac{\sin \theta
\cos \theta}{r } \hspace{-0.10 cm }
\left [ \frac{\mathrm{d } \tau } { \mathrm{d } \theta } \right ] \hspace{-0.10cm}. \end{split}\ ] ] since the @xmath2@xmath0 and @xmath2@xmath1 interfaces are planar , a notional derivative of @xmath4 with respect to the angle @xmath2 as introduced for the lens can not be defined in the present case ; instead @xmath276 $ ] is used . as in the case of the lens the relations between ( @xmath277 $ ] ,
@xmath98 $ ] ) and ( @xmath278 $ ] , @xmath279 $ ] ) are unique and the same is true for the corresponding notional derivatives .
+ similarly as for the lens we find that _
@xmath4 is independent of the choice of dividing interfaces .
_ this statement is true in the same sense as the corresponding one discussed for the lens below eq .
( [ lens6 ] ) . as in the case of the lens ,
higher order corrections to @xmath4 decreasing with increasing @xmath57 like , e.g. , @xmath183 such corrections are possible if @xmath4 is defined via eq .
( [ lensdrop1b ] ) instead of eq .
( [ eta ] ) depend on the choice of the dividing interfaces . correspondingly , the notional derivatives of @xmath4 appearing in eqs .
( [ drop3 ] ) and ( [ drop4 ] ) also depend on the choice of dividing interfaces and the corresponding terms are furthermore of the same order as the term @xmath280 with only the leading term in @xmath4 being kept .
+ we now analyse the transformation of the notional derivatives upon notional shifts of the dividing interfaces .
the same kind of arguments as for the lens leads to the following relations : @xmath281^{(2 ) } - \left [ \frac{\mathrm{d } \tau } { \mathrm{d } r } \right]^{(1 ) } & = & \frac{\sigma_{\alpha \beta } } { r } \left\ { \cos \theta \left[\mathrm{d } r \right ] - \left[\mathrm{d } h \right ] \right \ } \nonumber \\ \left [ \frac{\mathrm{d } \tau } { \mathrm{d } h } \right]^{(2 ) } - \left [ \frac{\mathrm{d } \tau } { \mathrm{d } h } \right]^{(1 ) } & = & - \frac{\sigma_{\alpha \beta } } { r } \left \ { \left[\mathrm{d } r \right ] - \cos \theta \left[\mathrm{d } h \right ] \right \ } \ , .
\nonumber \\ & & \label{drop5 } \end{aligned}\ ] ] changing the set of independent variables from ( @xmath269 , @xmath282 ) to ( @xmath57 , @xmath47 ) leads to the following form of the above transformation laws : @xmath283^{(2 ) } - \left [ \frac{\mathrm{d } \tau } { \mathrm{d } \theta } \right]^{(1 ) } & = & \sigma_{\alpha \beta } \left[\mathrm{d } r \right ] \nonumber \\ \left [ \frac{\mathrm{d } \tau } { \mathrm{d } r } \right]^{(2 ) } - \left [ \frac{\mathrm{d } \tau } { \mathrm{d } r } \right]^{(1 ) } & = & - \frac{\sigma_{\alpha \beta } \sin \theta } { r } \left[\mathrm{d } h \right ] \ , . \label{drop6 } \end{aligned}\ ] ] here the objective is to minimize the grand canonical potential @xmath198 with the constraint of fixed volume @xmath199 : @xmath284 in which no bulk terms are included since the laplace pressure and the volume of the @xmath1 phase are considered to be fixed ; the lagrange multiplier is denoted by @xmath204 .
we use @xmath47 and @xmath57 as the variables describing the system , and we introduce the stiffness constants @xmath285 and @xmath286 which are then expressed in terms of the natural stiffness constant @xmath287 with known properties . in addition , similarly as for the lens , we introduce the stiffness constants @xmath288 and @xmath289 describing the cost in free energy attributed to the three - phase - contact line resulting from variational changes of @xmath57 and @xmath47 at fixed thermodynamic conditions . in the following
we shall determine how these stiffness constants depend on the choice of dividing interfaces and how they are related to notional derivatives .
+ if we impose the relation ( for a justification see the discussion following , c.f . , eq .
( [ drop11 ] ) ) @xmath290 as in the case of the lens , the lagrange multiplier @xmath204 acquires the meaning of the negative laplace pressure : @xmath291 .
\end{split}\ ] ] equation ( [ drop8 ] ) corresponds to the relations in eqs .
( [ lens18 ] ) and ( [ lens19 ] ) which were introduced in the case of the lens for similar reasons .
we also remark that eq .
( [ drop8 ] ) is formally identical to the consistency relation in eq .
( [ drop3 ] ) . the second equation , obtained from minimizing @xmath292 and due to the independence of the variables @xmath47 and @xmath57 , has the following form ( compare eq .
( [ lens13 ] ) ) : @xmath293 ( an equation which expresses the fixed drop volume @xmath294/ [ ( 1 + \cos \theta ) \sin \theta ] $ ] in terms of @xmath57 and @xmath47 and which together with the other equations fixes @xmath57 , @xmath47 , and the lagrange parameter @xmath204 , does not contain information that could be used in the following . the same has been found for the lens . ) + similarly as for the lens , we deduce from a comparison of the two equations , which follow from eq .
( [ drop10 ] ) for two differently chosen dividing interfaces , the transformation laws between the stiffness constants of the three - phase - contact line : @xmath295 \nonumber \\ \frac{\mathrm{d } \tau } { \mathrm{d } r } \big\vert^{(2 ) } - \frac{\mathrm{d } \tau } { \mathrm{d } r } \big\vert^{(1 ) } & = & - \frac{\sigma_{\alpha \beta } \sin \theta } { r } \left[\mathrm{d } h \right ] \quad .
\label{drop11 } \end{aligned}\ ] ] actually , the transformation behavior of @xmath296 is already fixed independently via eq .
( [ drop8 ] ) , @xmath297 $ ] , and eq .
( [ lensdrop7 ] ) .
there is no contradiction between these two independent requirements because of the formal equivalence of the transformation laws in eqs .
( [ drop6 ] ) for the notional derivatives and those for the stiffness constants in eqs .
( [ drop11 ] ) , and because eqs .
( [ drop3 ] ) and ( [ drop8 ] ) are formally identical . in other words ,
one is indeed free to fix @xmath298 by eq .
( [ drop8 ] ) and thus to bestow the meaning of the negative laplace pressure @xmath299 on the lagrange multiplier @xmath204 .
+ equation ( [ drop10 ] ) can be simplified by using eq .
( [ drop8 ] ) : @xmath300 from a comparison of eqs .
( [ drop12 ] ) and ( [ drop4 ] ) we obtain @xmath301 \hspace{0.2 cm } , \hspace{0.2 cm } \frac{\mathrm{d } \tau } { \mathrm{d } r } \big\vert = \left [ \frac{\mathrm{d } \tau } { \mathrm{d } r } \right ] \quad.\ ] ] ( note that applying eq .
( [ drop8 ] ) to eq .
( [ drop13 ] ) in combination with eq .
( [ lensdrop8 ] ) yields @xmath302 . )
again , these relations are compatible with the transformation laws for both the notional derivatives and the stiffness constants .
at this point it is due to comment on the seeming contradiction between our statement that @xmath4 is independent of the choice of dividing interfaces and eq .
( [ deltaeta2 ] ) , following from a discussion of a straight three - phase - contact line at a wedge shaped fluid volume , which states the opposite and tells how @xmath4 should change upon shifting the fluid solid dividing interface .
a question related to that issue is whether one can use the line tensions of a gas liquid solid contact calculated from a certain microscopic theory in combination with a decomposition scheme in a model system like that considered in sect .
iii ( i.e. , for a fluid wedge geometry , see , e.g. , refs .
@xcite ) , and use them in the formulae derived above for the drop .
+ in order to answer these questions we reanalyze again the particular wedge geometry investigated already in sect .
iii in order to find out whether implicit assumptions in the analysis given there are at variance with prescriptions used in sect .
v for the drop .
we also resume our discussion of the drop system in order to see whether sensible alternative definitions of the line tension in that system do exist and whether the contradiction mentioned above can be resolved by using an alternative definition of @xmath4 .
if such an alternative exists the question about the relation between the two different line tensions has to be answered . in sect .
iii we have calculated the changes of the areas of the liquid
gas and of the solid fluid interfaces due to a notional shift of the position of the solid fluid interface within the box with rectangular cross section , as shown in fig .
2 , implicitly assuming that these area changes are representative for an entire system from which fig .
2 actually shows only a small part . in this procedure
a possible source of errors might be hidden .
we therefore discuss now the underlying general procedure implicitly used in sect .
first , the total system is decomposed into two subsystems i and ii such that subsystem i contains the three - phase - contact line as the object of interest .
on the other hand subsystem i is surrounded by a subsystem ii which itself does not contain any actual line contribution .
( in the case of straight contact lines this requires to use periodic boundary conditions in one direction ; this is of no relevance for the following discussion . )
the contribution of subsystem i to the grand potential is then analyzed in terms of volume , interfacial , and line contributions _ ignoring unphysical edge terms_. the same is done for subsystem ii except that no actual line contribution appears in the decomposition . at the end all contributions from both subsystems are put together in order to calculate the grand potential of the total system .
it is expected that the grand potential ( and the decomposition into volume , interfacial , and line contributions ) calculated in that way is equal to what one obtains if one decomposes the grand potential of the whole system into volume , interfacial , and line contributions and then adds up these contributions without first partitioning the system . in what follows
we discuss why it may be sometimes misleading to ignore the unphysical edge terms .
the decomposition introduced above is not unique .
two examples of such decompositions are indicated in fig .
[ fig9 ] by two boxes of different shapes enclosing and defining the respective subsystem i. the structure is assumed to be translationally invariant in the direction perpendicular to the plane shown in fig .
the cross section of the box has the shape of a rectangle for one of the chosen decompositions and in the other case it has the shape of a parallelogram such that two of the faces of the box are perpendicular to the liquid
gas interface .
the embedding system ii is not specified in any detail except that it is understood that the structures found within the box , e.g. , the liquid gas interface , extend beyond the box boundaries into the embedding system ii .
for that reason interfacial energies at the boundaries of the boxes do not arise .
( this statement also comprises the solid liquid and the solid gas interfaces since the respective boundary of the box can be placed anywhere inside the solid . )
there are also no actual line contributions to the free energy from the edges of the box or from the line at which the liquid
gas interface intersects a face of the box .
the environments of these lines are not different from bulk or from a corresponding region in an infinitely extended interface .
inspite of the absence of actual line contributions , artificial non - physical line contributions at the lines just mentioned have to be included in the expressions for the grand canonical potentials characterizing the subsystems .
that there is a need to introduce these artificial line terms becomes very plausible if one looks at the two possible decompositions of a system into two subsystems shown in fig .
evidently interfaces are cut in different ways by the faces defining subsystem i for the two considered options .
therefore , contributions to the grand canonical potential stemming from certain parts of the interfaces are either attributed to subsystem i or subsystem ii depending on how the decomposition is done .
only if one includes the artificial line terms in the analysis one can keep track of such traces of the decomposition ; otherwise one runs into inconsistencies . before we proceed with this analysis , which ultimately aims at finding out whether and how the true line tension may be determined from microscopic calculations for the most simple geometries , we continue with the additive decomposition of the grand canonical potential @xmath5 of a system @xmath303 where @xmath304 and @xmath305 are the contributions to the grand canonical potential attributed to subsystem i and ii , respectively , which , e.g. , in the spirit of density functional theory can be thought of to be calculated from the density distributions characterizing the total system .
already this decomposition is problematic and not unique although obvious choices do exist .
for instance , if we base our theoretical description on a type of model in which the grand canonical potential is expressed in terms of a local functional of the number density ( of , e.g. , the ginzburg
landau type ) , @xmath304 is obtained by integrating the local density of the grand canonical potential over the volume of the box defining the subsystem i. the local density of the grand canonical potential is obtained from particle - density distributions which may be obtained from calculations restricted to the interior of the box .
the boundary conditions have to correspond to a seamless continuation of the structures in the interior of subsystem i into the surrounding subsystem ii .
the decomposition procedure becomes less obvious if a non - local density - functional description is used . in that case interactions across the boundaries of the subsystems
occur . in order to calculate the contribution to @xmath304 coming from two - center integrals one might , e.g. ,
restrict one of the integrations to the interior of the box , whereas the second integration runs over the total volume of the whole system .
the value obtained for @xmath304 is independent of size and shape of the surrounding subsystem ii provided it is sufficiently big and the interactions between particles decay sufficiently rapidly as a function of their distance .
artificial interfacial contributions from the surfaces of the box are eliminated automatically by chosing the described procedure . by carrying out the integrations in the way outlined above
we basically mimic a local density of @xmath5 allowing , however , for the embedding of the subsystem into a global system . in the sense of the previous paragraph @xmath5
can be decomposed into a sum of two contributions originating from two subsystems i and ii . since for this decomposition
no use has been made of the concept of dividing interfaces between phases , in particular between solid and liquid or solid and gas , it is also clear that the two contributions @xmath304 and @xmath305 must be independent of choices for the dividing interfaces between solid and fluid .
+ in the next step @xmath304 and @xmath305 are further decomposed into volume , interface , and line contributions , i.e. , @xmath306 and @xmath307 where @xmath308 and @xmath309 are the areas of the l g interfaces within subsystem i and ii , respectively , @xmath310 and @xmath311 are those of the s l interfaces , and @xmath312 and @xmath313 are the areas of the s
g interfaces ( l : liquid , g : gas , s : solid or wall ) .
@xmath4 is the line tension of the actual three - phase - contact line and thus it appears only in subsystem i. the terms @xmath314 or @xmath315 characterize the artificial line contributions resulting from the decomposition into subsystems .
the quantities @xmath316 are the lengths of the artificial lines , which are equal in both subsystems . since @xmath317 and
because the only actual line inhomogeneity , which is present in the total system , is accounted for by the line tension @xmath4 , one has @xmath318 i.e. , the artificial line contributions of the two subsystems cancel each other .
+ two possibilities for cutting a subsystem i out of a total system containing a l g interface and a three - phase - contact line with the solid s. these two possibilities are indicated by two boxes ( full and dotted lines ) enclosing the l
s contact line .
the structure is assumed to be translationally invariant perpendicular to the plane of the figure .
two possible choices for the solid fluid dividing interface are indicated by solid ( 1 ) and dashed ( 2 ) lines .
the two dividing interfaces are separated by a distance @xmath31 .
the areas of the solid liquid ( solid gas ) interfaces within subsystem i change when the solid fluid dividing interface is shifted .
the values of these area changes depend on the shape of the box defining subsystem i. this dependence is indicated by the two horizontal double arrows .
the short double arrow indicates the changes , due to the shift @xmath31 , of the areas of the solid gas and the solid liquid interfaces within the box of rectangular cross section , the long double arrow the changes of these areas within the box with parallelogram - shaped cross section . ] since @xmath304 is independent of the choice of the dividing interfaces between phases , in particular of the solid fluid dividing interface , the same arguments as used in sec .
3 can now be applied to the subsystem i and to @xmath304 in order to study the influence of a shift by an amount @xmath31 of the s
g and s l dividing interfaces .
( we restrict the discussion to a structure that is translationally invariant in the direction parallel to the line , i.e. , @xmath319 . ) if the rectangular box indicated in fig .
[ fig9 ] is chosen one obtains @xmath320 ^{(2 ) } - \left [ \tau + \sum _
i l _ i \tau _ i^{\mathrm{art , i } } \right ] ^{(1 ) } = \sigma_{\mathrm{lg } } \sin \theta \delta h\ ] ] for the difference of the expression in square brackets for the two choices ( 2 ) and ( 1 ) of the dividing interfaces ( compare fig .
2 ) . the only but essential difference to eq .
( [ deltaeta2 ] ) is that eq .
( [ decomp - gen-5 ] ) gives the difference not only for @xmath4 but for the sum of the actual and the artificial line energies .
there seems to be a chance that the seeming contradiction between the result that @xmath4 of the actual three - phase - contact line is independent of the choice of dividing interfaces and eq .
( [ deltaeta2 ] ) can be resolved by including the artificial line contributions and replacing eq .
( [ deltaeta2 ] ) by eq .
( [ decomp - gen-5 ] ) .
of course eq .
( [ decomp - gen-5 ] ) is only valid for the decomposition employing the rectangular box .
if instead of a rectangular box a box with a cross section of the shape of a parallelogram ( as shown in fig .
[ fig9 ] ) is chosen eq .
( [ decomp - gen-5 ] ) is replaced by @xmath321 ^{(2 ) } - \left [ \tau + \sum _ i l _ i \tau _ i^{\mathrm{art , i } } \right ] ^{(1 ) } = 0 \quad , \ ] ] because the changes of the solid liquid ( solid gas ) interfaces within subsystem i , resulting from the shift of the solid fluid dividing interface , are computed in the way indicated by the long dotted double arrow in fig .
[ fig9 ] and not in the way indicated by the short solid double arrow in fig .
[ fig9 ] which is the correct way to compute those changes within the rectangular box . from the comparison of eq .
( [ decomp - gen-5 ] ) and eq .
( [ decomp - gen-6 ] ) it becomes obvious that in order to deduce @xmath4 from a calulated expression for @xmath304 one has to subtract not only volume and interface contributions but also the artificial line contributions .
such artificial line contributions are generated if a box boundary cuts through the inhomogeneous structure of an interface in a non - adapted way . within an interfacial region , which has a macroscopic extension in the lateral directions , the density profiles of the constituents of the fluid
only depend on the coordinate perpendicular to the interface .
if the box boundary defining subsystem i cuts perpendicularly through the interface , the cut is adapted to the density profiles and no artificial line contribution is generated .
if by contrast , a box boundary cuts through the interface with an angle deviating from @xmath322 , a spatial region of columnar shape , filled with an inhomogeneous fluid , is generated above or below the interface , which no longer belongs to subsystem i although according to the interfacial area attributed to subsystem i it should contribute to the interfacial energy within that subsystem .
vice versa a volume of columnar shape is attributed to subsystem i although the fluid within that volume should contribute to the interfacial energy attributed to subsystem ii . since in general
the free energy contributions from these two columnar spatial regions do not compensate one is left with an artificial line contribution .
for that reason we discuss still another possibility ( fig .
[ fig - wedge - b ] ) for cutting a subsystem i out of a total system containing a l g interface and a three - phase - contact line with the solid s. for this choice the box boundaries cut perpendicularly , i.e. , in an adapted manner through both interfaces .
this choice avoids the appearance of artificial line terms at the cuts of the box boundaries through the liquid gas or the solid fluid interfaces .
an additional line generated by this particular choice of the box boundaries is placed such , that the system is completely homogeneous in a big volume around that line .
therefore , no net artificial line contribution is associated with that line , too . carrying out the same analysis as for the two other box shapes we
again obtain eq .
( [ decomp - gen-5 ] ) but the artificial line energies should now vanish leading back to eq .
( [ deltaeta2 ] ) . after our previous analyses for different box shapes defining subsystem
i one might come to the following suggestion to cope with artificial line contributions . for the rectangular box ( which leads to eq .
( [ decomp - gen-5 ] ) ) the only artificial line contribution stems from the intersection of the liquid gas interface with the box boundary .
no artificial line contributions arise at the cuts of the box boundaries with the solid fluid interface , because the box boundaries cut perpendicularly through that interface .
there are also no artificial line contributions from the edges of the box located in homogeneous regions of the fluid .
if we now argue that the artificial line contribution does not change upon a notional shift of the solid fluid interface because the corresponding line is far away from that interface , the terms related to the artificial line contribution would cancel in eq .
( [ decomp - gen-5 ] ) and we would again be led back to eq .
( [ deltaeta2 ] ) . in the case of the parallelogram shaped box and eq .
( [ decomp - gen-6 ] ) we might argue using the same type of arguments , that only the two line contributions due to the intersection of the box boundaries with the solid fluid interface do arise .
but now one would be inclined to admit that these artificial line contributions might change upon a a notional shift of the solid fluid interface because the lines are spatially associated with the shifted interface . therefore
, one would not draw conclusions about notional changes of @xmath4 from eq .
( [ decomp - gen-6 ] ) .
however , one should not rely too heavily on such type of arguments .
after all , notional shifts of dividing interfaces lead just to changes in interfacial areas and thus to changes in interfacial contributions to the free energy which are then subtracted from the total free energy in order to define the line tension(s ) . from the viewpoint of a macroscopic
theory it is not clear how these changes of interfacial free energies should be split up among the real and the artificial line contributions . + in the present subsection we have discussed how a line tension @xmath4 and its dependence on the choice of dividing interfaces can be determined unambigously from calculations for a simple model system , i.e. , a liquid wedge in contact with a gas phase on a solid substrate . in order to be able to proceed
, one has to carry out the calculations within a finite subsystem ( box ) cut out from the unbounded model system .
the seamless continuation into the embedding system has to be incorporated into the calculation by choosing proper boundary conditions and by taking into account certain interactions of the finite subsystem with the embedding system .
in addition , the box boundaries have to be choosen such , that they cut through interfaces perpendicularly and artificial box edges have to be placed into homogeneous regions of the total system ( a possible choice of the box is shown in fig .
[ fig - wedge - b ] ) .
with all these precautions we find that the dependence of @xmath4 on the choice of the solid fluid dividing interface is definitely given by eq .
( [ deltaeta2 ] ) .
another possibility for cutting a subsystem i out of a total system containing a l g interface and a three - phase - contact line with the solid s. the structure is assumed to be translationally invariant perpendicular to the plane of the figure .
two possible choices for the solid fluid dividing interface are indicated by solid ( 1 ) and dashed ( 2 ) lines .
the two dividing interfaces are separated by a distance @xmath31 .
the areas of the solid liquid ( solid gas ) interfaces within subsystem i change ( horizontal double arrow ) if the solid fluid dividing interface is shifted .
these changes in areas are identical to those obtained if the box of rectangular cross section is chosen in fig .
[ fig9 ] . ] after the discussions in the previous subsection we are still left with the contradiction between eq .
( [ deltaeta2 ] ) , which tells us how @xmath4 should depend on the choice of dividing interfaces , and our statement in sect .
v which says that @xmath4 should be independent of that choice .
we would like to stress that this latter statement is based on a well defined decomposition scheme of the grand potential of a drop on a substrate ( or a liquid lens at a fluid interface ) .
+ in order to gain further insight into this problem we now discuss a finite fluid system enclosed in a container with solid walls . this way we avoid from the outset the necessity to cut out a finite subsystem and we stay away from the danger to pick up unphysical edge or line contributions .
our treatment of closed finite containers gives us also the opportunity to discuss a further aspect of solid fluid systems , i.e. , the curvature of the solid fluid interface .
we further chose the boundary conditions such that the liquid vapor interface is planar and thus no complications occur related to the laplace pressure . in addition
there is no curvature correction to the liquid vapor surface tension . on the other hand if a container with only planar walls is partially filled with a liquid in contact with its vapor phase , in addition to the liquid
solid contact line further edge ( line ) contributions do appear which scale with the same linear dimension of the container as the three - phase - contact line we are interested in .
therefore , the different line ( edge ) contributions can not be separated unless these edge contributions have been determined independently from investigations of reference configurations not containing a liquid vapor solid contact line . as a reference configuration the same container but
completely filled with either liquid or gas ( vapor ) is chosen .
but then one encounters the problem that there is no unique description how to extract all additional edge contributions individually and in particular it is impossible to find the transformations upon notional shifts of dividing interfaces for each of these edge contributions individually .
+ in order to facilitate to vary the length of the liquid
solid contact line indepently from the lengths of container edges we consider a biconical container composed of two identical but oppositely oriented right cones with circular base which are glued together base to base along the circumference of the base as shown in fig .
. cut through a biconical container ( circular base ) with an opening angle 180@xmath323 .
the apexes of the container are filled with the liquid phase ( l ) of a fluid up to the planar liquid gas interfaces .
the central part of the container is filled with the coexisting gas phase .
the filling height is determined by a prescribed amount of liquid .
two possible choices for the dividing interface between solid ( wall ) and fluid are indicated by solid and dashed lines .
the two possible dividing interfaces are separated by a distance @xmath324 .
the two corresponding positions of the liquid gas
solid contact lines denoted as l are indicated by small and large dots . at the joint between the two cones
there is another circular line l@xmath325 formed . ]
all container walls are composed of the same solid material s. the cones are filled with liquid l up to a certain height , which is determined by the amount of liquid provided , and the liquid is in contact with its vapor phase g. the opening angle of the cone is chosen to be @xmath326 such that the liquid vapor interface is planar .
the system contains two structural units characterized by lines .
the first line is at the joint of the two cones and is termed l@xmath325 ( fig .
[ fig7 ] ) .
it has the shape of a circle with a radius denoted as @xmath57 . in the following the related line ( edge ) tension
will be called @xmath327 . the second line
, the one we are interested in , is the three - phase - contact line between solid , vapor , and liquid .
it also has a circular shape with a radius denoted as @xmath328 .
the related line tension will be called @xmath4 . again a parallel displacement of the dividing interfaces between solid and fluid by an amount @xmath324 is considered .
this parallel displacement is indicated in fig .
this set - up allows one to vary the length of the three - phase - contact line corresponding to @xmath4 independently from the container dimensions .
it is tempting to conclude that thus the line tension @xmath4 and its dependence on the choice of dividing interfaces can be separated unambigously from the line contributions related to the geometry of the container .
however , since the container walls are curved , we also have to take into account curvature corrections to the solid fluid interface tensions .
as we shall see , these curvature corrections bring about contributions to the grand potential which scale with the length of the liquid vapor solid contact line .
therefore , for this geometry one also has to study at first reference configurations not containing a liquid
vapor solid contact line , i.e. , the bicone being homogeneously filled with either liquid or vapor , in order to determine how the curvature contributions depend on the choice of dividing interfaces .
this is actually possible if an additional albeit plausible assumption is made ( see below ) . in the next step
the bicone filled partially with liquid and partially with vapor is analyzed .
what is found for the curvature terms in the previous step for the container filled with an homogeneous fluid is then transferred to the new situation of a container filled with an inhomogeneous fluid without any modification .
we decompose the grand canonical potential into volume , interfacial , line contributions , and we include curvature contributions due to the curvature of the solid fluid wall ( see , e.g. , @xcite . the curvature contributions could be also expressed as a curvature correction to the solid fluid interfacial tension .
terms which do not scale at least with a linear container dimension are disregarded : @xmath329 which is the grand canonical potential for one half of the system and @xmath121 is the pressure of the fluid , @xmath330 the volume of one cone , @xmath331 the area of the side surface of one cone , @xmath332 is the interface tension of a planar solid gas or solid liquid interface , respectively , @xmath333 is the mean curvature of the side surface integrated over the whole area of the side surface of one cone with @xmath334 as the corresponding thermodynamic coefficient , and @xmath335 is the line ( edge ) tension associated with the joint between the two cones ( see fig . [ fig7 ] ) of length @xmath336 . the superscript @xmath337 indicates that geometric quantities as well as the interfacial and line tensions and the coefficient of the curvature term depend on the choice ( @xmath338 ) of the dividing interface between solid and fluid .
the integrated mean curvature of the side surface of a cone with base radius @xmath339 is @xmath340 if we again compare the decompositions for the two choices ( 1 ) and ( 2 ) of the dividing interface shown in fig .
[ fig7 ] and if we use @xmath341 ( a special variant of eq .
( [ deltasigmasf ] ) not restricting the generality of the following arguments ) and if we neglect terms which do not scale at least with @xmath339 , we obtain the relation @xmath342 without a further assumption it is of course not possible to obtain the transformation behavior upon shifting the dividing interface for @xmath343 and @xmath327 independently . in order to proceed we consider a rounding of the edge at l@xmath325 around the joint of the two cones and attribute to this edge a total ( integrated ) mean curvature based on the following arguments . for a container with smooth walls without any edge , differential geometry provides the general relations ( see , e.g. , refs .
@xcite ) @xmath344 and @xmath345 expressing the changes in volume and surface area of a container , associated with an infinitesimal parallel shift @xmath324 of its surface , in terms of the area @xmath107 , the total mean curvature @xmath346 , and the total gaussian curvature @xmath347 . from the calculated volume and area changes and from eqs .
( [ curvature-2 ] ) and ( [ curvature-3 ] ) , the total mean curvature for the bicone ( with rounded edge ) turns out to be @xmath348 whereas adding up the integrated mean curvatures of the side surfaces of the two cones ( eq .
( [ curvature-1 ] ) ) gives @xmath349 this means that a contribution @xmath350 to the total mean curvature of the container is missing .
obviously this contribution can be attributed to the line l@xmath325 at which the two cones are glued together and the curvature attributed to that line can be realized by deforming the surface in its vicinity into one that is differentiable .
we use now this observation in order to replace the line ( edge ) term by an equivalent curvature term : @xmath351 combining eq .
( [ bicone - hom-2 ] ) and eq .
( [ bicone - hom-3 ] ) one finds @xmath352 the grand canonical potential of the bicone filled partially with liquid and partially with vapor ( see fig .
[ fig7 ] ) decomposes as @xmath353 where @xmath354 and @xmath355 are those areas of the cone side surface which are in contact with the gas and the liquid phase , respectively . @xmath356 and @xmath357 are the interface tensions of the planar solid gas and solid liquid interfaces . @xmath358 and @xmath359 are the mean curvatures of the cone side surface integrated over that part of the surface which is in contact with the gas phase and the liquid phase , respectively . @xmath360 and @xmath361 are the corresponding curvature coefficients .
@xmath362 is the area of the planar liqid
gas interface and @xmath363 the surface tension of a planar liquid gas interface .
the last but one term in eq .
( [ bicone - inhom-1 ] ) , already familiar from the homogeneously filled bicone , gives the line contribution from the edge along which the two cones are glued together .
the last term is the contribution from the solid liquid gas three - phase - contact line .
the superscript @xmath337 indicates which of the quantities depends on the choice of the solid fluid dividing interface ( @xmath364 is independent of such choices ) . as in the previous subsection
we compare the decompositions for the two different choices of dividing interfaces indicated in fig .
[ fig7 ] , we neglect all higher order terms which do not scale with a characteristic system size , we further use eq .
( [ bicone - hom-2 ] ) which has been obtained for the homogeneously filled bicone , and finally end up with the relation @xmath365 equation ( [ bicone - inhom-2 ] ) states that a certain combination of curvature and line contributions is invariant under a notional shift of the solid fluid dividing interface . in order to fix the transformation behavior of the individual contributions under a notional shift ,
an additional convention is required . insisting on using eq .
( [ curvature-7 ] ) , which appears to be a most plausible choice for the kind of systems considered , fixes the transformation behavior of the difference of curvature coefficients which appear in eq .
( [ bicone - inhom-2 ] ) .
this leads to eq .
( [ deltaeta2 ] ) which tells that @xmath4 does depend on the choice of the solid fluid dividing interface .
the surface tensions at planar solid fluid interfaces depend on the position of the dividing interfaces as given in eq .
( [ deltasigmasf ] ) .
therefore , the difference @xmath366 of the gas solid and the liquid solid interface tensions at coexistence of the gas and the liquid phase does not depend on the choice of the solid fluid dividing interface . in our analysis of the drop in sect . v we could make use of that property because in defining @xmath4 we have used solid
fluid interface tensions at the very same pressure irrespective of whether the fluid phase is the gaseous phase outside the drop or the liquid phase in the interior of the drop . we have also discussed the possibility of introducing instead a quantity @xmath367 in which the two interface tensions are meassured at different pressures .
we also have given reasons why we have not pursued this possibility further . a sessile liquid drop on a planar substrate ( @xmath2 ) .
two choices for the substrate fluid dividing interface separated by a distance @xmath31 are shown .
the contact angle @xmath47 is only shown for that choice of the dividing interface , which is indicated by the solid horizontal line .
the pressure in the liquid phase ( @xmath1 ) differs by @xmath137 from that in the gas phase ( @xmath0 ) .
the drop is a spherical cap with radius @xmath368 , area @xmath369 , and volume @xmath370 . ] however , in view of the conflicting results for the transformation behavior under notional shifts of dividing interfaces between different definitions of line tensions we now seriously consider this option . + in order to make our reasoning as transparent as possible we discuss now only a notional shift of the substrate fluid dividing interface as shown in fig .
[ fig-drop.eps ] , but leave the liquid gas dividing interface fixed .
we again use the decomposition in eq .
( [ lensdrop4b ] ) for two positions of the substrate fluid dividing interface and use at first the relation @xmath371 which is valid if both interface tensions are measured at the same pressure ( and if for both the @xmath0@xmath2 and the @xmath1@xmath2 interface the dividing interface is positioned at the same height ) . after neglecting terms which do not scale with at least a linear extension of the system one arrives at @xmath372 since in eq .
( [ drop - again-1 ] ) neither the liquid gas surface tension @xmath373 ( in the relevant order ) nor the difference @xmath374 depend on the choice of the substrate fluid dividing interface , the line tension @xmath4 is the only physical quantity that can pick up all the notional changes of volume and interfacial contributions . in particular , the notional change of the volume contribution ( in leading order @xmath375 and @xmath376 )
does also contribute to the notional change of @xmath4 .
evaluating eq .
( [ drop - again-1 ] ) up to leading order , i.e. , neglecting corrections to @xmath4 of the order of @xmath183 , one inevitably comes to the conclusion that @xmath4 is independent of the choice of the substrate fluid dividing interface .
+ the only sensible way how an alternative definition of the line tension @xmath4 could be introduced is to use in the decomposition scheme of eq .
( [ lensdrop4b ] ) a solid liquid interfacial tension @xmath377 taken at the pressure @xmath378 of a bulk liquid in contact with the solid , whereas the solid gas interfacial tension @xmath379 is taken at the pressure @xmath121 .
one may argue that this way one mimics the actual conditions at the interfaces of the liquid - drop vapor solid system . on the other hand the interfacial tensions @xmath116 and @xmath115 are not measurable individually , whereas the difference @xmath380 is a measurable quantity .
nevertheless , we now seriously consider that definition of @xmath4 . in order to study the transformation of this newly defined @xmath4 under notional changes we proceed as at the beginning of this subsection with the only difference that instead of eq .
( [ interface - tension - contrast ] ) we have to use now @xmath381 if we carry out the analysis neglecting all terms which obviously can only give rise to corrections to @xmath4 of the order @xmath183 , i.e. , terms which can not contribute to a line tension in its strict sense , we arrive at @xmath382 \quad .\end{gathered}\ ] ] we now replace @xmath383 in eq .
( [ drop - again-2 ] ) ) by @xmath384 because the difference of the two quantities is of the order of @xmath183 ( we then can skip the superscript @xmath385 in the difference between the solid liquid and the solid gas surface tensions ) and with the same argument we replace @xmath386 by its planar limit @xmath387 .
we then can use young s law and finally arrive at @xmath388 i.e. , we again find the transformation in eq .
( [ deltaeta2 ] ) , as we did repeatedly in discussing systems with planar liquid gas interfaces .
the reason why this second definition of the line tension leads us again to the transformation in eq .
( [ deltaeta2 ] ) can be traced back to the fact that in the second definition a contribution to the notional change of the volume contribution @xmath389 is absorbed by a corresponding notional change of @xmath390 , whereas in the first definition this possibility does not exist .
we now derive a relation between the two values of @xmath4 determined for the same system under identical conditions , but using the two different definitions introduced above . in order to do
so we compare the two decompositions of the grand canonical potential , on which the two definitions are based on , for a given , but otherwise arbitrary , choice of dividing interfaces . in order to distinguish the two definitions we denote with @xmath4 the value of the line tension determined according to the decomposition scheme introduced in sect .
v ( see eq .
( [ lensdrop4 ] ) with @xmath391 and @xmath392 ) , which we sketched again at the beginning of this subsection .
with @xmath393 we denote the line tension determined via the second decomposition scheme ( see eq .
( [ lensdrop4 ] ) but with @xmath391 and @xmath394 ) . on the basis of its transformation behavior and
because effects due to the laplace pressure , which do not occur for planar liquid gas interfaces , are extracted from its definition , we identify @xmath393 with the line tension as defined for the previously discussed systems with planar liquid gas interfaces
( it shall be understood that the identification applies to the line tension in its strict sense , i.e. , disregarding subleading terms ) .
the comparison leads to the relation @xmath395
\quad .\ ] ] for a one - component fluid this relation can be re - expressed by using @xmath396 where @xmath18 is the chemical potential , and @xmath397 where @xmath398 is the bulk density of the fluid and @xmath399 \quad , \ ] ] whith @xmath400 as the excess adsorption .
equation ( [ drop - again-7 ] ) can be easily derived from the definition of the interface tension @xmath401 + pv \right ) , \ ] ] where @xmath402 is the equilibrium number density minimizing @xmath5 , and the general functional form of the grand canonical potential : @xmath403 = { \cal{f } } \left ( \rho ( { \bf r } ) \right ) + \int \rho ( { \bf r } ) \left ( v_{\mathrm{ext } } - \mu \right)\mathrm{d}^3r \quad .\ ] ] combining eqs .
( [ drop - again-4 ] ) ( [ drop - again-9 ] ) and expressing the laplace pressure @xmath137 in terms of the liquid
gas surface tension one obtains ( @xmath404 is the bulk density in the @xmath1 phase ) : @xmath405 @xmath406 is the excess adsorption at the planar @xmath1@xmath2 interface . in eq .
( [ drop - again-10 ] ) both terms on the right hand side depend on the choice of the liquid solid dividing ( @xmath1@xmath2 ) interface but these dependences cancel and thus @xmath4 is independent of that choice as it should .
+ the discussion in the present section has shown that the contradiction between our statement in sect .
v , that @xmath4 is independent of the choice of dividing interfaces , and eq .
( [ deltaeta2 ] ) is resolved as follows .
we have found that the definitions of the line tension at a substrate fluid fluid interface , which have been chosen on one hand in sect . v and on the other hand in eq .
( [ deltaeta2 ] ) , are different .
the line tension , which appears in eq .
( [ deltaeta2 ] ) , is @xmath393 and differs from @xmath4 as defined in sect .
v. the possibility to define two different line tensions arises because it is not obvious how to subtract the contribution to the grand canonical potential stemming from the planar @xmath1@xmath2 ( liquid substrate ) interface of a drop of @xmath1-phase in contact with a substrate .
the liquid
substrate interface tension @xmath115 is influenced by the laplace pressure , but @xmath115 is not accessible .
however , the difference @xmath407 of the fluid
substrate interface tensions is measurable via the contact angle for large drops .
if in the decomposition scheme defining the line tension we use an interface tension at a pressure deviating from the true pressure in the interior of the drop , i.e. , @xmath392 , in order to avoid the somewhat artificial combination @xmath408 , one is led to the definition @xmath4 for the line tension . if instead one uses @xmath394 in the decomposition one arrives at the definition @xmath393 .
corresponding alternative choices do not exist in the case of the contact of three fluid phases , since all fluid fluid interface tensions are measurable individually and since the dependence of interface tensions on curvature radii parametrize their dependence on the laplace pressure in a natural way .
by contrast , the geometry of the solid liquid interface is not influenced by the laplace pressure and it stays planar for all values of the laplace pressure .
the definition @xmath393 for the line tension has the merit that interface contributions due to the modification of the @xmath1@xmath2 interface tension caused by the laplace pressure , giving rise to a contribution proportional to the linear extension of the system , are not implicitly included in @xmath393 .
however , in order to use @xmath393 in equations determining the contact angle , it is necessary to re - express a number of equations given for the drop in sect .
v. equation ( [ drop1 ] ) expressed in terms of @xmath393 and its notional derivatives turns into @xmath409 + \cos \theta \left [ \frac{\mathrm{d } \tau_{\mathrm{w } } } { \mathrm{d } r } \right ] = \sin \theta \sigma_{\alpha \beta } \ ] ] and instead of eq .
( [ drop2 ] ) one obtains @xmath410 , \ ] ] or in an alternative notation @xmath411 \quad.\end{gathered}\ ] ] the relations between the notional derivatives of @xmath4 and @xmath393 are @xmath412 = \left[\frac{\mathrm{d } \tau_{\mathrm{w } } } { \mathrm{d } r } \right ] + \frac{1}{2 \sin \theta } \left ( \sigma_{\beta \gamma}(p+\delta p ) - \sigma_{\beta \gamma } ( p ) \right ) \ ] ] and @xmath413 = \left [ \frac{\mathrm{d } \tau_{\mathrm{w}}}{\mathrm{d } h } \right ] - \frac{\cos \theta}{2 \sin \theta } \left ( \sigma_{\beta \gamma}(p+\delta p ) - \sigma_{\beta \gamma } ( p ) \right ) - \frac{r \delta p}{2 } .\ ] ] in terms of the variables @xmath47 and @xmath57 one obtains @xmath414 = - \frac{r}{\sin \theta } \sigma_{\alpha \beta}\ ] ] and @xmath415 - \frac{\sin \theta \cos \theta}{r } \left [ \frac{\mathrm{d } \tau_{\mathrm{w } } } { \mathrm{d } \theta } \right ] , \end{split}\ ] ] or @xmath416 , \ ] ] which replace eqs . ( [ drop3 ] ) and ( [ drop4 ] ) . it is interesting to see that in eq .
( [ drop4-neu - b ] ) the term @xmath417 drops out due to the identity eq .
( [ drop3-neu ] ) , i.e. , this term must now be included in the term @xmath418 $ ] .
this is displayed in the relation @xmath419 = \sigma_{\alpha \beta } \cos \theta - \frac{1}{2 } ( \sigma_{\beta \gamma}(p+\delta p ) - \sigma_{\beta \gamma } ( p ) ) \\
+ \left [ \frac{\mathrm{d } \tau } { \mathrm{d } r } \right ] + \frac{\sin \theta \cos \theta}{r } \left [ \frac{\mathrm{d } \tau } { \mathrm{d } \theta } \right ] .
\end{split}\ ] ] equations ( [ drop3-neu ] ) - ( [ drop4-neu - c ] ) show that using @xmath393 instead of @xmath4 spoils the clear hierarchy in the various terms describing notional changes . + for completeness we provide also the transformation laws for the notional derivatives of @xmath393 : @xmath420^{(2 ) } - \left [ \frac{\mathrm{d } \tau_{\mathrm{w } } } { \mathrm{d } r } \right]^{(1 ) } = \frac{\sigma_{\alpha \beta } } { r } \cos \theta [ \mathrm{d } r ] , \ ] ] @xmath421^{(2 ) } - \left [ \frac{\mathrm{d } \tau_{\mathrm{w } } } { \mathrm{d } h } \right]^{(1 ) } = - \frac{\sigma_{\alpha \beta } } { r } \cos \theta [ \mathrm{d } h ] , \ ] ] @xmath422^{(2 ) } - \left [ \frac{\mathrm{d } \tau_{\mathrm{w } } } { \mathrm{d } \theta } \right]^{(1 ) } = - \sigma_{\alpha \beta } [ \mathrm{d } r ] , \ ] ] and @xmath423^{(2 ) } - \left [ \frac{\mathrm{d } \tau_{\mathrm{w } } } { \mathrm{d } r } \right]^{(1 ) } = \frac{\sigma_{\alpha \beta } \sin \theta}{r } [ \mathrm{d } h ] .\ ] ] we make now again contact with a variational approach with the constraint of fixed volume .
of course one could stick to the variational approach as discussed in sect .
v without any modifications . in that case one just might want to rewrite eqs .
( [ drop12 ] ) and ( [ drop13 ] ) in terms of @xmath393 and its notional derivatives .
alternatively , one may slightly modify the variational approach by introducing a @xmath1@xmath2 interface tension measured at a pressure @xmath378 in the same way as in the definition of @xmath393 . as in sect .
v we further introduce stiffnesses @xmath424 and @xmath425 but we do not introduce any further stiffness .
( one might contemplate to endow the @xmath1@xmath2 interface with a stiffness but there is no geometric measure characterizing that interface which could be related to such a stiffness ; the area is already used and the radius of cirumference is better attributed to the contact line . ) in that way one arrives at @xmath426 and @xmath427 one also obtains the relations @xmath428 - \sigma_{\alpha \beta } \cos \theta - \frac{r \cos \theta}{2 \sin \theta } \left [ \frac{\mathrm{d } \sigma_{\alpha \beta } } { \mathrm{d } r } \right ] , \ ] ] @xmath429 @xmath430 and the transformation law @xmath431
we have discussed some conceptual issues that arise in a macroscopic thermodynamic description of the three - phase contact of either three fluid phases or of two fluid phases meeting an inert solid substrate .
we have pointed out that the concept of a line tension accompanying the contact line has to be used with great care .
the conceptual difficulties arise because the interfaces between two phases are always diffuse and never sharp .
therefore a certain degree of freedom exists in positioning , somewhere in the transition region , an idealized mathematical interface ( the so called gibbs dividing interface ) separating the two adjacent phases . as a consequence ,
a similar degree of freedom exists in the position of the contact line defined as the common intersection of three dividing interfaces .
we have analyzed implications for a consistent description of the contact line following from the existence of these degrees of freedom .
for that purpose we have investigated two representative systems of three - phase contact : a liquid lens at a fluid fluid interface and a liquid drop on top of a smooth substrate .
both systems are used in experimental attempts to determine a line tension @xmath4 from the dependence of contact angles on the system size .
we have defined a prescription for decomposing the grand canonical free energy of a liquid lens or drop into volume , interface , and line contributions which renders a line tension @xmath4 independent of a particular choice of the dividing interfaces .
the prescription rests on geometrical definitions and on the notion that those interfaces of a lens or drop system , which are spherical segments , should be described in the same way as the interfaces of completely spherical drops . in particular
this means that the pressure drop across the curved interface is related to the surface tension ( and surface stiffness against changes of the radius of curvature @xmath282 ) via the generalized laplace equation ( [ lensdrop6 ] ) and that the surface tension of a spherical interface is independent of a notional shift @xmath97 $ ] of the dividing interface up to and including contibutions proportional to @xmath432 for large @xmath282 .
we have also used the fact that interface tensions of planar interfaces between two fluids in thermal equilibrium are independent of the position of the dividing interfaces .
the same is implied for the difference between the substrate
gas ( substrate vapor ) and the substrate
liquid interface tensions although the substrate on one hand and the fluid phases on the other are not in equilibrium . in fact this is true if the substrate liquid and the substrate gas interface tensions used in the decomposition scheme correspond to the same pressure , a prescription which has the advantage of relating the difference of these two quantities directly to the contact angle of macroscopic drops which is a measurable quantity .
our result , that @xmath4 is independent of the choice of dividing surfaces , is first of all a generalization to _ curved _ interfaces and _ curved _ contact lines of what is known for the line tension of a _
straight _ contact line at the intersection of three _ planar _ interfaces in a genuine three - phase contact @xcite .
a further generalization is the one from a genuine three - phase contact to a contact between two fluid phases and an inert solid phase .
it should be noted , however , that in the latter case an alternative definition of the line tension , denoted as @xmath393 , is possible which seems to be more natural for systems containing planar liquid gas interfaces and planar or curved solid walls and more useful for the purpose of computing line tensions by exploiting the simplest possible geometries . it should be emphasized that @xmath393 does depend on the choice of dividing interfaces .
the difference in the definition of this alternative line tension relative to the previous one rests on choosing in the decomposition scheme of the grand canonical potential a different substrate
liquid interface tension which is not taken at the pressure of the gas phase , which is implied for the substrate gas interface tension , but at a pressure which is enhanced by the laplace pressure .
we have provided a simple relation between the values of the line tensions corresponding to the two alternative definitions ( eqs .
( [ drop - again-4 ] ) and ( [ drop - again-10 ] ) ) .
we have further pointed out that the generalized neumann or young equations obtained from a minimization principle by simply adding a line - tension contribution to the free energy suffer from internal inconsistencies .
the purely geometric relations between two sets of contact angles obtained for two different choices of the dividing interfaces are at variance with those equations and with our result from the decomposition of the free energy which states that the line tension @xmath4 is independent of such descriptive ambiguities . in the case of a drop on a solid substrate using the alternative definition @xmath393 of a line tension can not resolve those inconsistencies . in order to find equations for the contact angles which are internally consistent we have followed two different routes . the first one is determined by the observation that the grand canonical free energy must be independent of notional ( descriptive ) changes of the system , i.e. , of parallel shifts of the dividing interfaces at fixed physical configurations . in the mathematical formulation we introduced notional changes of the line tension and corresponding notional derivatives of @xmath4 with respect to contact angle(s ) and to the contact - line radius in addition to the well established notional derivatives of surface tensions with respect to radii of curvature .
the second route follows a minimization of the grand canonical free energy under the constraint of fixed drop or lens volume .
in contrast to the previous formulations we have included into the free energy contributions from the stiffness of the interfaces with respect to a change of the radius of curvature and from stiffnesses of the contact line with respect to changes in contact angle(s ) and to the contact - line radius , all at fixed thermodynamic variables .
the equations following from these two routes have been compared and combined into one ( set ) of equations . for the lens these are eqs .
( [ lens22 ] ) and ( [ lens23 ] ) together with the relations in eqs .
( [ lens25 ] ) ( [ lens27 ] ) and also eqs .
( [ lens18 ] ) and ( [ lens19 ] ) relating the stiffness constants and notional derivatives .
for the drop the corresponding equations are given by eq .
( [ drop12 ] ) and the relations in eqs .
( [ drop13 ] ) and ( [ drop8 ] ) .
furthermore we have found that the actual values of the stiffness constants depend on the choice of the dividing interfaces .
the relations between two sets of stiffness constants for two different sets of dividing interfaces are given by eqs .
( [ lens16 ] ) and ( [ lens17 ] ) for the lens and by eq .
( [ drop11 ] ) for the drop .
( in case of the drop one could also use the alternative eqs .
( [ drop8-neu ] ) to ( [ drop12-neu - e ] ) . ) at this point it is interesting to see that for reasons similar to those which compelled us to introduce notional derivatives of @xmath4 with respect to contact angles or stiffnesses of @xmath4 with respect to changes of contact angles , djikaev and widom @xcite ( see also ref .
@xcite ) introduced a kind of derivative of @xmath4 with respect to contact angle in order to restore invariance against notional shifts of dividing interfaces of their linear adsorption equation for a straight contact line at a genuine three - phase contact .
even closer to our discussion is the work by rusanov et al .
@xcite , which differs , however , from ours in a number of important points .
first , rusanov et al .
discuss only drops on a substrate , whereas we discuss both lenses and drops .
second , in contrast to rusanov et al .
we have included notional shifts of the substrate fluid interfaces .
third , rusanov et al .
have not used a standard variational principle at constant volume ( typical examples for the use of this principle are given , e.g. , in refs .
@xcite ) and they did not give transformation laws between values of stiffness constants for different choices of dividing interfaces .
finally we have tried to make clear at which points in our line of arguments there is still room for chosing different conventions and what follows as a necessity for any sensible convention .
next we discuss the consequences of our investigations for interpreting experimental data . before doing so
, we provide explicit expressions for the change in contact angles @xmath433 or @xmath47 with respect to their limiting values @xmath434 , @xmath435 , @xmath61 or @xmath267 for macroscopicly large lenses or drops , respectively ( see fig .
3 ) . from eqs .
( [ lens23 ] ) and ( [ lens22 ] ) together with eq .
( [ lensdrop5 ] ) and the corresponding equation for @xmath115 we obtain @xmath436 similarly , eq .
( [ drop12 ] ) together with eq .
( [ lensdrop5 ] ) yields @xmath437 the leading type , @xmath438 , of correction terms arises due to algebraically decaying dispersion forces among the particles ( see ref .
@xcite ) . in eqs . ( [ lens28 ] ) and ( [ drop14 ] )
@xmath57 is the radius of the circular contact line , @xmath439 the interface tension of the planar @xmath440@xmath441 interface , @xmath442 the tolman length of the @xmath440@xmath441 interface , @xmath443 etc .
are stiffnesses against changes of contact angles or of the radius of the line , which are attributed to the line , and @xmath4 is the line tension . in the traditional analyses of size dependent
contact angles only the term proportional to @xmath4 is included . however , the omitted terms give contributions to the right hand sides of eqs .
( [ lens28 ] ) and ( [ drop14 ] ) which are comparable in magnitude to that term .
although @xmath4 is independent of the chosen dividing interfaces , the stiffnesses @xmath443 and @xmath444 in the case of the lens and @xmath250 and @xmath445 in the case of the drop are not . therefore , the additional terms in eqs .
( [ lens28 ] ) and ( [ drop14 ] ) containing these stiffnesses even depend on the positions of dividing interfaces which may be chosen arbitrarily within a certain range .
this is in accordance with the fact that this dependence is also present for the contact angles on the lhs of eqs .
( [ lens28 ] ) and ( [ drop14 ] ) ( see step 1 in the protocol given below ) .
moreover , the changes in the values of these particular terms with the position of dividing interfaces are as big as the term proportional to @xmath4 itself . in order to demonstrate this we compare @xmath4 with @xmath446 , @xmath447 , @xmath445 , or @xmath250 as well as with the terms of the form @xmath448 involving the tolman length @xmath449
the change , e.g. , of @xmath446 with a shift of the @xmath0@xmath1 dividing surface by @xmath450 $ ] is equal to @xmath451 $ ] for typical values of @xmath114 of the order of @xmath452 , and for @xmath450 $ ] of the order of @xmath453 one obtains a change in the value of @xmath446 of the order of @xmath454 , which is just the typical value of calculated or measured line tensions .
similar estimates hold for the other terms in particular also for those of the form @xmath448 if the reasonable assumption is made that @xmath455 is not much smaller than @xmath453 ( typical values of @xmath455 obtained theoretically are of the order of half a molecular diameter , see , e.g. , refs .
@xcite ) . 1 . giving experimental values of contact angles
@xmath1 or @xmath47 requires to define the dividing interfaces relative to which the contact angles are measured .
( so far this information is missing for basically all published experimental efforts to measure the line tension . ) after the dividing interfaces have been chosen ( iso - density surfaces ) , the spherical parts of the interface profiles of the drop or lens have to be determined by fitting spheres to their central parts ( e.g. , to the top of the drop ) . in case of the drop , one such sphere intersects with the planar solid
fluid interface ( chosen by convention ) at the three - phase - contact line .
the contact angle at this line is obtained by a tangential - plane construction to the sphere , i.e. , to the spherical extrapolation of the liquid
gas interface down to the solid fluid interface . in the case of the lens ,
two spheres are defined by the central part of the lens .
these spheres intersect at the three - phase - contact line , at which tangential planes to the spheres define the contact angles .
2 . the dependence of these angles on the lens or drop sizes , as characterized by @xmath57 , has to be studied in the _ leading _ asymptotic behavior for large @xmath57 .
this dependence is not only determined by the line tension @xmath4 calculated for a straight contact line but it depends on additional material parameters .
therefore , the common practice of infering @xmath4 from @xmath456 or @xmath457 according to eqs .
( [ mye10 ] ) or ( [ mye2 ] ) is not valid .
3 . if one would compare the rhs of eqs .
( [ lens28 ] ) and ( [ drop14 ] ) with theoretical quantities , these would have to be evaluated for that choice of the dividing interfaces for which the experimental data on the lhs are given .
4 . the values of certain stiffness constants appearing on the rhs of eqs . ( [ lens28 ] ) and ( [ drop14 ] )
depend sensitively on the choice of dividing interfaces .
however , the transformation behavior of these material parameters between different such choices is known and given by eqs .
( [ lens16 ] ) and ( [ lens17 ] ) for the lens and eq .
( [ drop11 ] ) for the drop .
all quantities on the rhs of eqs .
( [ lens28 ] ) and ( [ drop14 ] ) are accessible to theoretical computations , e.g. , based on density functional theory .
methods that can be used to calculate the tolman lengths are described in the literature ( see , e.g. , refs .
@xcite ) . + in calculations of the line tension much care has to be taken not to pick up additional artificial line or edge contributions . in the likely case that within a theoretical approach one has calculated @xmath393 , its relation to @xmath4 is given by eqs .
( [ drop - again-4 ] ) and ( [ drop - again-10 ] ) , provided both quantities apply to the same thermodynamic conditions .
+ once the density profiles for a lens or drop have been computed , the stiffness constants @xmath445 , @xmath250 etc . may be determined from their relations to the notional derivatives of @xmath4 ( see , e.g. , eq .
( [ drop13 ] ) ) or in the case of @xmath445 also to the stiffness constant of the interface ( see , e.g. , eq .
( [ drop8 ] ) ) which again is given by a notional derivative ( eq .
( [ lensdrop8 ] ) ) .
similar equations which apply to the case of the lens are eqs .
( [ lens18 ] ) , ( [ lens19 ] ) , ( [ lens25 ] ) , ( [ lens26 ] ) , and ( [ lens27 ] ) . @xmath4 itself as well as the notional derivatives of @xmath4 follow from decomposing the grand canonical potential for a series of different choices for the dividing interfaces .
for instance one might decompose the grand canonical potential @xmath5 of a given lens for a series of different values of @xmath458 $ ] and @xmath459 $ ] . carrying out this decomposition up to the order @xmath183 and making use of previous knowledge about the notional derivatives of the surface tensions ( see eq .
( [ lensdrop7 ] ) ) , the notional derivatives @xmath460 $ ] and @xmath461 $ ] can be determined from the dependence of @xmath4 on @xmath458 $ ] and @xmath459 $ ] .
these notional derivatives may be then converted into the notional derivatives @xmath462 $ ] and @xmath463 $ ] via the relations @xmath460
= \frac { \sin \alpha \cos ( \alpha + \gamma ) } { r \sin ( \alpha + \gamma ) } \left [ \frac{\mathrm{d } \tau}{\mathrm{d } \alpha } \right ] + \frac { \sin \gamma } { r \sin ( \alpha + \gamma ) } \left [ \frac{\mathrm{d } \tau}{\mathrm{d } \gamma } \right ] $ ] and @xmath461 = \frac { \sin \alpha } { r \sin ( \alpha + \gamma ) } \left [ \frac{\mathrm{d } \tau}{\mathrm{d } \alpha } \right ] + \frac { \sin \gamma \cos ( \alpha + \gamma ) } { r \sin ( \alpha + \gamma ) } \left [ \frac{\mathrm{d } \tau}{\mathrm{d } \gamma } \right ] $ ] .
similarly the notional derivatives @xmath464 $ ] and @xmath465 $ ] can be determined in case of the drop and converted into @xmath466 $ ] and @xmath467 $ ] via the relations @xmath464 = \frac { \cos \theta } { r } \left [ \frac{\mathrm{d } \tau}{\mathrm{d } \theta } \right ] + \frac { 1 } { \sin \theta } \left [ \frac{\mathrm{d } \tau}{\mathrm{d } h } \right ] $ ] and @xmath465 = - \frac { 1 } { r } \left [ \frac{\mathrm{d } \tau}{\mathrm{d } \theta } \right ] - \frac { \cos \theta } { \sin \theta } \left [ \frac{\mathrm{d } \tau}{\mathrm{d } h } \right ] $ ] .
although each individual quantity entering the rhs of eqs .
( [ lens28 ] ) and ( [ drop14 ] ) can be calculated separately in the way indicated above , measurements of the size dependent contact angles of drops or lenses can provide only certain combinations of material parameters .
their values depend sensitively but in a known way on the choice of dividing interfaces . 1 . how can the various stiffness constants attributed to the contact line be measured ?
the answer to this question requires the extension of the principles developed here for simple geometries to more general geometries . how do the stiffness constants of the contact line influence the equilibrium shapes of more complex structures , e.g. , of liquid bridges between solid substrates or of lenses and drops distorted from their ideal shape due to external forces ? 2 .
how can the stiffness constants of the contact line introduced here be related to the derivatives of @xmath4 with respect to contact angles which were introduced by djikaev and widom @xcite in order to restore invariance of their linear adsorption equation against notional shifts of dividing interfaces ?
the authors thank r. roth and b. widom for helpful comments .
one of the authors m.n . expresses his thanks for the hospitality at the max - planck - institut fr metallforschung where most of this work has been done ; he also acknowledges the support via the polish ministry of science and higher education grant n 202 076 31/0108 . 199 , _
lett . _ * 80 * , 3863 ( 1998 ) . , _ science _ * 283 * , 41 ( 1999 ) . ,
_ science _ * 283 * , 346 ( 1999 ) . , _ j. micromech .
_ * 3 * , 168 ( 1993 ) .
j. w. gibbs , _ the scientific papers _ ( dover , new york , 1961 ) , vol . i. j. langmuir , _ j. chem . phys . _
* 1 * , 756 ( 1933 ) .
r. rowlinson and b. widom , _ molecular theory of capillarity _
( oxford university press , new york , 1982 ) . , in _ proceedings of the gibbs symposium _ ( yale university press , 1989 ) , p.73 . ,
_ physica a _ * 263 * , 500 ( 1999 ) . ,
_ j. physique lett . _ * 44 * , l585 ( 1983 ) . ,
_ j. chem .
phys . _ * 81 * , 552 ( 1984 ) . , _
j. colloid interface sci . _ * 104 * , 477 ( 1985 ) .
phys . _ * 57 * , 827 ( 1985 ) . , _ adv .
colloid interface sci . _ * 110 * , 121 ( 2004 ) .
, in _ phase transitions and critical phenomena _ , edited by c. domb and j. l. lebowitz ( academic , london , 1988 ) , vol . 12 , p.1 . , in _ condensed matter theories _ , edited by l. blum and f.b
malik ( plenum , new york , 1993 ) , vol . 8 , p. 589 ; _ proceedings of the 9th int . symp . on surfactants in solution , varna , bulgaria , 1992 _
( unpublished ) .
b _ * 8 * , 309 ( 1994 ) . ,
_ j. phys .
chem . _ * 99 * , 2803 ( 1995 ) .
e _ * 57 * , 655 ( 1998 ) . c. bauer and s. dietrich , _ eur .
j. b _ * 10 * , 767 ( 1999 ) . ,
_ physica a _ * 271 * , 36 ( 1999 ) .
b _ * 13 * , 3255 ( 1999 ) . ,
_ j. chem .
faraday transactions i _ * 73 * , 390 ( 1977 ) . , _ ann . rev .
fluid mech . _ * 11 * , 371 ( 1979 ) . , _
j. colloid interface sci . _ * 77 * , 495 ( 1980 ) .
, _ europhys .
* 21 * , 483 ( 1993 ) . , _ adv .
colloid interface sci .
_ * 40 * , 157 ( 1992 ) .
, _ colloids and surfaces a _ * 116 * , 43 ( 1996 ) . ,
_ colloids and surfaces a _ * 156 * , 315 ( 1999 ) .
rep . _ * 58 * , 111 ( 2005 ) . , _
j. colloid interface sci .
_ * 120 * , 76 ( 1987 ) .
, _ langmuir _ * 4 * , 489 ( 1988 ) . , _ phys . rev . lett . _ * 72 * , 1698 ( 1994 ) . , _
j. colloid interface sci .
_ * 169 * , 256 ( 1995 ) . , _ langmuir _ * 13 * , 581 ( 1995 ) . , in _ applied surface thermodynamics _
neumann and j.k .
spelt ( marcel dekker , new york , 1996 ) , p.168 . , _ phys . rev .
lett . _ * 83 * , 3677 ( 1999 ) . ,
_ colloids and surfaces a _ * 156 * , 323 ( 1999 ) . ,
_ j. adhesion sci .
technol . _
* 13 * , 1155 ( 1999 ) .
lett . _ * 85 * , 1930 ( 2000 ) . , _ langmuir _ * 16 * , 2024 ( 2000 ) . , _ j. phys . : condens
* 12 * , a57 ( 2000 ) . ,
_ j. adhesion sci .
technol . _
* 14 * , 1767 ( 2000 ) . , _ phys .
* 63 * , 031601 ( 2001 ) . ,
_ j. phys . :
matter _ * 13 * , 4915 ( 2001 ) .
lett . _ * 89 * , 076102 ( 2002 ) . ,
_ langmuir _ * 18 * , 9771 ( 2002 ) . , _ phys .
lett . _ * 91 * , 186101 ( 2003 ) . ,
_ langmuir _ * 22 * , 116 ( 2006 ) . , _
j. colloid interface sci . _ * 265 * , 152 ( 2003 ) . , _ adv .
colloid interface sci .
_ * 114115 * , 103 ( 2005 ) . , _
j. colloid interface sci . _ * 142 * , 224 ( 1991 ) .
, _ langmuir _ * 13 * , 581 ( 1997 ) . , _ colloids and surfaces a _ * 129 - 130 * , 45 ( 1997 ) . ,
_ langmuir _ * 13 * , 3035 ( 1997 ) . ,
_ colloids and surfaces a _ * 146 * , 95 ( 1999 ) . , _ langmuir _ * 21 * , 8594 ( 2005 ) . , _ j. chem .
, faraday trans .
i _ * 72 * , 2815 ( 1976 ) . , _
j. colloid interface sci . _
* 75 * , 612 ( 1980 ) . , _
j. colloid interface sci . _ * 112 * , 97 ( 1986 ) ; p.a .
kralchevsky , i.b .
ivanov , and a.d .
nikolov , ibid * 112 * , 108 ( 1986 ) ; a.d .
nikolov , p.a .
kralchevsky , i.b .
ivanov , ibid * 112 * , 122 ( 1986 ) ; p.a .
kralchevsky , a.d .
nikolov , and i.b .
ivanov , ibid * 112 * , 132 ( 1986 ) . , _
j. colloid interface sci . _ * 151 * , 462 ( 1991 ) . ,
_ j. chem .
, faraday trans .
i _ * 91 * , 2681 ( 1995 ) . , _
j. colloid interface sci . _ * 208 * , 468 ( 1998 ) . ,
_ colloids and surfaces a _ * 156 * , 519 ( 1999 ) . ,
_ j. adhesion sci .
technol . _
* 14 * , 1783 ( 2000 ) . , _ colloid journal _ * 63 * , 518 ( 2001 ) . , _ phys .
lett . _ * 90 * , 128304 ( 2001 ) . , _ appl
a _ * 80 * , 445 ( 2005 ) . , _ j. chem .
* 26 * , 23 ( 1957 ) .
chem . _ * 46 * , 14 ( 1961 ) . , _ j. chem .
phys . _ * 45 * , 3160 ( 1966 ) . ,
_ j. chem .
phys . _ * 66 * , 5464 ( 1977 ) . ,
_ langmuir _ * 12 * , 5956 ( 1996 ) . , _ j. colloid interface sci . _ * 186 * , 462 ( 1997 ) . , _ colloids and surfaces a _ * 156 * , 423 ( 1999 ) . , _
j. colloid interface sci . _ * 218 * , 122 ( 1999 ) . ,
_ j. phys . : condens .
* 12 * , 9221 ( 2000 ) . ,
_ colloids and surfaces a _ * 161 * , 23 ( 2000 ) .
, _ langmuir _ * 18 * , 8919 ( 2002 ) . ,
_ physica a _ * 319 * , 151 ( 2003 ) . , _
j. colloid interface sci . _
* 261 * , 464 ( 2003 ) . , _ j. phys . a : math .
gen . _ * 37 * , 11547 ( 2004 ) . ,
_ j. phys . : condens
* 17 * , s537 ( 2005 ) . , _ langmuir _ * 22 * , 11041 ( 2006 ) . ,
_ j. phys . a : math .
_ * 40 * , 19 ( 2007 ) . , _ j. chem
. phys . _ * 75 * , 2441 ( 1981 ) . , _ j. chem .
phys . _ * 75 * , 3114 ( 1981 ) . ,
_ physica a _ * 117 * , 575 ( 1983 ) . , _
j. colloid interface sci . _ * 111 * , 94 ( 1986 ) .
, _ physica a _ * 168 * , 149 ( 1990 ) . ,
_ physica a _ * 173 * , 72 ( 1991 ) . , _ physica a _ * 183 * , 12 ( 1991 ) .
phys . _ * 80 * , 729 ( 1993 ) . ,
_ physica a _ * 251 * , 25 ( 1998 ) .
, _ langmuir _ * 15 * , 2586 ( 1999 ) . , _ j. stat
. phys . _ * 126 * , 133 ( 2007 ) . ,
phys . _ * 75 * , 925 ( 1992 ) .
, _ physica a _ * 183 * , 439 ( 1992 ) .
a _ * 45 * , 2645 ( 1992 ) . , _ physica a _ * 196 * , 335 ( 1993 ) .
e _ * 47 * , 4607 ( 1993 ) .
, _ physica a _ * 201 * , 457 ( 1993 ) . ,
_ j. phys .
i france _ * 4 * , 77 ( 1994 ) .
, _ europhys .
lett . _ * 25 * , 17 ( 1994 ) .
e _ * 52 * , 1223 ( 1995 ) . ,
_ j. chem .
phys . _ * 102 * , 400 ( 1995 ) . ,
_ physica a _ * 358 * , 492 ( 2005 ) . , _ j. phys . :
condens . matter _
* 16 * , 6917 ( 2004 ) .
lett . _ * 87 * , 106101 ( 2001 ) . ,
_ j. chem .
phys . _ * 102 * , 7584 ( 1995 ) . , _ phys
lett . _ * 95 * , 226107 ( 2005 ) . ,
* 22 * , 787 ( 2005 ) . ,
_ j. chem .
phys . _ * 110 * , 3536 ( 1998 ) . ,
_ j. phys .
b _ * 107 * , 1345 ( 2003 ) . , _
j. chem . phys . _ * 123 * , 184704 ( 2005 ) . ,
_ j. chem .
* 17 * , 333 ( 1949 ) . , _ j. chem .
* 19 * , 1591 ( 1951 ) . ,
_ j. phys .
chem . _ * 56 * , 526 ( 1952 ) .
, in _ encyclopedia of physics _ , vol . * x * , ed . by s. flgge ( springer , berlin , 1960 ) , p. 134 .
, in _ fluid interfacial phenomena _ , ed .
by c. a. croxton ( wiley , new york , 1986 ) , p. 555 . , _ statistical physics of fluids , basic concepts and applications _ ( springer , berlin , 2001 ) . , _ physica a _ * 184 * , 42 ( 1992 ) . , _ j. chem .
phys . _ * 100 * , 556 ( 1994 ) .
e _ * 53 * , 881 ( 1997 ) . ,
_ j. chem .
phys . _ * 108 * , 1148 ( 1998 ) . ,
_ j. chem .
* 109 * , 4063 ( 1998 ) . ,
_ j. chem .
_ * 109 * , 9901 ( 1998 ) . ,
_ j. chem .
phys . _ * 109 * , 9660 ( 1998 ) . , _ phys
e _ * 59 * , 469 ( 1999 ) . ,
_ j. chem .
phys . _ * 111 * , 5938 ( 1999 ) . , _ langmuir _ * 16 * , 3502 ( 2000 ) . , _ j. chem
. phys . _ * 111 * , 3705 ( 1999 ) ; ibid * 111 * , 10602 ( 1999 ) ; _ j. phys .
b _ * 105 * , 11586 ( 2001 ) . , _ j. phys
b _ * 105 * , 1050 ( 2001 ) . ,
_ j. chem .
phys . _ * 114 * , 5796 ( 2001 ) . ,
_ j. phys .
b _ * 105 * , 11615 ( 2001 ) . ,
_ j. chem .
phys . _ * 116 * , 302 ( 2002 ) . ,
_ colloids and surfaces a _ * 250 * , 263 ( 2004 ) . ,
_ colloids and surfaces a _ * 239 * , 141 ( 2004 ) . , _ j. chem .
phys . _ * 121 * , 5602 ( 2004 ) . ,
phys . _ * 103 * , 647 ( 2005 ) . ,
_ j. phys .
b _ * 110 * , 22125 ( 2006 ) . ,
phys . _ * 104 * , 3469 ( 2006 ) . ,
_ physica a _ * 372 * , 169 ( 2006 ) .
b _ * 46 * , 7312 ( 1992 ) . , _ phys . rev . lett . _ * 71 * , 404 ( 1993 ) . ,
_ j. chem .
phys . _ * 96 * , 9073 ( 1992 ) . , _ phys
e _ * 70 * , 031603 ( 2004 ) .
e _ * 73 * , 021602 ( 2006 ) . , _ colloids and surfaces a _ * 116 * , 25 ( 1996 ) . , _ phys .
e _ * 47 * , 1836 ( 1993 ) . ,
_ j. low temp .
* 101 * , 331 ( 1995 ) .
t. bieker and s. dietrich , _ physica a _ * 252 * , 85 ( 1998 ) ; * 259 * , 466 ( 1998 ) . ,
_ physica a _ * 202 * , 402 ( 1994 ) .
lett . _ * 98 * , 035702 ( 2007 ) . ,
phys . _ * 86 * , 891 ( 1995 ) . ,
_ j. chem .
phys . _ * 112 * , 5985 ( 2000 ) . ,
_ j. chem .
phys . _ * 81 * , 530 ( 1984 ) .
lett . _ * 93 * , 160601 ( 2004 ) . , _ europhys
. lett . _ * 69 * , 832 ( 2005 ) . , _ phys
e _ * 74 * , 041404 ( 2006 ) .
lett . _ * 97 * , 078101 ( 2006 ) . , _ vorlesungen ber inhalt , oberflche und isoperimetrie _ ( springer , berlin , 1957 )
. , _ lehrgang der hheren mathematik ii , v 13 _ ( veb deutscher verlag der wissenschaften , berlin , 1977 ) . | we analyze two representative systems containing a three - phase - contact line : a liquid lens at a fluid fluid interface and a liquid drop in contact with a gas phase residing on a solid substrate .
in addition we study a system containing a planar liquid gas interface in contact with a solid substrate .
we discuss to which extent the decomposition of the grand canonical free energy of such systems into volume , surface , and line contributions is unique in spite of the freedom one has in positioning the gibbs dividing interfaces . before being able to define the line contribution to the grand canonical free energy per unit length , the so - called line tension , first certain properties have to be attributed to the bulk phases and to the interfaces ;
these are chosen in agreement with what is known for spherical drops and for planar interfaces . in the case of a lens
it is found that the line tension is independent of arbitrary choices of the gibbs dividing interfaces . in the case of a drop
, however , one arrives at two different possible definitions of the line tension .
one of them corresponds seamlessly to that applicable to the lens .
the line tension defined this way turns out to be independent of choices of the gibbs dividing interfaces . in the case of the second definition ,
however , the line tension does depend on the choice of the gibbs dividing interfaces .
we also provide equations for the equilibrium contact angles which are form - invariant with respect to notional shifts of dividing interfaces which change the description of the system but leave the density configurations unchanged . in order to reconcile the transformation laws for the line tensions under notional changes of dividing interfaces , which follow from the decomposition of the grand canonical free energy , with the principle of form invariance and moreover with the invariance of observables under notional changes , additional stiffness constants attributed to the line
must be introduced .
the choice of the dividing interfaces influences the actual values of the stiffness constants .
we show how these constants transform as a function of the relative displacements of the dividing interfaces .
our formulation provides a clearly defined scheme to determine line properties from measured dependences of the contact angles on lens or drop volumes .
this scheme implies relations different from the modified neumann or young equations , which currently are the basis for extracting line tensions from experimental data .
these relations show that the experiments do not render the line tension alone but a combination of the line tension , the tolman length , and the stiffness constants of the line .
in contrast to previous approaches our scheme works consistently for any choice of the dividing interfaces .
it further allows us to compare results obtained by different experimental or theoretical methods , based on different conventions of choosing the dividing interfaces . |
since the seminal paper by yang and mills @xcite , the study of nonabelian gauge theories has attracted a lot of interest in several sectors of physics . focusing on the gauge theory of strong interactions ( qcd ) , thanks to improvement of computer facilities which allow to perform huge simulations of yang - mills theories , as well as to the possibility to create in laboratories extremely hot environments by means of heavy ion collisions , the interest in understanding the thermodynamical properties of strong nonabelian mediums has noticeably increased .
lattice simulations of @xmath0 yang - mills theories have been performed , see for example refs .
all of these studies agree about the onset of the deconfinenemt phase transition at @xmath2 mev .
below @xmath3 , the thermodynamics of the yang - mills theory is dominated by the lowest lying glueballs . for what concerns
the high temperature phase , the picture is not so clear . as a matter of fact , one of the most intriguing aspects is the nonperturbative nature of the gluon medium above the critical temperature @xcite .
for example , in @xcite it is pointed out that the perturbative regime is realized for temperatures @xmath4 ( more concretely , at @xmath5 both pressure and energy density are below the stefan - boltzmann limit , which corresponds to the ideal gas case , by about @xmath6 ) .
this makes the identification of the correct degrees of freedom of the gluon plasma , in proximity as well as well beyond the critical temperature , a very complicated task .
resummation schemes have been proposed , based for example on the hard thermal loop ( htl ) approach @xcite . more recently within a nnlo in the htl perturbation theory @xcite it has been shown that an agreement with the lattice results can be obtained also for the trace anomaly down to @xmath7 but with large uncertainties . at these high temperatures ,
the htl approach motivates and justifies a picture of weakly interacting quasi - particles , as determined by the htl propagators . within this theoretical framework ,
the quasiparticle approach to the thermodynamics of qcd has attracted a discreet interest recently , see for example refs .
@xcite and references therein .
in such an approach , one identifies the degrees of freedom of the deconfinement phase with transverse gluons ; the strong interaction in such a non perturbative regime is taken into account through an effective temperature - dependent mass for the gluons themselves .
this theoretical approach is similar to the one introduced by landau @xcite to describe strongly correlated systems . generally speaking ,
it is not possible to compute the self - energy of gluons exactly in the range of temperature of interest ; as an obvious consequence , the location of the poles of the propagator in the complex momentum plane , hence ( roughly speaking ) the gluon mass , is an unknown function .
for this reason , one usually assumes an analytic dependence of the gluon mass on the temperature , leaving few free parameters which are then fixed by fitting the thermodynamical data of lattice simulations .
the advantage of such an approach is , at least , twofolds .
firstly , it is of a theoretical interest in itself to understand which are the effective degrees of freedom of the yang - mills theory at finite temperature and in particular the evolution of non perturbative effects with temperature .
furthermore , once a microscopic description of the plasma is identified , it is also possible to include the implied dynamics into a transport theory capable to directly simulate the expanding fireball produced in heavy ion collisions computing the collective properties , as well as the chemical composition of the fireball as a function of time @xcite . in this article
we extend the quasiparticle picture of the finite temperature gluon medium , by adding an interaction of gluons with a background polyakov loop .
it is well known that the latter is an exact order parameter for the confinement - deconfinement transition in the pure gauge theory @xcite .
therefore , having access to the polyakov loop immediately leads us to the possibility to study the effect of the phase transition on the dynamical properties of gluons , mainly on their mass .
the standard quasiparticle approach , instead , tries to account for all the dynamics only by mean of temperature dependent masses @xcite .
this always leads to the necessity of diverging or steadily increasing masses as @xmath8 not only of the case of pure gauge theory , but also for the qcd case which includes quarks .
we show that combining a t - dependent quasiparticle mass , @xmath9 , with the polyakov loop dynamics results in a quite different behavior of @xmath9 as @xmath8 . the combined approach of the quasiparticle picture with the polyakov loop thermodynamics has been poorly investigated . to our knowledge
, there is only a first preliminary study about the role of the polyakov loop in the context of the quasiparticle description of the gluon plasma @xcite , which has some overlap with our study . in the present article
, we take a simple matrix model action as the effective potential for the polyakov loop . then , we use the weiss mean field procedure to compute averages .
such a procedure has been used in the study of the qcd phase diagram by means of the polyakov extended nambu - jona lasinio model @xcite , see @xcite . in the next section
we describe the technical details , it suffices to say here that the main difference between this procedure and the one commonly used in the literature is that , in the former case , averages are computed in terms of integrals over the gauge group with the appropriate invariant gauge measure having the pure gauge mean field action as the integration weight
. we can anticipate that in such a way we are not only able to reproduce the large trace anomaly , @xmath10 , with a quite good accuracy , but also identify the different roles of quasiparticles masses and polyakov loop in yang - mills thermodynamics .
the paper is organized as follows . in section
ii , we describe the matrix model in the weiss mean field potential and its coupling to quasiparticle masses , showing the explicit derivation for @xmath0 gauge group . in section iii , we compare the results of our approach to the lqcd thermodynamics discussing in particular the specific contributions to the pressure and to the interaction measure of the polyakov mean filed and of the quasiparticle masses . in section
iv we draw first conclusions of our work along with future possible investigations and developments of the present approach .
the system we consider in this article consists of a gas of gluon quasiparticles , propagating in a background of polyakov loop .
the free energy of the model is expressed as a linear combination of two contributions : the first one describes the thermodynamics of the polyakov loop ; the second one , on the other hand , is the contribution of gluon quasiparticles coupled to the polyakov loop .
we specify the two terms of the effective potential below . in this section
we follow closely the notations of @xcite . the starting point is the action of a matrix model @xcite , @xmath11 = -n_c^2 e^{-a / t}\bm\sum_{\bm x,\bm y } \ell_f(\bm x)\ell_f^*(\bm x + \bm y)~ , \label{eq : sc1}\ ] ] where @xmath12 denote the lattice site and its nearest neighbors respectively .
moreover , @xmath13 corresponds to the traced polyakov loop in the fundamental representation . the polyakov line in the representation @xmath14 of the gauge group is given by @xmath15~ , \label{eq : rrr}\ ] ] where @xmath16 are the gauge group generators in the representation @xmath14 .
the model specified by eq . corresponds to a matrix model , the untraced polyakov loops @xmath17 corresponding to the dynamical degrees of freedom .
the partition function can be written as @xmath18)~. \label{eq : pf}\ ] ] here , @xmath19 is the group invariant measure on the @xmath20 gauge group . it can be written as @xmath21 and the integration over the variables @xmath22 is performed over the range @xmath23 $ ] . adopting the weiss mean field procedure , the action in eq .
can be written as @xmath24~ , \label{eq : mfact}\ ] ] where @xmath25 denote the mean fields of the polyakov loop and of its conjugate respectively . in terms of @xmath26 ,
the average of an operator @xmath27 , namely @xmath28 , is defined as @xmath29e^{-s_{mf } } } { \int\bm\prod_{\bm x } dl(\bm x ) e^{-s_{mf}}}~. \label{eq : mf}\ ] ] the thermodynamic potential @xmath30 can be written by adding and subtracting the mean field action @xmath26 in the exponent of eq . .
we have @xmath31 where the average has to be performed as in eq . .
equation is exact . in the mean field approximation
, we replace @xmath32 with @xmath33 . in this way
, the thermodynamic potential depends explicitely on the mean fields @xmath25 , and their values are determined selfconsistently by the stationarity condition . in the mean field approximation the thermodynamic potential can thus be written as @xcite @xmath34 \nonumber \\ & & -\log\int dl~e^{n_c\left[\alpha\re\ell_f + i\beta\im\ell_f\right]}~. \label{eq : mfw}\end{aligned}\ ] ] in the previous equation , @xmath35 corresponds to the spacetime dimension , so that @xmath36 is the coordination number of the cubic lattice , that is the number of nearest neighbors of the site @xmath37 .
furthermore , @xmath38 is the lattice spacing , and @xmath39 is the volume of the lattice ; as a consequence , @xmath40 is the number of lattice sites .
the thermal average of the polyakov loop and of its conjugate are given by @xmath41 } { \int dl~e^{n_c\alpha\re\ell_f}\cos(n_c\beta\im\ell_f)}~,\\ \langle\ell_f^*\rangle & = & \frac{\int dl~e^{n_c\alpha\re\ell_f } \left[\cos(n_c\beta\im\ell_f)\re\ell_f + \sin(n_c\beta\im\ell_f)\im\ell_f\right ] } { \int dl~e^{n_c\alpha\re\ell_f}\cos(n_c\beta\im\ell_f)}~ ; \label{eq : bla}\end{aligned}\ ] ] from the above equations it is then clear that @xmath42 and @xmath43 are different when @xmath44 .
this happens for example in theories with fermions at finite baryon chemical potential . in the following we focus on the case of the pure gauge theory , in which @xmath45 .
we have thus @xmath46 , and eq .
can be written as @xmath47 -\log\int dl~e^{n_c\left[\alpha \re \ell_f \right]}\right)~ , \label{eq : mfwreal}\end{aligned}\ ] ] where we have used the shorthand notation @xmath48 . in the following , @xmath49 and @xmath50 in the above equation
are treated as free parameters , which will be fixed , together with the remaining two parameters of the model which will be described in the next section , by a minimization procedure of the mean quadratic deviation between the theoretical computations and the lattice data or pressure , energy density and interaction measure . for the sake of future reference
, we call the model described by the above thermodynamic potential as the pure matrix model ( pmm ) . the gluon quasiparticle contribution to
the thermodynamic potential reads @xcite @xmath51 in the above equation , @xmath52 corresponds to the polyakov line in the adjoint representation , as defined in eq . , and the trace is taken over the indices of the adjoint representation of the gauge group .
moreover , the quasiparticle energy is given by @xmath53 , where @xmath54 is supposed to arise from non - perturbative medium effects .
in the case of very high temperature , where the gauge theory is in the perturbative regime and thus @xmath55 , it can be proved that eq .
is the effective potential for the adjoint polyakov loop @xcite . at lower temperatures , where non - perturbative effects are important , eq . must be postulated as a starting point for a phenomenological description of the thermodynamics of the gluon plasma . in this article , we make use of a temperature depentent mass whose analytic form is given by @xcite @xmath56 ^ 2}~. \label{eq : mass}\ ] ] the parameters @xmath57 are determined by requiring that , for pressure and energy density , the mean quadratic deviation between our theoretical computation and the lattice data of @xcite is a minimum . in particular , we find @xmath58 mev and @xmath59 , see section iii .
the temperature dependent quasi - particle mass in eq .
has several motivations : firstly , in the very large temperature regime in which qcd is perturbative , one has @xmath60 ; moreover , for @xmath61 close to @xmath3 there are remnants of the confining mechanism and , a priori , there is no reason to reduce these effects to the polyakov loop dynamics ; finally , for @xmath62 , the glueball mass decreases @xcite , @xmath63 gev at @xmath64 , and in a valence gluon model one expects an effective mass of the quasiparticle @xmath65 mev at the melting temperature . in order to have a combined description of the gluon plasma around the critical temperature , in terms of the polyakov loop ( which acts as a background field ) and of gluons quasiparticles ( which propagate in the polyakov loop background ) , we add eq . to the strong coupling inspired potential in eq . .
in order to do this , and to be consistent at the same time with the weiss mean field procedure outlined above , we follow the lines of @xcite and write the thermodynamic potential as @xmath66 where the average must be understood as defined in eq . .
in taking the average , we use the same approximation of @xcite and write @xmath67 this approximation is also discussed in the appendix b of @xcite , where it is proved that taking the average of the log ( that is , @xmath68 ) instead of the argument of the log , would lead to unphysical artifacts in the case the particles are taken in representations different from the fundamental one .
we will discuss this aspect in more detail in the next section .
in the case of the color group @xmath0 the polyakov line can be parametrized in terms of two angles , @xmath69 and @xmath70 , namely @xmath71 ; the invariant haar measure reads : @xmath72 ^ 2~ , \label{eq : haarsu3}\ ] ] with @xmath73 .
we define @xmath74 where @xmath75 moreover we have @xmath76 the thermodynamic potential for the pure matrix model in this case then reads @xmath77 the quasiparticle contribution for the @xmath0 case in the mean field approximation reads @xmath78 where the determinant is taken in the adjoint color representation , and the adjoint polyakov line is given by @xmath79 $ ] , with @xmath80 the appropriate generators in the adjoint representation . we make use of the following equation to express the adjoint loop in terms of the fundamental one : @xmath81~ , \label{eq : marco1}\ ] ] where @xmath82 , @xmath83 correspond to the generators of @xmath0 in the fundamental representation , normalized with the condition @xmath84 = \delta_{ab}/2 $ ] . to compute the functional determinant in eq . , we chose a representation in which @xmath52 is diagonal ; the eigenvalues of @xmath52 are @xmath85 , @xmath86 , @xmath87 , and finally @xmath88 , with @xmath73 .
as a result we find @xmath89 where we have defined @xmath90 and the @xmath91 denotes the average over the gauge group elements as defined as in eq . .
our expression for the functional determinant is in agreement with eq .
( 4.5 ) of @xcite , once we put @xmath92 and change @xmath93 in eq . to take into account
that in @xcite , fermions in the adjoint representation are considered instead of bosons .
it can be proved analitically that @xmath94 moreover , numerically we have checked that @xmath95 for the sake of reference , we write down the equation for the traced fundamental and adjoint polyakov loops in terms of the angles @xmath69 and @xmath70 , namely @xmath96 we have checked numerically that @xmath97 in the asymptotic limit @xmath98 and @xmath99 for @xmath100 ( the latter step can be proved analitically as well ) .
the asymptotic behavior of the averages of the @xmath101 in eq .
implies that in the deconfinement phase , at very large temperature where @xmath102 , the thermal distribution of the quasigluons approaches that of a perfect gas of massive particles .
however the thermodynamics of the system remains different from the one of a mere massive gas because of the polyakov background mean field .
_ upper panel . _
pressure ( indigo , solid line ) and energy density ( orange , dashed line ) , normalized to the respective stefan - boltzmann values , as a function of temperature ( measured in units of the critical temperature @xmath103 mev ) , for the su(3 ) case .
dots correspond to lattice data taken from ref .
@xcite ; solid lines are the result of our numerical computation within the quasiparticle model . _
lower panel .
_ interaction measure per degree of freedon as a function of temperature .
the dashed orange line corresponds to the result of ref .
@xcite.,title="fig:",width=283 ] + _ upper panel . _
pressure ( indigo , solid line ) and energy density ( orange , dashed line ) , normalized to the respective stefan - boltzmann values , as a function of temperature ( measured in units of the critical temperature @xmath103 mev ) , for the su(3 ) case .
dots correspond to lattice data taken from ref .
@xcite ; solid lines are the result of our numerical computation within the quasiparticle model . _
lower panel .
_ interaction measure per degree of freedon as a function of temperature .
the dashed orange line corresponds to the result of ref .
@xcite.,title="fig:",width=283 ] in fig .
[ fig : compsu3 ] we plot the pressure @xmath104 ( red data ) and energy density ( green data ) , normalized to the stefan - boltzmann value , as a function of temperature ( measured in units of the critical temperature ) , for the su(3 ) case .
given the pressure , the energy density is computed by virtue of the thermodynamical relation performing the pertinent total derivative of the temperature , @xmath105 .
of the four parameters in our model , one of them is fixed in order to reproduce the first order phase transition at @xmath106 mev ; the remaining three parameters are fixed in order to require that the mean quadratic deviation for pressure , energy density and interaction measure between our computation and the lattice data is minimized .
this procedure leads to the numerical values @xmath107 mev , @xmath108 mev@xmath109 , @xmath58 mev and finally @xmath59 .
these parameters produce a gluon mass @xmath110mev at @xmath111 . in the figure , @xmath112 and @xmath113 where @xmath114 is the dimension of the adjoint representation .
dots correspond to lattice data taken from ref .
@xcite ; solid lines are the result of our numerical computation within the quasiparticle model . in the lower panel of fig .
[ fig : compsu3 ] we compare our results for the interaction measure , @xmath115 , with the lattice data ( represented by dots in the figure ) .
we notice that our model reproduces fairly well the peak of the interaction measure observed on the lattice in the critical region .
we also compare our results with the ones obtained in @xcite , where a phenomenological potential for the polyakov loop is considered instead of our matrix model , and the usual mean field approximation is used .
the comparison shows that the peak is much better reproduced in our case .
however one could argue that the tail at high @xmath61 in instead not so well reproduced ; in the next section we will see that such a drawback can be cured by a slightly different @xmath116 behavior . _ upper panel_. ratio @xmath117 for our model ( indigo , solid line ) and for the model without polyakov loop ( orange , dashed line ) of @xcite , against temperature , the latter measured in units of the critical temperature @xmath106 mev . _
lower panel_. expectation value of the fundamental ( indigo , solid line ) and adjoint ( orange , dashed line ) loops against temperature.,title="fig:",width=283 ] + _ upper panel_. ratio @xmath117 for our model ( indigo , solid line ) and for the model without polyakov loop ( orange , dashed line ) of @xcite , against temperature , the latter measured in units of the critical temperature @xmath106 mev . _
lower panel_. expectation value of the fundamental ( indigo , solid line ) and adjoint ( orange , dashed line ) loops against temperature.,title="fig:",width=283 ] in the upper panel of fig .
[ fig : masses ] we plot our result for the gluon mass against temperature , and we compare it with the result of @xcite obtained without the polyakov loop dynamics . in ref .
@xcite the following ansatz is used : @xmath118 with @xmath119 .
numerical values of the parameters are @xmath120 , @xmath121 , @xmath122 and @xmath123 .
the first addendum on the r.h.s . of the above equation , which is interpreted as
the energy contained in the correlation volume , is important close to @xmath3 , and it is necessary to have a large mass as @xmath61 approaches @xmath3 from above .
in fact , when the polyakov loop is not introduced , a large mass is needed in order to suppress pressure and energy density around @xmath3 . in the case
the phase transition is of the second order , as it happens for @xmath124 , correlation length diverges at @xmath3 ; therefore , @xmath125 since the mass diverges as well .
when the phase transition is of the first order , the correlation lenght ( hence the correlation volume ) does not diverge at @xmath3 , but it is expected to be large ; as a consequence , the mass should increase as @xmath61 approaches @xmath3 from above .
for this reason @xmath126 in eq .
is slightly less than unity @xmath127 .
the second addendum on the r.h.s . of eq
. dominates the mass at large temperature , and its origin is of perturbative nature .
_ upper panel .
_ full pressure of our model ( indigo , solid line ) , perfect gas pressure ( orange , dashed line ) and quasiparticle contribution ( teal , dot dashed line ) against temperature .
all the data are measured in units of the stefan - boltzmann pressure . _
. _ interaction measure of our model ( solid , indigo line ) compared with the pure quasiparticle model ( orange , dashed line ) . in the figure
, the dot dashed teal line corresponds to the interaction measure we would obtain neglecting the derivative of the gluon mass , see the text for more detail.,title="fig:",width=283 ] + _ upper panel . _
full pressure of our model ( indigo , solid line ) , perfect gas pressure ( orange , dashed line ) and quasiparticle contribution ( teal , dot dashed line ) against temperature .
all the data are measured in units of the stefan - boltzmann pressure . _
. _ interaction measure of our model ( solid , indigo line ) compared with the pure quasiparticle model ( orange , dashed line ) . in the figure
, the dot dashed teal line corresponds to the interaction measure we would obtain neglecting the derivative of the gluon mass , see the text for more detail.,title="fig:",width=283 ] in our case , the statistical suppression of states below and around @xmath3 is achieved by virtue of the polyakov loop .
for this reason , we do not need to have a large mass as @xmath61 approaches @xmath3 .
this mechanism is similar to the statistical confinement that takes place in the pnjl model @xcite . to show this aspect of the model clearly , we have plotted in fig . [
fig : yulia ] the full pressure obtained within our model ( indigo solid line ) , and we have compared this result with the pressure of a perfect gas having the temperature dependent mass we have obtained in our computations ( orange dashed line ) . from the data shown in the figure ,
the role of the polyakov loop to suppress thermodynamically the contribution of states is noticeable . for completeness
, we have also plotted the quasiparticle gas contribution to the pressure , @xmath128 , with polyakov loop taken into account ( the latter has been selfconsistently computed at any temperature ) .
the role of the polyakov loop is apparent also in the lower panel of fig .
[ fig : yulia ] , where we compare our results for the interaction measure , which has polyakov loop thermodynamics built into it , and the interaction measure for a gas of quasiparticle with mass given by eq . .
for a matter of simplicity , the latter is estimated by using the relativistic boltzmann distribution function for gluons , namely @xcite @xmath129 k_1(m / t)~ , \label{eq : bolx}\ ] ] the analytical expression allows to identify the two contributions to @xmath130 , we have verified that turning to the bose - einstein statistics does not lead to significant different results .
it is evident that the polyakov loop gives a relevant contribution to the interaction measure .
moreover , it is instructive to show the concrete effect of a temperature dependent mass on the interaction measure . to this end , in the figure we have plotted eq .
in which we neglect the term proportional to @xmath131 : the result is shown by the teal line .
since @xmath131 in our case is positive , its contribution suppresses the interaction measure , as it is clear from the data shown in the figure . _ upper panel .
_ normalized pressure and energy density against temperature .
_ middle panel .
_ interaction measure per degree of freedon as a function of temperature .
_ lower panel .
_ the ratio @xmath117 against temperature.,title="fig:",width=283 ] + _ upper panel .
_ normalized pressure and energy density against temperature .
_ middle panel .
_ interaction measure per degree of freedon as a function of temperature .
_ lower panel .
_ the ratio @xmath117 against temperature.,title="fig:",width=283 ] + _ upper panel
. _ normalized pressure and energy density against temperature .
_ middle panel .
_ interaction measure per degree of freedon as a function of temperature .
_ lower panel .
_ the ratio @xmath117 against temperature.,title="fig:",width=283 ] as a final investigation , we have considered another ansatz for the thermal gluon mass , that is @xmath132 where @xmath133 mev and @xmath134 mev
( these two coefficients have been fixed once again by a minimization procedure of the quadratic relative deviations ) .
the results are summarized in fig .
[ fig : compsu3lin ] .
few comments follow .
secondly , the peak of the interaction measure is slightly suppressed : in this case the suppression is of about @xmath6 , which has to be compared with the case of the logharithmic mass ( less than @xmath135 ) .
however the tail of @xmath136 is better reproduced , because of the combined effect of mass , polyakov loop and @xmath131 .
the role of the polyakov loop in avoiding the divergence of the gluon mass as @xmath3 is approached from above remains .
we leave a more systematic study of the several possible ansatzes to a future work .
in this article , we have studied a combined description of the deconfinement phase of the @xmath0 gluon plasma , in terms of gluon quasiparticles and polyakov loop .
the two are mixed by a phenomenological potential , see eq .
, which is inspired by weak coupling computations @xcite .
to the quasiparticle contribution , we have added a potential for the polyakov loop given by eq . ; its analytic form is inspired by the leading order strong coupling expansion , and it has been used in the literature in the context of the pnjl model @xcite .
our main purpose is twofold .
firstly , we are interested to a simple description of the lattice data about the thermodynamics of the gluon plasma .
this is interesting because it allows to understand which are the relevant degrees of freedom in the deconfinement phase of the theory .
moreover , once the proper degrees of freedom are identified , the effective description studied here can be completed by adding dynamical quarks .
employing a relativistic transport theory , this can allow also a direct connection between the developments of effective models and the study of the quark - gluon plasma created in relativistic heavy - ion collisions .
within our model , we are able to reproduce fairly well the lattice data about the gluon plasma thermodynamics in the critical region .
in particular , we reproduce the peak of the interaction measure within a 2% , see lower panel in fig .
[ fig : compsu3 ] .
one of the main conclusions of our work is that because of the coupling to the polyakov loop , a gluon mass of the order of the critical temperature , @xmath137 , is enough to reproduce the lattice data in the critical region .
this is different from what is found in standard quasi - particle models .
the polyakov loop is neglected and the gluon plasma is described in terms of a perfect gas of massive gluons . as a consequence ,
a mass rapidly increasing with @xmath61 is needed to suppress pressure and energy density in the critical region . on the other hand , in the case studied in this article , the suppression of the pressure and energy density is mainly caused by the polyakov loop , as shown in fig .
[ fig : yulia ] .
therefore , lighter quasiparticles with a smooth @xmath138dependence describe fairly well the lattice data in the critical region .
there are several interesting directions to follow to extend the present work .
firstly , having in mind the study of the quark - gluon plasma , we plan to add dynamical massive quarks to the picture .
it is also of a certain interest to span the parameter space in more detail , eventually analyzing several functional forms of the gluon mass .
moreover , even in the case of the pure gauge theory , it would be interesting to investigate the way to adjust the interaction measure in order to reproduce lattice data in the regime of very large temperature .
even more , it would be of a certain interest to extend our study to the case of different gauge groups . c. -n .
yang and r. l. mills , phys .
* 96 * , 191 ( 1954 ) .
s. datta and s. gupta , phys .
d * 82 * , 114505 ( 2010 ) [ arxiv:1006.0938 [ hep - lat ] ] .
g. boyd , j. engels , f. karsch , e. laermann , c. legeland , m. lutgemeier and b. petersson , nucl .
b * 469 * , 419 ( 1996 ) [ hep - lat/9602007 ] .
g. boyd , j. engels , f. karsch , e. laermann , c. legeland , m. lutgemeier and b. petersson , phys .
lett . * 75 * , 4169 ( 1995 ) [ hep - lat/9506025 ] .
s. .borsanyi , g. endrodi , z. fodor , s. d. katz and k. k. szabo , arxiv:1104.0013 [ hep - ph ] .
d.zwanziger , nucl .
* b 485*(1997)185 .
e. braaten and r. d. pisarski , phys .
d * 45 * , r1827 ( 1992 ) .
j. p. blaizot and e. iancu , nucl .
b * 417 * , 608 ( 1994 ) .
j. o. andersen , e. braaten , and m. strickland , phys .
lett . * 83 * , 2139 ( 1999 ) .
j. p. blaizot , e. iancu , and a. rebhan , phys .
lett . * 83 * , 2906 ( 1999 ) .
n. meisinger , m. c. ogilvie and t. r. miller , phys .
b * 585 * , 149 ( 2004 ) [ hep - ph/0312272 ] .
p. castorina , v. greco , d. jaccarino and d. zappala , eur .
j. c * 71 * , 1826 ( 2011 ) [ arxiv:1105.5902 [ hep - ph ] ] . f. giacosa , phys .
d * 83 * , 114002 ( 2011 ) [ arxiv:1009.4588 [ hep - ph ] ] .
a. peshier and w. cassing , phys .
lett . * 94 * , 172301 ( 2005 ) [ hep - ph/0502138 ] .
s. plumari , w. m. alberico , v. greco and c. ratti , phys .
d * 84 * , 094004 ( 2011 ) [ arxiv:1103.5611 [ hep - ph ] ] .
v. s. filinov , v. e. fortov , p. r. levashov , y. .b . ivanov and m. bonitz ,
phys . lett .
a * 376 * , 1096 ( 2012 ) [ arxiv:1203.2191 [ hep - ph ] ] .
p. castorina and m. mannarelli , phys .
c * 75 * , 054901 ( 2007 ) [ hep - ph/0701206 [ hep - ph ] ] ; phys .
b * 644 * , 336 ( 2007 ) [ hep - ph/0510349 ] .
p. castorina , d. e. miller and h. satz , eur .
j. c * 71 * , 1673 ( 2011 ) [ arxiv:1101.1255 [ hep - ph ] ] .
f. brau and f. buisseret , phys . rev .
d * 79 * , 114007 ( 2009 ) [ arxiv:0902.4836 [ hep - ph ] ]
. j. cao , y. jiang , w. -m .
sun and h. -s .
zong , phys .
b * 711 * , 65 ( 2012 ) [ arxiv:1203.6460 [ hep - ph ] ] . v. m. bannur , phys .
b * 647 * , 271 ( 2007 ) ; v. m. bannur , eur .
j. c * 50 * , 629 ( 2007 ) ; e. m. lifshitz , l. p. pitaevskii , _ statistical physics , part 2 : volume 9 ( course of theoretical physics vol .
9 ) _ , butterworth - heinemann ( 1980 ) .
f. scardina , m. colonna , s. plumari and v. greco , arxiv:1202.2262 [ nucl - th ] .
plumari s , baran v , di toro m , ferini g and greco v 2010 _ phys.lett .
_ b * 689 * 18 a. m. polyakov , phys .
b * 72 * , 477 ( 1978 ) .
l. susskind , phys .
d * 20 * , 2610 ( 1979 ) .
b. svetitsky and l. g. yaffe , nucl .
b * 210 * , 423 ( 1982 ) .
b. svetitsky , phys .
* 132 * , 1 ( 1986 ) .
k. fukushima , phys .
b * 591 * , 277 ( 2004 ) [ hep - ph/0310121 ] . c. ratti , m. a. thaler and w. weise , phys .
d * 73 * , 014019 ( 2006 ) [ hep - ph/0506234 ] .
h. abuki and k. fukushima , phys .
b * 676 * , 57 ( 2009 ) [ arxiv:0901.4821 [ hep - ph ] ] .
t. zhang , t. brauner and d. h. rischke , jhep * 1006 * , 064 ( 2010 ) [ arxiv:1005.2928 [ hep - ph ] ] .
s. gupta , k. huebner and o. kaczmarek , phys .
d * 77 * , 034503 ( 2008 ) [ arxiv:0711.2251 [ hep - lat ] ] .
k. fukushima , phys .
d * 77 * , 114028 ( 2008 ) [ erratum - ibid .
d * 78 * , 039902 ( 2008 ) ] [ arxiv:0803.3318 [ hep - ph ] ] .
h. abuki , r. anglani , r. gatto , g. nardulli and m. ruggieri , phys .
d * 78 * , 034034 ( 2008 ) [ arxiv:0805.1509 [ hep - ph ] ] . | we study the interpretation of lattice data about the thermodynamics of the deconfinement phase of @xmath0 yang - mills theory , in terms of gluon quasiparticles propagating in a background of a polyakov loop .
a potential for the polyakov loop , inspired by the strong coupling expansion of the qcd action , is introduced ; the polyakov loop is coupled to tranverse gluon quasiparticles by means of a gas - like effective potential .
this study is useful to identify the effective degrees of freedom propagating in the gluon medium above the critical temperature .
a main general finding is that a dominant part of the phase transition dynamics is accounted for by the polyakov loop dynamics , hence the thermodynamics can be described without the need for diverging or exponentially increasing quasiparticle masses as @xmath1 , at variance respect to standard quasiparticle models . |
in surface design , the user often needs to perform rounding or filleting between two intersecting surfaces .
mathematically , the surface used in making the rounding is defined as the envelope of a family of spheres which are tangent to both surfaces .
this envelope of spheres centered at @xmath1 with radius @xmath2 , where @xmath3 and @xmath2 are rational functions , is called a canal surface with spine curve @xmath4 .
if the radius @xmath2 is constant the surface is called a pipe surface . moreover , if additionally we reduce the dimension ( take @xmath3 in a plane and consider circles instead of spheres ) we obtain the offset to the curve .
canal surfaces are very popular in geometric modelling , as they can be used as a blending surface between two surfaces .
for example , any two circular cones with a common inscribed sphere can be blended by a part of a dupin cyclide bounded by two circles as it was shown by @xcite ( see figure [ fig : cyclide ] ) .
cyclides are envelopes of special quadratic families of spheres . for other examples of blending with canal surfaces we refer to @xcite .
= 3.5 cm here we study the implicit equation of a canal surface @xmath5 and its implicit degree .
the implicit equation of a canal surface can be obtained after elimination of the family variable @xmath6 from the system of two equations @xmath7 ( here @xmath8 are quadratic in the variables @xmath9 ) , i.e. by taking the resultant with respect to @xmath6 . however , this resultant can have extraneous factors . in the paper
we explain how these factors appear and how we can eliminate them . by using lie and laguerre geometry
, we see that the above system of equations is related to a system @xmath10 , where @xmath11 are linear in the variables @xmath12 and @xmath13 is the lie quadric ( for the exact definition see formula ( [ liequadric ] ) ) .
it turns out that the variety defined by the system of equations @xmath14 is a dual variety to the curve @xmath15 , where @xmath16 is a curve on the lie quadric determined by the spine curve @xmath17 ( for the explicit definition see formula ( [ bar_e ] ) ) . for the dual variety @xmath18
we define the @xmath0-basis , which consists of two polynomials @xmath19 which are linear in @xmath20 , and of degree @xmath21 in @xmath6 are such that @xmath22 is minimal .
it turns out that the resultant of @xmath23 and @xmath24 with respect to @xmath6 gives the implicit equation of the variety @xmath18 .
there is a simple substitution formula ( see the algorithm at the end of section [ dualvariety ] ) to compute the implicit equation of the canal surface from the implicit equation of the variety @xmath18 .
partial solutions to the problem of finding the implicit equation ( and degree ) for canal surfaces have been given in other papers .
for instance , the degree of offsets to curves is studied in @xcite . in @xcite , there is a degree formula for the implicit equation of a polynomial canal surface .
quadratic canal surfaces ( parametric and implicit representation ) have been studied in @xcite .
we close the introduction by noting that the implicit degree of a canal surface is important for the parametric degree .
our observation is that if the canal surface has the minimal parametrization of bi - degree @xmath25 then its implicit degree is close to @xmath26 . on the minimal bi - degree @xmath25 parametrizations of the canal surface
we refer to @xcite .
the paper is organized as follows . in the next section
we develop some algebraic formalism about modules with two quasi - generators .
we define the @xmath0-basis for these modules and present an algorithm for its computation . in the following section , we recall some needed facts about lie and laguerre sphere geometry . then using lie and laguerre geometry we describe the canal surface explicitly
also , we introduce the @xmath27-hypersurface which contains all @xmath28-offsets to the canal surface . using the @xmath0-basis algorithm
we compute the implicit equations of the dual variety @xmath18 , the @xmath27-hypersurface and the canal surface @xmath5 .
next we apply the results of the previous section to the dual variety @xmath18 of the curve and explain how to compute the implicit degree of the @xmath27-hypersurface ( without computation of the implicit equation ) .
finally , we give some computational examples .
0.1 cm let @xmath29 $ ] be polynomial ring over the field of real numbers , and denote @xmath29^d$ ] the @xmath30-module of @xmath28-dimensional row vectors with entries in @xmath29 $ ] .
let @xmath31 be the field of rational functions in @xmath6 .
for a pair of vectors @xmath32,@xmath33^d$ ] the set @xmath34^d\ |\ a , b\in{\mathbb{r}}(t),a , b\in{\mathbb{r}}[t]^d\}\subset{\mathbb{r}}[t]^d\ ] ] is the @xmath29$]-module with two polynomial _ quasi - generators _ @xmath35 . here
, we assume that @xmath35 are @xmath29$]-linearly independent , i.e. @xmath36 with @xmath37 $ ] if and only if @xmath38 .
* remark : * note that the vectors @xmath35 may not be generators of the module @xmath39 over @xmath29 $ ] because @xmath40 and @xmath41 in the definition ( [ module ] ) are from the field @xmath31 of rational functions . for example , if @xmath42 with @xmath43,d\in{\mathbb{r}}[t]^d$ ] and @xmath44 then @xmath35 are not generators of the module @xmath39 . + for @xmath32,@xmath33^d$ ]
we define the pl@xmath45cker coordinate vector @xmath46 as follows : @xmath47,[1,3], ... ,[d-1,d])\in{\mathbb{r}}[t]^{d(d-1)/2},\ \mbox{where}\ [ i , j]=a_ib_j - a_jb_i .
\nonumber\ ] ] in other words , @xmath46 is the vector of @xmath48-minors of the matrix @xmath49 and we denote by @xmath50 the degree of the pl@xmath45cker coordinate vector , i.e. the maximal degree of a 2-minor of @xmath51 .
let a polynomial vector @xmath52^d$ ] be presented as @xmath53 we denote the leading vector @xmath54 by @xmath55 and the degree of @xmath56 by @xmath57 .
note that if @xmath58 and @xmath59 are linearly independent over @xmath60 then @xmath61 and @xmath62 .
we define @xmath63^d \ \
\mbox{such that}\ \langle \tilde a,\tilde b \rangle
= m \}\ ] ] to be the degree of the module @xmath39 with two quasi - generators .
two quasi - generators @xmath64 of the module @xmath65 are called a @xmath0-basis of the module @xmath39 if @xmath66 .
as we always have the inequality @xmath67 , this means in particular that the sum @xmath68 is minimal .
a @xmath0-basis always exists , as we shall see at the end of the section .
let us explain the geometric motivation behind this definition . by abuse of notation , we will continue to denote parameters @xmath6 , however in the geometric definitions that follow , they should be understood as parameters @xmath69 and polynomials in @xmath29 $ ]
should be thought of as homogenized with respect to a new variable @xmath70 .
we define the following subspace of @xmath71 for the module @xmath72 .
@xmath73 where @xmath74^d$ ] , @xmath75 and @xmath76 .
we have the inequality @xmath77 , because the module @xmath39 has only two quasi - generators .
in fact , we have @xmath78 for all @xmath79 , as we will see in proposition [ mu - basis - prop].2 .
whenever two vectors @xmath80 and @xmath81 are linearly independent in @xmath71 then @xmath82 is the intersection of two hyperspaces @xmath83 and @xmath84 . using those subspaces , we can associate a hypersurface @xmath85 in the real projective space @xmath86 with the module @xmath39 @xmath87 note that this definition and the definition of @xmath82 depend only on the module @xmath39 and not on the choice of quasi - generators .
it is useful to compare the hypersurface @xmath88 with the hypersurface @xmath89 defined as @xmath90 where @xmath91 are quasi - generators of @xmath39 . by definition , this is the variety defined by @xmath92 and it is clear that @xmath93 . if the vectors @xmath94 are linearly dependent , then @xmath95 is a subspace of codimension one .
note that in this case the implicit equation @xmath92 contains the factor @xmath96 . as a matter of fact
, this happens if and only if @xmath97 has rank one , which is equivalent to saying that @xmath79 is a zero of the ideal generated by the pl@xmath45cker coordinates .
in fact , we will see in proposition [ s_m ] that this phenomenon does not occur for @xmath0-bases , i.e. if @xmath98 is a @xmath0-basis of the module @xmath39 then @xmath99 and there are no extraneous factors as before .
we should explain why we use the term @xmath0-basis .
the above definition is a generalization of the usual definition for the @xmath0-basis of a rational ruled surface ( as in @xcite , @xcite or @xcite ) .
they coincide in the special case @xmath100 .
@xmath39 is the analogue of the syzygy module ( i.e. the module of moving planes following the parametrization of the ruled surface ) and the subspaces @xmath101 , which in this case are 2-dimensional and hence define projective lines , are exactly the family of lines which constitute the ruled surface .
similarly , the case @xmath102 corresponds at the theory of @xmath0-bases for rational curves and our definition is equivalent to the usual definition as in ( * ? ? ?
* theorem 3 , condition 3 ) . however , the approach used here is actually inverse to the approach in the cited papers . in the latter
the ruled surface is defined by a parametrization and then the module of moving planes is studied , whereas here we fix a module that `` looks like '' such a moving plane module and then study the ( generalized ) ruled surface that corresponds to it .
note that by definition of the subspaces @xmath101 any element @xmath103 of @xmath39 can be considered a moving plane following @xmath88 , in the sense that for all @xmath104 there is a parameter @xmath6 such that @xmath105 .
note that @xmath46 defines the so - called plcker curve @xmath106 in @xmath107 by @xmath108:[1,3]: ... :[d-1,d ] ) \end{array}\ ] ] where @xmath109=a_ib_j - a_jb_i$ ] .
we will denote @xmath110 the degree of the parametrization , which is the cardinality of the fiber of a generic point in the image of @xmath111 .
note that @xmath111 and @xmath112 are the same for any choice of quasi - generators of @xmath39 .
[ degreeofsm ] for any pair of quasi - generators @xmath35 of @xmath39 we have the degree formula @xmath113k _ m = ( ab)-q_a , b,@xmath113 where @xmath114 and @xmath115 .
moreover , we have @xmath116 . the proposition and the proof are similar to lemma 1 in @xcite and to theorem 5.3 in @xcite .
the implicit degree of the hypersurface @xmath117 is the number of intersections between a generic line and the hypersurface .
the generic line @xmath118 is defined by two points in the space @xmath119 , where @xmath120 .
the line @xmath118 intersects the hyperplane @xmath121 if and only if @xmath122 . since the line @xmath118 should intersect the hyperplane @xmath123 too
, we see that the implicit degree is the number of intersections of two curves in the @xmath124 plane : @xmath125 @xmath126 eliminating @xmath70 from the above equation we have @xmath127 where @xmath128 means a standard scalar product of two vectors @xmath129 .
the number of solutions of is the number of intersection points of the plcker curve with a generic hyperplane in @xmath107 , so @xmath130 .
now it is known ( see for example ( * ? ? ?
* theorem 1 ) ) that @xmath131 and the proposition follows .
we have yet to show the existence of the @xmath0-basis .
to this end , we propose an algorithm for its computation , the basic idea of which is to reduce @xmath114 to a constant using the so - called smith form of the @xmath132 matrix @xmath49 and then render the leading vectors linearly independent by a simple degree reduction .
the smith form is a decomposition @xmath133 , with unimodular @xmath134^{2 \times 2}$ ] , @xmath135^{d \times d}$ ] , and @xmath136^{2 \times d}\ ] ] it always exists and can be computed efficiently by standard computer algebra systems .
+ * algorithm ( @xmath0-basis ) * 1 .
input : quasi - generators @xmath32,@xmath33^d$ ] of the module @xmath39 2 .
set @xmath137 3 .
compute a smith form @xmath138 with unimodular @xmath139^{2 \times 2}$],@xmath140^{d \times d}$ ] .
4 . set @xmath141 to be the @xmath132-submatrix consisting of the first two rows of @xmath142 . 5 .
if the vector of leading terms ( with respect to the variable @xmath6 ) of the first row is @xmath143 times the one of the second row , @xmath144 $ ] , set @xmath145 . 6 . if the vector of leading terms ( with respect to the variable @xmath6 ) of the second row is @xmath143 times the one of the first row , @xmath146 $ ] , set @xmath147 .
if the preceding two steps changed @xmath141 go back to step 5 .
set @xmath64 to be the rows of @xmath141
output : a @xmath0-basis @xmath64 of the module @xmath39 * * + as we shall see in section [ dualvariety ] , the case we are interested in is the case @xmath148 , so we are dealing with very small matrices and the computations are extremely fast .
note that we actually only need the first two rows of @xmath142 , so we could optimize the algorithm by modifying the smith form algorithm used as not to compute the unnecessary entries of the matrices @xmath149 and @xmath142 . generally , the number of elementary matrix operations in step 5 and 6 is very low . in the worst case ,
it is bounded by the maximal degree of the entries of the matrix @xmath141 in step 4 of the algorithm , since each step reduces the maximal degree in one of the rows of @xmath141 .
next , we will show that the output of the above algorithm is a @xmath0-basis and that the resultant of a @xmath0-basis @xmath64 of the module @xmath72 is an implicit equation of @xmath88 . in section [ dualvariety ]
, we will use these results for a special choice of @xmath56 and @xmath150 to compute the implicit equation of a canal surface .
[ mubadeg ] the output of the above algorithm is a @xmath0-basis and we have @xmath151 , where @xmath115 .
let @xmath152 be the output of the above algorithm . by construction
it is clear that @xmath152 are quasi - generators of @xmath39 and that @xmath153 .
furthermore , we have @xmath154 , because the vectors of leading terms of @xmath155 and @xmath156 are linearly independent .
so by proposition [ degreeofsm ] we deduce @xmath157 moreover , by definition we have @xmath158 and if @xmath35 are quasi - generators such that @xmath159 is minimal , the degree formula gives @xmath160 , which shows that @xmath151 , and as a consequence that @xmath152 is indeed a @xmath0-basis .
[ s_m ] let @xmath152 be a @xmath0-basis of @xmath39 and @xmath75 variables
. then @xmath161 where @xmath162 is the implicit equation of the hypersurface @xmath88 .
first , we will show in the same way as in ( * ? ? ?
* theorem 9 ) that @xmath163 is geometrically irreducible , i.e. the power of an irreducible polynomial . as we shall see in proposition [ mu - basis - prop ] , the intersection of the hyperplanes @xmath164 and @xmath165 is of codimension 2 for any parameter @xmath166 .
so the incidence variety @xmath167 is a vector bundle over @xmath168 and hence irreducible .
so the projection on @xmath169 is irreducible as well and its equation , which is by definition the hypersurface defined by @xmath163 , is a power of an irreducible polynomial .
as we have remarked earlier , the resultant of two quasi - generators is always a multiple of the implicit equation of @xmath88 , so @xmath163 is a power of @xmath162 .
but using the degree property above we see @xmath170 which implies that @xmath171 equals @xmath172 .
it is known that the plcker curve @xmath106 can be properly reparametrized , i.e. there exists a rational function @xmath143 of degree @xmath112 such that @xmath173 , where @xmath103 is a proper parametrization of @xmath106 .
it is tempting to use this proper reparametrization in order to represent the implicit equation @xmath162 of @xmath88 directly as a resultant as in the proof of ( * ? ? ?
* theorem 3 ) .
however , @xmath143 does not necessarily factorize @xmath56 and @xmath150 , i.e. it is not sure that there exist @xmath174 and @xmath175 with @xmath176 and @xmath177 , which would be needed to do this . in the following we present some properties of @xmath0-bases .
note that the properties in propositions [ mu - basis - prop ] , [ mu - basis - prop-2 ] are similar to @xcite theorems 1,3 .
however , we give different proofs by deducing them from the degree formula and lemma [ mubadeg ] .
[ mu - basis - prop ] let @xmath72 and let @xmath98 be a @xmath0-basis of the module @xmath39 .
then the following properties hold : + 1 .
the vectors @xmath178 are linearly independent .
@xmath179 are linearly independent over @xmath180 for any parameter value @xmath181 .
if @xmath178 were linearly dependent , this would imply that @xmath182 which is a contradiction to lemma [ mubadeg ] .
suppose that @xmath179 are linearly dependent for some @xmath181 .
this is equivalent to saying that the matrix @xmath183 is not of full rank , which means that all 2-minors vanish .
so @xmath79 is a root of @xmath184 and as above we deduce @xmath185 which is again a contradiction to lemma [ mubadeg ] .
[ mu - basis - prop-2 ] let @xmath186 and assume that @xmath98 satisfy conditions 1,2 from proposition [ mu - basis - prop ] . then any element @xmath187 has the following expression : @xmath188 for some @xmath189 $ ] , i.e. @xmath98 are generators of the module @xmath39 over the polynomial ring @xmath29 $ ] .
moreover , the pair @xmath98 is a @xmath0-basis of the module @xmath39 .
let @xmath190 , it can be expressed as @xmath191 with @xmath192 $ ] and co - prime numerators and denominators in the rational functions @xmath193 and @xmath194 .
furthermore , we may assume that @xmath195 , because if @xmath196 is a linear combination of @xmath98 , then so is @xmath197 .
multiplying both sides of the above equation with @xmath198 we obtain @xmath199 or equivalently @xmath200 and since @xmath41 divides neither @xmath40 nor @xmath201 ( if it divided @xmath201 , for any root @xmath79 of @xmath41 and any constant @xmath202 we would deduce the relation @xmath203 , which contradicts property 2 in proposition [ mu - basis - prop ] ) , one concludes that @xmath41 divides @xmath28 and by a symmetric argument that @xmath28 divides @xmath41 , so we may assume @xmath204 .
so we have @xmath205 and plugging a root @xmath79 of @xmath41 into the equation , we would obtain a non - trivial linear relation between @xmath201 and @xmath206 , again a contradiction to proposition [ mu - basis - prop ] .
this implies that @xmath41 and @xmath28 are constant , which shows that any @xmath190 can be expressed as linear combination of @xmath201 and @xmath206 over @xmath29 $ ] . in other words : @xmath201 and @xmath206 are not only quasi - generators of @xmath39 , but actually generators in the usual sense , i.e. over @xmath29 $ ] .
suppose that @xmath207 and let @xmath208 .
then we proved that @xmath209 for some polynomials @xmath210 $ ] .
since @xmath178 are linearly independent @xmath211 and @xmath212 do not cancel each other .
therefore , @xmath213 ( if @xmath214 ) or @xmath215 ( if @xmath216 ) . also @xmath217 and @xmath218 .
so , we see that @xmath219 , i.e. a pair @xmath98 is a @xmath0-basis of the module @xmath39 .
here we shortly recall the elements of lie and laguerre sphere geometry ( cf .
@xcite ) . we start from the construction of lie s geometry of oriented spheres and planes in @xmath220 .
let @xmath221 , @xmath222 .
the oriented sphere @xmath223 in @xmath220 is the set @xmath224 where by @xmath225 we denote the standard positive definite scalar product in @xmath220 .
the orientation is determined by the sign of @xmath226 : the normals are pointing outwards if @xmath227 . if @xmath228 then @xmath229 is a point .
let @xmath230 with @xmath231 and @xmath232 .
the oriented plane @xmath233 in @xmath220 is the set @xmath234 the _ lie scalar _ product with signature @xmath235 in @xmath236 is defined by the formula @xmath237=\frac{-x_1z_2-x_2z_1}{2}+x_3z_3+x_4z_4+x_5z_5-x_6z_6 .
\nonumber\ ] ] for @xmath238 and @xmath239 . in matrix notation
we have @xmath240=xcz^t,\ \mbox { where}\ xc=(-x_2/2,-x_1/2,x_3,x_4,x_5,-x_6).\ ] ] denote @xmath241 and define the quadric @xmath242=-uy_0+y_1 ^ 2+y_2 ^ 2+y_3 ^ 2-y_4 ^ 2=0\}\ ] ] where @xmath243 $ ] is the obvious extension of the lie scalar product to @xmath244 .
@xmath245 is called _
lie quadric_. we represent an oriented sphere @xmath223 ( or an oriented plane @xmath233 ) as a point @xmath246 ( resp .
@xmath247 ) on the lie quadric : @xmath248 it is easy to see that we have determined a bijective correspondence between the set of points on the lie quadric @xmath245 and the set of all oriented spheres / planes in @xmath220 . here
we assume that a point @xmath249 on the lie quadric @xmath245 corresponds to an infinity , i.e. to a point in the compactification of @xmath220 .
we say that @xmath250 is the improper point on the lie quadric .
notice that oriented planes in @xmath220 correspond to points @xmath251 , where @xmath252 is a tangent hyperplane to the lie quadric at the improper point @xmath253 .
two oriented spheres @xmath254 are in _ oriented contact _ if they are tangent and have the same orientation at the point of contact .
the analytic condition for oriented contact is @xmath255 where @xmath256 denotes the usual distance between two points in the euclidean space @xmath220 .
one can check directly that the analytical condition of oriented contact on the lie quadric is equivalent to the equation @xmath257=0 .
\nonumber\ ] ] it is known that the lie quadric contains projective lines but no linear subspaces of higher dimension ( chapter 1 , corollary 5.2 in @xcite ) . moreover , the line in @xmath244 determined by two points @xmath258 of @xmath245 lies on @xmath245 if and only @xmath259=0 $ ] , i.e. the corresponding spheres to @xmath258 are in an oriented contact ( chapter 1 , theorem 1.5.4 in @xcite ) .
the points on a line on @xmath245 form so called _ parabolic pencil _ of spheres .
all spheres which correspond to a line on @xmath245 are precisely the set of all spheres in an oriented contact . here
we use a slightly different coordinate system in lie geometry than in the book @xcite .
the scalar product as in @xcite may be obtained applying the following transformation : + @xmath260 we show now that the set of points @xmath20 in @xmath245 with @xmath261 is naturally diffeomorphic to the affine space @xmath262 .
this diffeomorphism is defined by the map @xmath263 where @xmath264 as before , i.e. the tangent hyperplane to the lie quadric @xmath245 at the improper point @xmath250 .
let @xmath265 and denote by @xmath266 the lorentz scalar product on @xmath262 , which can be seen as the restriction of the lie scalar product @xmath243 $ ] to @xmath262 .
the affine space @xmath262 with the lorentz scalar product is called the lorentz space and denoted by @xmath267 .
let @xmath268 .
one can check that inverse map of @xmath269 is given by the formula : @xmath270 * * + notice , that @xmath271 , i.e. the sphere @xmath272 corresponds to a point @xmath273 .
the map @xmath269 can be extended to a linear projection @xmath274 from @xmath275 to @xmath276 defined as @xmath277 the points of @xmath251 can be represented as @xmath278 and these points correspond to planes in @xmath220 . note that @xmath279 are infinite points to the natural extension of @xmath262 to @xmath276 which correspond to a pencil of parallel planes in @xmath220 .
the quadric @xmath280 is called _
absolute quadric_. the preimage of the map @xmath281 has the following form @xmath282 * * + where @xmath283 and @xmath9 as before .
a direct computation shows that for @xmath284 @xmath285=\langle v - w , v - w\rangle\ ] ] the formula shows that two oriented spheres defined by @xmath286 , @xmath287 ( i.e. spheres @xmath288 and @xmath289 ) are in oriented contact if and only if @xmath290 .
let us define two maps : an embedding @xmath291 and a projection @xmath292 , where @xmath222 .
we will treat points @xmath293 as spheres with zero radius and identify them with @xmath220 .
all interrelations between the spaces introduced above can be described in the following diagram @xmath294 for an oriented surface ( curve or point ) @xmath295 define an _
isotropic hypersurface _
@xmath296 as the union of all points in @xmath262 which correspond to oriented tangent spheres of @xmath297 .
let @xmath298 be a variety which corresponds to tangent spheres with radius @xmath28 of @xmath297 .
the set @xmath299 are centers of spheres with radius @xmath28 tangent to @xmath297 .
the set @xmath300 is called @xmath28-envelope of the variety @xmath297 . since @xmath301
we can treat the isotropic hypersurface @xmath302 as the union of all @xmath28-envelope to the variety @xmath297 .
if @xmath303 , we define a function @xmath304 let @xmath305 be such that @xmath306 .
by the formula ( [ twoproducts ] ) we see that spheres @xmath307 and @xmath308 are in oriented contact . therefore , in the same manner as previously , we define the isotropic hypersurface @xmath309 as follows @xmath310 in fact , @xmath309 is a quadratic cone with a singular point at a vertex @xmath311 and may be viewed as the set of all spheres which touches the fixed sphere @xmath312 after the restriction to the linear subspace @xmath313 this hypersurface consists of all spheres with radius @xmath28 which are in oriented contact with the sphere @xmath307 which we denote as @xmath314 we notice that @xmath315 is defined by the equation @xmath316 , i.e. @xmath317 therefore , in this case , the isotropic hypersurface @xmath318 may be treated as a union all envelopes to the sphere @xmath319 in the next section we generalize the definition of the isotropic hypersurface @xmath302 for a curve @xmath297 in @xmath262 ( or @xmath276 ) .
all lines in @xmath267 with directional vectors @xmath286 can be classified into three types depending on the sign of @xmath320 : @xmath321-lines , @xmath322-lines ( also called _ isotropic _ lines ) , and @xmath323-lines .
in this section , we will see that the definition of the canal surface is not obvious and and we will introduce some geometrical object related to it .
a canal surface is given by a so - called _ spine curve _ @xmath17 , which is the closed image ( with respect to the zariski topology ) of a rational map @xmath324 with polynomials @xmath325 $ ] such that @xmath326 .
for abbreviation , we usually skip the variable @xmath6 in the notations .
the spine curve describes a family of spheres @xmath327 whose centers are given by the first three coordinates @xmath328 and whose radii are given by the last coordinate @xmath329 . intuitively , the canal surface is the envelope of this family of spheres , but there are some subtleties to consider before we can make a precise definition .
we can also consider the spine curve as a projective curve @xmath330 given as the closed image of a parametrization @xmath331 with the non - restrictive condition @xmath332 , which means that there are no base - points ( i.e. parameters for which the map is not well - defined ) .
note that in this case the polynomials @xmath333 are actually to be considered as homogenized to the same degree @xmath334 with respect to a new variable @xmath70 .
as there is a one - to - one correspondence between the univariate polynomials of a certain degree and their homogeneous counterparts , we will keep the notation from above and distinguish between the affine and projective case only where it is necessary to avoid confusion . in
the following we use the notations @xmath335 @xmath336 we first proceed to define a hypersurface in @xmath276 which is closely related to the canal surface .
the _ isotropic hypersurface _ @xmath337 associated with the ( projective ) spine curve @xmath330 is the variety in @xmath276 defined by the polynomial @xmath338 where @xmath339
so , we define @xmath340 as the envelope of the family of isotropic hypersurfaces @xmath341 . in the previous section we showed that @xmath342 .
this interpretation leads to the following definition .
@xmath28-envelope _ associated with the ( projective ) spine curve @xmath330 is defined as the hypersurface @xmath343 given by the implicit equation @xmath344 i.e. the equation obtained by replacing @xmath345 in @xmath346 by @xmath347 , where @xmath348 . the _ affine envelope _ @xmath349 at distance @xmath28 is the restriction of @xmath350 to the affine space @xmath220 , defined by the equation @xmath351 , i.e. by setting @xmath352 .
so @xmath340 contains all offsets associated with the spine curve @xmath330 .
indeed , the surface @xmath353 is a hyperplane section of @xmath340 , which can be interpreted as a parametrization of all offsets ( with respect to the parameter @xmath345 ) .
the special case @xmath354 is particularly important . for the real part of @xmath355 to be non - empty
, one has to suppose that @xmath17 has tangent @xmath321-lines almost everywhere , or equivalently that @xmath356 almost everywhere .
@xmath355 is the envelope of the family of spheres in @xmath220 given by the spine curve @xmath17 and @xmath349 is the envelope of the same family of spheres with radii augmented by @xmath28 .
for instance , circular cylinders or circular cones ( call them just _ cones _ ) are envelopes @xmath357 of @xmath321-lines @xmath358 and vice versa . in the literature , the canal surface @xmath5 is usually defined as this envelope @xmath355 .
however , we will show in an example that these envelopes can contain `` unwanted '' extraneous factors , which are geometrically counterintuitive .
consider the spine curve @xmath17 given by @xmath359 the first three coordinates describe a circle in the plane and moving spheres of constant radius along this curve , so intuitively the envelope should be a torus @xmath360 .
but it turns out that the implicit equation of @xmath355 is up to a constant computed as @xmath361 where @xmath362 is indeed the equation of the torus . to understand where the other factors come from , consider the following : for a given parameter @xmath6 , the equations @xmath363 and @xmath364 define spheres @xmath365 and @xmath366 in @xmath220 and @xmath367 of the intersections of these spheres ( actually this is nothing else than the geometric definition of the resultant ) .
now , while for almost all @xmath6 this intersection is a transversal circle on the torus ( often called characteristic circle in the literature ) , it can happen that the spheres degenerate either to planes or to the whole space .
in our example , for the parameters @xmath368 and @xmath369 we have @xmath370 , @xmath371 and @xmath372 , so the intersection in those parameters actually degenerates to ( complex ) planes which correspond to the factor @xmath373 . in the parameter @xmath374 ,
both @xmath363 and @xmath364 define the same sphere whose equation @xmath375 is the other extraneous factor .
this kind of phenomenon can also happen for real parameter values , but it is interesting to remark that even though we consider a real parametrization , non - real parameters can interfere with the envelope , because the resultant `` knows '' about them .
this example shows that @xmath355 is not a suitable definition for the canal surface @xmath5 and we will later develop one that avoids the kind of extraneous components we have observed .
sometimes , in the literature , @xmath355 is defined in affine space as the resultant @xmath376 or in other words by deriving the affine equation of the sphere after the substitutions @xmath377 and homogenizing afterwards .
note that in this case @xmath378 and @xmath379 are linear in @xmath380 .
let @xmath381 and @xmath382 .
an easy computation shows that we have the following equalities @xmath383 therefore , by properties of the resultant ( [ res - product]),([res - divisor ] ) , we have @xmath384 hence , we have @xmath385 , so there are even more extraneous factors than before due to the roots of @xmath386 .
the main idea to understand and eliminate the extraneous components that appeared in the example is to linearize the equations @xmath363 and @xmath364 by replacing the quadratic term @xmath387 by a new variable @xmath388 ( or more precisely @xmath389 to keep the equations homogeneous ) .
this will make the results developed in section 2 applicable .
geometrically , this means that we will pull back the spine curve to @xmath245 via the correspondence @xmath281 . for a spine curve @xmath390
we define a _ proper _ pre - image @xmath16 in the lie quadric @xmath245 as the closure of the set @xmath391 in @xmath245 .
it is immediate by that the parametrization of @xmath16 is @xmath392 we can now define the envelopes associated with this new spine curve as follows .
the variety @xmath393 associated with @xmath16 is the hypersurface in @xmath244 defined by the implicit equation @xmath394 where @xmath395 and @xmath396=u e_0 ^ 2+y_0\langle e , e\rangle-2\langle e_0e , y\rangle , \nonumber \\ h_2({\hat{y}},t)&= & \frac{\partial h_1({\hat{y}},t)}{\partial t } = -2[{\hat{y } } , { { \hat{\mathcal{e}}}}'(t)]=2(u e_0e_0'+y_0\langle e',e\rangle-\langle ( e_0e)',y\rangle ) .
\nonumber\end{aligned}\ ] ] similarly , the variety @xmath397 is defined by the implicit equation @xmath398 i.e. the equation obtained by replacing @xmath345 in @xmath399 by @xmath347 , where @xmath348 .
of course this is nothing else than substituting @xmath400 in @xmath363 and @xmath364 by @xmath389 and dividing by @xmath401 , so @xmath402 , i.e. @xmath403 .
now as an immediate corollary we obtain [ phi]with the notations as above we have @xmath404 and @xmath405 to sum up , we have defined two hypersurfaces as resultants of two quadratic forms : @xmath343 , which are the offsets to the spine curve @xmath330 , and @xmath406 , which can be interpreted as a parametrization of those offsets . as seen in an example
, these definitions can lead to additional components which are against the geometric intuition , so it is desirable to give another definition which avoids those extra factors . to this end , we have linearized the problem by replacing the quadratic polynomials @xmath363 and @xmath364 by linear forms @xmath407 and @xmath408 by substituting the quadratic term by a new variable and have seen how to reverse this substitution .
geometrically , this means that we replace the hypersurfaces @xmath350 and @xmath340 by hypersurfaces @xmath409 and @xmath410 in one dimension higher .
this has the advantage that we can now apply the technique of @xmath0-bases developed earlier to understand and eliminate the extraneous factors of @xmath409 and @xmath410 and then come back to @xmath411 ( resp .
@xmath276 ) with the substitution formulae of proposition [ phi ] .
in this section , we will finally be able to define the canal surface @xmath5 ( and more general offsets to it ) and the so - called dual variety @xmath412 , which can be seen as a parametrization of the offsets to @xmath5 .
up to the constant @xmath413 the system @xmath414 is equal to @xmath415 = { \hat{\mathcal{e}}}(t)c { \hat{y}}^t = 0 , \\ \left[{\hat{y}},{\hat{\mathcal{e}}}'(t)\right ] = { \hat{\mathcal{e}}}'(t)c { \hat{y}}^t = 0 , \\
\end{array } \right.\ ] ] where the matrix @xmath103 is defined by the formula .
we can interpret the variety @xmath410 defined by as a dual variety to the curve @xmath16 with respect to the lie quadric @xmath245 , i.e. the dual variety to the curve @xmath416 . indeed , this dual variety consists of the hyperplanes which touch the curve @xmath416 .
the first equation in ( [ gamma_lie ] ) means that the hyperplane contains the point @xmath416 , the second equation means that the hyperplane contains the tangent vector @xmath417 to the curve @xmath416 . in order to simplify notation
we denote @xmath418 and we have that @xmath419 by .
as we have seen in section 2 , this surface contains extraneous factors which correspond to the roots of the @xmath48-minors of the matrix @xmath420 , but which can be eliminated by replacing @xmath421 by a @xmath0-basis of the module @xmath422 .
it is thus natural to make the following definition .
we define the dual variety @xmath423 to the curve @xmath16 as the hypersurface @xmath424 where @xmath422 is the module quasi - generated by @xmath425 and @xmath426 . by the results of section [ quasigens ]
, it is immediate that @xmath427 does not contain the components of @xmath410 caused by parameters @xmath6 where @xmath428 is not of full rank or equivalently , where the intersection of the hyperplanes defined by @xmath407 and @xmath408 is of codimension 1 , i.e. the hyperplanes coincide .
so we can deduce [ impleq]let @xmath429 be a @xmath0-basis of the module quasi - generated by @xmath425 and @xmath426 and let @xmath112 be the degree of the parametrization @xmath430 as in section [ quasigens ] .
then @xmath431 where @xmath432 and @xmath433 where @xmath434 is the implicit equation of @xmath435 .
it follows directly from propositions [ degreeofsm ] and [ s_m ] .
of course , the same considerations can be applied to the hypersurfaces @xmath409 and we make the analogous definitions . substituting @xmath436 in @xmath407 and @xmath408 corresponds to replacing @xmath425 and @xmath426 by two linear forms @xmath437 with @xmath438 .
now @xmath439 and one makes an analogous definition : we define the hypersurface @xmath440 as @xmath441 where @xmath442 is the module quasi - generated by @xmath197 and @xmath443 . in this case also , @xmath444 does not contain extraneous factors due to the parameters @xmath6 where the rank of @xmath445 drops . at this point , it should be remarked that while we clearly always have @xmath446 this inclusion is not necessarily an equality ( note that we had @xmath447 for the corresponding varieties ) .
analogously to proposition [ impleq ] the following holds .
[ impleq1 ] let @xmath448 be a @xmath0-basis of the module quasi - generated by @xmath197 and @xmath443 and let @xmath112 be the degree of the parametrization @xmath449 as in section [ quasigens ] .
then @xmath450 where @xmath451 and @xmath452 where @xmath453 is the implicit equation of @xmath440 .
it follows directly from propositions [ degreeofsm ] and [ s_m ] .
finally , we can use the correspondance of proposition [ phi ] to define the canal surface .
[ gamma - definition ] the @xmath454-hypersurface is defined as @xmath455 and the offset @xmath456 at distance d to the canal surface @xmath5 is @xmath457 where @xmath458 and @xmath459 .
the canal surface itself is the special case @xmath354 or in other words @xmath460 .
note that the extraneous factors of @xmath409 and @xmath410 are in one - to - one correspondence with the extraneous factors of the corresponding hypersurfaces @xmath350 and @xmath340 since they are caused by parameter values where the intersection of @xmath407 and @xmath408 ( resp . @xmath363 and
@xmath364 ) is of codimension one .
so @xmath412 and @xmath461 contain no such factors .
+ in this section and the previous one , many different geometric objects have been defined .
we illustrate in the following diagram how they are related in order to make the situation clearer . @xmath462 note that the hypersurfaces in the third row are included in the corresponding hypersurfaces in the second row .
the first column is the naive definition of the objects to be studied : @xmath463 is more or less a @xmath28-offset to the canal offsets and @xmath340 a hypersurface in one dimension higher containing all those offsets .
however , they contain extraneous factors .
so by passing to the second column , we linearize the hypersurfaces ( i.e. we express them as resultants of linear forms ) and can apply @xmath0-bases to eliminate the extraneous factor , which gives the third column and finally go back down in dimension ( by intersecting with @xmath245 and applying @xmath274 to obtain the objects we are interested in : the offsets @xmath456 ( in particular the canal surface @xmath460 ) and the @xmath454-hypersurface .
we can now describe how to compute powers of the implicit equations of the dual varieties @xmath435 and @xmath440 , the hypersurface @xmath412 and the offsets surface @xmath461 .
we should remark that these powers ( which are the degrees of the parametrizations of the corresponding plcker curves ) are in a way inherent to the geometry of the problem , as we shall illustrate in example [ doubleellipsoid ] .
they can be interpreted as the number of times the surface is traced by the spine curve .
note also that this not necessarily due to the non - properness of the spine curve : even for a proper spine curve it can happen that the canal surface ( or its offsets ) is multiply traced , as in example [ doubleellipsoid ] . + * algorithm ( implicit equations ) * 1 .
input : a rational vector @xmath464 as in formula ( [ eee ] ) .
2 . define @xmath465 ^ 6 $ ] as in formula and @xmath466 ^ 5 $ ] as in formula .
3 . compute a @xmath0-basis @xmath429 of the module @xmath422 and a @xmath0-basis @xmath448 of the module @xmath442 using the algorithm in section [ quasigens ] .
4 . set @xmath467 and @xmath468 .
5 . let @xmath469 , where @xmath112 is a minimal integer such that @xmath470 is a polynomial .
similarly , set + @xmath471 .
output : @xmath434 , @xmath453 , @xmath470 , and @xmath472 , which are powers of the implicit equation of the varieties @xmath435 , @xmath440 , @xmath412 and @xmath461 * * + note that the affine parts of these equations can be obtained by replacing @xmath352 before the resultant computation .
we can describe the parametrization of @xmath435 .
the hyperplane defined by the equation @xmath473 is tangent to the curve @xmath416 , @xmath474 are three arbitrary points . by the definition a point on the dual variety @xmath18
is @xmath475 . define @xmath476 to be the @xmath477 matrix with five rows @xmath478 . and
let @xmath479 be @xmath480 matrices obtained from @xmath197 by removing the i - th column . then using the laplacian expansion by minors for the first row of the determinant ( [ tang ] ) we obtain the parametrization of @xmath435 as follows : @xmath481 where @xmath482 . here
@xmath483 are arbitrary parameters .
the aim of this section is to get some formula for the implicit degree of the hypersurface @xmath412 in terms of the rational spine curve @xmath484 .
notice that the implicit degree of the canal surface @xmath5 is less or equal than @xmath485 because we always have the inclusion @xmath486 so this formula gives upper bound for the degree of the canal surface . in the case of a polynomial spine curve the upper bound was obtained in the paper @xcite .
note that for the computation of the implicit degree we do not need the implicit equation of the hypersurface .
we believe that this formula is useful for higher degree spine curves because the computation of the implicit equation may be very difficult in practice .
let us remind that the pre - image @xmath487 is defined by the intersection of two varieties @xmath488 .
let denote by @xmath489 the equation of @xmath435 .
the lie quadric has the equation @xmath490 ( recall that @xmath491 ) . by the definition ( [ gamma - definition ] )
the equation of @xmath492 is obtained after the elimination of the variable @xmath388 from the equations of @xmath245 and @xmath435 , i.e. @xmath493 where @xmath112 is a minimal integer such that the left side of the equation ( [ geq ] ) is polynomial .
we introduce the following weighted degree @xmath494 where @xmath495 is a linear combination of the monomials @xmath496 . using this notation we have @xmath497 .
let us assume that the curve @xmath498 has a homogeneous parametrization .
we introduce the following notations @xmath499 where @xmath500 means derivative with respect to @xmath6 .
we will say that the curve @xmath501 is of _ general _ type if : @xmath502 ( i.e. @xmath503 ) , where @xmath504 and @xmath505 as in formula ( [ eeprime ] ) . the first equation in the system ( [ gamma_lie ] )
has the following form : @xmath506 the polynomial @xmath407 is linear in the variables @xmath507 and has degree @xmath508 in the variable @xmath6 . the elimination of the variable @xmath6 from the system @xmath509 is the reducible polynomial @xmath510 . by definition one of those factors is the equation of the dual variety @xmath511 .
[ degree_of_g ] if @xmath330 is a curve of general type then @xmath512 , where @xmath513 .
moreover , we have @xmath514 , where @xmath515 is the leading coefficient of the polynomial @xmath407 with respect to the variable @xmath6 and @xmath516 . for the curve of general type by the proposition [ impleq ] we have + @xmath517 we can compute components of the pl@xmath45cker vector @xmath518:[1,3]: ...
:[5,6])\in{\mathbb{p}}^{14}$ ] .
for example , @xmath519=e_0 ^ 2w_j/2 , \ j=1,2,3,4 $ ] and @xmath520=(-\langle e , e\rangle ( e_0e_1)'+e_0e_1\langle e , e\rangle')/2 $ ] . by assumption ( [ assump ] )
we see + @xmath521,[1,4],[1,5],[1,6])=e_0 ^ 2 $ ] and @xmath522)=1 $ ] .
therefore @xmath523 .
so we have @xmath524 .
notice that @xmath525 .
therefore we see @xmath526 , i.e. @xmath527 is the implicit equation of @xmath435 .
thereinafter , we will show that for the curve of general type , we have @xmath528 , where @xmath513 . for this we consider another resultant @xmath529 and show that @xmath530 .
we define @xmath408 in the following way .
let @xmath531 then we have the following equality @xmath532 , where @xmath533 in other words , @xmath408 is the numerator of a rational function @xmath534 .
we often use the following properties of the resultant @xmath535 where @xmath536 .
we need an explicit formula for factors of the resultant @xmath537 where @xmath515 is a leading coefficient with respect to variable @xmath6 of the polynomial @xmath407 . indeed , since @xmath538 then @xmath539 .
thus we obtain the formula in ( [ h1 g ] ) from the property ( [ res - divisor ] ) .
the left side of the formula ( [ h1 g ] ) is equal to @xmath540 .
otherwise , the right side of this formula is equal to @xmath541 .
also , from the property ( [ res - divisor ] ) follows that @xmath542 .
therefore we have and @xmath530 . in the lemma
below , we prove that @xmath543 .
we summarize our computations in the following [ degree ] the degree of the hypersurface @xmath412 with the spine curve @xmath330 which satisfies the assumption ( [ assump ] ) is equal to @xmath544 , where @xmath513 . with the notation as above
we suppose that the conditions ( [ assump ] ) are satisfied .
then we have the equality @xmath545 , where @xmath513 . [ lemma1 ] the weighted degree @xmath546
may be viewed as a degree of a variety @xmath547 , where @xmath548 is a linear projection from the improper point @xmath549 on the lie quadric ( see explicit formula ( [ prj ] ) ) .
the degree of the variety @xmath550 can be computed constructively by counting points of intersection with a general line @xmath551 .
the pre - image of the line @xmath552 is a conic on the lie quadric @xmath245 which passes through the improper point @xmath253 .
hence the degree of @xmath550 is @xmath553 , where @xmath554 is the multiplicity of the intersection @xmath555 at the point @xmath253 .
we need a parametric representation of the general conic @xmath556 .
assume that the conic @xmath103 is in the parameterized plane @xmath557 , where @xmath558 .
the plane @xmath559 intersects a singular cone @xmath560 on two lines .
we choose two vectors @xmath258 in these lines so that the first coordinate is zero , i.e. @xmath561 such that @xmath562 . with the notations as above the general conic
@xmath563 has the following parametrization : @xmath564 since @xmath565 we can compute the multiplicity @xmath566 as follows .
lets denote by @xmath567 the restriction of polynomials @xmath11 to the conic @xmath103 : @xmath568 the computation of the resultant gives @xmath569 , here @xmath570 . on the other side we can compute the number of common points @xmath571 on curves @xmath572 and @xmath573 counted with multiplicities .
this number coincides with @xmath574 ( see @xcite , proposition 5 ) .
the second curve @xmath575 is reducible .
therefore , the resultant with respect to @xmath6 is @xmath576 the second factor has a representation @xmath577 , where @xmath578 and @xmath579 .
it is easy to see that @xmath580 .
if @xmath581 then @xmath582 .
the first curve @xmath572 is hyper - elliptic , i.e. the projection to @xmath6 axes @xmath583 is a map two - to - one .
the curve @xmath572 has the following discriminant with respect to @xmath388 : @xmath584 it is ease to see that point @xmath571 is a singular point on the curve @xmath572 if and only if @xmath582 .
therefore the point @xmath571 has multiplicity at least two as a point of the intersection of two curves @xmath585 .
we will prove that the multiplicity of the intersection of two curves @xmath585 at the point @xmath571 equals to two if @xmath582 .
for simplicity we assume that @xmath586 . the first equation ( [ h1conic ] ) in the local ring @xmath587_{\langle u , t\rangle}$ ] is @xmath588 = [ \tilde{e}_{0}k \langle e , a\rangle , -f k/2,-{\tilde{e}_{{0}}}^{2}\,k/2,\tilde { e}_{{0}}\langle e,{a - b}\rangle , 1/2\,\tilde{e}_{{0}}^{2}]\ \end{aligned}\ ] ] and @xmath589 .
the second equation ( [ h2conic ] ) in the local ring @xmath30 is @xmath590= \nonumber \\ & & [ \tilde{e}_{0}^{2}k\langle e ' , a\rangle,-{f } ' \tilde{e}_{{0}}-\tilde{e}_{0}{e}'_{{0}}k\langle e , a\rangle , k f \left ( { e}'_{{0}}+\,{e}'_{{0}}\right ) , - \tilde{e}_{0}^2k \langle e',{a - b}\rangle , \tilde{e}_{0}{e}'_{{0 } } \langle e,{a - b}\rangle].\nonumber\end{aligned}\ ] ] an easy computation with maple shows that @xmath591 therefore the point @xmath592 has multiplicity two if and only if @xmath593 , i.e. @xmath594 is a unit in the local ring @xmath30 .
a straightforward computation shows that @xmath595 since @xmath596 and @xmath597 by the condition ( [ assump ] ) we conclude that @xmath593 for a general conic .
hence , the multiplicity @xmath598 is equal to @xmath599 . therefore , we have @xmath600 we conjecture that the degree of the hypersurface @xmath412 is @xmath601 , where @xmath287 is defined by the formula ( [ seq - w ] ) .
let @xmath334 be the degree of the spine curve @xmath602 . and let @xmath603 be the degree of the hypersurface @xmath412 with the spine curve @xmath17 . for the general spine curve we have @xmath607 , i.e. @xmath608 .
the lower bound is not clear .
there are examples of spine curves with the following degrees : @xmath609 it seems that there does not exist a spine curve such that @xmath610 .
[ doubleellipsoid ] let us consider the following spine curve : @xmath611 this is a proper parametrization of an ellipse in @xmath262 .
we find the pl@xmath45cker coordinate vector @xmath612 and @xmath613 .
also , we see that @xmath614 and @xmath615 . if we run the @xmath0-basis algorithm with two input vectors @xmath616 we get the output two vectors @xmath617 and @xmath618 : @xmath619 now we can find the implicit equation @xmath620 of the dual variety @xmath621 .
the polynomial @xmath622 contains 26 monomials and has degree 6 ( as in proposition [ degree_of_g ] ) .
the equation of the hypersurface @xmath412 is defined by the polynomial + @xmath623 of degree 8 . since , @xmath624 the 0-envelope of the canal surface @xmath625 is reducible .
the canal surface @xmath5 is the double ellipsoid of revolution @xmath626 . indeed , for the computation of @xmath627 we should assume that the variable @xmath628 and to repeat the same steps as above .
we should consider only the first 5 coordinates of the vectors @xmath629 .
let us denote these two vectors with 5 coordinates by @xmath630 . but this time we see that the pl@xmath45cker vector @xmath631 has a non - trivial common divisor , i.e. @xmath632 .
so , using the @xmath0-basis algorithm we find the @xmath0-basis @xmath633 for the input @xmath634 . in this case
we see that @xmath635 .
now we find the resultant @xmath636 , where @xmath637 .
after the substitution @xmath638 we obtain the implicit equation of the canal surface the double ellipsoid @xmath626 .
we can see this geometrically , too .
the point @xmath639 corresponds to the sphere @xmath640 with a center on the @xmath641-axis .
if @xmath642 $ ] then the sphere @xmath643 is tangent to the ellipsoid @xmath644 , and inside this ellipsoid .
moreover , the real envelope of the family @xmath645 $ ] is the ellipsoid @xmath646 .
note that the sphere @xmath647 has the same center but the opposite radius to the sphere @xmath643 , i.e. it has the opposite orientation
. therefore , the real envelope of the family @xmath648\cap [ 2,\infty)$ ] is the same ellipsoid @xmath646 .
hence , from the point of laguerre geometry the envelope of the whole family @xmath643 is the double ellipsoid @xmath649 .
note , that the d - offset to the canal surface , in this case is the d - offset to ellipsoid @xmath646 and it has degree 8 .
also , we can check that by theorem [ degree ] the degree of the @xmath605 hypersurface is 8 , too . for a detailed study and other examples of canal surfaces with a quadratic spine curve we recommend to look at the paper @xcite .
consider the polynomial spine curve @xmath650 @xmath651 we find the pl@xmath45cker coordinate vector @xmath652 and @xmath613 .
also , we see that @xmath653 and @xmath654 .
if we run the @xmath0-basis algorithm with two input vectors @xmath655 we get the output of two vectors @xmath656 and find the implicit equation @xmath622 of the dual variety @xmath621 ( it contains 54 monomials , so we do not present an explicit formula ) .
finally , we find that @xmath657 . for this example , we have @xmath658 , i.e. the implicit degree of the canal surface is 5 . note
that this contradicts theorem 4 in @xcite , because in this example the degree of the canal surface is an odd number .
it seems that the mentioned theorem gives only an upper bound estimation , but not the exact degree formula of canal surfaces with polynomial spine curve . in the next example
we take the following spine curve : + @xmath659 .
the first three coordinates define the viviani curve , i.e. it is intersection curve of the sphere and the tangent cylinder .
we find the pl@xmath45cker coordinate vector @xmath660 and @xmath613 .
also , we see that @xmath661 .
if we run the @xmath0-basis algorithm with two input vectors @xmath662 we get output of two vectors @xmath663 both of degree 3 and find the implicit equation @xmath622 of the dual variety @xmath621 ( it contains 58 monomials ) .
finally , we find that @xmath664 . for this example , we have @xmath665 , i.e. the implicit degree of the canal surface is 10 .
m. dohm , implicitization of rational ruled surfaces with @xmath0-bases , accepted for publication in journal of symbolic computation , special issue eaca 2006 , preprint available at http://arxiv.org/pdf/math/0702658 r. krasauskas , s. zube , canal surfaces defined by quadratic families of spheres , in : geometric modeling and algebraic geometry , b. jttler , r. piene ( eds . ) , springer , 2007 , pp .
138 - 150 .
( publication : november , 2007 ) | a canal surface is an envelope of a one parameter family of spheres . in this paper
we present an efficient algorithm for computing the implicit equation of a canal surface generated by a rational family of spheres . by using laguerre and lie geometries , we relate the equation of the canal surface to the equation of a dual variety of a certain curve in 5-dimensional projective space .
we define the @xmath0-basis for arbitrary dimension and give a simple algorithm for its computation .
this is then applied to the dual variety , which allows us to deduce the implicit equations of the the dual variety , the canal surface and any offset to the canal surface .
+ _ key words : _ canal surface , implicit equation , resultant , @xmath0-basis , offset [ thm]lemma [ thm]corollary [ thm]conjecture [ thm]proposition [ thm]definition [ thm]remark [ thm]assumption [ thm]example |
-0.1 cm -0.1 cm [ fig : introduction ] -0.5 cm lossy compression ( jpeg , webp and hevc - msp ) is one class of data encoding methods that uses inexact approximations for representing the encoded content . in this age of information explosion ,
lossy compression is indispensable and inevitable for companies ( _ twitter _ and _ facebook _ ) to save bandwidth and storage space .
however , compression in its nature will introduce undesired complex artifacts , which will severely reduce the user experience ( figure [ fig : introduction ] ) .
all these artifacts not only decrease perceptual visual quality , but also adversely affect various low - level image processing routines that take compressed images as input , contrast enhancement @xcite , super - resolution @xcite , and edge detection @xcite . however , under such a huge demand , effective compression artifacts reduction remains an open problem .
we take jpeg compression as an example to explain compression artifacts .
jpeg compression scheme divides an image into 8@xmath08 pixel blocks and applies block discrete cosine transformation ( dct ) on each block individually .
quantization is then applied on the dct coefficients to save storage space .
this step will cause a complex combination of different artifacts , as depicted in figure [ fig : introductiona ] . _ blocking artifacts _ arise when each block is encoded without considering the correlation with the adjacent blocks , resulting in discontinuities at the 8@xmath08 borders . _ ringing effects _ along the edges occur due to the coarse quantization of the high - frequency components ( also known as gibbs phenomenon @xcite ) .
_ blurring _ happens due to the loss of high - frequency components . to cope with the various compression artifacts ,
different approaches have been proposed , some of which can only deal with certain types of artifacts .
for instance , deblocking oriented approaches @xcite perform filtering along the block boundaries to reduce only blocking artifacts .
liew @xcite and foi @xcite use thresholding by wavelet transform and shape - adaptive dct transform , respectively .
these approaches are good at removing blocking and ringing artifacts , but tend to produce blurred output .
jung @xcite propose restoration method based on sparse representation .
they produce sharpened images but accompanied with noisy edges and unnatural smooth regions .
to date , deep learning has shown impressive results on both high - level and low - level vision problems .
in particular , the srcnn proposed by dong @xcite shows the great potential of an end - to - end dcn in image super - resolution .
the study also points out that conventional sparse - coding - based image restoration model can be equally seen as a deep model .
however , we find that the three - layer network is not well suited in restoring the compressed images , especially in dealing with blocking artifacts and handling smooth regions . as various artifacts are coupled together , features extracted by the first layer is noisy , causing undesirable noisy patterns in reconstruction . to eliminate the undesired artifacts , we improve the srcnn by embedding one or more `` feature enhancement '' layers after the first layer to clean the noisy features .
experiments show that the improved model , namely `` artifacts reduction convolutional neural networks ( ar - cnn ) '' , is exceptionally effective in suppressing blocking artifacts while retaining edge patterns and sharp details ( see figure [ fig : introduction ] ) . however , we are met with training difficulties in training a deeper dcn .
`` deeper is better '' is widely observed in high - level vision problems , but not in low - level vision tasks .
specifically , `` deeper is not better '' has been pointed out in super - resolution @xcite , where training a five - layer network becomes a bottleneck .
the difficulty of training is partially due to the sub - optimal initialization settings .
the aforementioned difficulty motivates us to investigate a better way to train a deeper model for low - level vision problems .
we find that this can be effectively solved by transferring the features learned in a shallow network to a deeper one and fine - tuning simultaneously .
this strategy has also been proven successful in learning a deeper cnn for image classification @xcite . following a similar general intuitive idea , _ easy to hard _
, we discover other interesting transfer settings in this low - level vision task : ( 1 ) we transfer the features learned in a high - quality compression model ( easier ) to a low - quality one ( harder ) , and find that it converges faster than random initialization .
( 2 ) in the real use case , companies tend to apply different compression strategies ( including re - scaling ) according to their purposes ( figure [ fig : introductionb ] ) .
we transfer the features learned in a standard compression model ( easier ) to a real use case ( harder ) , and find that it performs better than learning from scratch .
the contributions of this study are three - fold : ( 1 ) we formulate a new deep convolutional network for efficient reduction of various compression artifacts .
extensive experiments , including that on real use cases , demonstrate the effectiveness of our method over state - of - the - art methods @xcite both perceptually and quantitatively .
( 2 ) we verify that reusing the features in shallow networks is helpful in learning a deeper model for compression artifact reduction . under the same intuitive idea
_ easy to hard _ , we reveal a number of interesting and practical transfer settings .
our study is the first attempt to show the effectiveness of feature transfer in a low - level vision problem .
( 3 ) we show the effectiveness of ar - cnn in facilitating other low - level vision routines ( super - resolution and contrast enhancement ) , when they take jpeg images as input .
existing algorithms can be classified into deblocking oriented and restoration oriented methods .
the deblocking oriented methods focus on removing blocking and ringing artifacts . in the spatial domain ,
different kinds of filters @xcite have been proposed to adaptively deal with blocking artifacts in specific regions ( , edge , texture , and smooth regions ) . in the frequency domain ,
liew @xcite utilize wavelet transform and derive thresholds at different wavelet scales for denoising .
the most successful deblocking oriented method is perhaps the pointwise shape - adaptive dct ( sa - dct ) @xcite , which is widely acknowledged as the state - of - the - art approach @xcite .
however , as most deblocking oriented methods , sa - dct could not reproduce sharp edges , and tend to overly smooth texture regions .
the restoration oriented methods regard the compression operation as distortion and propose restoration algorithms .
they include projection on convex sets based method ( pocs ) @xcite , solving an map problem ( foe ) @xcite , sparse - coding - based method @xcite and the regression tree fields based method ( rtf ) @xcite , which is the new state - of - the art method .
the rtf takes the results of sa - dct @xcite as bases and produces globally consistent image reconstructions with a regression tree field model .
it could also be optimized for any differentiable loss functions ( ssim ) , but often at the cost of other evaluation metrics . super - resolution convolutional neural network ( srcnn )
@xcite is closely related to our work . in the study ,
independent steps in the sparse - coding - based method are formulated as different convolutional layers and optimized in a unified network .
it shows the potential of deep model in low - level vision problems like super - resolution .
however , the model of compression is different from super - resolution in that it consists of different kinds of artifacts .
designing a deep model for compression restoration requires a deep understanding into the different artifacts .
we show that directly applying the srcnn architecture for compression restoration will result in undesired noisy patterns in the reconstructed image .
transfer learning in deep neural networks becomes popular since the success of deep learning in image classification @xcite .
the features learned from the imagenet show good generalization ability @xcite and become a powerful tool for several high - level vision problems , such as pascal voc image classification @xcite and object detection @xcite .
yosinski @xcite have also tried to quantify the degree to which a particular layer is general or specific .
overall , transfer learning has been systematically investigated in high - level vision problems , but not in low - level vision tasks . in this study , we explore several transfer settings on compression artifacts reduction and show the effectiveness of transfer learning in low - level vision problems .
our proposed approach is based on the current successful low - level vision model srcnn @xcite . to have a better understanding of our work
, we first give a brief overview of srcnn .
then we explain the insights that lead to a deeper network and present our new model .
subsequently , we explore three types of transfer learning strategies that help in training a deeper and better network .
the srcnn aims at learning an end - to - end mapping , which takes the low - resolution image @xmath1 ( after interpolation ) as input and directly outputs the high - resolution one @xmath2 .
the network contains three convolutional layers , each of which is responsible for a specific task .
specifically , the first layer performs * patch extraction and representation * , which extracts overlapping patches from the input image and represents each patch as a high - dimensional vector
. then the * non - linear mapping * layer maps each high - dimensional vector of the first layer to another high - dimensional vector , which is conceptually the representation of a high - resolution patch . at last , the * reconstruction * layer aggregates the patch - wise representations to generate the final output .
the network can be expressed as : @xmath3 where @xmath4 and @xmath5 represent the filters and biases of the @xmath6th layer respectively , @xmath7 is the output feature maps and @xmath8 denotes the convolution operation .
the @xmath4 contains @xmath9 filters of support @xmath10 , where @xmath11 is the spatial support of a filter , @xmath9 is the number of filters , and @xmath12 is the number of channels in the input image . note that there is no pooling or full - connected layers in srcnn , so the final output @xmath2 is of the same size as the input image .
rectified linear unit ( relu , @xmath13 ) @xcite is applied on the filter responses .
these three steps are analogous to the basic operations in the sparse - coding - based super - resolution methods @xcite , and this close relationship lays theoretical foundation for its successful application in super - resolution .
details can be found in the paper @xcite .
* insights . * in sparse - coding - based methods and srcnn , the first step feature extraction determines what should be emphasized and restored in the following stages .
however , as various compression artifacts are coupled together , the extracted features are usually noisy and ambiguous for accurate mapping . in the experiments of reducing jpeg compression artifacts ( see section [ sec : exp_srcnn ] ) , we find that some quantization noises coupled with high frequency details are further enhanced , bringing unexpected noisy patterns around sharp edges .
moreover , blocking artifacts in flat areas are misrecognized as normal edges , causing abrupt intensity changes in smooth regions .
inspired by the feature enhancement step in super - resolution @xcite , we introduce a feature enhancement layer after the feature extraction layer in srcnn to form a new and deeper network ar - cnn .
this layer maps the `` noisy '' features to a relatively `` cleaner '' feature space , which is equivalent to denoising the feature maps . *
* the overview of the new network ar - cnn is shown in figure [ fig : framework ] .
the three layers of srcnn remain unchanged in the new model .
we also use the same annotations as in section [ sec : srcnn ] . to conduct feature enhancement
, we extract new features from the @xmath14 feature maps of the first layer , and combine them to form another set of feature maps .
this operation @xmath15 can also be formulated as a convolutional layer : @xmath16 where @xmath17 corresponds to @xmath18 filters with size @xmath19 .
@xmath20 is an @xmath18-dimensional bias vector , and the output @xmath21 consists of @xmath18 feature maps .
overall , the ar - cnn consists of four layers , namely the feature extraction , feature enhancement , mapping and reconstruction layer .
it is worth noticing that ar - cnn is not equal to a deeper srcnn that contains more than one non - linear mapping layers . rather than imposing more non - linearity in the mapping stage
, ar - cnn improves the mapping accuracy by enhancing the extracted low - level features .
experimental results of ar - cnn , srcnn and deeper srcnn will be shown in section [ sec : exp_srcnn ] given a set of ground truth images @xmath22 and their corresponding compressed images @xmath23 , we use mean squared error ( mse ) as the loss function : @xmath24 where @xmath25 , @xmath26 is the number of training samples .
the loss is minimized using stochastic gradient descent with the standard backpropagation .
we adopt a batch - mode learning method with a batch size of 128 .
second row : the 5-layer ar - cnn targeted at _ dataa_-@xmath27 .
third row : the ar - cnn targeted at _ dataa_-@xmath28 .
fourth row : the ar - cnn targeted at _ twitter _ data .
green boxes indicate the transferred features from the base network , and gray boxes represent random initialization .
the ellipsoidal bars between weight vectors represent the activation functions.,title="fig : " ] -3.9 cm transfer learning in deep models provides an effective way of initialization .
in fact , conventional initialization strategies ( randomly drawn from gaussian distributions with fixed standard deviations @xcite ) are found not suitable for training a very deep model , as reported in @xcite . to address this issue
, he @xcite derive a robust initialization method for rectifier nonlinearities , simonyan @xcite propose to use the pre - trained features on a shallow network for initialization . in low - level vision problems ( super resolution )
, it is observed that training a network beyond 4 layers would encounter the problem of convergence , even that a large number of training images ( imagenet ) are provided @xcite .
we are also met with this difficulty during the training process of ar - cnn . to this end
, we systematically investigate several transfer settings in training a low - level vision network following an intuitive idea of `` easy - hard transfer '' .
specifically , we attempt to reuse the features learned in a relatively easier task to initialize a deeper or harder network .
interestingly , the concept `` easy - hard transfer '' has already been pointed out in neuro - computation study @xcite , where the prior training on an easy discrimination can help learn a second harder one .
formally , we define the base ( or source ) task as _ a _ and the target tasks as _ @xmath29 _ , @xmath30 .
as shown in figure [ fig : easy - hard1 ] , the base network _
basea _ is a four - layer ar - cnn trained on a large dataset _
of which images are compressed using a standard compression scheme with the compression quality _ qa_. all layers in _ basea _ are randomly initialized from a gaussian distribution .
we will transfer one or two layers of _ basea _ to different target tasks ( see figure [ fig : easy - hard1 ] ) .
such transfers can be described as follows .
* transfer shallow to deeper model . * as indicated by @xcite , a five - layer network is sensitive to the initialization parameters and learning rate .
thus we transfer the first two layers of _ basea _ to a five - layer network _
target@xmath31_. then we randomly initialize its remaining layers _ , and _
@xmath32_. ] and train all layers toward the same dataset _
dataa_. this is conceptually similar to that applied in image classification @xcite , but this approach has never been validated in low - level vision problems .
* transfer high to low quality .
* images of low compression quality contain more complex artifacts .
here we use the features learned from high compression quality images as a starting point to help learn more complicated features in the dcn .
specifically , the first layer of _
target@xmath33 _ are copied from _
basea _ and trained on images that are compressed with a lower compression quality _
* transfer standard to real use case .
* we then explore whether the features learned under a standard compression scheme can be generalized to other real use cases , which often contain more complex artifacts due to different levels of re - scaling and compression .
we transfer the first layer of _ basea _ to the network _
target@xmath32 _ , and train all layers on the new dataset . * discussion .
* why the features learned from relatively easy tasks are helpful ?
first , the features from a well - trained network can provide a good starting point .
then the rest of a deeper model can be regarded as shallow one , which is easier to converge .
second , features learned in different tasks always have a lot in common .
for instance , figure [ fig : features ] shows the features learned under different jpeg compression qualities .
obviously , filters @xmath34 of high quality are very similar to filters @xmath35 of low quality .
this kind of features can be reused or improved during fine - tuning , making the convergence faster and more stable .
furthermore , a deep network for a hard problem can be seen as an insufficiently biased learner with overly large hypothesis space to search , and therefore is prone to overfitting .
these few transfer settings we investigate introduce good bias to enable the learner to acquire a concept with greater generality .
experimental results in section [ sec : transfer ] validate the above analysis .
-0.2 cm -0.16 cm [ fig : features ] -0.4 cm
we use the bsds500 database @xcite as our base training set .
specifically , its disjoint training set ( 200 images ) and test set ( 200 images ) are all used for training , and its validation set ( 100 images ) is used for validation . as in other compression artifacts reduction methods ( rtf @xcite ) ,
we apply the standard jpeg compression scheme , and use the jpeg quality settings @xmath36 ( mid quality ) and @xmath37 ( low quality ) in matlab jpeg encoder .
we only focus on the restoration of the luminance channel ( in ycrcb space ) in this paper .
the training image pairs @xmath38 are prepared as follows images in the training set are decomposed into @xmath39 sub - images @xmath40 . then the compressed samples @xmath41 are generated from the training samples with matlab jpeg encoder @xcite .
the sub - images are extracted from the ground truth images with a stride of 10 .
thus the 400 training images could provide 537,600 training samples . to avoid the border effects caused by convolution , ar - cnn produces a @xmath42 output given a @xmath43 input @xmath44 . hence , the loss ( eqn .
( [ eqn : loss ] ) ) was computed by comparing against the center @xmath42 pixels of the ground truth sub - image @xmath45 . in the training phase ,
we follow @xcite and use a smaller learning rate ( @xmath46 ) in the last layer and a comparably larger one ( @xmath47 ) in the remaining layers .
we use the live1 dataset @xcite ( 29 images ) as test set to evaluate both the quantitative and qualitative performance .
the live1 dataset contains images with diverse properties .
it is widely used in image quality assessment @xcite as well as in super - resolution @xcite . to have a comprehensive qualitative evaluation
, we apply the psnr , structural similarity ( ssim ) @xcite windows as in @xcite .
] , and psnr - b @xcite for quality assessment .
we want to emphasize the use of psnr - b .
it is designed specifically to assess blocky and deblocked images , thus is more sensitive to blocking artifacts than the perceptual - aware ssim index .
the network settings are @xmath48 , @xmath49 , @xmath50 , @xmath51 , @xmath52 , @xmath53 , @xmath54 and @xmath55 , denoted as ar - cnn ( 9 - 7 - 1 - 5 ) or simply ar - cnn .
a specific network is trained for each jpeg quality .
parameters are randomly initialized from a gaussian distribution with a standard deviation of 0.001 .
.the average results of psnr ( db ) , ssim , psnr - b ( db ) on the live1 dataset . [ cols="^,^,^,^,^",options="header " , ] -0.15 cm in table [ tab : transfer ] , we denote a deeper ( five - layer ) ar - cnn as `` 9 - 7 - 3 - 1 - 5 '' , which contains another feature enhancement layer ( @xmath56 and @xmath57 ) . results in figure [ fig : transfer1 ] show that the transferred features from a four - layer network enable us to train a five - layer network successfully .
note that directly training a five - layer network using conventional initialization ways is unreliable .
specifically , we have exhaustively tried different groups of learning rates , but still have not observed convergence .
furthermore , the `` transfer deeper '' converges faster and achieves better performance than using he s method @xcite , which is also very effective in training a deep model .
we have also conducted comparative experiments with the structure `` 9 - 7 - 1 - 1 - 5 '' and observed the same trend . -0.15 cm
-0.15 cm results are shown in figure [ fig : transfer2 ] .
obviously , the two networks with transferred features converge faster than that training from scratch .
for example , to reach an average psnr of 27.77db , the `` transfer 1 layer '' takes only @xmath58 backprops , which are roughly a half of that for `` base - q10 '' .
moreover , the `` transfer 1 layer '' also outperforms the ` base - q10' by a slight margin throughout the training phase .
one reason for this is that only initializing the first layer provides the network with more flexibility in adapting to a new dataset .
this also indicates that a good starting point could help train a better network with higher convergence speed .
online social media like _ twitter _ are popular platforms for message posting .
however , _ twitter _ will compress the uploaded images on the server - side .
for instance , a typical 8 mega - pixel ( mp ) image ( @xmath59 ) will result in a compressed and re - scaled version with a fixed resolution of @xmath60 .
such re - scaling and compression will introduce very complex artifacts , making restoration difficult for existing deblocking algorithms ( sa - dct ) .
however , ar - cnn can fit to the new data easily .
further , we want to show that features learned under standard compression schemes could also facilitate training on a completely different dataset .
we use 40 photos of resolution @xmath59 taken by mobile phones ( totally 335,209 training subimages ) and their _ twitter_-compressed version to train three networks with initialization settings listed in table [ tab : transfer ] . -0.2 cm [ fig : last - fig ] -1.3 cm from figure [ fig : transfer3 ]
, we observe that the `` transfer @xmath61 '' and `` transfer @xmath62 '' networks converge much faster than the `` base - twitter '' trained from scratch .
specifically , the `` transfer @xmath61 '' takes @xmath63 backprops to achieve 25.1db , while the `` base - twitter '' uses @xmath64 backprops . despite of fast convergence ,
transferred features also lead to higher psnr values compared with `` base - twitter '' .
this observation suggests that features learned under standard compression schemes are also transferrable to tackle real use case problems .
some restoration results are shown in figure .
we could see that both networks achieve satisfactory quality improvements over the compressed version .
in the real application , many image processing routines are affected when they take jpeg images as input .
blocking artifacts could be either super - resolved or enhanced , causing significant performance decrease . in this section ,
we show the potential of ar - cnn in facilitating other low - level vision studies , super - resolution and contrast enhancement . to illustrate this , we use srcnn @xcite for super - resolution and tone - curve adjustment @xcite for contrast enhancement @xcite , and show example results when the input is a jpeg image , sa - dct deblocked image , and ar - cnn restored image . from results shown in figure [ fig : application ]
, we could see that jpeg compression artifacts have greatly distorted the visual quality in super - resolution and contrast enhancement .
nevertheless , with the help of ar - cnn , these effects have been largely eliminated . moreover
, ar - cnn achieves much better results than sa - dct .
the differences between them are more evident after these low - level vision processing .
applying deep model on low - level vision problems requires deep understanding of the problem itself . in this paper
, we carefully study the compression process and propose a four - layer convolutional network , ar - cnn , which is extremely effective in dealing with various compression artifacts .
we further systematically investigate several _ easy - to - hard _ transfer settings that could facilitate training a deeper or better network , and verify the effectiveness of transfer learning in low - level vision problems . as discussed in srcnn @xcite
, we find that larger filter sizes also help improve the performance .
we will leave them to further work . | lossy compression introduces complex compression artifacts , particularly the blocking artifacts , ringing effects and blurring .
existing algorithms either focus on removing blocking artifacts and produce blurred output , or restores sharpened images that are accompanied with ringing effects .
inspired by the deep convolutional networks ( dcn ) on super - resolution @xcite , we formulate a compact and efficient network for seamless attenuation of different compression artifacts .
we also demonstrate that a deeper model can be effectively trained with the features learned in a shallow network . following a similar `` easy to hard '' idea
, we systematically investigate several practical transfer settings and show the effectiveness of transfer learning in low - level vision problems .
our method shows superior performance than the state - of - the - arts both on the benchmark datasets and the real - world use case ( _ twitter _ ) .
in addition , we show that our method can be applied as pre - processing to facilitate other low - level vision routines when they take compressed images as input . |
the topological charge of gauge field theory is defined as @xmath1,\ ] ] where @xmath2 and @xmath3 refer to the field strength tensor and its dual , respectively .
it is well - known that quantum chromodynamics(qcd ) contains nontrivial gauge configurations carrying the topological charge with an integer number@xcite .
these configurations interpolate between the topologically different vacua of qcd characterized by different chern - simons numbers . for a long time the experimental evidence for the existence of the topological gluon configurations only comes indirectly from the meson spectrum@xcite .
recently , the study of the chiral magnetic effect(cme)@xcite has received much attentions since the topological configurations of qcd may be observed directly in the noncentral heavy - ion collisions .
the essence of the cme is the chiral imbalance induced by the nonzero topological charge through the axial anomaly of qcd @xmath4 where @xmath5 denotes the quark number of the right - handed(left - handed ) chirality in the chiral limit .
( [ eq : int1aaa ] ) indicates the chiral asymmetry is generated by the nonzero topological excitations . in a strong magnetic field which may be produced in the noncentral heavy - ion collisions , a net electric current will be induced along the direction of the magnetic field due to the chirality imbalance . in order to study the effect of the chiral asymmetry ,
usually the axial chemical potential @xmath6 is introduced in a grand canonical ensemble@xcite .
the physical meaning of @xmath6 is the difference of the right- and left - handed quark chemical potentials .
this quantity can be identified as the time derivative of the @xmath7 angle of qcd , namely @xmath8@xcite . in the presence of an external magnetic field @xmath9 ,
the analytical expression of the induced electric current density for the cme at finite @xmath6 has been given in @xcite which takes the form @xmath10 where @xmath11 .
note that all quarks carrying the same electric charge is assumed in .
obviously , this equation explicitly breaks the @xmath12- and @xmath13-symmetry .
equation.([eq : jz - zeroga ] ) indicates that the hot quark - gluon plasma(hqgp ) is an ideal place to study the physics of @xmath12- and @xmath13-odd excitations of qcd .
first , the production probability of the configurations with nonzero winding numbers becomes higher since jumping over the potential barrier between distinct classical vacua is probable at high temperatures .
the configurations responsible for such thermal transitions are called sphalerons@xcite in qcd , which might occur at a copious rate at high temperature compared to the low - temperature tunneling by the instanton .
second , a strong magnetic field can be produced in non - central heavy - ion collisions@xcite .
the study based on the urqmd model@xcite suggests that the @xmath14 produced at rhic and lhc can be as large as @xmath15 and @xmath16 , respectively , where @xmath17 . ] .
so in a noncentral heavy - ion collision , the induced electric current may lead to an excess of the positive electric charge on one side of the reaction plane and the negative electric charge on the other .
such a charge separation effect may be observed experimentally .
recently , a conclusive observation of charge azimuthal correlation has been presented by the star collaboration @xcite which may result from the cme with local @xmath12- and @xmath13-violation . theoretically , the investigation of the cme has been performed recently in the lattice qcd simulation by introducing a finite @xmath6@xcite .
the advantage of the lattice simulation at finite axial chemical potential is that there is no sign problem comparing with the case at finite baryon chemical potential . in this study ,
the cme described by eq . has been confirmed qualitatively .
but at the quantitative level , the obtained electric current density is significantly suppressed @xcite compared to the analytical formula ( [ eq : jz - zeroga ] ) .
the further study at the quenched level of qcd suggests that the lattice data are sensitive to the lattice spacing@xcite and the obtained data increase significantly compared to the ones in@xcite .
however , the deviation of the new lattice data from the analytical formula is still sizable even some systematic errors have been taking into account@xcite .
the suppression of the lattice data may stem from some errors of the lattice simulation and/or the non - perturbative effects of qcd . accordingly , besides taking more improved lattice calculations , physically understanding the current lattice result through other methods of qcd is very important and necessary . on the other hand , due to the limitation of the lattice calculation , investigating the non - perturbative corrections and making new predictions on the cme
are also very required by using the effective theories or models of qcd .
the purpose of this paper has two aspects : first , we will try to disclose the possible mechanism for the deviation of the present lattice data from the analytical result by taking into account some qcd corrections ; second , we will explore the non - perturbative influence of the instanton molecules on the cme and make some predictions .
note that both aspects are closely related to the effective four - quark interaction in the axial - vector channel , where the chirality density - density correlation plays an important role .
the paper is organized as follows : in section [ sec : effective ] , we give the effective quark interactions of qcd for @xmath18 . in section
[ sec : correction ] , we present the correction to the cme from the axial - vector interaction . section [ sec : numerical ] is devoted to the numerical results and related discussions .
the last section is the summary and conclusion .
the experimental data of rhic suggests that strongly coupled quark - gluon plasma(sqgp ) is formed for @xmath18 , which exhibits the behavior of a perfect fluid .
now it is widely believed that the non - perturbative qcd still plays very important roles on the hqgp up to @xmath19 . since the chiral magnetic effect may be observed directly in heavy - ion collisions
, one naturally expects that the role of the non - perturbative aspects of qcd on the cme should be quite remarkable .
one of the possible non - perturbative effects may arise from the instanton - anti - instanton ( @xmath20 ) molecules at finite temperature near @xmath21 . as a major component of the qcd vacuum
, the instantons can produce the quark condensate , the low - lying hadron state and other non - perturbative features of qcd , according to the interacting instanton liquid model ( iilm ) @xcite . at zero and low temperatures ,
the instanton ensemble is in a random liquid state which is responsible for the spontaneous chiral symmetry breaking and the axial anomaly .
naively one would expect that the instantons might give a little contribution to the cme due to their significant suppression at high temperature .
however , it has been suggested that the hot chiral transition is not driven by the vanishing of the instantons and anti - instantons but by their rearrangement around @xmath21 @xcite .
namely , when approaching to @xmath21 from below , the random instantons and anti - instantons are paired up into the ordered @xmath20molecules .
this picture implies that the instantons still keeps sizable density near @xmath21 with the molecular forms .
moreover , a recent investigation indicates that the @xmath20molecules can even survive up to @xmath22@xcite .
the rearrangement picture indicates that the non - perturbative effects induced by the instanton configurations will persist into the high temperature region . in the chiral symmetric phase
, the interacting @xmath20molecules can be regarded as one of the possible mechanisms responsible for the sqgp formed in @xmath23@xcite .
so even the chiral restoration is one of the necessary conditions for the appearance of the electric current @xmath24 , the instantons may still exert important influence on the chiral magnetic effect via the molecular forms .
it is well known that the random instanton configurations can lead to the famous t hooft interactions among quarks with different flavors @xmath25 which explicitly breaks the @xmath26 symmetry due to the axial anomaly .
similarly , the ordered @xmath20molecules also result in effective quark interactions , but only with the four - fermion coupling forms . for two - flavor case , the fierz - invariant quark interactions induced by @xmath20molecules has been derived in @xcite with the form @xmath27\nonumber\\ & + \frac{2}{n^2_{c}}(\bar{\psi}\gamma_{\mu}\gamma_5\psi)^2-\frac{1}{2n^2_{c}}\sum_{a=0}^{3}[(\bar{\psi}\tau^{a}\gamma_{\mu}\psi)^2\nonumber\\ & + ( \bar{\psi}\tau^{a}\gamma_{\mu}\gamma_5\psi)^2 ] \big\}+\mathcal{l}_8 , \label{eq : molecule}\end{aligned}\ ] ] where @xmath28 and @xmath29 are unit and pauli matrices in flavor space respectively and @xmath30 is the coupling constant .
the @xmath31 in refers the quark interactions in the color octet channel . in the @xmath20molecular picture
, it is expected that the dominant quark interactions for @xmath18 are the molecular ones rather than the t hooft interactions @xcite region according to @xcite .
such interactions should have little influence on the cme since the electric current only appears in the chirally symmetric phase . ] .
different from ( [ eq : thooft ] ) , the lagrangian respects not only the @xmath32 symmetry but also the @xmath33 symmetry .
this is because the @xmath20pair gives vanishing topological charges@xcite .
note that the most general fierz - invariant form for the four - quark interactions under the invariance of @xmath34 symmetry including both the scalar(psudo - scalar ) and vector(axial - vector ) currents is @xmath35\nonumber\\ & -\frac{1}{2}g_2\sum_{a=0}^{3}[(\bar{\psi}\tau^{a}\gamma_{\mu}\psi)^2+(\bar{\psi}\tau^{a}\gamma_{\mu}\gamma_5\psi)^2]\nonumber\\ & -\frac{1}{2}g_3[(\bar{\psi}\tau^{0}\gamma_{\mu}\psi)^2+(\bar{\psi}\tau^{0}\gamma_{\mu}\gamma_5\psi)^2]\nonumber\\ & -\frac{1}{2}g_4[(\bar{\psi}\tau^{0}\gamma_{\mu}\psi)^2-(\bar{\psi}\tau^{0}\gamma_{\mu}\gamma_5\psi)^2]+\mathcal{l}^{(4)}_{8 } , \label{eq : full4f}\end{aligned}\ ] ] where @xmath36 are four independent coupling constants @xcite .
comparing with , one can find that the interacting @xmath37 molecule model ( iimm ) gives prediction of these coupling constants with @xmath38 which are all dependent on the only parameter @xmath30 .
we shall show below that nonzero @xmath39 and @xmath40 give nontrivial contributions to the induced current @xmath24 for the cme .
we stress that even we are very interested in the possible non - perturbative influence of the instantons on the cme , we do not limit our study within the only formalism , which is just a special case of . without loss of generality
, we will assume that the lagrangian density with the four - quark interactions works for @xmath23 .
this lagrangian can be regarded as an extension of the traditional njl model of qcd ( in the next section , the degree of freedom of the polyakov - loop is also included ) . for the physics related to the cme
, we are particularly interested in the effective four - quark interactions in the vector isospin - scalar and axial - vector isospin - scalar channels @xmath41 which appears in both and .
the significance of @xmath42 to the cme is attributed to the appearance of two special vacuum expectation values ( vevs ) in the chirally asymmetric system under the influence of an external magnetic field @xmath9 , namely , @xmath43 and @xmath44@xcite .
the former is the vev of the vector current @xmath45 along the direction of @xmath9 and the latter the chiral charge density conjugate to @xmath6 . as a consequence
, the chiral magnetic effect might be sensitive to the couplings @xmath46 and @xmath47 in .
the prediction of the iimm gives @xmath48 equation indicates two unconventional points induced by the @xmath20molecules : ( 1 ) the vector interaction is attractive while the negative @xmath49 implies that the axial - vector interaction is repulsive ; ( 2 ) the magnitude of @xmath49 is three times that of @xmath50 ( for the totally polarized case , the coupling @xmath49 in the temporal direction is even 12 times that of @xmath50@xcite ) .
this is quite different from the conventional four - quark interactions derived from the fierz transformation of the colored quark - antiquark current - current interaction @xmath51 which arises from the one gluon exchange approximation ( ogea ) of qcd . in the ogea ,
both @xmath50 and @xmath49 are positive with @xmath52 where @xmath53 is the coupling constant in the scalar iso - scalar channel .
note that it is very likely that the effective quark interactions stemming from the ogea can persist into the non - perturbative region of qcd with relatively strong couplings .
a well - known example in the literature is the global color model ( gcm ) of qcd , which is based on the nonlocal colored quark - antiquark current - current interaction @xmath54 where @xmath55 is a function of the strong coupling@xcite .
this model can successfully describe the low - lying hadrons@xcite and give reasonable qcd condensates in the vacuum and at finite temperature and density@xcite . since the axial - vector interaction may be repulsive ( as in the iimm ) or attractive ( as in the ogea or gcm )
, we will treat the coupling @xmath49 in as a free parameter .
the realistic @xmath49 may include contributions from both the @xmath20molecules and the ogea , which should depend on the temperature in general . on the other hand , since the vector interaction is attractive in both the iimm and gcm , we only consider the positive @xmath50 in the following .
working at the mean field level , the lagrangian can be rewritten as @xmath56 we can read two nontrivial points from eq .. first , both squared terms of the condensates @xmath57 and @xmath58 give contributions to the thermodynamical potential .
second , the four - quark interactions in the axial - vector and vector channels lead to a dynamical axial chemical potential @xmath59 , and an effective gauge field @xmath60 along the direction of @xmath9 , respectively . note that both points are not taken into account in @xcite to derive the analytical formula for the cme current . to investigate the dynamical influence of @xmath42 on the cme , we add the kinetic lagrangian for free - interacting quarks , namely @xmath61 to the lagrangian .
we will work in the chiral limit throughout this paper . to describe the confinement - deconfinement transition of qcd
, we also include the dynamics of the polyakov - loop , the action of which in the pure gauge theory denotes as @xmath62 .
our model can be looked as the polyakov - loop - enhanced iimm but with a varying axial - vector coupling @xmath49 . in form
, it is very similar to the so called two - flavor pnjl model@xcite .
one then gets the mean field thermodynamical potential @xmath63 where @xmath64 and @xmath65 is the uv regulator function . @xmath66 and @xmath67 in correspond to the normalized traced polyakov loop and its hermitean conjugate respectively . in
, @xmath68 refers to the spin and @xmath69 denotes the landau level .
the dispersion relations of the quasi - particles derived from the eigenvalues of the dirac operator take the form @xmath70 ^ 2 \ ; , \label{eq : dispersion}\ ] ] where @xmath71 and @xmath72 stands for the electric charge for @xmath73(u or d ) quark . in the above equations ,
@xmath74 and @xmath75 refer to the effective axial chemical potential and the modified momentum along @xmath76 respectively , namely @xmath77 the current @xmath78 , the chirality density @xmath79 , the quark condensate for nonzero @xmath6 . ] @xmath80 and the vev of the polyakov - loop @xmath66 can be determined self - consistently by solving the saddle point equations @xmath81 note that the influence of @xmath50 on the cme has been investigated in @xcite , where only the vector interaction is introduced phenomenologically .
the main conclusion is that @xmath50 may lead to a dielectric correction to the cme current density @xmath82 where @xmath83 @xcite .
when including the axial - vector interaction , the dispersion relation suggests that eq . will be further modified as @xmath84 however , a subtlety on the calculation of the polarization tensor or susceptibility in @xcite has been pointed out by kenji fukushima @xcite : according to eqs.(8 - 10 ) in @xcite , the quantity @xmath85 should be zero since the corresponding static susceptibility @xmath86 is always vanishing .
this implies that the vector interaction does not give any correction to the cme .
actually , this point is also consistent with the nielsen - ninomiya s argument @xcite which is only based on the nonrenormalization of triangle anomalies and the energy conservation . as shown in eq .
( or eq . ) , the nonzero @xmath85 will lead to a modification of the coefficient before @xmath87 ( or @xmath88 in our case ) . nevertheless , the argument proposed by nielsen and ninomiya @xcite does not support the coefficient receiving any correction .
this is because such a coefficient is determined by the the nonrenormalization of triangle anomalies , which is known exactly due to the topological nature of anomalies ( more details of this argument see the first part of sec .
iii in @xcite ) . in light of this ,
the correct formula for the current density should be @xmath89 rather than eq . when considering the axial - vector interaction .
we note that , unlike eqs . and , eq . does not contradict nielsen - ninomiya s argument .
this is because the energy required to remove a particle from the left - handed fermi - surface and add it to the right - handed fermi - surface is no longer @xmath90 but @xmath91 due to the influence of the axial - vector interaction .
in fact , eq . will be obtained naturally if we take the same method as part a of sec .
iii in @xcite but with the replacement of @xmath92 with @xmath74 . in the following
, we will ignore the vector interaction and only consider the effect of the axial - vector interaction on the cme .
using the method @xmath93 and the replacement @xmath94 one can get the integral expression for the calculation of the electric current density in the two - flavor case @xmath95 \;.
\label{eq : jemint}\ ] ] as the case of zero @xmath49 , only the lowest landau level ( lll ) gives a non - vanishing contribution in .
ignoring the correction from @xmath50 , the induced electric current density @xmath24 for the two - flavor case under the influence of @xmath49 becomes @xmath96 as it should be , eq .
indicates that the induced electric current density is ultraviolet finite . unlike the analytical formula
, the chirality density @xmath58 appears directly in eq . .
this density can be obtained self - consistently by @xmath97 according to .
this equation is similar to the determination of the baryon number density at finite @xmath98 when considering the vector interaction in the njl - type model@xcite .
we note that even the formula is obtained from the lll integration , it does nt mean that @xmath24 only depends on the lll for nonzero @xmath49 since the high landau levels also give contributions to the chirality density @xmath58 .
the modified expression of @xmath24 implies that the induced electric current density under the influence of the axial - vector interaction may deviate from the analytical result in the following aspects .
first , the current density @xmath24 is no longer directly proportional to the axial chemical potential @xmath6 and the magnetic field @xmath9 .
the reason is that in general , the dynamical chemical potential @xmath99 does nt linearly depend on @xmath6 and @xmath9 .
second , @xmath99 is @xmath100-dependent ( since @xmath49 and @xmath58 are both @xmath100-dependent ) , so the current density @xmath24 also depends on the temperature .
third , the iimm predicts that the cme will be enhanced significantly : ( 1 ) @xmath99 has the same sign as @xmath6 because of the negative @xmath49 ; ( 2 ) @xmath99 may be sizable since @xmath49 is strong as @xmath53 .
in contrast , the ogea ( or gcm ) predicts that the cme will be weakened by the attractive axial - vector interaction compared to the analytical result .
in this section , we provide numerical results for the influence of the axial - vector interaction on the chirally imbalanced matter in a constant magnetic field .
the coupling @xmath53 , the uv regulator function @xmath101 and the polyakov - loop model @xmath62 in ( [ eq : omega ] ) are all adopted from @xcite , in which the pnjl model was used to study the cme .
the model parameters of the njl part are fitted by reproducing the pion decay constant and the quark condensate in the vacuum .
the @xmath21 for vanishing @xmath6 , @xmath9 , and @xmath49 is @xmath102 with the influence of the polyakov - loop dynamics@xcite .
we will focus on the variation of the ratio @xmath103 with the impact of the coupling @xmath49 .
the quantity @xmath104 corresponds to the ratio @xmath105 , which is the focus of the recent lattice qcd calculations @xcite .
the deviation of @xmath106 from the unity reflects the influence of the axial - vector interaction on the cme compared to the analytical formula .
the ratio of @xmath107 to @xmath108 as a function of @xmath49 for different @xmath6 .
the @xmath109 and @xmath100 are fixed as @xmath110 and @xmath111 , respectively.,title="fig : " ] + in fig.[fig:1 ] , we first show the dependence of @xmath106 on the coupling @xmath49 for different @xmath6 with fixed @xmath112 and @xmath113
. the density of @xmath20molecules might be sizable at this temperature since it is just slightly greater than @xmath21 ( the catalytic effect of the magnetic field on the chiral symmetry breaking can be ignored for such an @xmath109 @xcite ) .
we see that @xmath106 decreases monotonically with increasing @xmath49 .
the electric current density is enhanced by @xmath114 and weakened by @xmath115 .
its deviation from the analytical value is quite significant : for the moderate coupling @xmath116 ( @xmath117 ) , the electric current density increases ( decreases ) by @xmath11862 ( 27 ) percent at @xmath119 ; while for the strong coupling @xmath120 ( @xmath121 ) , the electric current density increases ( decreases ) by @xmath118140 ( 34 ) percent at the same @xmath6 .
this is in contrast to the minor modification of @xmath24 caused by the vector interaction@xcite .
upper panel : the current @xmath107 as a function of @xmath6 for different @xmath49 .
lower panel : the ratio of @xmath107 to @xmath108 as a function of @xmath6 for different @xmath49 .
the @xmath109 and @xmath100 are fixed as @xmath110 and @xmath111 , respectively.,title="fig : " ] + upper panel : the current @xmath107 as a function of @xmath6 for different @xmath49 .
lower panel : the ratio of @xmath107 to @xmath108 as a function of @xmath6 for different @xmath49 .
the @xmath109 and @xmath100 are fixed as @xmath110 and @xmath111 , respectively.,title="fig : " ] figure.[fig:2 ] shows the electric current density @xmath24 and the ratio @xmath106 as functions of @xmath6 for different @xmath49 , where the same @xmath100 and @xmath9 are used as in fig.[fig:1 ] . for @xmath122
, the upper panel indicates that the electric current density is nearly proportional to @xmath6 in the small @xmath6 region , but its deviation from the linear behavior becomes more and more significantly with increasing @xmath6 .
different from this , the electric current density almost linearly depends on @xmath6 up to the moderate @xmath6 region for @xmath123 .
these features are demonstrated more clearly in the lower panel of fig.[fig:2 ] : the ratio @xmath106 monotonically increases with @xmath6 for @xmath122 while almost remains as a constant for @xmath123 .
the same information can be read from fig.[fig:1 ] .
this is because the chirality density rises more rapidly with growing @xmath6 for the repulsive axial - vector interaction .
the dependence of @xmath106 on the temperature with @xmath124 and @xmath125 is displayed in fig.[fig:3 ] . for @xmath114 ,
the ratio @xmath106 monotonically increases with increasing @xmath100 and its deviation from one becomes huge for the strong coupling . for @xmath115 ,
the ratio @xmath106 decreases relatively slowly with growing @xmath100 for both the moderate and strong couplings .
though the change of @xmath106 by @xmath100 is not so significant as @xmath114 , the ratio @xmath106 still decreases by @xmath11810 ( 15 ) percent from @xmath126 to @xmath127 for @xmath128 ( @xmath129 ) .
hence , the temperature has a noticeable effect on the induced current .
the reason is that the density @xmath58 is a monotonically increasing function of @xmath100 , as reported in @xcite .
the ratio of @xmath107 to @xmath108 as a function of @xmath100 for different @xmath49 .
the @xmath6 and @xmath109 are fixed as @xmath130 and @xmath131 , respectively .
, title="fig : " ] + figure.[fig:4 ] indicates that the ratio @xmath106 is not so sensitive to the external magnetic field but still slowly rises(drops ) with @xmath109 for @xmath114(@xmath115 ) .
this is because that the chirality density is enhanced by the magnetic field through the increase of the thermodynamical potential@xcite .
note that such a tendency will slow down or cease for large enough @xmath132 since its catalytic effect on the chiral symmetry breaking : the enhancement of the chiral condensate by the magnetic field will suppress the chirality density .
the ratio of @xmath107 to @xmath108 as a function of @xmath109 for different @xmath49 .
the @xmath6 and @xmath100 are fixed as @xmath133 and @xmath134 , respectively.,title="fig : " ] + figures.1 - 4 imply that the chirality density - density correlation has an important impact on the chiral magnetic effect , no matter the axial - vector interaction arises from the @xmath20pairings , the ogea , or the gcm .
the ratio @xmath106 can deviate significantly from the unity under the combined effects of @xmath49 , @xmath6 , @xmath100 and @xmath9
. the deviation may be positive or negative which depends on the sign of @xmath49 .
this contrasts with the role of @xmath50 on the cme : it only affects the cme obviously when the magnetic field @xmath132 is very large@xcite ; its correction to the cme has nothing to do with @xmath6 and @xmath100 .
let us compare our results with the recent lattice qcd data . in @xcite , the cme
was qualitatively confirmed by the lattice simulation via introducing a finite @xmath6 .
however , the obtained coefficient @xmath135 is around @xmath136 , which is far less than the analytical value 1/@[email protected] .
the further study using the quenched simulation suggests that the suppression is sensitive to the lattice spacing@xcite .
the new lattice coefficient @xmath135 is roughly in the range 0.02 - 0.03 @xcite , which increases greatly compared to the former results .
however , it still much less than the standard analytical value .
moreover , another new point in @xcite is that the coefficient @xmath135 might be dependent on the temperature .
regardless the possible errors in @xcite , we stress that the current lattice data are quite natural when considering the effect of the attractive axial - vector interaction : ( 1 ) according to , the induced electric current is proportional to the effective axial chemical potential @xmath74 , which may be suppressed significantly compared to @xmath6 by the positive @xmath49 . as demonstrated in figs.1 - 4
, the suppression is really remarkable for the moderate and strong couplings .
so the reduction of @xmath74 by the attractive axial - vector interaction can explain why the lattice data of the induced electric current are much less than the analytical value .
let us make a rough comparison . according to fig.7 in @xcite , the lattice ratio @xmath105 is located in @xmath139 for @xmath140(@xmath141 ) .
our numerical results shown in fig.[fig:3 ] suggest that the ratio @xmath106 is in the range of @xmath142 for @xmath140(@xmath143 ) if @xmath144 varies from @xmath129 to @xmath117 .
we see that the lattice data are quite consistent with our results if the axial - vector interaction is attractive and strong enough .
( 2 ) the lattice data suggest that the coefficient @xmath135 or the ratio @xmath105 is insensitive to @xmath6 and @xmath14 .
this feature is also in agreement with our results for the attractive axial - vector interaction , as shown in fig.[fig:2 ] and fig.[fig:4 ] .
( 3 ) one set of data ( @xmath145 with @xmath146 ) in fig.7 of @xcite indicates that the coefficient @xmath135 decreases with growing @xmath100 .
the author has argued that such a behavior might be a lattice artifact .
however , that the induced current density @xmath24 may be suppressed by @xmath100 can be well interpreted by the attractive coupling @xmath49 as demonstrated in fig.[fig:3 ] . according to fig.7 of @xcite , the ratio @xmath105 decreases from @xmath147 to @xmath148 when
increasing @xmath100 from @xmath149 to @xmath150 ; or in other words , the ratio @xmath105 decreases roughly by @xmath151 per @xmath152 .
this is not so far away from the decline rate @xmath153 per @xmath152 for @xmath128 , as indicated in fig.[fig:3 ] .
should change more mildly since the coupling @xmath49 becomes weaker with increasing @xmath100 .
] so to judge whether the coefficient @xmath135 is sensitive to the temperature or not needs more detailed lattice investigation . on the contrary
, the iimm predicts that the cme current or the coefficient @xmath135 will be enhanced by the @xmath20molecules .
clearly , this prediction deviates distinctly from the present lattice data .
this might suggest that the dominant axial - vector interaction for @xmath18 is coming from the ogea .
nevertheless , the current lattice simulation is still rough and far from conclusive(at leat at the quantitative level ) .
first , that the current lattice data are sensitive to the lattice spacing @xcite suggests that the continuum limit is very important for quantitatively understanding the chiral magnetic effect .
second , the main simulations in @xcite are still at the quenched level while the discretization error of the fermion action is larger than that of the gauge action . therefore , more improved quantitative lattice investigation will shed light on the role of the @xmath20molecules on the chiral magnetic effect . in any case
, we stress that , due to the contribution of the dynamical axial chemical potential , the lattice result of the induced current for a finite @xmath6 does nt need to be quantitatively consistent with the analytical formula . if no other qcd modifications , the coefficient @xmath135 or the ratio @xmath106 should be greater ( less ) than the standard analytical value for @xmath114 ( @xmath115 ) .
the chiral tricritical point ( tcp ) on the @xmath0 phase diagram has been found in @xcite , which locates at the relatively larger @xmath6 region ( in the chiral limit for vanishing quark number density ) .
the tcp is also confirmed in the linear sigma model , whose location is very close to the @xmath100-axis@xcite .
these studies show that the @xmath0 phase diagram is somehow similar to the @xmath154 phase diagram of qcd . here
we will demonstrate the role of the axial - vector interaction on the @xmath0 phase diagram .
we are particularly interested in its influence on the tcp . in fig.[fig:5 ] , we show the locations of the tcp for different axial - vector interactions . we see that the tcp is quite sensitive to the axial - vector coupling .
the negative @xmath49 shifts the tcp towards the smaller @xmath6 region while the positive @xmath49 moves it towards the larger @xmath6 area .
this implies that the first - order chiral transition is strengthened by the repulsive axial - vector interaction induced by @xmath20 molecules while weakened by the attractive axial - vector interaction stemming from the ogea .
the locations of the chiral tcp in the @xmath0 plane for different @xmath49.,title="fig : " ] + it is well - known that the attractive vector interaction can soften the chiral transition and move the critical point(cp ) towards the larger @xmath98 area via generating a dynamical quark chemical potential @xcite .
especially , when taking into account the two - flavor color superconductivity , the vector interaction combined with the charge - neutrality constraint and/or the axial anomaly can even lead to multiple chiral critical points @xcite .
we observe from fig.[fig:5 ] that the attractive axial - vector interaction at finite @xmath6 has a similar influence on the chiral phase transition as the vector interaction does at finite @xmath98 .
however , in the instanton molecule model , the axial - vector interaction at finite @xmath6 and the vector interaction at finite @xmath98 have opposite effects on the chiral transition . since the induced @xmath49 and @xmath50 have different signs , the chiral transition will be strengthened ( softened ) at finite @xmath6 ( @xmath98 ) . in addition ,
the influence of the axial - vector interaction on the chiral transition at finite @xmath6 could be more remarkable since the coupling @xmath49 is much stronger than @xmath50 in the instanton molecule picture .
locating the cp on the @xmath154 phase diagram is a focus of the current studies on the qcd phase transition and heavy - ion collisions .
so far , the location and even the existence of the cp is still under debate .
since there is no sign problem at finite @xmath6 , it is very interesting to explore the @xmath0 phase diagram of the chirally imbalanced matter and locate the tcp by means of the lattice simulation .
such an investigation may provide useful information on the true qcd phase transition .
moreover , studying the @xmath0 phase diagram through lattice qcd may shed light on the non - perturbative features of sqgp .
for example , we can test the instanton molecule picture of qcd by investigating the @xmath0 phase diagram .
in this article , we have first investigated the influence of the axial - vector interaction on the chiral magnetic effect by introducing an axial chemical potential @xmath6 .
such an interaction can be induced by the instanton molecules or derived from the one gluon exchange approximation of qcd . in the presence of a finite chirality density @xmath58 , a dynamical axial chemical potential @xmath99
is generated through the density - density correlation .
we derived that the induced electric current density @xmath24 in an external magnetic field @xmath9 linearly depends on the effective axial chemical potential @xmath74 which including the contribution of @xmath99 .
accordingly , the obtained @xmath24 deviates from the standard analytical formula and explicitly depends on the chirality density @xmath58 . in general
, the deviation relies on the axial - vector coupling @xmath49 , the axial chemical potential @xmath6 , the temperature @xmath100 , and the magnetic field @xmath132 . for @xmath115 from the one gluon exchange approximation
, the induced current density can be significantly suppressed comparing with the analytical value .
we also find that the ratio @xmath155 is insensitive to the axial chemical potential and the magnetic field but decreases with increasing temperature if @xmath115 .
these features are quite consistent with the recent lattice results at finite @xmath6 .
hence , the suppression of the lattice data comparing with the analytical formula can be attributed to the influence of the attractive axial - vector interaction of qcd . actually , it is very likely that the deviation of the lattice result originates from the difference between @xmath6 and @xmath74 .
the reason is that nielsen - ninomiya s argument does not support the renormalization of cme if @xmath6 is not shifted dynamically . on the contrary , an unconventional prediction from the instanton molecule model
is that the axial - vector interaction is repulsive with @xmath114 and much stronger than the vector interaction . as a consequence
, the chiral magnetic effect will be enhanced significantly by the instanton molecules near @xmath21 .
in such a picture , the ratio @xmath156 increases with both @xmath6 and @xmath100 .
these features deviate obviously from the current lattice results .
since the present lattice simulation at finite @xmath6 is still rough and sensitive to the lattice spacing , we anticipate the more improved lattice calculation to test these predictions .
in addition , we also demonstrated that the axial - vector interaction plays an important role on the @xmath0 phase diagram of the chirally imbalanced matter .
the repulsive ( attractive ) axial - vector interaction can effectively strengthen ( weaken ) the chiral phase transition if the coupling is relatively strong .
the chiral tcp on the @xmath0 plane is shifted towards the lower ( higher ) @xmath6 region by the repulsive ( attractive ) axial - vector coupling .
this is quite different from ( similar to ) the role of the vector interaction on the chiral phase transition at finite baryon chemical potential @xmath98 .
since there is no sign problem at finite @xmath6 , the improved lattice simulation is still very welcome for disclosing the non - perturbative qcd effect on the chiral magnetic effect .
moreover , studying the deviation of the induced current density from the analytic result and locating the tcp in @xmath0 plane by the lattice simulation can be used to test the instanton molecule picture for @xmath18 .
we note that if @xmath6 is introduced by coupling it only to the chiral density , it ca nt be considered as a true chemical potential .
the reason is that the axial charge is not a conserved quantity due to the axial anomaly .
recently , it is proposed by robakov @xcite that @xmath6 should be conjugated to a proper combination of the chiral density and a chern - simon term , which is a conserved quantity .
it is then claimed in @xcite that the fact @xmath6 must be associated with a conserved charge is essential in the discussion for the nonrenormalization of the cme current .
since there exists a subtlety on the definition of @xmath6 , a question naturally arises that whether the dynamical axial chemical potential @xmath157 defined in this paper can be associated with a conserved charge .
we point out that @xmath157 does nt need to be conjugated to a conserved charge if @xmath6 could be really well - defined .
the reason is that @xmath157 is an induced quantity generated dynamically by the axial - vector interaction for nonzero @xmath58 .
this is different from the parameter @xmath6 ( assuming it has been well - defined as a true chemical potential ) , which should be associated with a conserved charge in principle .
we note that no matter how @xmath6 is introduced , it must give rise to nonzero @xmath58 .
as demonstrated in this paper , the @xmath157 induced by @xmath58 will modify the chiral imbalance described originally by @xmath6 .
a natural extension of this work is to investigate the role of the axial - vector interaction at finite @xmath98 in the presence of an external magnetic field .
conjugate to the induced current for the chiral magnetic effect at finite @xmath6 , an axial - vector current along the direction of the external magnetic field is generated at finite @xmath98@xcite .
another related topic at finite @xmath98 is the so called chiral shift parameter@xcite in a magnetic field .
the roles of the axial - vector interaction in these cases will be reported elsewhere@xcite .
99 a. a. belavin , a. m. polyakov , a. s. schwartz and yu .
s. tyupkin , phys . lett .
b * 59 * , 85 ( 1975 ) .
e. witten , nucl .
b * 156 * , 269 ( 1979 ) ; g. veneziano , nucl .
b * 159 * , 213 ( 1979 ) .
d. e. kharzeev , l. d. mclerran and h. j. warringa , nucl .
a * 803 * , 227 ( 2008 ) [ arxiv:0711.0950 [ hep - ph ] ] .
l. d. mclerran , e. mottola and m. e. shaposhnikov , phys . rev .
d * 43 * , 2027 ( 1991 ) .
g. d. moore , phys .
b * 412 * , 359 ( 1997 ) ; g. d. moore and k. rummukainen , phys . rev .
d * 61 * , 105008 ( 2000 ) ; d. bdeker , g. d. moore and k. rummukainen , phys .
d * 61 * , 056003 ( 2000 ) ; g. d. moore , arxiv : hep - ph/0009161
. v. skokov , a. y. illarionov and v. toneev , int .
j. mod .
a * 24 * , 5925 ( 2009 ) [ arxiv:0907.1396 [ nucl - th ] ] .
a. yamamoto , phys .
* 107 * , 031601 ( 2011 ) [ arxiv:1105.0385 [ hep - lat ] ] .
a. yamamoto , arxiv:1111.4681 [ hep - lat ] .
t. schafer and e. v. shuryak , rev .
phys . * 70 * , 323 ( 1998 ) [ arxiv : hep - ph/9610451 ] .
e. m. ilgenfritz and e. v. shuryak , phys .
b * 325 * , 263 ( 1994 ) [ arxiv : hep - ph/9401285 ] .
t. schafer , e. v. shuryak and j. j. m. verbaarschot , phys . rev .
d * 51 * , 1267 ( 1995 ) [ arxiv : hep - ph/9406210 ] .
o. wantz and e. p. s. shellard , nucl .
b * 829 * , 110 ( 2010 ) [ arxiv:0908.0324 [ hep - ph ] ] .
s. klimt , m. f. m. lutz , u. vogl and w. weise , nucl .
a * 516 * , 429 ( 1990 ) .
r. t. cahill and c. d. roberts , phys .
d * 32 * , 2419 ( 1985 ) .
p. c. tandy , prog .
* 39 * , 117 ( 1997 ) [ nucl - th/9705018 ] .
z. zhang and w. -q .
zhao , phys .
b * 610 * , 235 ( 2005 ) [ hep - ph/0406210 ] ; phys .
b * 612 * , 207 ( 2005 ) [ hep - ph/0409100 ] ; phys .
b * 617 * , 157 ( 2005 ) [ hep - ph/0603183 ] ; z. zhang and y. -x .
liu , phys .
c * 75 * , 035201 ( 2007 ) [ hep - ph/0603252 ] .
k. fukushima , phys .
b * 591 * , 277 ( 2004 ) [ hep - ph/0310121 ] . c. ratti , m. a. thaler and w. weise , phys .
d * 73 * , 014019 ( 2006 ) [ hep - ph/0506234 ] .
s. roessner , c. ratti and w. weise , phys .
d * 75 * , 034007 ( 2007 ) [ hep - ph/0609281 ] .
z. zhang and y. -x .
liu , phys .
c * 75 * , 064910 ( 2007 ) [ hep - ph/0610221 ] .
k. fukushima , m. ruggieri , phys .
* d82 * , 054001 ( 2010 ) .
[ arxiv:1004.2769 [ hep - ph ] ] .
k. fukushima , m. ruggieri and r. gatto , phys .
d * 81 * , 114031 ( 2010 ) [ arxiv:1003.0047 [ hep - ph ] ] .
m. n. chernodub and a. s. nedelin , phys .
d * 83 * , 105008 ( 2011 ) [ arxiv:1102.0188 [ hep - ph ] ] .
m. kitazawa , t. koide , t. kunihiro , y. nemoto , prog .
* 108 * , 929 - 951 ( 2002 ) .
[ hep - ph/0207255 , hep - ph/0307278 ] .
z. zhang , k. fukushima , t. kunihiro , phys .
rev . * d79 * , 014004 ( 2009 ) .
[ arxiv:0808.3371 [ hep - ph ] ] .
z. zhang , t. kunihiro , phys .
rev . * d80 * , 014015 ( 2009 ) .
[ arxiv:0904.1062 [ hep - ph ] ] ; phys . rev . *
d83 * , 114003 ( 2011 ) .
[ arxiv:1102.3263 [ hep - ph ] ] ; t. kunihiro , y. minami and z. zhang , prog .
suppl . * 186 * , 447 ( 2010 ) .
t. hatsuda , m. tachibana , n. yamamoto and g. baym , phys .
lett . * 97 * , 122001 ( 2006 ) [ hep - ph/0605018 ] .
h. abuki , g. baym , t. hatsuda and n. yamamoto , phys .
d * 81 * , 125010 ( 2010 ) [ arxiv:1003.0408 [ hep - ph ] ] .
v. a. rubakov , arxiv:1005.1888 [ hep - ph ] .
m. a. metlitski and a. r. zhitnitsky , phys .
d * 72 * , 045011 ( 2005 ) [ arxiv : hep - ph/0505072 ] . | the recent lattice calculation at finite axial chemical potential suggests that the induced current density of the chiral magnetic effect ( cme ) is somehow suppressed comparing with the standard analytical formula .
we show in a njl - type model of qcd that such a suppression is a natural result when considering the influence of the attractive axial - vector interaction .
we point out that the lattice result does nt need to be quantitatively consistent with the analytical formula due to the chirality density - density correlation .
we also investigate the nonperturbative effect of instanton molecules on the cme .
since an unconventional repulsive axial - vector interaction is induced , the cme will be enhanced significantly by the instanton - anti - instanton pairings .
such a prediction needs to be tested by more improved lattice simulations .
we further demonstrate that the axial - vector interaction plays an important role on the @xmath0 phase diagram . |
the concept of full counting statistics has been introduced in very early days of quantum transport .
that time , nobody ever thought of such an abstract and complicated problem as the evaluating and measuring the higher cumulants of electronic noise .
the research has been driven by pure curiosity and has resulted in compact and deep levitov formula.@xcite the significance of this contribution has been underappreciated for a number of years .
the direct experimental verification seemed to be out of question , while the theorists opted to address more primitive and standard problems .
nowadays full counting statistics is a reasonably established field attracting attention of many , although the peak of interest is probably in the past.@xcite there are beautiful experiments where the higher cumulants have been measured @xcite and even single electron transfers have been actually counted.@xcite theoretically , the full counting statistics has been evaluated for virtually any important electron transport system including even so generic as anderson impurity model.@xcite still it remains underappreciated that the full counting statistics is eventually not about the marginal deviations of electric currents .
let as draw a parallel with general relativity .
the general relativity is not an art to calculate ridiculously small corrections to newton s law , although only those can be verified experimentally .
the general relativity brings us true knowledge about the universe .
similar to that , full counting statistics lies near the heart of quantum theory of electricity and is in fact an indispensible element for this .
the present contribution aims to explicate the link between full counting statistics and quantum field theory of electric circuits .
the most established example of such theory is the quantum theory of a superconducting josephson junction in a dissipative electromagnetic environment @xcite that is readily reduced to a single - variable field theory for the superconducting phase across the junction @xcite .
the action of the same type governs coulomb blockade phenomena in non - superconducting systems.@xcite the physics of one - dimensional interacting electrons in the framework of luttinger model is often reduced to similar schemes @xcite , where the variable is a drop of a phase over the barrier present in the one - dimensional setup and can be associated with the voltage drop at the barrier .
the structure of the article is as follows .
we start ( sections 2,3 ) by formulating a general quantum theory of a simplest electric circuit and see the need and advantage of the fcs in this respect .
we show that the classical limit of the field theory is not trivial as far as fcs is concerned and obtain the fcs at current bias ( section 4 ) .
quantum effects at big conductances ( as compared with the conductance quantum @xmath0 ) can be incorporated by the renormalization of the action parameters .
this is frequently the case in field theories @xcite we perform the renormalization procedure explicitly in section 5 for a quantum contact in series with an ohmic one and for a set of quantum contacts connected to a single node ( section 6 ) .
the concrete results revisited here were first published in @xcite,@xcite , and @xcite .
let us first start with elementary electric circuit theory and reason the quantum extension of it .
a circuit is made of three archetypal elements : terminals , connectors , and nodes .
the voltage is fixed in terminals .
a connector is characterized by its @xmath1 characteristics : @xmath2 where we have assumed general relation between the voltage and current so that current at the time moment @xmath3 depends on time - dependent voltage at all ( previous ) time moments , @xmath4 .
a simplest circuit contains two connectors ( @xmath5 and @xmath6 ) in series , so that a single node and two terminals .
1a ) connecting elements in this way brings about an extra variable : voltage @xmath7 in the node . in the elementary circuit theory under consideration ,
this voltage @xmath7 is determined from the current conservation in the node , @xmath8 assuming the terminal voltages are fixed to @xmath9 and @xmath10 .
once the voltage is determined , one finds the @xmath1 characteristics of the whole circuit .
thereby , the full description of the system naturally emerges from the two descriptions of the separate connectors . ) in series with ohmic one ( characterized by the frequency - dependent impedance @xmath11 ) .
, width=453 ] let us try to reason a quantum extension of this theory .
first of all , it is convenient to change from voltages to phases defined as @xmath12 .
this allows to treat superconducting and non - superconducting systems on equal footing .
the phases(voltages ) of the terminals can be regarded as time - dependent external parameters while the phase of the node becomes a real quantum variable , that is , an operator @xmath13 .
however , it is hardly an option to formulate the theory in operator formalism since finding the classical correspondence becomes a formidable task .
rather , we shall opt for feynman - vermon or keldysh - action description of the system where all observables can be presented as path integrals over the time - dependent non - operator variable @xmath14 . the price to pay for this convenience
is quite known : the variable `` doubles '' .
the point is that the path integral should performed over two parts of the keldysh contour that correspond to coherent evolution of `` kets '' and `` bra '' s .
the variable in principle takes different values @xmath15 at these two contours .
we still want the description of the compound system under consideration to emerge from the two descriptions of the separate connectors @xmath5 and @xmath6 .
what descriptions ?
since the variable has `` doubled '' , the simple @xmath1 characteristics does not suffice .
a connector has to be described by something that depends on both variables @xmath15 .
it is also clear that from all possible quantum variables of the connector such description should involve only one : the operator of electric current through the connector , @xmath16 , and present the reaction of the current on the variables @xmath15 . the proper description of the connector is thus provided by the feynman - vernon influence functional of two variables @xmath17 , @xmath18=\left\langle\overleftarrow{\rm t}\exp\left\{\frac{i}{e}\int dt\bigl[\phi(t)+{\textstyle\frac{1}{2}}\chi(t)\bigr]\hat{i}(t)\right\ } \hat{\rho } \overrightarrow{\rm t}\exp\left\{\frac{i}{e}\int dt\bigl[-\phi(t)+{\textstyle\frac{1}{2}}\chi(t)\bigr ] \hat{i}(t)\right\}\right\rangle.\label{zphichi}\end{aligned}\ ] ] here , the trace is over the electronic degrees of freedom specific for the connector :
thereby they are `` traced out '' and never explicitly enter our field theory .
the notation @xmath19 denotes time - ordering of the exponentials in ascending ( descending ) order , these exponents presenting the quantum evolution of density matrix @xmath20 subject to the fields @xmath21 .
the easiest way to understand the functional is to expand its log in terms of @xmath22 at @xmath23 .
the coefficents of the expansion present the cumulants of the time - ordered current operators in the connector that is subject to the external classical phase @xmath14 .
in particular , @xmath24 , @xmath25 , @xmath26 and so on . therefore , the functional @xmath27 is nothing but the generating function of the current fluctuations . for quantum conductors and slow - varying @xmath21
it has been first considered in @xcite now we are ready to build up the quantum description of the whole circuit .
since the circuit is nothing but a compound connector , it has to be characterized by the similar generating function @xmath28 that depends on the drop of the fields @xmath21 over the circuit .
such functional is nothing but a path integral convolution of the functionals of the separate connectors @xmath29 and @xmath30 , @xmath31=\int{\cal d}\phi_{1}{\cal d}\chi_{1}\,{\cal z}_{a}[\phi_1,\chi_1]{\cal z}_{b}[\phi-\phi_{1},\chi-\chi_{1}].\label{concatenation}\ ] ] where the path integration measure @xmath32 .
the overall generating function is the average over fluctuating phases @xmath33 at the node of the circuit shared by both conductors .
such convolution law is nothing but the presentation of the current conservation in the node .
one can see it if one substitutes @xmath28 in the form ( [ zphichi ] ) and carries out the integration over @xmath34 .
this is the field theory of the simplest single - node circuit .
the extension to a more complicated circuit is straightforward .
the functional is a product of @xmath27 s of all conductors integrated over the extra variables @xmath21 defined in each node of the circuit .
this functional depends on the phases @xmath35 applied in each terminal and counting fields @xmath22 defined in terminals so that it gives statistics of the currents to / from each terminal of the circuit .
the difference between classical and quantum effects is not readily manifested in the field theory under consideration .
indeed , the current fluctuations in the connectors may be of classical as well as of quantum origin but the presentation of these fluctuations is almost the same .
for the theory in hand , it is constructive to define the difference between classical and quantum as the difference between _ low_-frequency and _ high_-frequency regimes .
this is very much like the theory of frequency - dependent noise : it is known that at sufficiently low frequencies ( @xmath36 for equilibrium systems ) any noise can be regarded as classical irrespective of its origin while quantum mechanics becomes important at higher frequencies .
in general the functional dependence of a connector functional @xmath27 one the phases may be complicated and non - local in time .
however , one expects on physical grounds the non - locality to vanish at sufficiently low frequencies : the current and its statistics at the time moment @xmath3 would only depend on the voltage at the same moment .
therefore , at sufficiently slow realizations of the fluctuating phases the functional @xmath27 can be expressed in terms of a single function @xmath37 , @xmath38=\int dt\,s\bigl(\dot{\phi}(t),\chi(t)\bigr),\label{zadef}\ ] ] ( here we for simplicity specify to non - superconducting systems ) .
let us now concatenate two connectors and try to assess the _ low - frequency _ limit of the resulting @xmath28 taking the path integral in ( [ zphichi ] ) .
there can be two importantly distinct cases
. it may be that the path integral is dominated by low - frequency @xmath34 so that the result is determined by low - frequency limits of @xmath39 .
moreover , in this case the path integral can be evaluated in the saddle - point approximation with the result @xmath40 here @xmath41 and @xmath42 stand for the ( generally complex ) values of @xmath43 and @xmath44 at the saddle point where the derivatives with respect to these phases vanish . since in field theory the relevance of saddle - point approximation
is normally associated with classical behavior , we call this classical limit . generally , high - frequency fluctuations of the fields also contribute to the path integral
. the effect of this contribution is that the actual @xmath45 deviates from the @xmath46 evaluated with the saddle - point method .
adopting again the common field theory terms , we call the deviation the _ quantum correction_. depending on the connector parameters , the correction can be vanishingly small or overwhelming
. the next statement might look less obvious : the quantum correction present the effect of electron - electron interaction in the system and in the context of quantum transport is commonly referred to as ( dynamical ) coulomb blockade effect .
first puzzled question : where is a capacitor of capacitance @xmath47 providing the charging energy @xmath48 that should accompany any passage about coulomb blockade ?
well , it always present ( fig .
1b ) as a capacitance between the node and ground .
the capacitive connector contributes to ( [ zphichi ] ) with @xcite @xmath49 this term suppresses the high - frequency fluctuations of the fields in the node and is usually needed for proper _ regularization _ of the theory since it provides the high - frequency cut - off .
common wisdom of quantum transport suggests that coulomb interaction is weak provided the typical conductance of the connectors @xmath50 exceeds by far the conductance quantum and is dominating otherwise .
one can see this from the estimation of phase fluctuations around the saddle point , @xmath51 .
the fluctuation is @xmath52 provided @xmath53 .
we will see that in this regime the quantum corrections are small although logarithmically diverge at small frequency .
in opposite case our system develops a strong coulomb gap and becomes a sort of set transistor .
we end this section with concrete examples of connectors .
ohmic connectors are linear conductors and exhibit gaussian current fluctuations . in terms of fourier components of the fields , @xmath54 @xmath11 being the frequency - dependent impedance ( resistance ) of the connector .
since the action is of gaussian type , the field theory is completely trivial .
any circuits made of ohmic connectors are reduced to ohmic connectors as well . the ref .
@xcite has provided the connector action for a tunnel connector between the superconducting and/or normal leads , that appeared to be non - gaussian even for normal case . in the simplest low - energy limit
such connector is a josephson junction characterized by josephson energy @xmath55 and corresponding action @xmath56 an arbitrary coherent quantum connector is characterized by the set of transmission coefficients @xmath57 . the fcs studies @xcite have demonstrated that in low - frequency limit the functional corresponds to @xmath58 .
\label{sxi0}\ ] ]
let us investigate the saddle point of our field theory for a quantum conductor ( a ) of conductance @xmath50 in series with an ohmic conductor ( b ) of low - frequency resistance @xmath59 .
it is instructive to assume that the ohmic conductor is noiseless , that is , is kept at vanishing temperature .
its contribution to the action thus reads ( see ( [ ohmic ] ) @xmath60 the subject of interest is how the current fluctuations produced in the quantum conductor are distributed in the whole circuit .
if the ohmic resistor is small , @xmath61 , we have fcs of the quantum conductor at voltage bias . increasing @xmath59 to @xmath62
, we achieve the current bias for this quantum conductor and access the fcs in this limit .
we apply general eqs .
( [ zadef ] ) and ( [ saddlepoint ] ) to our specific circuit that is driven by the voltage source @xmath63 . to avoid significant quantum corrections to this field theory ( coulomb blockade effects @xcite ) we assume that @xmath64 at frequencies @xmath65 where the quantum corrections start to form .
the zero - frequency impedance @xmath59 can have any value .
both the voltage drop @xmath66 at the quantum conductor and the current @xmath67 through the conductor fluctuate in time for finite @xmath59 , with averages @xmath68 , @xmath69 .
voltage bias corresponds to @xmath70 and current bias to @xmath71 , with @xmath72 the imposed current .
there are three characteristic time scales : @xmath73 , @xmath74 , and the @xmath75-time of the circuit .
the low - frequency regime on which we concentrate is reached for current and voltage fluctuations that are slow on any of these time scales .
we seek the cumulant generating function of charge @xmath76 where @xmath77 is the @xmath78-th cumulant of the charge transferred during the time interval @xmath79 .
it is directly related to the keldysh action in the saddle point ( [ saddlepoint ] ) by @xmath80 to characterize the fluctuations of the voltage across the quantum contact , we will also need the cumulant generating function of phase , @xmath81 .
we use that in the absence of noise in the ohmic connector @xmath82 .
therefore , @xmath83 is related to @xmath84 by a change of variables .
the relation is @xmath85 @xmath86 being the phase change induced by external voltage in time interval @xmath79 , @xmath87 . in the limit @xmath88 of voltage bias
the saddle point of the keldysh action is at @xmath89 , @xmath90 , and from eqs .
( [ saddlepoint ] ) , ( [ fxidef ] ) , and ( [ gxidef ] ) one recovers the results of ref .
@xcite : the cumulant generating function @xmath91 and the corresponding probability distribution @xmath92^{\phi_{0}}. \label{pq0}\ ] ] we note that the parameter @xmath93 is in fact the number of attempted transmissions per channel .
the first few cumulants are @xmath94 , @xmath95 , @xmath96 . in the single - channel case ( @xmath97 )
the distribution ( [ pq0 ] ) has the binomial form ( [ pqresult ] ) . after these preparations we are now ready to generalize all of this to finite @xmath59 , and in particular to derive the dual distribution of phase ( [ pphiresult ] ) under current bias . calculating saddle - point values of @xmath34 from eqs.([saddlepoint ] ) and ( [ calfdef ] )
we observe that ( @xmath98 ) @xmath99,\;\ ; \sigma+z{\cal s}(\sigma)=\xi.\label{fxi}\ ] ] the implicit function @xmath100 ( which is determined from the saddle point of the action ) provides the cumulant generating function of charge @xmath101 for arbitrary series resistance @xmath59 .
one readily checks that @xmath102 in the limit @xmath103 , as it should . by expanding eq .
( [ fxi ] ) in powers of @xmath104 we obtain a series of relations between the cumulants @xmath77 of charge at @xmath105 and the cumulants @xmath106 at @xmath107 .
for example , to linear order we find @xmath108 , which nothing but the trivial division of voltage : the mean current @xmath109 is rescaled by a factor @xmath110 coming from the series resistance
. naively , one may assume that the same rescaling applies to the fluctuations .
indeed , to second order one finds @xmath111 , in agreement with elementary circuit theory .
however , if we go to higher cumulants we find that other terms appear , which can not be incorporated by any rescaling .
for example ,
( [ fxi ] ) gives for the third cumulant @xmath112 while the first term on the the right - hand - side has the expected scaling form , the second term does not .
this is generic for @xmath113 : @xmath114 plus a non - linear ( rational ) function of lower cumulants @xcite .
all terms are of the same order of magnitude in @xmath115 . turning now to the limit @xmath116 of current bias , we see from eq .
( [ fxi ] ) that @xmath117 with @xmath118 defined in terms of the functional inverse @xmath119 of @xmath120 .
the parameter @xmath121 ( which assumed to be an integer @xmath122 ) is the number of charges transferred by the bias current @xmath123 during the time interval @xmath79 . transforming from charge to phase variables by means of eq.([gxidef ] )
, we find that @xmath124 with @xmath125 it is interesting to discuss a single - channel conductor ( transmission @xmath126 ) separately . in this case the functional inverse gives the function of a similar form .
( [ ginfty ] ) reduces to @xmath127 $ ] , corresponding to the pascal distribution ( [ pphiresult ] ) .
the first three cumulants are @xmath128 , @xmath129 , @xmath130 . while the charge @xmath131 for voltage bias @xmath132 is known to have the binomial distribution @xcite @xmath133 we find that the dual distribution of phase @xmath134 for current bias @xmath135 is the pascal distribution @xcite @xmath136 ( both @xmath137 and @xmath138 are integers for integer @xmath93 and @xmath139 . ) in the more general case not depending on the type of the quantum conductor we have found that the distributions of charge and phase are related in a remarkably simple fashion for @xmath140 : @xmath141 ( the remainder @xmath142 equals @xmath143 in the shot - noise limit . )
this formula , which valid with logarithmic accuracy , is a manifestation of charge - phase duality,@xcite and holds for any conductors .
the binomial distribution ( [ pqresult ] ) for voltage bias has the interpretation @xcite that electrons hit the barrier with frequency @xmath144 and are transmitted independently with probability @xmath126 . for current bias
the transmission rate is fixed at @xmath145 .
deviations due to the probabilistic nature of the transmission process are compensated for by an adjustment of the voltage drop over the barrier .
if the transmission rate is too low , the voltage @xmath10 rises so that electrons hit the barrier with higher frequency . the number of transmission attempts ( `` trials '' ) in a time @xmath79 is given by @xmath146 .
the statistics of the accumulated phase @xmath138 is therefore given by the statistics of the number of trials needed for @xmath147 successful transmission events .
this stochastic process has the pascal distribution ( [ pphiresult ] ) . for the general multi - channel case a simple expression for @xmath148 can be obtained in the ballistic limit ( all @xmath149 s close to 1 ) and in the tunneling limit ( all @xmath149 s close to 0 ) . in the ballistic limit one has @xmath150 , corresponding to a poisson distribution in the discrete variable @xmath151 .
in the tunneling limit @xmath152 , corresponding to a chi - square distribution @xmath153 in the continuous variable @xmath154 .
in contrast , the charge distribution @xmath155 is poissonian both in the tunneling limit ( in the variable @xmath137 ) and in the ballistic limit ( in the variable @xmath156 ) .
for large @xmath139 and @xmath138 , when the discreteness of these variables can be ignored , we may calculate @xmath148 from @xmath157 in saddle - point approximation .
if we also calculate @xmath155 from @xmath158 in the same approximation ( valid for large @xmath93 and @xmath137 ) , we find that the two distributions have a remarkably similar form : @xmath159 , \label{pphi0q}\\ p_{q_{0}}(\phi)&=&n_{q_{0}}(\phi)\exp[\tau\sigma(2\pi\phi/\tau , q_0/\tau ) ] \label{pq0phi}.\end{aligned}\ ] ] the same exponential function @xmath160 appears in both distributions ( with @xmath161 the location of the saddle point ) .
the pre - exponential functions @xmath162 and @xmath163 are different , determined by the gaussian integration around the saddle point . since these two functions vary only algebraically , rather than exponentially , we conclude that eq .
( [ pcp ] ) holds with the remainder @xmath164 obtained by evaluating @xmath165 $ ] at @xmath166 , @xmath167 .
we consider the same circuit and turn to analysis of quantum corrections assuming @xmath168 in the relevant frequency region .
we demonstrate that the main effect of the corrections can be incorporated into the renormalization of energy dependence of the transmission eigenvalues of the quantum connector .
we study this dependence in a non - perturbative limit to obtain an unexpected result : owing to accumulation of quantum corrections , all quantum conductors behave at low energies like either a single or a double tunnel junction , which divides them into two broad classes .
it has been shown that at low energy scales the relevant part of the electron - electron interaction in mesoscopic conductors comes from their electromagnetic environment @xcite .
the resulting dynamical coulomb blockade has been thoroughly investigated for tunnel junctions @xcite .
the measure of the interaction strength is the external impedance @xmath11 at the frequency scale @xmath169 determined by either the voltage @xmath66 at the conductor or its temperature @xmath170 .
if @xmath171 the interaction is weak , otherwise coulomb effects strongly suppress electron transport .
a tunnel junction is the simplest quantum conductor with all transmission eigenvalues @xmath172 .
interaction effects for general connectors with @xmath173 are difficult to quantify for arbitrary @xmath174 . for @xmath175
, one can employ perturbation theory to first order in @xmath174 @xcite . the contributions @xcite associated the resulting interaction correction to the conductance with shot noise properties of the conductor , while the interaction correction to noise has been associated with the third cumulant of charge transfer @xcite . to sort this out
, one shall proceed in the framework of the field theory outlined where all cumulants are incorporated into functional dependence of the action on the field @xmath22 . the recent experiment @xcite addresses the correction to the conductance at arbitrary transmission .
a tunnel junction in the presence of an electromagnetic environment exhibits an anomalous power - law i - v characteristic , @xmath176 . the same power law behavior is typical for tunnel contacts between one - dimensional interacting electron systems , the so - called luttinger liquids @xcite .
it has also been found for contacts with arbitrary transmission between single - channel conductors in the limit of weak interactions @xcite . in this case , the interactions have been found to renormalize the transmission . in our model of a quantum connector , its transmission probabilities @xmath57
are energy independent in the absence of interactions .
we first analyze the quantum correction to first order in @xmath174 .
we identify an elastic and an inelastic contribution .
the elastic contribution comes with a logarithmic factor that diverges at low energies suggesting that even weak interactions can suppress electron transport at sufficiently low energies . to quantify this we sum up quantum corrections to the action in all orders in @xmath174 by a renormalization group analysis .
we show that the result is best understood as a renormalization of the transmission eigenvalues similar to that proposed in @xcite .
the renormalization brings about an energy dependence of the transmission eigenvalues according to the flow equation @xmath177.\ ] ] to calculate transport properties in the presence of interactions , one evaluates @xmath178 at the energy @xmath179 . with relation ( [ main ] )
we explore the effect of quantum corrections on the distributions of transmission probabilities for various types of mesoscopic conductors . in general
, their conductance @xmath50 and their noise properties display a complicated behavior at @xmath180 that depends on details of the conductor .
however , in the limit of very low energies @xmath181 we find only two possible scenarios .
the first one is that the conductor behaves like a _ single _ tunnel junction with @xmath182 . in the other scenario ,
the transmission distribution approaches that of a symmetric _
double _ tunnel junction .
the conductance scales then as @xmath183 .
any given conductor follows one of the two scenarios .
this divides all mesoscopic conductors into two broad classes .
we still analyze a simple circuit that consists of a mesoscopic conductor in series with an external resistor @xmath11 biased with a slow - varying voltage source @xmath184 ( fig .
1 ) but now concentrate on quantum corrections
. as we have already done , we present the generating function @xmath185 ) $ ] of the low - frequency current fluctuations in the circuit is as a path integral over the fields @xmath186(eq .
[ concatenation ] ) .
it is convenient for us to change the order of the connectors so that @xmath187 \ln{\cal z}_{ohm } \left[\phi_1,\chi_1\right]\right\ } } \end{aligned}\ ] ] where @xmath188 and @xmath189 is given by ( [ ohmic ] ) .
let us assume that @xmath190 are slow fields .
in order not to repeat the considerations of the previous section , we will simply set @xmath191 to 0 . in this case , the saddle point is trivial : @xmath192 . in physical terms , all the voltage drops on the quantum contact .
we start the renormalization by concentrating on the `` fast '' part of the fields @xmath34 and expanding the action till quadratic terms in these fields .
doing so , we neglect the time - dependence of slow fields in comparison with that of fast fields , so the corresponding part of the action reads @xmath193 where @xmath194 , and @xmath195 presents ( [ ohmic ] ) .
we require that @xmath196 at any frequency . under these conditions ,
the fluctuations of the fast fields are determined by the ohmic term while the part of the action that comes from the fluctuation and _ does _ depend on the slow fields @xmath190 is determined by the quantum conductor . indeed , taking the gaussian integral ( [ gaussian - renormalization ] ) we obtain the contribution to the action @xmath197
this is the renormalization sought .
the correction to the conductance of the quantum conductor it gives is of the order of @xmath198 .
to proceed , we need the action of the quantum conductor at fast fields , not just at slow ones as given by eq .
it is expressed in terms of keldysh green functions @xmath199 ( the `` check '' denotes @xmath200 matrices in keldysh space ) of electrons in the two reservoirs adjacent to the conductor @xcite .
it takes the form of a trace over frequency and keldysh indices , @xmath201\end{aligned}\ ] ] and depends on the set of transmission eigenvalues @xmath57 that characterizes the conductor . the fields @xmath202 enter the expression as a gauge transform of @xmath203 in one of the reservoirs , @xmath204 } { \check g}^{\rm res}(t - t ' ) { \textstyle \left [ \begin{array}{cc } e^{-i\phi^{+}_c(t ' ) } & 0 \cr 0 & e^{-i\phi^{-}_c(t ' ) } \end{array}\right ] } , \nonumber \\\end{aligned}\ ] ] @xmath205 being the drop of the phase over the quantum conductor , @xmath206 being the equilibrium keldysh green function @xmath207 & 2 f(\epsilon)-1 \end{array } \right),\ ] ] at a given equilibrium electron distribution function @xmath208 . to zeroth order in @xmath174 the fields @xmath202 do not fluctuate and are fixed to @xmath209 . substituting this into eq .
( [ eq : action ] ) we recover the slow - field action ( [ sxi0 ] ) @xmath210 \right\ } } \end{aligned}\ ] ] ( @xmath211 and @xmath212 ) . to assess the renormalization correction
, we expand the non - linear @xmath213 to second order in the fluctuating fields @xmath214 and use ( [ renorm ] ) .
the expression for the correction can be presented as @xmath215 { s}^{(1)}_{\rm el } \label{eq:1st1 } + n(\omega ) { s}^{(1)}_{\rm in}(\omega ) + [ n(\omega)+1 ] { s}^{(1)}_{\rm in}(-\omega)\right\}}. \end{aligned}\ ] ] the three terms in square brackets correspond to _ elastic _ electron transfer , inelastic transfer with absorption of energy @xmath216 from the environment , and inelastic electron transfer with emission of this energy respectively .
it is crucial to note that inelastic processes can only occur at frequencies @xmath217 and that their contribution to the integral is thus restricted to this frequency range .
in contrast , elastic contributions come primarily from frequencies exceeding the scale @xmath218 . if @xmath219 for @xmath220 , the elastic correction diverges logarithmically , its magnitude being @xmath221 .
this suggests that i. the elastic correction is more important than the inelastic one and ii . a small value of @xmath174 can be compensated for by a large logarithm , indicating the breakdown of perturbation theory . the upper cut - off energy @xmath222
is set either by the inverse @xmath75-time of the environment circuit or the thouless energy of the electrons in the mesoscopic conductor .
the concrete expression for @xmath223 reads @xmath224 where we have introduced the functions @xmath225\right\}^{-1 } \end{aligned}\ ] ] and the notation @xmath226 we do not analyze @xmath227 further and instead turn to the analysis of the elastic correction .
it is important that the explicit form of this correction can be presented as @xmath228 this suggests that the main effect of renormalization is _ to change the transmission coefficients _ @xmath57 .
it also suggests that we can go beyond perturbation theory by a renormalization group analysis that involves the @xmath57 only .
in such an analysis one concentrates at each renormalization step on the `` fast '' components of @xmath229 with frequencies in a narrow interval @xmath230 around the running cut - off frequency @xmath231 . integrating out these fields one obtains a new action for the slow fields .
subsequently one reduces @xmath231 by @xmath232 and repeats the procedure until the running cut - off approaches @xmath218 .
we find that at each step of renormalization the action indeed retains the form given by eq .
( [ eq : action ] ) and only the @xmath57 change , provided @xmath233 .
the resulting energy dependence of the @xmath57 obeys eq .
( [ main ] ) .
the approximations that we make in this renormalization procedure amount to a summation of the leading logarithms in every order of the perturbation series . in the rest of the section we analyze the consequences of eq .
( [ main ] ) for various mesoscopic conductors .
equation ( [ main ] ) can be explicitly integrated to obtain @xmath234 in terms of the `` high energy '' ( non - interacting ) transmission eigenvalues @xmath235 .
a mesoscopic conductor containing many transport channels is most conveniently characterized by the distribution @xmath236 of its transmission eigenvalues @xcite .
it follows from eq .
( [ eq : tofe ] ) that the effective transmission distribution at the energy scale @xmath231 reads @xmath237 ^ 2 } \rho_{\lambda}\left ( \frac{t}{\xi+t(1-\xi)}\right).\ ] ] we now analyze its low energy limit @xmath238 .
any given transmission eigenvalue will approach zero in this limit .
seemingly this implies that for any conductor the transmission distribution approaches that of a tunnel junction , so that all @xmath172 .
the overall conductance would be proportional to @xmath104 in accordance with ref .
@xcite .
indeed , this is one of the possible scenarios .
a remarkable exception is the case that the non - interacting @xmath239 has an inverse square - root singularity at @xmath240 .
many mesoscopic conductors display this feature , most importantly diffusive ones @xcite . in this case
, the low - energy transmission distribution approaches a limiting function @xmath241 the conductance scales like @xmath242 .
@xmath243 is known to be the transmission distribution of a _
double _ tunnel junction : two identical tunnel junctions in series @xcite .
indeed , one checks that for a double tunnel junction the form of the transmission distribution is unaffected by interactions .
this sets an alternative low - energy scenario .
we are not aware of transmission distributions that would give rise to other scenarios .
we believe that this is an important general result in the theory of quantum transport and suggest now a qualitative explanation .
the statement is that the conductance of a phase - coherent conductor at low voltage and temperature @xmath244 asymptotically obeys a power law with an exponent that generically takes two values , @xmath245 for tunneling electrons the exponent is @xmath246 .
an electron traverses the conductor in a single leap .
the second possible exponent @xmath174 has been discussed in the literature as well , in connection with _
resonant _ tunneling through a _
double _ tunnel barrier in the presence of interactions @xcite .
this resonant tunneling takes place via intermediate discrete states contained between the two tunnel barriers .
the halved exponent @xmath247 occurs in the regime of the so - called _ successive _ electron tunneling . in this case
, the electron first jumps over one of the barriers ending up in a discrete state . only in a second jump over the second barrier
the charge transfer is completed .
since it takes two jumps to transfer a charge , the electron feels only half the counter voltage due to interactions with electrons in the environmental impedance @xmath59 at each hop .
consequently , the exponent at each jump takes half the value for direct tunneling .
our results strongly suggest that this transport mechanism is not restricted to resonant tunneling systems , or , in other words , that resonant tunneling can occur in systems of a more generic nature than generally believed .
as far as transport is concerned , a mesoscopic conductor is characterized by its scattering matrix regardless of the details of its inner structure . in this approach
it is not even obvious that the conductor can accommodate discrete states .
nevertheless , the transmission distribution of this scattering matrix does depend on the internal structure of the conductor .
the inverse square root singularity of this distribution at @xmath248 for a double tunnel barrier is due to the formation of fabry - perot resonances between the two barriers .
probably similar resonances are at the origin of the same singularity for more complicated mesoscopic conductors with multiple scattering .
they are then the intermediate discrete states that give rise to the modified scaling of the conductance in presence of interactions .
one may speculate that in diffusive conductors these resonances are the so - called nearly localized states found in @xcite . from equation ( [ eq : tofe ] )
one concludes that the resonant tunneling scaling holds only if @xmath249 so that many transport channels contribute to the conductance . at sufficiently small energies
, @xmath250 becomes of the order of @xmath251 .
all transmission eigenvalues are then small and the conductance crosses over to the tunneling scaling .
one could wonder about the generality of the results obtained in the previous section and expressed by eq .
[ main ] indeed , it has been proven under rather restrictive assumptions of an ohmic connector of negligible resistance . here
, we consider a more general model of several quantum connectors coming together in a single node ( fig .
[ coulomb - island ] ) .
each connector labeled by @xmath252 is characterized by the set of the transmission eigenvalues @xmath253}$ ] . in traditional coulomb blockade situation ( @xmath254 )
this setup is called coulomb island or set transistor @xcite and is seen very different from the junction - in - the environment setup considered in the previous section .
however , we show that in the limit of large conductance @xmath53 the coulomb island is governed by very similar renormalization equations : @xmath255}}{d\ln e } = \frac{2\ , t_n^{[k]}(1- t_n^{[k]})}{\sum_{n , k } t_n^{[k]}}. \label{rgequation}\ ] ] it looks like each quantum channel sees all others as an `` environment '' characterized by the effective island conductance @xmath256}$ ] . if we use @xmath257 , eqs . [ rgequation ] and
[ main ] are identical .
the difference is that @xmath258 by itself is subject to renormalization .
this results in very different low - energy behavior .
in contrast to the considerations of the previous section , the renormalization of all transmission eigenvalues may break down at finite energy effective coulomb gap @xmath259 , @xmath260 being a numerical factor depending on the details of the initial transmission distribution , @xmath261 is the island conductance at high energies @xmath262 .
remarkably , @xmath263 coincides with the effective charging energy evaluated with instanton technique.@xcite however , the renormalization stops at the effective thouless energy @xmath264 , @xmath265 being mean level spacing in the island .
this gives rise to _ two _ distinct scenarios at low energy .
if @xmath266 , coulomb blockade does not occur with zero - bias conductance being saturated at the value @xmath267 .
alternatively , @xmath268 and @xmath263 defines the coulomb gap .
let us give the details of the model in use .
the coulomb island is characterized by two parameters : charging energy @xmath269 and the mean level spacing @xmath265 , @xmath270 .
@xcite the island is connected to @xmath271 external leads by means of @xmath272 arbitrary quantum connectors ( fig .
[ coulomb - island ] ) characterized by the set of transmission eigenvalues @xmath273}$ ] .
we assume that the island is strongly coupled to the leads , @xmath274 } \gg 1 $ ] .
our goal is to evaluate the functional @xmath27 for the whole circuit that now depends on voltages and counting fields in each terminal , @xmath275))$ ] . to evaluate this for the coulomb island
, we have extended the semiclassical approach for the fcs of the non - interacting electrons @xcite .
the node houses a dynamical phase variable @xmath276 @xcite , its time derivative , @xmath277 , presents the fluctuating electrostatic potential of the island . according to the rules of our field theory , the functional is represented in the form of a real - time path integral over the fields @xmath202 residing at two branches of the keldysh contour @xmath278}_{\rm con}\bigr(\{\hat g,\hat g_k^\chi\}\bigl ) \,-\ , i\pi\delta^{-1}{\rm tr}\{(i\partial_t - \dot\phi)\hat g\ } \big\ } \label{model } \end{aligned}\ ] ] here @xmath279 is the matrix in keldysh space , @xmath280 matrix @xmath281 presents the electron green function in the island that implicitly depends on @xmath202 .
the trace operation includes the summation over keldysh indices and the integration in time .
the contribution of each connector @xmath282}_{\rm con}$ ] has a form ( [ eq : action ] ) @xmath283 } = -\frac{1}{2}\sum_{n } { \rm tr } \ln\left[1 + \frac{1}{4}t_n^{[k ] } ( \{\hat g,\hat g_k^\chi\ } - 2)\right ] \label{sk}\ ] ] @xmath284 denoting the anticommutator of the green functions with respect to both keldysh and time indices .
the green functions in the leads @xmath285 are obtained by @xmath22-dependent gauge transformation @xcite of the equilibrium green functions in the reservoir @xmath252 , @xmath286}$ ] , @xmath287 , where @xmath286}$ ] are given by @xmath288 } = \left ( \begin{array}{cc } 1 - 2f_k & -2 f_k \\
-2(1-f_k ) & 2f_k-1 \end{array } \right ) $ ] . here
@xmath289 presents the electron distribution function in the @xmath252-th reservoir . the expression ( [ sk ] )
is valid under assumption of instantaneous electron transfer via a connector , thus corresponding to energy - independent @xmath290}_n$ ] . in order to find @xmath281 at given @xmath291 , we minimize the action with respect to all @xmath281 subject to the constrain @xmath292 .
this yields the saddle point equation for @xmath281 : @xmath293}_n [ \hat g_k^\chi,\hat g ] } { 4+t^{[\,k]}_n\left(\{\hat g_k^\chi , \hat g\ } - 2 \right ) } = i\pi\delta^{-1}[\,i\partial_t - \dot\phi , \hat g ] \label{saddle}\ ] ] where @xmath294 $ ] denotes the commutator in the keldysh - time space .
this relation expresses @xmath295)$ ] via the reservoir green functions @xmath296}$ ] .
this circuit theory relation is similar to obtained in @xcite .
it disregards the mesoscopic fluctuations , since those lead to corrections of the order of @xmath297 at all energies , whereas the interaction corrections are of the order of @xmath298 tending to diverge at small energies . if @xmath299 , eq .
( [ saddle ] ) separates in energy representation and coincides with that of ref .
@xcite .
this sets the model .
we start the analysis of the model with perturbation theory in @xmath300 around the semiclassical saddle point @xmath301 , @xmath299 .
the phase fluctuations are small , @xmath302 , so we keep only quadratic terms to the action ( [ model ] ) .
the resulting gaussian path integral over @xmath300 can be readily done .
this procedure is equivalent to the summation of all one - loop diagrams of the conventional perturbation theory , i.e. to the `` random - phase approximation '' ( rpa ) .
we restrict ourselves to the most interesting low voltage / temperature limit , @xmath303 . in this limit
, we evaluate the interaction correction to the cgf with the logarithmic accuracy .
it reads @xmath304}_n(1-t^{[\,k]}_n)\bigl(\{\hat g_k^\chi,\hat g_0 \}-2\bigr ) } { 4+t^{[\,k]}_n\bigl(\{\hat g_k^\chi , \hat g_0\ } - 2 \bigr ) } \nonumber\end{aligned}\ ] ] provided @xmath305 , where @xmath306 is the thouless energy of the island . in the opposite case , @xmath307
, the voltage / temperature should be replaced with @xmath308 .
note , that the correction ( [ s2 ] ) is contributed by only virtual inelastic processes that change the probabilities of real elastic scatterings . for simplicity
, we consider the shot - noise limit @xmath309 only .
then the magnitude of the correction shall be compared with the zero - order cgf @xmath310}_\chi\sim t_0 ev g_0 $ ] .
this implies that the perturbative rpa result ( [ s2 ] ) is applicable only if @xmath311 . at lower voltages @xmath312 logarithmically diverges .
this indicates that we should proceed with a renormalization group ( rg ) analysis .
we perform the rg analysis of the action ( [ model ] ) along the lines of the previous section decomposing @xmath202 onto the fast and slow parts . on each step of rg procedure
we eliminate the fast degrees of freedom in the energy range @xmath313 to obtain new action @xmath314 $ ] , @xmath231 being the current ultraviolet cutoff .
our key result is that the change in the action at each step of rg procedure can be presented as a change of transmission eigenvalues @xmath253}$ ] .
therefore , the rg equations can be written directly for transmission eigenvalues and take a simple form ( [ rgequation ] ) .
the equations are to be solved with initial conditions at the upper cutoff energy @xmath315 , those are given by `` bare '' transmission eigenvalues @xmath253}(e = g_0
e_c ) = t_n^{[k]}$ ] .
the rg equations resemble those for the transmission coefficient for a scatterer in the weakly interacting one - dimensional electron gas @xcite and for a single multi - channel scatterer in the electromagnetic environment @xcite .
the effective impedance @xmath59 is just replaced by inverse conductance of the island to all reservoirs , @xmath316}(e)$ ] .
the important difference is that this conductance is itself subject to renormalization .
the difference becomes most evident in the case when all contacts are tunnel junctions , @xmath253}\ll 1 $ ] . in this case
, one can sum up over @xmath317 in eqs .
( [ rgequation ] ) to obtain the rg for the conductance only : @xmath318 .
this renormalization law @xcite was recently applied to conductance of granular metals .
the eqs .
( [ rgequation ] ) could be also derived in the framework of functional rg approach to @xmath319-model of disordered metal .
@xcite .
we solve the rg eqs .
( [ rgequation ] ) in general case to obtain @xmath320}(e ) & = & t_n^{[k]}y/\left(1-t_n^{[k]}(1-y)\right ) ,
\label{solution}\\ \ln({g_0 e_c}/e ) & = & -\frac{1}{2}\sum_{n,\,k } \ln(1-t_n^{[k]}(1-y))\end{aligned}\ ] ] the first equation gives the renormalized transmission eigenvalues at a given value @xmath231 of the upper cutoff in terms of variable @xmath321 , @xmath322 .
the second equation implicitly expresses @xmath321 . .
arrows show the energy scale @xmath323 .
pane ( a ) : tunnel connectors , @xmath261 changes from @xmath324 ( upper curve ) to @xmath325 ( lowermost curve ) with the step @xmath326 .
pane ( b ) : diffusive connectors , @xmath261 changes from @xmath327 to @xmath328 with the step @xmath329 .
the conductance either hits @xmath9 manifesting the coulomb gap or saturates at finite value . ,
width=432 ] we note that the energy dependence of transmission coefficients induced by interaction is very weak provided @xmath249 : if energy is changed by a factor of two , the conductance is changed by @xmath330 . to use the equations for evaluation of fcs at given voltages @xmath331}$ ] of the leads
, one takes @xmath253}(e)$ ] at upper cutoff @xmath332})$ ] , and further disregards their energy dependence
. then one can follow the lines of ref .
@xcite : it is convenient to introduce the function @xmath282}(x ) = -\sum_n\ln[1+\frac{1}{2}t_n^{[k]}(x-1)]$ ] to incorporate all required information about transmission eigenvalues .
the renormalization of @xmath282}$ ] in terms of @xmath333 is especially simple : @xmath282}(x , y ) = s^{[k]}((x + 1)y-1)-s^{[k]}(2y-1)$ ] . from this
one readily finds the conductance of each scatterer , @xmath334}(y ) = 2 g_q \partial s^{[k]}/\partial x(1,y)$ ] , as well as the renormalized transmission distribution @xmath335}(t , y ) = ( { 2}/{\pi}){\rm i m } \{\partial s^{[k]}/\partial x\,(1 - 2/t - i0,y)\}$ ] .
the rg equations ( 1 ) have a fixed point at @xmath290}_{n}=0,y=0 $ ] that occur at finite energy @xmath336}_n)^{1/2}\ ] ] this indicates the breakdown of rg and formation of coulomb blockade with the exponentially small gap @xmath263 .
the same energy scale was obtained from equilibrium instanton calculation of ref .
for a field theory , one generally expects different physics and different energy scales for instantons and perturbative rg .
the fact that these scales are the same shows a hidden symmetry of the model which is yet to understand .
alternative low - energy behavior is realized if the current cut - off reaches @xmath337 .
3 ) the log renormalization of the transmission eigenvalues stops at this point and their values saturate .
we thus predict a sharp crossover between the two alternative scenarios , that occur at value of @xmath338 corresponding to @xmath339 .
this value equals @xmath340 , where @xmath341})$ ] , and depends on transmission distribution of all connectors . if all connectors are tunnel junctions , @xmath342 . for diffusive connectors , @xmath343 and the energy dependence of the total conductivity
is given by @xmath344 , @xmath345 .
3 ) 10 l. s. levitov and g. b. lesovik , jetp lett . * 58 * , 230 ( 1993 ) ; cond - mat/9401004 ; l. s. levitov , h. lee , and g. b. lesovik , j. math phys . * 37 * , 4845 ( 1996 ) . yu .
v. nazarov ( editor ) , nato asi series ii , * 97 * , _ quantum noise in mesoscopic physics _ , ( dordrecht , kluwer , 2003 ) . b. reulet , j. senzier , and d. e. prober , phys .
* 91 * , 196601 ( 2003 ) .
j. bylander , t. duty , p. delsing , nature * 434 * , 361 ( 2005 ) ; t. fujisawa , t. hayashi , r. tomita , y. hirayama , s. gustavsson , r. leturcq , t. ihn , et al .
science * 312 * , 1634 ( 2006 ) ; phys .
b * 75 * 075314 ( 2007 ) .
a. o. gogolin , a. komnik , phys .
b * 73 * , 195301 ( 2006 ) .
v. ambegaokar , u. eckern , g. schn , phys .
lett . * 48 * , 1745 ( 1982 ) .
h. grabert , p. schramm , g. l. ingold , phys . rep . * 168 * , 115 ( 1988 ) .
d. v. averin and k. k. likharev , in _ mesoscopic phenomena in solids _ , edited by b. l. altshuler , p. a. lee , and r. a. webb ( elsevier , amsterdam , 1991 ) .
ingold and yu .
v. nazarov , in _ single charge tunneling _ , edited by h. grabert and m. h. devoret , nato asi series b294 lenum , new york , 1992 ) .
m. h. devoret , d. esteve , h. grabert , g. l. ingold , h. pothier , c. urbina phys .
* 64 * , 1824 ( 1990 ) . c. l. kane and m. p. a. fisher , phys .
rev . lett . * 79 * 5086 ( 1992 ) .
r. egger , h. grabert , phys .
* 75 * , 3505 ( 1995 ) .
j. zinn - justin , _ quantum field theory and critical phenomena _
( clarendon press , oxford , 1993 ) .
m. kindermann , yu . v. nazarov and c. w. j. beenakker , phys .
90 * , 246805 ( 2003 ) .
m. kindermann and yu .
v. nazarov , phys .
lett . * 91 * , 136802 ( 2003 ) .
d. a. bagrets and yu . v. nazarov , phys .
lett * 94 * , 056801 ( 2005 ) . yu .
v. nazarov , ann .
( leipzig ) * 8 * , 507 ( 1999 ) ; yu .
v. nazarov and m. kindermann , in ref .
g. schn and a. d. zaikin , phys .
rep . * 198 * , 237 ( 1990 ) .
we record the result for the fourth cumulant , obtained by expansion of eq .
( [ fxi ] ) to order @xmath346 : @xmath347 , where we have abbreviated @xmath348 .
the pascal distribution @xmath349 is also called the `` binomial waiting - time distribution '' , since it gives the probability of the number @xmath350 of independent trials ( with success probability @xmath351 ) that one has to wait until the @xmath272-th success .
it is related to the negative - binomial distribution @xmath352 by the displacement @xmath353 . yu .
v. nazarov , zh .
95 , 975 ( 1989 ) ; yu .
v. nazarov , sol .
. comm . * 75 * , 669 ( 1990 ) .
d. s. golubev , a. d. zaikin , phys .
86 , 4887 ( 2001 ) .
a. levy yeyati _
et al . _ ,
lett . * 87 * , 046802 ( 2001 ) .
galaktionov , d.s .
golubev , and a. d. zaikin , phys .
b * 72 * , 205417 ( 2005 ) .
_ , in _ electronic correlations : from meso - to nano - physics _ , ed . by t. martin , g. montambaux , j. trn thanh vn , edp sciences , 2001 , p. 17
k. a. matveev , d. yue , and l. i. glazman , phys .
71 , 3351 ( 1993 ) .
v. nazarov , in : _ quantum dynamics of submicron structures _ , edited by h. a. cerdeira , b. kramer , g. schn , kluwer , 1995 , p. 687 ; cond - mat/9410011 .
j. a. melsen and c. w. j. beenakker , physica b * 203 * , 219 ( 1994 ) .
b. a. muzykantskii and d. e. khmelnitskii , phys .
b * 51 * , 5480 ( 1995 ) .
v. nazarov , phys .
lett . * 82 * , 1245 ( 1999 ) .
strictly speaking , a purely electric field theory corresponds to @xmath354 .
a finite @xmath355 enables electrons rather than electricity to escape the node and may provide coherent electron transfer through two connectors .
yu . v. nazarov and d. a. bagrets , phys .
* 88 * , 196801 ( 2002 ) .
w. belzig and yu .
v. nazarov , phys .
lett . * 87 * , j. m. kosterlitz , phys .
lett * 37 * , 1577 ( 1976 ) ; | we review the relations between the full counting statistics and the field theory of electric circuits .
we demonstrate that for large conductances the counting statistics is determined by non - trivial saddle - point of the field .
coulomb effects in this limit are presented as quantum corrections that can stongly renormalize the action at low energies . |
the mergers of neutron star , black hole , and neutron star - black hole binaries provide the most promising sources of gravitational waves for detection by laser interferometers now being constructed , such as ligo and virgo .
a variety of additional factors also motivate the study of binary neutron star systems : for example , such systems are known to exist with several having been detected in our galaxy in the form of binary radio pulsars ( eg . @xcite and references therein ) , and coalescing neutron stars may be a source of gamma - ray bursts at cosmological distances @xcite .
these facts have motivated numerous theoretical studies of the inspiral and coalescence of compact binaries . until recently ,
such calculations have involved either post - newtonian ( pn ) point mass approximations or purely newtonian hydrodynamic treatments .
analytic analyses of the gravitational radiation waveforms have been performed in the pn approximation with the binary components , whether black holes or neutron stars , being treated as point masses @xcite . these investigations are therefore appropriate in the early inspiral phase when the stars are still well separated and tidal effects can be safely neglected .
semi - analytic @xcite and numerical @xcite studies of the coalescence waveform from binary neutron stars also have been performed using newtonian hydrodynamics and the quadrupole radiation approximation .
these studies treat newtonian effects arising from the finite size of the interacting fluid stars and focus on their tidal distortion , disruption , and eventual merger .
numerical simulations of coalescing black hole and neutron star binaries are also underway in full general relativity ( eg .
@xcite ) . complicated by the challenges of treating both matter and strong gravitational fields in ( 3 + 1 ) dimensional spacetime , these simulations are essential for obtaining definitive , quantitative results for the highly relativistic interactions which generate the late inspiral and final coalescence waveforms .
there is an important distinction between binaries in newtonian theory and those in general relativity : in newtonian theory a binary can exist in true dynamic equilibrium , but in general relativity true equilibrium can not exist because of the generation of gravitational waves . for binaries with sufficiently large orbital separation ( @xmath4 ) ,
the orbital decay time due to emission of gravitational waves is much longer than the orbital timescale , so that the inspiral is quasi - static through a sequence of quasi - equilibrium configurations .
the inspiral becomes dynamic and quasi - equilibrium is destroyed with the onset of relativistic orbital instabilities ( @xmath5 few @xmath6 ) .
dynamic behavior may also commence when the purely newtonian tidal instability is reached near @xmath5 few @xmath7 ( eg . , @xcite ) .
for these reasons , both newtonian and relativistic effects are important . in the relativistic simulations of wilson , mathews , and marronetti @xcite ,
the time - dependent equations of relativistic hydrodynamics are solved in a simplified , strong - field background , and the system is relaxed to find quasi - static equilibrium configurations of corotating neutron star binaries for a specific equation of state .
these authors claim that neutron stars just below the maximum ( isolated ) mass become unstable and collapse to black holes when placed in close binary orbits , in contrast to the newtonian result that the tidal field tends to stabilize a star ( see fig .
15 of ref .
@xcite ; also ref .
such collapse could significantly affect the inspiral and the associated gravitational waveform , and therefore deserves careful attention .
recently , baumgarte et al .
@xcite have constructed quasi - static binary equilibrium models in full general relativity . however , their numerical models , based on a polytropic equation of state , are restricted to be corotating .
this work complements the numerical analysis of cook @xcite , who analyzed binary black holes with zero spins .
a fully general relativistic calculation of a binary neutron star configuration is a computationally intensive problem .
by contrast , an ellipsoidal figure of equilibrium ( efe ) treatment , while only approximate , can find an equilibrium configuration in less than a second on a typical workstation .
this speed affords a quick means of gaining a qualitative understanding of the stellar models and equilibrium sequence .
these sequences can help provide a better understanding of the general relativistic solutions which are now being generated ( eg . ref .
@xcite ) . in addition , while general relativistic codes are currently limited to corotation , or near corotation , an efe treatment allows straightforward construction and comparison of both corotating and irrotational sequences .
the second order variation of the functional can be used to identify instabilities , e.g. the innermost stable circular orbit ( isco ) or the collapse of a massive neutron star in a binary to a black hole .
furthermore , both secular and dynamical instabilities can be identified and distinguished . taniguchi and nakamura @xcite
use the efe approach to model black hole - neutron star binaries in the roche approximation , using a pseudo - potential to model the strong - field orbital motion .
lai and wiseman @xcite insert efe terms in the hybrid - p@xmath8n dynamical equations of motion , which are then integrated to identify the isco .
the efe approach has also recently been extended to first pn order in treatments of corotating binaries @xcite . in this paper
, we adopt the efe energy variational method , extended to first pn order .
our energy functional includes terms for both the internal hydrodynamic structure as well as the external orbital motion .
we add the leading pn correction to the internal and self - gravitational energy of the stars .
we adopt the hybrid expressions of @xcite , to pn order , to give the orbital terms ; these expressions are exact in the test mass limit and correct to pn order for arbitrary masses .
the energy functional is varied to find quasi - equilibrium configurations of both corotating and irrotational binary sequences , parameterized by orbital separation ( or orbital angular velocity ) .
in [ approximations ] we outline the basic approximations of our efe scheme . in
iii we give the various energy , angular momentum , and circulation terms which comprise our functionals .
the method of solution is discussed in
iv , and results are presented in
v. finally , we discuss the implications of our results and directions for future work in vi . in the appendices ,
we derive the pn energy contributions for an isolated spherical star obeying a polytropic equation of state .
in this section , we describe the fundamental approximations and assumptions behind our energy variational method , which we apply to construct general darwin - riemann equilibrium models .
further details and applications of our method can be found in @xcite . in steady state ,
an isolated , self - gravitating binary system is characterized by conserved global quantities such as the rest ( baryonic ) mass @xmath9 and @xmath10 of each star and the total angular momentum @xmath11 .
the total energy of the system can be written as a functional of the rest mass density and velocity distributions @xmath12 and @xmath13 , respectively , even if the system is not in equilibrium . in principle , an equilibrium configuration can be determined by extremizing this energy functional with respect to all variations of @xmath12 and @xmath14 that leave the conserved quantities unchanged .
the fundamental approximation of our method is to replace the infinite number of degrees of freedom contained in @xmath12 and @xmath14 by a finite number of parameters @xmath15 , so that the total energy is a function only of these parameters : @xmath16 an equilibrium configuration is then determined by extremizing the energy according to @xmath17 where the partial derivatives are taken holding the conserved quantities @xmath18 constant .
an expression like equation ( [ lrs4.2.1 ] ) can be written down only with the help of simplifying assumptions .
we adopt the compressible ellipsoidal approximation to model the neutron stars , i.e. we assume they are triaxial ellipsoids supported in hydrostatic equilibrium by a polytropic equation of state @xcite
. surfaces of constant density are assumed to be concentric self - similar ellipsoids .
these approximations become exact in the limit that the fluid is incompressible and the gravitational potential is strictly newtonian and truncated beyond the tidal term @xcite . for compressible configurations ,
this simplification is equivalent to using an ellipsoidal trial function in an energy functional and then extremizing the functional to find the equilibrium solution .
comparisons with numerical simulations show that the ellipsoidal model is reliable , particularly for stars obeying stiff equations of state @xcite .
the velocity field of the fluid is taken to be that of a riemann - s ellipsoid , in which all internal velocities are linear functions of the coordinates ( see eg .
@xcite , 5 , for details ) .
the vorticity vector is assumed to be everywhere parallel to the orbital rotation axis . since the viscosity in neutron stars is thought to be negligible
, the two stars will conserve circulation as they inspiral @xcite , so that slowly spinning neutron stars at @xmath19 will maintain a small , constant circulation .
however , current general relativistic quasi - equilibrium codes are forced to consider corotating , or nearly corotating , sequences rather than sequences of constant circulation .
we therefore concentrate on two interesting limiting cases : irrotational sequences ( for non - synchronized systems with zero circulation ) and corotating sequences ( corresponding to synchronized binary systems in uniform rotation ) .
consider a binary system composed of two stars of rest mass @xmath9 and @xmath10 in circular orbit . throughout this paper
unprimed quantities refer to the star of rest mass @xmath9 while primed quantities refer to the star of mass @xmath10 . following ref .
@xcite , we denote the mass ratio as @xmath20 . the density and pressure are related by @xmath21 note that our treatment allows the two stars to have both distinct polytropic indices ( @xmath22 ) and polytropic constants ( @xmath23 ) , although in this paper we concentrate on the case of two identical stars .
the binary separation is denoted by @xmath24 , and the principal axes of the two ellipsoids by @xmath25 , @xmath26 , @xmath27 , and @xmath28 , @xmath29 , @xmath30 .
the orientation is such that @xmath25 and @xmath28 are measured along the binary axis , @xmath26 and @xmath29 in the direction of the orbital motion , and @xmath27 and @xmath30 along the rotation axis . in place of the three principal axes @xmath31 , it is often convenient to introduce as independent variables the central rest mass density @xmath32 , and two oblateness parameters defined as @xmath33 similarly we can introduce @xmath34 , @xmath35 , and @xmath36 in place of the three @xmath37 .
thus the seven independent variables which parameterize our models are @xmath38 . as is standard in efe treatments , we assume that the density profile @xmath39 , where @xmath40 is the mass interior to an isodensity surface , is identical to that of a spherical polytrope with the same @xmath41 and @xmath0 , but with radius @xmath42 [ see eq .
( [ ov - based ] ) ] .
in particular , to pn order , a star with semi - major axes @xmath31 has the same central density as a spherical star ( not necessarily in equilibrium ) with radius @xmath7 in schwarzschild coordinates , @xmath43 \right\ } , \label{rhoofr}\ ] ] where we have used equations ( [ surface ] ) and ( [ final.aofm ] ) .
here @xmath44 and @xmath45 are the usual lane - emden variables for a polytrope ( see , e.g. , ref .
@xcite ) with @xmath46 , and where the quantities @xmath47 and @xmath48 are defined in appendix [ pn.expand.ov ] and depend only on the polytropic index @xmath0 ( for @xmath2 we have @xmath49 and @xmath50 , while for @xmath51 we have @xmath52 and @xmath53 ) . in the newtonian limit ( @xmath54 ) , equation ( [ rhoofr ] ) reduces to the expression used in the newtonian treatment of @xcite .
although equation ( [ rhoofr ] ) originates from a pn expansion , our code treats this relation as exact .
we now review the energy , angular momentum , and circulation terms in the newtonian efe approximation , which has been covered in detail in @xcite .
when similar expressions can be written for the two stars , we give only the one corresponding to @xmath9 , the other being obtained simply by replacing unprimed with primed quantities .
the usual newtonian orbital contributions to the energy and angular momentum are @xmath55 where @xmath56 is the reduced mass and @xmath57 is the orbital angular velocity . the gravitational tidal ( quadrupole ) interaction energy @xmath58 between the two stars is ( see appendix b of ref .
@xcite ) @xmath59 here we have defined @xmath60 where @xmath61 with analogous relations for @xmath62 and @xmath63 .
we will refer to key relations in ref .
@xcite by their equation number preceded by an `` i '' ; for example , equations ( [ lrs1.b8 ] ) and ( [ lrs1.b9 ] ) are also equations ( i.b8 ) and ( i.b9 ) , respectively .
the dimensionless structure coefficient @xmath64 appearing in equation ( [ lrs1.b8 ] ) is defined by [ cf .
( i.3.17 ) ] @xmath65 where @xmath66 so that @xmath67 for @xmath68 .
values of @xmath64 and @xmath69 for various @xmath0 are tabulated in table i of @xcite .
@xmath79 and @xmath80 are tabulated in ref .
@xcite .
we permit the star to be differentially rotating with uniform vorticity @xmath82 as measured in the corotating frame of the binary : @xmath83 here the velocity field @xmath84 in the corotating frame is given by @xmath85 with @xmath86 the quantity @xmath87 is the angular velocity of the internal fluid motions in the corotating frame . here
@xmath88 is along the binary axis , directed from @xmath9 to @xmath10 , @xmath89 is in the direction of the orbital velocity , and @xmath90 is perpendicular to the orbital plane .
the velocity field in the inertial frame is given by @xmath91 and the vorticity in this frame is @xmath92 where @xmath93 corotation ( synchronization ) corresponds to @xmath94 , while an irrotational velocity field is obtained when @xmath95 .
note that the geometric outer shapes of the two stars always rotate at the orbital angular velocity @xmath96 in the inertial frame , regardless of the internal velocity field .
the newtonian spin kinetic energy @xmath97 ( i.e. , the kinetic energy in internal fluid motions ) in the inertial frame can be obtained from equations ( [ lrs4.2.12])([lrs4.2.14 ] ) . assuming a lane - emden trial density function , one finds [ cf .
( i.5.6 ) ] @xmath98 where @xmath99 is the moment of inertia .
similarly , the newtonian spin angular momentum @xmath100 can be written [ cf .
( i.5.5 ) ] @xmath101 as given by @xcite , the purely newtonian total energy of the system , not necessary in equilibrium , is therefore @xmath102 similarly , the total angular momentum to newtonian order is @xmath103 another important quantity , conserved in the absence of viscosity , is the fluid circulation @xmath104 along the equator of the star ( see 5.1 of ref .
we define @xmath105 the quantity @xmath106 has the dimensions of angular momentum but is proportional to the conserved circulation @xmath104 . for convenience , we typically refer to @xmath106 itself as the circulation . in this paper , we augment the newtonian terms of [ newtonianefe ] with pn contributions ( a ) to the orbital energy and angular momentum , ( b ) to the internal and self - gravitational energy of a star , and ( c ) to the coupling energy and angular momentum between the internal structure and the orbital motion .
we replace of the newtonian contributions to orbital energy and angular momentum [ eq .
( [ newtonianorbitalej ] ) ] with the hybrid - pn expressions of @xcite ( with @xmath107 , valid for circular orbits ) : @xmath108 where the schwarzschild , or test - mass , contributions are ( with @xmath109 for these two equations only ) @xmath110^{-1/2}-1\right\ } , \\ % j_t & = & \mu vr\left\{\left[{(r+m)^5 \over r^4(r - m)}\right]^{1/2}\left[1-v^2{(r+m)^3 \over r^2(r - m)}\right]^{-1/2}\right\},\end{aligned}\ ] ] and the pn corrections unaccounted for by the schwarzschild terms are @xmath111 here @xmath112 is the orbital velocity , the total rest mass @xmath113 , and the mass parameter @xmath114 . in the notation of ref .
@xcite , @xmath40 also includes contributions from the self energy of the stars and is not simply the total rest mass ; however , we account for these contributions below [ see eqs .
( [ eoi ] ) and ( [ joi ] ) ] as a coupling between the internal structure and the orbital motion . with our definitions ,
the quantities @xmath40 and @xmath115 remain strictly constant along the sequence .
the radius @xmath24 is the so - called harmonic or de donder coordinate .
when expanded in @xmath116 , the above expressions for @xmath117 and @xmath118 give the newtonian terms [ eq .
( [ newtonianorbitalej ] ) ] to lowest order and the correct orbital pn terms to first order .
the pn contributions to the internal and self - gravitational energy of a spherical polytrope are derived in appendix [ derive.l1.and.l2 ] : @xmath119 where the coefficients @xmath120 and @xmath121 are defined by equations ( [ final.l1 ] ) and ( [ final.l2 ] ) , respectively , and depend only on the polytropic index @xmath0 .
values for @xmath120 and @xmath121 are listed in table [ structureconstants ] .
we also include the analogous expressions for the pn energy contributions @xmath122 and @xmath123 in the second star . for isolated stars ,
variation of the energy function @xmath124 does a good job of approximating the exact equilibrium sequence obtained by integrating the oppenheimer - volkoff ( ov ) equation , especially for soft equations of state ( see figure [ isolated.spherical ] ) .
we neglect pn corrections due to asphericity ; for small perturbations from spherical , equations ( [ upn ] ) and ( [ wpn ] ) are valid to lowest order .
to account completely for all pn corrections to the internal and self - gravitational energy , equations ( [ upn ] ) and ( [ wpn ] ) would need to be multiplied by a form factor analogous to @xmath125 in equation ( [ lrs4.2.7 ] ) .
we also neglect p@xmath8n and higher order corrections to the internal and self - gravitational energy of the neutron stars .
the final pn correction terms we include involve the coupling between the orbital motion and the internal structure .
effacement of internal structure ( see ref .
@xcite and references therein ) requires that we substitute the effective masses @xmath126 and @xmath127 _ in place of _ the rest masses @xmath9 and @xmath10 in the newtonian equation ( [ newtonianorbitalej ] ) if we wish to maintain pn order accuracy . here the self energy of the mass @xmath9 neutron star is @xmath128 .
we have chosen to exclude the spin kinetic energy @xmath97 from the self energy @xmath129 in order to ensure that our binary configurations satisfy the identity @xmath130 exactly ( see [ find.eq.soln ] ) .
we speculate that additional higher order terms which we are neglecting would restore this identity when we include @xmath97 .
furthermore , @xmath97 is typically small , especially for irrotational sequences , so that we are neglecting a term which is smaller than our order of accuracy . expanding to pn order , the resulting corrections to the newtonian point - mass energy and angular momentum are @xmath131 with analogous expressions @xmath132 and @xmath133 due to the corrections arising from the self energy of the neutron star of mass @xmath10 .
the subscript oi serves as a reminder that these terms couple the orbital motion to the internal structure of the neutron stars .
although including @xmath134 and @xmath135 as part of @xmath129 in equations ( [ eoi ] ) and ( [ joi ] ) amounts to keeping terms which are formally of p@xmath8n order , doing so keeps our energy functional accurate even for stars at @xmath19 but with @xmath136 ; excluding @xmath134 and @xmath135 from @xmath129 would lead to an energy functional which yields manifestly incorrect results for isolated stars ( in particular , the stability of a star would depend on the velocity of the observer ) .
we neglect pn effects in the internal motion which would appear as corrections to the spin kinetic energy @xmath97 , the spin angular momentum @xmath100 , and the circulation @xmath104 . for corotating sequences , these corrections are on the order of @xmath137 times the pn orbital terms , so that we are justified in neglecting pn internal motion effects all the way down to roughly the isco ( @xmath5 few @xmath7 ) . for irrotational sequences ,
pn internal motion is especially small , since the fluid motion in the inertial frame is always predominately set by the orbital motion .
we are also neglecting all pn tidal terms ( but see [ future.work ] ) .
the total energy and angular momentum functionals used in our treatment are therefore @xmath138 and @xmath139 since we are approximating the internal fluid motion as newtonian , we continue to use equation ( [ lrs4.2.20 ] ) for the circulation .
in order to find an equilibrium solution , we must extremize the energy function @xmath140 while holding all the quantities @xmath9 , @xmath10 , @xmath141 , @xmath142 , and @xmath11 constant .
we are therefore identifying those configurations for which the energy is unchanged to first order by any perturbation which conserves rest mass , circulation , and total angular momentum .
an equilibrium sequence is constructed by repeating this extremization at various values of the orbital separation @xmath24 . in the case of corotation , the stars are members of the riemann - s family with vorticity @xmath143 , and although the perturbations considered conserve circulation , the circulation does indeed vary along the equilibrium sequence itself @xcite .
unfortunately , the various energy terms in [ ejcterms ] do not depend explicitly on @xmath144 , and @xmath142 . instead
, the energy has been written in terms of the variables @xmath145 from the set @xmath146 , and these variables are not independent .
the lagrange multiplier method allows us straightforwardly to minimize the total energy @xmath147 subject to the constraints that the total angular momentum @xmath148 and the circulations @xmath149 and @xmath150 are constant .
the equations which determine an equilibrium configuration are @xmath151 where @xmath152 , @xmath153 , and @xmath154 are the lagrange multipliers . here the partial differentiation with respect to @xmath145 is done with all the remaining variables @xmath155 ( @xmath156 ) being held constant .
although equations ( [ lagrange ] ) are accurate only to the order of our expansion of @xmath140 , @xmath11 , @xmath106 , and @xmath157 , our numerical code treats them as exact .
we expanded and restructured a version of the code developed in ref .
@xcite to solve these coupled equations numerically . for a given value of @xmath24
, we use a newton - raphson scheme to solve the 10 equations ( [ lagrange ] ) and find the variables @xmath158 , and @xmath154 . in practice ,
our code uses cramer s rule on the @xmath159 equations of ( [ lagrange ] ) to find the lagrange multipliers : @xmath160 where the determinant @xmath161 using equations ( [ lagrangemultipliers ] ) , we substitute for the lagrange multipliers in the remaining 7 equations of ( [ lagrange ] ) to find the orbital velocity as well as the shape and structure of the neutron stars .
the lagrange multipliers can even be determined analytically , since expressions for the various derivatives appearing in equations ( [ lagrangemultipliers ] ) and ( [ denom ] ) can be derived from equations ( [ efunctional ] ) , ( [ jfunctional ] ) , and ( [ lrs4.2.20 ] ) .
the algebra is simplified somewhat by @xmath162 and @xmath163 .
we find @xmath164 , @xmath165 , and @xmath166 as mathematical identities , for any @xmath167 and not just along equilibrium sequences .
this provides a convenient check of a large portion of the code , since the lagrange variables are determined by evaluating the determinants in equations ( [ lagrangemultipliers ] ) and ( [ denom ] ) directly at every iteration in the newton - raphson scheme .
furthermore , equation ( [ lagrange ] ) then implies @xmath168 so that @xmath130 along both corotating ( @xmath169 ) and constant circulation ( @xmath170 ) sequences .
consequently , the minima of @xmath11 and @xmath140 always coincide precisely on our equilibrium sequences .
we concentrate on binaries composed of identical stars , each with polytropic index @xmath2 or @xmath171 , and with a compactness parameter @xmath172 or @xmath173 in isolation .
these values of @xmath0 and @xmath1 cover an interesting region of parameter space relevant to neutron stars . for the av14+uvii equation of state of ref .
@xcite , stars in the mass range @xmath174 have compactness parameters @xmath175 with corresponding effective polytropic indices @xmath176 ( see table 3 of ref .
@xcite ) . for an isolated neutron star with @xmath177 , the nine equations of state in ref .
@xcite which allow this mass all yield a compactness parameter in the range @xmath178 .
some physical quantities describing our neutron stars in isolation are listed in table [ isolated.spherical.models ] .
these quantities can be converted to physical units once a value of the polytropic constant @xmath41 is chosen .
for instance , for @xmath2 and @xmath179 @xmath180g@xmath181sec@xmath8 , the maximum gravitational mass @xmath182 , so that @xmath183 , and @xmath184 and @xmath185 , and @xmath186 km for @xmath187 , and @xmath173 , respectively . for @xmath3 and @xmath188 @xmath189g@xmath190sec@xmath8 , the maximum mass @xmath191 so that @xmath192 , and @xmath193 and @xmath194 , and @xmath195 km for @xmath187 , and @xmath173 , respectively .
any of our models can be scaled to a particular mass ( say @xmath196 ) by an appropriate choice of the polytropic constant @xmath41 .
integration of the ov equation reveals that the true maximum masses of @xmath2 and @xmath171 sequences occur at log @xmath197 and @xmath198 , respectively , at which point @xmath199 and @xmath200 .
we note , however , that our @xmath51 isolated stars with @xmath201 are indeed stable in our pn - accurate model , since our energy functional [ eq . ( [ isolated.energy.functional ] ) ] gives @xmath202 at the maximum isolated mass ( see table [ isolated.spherical.models ] ) .
when expressing our results , the orbital angular velocity , angular momentum , energy , and circulation are listed in terms of the dimensionless ratios @xmath203 where @xmath204 . here
@xmath205 is defined to be the radius that an isolated polytrope , with the same @xmath9 , @xmath41 , and @xmath0 , would have in newtonian theory ( ie . without the @xmath134 and @xmath135 corrections ) : @xmath206^{n/(3-n ) } \left({m_o\over4\pi\xi_1 ^ 2|\theta'_1|}\right)^{(1-n)/(3-n)}.\label{ro}\ ] ] the radius @xmath205 is a convenient unit of length because it is constant along equilibrium sequences .
the orbital separation @xmath24 is measured in harmonic coordinates while the semi - major axes @xmath31 are measured in schwarzschild coordinates , which is permissible because we have two distinct small parameters in our pn expansion , @xmath116 and @xmath1 .
we determine when the neutron stars touch by the pn accurate relation @xmath207 , where @xmath208 is the total mass energy of the system .
since the radial coordinate @xmath24 is not a gauge invariant quantity , we choose to make our plots as a function of the orbital angular frequency @xmath209 ( @xmath210 the gravitational wave frequency ) .
the non - dimensionalized orbital velocity @xmath211 can be converted to a frequency @xmath72 in hz via the relation @xmath212 where @xmath213 . for given values of @xmath0 and @xmath1 ,
the corresponding values of @xmath214 and @xmath215 can be determined from table [ isolated.spherical.models ] .
for example , an @xmath3 model with @xmath216 corresponds to @xmath217 and @xmath218 . while @xmath9 and @xmath10 are constant along our equilibrium sequences , the ratio @xmath1 does indeed vary ; however , whenever we specify a particular value of @xmath1 we are referring to that value at infinity ( ie . for isolated neutron stars ) . figures [ evsw2irrotational ] and [ evsw2irrotational.n=1 ] present irrotational ( @xmath219 ) sequences in the equal mass case with @xmath2 and @xmath3 , respectively , showing the total energy @xmath140 , relative to its value @xmath220 at infinity , for @xmath2210.125 , 0.2 and 0.25 ( thick solid curves ) .
the energy @xmath220 can be calculated from equation ( [ isolated.energy.functional ] ) for each star .
also shown in this figure are the point - mass newtonian equilibrium energy ( thin dashed curve ) , as well as the results of a newtonian efe treatment as in @xcite ( thick dashed line ) .
figure [ eosnew ] shows the dependence of @xmath1 on the orbital frequency @xmath222 at the isco for @xmath2 and @xmath3 irrotational sequences . for each @xmath0 ,
the curves are terminated at the value @xmath1 corresponding to the maximum mass . table [ irrotationalwill ] presents physical quantities along selected irrotational sequences for @xmath2 and @xmath171 , and various values of @xmath1 . for each sequence ,
the value of @xmath223 marked by a single asterisk ( * ) identifies the equilibrium configuration for which @xmath140 ( and @xmath11 ) is a minimum , ie .
as expected , the size of the neutron star @xmath224 decreases as @xmath1 increases .
our @xmath225 sequences agree with the newtonian results of ref .
@xcite ( see their table 1 )
. tables [ irrotationale ] and [ irrotationalj ] show the sizes of the individual energy and angular momentum terms appearing in equations ( [ efunctional ] ) and ( [ jfunctional ] ) for the specific sequence @xmath216 , @xmath3 .
note that some of the relativistic corrections are included in @xmath117 and @xmath118 : for instance , at the isco , @xmath117 is @xmath226 times smaller than what the newtonian expression [ eq.([newtonianorbitalej ] ) ] would give .
the importance of the pn terms increases as @xmath1 increases , and the stars behave more like point masses . fully relativistic calculations of binary neutron star systems ( eg .
@xcite ) provide a nice test of our pn models .
however , such calculations are currently limited to corotating , or nearly corotating , stars .
figures [ evsw2corotating ] and [ evsw2corotating.n=1 ] present corotating ( @xmath169 ) sequences with @xmath2 and @xmath3 , respectively , showing the total energy @xmath140 , relative to its value @xmath220 at infinity , for @xmath172 and 0.25 ( thick solid curves ) . tables [ corotatingwill ] presents physical quantities along selected corotating sequences . for each sequence ,
the value of @xmath223 marked by a single asterisk ( * ) identifies the equilibrium configuration for which @xmath140 ( and @xmath11 ) is a minimum , the innermost _ secularly _ stable circular orbit , while the value marked by a double asterisk ( * * ) identifies the innermost _ dynamically _ stable circular orbit .
the dynamical isco occurs when @xmath227 at constant circulation along an equilibrium sequence ( see ref .
we numerically evaluated @xmath228 simply by finding equilibrium configurations at two slightly different orbital separations and then finite differencing .
as in our irrotational sequences , the overall size of the neutron star @xmath224 does indeed decrease as @xmath1 is increased . for corotation , the spin of each star causes a rotational bulge such that @xmath229 .
the circulation @xmath230 grows in magnitude as the radial separation of the neutron stars decreases .
table [ compare.to.lw96 ] presents the total angular momentum @xmath11 , the total mass energy @xmath208 , @xmath231 , @xmath232 and the orbital frequency @xmath72 , all at the isco of selected irrotational sequences . for
any given @xmath0 , the orbital frequency @xmath72 can be converted into units of hz if either the gravitational mass @xmath6 or the equation of state ( ie . the polytropic constant @xmath41 ) is specified .
it is therefore useful to express the isco frequency both as @xmath233 [ hz ] ( when @xmath6 is known ) and @xmath234 [ hz ] ( when the equation of state is known ) , where @xmath235 and @xmath236 .
here @xmath237 is the maximum gravitational mass in isolation for a given equation of state .
the conversion factor between @xmath233 and @xmath234 depends on the quantity @xmath238 , which can be determined from the last two columns of table [ isolated.spherical.models ] .
table [ corotating.isco ] lists the orbital frequency @xmath239 at the innermost secularly stable orbit as well as the frequency @xmath240 at the innermost dynamically stable orbit . for the six combinations of @xmath0 and @xmath1 listed in this table ,
the frequency @xmath240 differs by less than 4% from the @xmath222 values of the corresponding irrotational sequences . that the spin has only a small effect on the orbital dynamics
is consistent with the small values of @xmath241 ( typically on the order of 0.01 near the isco ) listed in table [ corotatingwill ] .
an important issue is whether the maximum mass of each neutron star increases or decreases as the binary components are brought together .
figure [ variousf ] shows the rest mass @xmath9 , normalized to the maximum rest mass @xmath242 at infinite separation , as a function of the central density parameter @xmath243 for an @xmath3 corotating binary system and at three different values of the orbital angular frequency : @xmath244 ( solid curve ) corresponds to spherical isolated stars , @xmath245 hz ( long dashed curve ) corresponds to the innermost dynamically stable orbit for a star with @xmath246 ( @xmath216 at infinity ) , and @xmath247 hz ( short dashed curve ) corresponds to the innermost dynamically stable orbit for a star with @xmath248 ( @xmath202 at infinity ) .
as they inspiral , stars which begin on the stable side ( @xmath249 ) of the @xmath244 curve move along horizontal lines ( since @xmath9 is conserved ) to the left on this plot .
figure [ mvsomega4 ] shows the maximum equilibrium rest mass @xmath250 as a function of the orbital angular frequency for corotating ( dashed curves ) and irrotational ( solid curves ) sequences .
for example , the maxima of the three curves in fig .
[ variousf ] gives three data points for the @xmath51 corotating curve in figure [ mvsomega4 ] .
the curves are terminated at the values of @xmath72 for which stars of rest mass @xmath251 acquire a dynamically unstable orbit .
we see that the maximum equilibrium mass increases as the orbit decays , regardless of the sequence considered . we conclude that if the maximum equilibrium mass does ever decrease in a binary system , it can do so only through pn tidal terms , or higher order pn terms , which we are neglecting in this paper ( see [ future.work ] ) .
note that we have not determined from our equilibrium analysis , which of our equilibria are stable .
the exception is the @xmath244 ( infinite separation ) sequence , for which the turning point on the @xmath6 versus @xmath32 plot marks the onset of radial instability to collapse .
we have extended the work of previous studies in a number of ways .
we have improved upon purely newtonian efe treatments by including pn effects , and improved upon point - mass pn treatments by including both newtonian and pn finite - size effects .
the main message of appendix b is that equation ( [ isolated.energy.functional ] ) is the pn accurate energy functional of an isolated non - rotating star of arbitrary polytropic index @xmath0 .
nowhere in the derivation of equation ( [ isolated.energy.functional ] ) is equilibrium assumed , but in equilibrium the trial density function we implement agrees to pn order with the density profile obtained by integrating the ov equation .
the pn coefficients @xmath120 and @xmath121 appearing in equation ( [ isolated.energy.functional ] ) are listed for various polytropic indices @xmath0 in table [ structureconstants ] .
figure [ isolated.spherical ] demonstrates that using equation ( [ isolated.energy.functional ] ) yields reliable results ( compare the long dashed curve to the ov results given by the solid curve ) , especially for soft equations of state ( which have a maximum mass at small @xmath32 ) .
one particularly nice feature of our energy functional is that it correctly exhibits a maximum mass even for stiff equations of state .
various physical quantities for our isolated spherical models are listed in table [ isolated.spherical.models ] .
figures [ evsw2irrotational ] and [ evsw2irrotational.n=1 ] , present the equilibrium energy @xmath140 as a function of the orbital frequency @xmath72 along irrotational sequences with various values of the polytropic index @xmath0 and compactness parameters @xmath1 . note that for a given @xmath6 and @xmath1 , @xmath3 polytropes are more centrally concentrated than @xmath2 polytropes , so that the former behave more like point masses ( ie .
newtonian effects which destabilize the orbit are less important for @xmath3 ) and have a slightly larger @xmath233 .
this same tendency can be seen by comparing the @xmath2 and @xmath51 corotating sequences presented in figures [ evsw2corotating ] and [ evsw2corotating.n=1 ] , respectively . in these plots
the minimum of the equilibrium energy @xmath140 marks the onset of an orbital secular instability .
the dynamical instability sets in at higher orbital frequency , as can be seen from table [ corotatingwill ] or [ corotating.isco ] .
tables [ irrotationalwill ] and [ corotatingwill ] present various quantities along selected sequences for comparisons with future studies .
the tables [ irrotationale ] and [ irrotationalj ] present the relatives sizes of the terms which make up our energy and angular momentum functionals ; these tables allow us to explicitly examine the relative importance of the various effects which determine the orbital dynamics and the interior structure . comparing the magnitudes of @xmath58 and @xmath252 ,
it is clear that both tidal and relativistic terms play an important role .
tables [ compare.to.lw96 ] and [ corotating.isco ] summarize some of the important quantities at the isco of our irrotational and corotating sequences , respectively . note that for fixed @xmath0 , @xmath233 always increases with @xmath1 , since more compact stars behave more like point masses .
although real neutron star sequences are probably close to irrotational , fully relativistic calculations of binary systems are currently limited to corotating sequences . by looking for turning points in the equilibrium energy , it is straightforward to determine the secular isco frequency @xmath253 in such corotating sequences
. it is often assumed that the actual isco frequency @xmath222 of an irrotational sequence does not differ drastically from the frequency @xmath253 determined from corotating calculations .
our results allow us to quantify this difference . for @xmath2
we find that the irrotational @xmath222 is approximately 17% larger than the corresponding @xmath253 , while for @xmath3 the difference is approximately 20% ( compare entries in tables [ compare.to.lw96 ] and [ corotating.isco ] ) . for any polytropic index @xmath0 , this difference depends only very weakly on @xmath1 .
if multiple binary neutron star coalescence events are observed by instruments such as ligo , then tables like table [ compare.to.lw96 ] , or figures like figure [ eosnew ] , will give information regarding the equation of state .
for instance , consider a simple world in which binary components always had identical masses , although allow for varying neutron star mass from one binary to the next .
suppose further that all neutron stars obeyed the same ( but unknown ) polytropic equation of state .
a single coalescence observation would yield a value of @xmath233 , since the gravitational mass @xmath6 is encoded in the early inspiral waveform and since the isco frequency @xmath222 can be determined by the onset of the orbital plunge .
however , the equation of state and compactness parameter @xmath1 could not be uniquely determined from a single observation , since we do not yet know which curve ( if either ) in figure [ eosnew ] applies . for each value of @xmath0
being considered , the corresponding maximum gravitational mass @xmath237 can be immediately determined from the observed value of @xmath222 and from the @xmath254 column in table [ compare.to.lw96 ] .
an additional observation of a coalescence event with a different @xmath222 would yield another value of @xmath237 for each @xmath0 .
the polytropic index which consistently gives the same @xmath237 regardless of the observed @xmath222 must then represent the true equation of state .
tables similar to table [ compare.to.lw96 ] will therefore be helpful in testing realistic equations of state once coalescence observations are available .
there is only one equation of state for neutron stars , even if knowledge of it is still uncertain .
for a given equation of state , one might ask how @xmath222 depends on the mass of the binary components . for polytropes , fixing @xmath0 and @xmath41
also fixes the maximum gravitational mass @xmath237 in isolation . from the @xmath234 column of table [ compare.to.lw96 ]
, we see that for @xmath2 the general trend of @xmath222 is for it to decrease as the stellar mass @xmath6 ( or @xmath1 ) increases , while for @xmath3 the frequency @xmath222 increases with @xmath6 .
we therefore conclude that the relation between @xmath222 and @xmath6 depends strongly on the equation of state .
the reason for this dependence can be traced to two competing effects .
namely , as we consider more compact ( larger @xmath6 ) stars , pn orbital destabilization effects become stronger , while newtonian destabilization effects become weaker . for the cases listed in table [ compare.to.lw96 ] , the total angular momentum @xmath11 is @xmath255 to @xmath256 times the total energy squared @xmath257 ( in gravitational units , @xmath109 ) at the isco . in those cases with @xmath258 , angular momentum would need to be either radiated away or carried away by ejected matter before a black hole can form ( see ref .
@xcite for a discussion ) .
we find @xmath258 for all @xmath2 sequences with @xmath259 , while for @xmath3 this criterion is satisfied when @xmath260 .
figure [ variousf ] presents the equilibrium rest mass @xmath9 as function of the central density for various orbital frequencies @xmath72 for an @xmath3 corotating sequence .
we see that , for the terms we have included in our energy functional , the maximum mass in a close binary is slightly larger than in isolation .
figure [ mvsomega4 ] summarizes our maximum mass results for both irrotational and corotating sequences , and for @xmath2 and @xmath171 .
the irrotational sequences maintain a lower maximum mass than their corresponding corotating sequences , because corotating stars ( a ) have more rotational support and ( b ) are more ellipsoidal so that the newtonian tidal field has more of a stabilizing effect .
the relative increases in @xmath261 is more gradual for @xmath2 than @xmath51 , because @xmath2 stars near the maximum mass are much more centrally condensed than @xmath51 stars near their maximum mass : from table [ isolated.spherical.models ] we see that both the central density parameter @xmath262 and the compactness parameter @xmath1 are much larger for the maximum mass @xmath2 star ( log @xmath263 , @xmath264 ) than the @xmath3 star ( log @xmath265 , @xmath266 ) .
consequently , the @xmath2 stars near the maximum mass have smaller moments of inertia and behave more like point masses , so that stabilizing hydrodynamic effects play a less significant role .
our results suggest that the maximum equilibrium mass increases as the orbit decays , regardless of the sequence considered . however , if the maximum equilibrium mass does decrease in a binary system , it must be a result of the pn tidal terms , or of higher order pn corrections , that we are neglecting ( see [ future.work ] ) .
recently , taniguchi and nakamura @xcite have applied the ellipsoidal approximation to a neutron star - black hole binary .
they also compute irrotational and corotating sequences , but only for @xmath68 polytropes . in place of the hybrid - pn orbital energy contribution @xmath267 , they use a generalized pseudo - newtonian potential constructed to fit the iscos of the hybrid p@xmath8n treatment of kidder , will , and wiseman @xcite .
we improve upon the work of taniguchi and nakamura by including the relativistic corrections @xmath134 and @xmath135 to the internal energy of the neutron star , and by calculating sequences of fixed @xmath268 , instead of fixed @xmath269 which varies as two neutron stars inspiral . despite these differences our results do agree qualitatively : for a given equation of state more compact stars have larger isco frequencies , and for a fixed compactness parameter the isco along irrotational sequences is larger than the secular isco along corotating sequences .
taniguchi and shibata @xcite have recently presented pn efe analyses of binary neutron star systems which is rather similar to our own .
the main differences are that ( 1 ) they derive equilibrium conditions from a tensor virial method , whereas we use an energy variational method , ( 2 ) they include all terms , up to quadrupole order , which are formally pn order , while we have neglected pn tidal terms and pn corrections to the internal fluid motion , ( 3 ) they consider only corotating sequences , while we also treat the more realistic irrotational sequences , and ( 4 ) they implement a simple lane - emden trial density function , as opposed to our pn - accurate density function [ see eq .
[ ov - based ] ] .
our treatment does somewhat better than taniguchi and shibata s at matching the numerical , pn - accurate results of shibata @xcite .
for instance , for @xmath2 and @xmath270 , shibata finds secular isco values of @xmath271 , @xmath272 , and @xmath211 equal to @xmath273 , @xmath274 , and @xmath275 , respectively ; compare this to our values @xmath276 ,
@xmath277 , and @xmath278 , and to shibata and taniguchi s values @xmath279 , @xmath280 , and @xmath281 .
we find a similar level of agreement with shibata along other sequences : for @xmath51 and @xmath282 , shibata finds secular isco values of @xmath271 , @xmath272 , and @xmath211 equal to @xmath283 , @xmath284 , and @xmath285 , while our treatment gives @xmath286 , @xmath287 , and @xmath288 . in order to examine the isco in irrotational sequences , lai and wiseman @xcite
have combined the newtonian ellipsoidal equations of motion @xcite with the hybrid p@xmath8n point mass equations of motion of kidder , will , and wiseman .
this treatment does not include the pn coupling between the orbital motion and internal structure ( @xmath252 and @xmath289 in this paper ) , nor does it include pn corrections to the internal and self - gravitational energy ( @xmath134 and @xmath135 in this paper ) .
nevertheless , our results are in excellent agreement with those of lai and wiseman ( see the comparison in table [ compare.to.lw96 ] ) : for the three values of @xmath290 for which a direct comparison is possible , we find agreement to better than 4% in @xmath222 and to better than about 2% in the ratios @xmath231 and @xmath232 .
the comparisons of @xmath231 and of @xmath232 should be viewed with some caution because of the freedom to measure distances with various coordinate systems in general relativity . in the newtonian limit
our results agree precisely with the newtonian results of lai and wiseman [ see their equations ( 8) and ( 9 ) ] , as they must since both treatments are extensions of the same newtonian ellipsoidal approximation . in the point mass limit ,
the results of lai and wiseman are accurate to p@xmath8n order , whereas our results are accurate only to first pn order .
we have not implemented the point mass p@xmath8n orbital energy and angular momentum corrections in our energy functional , because doing so causes @xmath291 to hold to p@xmath8n order but not exactly along our sequences , and because our functional would be neglecting other p@xmath8n terms which are larger .
however , we have explicitly tested the importance of the point mass p@xmath8n corrections by adding them to our energy and angular momentum functions .
we find that for @xmath201 , the isco frequency for our @xmath2 and @xmath3 irrotational sequences is decreased by less than 4% , while for @xmath292 and @xmath293 this frequency is increased by about 2% or less .
baumgarte et al .
@xcite have recently calculated in full general relativity quasi - equilibrium corotating sequences of binary neutron stars obeying a polytropic equation of state .
their computations are the most reliable to date for the neutron star structure , maximum equilibrium mass , and determination of the secular isco . over the range of orbital separations they consider ,
they find the maximum equilibrium mass to increase slightly as the orbit decays , in agreement with the results presented in figures [ variousf ] and [ mvsomega4 ] .
this agreement is reassuring , since these relativistic calculations essentially represent the `` true '' solution .
figure [ compare.with.tom ] compares our binding energy results with those of ref .
@xcite for an @xmath3 corotating sequence .
the maximum equilibrium rest mass results presented in figures [ variousf ] and [ mvsomega4 ] are also in qualitative agreement with the results of lai @xcite : the maximum mass of a neutron star seems to increase as the orbit decays .
the main differences between our energy functional and that of lai are that ( a ) lai s adopted equation of state is based on a degenerate , non - relativistic fermi gas of nucleons ( @xmath294 polytrope ) , while we have concentrated on @xmath2 and @xmath3 polytropes , ( b ) lai has included a correction @xmath295 as in ref .
@xcite accounting for the tendency of neutrons to become mildly relativistic , and ( c ) we use equation ( [ isolated.energy.functional ] ) for the self energy of a polytrope , whereas lai uses equation ( [ st.eeq ] ) .
our maximum mass results are also completely consistent with the recent analytic treatments of brady and hughes @xcite and of wiseman @xcite , which show that there can be no decrease in the maximum equilibrium rest mass at first pn order , presuming tidal effects can be neglected .
however , we can not be sure that pn tidal effects would not change our results ( see [ future.work ] ) .
nevertheless , our results are consistent with the relativistic integrations of baumgarte et al .
( 1997 ) , which automatically include all pn tidal terms .
this agreement provides strong support of our analysis , at least in the case of corotating binaries .
recently , wilson , mathews , and marronetti @xcite have presented numerical simulations which suggest that otherwise stable neutron stars may individually collapse to black holes when brought into a close binary orbit .
our results suggest that the maximum equilibrium mass increases as the orbit decays . in this section , however , we discuss two important caveats which leave open the possibility that the stars could collapse prior to the orbital plunge : ( 1 ) the pn tidal terms neglected in this paper may affect the maximum equilibrium mass , and ( 2 ) even if the maximum equilibrium mass does increase in binaries , some of the allowable equilibria may be _ unstable _ to collapse . the largest uncertainty in our results may come from neglected pn tidal terms .
chandrasekhar @xcite and chandrasekhar and nutku @xcite have derived the pn corrections to the energy and angular momentum for arbitrary fluid configurations , which could be used to determine the pn corrections for our binary star system .
the tidal components of these corrections are absent from our present analysis .
for example , consider the pn correction from ref .
@xcite s equation ( 68 ) which is proportional to @xmath296 , where @xmath297 is the usual newtonian gravitational potential . inside a star
, the potential @xmath297 can be decomposed into an internal contribution @xmath298 due to that star , and an external contribution @xmath299 due to the orbital companion .
the integral @xmath300 then contributes to the self - gravitational pn energy @xmath135 , while the cross term @xmath301 gives a coupling as in @xmath252 .
the integral @xmath302 to lowest order helps give the point mass pn orbital energy like in ref .
@xcite , and for finite - size stars this integral can be evaluated in a multipole expansion with the higher order terms accounting for pn tidal effects .
a careful investigation of pn tidal effects seems worthwhile .
we also note that we have not tested our equilibrium configurations for stability against collapse to black holes : even if the maximum equilibrium mass does increase as the orbit decays , this does not necessarily mean the maximum stable mass follows the same trend .
the efe energy variational approach provides a straightforward means of testing the stability of equilibrium models , which could be exploited in future studies .
in particular , the second order variation of the energy functional can be used to identify any instability to collapse .
we would like to thank phil arras for a careful checking of the appendices , as well as thomas baumgarte , anna flanagan , larry kidder , dong lai , saul teukolsky , and alan wiseman for useful discussions .
this work was supported by nsf grant ast 91 - 19475 and nasa grant nag5 - 2809 at cornell , and by nsf grant ast 96 - 18524 and nasa grant nag5 - 3420 at illinois .
far is supported by an alfred p. sloan foundation fellowship .
in this appendix we expand the ov equation ( see , eg .
@xcite ) to pn order and thereby derive the first pn corrections to the newtonian lane - emden functions for an isolated spherical polytrope in equilibrium . our results are useful for generating trial density functions in [ ov.based.density.function ] accurate to pn order .
let @xmath303 be the rest mass density of a spherically symmetric distribution of matter at the schwarzschild radius @xmath24 .
let @xmath304 be the enclosed total mass - energy , and define the quantity @xmath45 and the non - dimensional functions @xmath305 and @xmath306 by @xmath307 where @xmath308 is the central rest mass density , so that @xmath309 . in the newtonian limit , the functions @xmath305 and @xmath306 reduce to the usual lane - emden functions @xmath44 and @xmath310 . here
@xmath311 is a scale factor , with @xmath7 being the radius of the star in schwarzschild coordinates and @xmath312 being defined by the condition @xmath313 . for a polytropic equation of state @xmath314
the ov equations yield a hydrostatic equilibrium solution if ( see , eg . ,
@xcite ) @xmath315 where we have defined @xmath316 evaluating the last equation of ( [ non.dimensional ] ) at the surface , we see that for an equilibrium configuration the scale factor @xmath317 satisfies @xmath318 where @xmath6 is the total mass - energy and @xmath319 . in the newtonian limit ( @xmath54 ) , @xmath320 , @xmath321 and @xmath322 , where @xmath44 is the usual lane - emden function .
if we define the functions @xmath323 and @xmath324 as the pn components in the expansions @xmath325 then equations ( [ ov1 ] ) and ( [ ov2 ] ) become @xmath326 in deriving equations ( [ dalpha ] ) and ( [ dbeta ] ) we have used the lane - emden equation to cancel all terms of newtonian order .
we note that the functions @xmath323 and @xmath324 depend only on the polytropic index @xmath0 and can be numerically integrated simultaneously with the lane - emden equation .
the numerical integrations can be started slightly away from @xmath327 with the help of the approximations , for @xmath328 : @xmath329 the surface occurs when @xmath330 , to pn order , where we have taylor expanded and used @xmath46 to obtain the last equality .
therefore , @xmath331 where @xmath332 .
similarly , taylor expanding equation ( [ beta ] ) about the surface at @xmath333 gives @xmath334 where @xmath319 and @xmath335 .
in the special case @xmath3 it is well known that the solution to the lane - emden equation is @xmath336 .
equations ( [ dalpha ] ) and ( [ dbeta ] ) can also be solved analytically when @xmath3 : @xmath337 these equations provide a convenient check of our numerical integrations in appendix b.
we now derive the pn self energy terms of a spherically symmetric fluid which is instantaneously at rest .
our approach generalizes the method of ref .
let us introduce the enclosed rest mass @xmath338 and the proper radial coordinate @xmath339 , which are related to the enclosed total mass - energy @xmath40 and the schwarzschild radius @xmath24 by @xmath340 where @xmath341 is the internal energy density .
to pn order , the energy of such a system , excluding rest mass energy , is [ see equations ( 6.9.3 ) , ( 6.9.8 ) , ( 6.9.10 ) and ( 6.9.18 ) of ref .
@xcite ] where @xmath9 is the total rest mass , and @xmath343 since we are working only to pn order , the quantities @xmath40 and @xmath24 can be replaced with @xmath338 and @xmath339 in @xmath344 to @xmath345 , if desired .
note that equations ( [ exact.energy ] ) and ( [ energy.functional ] ) do _ not _ require that the fluid be in equilibrium : the density profile @xmath346 and internal energy profile @xmath341 need to be spherically symmetric , but are otherwise arbitrary .
let us now restrict our attention to fluids obeying a polytropic equation of state @xmath347 where @xmath41 is a constant , but allow the density profile @xmath346 to remain arbitrary for the moment .
we wish to find the contribution to the energy beyond that given by the usual newtonian expressions for the internal energy @xmath348 and the self - gravitational energy @xmath349 [ see equations ( [ lrs4.2.5 ] ) and ( [ lrs4.2.6 ] ) ] , ie .
@xmath350 where it is understood that all terms of order p@xmath8n and higher are dropped .
inspection of equations ( [ energy.functional ] ) , ( [ eos ] ) and ( [ define.epn ] ) reveals that @xmath351 is composed of two types of terms : those which depend on @xmath41 and those which do not ; we write @xmath352 where we have defined @xmath353 the quantity @xmath354 is the energy of a spherical symmetric configuration with an _ arbitrary _ density profile , to pn order .
if we restrict our attention to a set of well - chosen density functions , then the integrals in equations ( [ upn.app ] ) and ( [ wpn.app ] ) can be evaluated . in particular ,
if we consider a family of density profiles of the form @xmath355 [ or equivalently @xmath356 , where @xmath357 [ or @xmath358 is arbitrary , then @xmath359 here the dimensionless coefficients @xmath120 and @xmath121 are determined by evaluating the integrals in equations ( [ upn.app ] ) and ( [ wpn.app ] ) for the chosen family of density profiles . as a concrete example
, consider density profiles which are related to the non - dimensional lane - emden function @xmath44 by @xmath360 where @xmath361 is the proper radius of the star and the central density @xmath32 does _
not _ necessarily have the value which would allow for equilibrium at newtonian order .
the superscript `` le '' appearing in equation ( [ lane - emden ] ) is a reminder that we are limiting ourselves to a single family of trial density profiles based upon lane - emden functions .
since density profiles of this form differ at least at pn order from the ov solution which minimizes the energy given by equation ( [ exact.energy ] ) , there is _ no _ value of @xmath32 for which the fluid is in equilibrium to pn order . using equation ( [ lane - emden ] ) in equations ( [ upn.app ] ) and ( [ wpn.app ] )
we obtain simply @xmath362 from here it is straightforward to write @xmath363 and @xmath364 as integrals over lane - emden functions , and these coefficients therefore depend only on the polytropic index @xmath0 . equations ( [ l1le ] ) and ( [ l2le ] ) define the coefficients used to produce the short dashed curves in figure [ isolated.spherical ] . in the following subsection
we go through a considerable amount of work to derive @xmath120 and @xmath121 for more accurate trial density functions .
however , the coefficients given by equations ( [ l1le ] ) and ( [ l2le ] ) already represent a significant improvement over that used previously in the literature ( compare the short - dashed and dot - dashed curves in figure [ isolated.spherical ] ) .
the reason for this improvement is that we have kept both of the pn terms [ eqs .
( [ l1 ] ) and ( [ l2 ] ) ] which result naturally from the subtraction in equation ( [ define.epn ] ) .
previous analyses have used equilibrium relations to eliminate @xmath41 in equation ( [ l1 ] ) and forced @xmath134 to scale like @xmath135 , resulting in an energy functional which is valid only along equilibrium sequences and therefore should not be used in an energy variational treatment [ see the discussion surrounding eq .
( [ st.eeq ] ) below ] .
nowhere in the derivation of equations ( [ l1le ] ) and ( [ l2le ] ) have we assumed the star is in equilibrium .
the disadvantage of being restricted to lane - emden trial density functions is that the energy minimum of equation ( [ energy.functional ] ) then differs at pn order from the actual energy minimum .
we know that fluid configurations which obey the ov equation are precisely those necessary to minimize the energy given by equation ( [ exact.energy ] ) .
this suggests that we construct a family of trial density functions based upon the solution to the ov equation , namely @xmath365 where @xmath7 is the stellar radius in schwarzschild coordinates , the central density @xmath32 is _ not _ necessarily the value required by the ov equation for equilibrium [ ie .
( [ equilibrium.rho ] ) and ( [ scale.factor ] ) need not be satisfied ] , and the function @xmath323 is obtained by solving equations ( [ dalpha ] ) and ( [ dbeta ] ) .
the expansion parameter @xmath366 is defined by equation ( [ qc ] ) with @xmath311 , ie .
@xmath367 an energy variational method based upon this family of trial density functions can identify the minimum of equation ( [ energy.functional ] ) to pn order .
there are an infinite number of other families which satisfy this criterion [ for instance , we could generate one other such family by replacing @xmath32 with its equilibrium value in eq .
( [ qc.ov-based ] ) ] .
however , the trial density functions defined by equations ( [ ov - based ] ) and ( [ qc.ov-based ] ) are independent of the polytropic constant @xmath41 and hence make it easy to track which terms do and do not depend on @xmath41 .
we note that the last equation of ( [ non.dimensional ] ) does not apply here , although by construction it is satisfied to pn order when @xmath32 has its equilibrium value .
[ to derive the enclosed total mass - energy even for non - equilibrium configurations , use @xmath368 with @xmath369 . ]
our plan is now to evaluate @xmath120 and @xmath121 as defined by equations ( [ l1 ] ) and ( [ l2 ] ) , using the family of trial density functions given in equation ( [ ov - based ] ) .
we begin with the first integral appearing in equation ( [ upn.app ] ) and work to pn order : @xmath370 where we have used equation ( [ differentials ] ) and expanded .
we now focus our attention on the integral @xmath371 note that since we are working only to pn order , it is justified to end all the integrals in this appendix at @xmath372 [ where @xmath372 is defined by @xmath46 ] .
for instance , since @xmath374 in the range @xmath375 , and since @xmath376 , @xmath377 therefore equation ( [ udm ] ) becomes @xmath378 in order to write our final answer in the form of equation ( [ deltaeself ] ) , we need to derive a relation between @xmath379 and the total rest mass @xmath9 . from our knowledge of newtonian polytropes we expect ( @xmath380 , and we can obtain a relation accurate to pn order by evaluating @xmath9 : @xmath381 making use of @xmath382 in equation ( [ aofm ] ) gives @xmath383 where @xmath335 . we evaluated the integral in equation ( [ almost.final.aofm ] ) with the help of equations ( [ dbeta ] ) and ( [ justified ] ) . equation ( [ final.aofm ] ) is our desired relation between @xmath384 and @xmath9 .
using equations ( [ final.aofm ] ) and ( [ final.udm ] ) in equation ( [ udmprime ] ) gives @xmath385 we note that the lowest order term is the usual internal energy contribution for a pure polytrope , @xmath386 , as expected ; the first order corrections will help give us our expression for the coefficient @xmath120 .
using equation ( [ final.udmprime ] ) and the definition of @xmath387 in equation ( [ upn.app ] ) gives @xmath388 or , after using equations ( [ qc.ov-based ] ) and ( [ final.aofm ] ) to lowest order and simplifying , @xmath389 the relation @xmath390 can be used to reduce the double integral in equation ( [ almost.final.l1 ] ) to a single integral : @xmath391 if desired , the identity [ which can be proved from equations ( [ dalpha ] ) and ( [ dbeta ] ) and appropriate integrations by parts ] @xmath392\ ] ] could be used to further manipulate equation ( [ final.l1 ] ) when @xmath393 , while for @xmath3 this integral can be evaluated using equation ( [ alpha.n=1 ] ) .
however , we find it more convenient to work with equation ( [ final.l1 ] ) directly . equation ( [ final.l1 ] ) is our final expression for the coefficient @xmath120 , and was obtained by evaluating equation ( [ upn.app ] ) with the ov - based trial density functions defined in equation ( [ ov - based ] ) .
we could proceed to calculate @xmath121 from equation ( [ wpn.app ] ) in a similar manner .
however , it is more straightforward to take the following approach . rearranging equations ( [ define.epn ] ) and ( [ deltaeself ] ) yields the pn accurate energy functional for an isolated spherical star , not necessarily in equilibrium , obeying a polytropic equation of state : @xmath394 along an _ equilibrium _ sequence we have @xmath395 , which gives , to lowest order , @xmath396 indeed , equation ( [ hydro.equil ] ) is the same relation obtained by writing the newtonian equation of hydrostatic equilibrium in terms of lane - emden functions and simplifying . to pn order , the energy along an equilibrium sequence therefore satisfies @xmath397 since here we are focusing on the equilibrium sequence , the first variational derivative of @xmath140 with respect to the density functional must vanish ; therefore , any variation in the density profile of pn order causes a variation in the energy of order p@xmath8n , so that it is sufficient to use lane - emden profiles when evaluating the newtonian order integrals in @xmath351 .
we therefore find that along an equilibrium sequence @xmath398 .
the newtonian equation of hydrostatic equilibrium , equation ( [ hydro.equil ] ) , can then be used to evaluate @xmath344 through @xmath345 in this case .
these integrals have been carried out by ref .
@xcite ( see their 6.9 ) , so that @xmath399 where ( see eq . [
6.9.31 ] of @xcite ) @xmath400^{7/3}}\left ( -{5 + 2n - n^2 \over n+1}2\int_0^{\xi_1}\xi^3\theta'\theta^{n+1}\,d\xi + { 3 \over 2}(n-1)\int_0^{\xi_1}\xi^4\theta'^2\theta^n\,d\xi \right ) \\ & = & { ( 4 \pi)^{2/3 } \over [ \xi_1 ^ 2|\theta_1'|]^{7/3}}\left ( { n-1\over 2(5-n)}\xi_1 ^ 4|\theta_1'|^3 -{(n+2)(n+5 ) \over 6(n+1)}\int_0^{\xi_1}\xi^3\theta'\theta^{n+1}\,d\xi \right ) .
\label{k4}\end{aligned}\ ] ] by integrating the ov equation exactly and extracting coefficients from the resulting equilibrium sequence energy @xmath401 , we have numerically verified that equation ( [ st.eeq ] ) holds to pn order .
we emphasize that equation ( [ eeq ] ) or ( [ st.eeq ] ) , in contrast to equation ( [ isolated.energy.functional ] ) , should not be used as energy functionals in an energy variational treatment , since they are valid only for equilibrium configurations ; this has not been noticed in previous analyses . comparing equations ( [ eeq ] ) and ( [ st.eeq ] ) we see immediately @xmath402^{7/3}}\left ( { n-10\over 6}\int_0^{\xi_1}\xi^3\theta'\theta^{n+1}\,d\xi \right .
\nonumber \\ & & \left .
+ { n\over n+1}\int_0^{\xi_1}\alpha\theta^n\xi^2\,d\xi + { n \beta_1 \xi_1 ^ 3 |\theta_1'|\over ( n+1)(5-n ) } + { 3n^2 - 10n-5\over2(5-n)^2}\xi_1 ^ 4|\theta_1'|^3\right ) .
\label{final.l2}\end{aligned}\ ] ] equations ( [ final.l1 ] ) and ( [ final.l2 ] ) are our final expressions for the coefficients @xmath120 and @xmath121 , and were used to generate the long dashed curves in figure [ isolated.spherical ] .
the newtonian curves in this figure , as well as the @xmath3 curves which use equation ( [ st.eeq ] ) , have a maximum in the gravitational mass @xmath6 which is not accompanied by a maximum in the rest mass @xmath9 ; this is because along these sequences @xmath403 vanishes due to @xmath404 .
the remaining sequences have maxima in @xmath6 and @xmath9 at the same @xmath32 , where @xmath405 , and these turning points mark the onset of radial instability .
the derivation of the coefficients @xmath120 and @xmath121 is formally invalid if @xmath68 , since then the first relation in equation ( [ le.relations ] ) is accurate only to newtonian order .
however , for @xmath68 one can analytically derive the coefficients exactly to be @xmath406 and @xmath407 , and the numerically obtained values given by equations ( [ final.l1 ] ) and ( [ final.l2 ] ) do indeed approach these limiting values as @xmath0 approaches zero . see table [ structureconstants ] for numerical values of these coefficients for various @xmath0 .
we note that the coefficient @xmath120 is fairly well approximated simply by @xmath408 .
also note that for @xmath409 , @xmath120 is roughly 10 or more times larger in magnitude than @xmath121 , so that we expect @xmath134 to be significantly larger than @xmath135 in magnitude for all but the stiffest equations of state .
r. epstein and r. v. wagoner , * 197 * , 717 ( 1975 ) ; c. w. lincoln and c. m. will , phys .
d. * 42 * , 1123 ( 1990 ) ; l. blanchet and t. damour , phys .
d * 46 * , 4304 ( 1992 ) ; c. cutler _
* 70 * , 2984 ( 1993 ) ; c. cutler and e. e. flanagan , phys . rev .
d * 49 * , 2658 ( 1994 ) ; t. a. apostolatos , c. cutler , g. j. sussman , and k. s. thorne , phys .
d * 49 * , 6274 ( 1994 ) .
k. oohara and t. nakamura , prog .
. phys . * 88 * , 307 ( 1992 ) ; m. shibata , t. nakamura , and k. oohara , _ ibid . _ * 88 * , 1079 ( 1992 ) ; rasio , f. a. , and shapiro , s. l. , apj , 401 , 226 ( 1992 ) ; rasio , f. a. , and shapiro , s. l. , apj , 432 , 242 ( 1994 ) ; m. b. davies , w. benz , t. piran , and f. k. thielemann , * 431 * , 742 ( 1994 ) ; x. zughe , j. m. centrella , and s. l. w. mcmillan , phys .
d * 50 * , 6247 ( 1994 ) ; m. ruffert , h. t. janka , and g. schfer , astrophys .
sci . * 231 * , 423 ( 1995 ) . for earlier work ,
see articles in _ sources of gravitational radiation _ , edited by l. smarr ( cambridge univ .
press , cambridge , 1979 ) , and in _ frontiers in numerical relativity _
, edited by c. r. evans , l. s. finn , and d. w. hobill ( cambridge univ .
press , cambridge 1989 ) . for more recent progress , see r. a. matzner _
et al . _ ,
science * 270 * , 941 ( 1995 ) .
cddccd @xmath0 & log @xmath262 & @xmath1 & @xmath214 & @xmath215 & @xmath410 + 0.5 & -1.0693 & 0.125 & 0.1234 & 0.9276 & 0.0580 + 0.5 & -1.0076 & 0.1380 & 0.1359 & 0.9198 & 0.0655 + 0.5 & -0.7687 & 0.1941 & 0.1883 & 0.8853 & 0.0984 + 0.5 & -0.7450 & 0.2 & 0.1936 & 0.8816 & 0.1019 + 0.5 & -0.5541 & 0.2481 & 0.2348 & 0.8517 & 0.1297 + 0.5 & -0.5464 & 0.25 & 0.2364 & 0.8505 & 0.1308 + 0.5 & 0.1212 * & 0.3917 & 0.3058 & 0.7875 & 0.1804 + + 1.0 & -1.9749 & 0.02 & 0.0196 & 0.9900 & 0.0246 + 1.0 & -1.0180 & 0.125 & 0.1074 & 0.9387 & 0.1346 + 1.0 & -0.8907 & 0.15 & 0.1232 & 0.9275 & 0.1545 + 1.0 & -0.6532 & 0.2 & 0.1467 & 0.9083 & 0.1838 + 1.0 & -0.4171 & 0.25 & 0.1563 & 0.8984 & 0.1960 + 1.0 & -0.3902 * & 0.2556 &
0.1564 & 0.8983 & 0.1961 + ccccrccccc @xmath223 & @xmath411 & @xmath412 & @xmath231 & @xmath232 & @xmath241 & @xmath413 & @xmath272 & @xmath271 & @xmath224 + + 6.0 & 6.052&@xmath414&0.9871&0.9874 & 0.371(-6)&0.1097 & 1.7401&-1.1937&1.0000 + 5.0 & 5.075&@xmath414&0.9776&0.9783 & 0.193(-5)&0.1430 & 1.5944&-1.2095&1.0000 + 4.0 & 4.120&@xmath414&0.9557&0.9579 & 0.146(-4)&0.1958 & 1.4392&-1.2320&1.0002 + 3.5 & 3.661&@xmath414&0.9331&0.9378 & 0.492(-4)&0.2343 & 1.3602&-1.2467&1.0004 + 3.2 & 3.396&@xmath414&0.9117&0.9193 & 0.111(-3)&0.2629 & 1.3145&-1.2565&1.0006 + 3.0 & 3.228&@xmath414&0.8921&0.9027 & 0.200(-3)&0.2846 & 1.2859&-1.2633&1.0009 + 2.8 & 3.068&@xmath414&0.8663&0.8814 & 0.373(-3)&0.3084 & 1.2603&-1.2698&1.0014 + 2.6 & 2.920&@xmath414&0.8322&0.8537 & 0.724(-3)&0.3340 & 1.2401&-1.2754&1.0023 + 2.344 * & 2.759&@xmath414&0.7716&0.8049 & 0.178(-2)&0.3681 & 1.2283&-1.2790&1.0042 + 2.0 & 2.623&@xmath414&0.6498&0.7045 & 0.635(-2)&0.4090 & 1.2651&-1.2663&1.0107 + + 6.0 & 5.173 & 13.36&0.9845&0.9848 & 0.424(-6)&0.1146 & 1.5553&-1.2860&0.8533 + 5.0 & 4.344 & 11.22&0.9732&0.9737 & 0.210(-5)&0.1458 & 1.4727&-1.2953&0.8534 + 4.0 & 3.537 & 9.14&0.9475&0.9495 & 0.146(-4)&0.1930 & 1.3956&-1.3064&0.8536 + 3.5 & 3.150 & 8.14&0.9221&0.9259 & 0.457(-4)&0.2260 & 1.3647&-1.3120&0.8539 + 3.2 & 2.928 & 7.56&0.8989&0.9049 & 0.972(-4)&0.2499 & 1.3520&-1.3146&0.8542 + 3.0 & 2.787 & 7.20&0.8784&0.8866 & 0.166(-3)&0.2679 & 1.3477&-1.3155&0.8547 + 2.922 * & 2.734 & 7.06&0.8690&0.8782 & 0.207(-3)&0.2753 & 1.3473&-1.3156&0.8549 + 2.8 & 2.652 & 6.85&0.8524&0.8636 & 0.293(-3)&0.2875 & 1.3485&-1.3153&0.8553 + 2.6 & 2.527 & 6.53&0.8192&0.8345 & 0.534(-3)&0.3090 & 1.3566&-1.3132&0.8563 + 2.0 & 2.258 & 5.83&0.6525&0.6899 & 0.396(-2)&0.3794 & 1.4748&-1.2770&0.8653 + 1.736 & 2.245 & 5.80&0.5351&0.5835 & 0.103(-1)&0.4036 & 1.6258&-1.2255&0.8772 + + 6.0 & 4.878 & 10.32&0.9836&0.9838 & 0.437(-6)&0.1171 & 1.4952&-1.3223&0.8042 + 5.0 & 4.099 & 8.67&0.9716&0.9721 & 0.213(-5)&0.1479 & 1.4296&-1.3297&0.8043 + 4.0 & 3.340 & 7.07&0.9446&0.9465 & 0.144(-4)&0.1938 & 1.3749&-1.3377&0.8045 + 3.5 & 2.978 & 6.30&0.9182&0.9218 & 0.444(-4)&0.2256 & 1.3588&-1.3406&0.8048 + 3.280 * & 2.825 & 5.98&0.9014&0.9064 & 0.760(-4)&0.2421 & 1.3566&-1.3410&0.8051 + 3.2 & 2.770 & 5.86&0.8943&0.8998 & 0.931(-4)&0.2486 & 1.3570&-1.3409&0.8053 + 3.0 & 2.638 & 5.58&0.8733&0.8808 & 0.157(-3)&0.2658 & 1.3612&-1.3400&0.8057 + 2.8 & 2.512 & 5.31&0.8470&0.8571 & 0.274(-3)&0.2847 & 1.3717&-1.3375&0.8065 + 2.6 & 2.396 & 5.07&0.8136&0.8274 & 0.493(-3)&0.3053 & 1.3913&-1.3324&0.8076 + 2.0 & 2.145 & 4.54&0.6490&0.6822 & 0.352(-2)&0.3748 & 1.5575&-1.2825&0.8174 + 1.682 & 2.146 & 4.54&0.5069&0.5517 & 0.112(-1)&0.4034 & 1.7850&-1.2052&0.8341 + + + + + + 6.0 & 4.026 & 13.72&0.9894&0.9895 & 0.144(-6)&0.1713 & 1.4318&-1.3369&0.6663 + 5.0 & 3.373 & 11.50&0.9816&0.9819 & 0.719(-6)&0.2188 & 1.3528&-1.3502&0.6663 + 4.0 & 2.732 & 9.31&0.9638&0.9648 & 0.506(-5)&0.2917 & 1.2766&-1.3668&0.6667 + 3.5 & 2.422 & 8.25&0.9459&0.9480 & 0.161(-4)&0.3434 & 1.2431&-1.3759&0.6672 + 3.2 & 2.241 & 7.64&0.9294&0.9327 & 0.346(-4)&0.3813 & 1.2266&-1.3811&0.6679 + 3.0 & 2.125 & 7.24&0.9147&0.9193 & 0.598(-4)&0.4098 & 1.2182&-1.3840&0.6686 + 2.8 & 2.013 & 6.86&0.8957&0.9022 & 0.107(-3)&0.4413 & 1.2127&-1.3860&0.6698 + 2.660 * & 1.939 & 6.61&0.8792&0.8874 & 0.163(-3)&0.4651 & 1.2115&-1.3865&0.6710 + 2.6 & 1.908 & 6.50&0.8712&0.8803 & 0.197(-3)&0.4757 & 1.2117&-1.3864&0.6717 + 2.0 & 1.664 & 5.67&0.7405&0.7667 & 0.159(-2)&0.5884 & 1.2636&-1.3615&0.6889 + 1.721 & 1.636 & 5.58&0.6317&0.6720 & 0.481(-2)&0.6266 & 1.3521&-1.3146&0.7143 + + 6.0 & 3.398 & 10.87&0.9890&0.9891 & 0.149(-6)&0.2128 & 1.3490&-1.3952&0.5621 + 5.0 & 2.850 & 9.11&0.9808&0.9811 & 0.739(-6)&0.2700 & 1.2860&-1.4083&0.5626 + 4.0 & 2.316 & 7.41&0.9623&0.9633 & 0.512(-5)&0.3557 & 1.2303&-1.4232&0.5647 + 3.5 & 2.063 & 6.60&0.9438&0.9458 & 0.160(-4)&0.4144 & 1.2099&-1.4299&0.5675 + 3.2 & 1.919 & 6.14&0.9269&0.9301 & 0.342(-4)&0.4559 & 1.2024&-1.4327&0.5706 + 3.0 & 1.827 & 5.84&0.9118&0.9162 & 0.587(-4)&0.4863 & 1.2006&-1.4334&0.5737 + 2.987 * & 1.822 & 5.83&0.9107&0.9152 & 0.608(-4)&0.4884 & 1.2006&-1.4335&0.5739 + 2.8 & 1.741 & 5.57&0.8926&0.8987 & 0.104(-3)&0.5189 & 1.2023&-1.4327&0.5779 + 2.6 & 1.662 & 5.32&0.8678&0.8765 & 0.190(-3)&0.5531 & 1.2089&-1.4296&0.5836 + 2.0 & 1.500 & 4.80&0.7382&0.7628 & 0.149(-2)&0.6538 & 1.2835&-1.3897&0.6194 + 1.688 & 1.514 & 4.84&0.6158&0.6556 & 0.515(-2)&0.6787 & 1.3948&-1.3250&0.6631 + ccccrcccccc @xmath223 & @xmath415 & @xmath416 & @xmath417 & @xmath418 & @xmath419 & @xmath420 & @xmath421 & @xmath422 & @xmath423 & @xmath271 + + 6.0 & 1.5147 & -2.1704 & 0.0000 & 0.0000 & -0.1234 & -0.0010 & -0.5927 & -0.0015 & 0.0374 & -1.3369 + 5.0 & 1.5142 & -2.1701 & 0.0000 & -0.0001 & -0.1440 & -0.0012 & -0.5924 & -0.0015 & 0.0450 & -1.3502 + 4.0 & 1.5121 & -2.1689 & 0.0000 & -0.0005 & -0.1714 & -0.0015 & -0.5913 & -0.0015 & 0.0562 & -1.3668 + 3.5 & 1.5086 & -2.1669 & 0.0000 & -0.0010 & -0.1879 & -0.0016 & -0.5895 & -0.0014 & 0.0638 & -1.3759 + 3.2 & 1.5042 & -2.1644 & 0.0001 & -0.0016 & -0.1981 & -0.0017 & -0.5872 & -0.0014 & 0.0692 & -1.3811 + 3.0 & 1.4993 & -2.1616 & 0.0001 & -0.0023 & -0.2047 & -0.0018 & -0.5847 & -0.0014 & 0.0731 & -1.3840 + 2.8 & 1.4917 & -2.1572 & 0.0002 & -0.0034 & -0.2106 & -0.0018 & -0.5807 & -0.0014 & 0.0774 & -1.3860 + 2.660 * & 1.4839 & -2.1526 & 0.0004 & -0.0045 & -0.2140 & -0.0019 & -0.5767 & -0.0014 & 0.0804 & -1.3865 + 2.6 & 1.4797 & -2.1502 & 0.0004 & -0.0051 & -0.2151 & -0.0019 & -0.5745 & -0.0014 & 0.0817 & -1.3864 + 2.0 & 1.3752 & -2.0864 & 0.0033 & -0.0182 & -0.2027 & -0.0027 & -0.5211 & -0.0014 & 0.0925 & -1.3615 + 1.696 & 1.2219 & -1.9844 & 0.0106 & -0.0349 & -0.1615 & -0.0041 & -0.4451 & -0.0013 & 0.0906 & -1.3082 + cccccc @xmath223 & @xmath424 & @xmath425 & @xmath426 & @xmath427 & @xmath272 + + 6.0 & 1.5437 & 0.0000 & -0.0017 & -0.1102 & 1.4318 + 5.0 & 1.4524 & 0.0000 & -0.0008 & -0.0988 & 1.3528 + 4.0 & 1.3623 & 0.0001 & 0.0006 & -0.0864 & 1.2766 + 3.5 & 1.3213 & 0.0002 & 0.0015 & -0.0799 & 1.2431 + 3.2 & 1.3000 & 0.0005 & 0.0022 & -0.0760 & 1.2266 + 3.0 & 1.2882 & 0.0007 & 0.0026 & -0.0734 & 1.2182 + 2.8 & 1.2794 & 0.0012 & 0.0030 & -0.0709 & 1.2127 + 2.660 * & 1.2757 & 0.0017 & 0.0032 & -0.0692 & 1.2115 + 2.6 & 1.2749 & 0.0021 & 0.0033 & -0.0685 & 1.2117 + 2.0 & 1.3112 & 0.0130 & 0.0032 & -0.0638 & 1.2636 + 1.696 & 1.3890 & 0.0388 & 0.0008 & -0.0647 & 1.3638 + cccccccccccc @xmath223 & @xmath411 & @xmath412 & @xmath231 & @xmath232 & @xmath241 & @xmath413 & @xmath272 & @xmath271 & @xmath224 & @xmath428 + + 6.0 & 6.073&@xmath414&0.9874&0.9794 & 0.221(-2)&0.1092 & 1.8057&-1.1904&1.0009&-0.0623 + 5.0 & 5.105&@xmath414&0.9784&0.9651 & 0.376(-2)&0.1418 & 1.6809&-1.2039&1.0016&-0.0816 + 4.0 & 4.162&@xmath414&0.9585&0.9348 & 0.711(-2)&0.1929 & 1.5603&-1.2211&1.0031&-0.1130 + 3.5 & 3.710&@xmath414&0.9390&0.9065 & 0.103(-1)&0.2299 & 1.5072&-1.2307&1.0046&-0.1369 + 3.2 & 3.450&@xmath414&0.9214&0.8819 & 0.131(-1)&0.2571 & 1.4810&-1.2363&1.0061&-0.1554 + 3.0 & 3.283&@xmath414&0.9058&0.8609 & 0.156(-1)&0.2777 & 1.4674&-1.2394&1.0074&-0.1701 + 2.8 & 3.124&@xmath414&0.8860&0.8352 & 0.186(-1)&0.3005 & 1.4583&-1.2417&1.0092&-0.1871 + 2.625 * & 2.992&@xmath414&0.8641&0.8079 & 0.219(-1)&0.3221 & 1.4553&-1.2425&1.0112&-0.2042 + 2.6 & 2.974&@xmath414&0.8605&0.8035 & 0.224(-1)&0.3254 & 1.4554&-1.2424&1.0115&-0.2068 + 2.115**&2.677&@xmath414&0.7609&0.6909 & 0.367(-1)&0.3915 & 1.4949&-1.2298&1.0218&-0.2696 + 2.0 & 2.630&@xmath414&0.7256&0.6543 & 0.416(-1)&0.4066 & 1.5208&-1.2208&1.0259&-0.2880 + + 6.0 & 5.189 & 13.40&0.9848&0.9782 & 0.175(-2)&0.1141 & 1.6096&-1.2832&0.8541&-0.0524 + 5.0 & 4.365 & 11.27&0.9739&0.9634 & 0.284(-2)&0.1449 & 1.5423&-1.2907&0.8546&-0.0671 + 4.0 & 3.565 & 9.21&0.9499&0.9324 & 0.503(-2)&0.1911 & 1.4893&-1.2983&0.8558&-0.0899 + 3.5 & 3.182 & 8.22&0.9269&0.9039 & 0.699(-2)&0.2232 & 1.4756&-1.3007&0.8571&-0.1067 + 3.345 * & 3.067 & 7.92&0.9172&0.8920 & 0.780(-2)&0.2348 & 1.4746&-1.3009&0.8576&-0.1129 + 3.2 & 2.962 & 7.65&0.9065&0.8792 & 0.867(-2)&0.2465 & 1.4756&-1.3007&0.8582&-0.1194 + 3.0 & 2.821 & 7.29&0.8887&0.8584 & 0.101(-1)&0.2640 & 1.4809&-1.2995&0.8593&-0.1294 + 2.842 * * & 2.714 & 7.01&0.8716&0.8387 & 0.115(-1)&0.2790 & 1.4893&-1.2975&0.8604&-0.1383 + 2.8 & 2.687 & 6.94&0.8666&0.8329 & 0.119(-1)&0.2832 & 1.4922&-1.2968&0.8608&-0.1408 + 2.6 & 2.560 & 6.61&0.8388&0.8017 & 0.141(-1)&0.3042 & 1.5117&-1.2918&0.8627&-0.1542 + 1.734 & 2.231 & 5.76&0.5976&0.5539 & 0.357(-1)&0.4054 & 1.8183&-1.1935&0.8899&-0.2491 + + 6.0 & 4.892 & 10.35&0.9839&0.9778 & 0.160(-2)&0.1167 & 1.5459&-1.3196&0.8050&-0.0492 + 5.0 & 4.117 & 8.71&0.9723&0.9627 & 0.257(-2)&0.1470 & 1.4941&-1.3255&0.8055&-0.0625 + 4.0 & 3.365 & 7.12&0.9470&0.9312 & 0.447(-2)&0.1921 & 1.4606&-1.3303&0.8066&-0.0830 + 3.706 * & 3.151 & 6.67&0.9341&0.9157 & 0.537(-2)&0.2096 & 1.4580&-1.3307&0.8073&-0.0912 + 3.5 & 3.006 & 6.36&0.9227&0.9022 & 0.615(-2)&0.2232 & 1.4596&-1.3304&0.8078&-0.0979 + 3.230 * * & 2.820 & 5.96&0.9038&0.8801 & 0.741(-2)&0.2432 & 1.4674&-1.3288&0.8089&-0.1079 + 3.2 & 2.800 & 5.92&0.9013&0.8773 & 0.757(-2)&0.2456 & 1.4688&-1.3285&0.8090&-0.1091 + 3.0 & 2.668 & 5.64&0.8828&0.8562 & 0.878(-2)&0.2624 & 1.4813&-1.3258&0.8101&-0.1180 + 2.8 & 2.542 & 5.38&0.8598&0.8305 & 0.103(-1)&0.2809 & 1.5010&-1.3211&0.8116&-0.1282 + 2.6 & 2.425 & 5.13&0.8312&0.7990 & 0.122(-1)&0.3012 & 1.5304&-1.3137&0.8136&-0.1401 + 1.679 & 2.131 & 4.51&0.5634&0.5264 & 0.338(-1)&0.4053 & 1.9611&-1.1760&0.8463&-0.2354 + + + + + 6.0 & 4.045 & 13.79&0.9895&0.9847 & 0.127(-2)&0.1702 & 1.4714&-1.3338&0.6683&-0.0373 + 5.0 & 3.398 & 11.58&0.9820&0.9741 & 0.208(-2)&0.2166 & 1.4040&-1.3450&0.6697&-0.0479 + 4.0 & 2.769 & 9.44&0.9651&0.9516 & 0.373(-2)&0.2867 & 1.3461&-1.3574&0.6728&-0.0645 + 3.5 & 2.466 & 8.41&0.9486&0.9303 & 0.523(-2)&0.3354 & 1.3260&-1.3628&0.6759&-0.0769 + 3.2 & 2.292 & 7.81&0.9337&0.9117 & 0.652(-2)&0.3706 & 1.3195&-1.3647&0.6787&-0.0863 + 3.056 * & 2.211 & 7.53&0.9246&0.9005 & 0.729(-2)&0.3893 & 1.3186&-1.3650&0.6805&-0.0916 + 3.0 & 2.180 & 7.43&0.9206&0.8957 & 0.762(-2)&0.3969 & 1.3188&-1.3650&0.6813&-0.0938 + 2.8 & 2.073 & 7.06&0.9041&0.8759 & 0.899(-2)&0.4255 & 1.3220&-1.3638&0.6848&-0.1025 + 2.6 & 1.972 & 6.72&0.8831&0.8512 & 0.107(-1)&0.4565 & 1.3307&-1.3605&0.6895&-0.1126 + 2.570 * * & 1.957 & 6.67&0.8795&0.8470 & 0.110(-1)&0.4613 & 1.3326&-1.3597&0.6903&-0.1143 + 1.730 & 1.703 & 5.81&0.6886&0.6420 & 0.276(-1)&0.5933 & 1.5182&-1.2710&0.7500&-0.1913 + + 6.0 & 3.460 & 11.06&0.9891&0.9845 & 0.122(-2)&0.2077 & 1.3903&-1.3908&0.5715&-0.0339 + 5.0 & 2.926 & 9.36&0.9813&0.9738 & 0.197(-2)&0.2607 & 1.3385&-1.4012&0.5763&-0.0435 + 4.0 & 2.409 & 7.70&0.9638&0.9510 & 0.350(-2)&0.3381 & 1.2990&-1.4113&0.5849&-0.0585 + 3.5 & 2.163 & 6.92&0.9468&0.9296 & 0.487(-2)&0.3901 & 1.2895&-1.4142&0.5921&-0.0698 + 3.371 * & 2.101 & 6.72&0.9408&0.9222 & 0.534(-2)&0.4054 & 1.2890&-1.4144&0.5945&-0.0733 + 3.2 & 2.022 & 6.47&0.9315&0.9108 & 0.606(-2)&0.4267 & 1.2900&-1.4140&0.5982&-0.0784 + 3.0 & 1.933 & 6.18&0.9181&0.8947 & 0.707(-2)&0.4534 & 1.2942&-1.4124&0.6033&-0.0853 + 2.839 * * & 1.864 & 5.96&0.9049&0.8791 & 0.805(-2)&0.4763 & 1.3006&-1.4098&0.6083&-0.0917 + 2.8 & 1.848 & 5.91&0.9013&0.8749 & 0.832(-2)&0.4820 & 1.3026&-1.4090&0.6097&-0.0934 + 2.6 & 1.769 & 5.66&0.8799&0.8502 & 0.990(-2)&0.5123 & 1.3167&-1.4029&0.6176&-0.1029 + 1.700 & 1.596 & 5.10&0.6739&0.6301 & 0.269(-1)&0.6335 & 1.5440&-1.2852&0.7053&-0.1832 + cddccccccccc & & & & & & & & & + & & & & & & & @xmath233 & @xmath429 & & & @xmath233 + @xmath0 & @xmath1 & @xmath290 & @xmath430 & @xmath431 & @xmath231 & @xmath232 & [ hz ] & [ hz ] & @xmath231 & @xmath232 & [ hz ] + 0.5&0.125 & 0.1145&3.7178&3.387 & 0.821 & 0.839 & 258 & 596 & & & + 0.5&0.1380 & 0.125 & 3.5656&3.326 & 0.828 & 0.845 & 289 & 597 & 0.830 & 0.850 & 279 + 0.5&0.1941 & 0.1667&3.0998&3.072 & 0.865 & 0.875 & 404 & 576 & 0.857 & 0.871 & 399 + 0.5&0.2 & 0.1707&3.0622&3.046 & 0.869 & 0.878 & 413 & 572 & & & + 0.5&0.2481 & 0.2 & 2.7997&2.841 & 0.900 & 0.905 & 471 & 531 & 0.880 & 0.891 & 488 + 0.5&0.25 & 0.2010&2.7904&2.833 & 0.901 & 0.906 & 473 & 529 & & & + + 1.0&0.125 & 0.1008 & 3.6758&3.463 & 0.829 & 0.846 & 297 & 362 & & & + 1.0&0.2 & 0.1332 & 3.1631&3.228 & 0.878 & 0.887 & 474 & 438 & & & + 1.0&0.25 & 0.1405 & 3.0364&3.154 & 0.911 & 0.915 & 542 & 475 & & & + cdcccc @xmath0 & @xmath1 & @xmath432 [ hz ] & @xmath433 [ hz ] & @xmath434 [ hz ] & @xmath435 [ hz ] + 0.5&0.125 & 220 & 267 & 508 & 617 + 0.5&0.2 & 353 & 419 & 488 & 580 + 0.5&0.25 & 409 & 475 & 458 & 531 + + 1.0&0.125 & 248 & 300 & 302 & 366 + 1.0&0.2 & 397 & 471 & 367 & 434 + 1.0&0.25 & 450 & 529 & 394 & 463 + | using an energy variational method , we calculate quasi - equilibrium configurations of binary neutron stars modeled as compressible triaxial ellipsoids obeying a polytropic equation of state .
our energy functional includes terms both for the internal hydrodynamics of the stars and for the external orbital motion .
we add the leading post - newtonian ( pn ) corrections to the internal and gravitational energies of the stars , and adopt hybrid orbital terms which are fully relativistic in the test - mass limit and always accurate to pn order .
the total energy functional is varied to find quasi - equilibrium sequences for both corotating and irrotational binaries in circular orbits .
we examine how the orbital frequency at the innermost stable circular orbit depends on the polytropic index @xmath0 and the compactness parameter @xmath1 .
we find that , for a given @xmath1 , the innermost stable circular orbit along an irrotational sequence is about 17% larger than the innermost secularly stable circular orbit along the corotating sequence when @xmath2 , and 20% larger when @xmath3 .
we also examine the dependence of the maximum neutron star mass on the orbital frequency and find that , if pn tidal effects can be neglected , the maximum equilibrium mass increases as the orbital separation decreases . |
[ intr ] during the last two decades - using balloon flights and space - borne missions - , the fluxes of galactic cosmic rays ( gcr ) and their energy distributions were observed in different phases of solar activity .
these data allow one to attempt a better understanding of processes related to the transport of gcrs through the heliosphere . furthermore , the study of propagation properties - i.e. , the effect of solar modulation on the fluxes - of gcrs may , in turn , provide a tool to determine demodulated local interstellar spectra ( lis ) of gcr components , for instance , protons , light - nuclei , electrons , positrons , anti - protons , etc .
, thus , a further understand of processes of generation , acceleration and diffusion within the milky way ( e.g. , see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in addition , an accurate determination of demodulated spectra may allow one to untangle features due to new physics - i.e. , dark matter ( e.g. , see * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * and references therein ) - or astrophysical sources ( e.g. , see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) .
recently , spectra of gcrs were obtained using dedicated spectrometers on space born - missions ( e.g. , see * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and balloon flights ( e.g. , see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
these spectra were measured i ) with an accuracy down to about or less than 30% and ii ) covering a time duration longer than a solar cycle , i.e. , these spectra were measured under solar conditions largely different .
these data can be hopefully exploited to determine a general treatment of solar modulation in the inner heliosphere to be used for different phases of solar activity and a better understanding of space radiation environment close to earth ( e.g. , see * ? ? ? * and references therein ) . in the near future
even more accurate and systematic data will be available from ams-02 .
this spectrometer is operational onboard of the international space station from may 2011 and is expected to collect data for more than a solar cycle @xcite .
these observations will allow one to obtain accurate spectra with different solar activity conditions from some hundreds mev up to very high energy ( a few tev s ) ; in addition , using the same experimental apparatus , systematic errors on measured fluxes are expected to be minimized .
furthermore , observations made by the ulysses spacecraft @xcite in the inner heliosphere could determine a latitudinal dependence of gcr ( mostly protons ) intensity with an equatorial southward offset minimum and a north polar excess ( e.g. , see * ? ? ?
finally , it has be remarked that modulation phenomena were observed at low energies ( i.e. , lower than 500mev / nucleon ) in the outer heliosphere ( e.g. , see * ? ? ?
* ) and are currently investigated , for instance , by @xcite ( see also references therein ) . in the present model , a two dimensional ( 2-d ) - i.e. , depending on the helio - colatitude and radial distance from the sun @xcite - monte carlo approach is adopted to solve the transport equation of propagation of gcrs down to the inner heliosphere , without addressing cr modulation observed in the outer .
the model exhibits a slow time dependence because of the ( almost ) monthly averages of solar activity parameters adopted for the i ) solar wind speed ( @xmath0 ) , ii ) tilt angle ( @xmath1 ) of the neutral sheet and iii ) diffusion parameter @xmath2 ( discussed in sect .
[ ffmodel ] ) .
furthermore , one has to remark that the solar wind usually takes of the order of or more than one year to reach the border of the heliosphere . as a consequence , the above parameters
are locally evaluated within the heliosphere , allowing the modulation treatment to better ( or _ dynamically _ ) account for the effects of solar activity as a function of the distance from the sun .
in addition , the current treatment accounts for effects due to the charge sign of particles ( i.e. , the so - called _ particle drift effect _ ) , e.g. , those related , for instance , to a ) the curvature and gradient of the interplanetary magnetic field ( imf ) and b ) the extension of the neutral current sheet inside the heliosphere .
thus the model introduces a dependence on the sign solar - field polarity ( @xmath3 ) ( e.g. , see * ? ? ?
* ; * ? ? ?
the present code allows the fluxes of protons ( as well antiprotons ) and helium nuclei to be modulated from the border of the heliosphere down to earth - but outside earth s magnetosphere @xcite - in order to compare them with the available experimental observations .
furthermore , electrons and positrons modulated spectra can be derived accounting for the additional collision , radiative and inverse compton energy - losses ( see * ? ? ?
* ) . in the next sections ,
the heliosphere , drifts , diffusion tensor , determination of the diffusion parameter , dependence of both the solar wind and imf on the radial distance and helio - colatitude , neutral current sheet are discussed ( sects .
[ model][antysymmetric_part ] ) . then , the implementation of the mathematical model and the parametrization with the dynamical treatment of heliosphere are treated ( sects .
[ dynamic ] , [ code ] ) .
finally , comparisons among obtained modulated spectra of differential intensities with those experimentally observed are performed and discussed ( sects .
[ results][heliosph_latid_dep ] ) .
the transport of galactic protons ( gp ) inside the heliosphere was initially treated by @xcite , who demonstrated that - in the framework of statistical physics - the random walk of the cosmic ray particles is a markoff process , describable by a fokker
planck equation ( hereafter fpe ) ( e.g. , see also @xcite , and also sections 4.1.2.4 of ( * ? ? ?
* and references therein ) ) .
thus ( at the time @xmath4 ) , the number density , at the position @xmath5 and time @xmath4 can be found , for instance , expressed in equation ( 1 ) of @xcite ( see also references therein ) . ]
@xmath6 of gps per unit interval of particle kinetic energy @xmath7 ( the so - called differential density ) can be obtained from the solution of the fpe : @xmath8 ( e.g. , see @xcite , equation ( 4.75 ) in section 4.1.2.6 of @xcite and references therein ) with @xmath9 the solar wind velocity along the axis @xmath10 , @xmath11 the drift velocity ( e.g. , see @xcite and also @xcite and references therein ) , @xmath12 and @xmath13 the antisymmetric and symmetric part of the diffusion tensor - respectively - , @xmath14 and @xmath15 the rest mass of the proton . the number density @xmath6 is related to the differential intensity @xmath16 as : @xmath17 where @xmath18 is the speed of the gcr particle
. equation ( [ eq_parker ] ) - as well known - describes i ) the diffusion of gcrs by magnetic irregularities , ii ) the so - called _ adiabatic - energy changes _ associated with expansions and compressions of cosmic radiation , iii ) the convection effect resulting from the solar wind with velocity @xmath19 and iv ) the drift effects related to the _ drift velocity _ ( @xmath20 ) . in turn
, the drift velocity is determined by the antisymmetric part of the diffusion tensor [ see eq . ( [ drift_antisym_ten ] ) and sect .
[ antysymmetric_part ] ] which accounts for gradient , curvature and current sheet drifts of particles in the imf , i.e. , it depends on the charge sign of particles . furthermore - as discussed by @xcite - , one can re - write eq.([eq_parker ] ) as @xmath21.\ ] ] thus , one obtains that drift effects are accounted for by a convection velocity in which the drift velocity is added to the solar wind velocity . in this way , the resulting _ effective convection velocity _ may non - negligibly differ from that due to the solar wind ; but - as remarked by @xcite - noting that @xmath22 , one finds that drift effects do not contribute to the adiabatic - energy changes [ second right - hand term of eqs .
( [ eq_parker ] , [ eq_parker1 ] ) ] . even if drift effects are included in eqs .
( [ eq_parker ] , [ eq_parker1 ] ) , some modulation models neglected it ( e.g. , see * ? ? ?
* ; * ? ? ?
* and references therein ) .
gradients of particle density can also result from the convection effect .
drift mechanisms can modify both the radial and ( solar ) latitude dependence of the gradient magnitude .
for instance , drift motions can affect modulated gcr spectra by redirecting particles within the heliosphere @xcite .
when the particle larmor radius is much shorter than the magnetic - field scale length , drift effects can be taken into account by evaluating the average distance in which a relevant field variation occurs .
drift effects affect particle motions over large distances due to the large scale variation of the imf strength .
different intensities of gcr modulation were observed in time periods with opposite field polarity , for instance , by @xcite .
thus , it is necessary to explicitly consider particle drifts inside the equation of propagation of gcr .
as well known for a reference system with the 3rd coordinate along the average magnetic field , the symmetric part of the diffusion tensor ( or coefficient ) - for an isotropic perpendicular diffusion - includes both the transverse ( @xmath23 ) and parallel ( @xmath24 ) components ( e.g. , see * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* ) . in turn , for a standard parker field [ eq . ( [ fun_parkerfield ] ) ] these two components
are related to the radial component in heliocentric spherical coordinates as @xmath25 with @xmath26 the angle between radial and magnetic field directions - the so - called _ spiral angle _ [ eq . ( [ spiral_angle ] ) ] - ( e.g. , see * ? ? ? * ; * ? ? ?
* ) and @xmath27 , where @xmath28 is the polar angle @xcite .
it has to be remarked that the general transformations of the symmetric and antisymmetric parts of the diffusion tensor from field - aligned to heliospheric ( spherical ) coordinates can be found in @xcite .
furthermore , it has to be remarked that a general discussion about the role of parallel and perpendicular diffusion is available in @xcite .
@xcite ( see also * ? ? ?
* ) suggested that the parallel diffusion coefficient is given by @xmath29\ ] ] with @xmath30 , @xmath18 the particle velocity and @xmath31 the speed of light ; the diffusion parameter @xmath32 accounts for the dependence on the solar activity and is treated in sect .
[ ffmodel ] ; @xmath33 ( typically @xmath34nt ) is the value of imf at earth s orbit but it varies as a function of the time ; @xmath35 is the magnitude of the large scale imf ( discussed in sect .
[ hmf ] ) , thus , it depends on the heliospheric region ( sect .
[ dynamic ] ) through which gcrs are transported ; finally , the term @xmath36 takes into account the dependence on the rigidity @xmath37 of the gcr particle and is usually expressed in gv .
to a first approximation , one can assume that @xmath38 for particle rigidities above a threshold value @xmath39 within the rigidity range ( 0.41.015)gv , as commonly supposed by many authors ( e.g. , see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in the present model , @xmath40
is assumed to be equal to the value of the rigidity ( @xmath37 ) above the upper limit of the @xmath39 range , i.e. , for proton kinetic energies @xmath41gev ( see sects .
[ h_a_results ] , [ r_l_tilt_angle ] ) .
below @xmath39 , it can be usually approximated to a constant ( e.g. , see * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * ) .
it has to be remarked that nowadays treatments resulting in a more complex dependence of the diffusion tensor on rigidity are proposed by several authors ( e.g. , see * ? ? ?
* ; * ? ? ?
* and references therein ) .
some of these studies are motivated from dealing with magnetohydrodynamic turbulence in the expanding solar wind and/or accounting observations carried out on data of low energy electrons collected using spacecrafts [ for instance , ( 310)mev from ulysses spacecraft in @xcite and 16mev from pioneer 10 in @xcite ] . in heliocentric
spherical coordinates , the perpendicular diffusion coefficient has two components , one along the radial direction , @xmath42 , the other one for the polar direction @xmath43 .
@xmath44 is the ratio between perpendicular ( in the radial direction ) and parallel diffusion coefficients , i.e. , @xmath45 . in the present model ,
we use @xmath46 : this value is in the mid of the range suggested by @xcite ( see also @xcite and section 6.3 of @xcite ) . the value of the perpendicular diffusion coefficient in the polar direction ( @xmath43 ) can be assumed to be almost equal to that radial ( @xmath42 ) ( e.g. , see * ? ? ? * and references therein ) . however , @xcite suggested the usage of an _ enhanced _
@xmath43 in the polar regions in order to reproduce the amplitude and rigidity dependence of the latitudinal gradients of gcr differential intensities for protons and electrons ( e.g. , see * ? ? ?
* ; * ? ? ?
he introduced a sharp transition ( via a transition function , e.g. , see figure 7 in that article ) in the colatitude regions @xmath47 and @xmath48 .
he also derived that @xmath43 has to be increased by a factor of about ( or larger than ) 10 ; @xcite used a factor of 8 . in the current code , @xmath43 is given by : @xmath49 where the polar regions correspond to colatitudes with @xmath50 or @xmath51 , while the equatorial region corresponds to colatitudes with @xmath52 .
the solar colatitudes of @xmath53 and @xmath54 correspond to those at which the sw speed becomes constant in periods not dominated by high solar activity [ eq .
( [ fun_swfunct1 ] ) ] . the usage of the transition function can be fully implemented in current treatment , but is not required with the present overall code accuracy .
in fact , the results obtained from the so - called `` l '' _ model _ ( i.e. , the one with a better agreement with data , see sects .
[ h_a_results ] , [ r_l_tilt_angle ] ) indicate that only in periods not dominated by high solar activity the enhancement of @xmath43 [ eq .
( [ enhaced_k ] ) ] slightly improve the overall agreement with data by a few percent .
finally , in appendix [ app1 ] the diffusion coefficients in heliocentric polar coordinates are expressed in terms of those parallel and perpendicular to the imf .
moreover , it has to be remarked that the diffusion tensor i ) is not well determined during solar maxima and ii ) can be adapted to better account for the complex structure of the imf - which depends on the solar activity - found with ulysses spacecraft ( e.g. , see * ? ? ?
* and references therein ) .
for instance , potgieter , burger and ferreira ( 2001 ) - see also references therein - discussed the so - called _ propagating diffusion barriers _ and suggested a time dependent model for the diffusion coefficients .
the latter are supposed to be @xmath55^n$ ] , where @xmath56 is the imf magnitude at the time @xmath4 and @xmath57nt is the average imf magnitude during minimum modulation conditions at earth @xcite and @xmath58 is the ratio between the actual tilt - angle value ( sect . [ hmf ] ) and that close to solar minimum ( @xmath59@xmath60 ) ( e.g. , see * ? ? ?
* ; * ? ? ?
however , in the current model the time dependence of diffusion coefficients is taken into account using a diffusion parameter , which is treated in sect .
[ ffmodel ] .
the agreement with data obtained during high solar activity is discussed in sect .
[ h_a_results ] . in the ffm ( e.g. ,
see @xcite and also section 4.1.2.4 of @xcite ) , gleeson & axford ( 1968 ) assumed that , at the time @xmath4 , i ) modulation effects can be expressed with a spherically symmetric modulated differential number density @xmath6 of gcrs , ii ) the diffusion coefficient reduces to a scalar ) , it is expressed by a tensor with a symmetric and an antisymmetric part ( see discussion in sect .
[ model ] ) . ] given by a separable function of @xmath61 ( the radial distance from the sun ) and @xmath37 ( the particle rigidity in gv ) : @xmath62 with @xmath63 from eq .
( [ kp_linear ] ) for particle rigidities above @xmath64gv , and iii ) the modulation occurs in a steady - state condition , i.e. , the relaxation time of the distribution is short with respect to the solar cycle duration so that one can assume that the partial derivative of @xmath6 with respect to time is zero .
they derived that the differential intensity [ eq . ( [ eq::j ] ) ] at a radial distance @xmath61 is given by the expression @xmath65,\ ] ] where @xmath66 is the undisturbed intensity beyond the solar wind termination located at a radial distance @xmath67 from the sun ; @xmath68 is the total energy of the particle with rest mass @xmath69 and , finally , @xmath70 is the so - called force - field energy loss @xcite .
when modulation is small [ i.e. , for @xmath71 @xcite , they determined that @xmath72 where @xmath73 is the particle charge and @xmath74 is the so - called _ modulation strength _ ( or _ modulation parameter _ ) usually expressed in units of gv ( or mv ) . assuming that @xmath0 ( the solar wind speed ) and @xmath32 are almost constant , @xmath74 is linearly dependent on @xmath75 ( e.g. , see equation ( 4.64 ) of * ? ? ? * ) , from which one gets that the diffusion parameter is given by @xmath76 i.e. , @xmath32 [ similarly to @xmath74 ] is linearly dependent on @xmath75 .
as already mentioned , in the ffm the diffusion coefficient @xmath78 is a scalar quantity and does not account for effects related to the charge sign of the transported particles .
@xmath79 is independent of the species of gcr particles ( e.g. , see discussion at page 1014 of @xcite or equation ( 1 ) of usoskin and collaborators 2005 , and also @xcite ) .
usoskin and collaborators 2005 monthly determined the values of the modulation strengths [ @xmath80 for the time period from 1951 up to 2004 using measurements of neutron monitors ( i.e. , located at @xmath81au ) ; while the values of solar wind speeds are available from nasa / gsfc s omni data set through omniweb .
( left side ) and percentage differences @xmath82 [ eq . ( [ rperc ] ) ] ( right side ) as a function of the ssn value ;
the central continuous lines are obtained from a fit of @xmath2 with respect to ssn values in the range @xmath83 ; the dashed and dotted lines are obtained adding ( top ) or subtracting ( bottom ) one standard deviation from the fitted values.,width=403 ] .parameters of the polynomial expression ( [ k0_ssn ] ) . [ cols="^,^,^,^,^ " , ] the sensitivity of this approach was estimated from the differences of the simulated intensities with radii of 80 , 90 , 110 and 120au with that with a radius of 100au for protons with energies above 30gev , i.e. , for an energy region in which the spectrum is unaffected by modulation and , thus , no difference is expected . for this purpose , we defined the quantity [ see also eqs .
( [ eta_rms ] , [ eta_i ] ) ] @xmath84 with @xmath85 where @xmath86 is the differential intensity of @xmath87-th energy bin ( above 30gev ) , @xmath88 is the error due to monte carlo uncertainties for @xmath87-th energy bin , @xmath89 is the differential intensity computed with a radius of 100au and , finally `` h '' indicates 80 , 90 , 110 and 120au .
@xmath90 resulted equal to about 2.3% for heliospheres with 80 , 90 , 110 and 120au .
thus , the modulated intensities for heliospheres with radii of 80 , 90 , 100 , 110au ( and also 120au for bess-2002 ) are in agreement among them and experimental data within the present sensitivity of about 2.3% of the current approach ; at 120au the simulated intensity is marginally non compatible with that obtained with 100au for pamela2006/08 .
these results indicated that , as expected , the diffusion parameter almost accounts for effects related to the variation of the physical dimensions of the heliosphere within the present approximations . observations made by the ulysses spacecraft @xcite in the inner heliosphere
could determine a latitudinal dependence of gcr ( mostly protons ) intensity with an equatorial southward offset minimum and a north polar excess .
this dependence was discussed also in terms of modulation models which were including particle drift effects ( e.g. , see * ? ? ?
* ; * ? ? ?
* ) . for protons with energies larger than 100mev
, @xcite expressed their results in terms of the solar latitude and found that i ) the latitudinal gradient is @xmath91 , ii ) the counting rate minimum is nearly constant in a latitudinal region of @xmath92@xmath93 [ figure 2 of @xcite ] at @xmath94au [ e.g. , see @xcite ] and iii ) the rate at the minimum is about @xmath95% with respect to that at @xmath96 .
@xcite have shown that the latitudinal dip - found with the ulysses fast scan - can be reproduced in a model using the parker standard field and a polar enhancement of the diffusion tensor .
calculated at 1au as a function function of solar latitude for a ) @xmath97mev , b ) @xmath98mev , c ) @xmath99mev and d ) @xmath100mev : left - hand side during 1995 ( @xmath101 ) , right - hand during 2007 ( @xmath102).,title="fig:",scaledwidth=40.0% ] calculated at 1au as a function function of solar latitude for a ) @xmath97mev , b ) @xmath98mev , c ) @xmath99mev and d ) @xmath100mev : left - hand side during 1995 ( @xmath101 ) , right - hand during 2007 ( @xmath102).,title="fig:",scaledwidth=40.0% ] using helmod code , we could investigate the latitudinal dependence of the differential intensity at 1au , above 444mev ( as so far treated ) up to 200gev .
the heliosphere was subdivided in 20 regions equally spaced with respect to the co - latitudinal parameter @xmath103 [ sect .
[ code ] ] .
the total fluxes obtained in each region were divided by the maximum flux occurring at the north pole , thus , @xmath104 represents the normalized flux ( to that at the north pole ) as a function of the co - latitude .
in addition , the values of @xmath104 were calculated for periods of opposite magnetic polarities and compatible with ulysses pole - to - pole fast scans , i.e. , for the years 1995 with @xmath105 and 2007 with @xmath106 .
@xmath104 can be equivalently expressed as a function of the solar latitude for a comparison with the results obtained by @xcite .
@xmath104 as a function the solar latitude is shown in fig . [ fig_simlat ] for the year 1995 ( left - hand side ) and 2007 ( right - hand side ) .
@xmath104 is also shown as a function of the minimum kinetic energy accounted for protons ( @xmath107 ) , i.e. , a ) @xmath97mev , b ) @xmath98mev , c ) @xmath108mev and d ) @xmath100mev . by inspection of fig .
[ fig_simlat ] , for the year 1995 one can remark that @xmath104 has 1 ) a latitudinal gradient of @xmath109 , 2 ) an equatorial southward offset minimum in the latitudinal region @xmath110@xmath93 with @xmath97mev and 3 ) at this minimum , the flux is about @xmath95% of that at the north pole .
thus , the simulated fluxes reproduce the features of the experimental data ( above 100mev ) exhibited in figures 2 and 3 of @xcite [ see also @xcite ] regarding the period of 1995 ulysses fast scan .
however with increasing @xmath107 , the latitudinal gradient decreases ( fig .
[ fig_simlat ] , left - hand side ) . in the year 2007 , a similar minimum is exhibited for @xmath97mev with a @xmath111% north south poles asymmetry ; with increasing @xmath107 , this asymmetry gradually disappears and the flux reduction on the equatorial region is less pronounced down to @xmath112% with @xmath113mev ( fig .
[ fig_simlat ] , right - hand side ) .
it is worthwhile to note that in helmod code the magnetic field structure is treated similarly in north and south hemisphere approximating parker s magnetic - field with that suggested by @xcite ( sect .
[ hmf ] ) . however , the current 2-d model uses the complete @xmath114 diffusion tensor ( see sects .
[ hmf ] , [ code ] and appendix [ app1 ] ) which contains both symmetric and antisymmetric components in the off - diagonal terms .
the symmetric component of the off - diagonal terms [ eq .
( [ tens_our5 ] ) ] is determined by the divergence - free imf used , which exhibits a latitudinal component arising from the modification by @xcite [ eqs .
( [ fun_parkerfield ] , [ lat_comp ] , [ delta_sin ] ) ] . in the framework of the present 2-d model ,
the north polar excess and equatorial southward offset minimum shown in fig .
[ fig_simlat ] originate from the non - zero symmetric component of the off - diagonal terms .
the actual extension of the dip is related to both the enhancement of the diffusion tensor in the polar regions [ eq . ( [ enhaced_k ] ) ] and drift effects .
a systematic investigation of the solar modulation effect on the propagation of cosmic protons through the heliosphere down to the earth was carried out comparing experimental observations performed during the solar cycle 23 and simulated differential intensities obtained using helmod code .
the simulated spectra were derived from a lis [ eqs .
( [ bhh_lis ] , [ j_helmod ] ) ] , whose normalization constant ( @xmath115 ) depends on the experimental set of data ( see discussion in sect.[lis_discussion ] ) . the stochastic 2d monte carlo ( helmod ) code includes i ) a fully treated diffusion tensor with symmetric and antisymmetric off - diagonal elements , b ) a diffusion parameter which is a function of the intensity of solar activity and varies with solar polarity and phase ( sects .
[ ffmodel ] , [ dynamic ] ) and c ) a magnetic - field which is modified with respect to parker s magnetic field in the polar region as proposed by @xcite ( sect .
[ hmf ] ) . for observations performed during high solar activity , the simulated intensities ( found with a better agreement to experimental data ) were obtained i ) using the `` l '' model for the tilt angle ( @xmath1 ) [ sect .
[ dynamic ] and @xcite ] , ii ) with @xmath116 independently of the latitude and iii ) inside a heliospheric region where solar latitudes are lower than @xmath117 . for observations performed when solar activity is no longer at the maximum , the simulated intensities ( found with a better agreement to experimental data ) were obtained a ) using the `` l '' model for the tilt angle ( @xmath1 ) [ sect .
[ dynamic ] and @xcite ] , b ) with an enhancement of the diffusion tensor along the polar direction ( @xmath43 ) [ eq .
( [ enhaced_k ] ) ] and c ) inside a heliospheric region where solar latitudes are lower than @xmath118 . in addition ( within 2.3% ) ,
the simulated differential intensities for spherical heliospheres with radii of 80 , 90 and 110au ( and also 120au for bess-2002 ) are compatible with that with a radius of 100au ; a slightly lower agreement was obtained for a spherical heliosphere with a radius of 120au for pamela2006/2008 .
these results indicated that , within the present approximations , the diffusion parameter almost accounts for effects related to the variation of the physical dimensions of the heliosphere .
the simulated modulated spectrum determined for the year 1995 exhibits a latitudinal gradient of @xmath119 , an equatorial southward offset minimum in the latitudinal region @xmath110@xmath93 with @xmath97mev and at this minimum the flux is about @xmath95% of that at north pole .
thus , the simulated fluxes reproduce the features of the experimental data from ulysses fast scan in 1995 @xcite .
although the treatment is highly simplified with respect to the complexity of physical mechanisms responsible for modulation effects , the overall satisfactory agreement found allows one to remark that the choice of parameters regarding the structure of imf , diffusion tensor , diffusion parameter and tilt angle is almost appropriate to describe the experimental data .
finally , the experimental data from accurate observations over a long duration ( like those from the ams-02 spectrometer ) will allow one to undertake a deeper systematic investigation of solar modulation effects over a period longer than a solar cycle .
thus , possibly , further advancements can be put forward in the present approximations on the transport of gcr s through the heliosphere , for instance those at low rigidities , the spatial and rigidity properties of diffusion tensor .
in a reference frame with the 3rd coordinates along the average magnetic field , the matrix of the diffusion tensor used in eqs .
( [ eq_parker ] , [ eq_parker1 ] ) is given by ( e.g. , see * ? ? ?
* ) : @xmath120 in heliocentric spherical coordinates @xmath121 , for instance those used in eqs .
( [ eq::app1][eq::app3 ] ) , the matrix elements of the @xmath122 tensor are found , for instance , in equation ( 17 ) of @xcite . in a 2-d approximation ,
the matrix elements of the resulting tensor are:@xmath123 with @xmath124 and @xmath125 [ see figure 6 of @xcite ] , where @xmath26 is the spiral angle ( for a standard parker imf @xmath126 reduces to eq .
( [ spiral_angle ] ) and @xmath127 ) , and with @xmath128 , \\
\label{a_kperp } & & k_{\perp r } = \rho_k k_{||},\\ \label{a_kperpt } & & k_{\perp \theta } = \iota(\theta)\rho_k k_{||},\end{aligned}\ ] ] where @xmath129 and @xmath130 is a step function that is 1 in equatorial region and 10 in polar region ( sect . [ model ] ) .
the diffusion parameter @xmath131 is replaced by @xmath132 for an effective heliosphere of 100au ( sect .
[ ffmodel ] ) . furthermore , the matrix elements of the later tensor consist of a symmetric ( @xmath133 ) and antisymmetric ( @xmath134 ) part : @xmath135 with the antisymmetric part related to drift velocity [ eq .
( [ v_drift ] ) ] , @xmath136 , and treated in sect .
[ antysymmetric_part ] . finally , using eqs .
( [ a_kparr][a_split ] ) , eqs .
( [ tens_generic1][tens_generic4 ] ) can be re - written as:@xmath137 \ !
\left [
\iota(\theta)\rho_k\,\sin^2\xi + \cos^2\xi\!\left(\cos^2\psi + \rho_k\,\sin^2\psi\right ) \right],\\ \label{tens_our2 } & & k_{\theta\theta}=k^s_{\theta\theta}= \beta k_0 \,k_{p}(p , t)\!\left[\frac{b_{\oplus}}{3b}\right ] \!\left [ \iota(\theta)\rho_k\ , \cos^2\xi + \sin^2\xi\!\left(\,\cos^2\psi + \rho_k\,\sin^2\psi\right ) \right],\\ \label{tens_our3 } & & k_{r\theta}= k^a_{r\theta}+ k^s_{r\theta},\\ \label{tens_our4 } & & k_{\theta r}= k^a_{\theta r}+k^s_{r\theta } \end{aligned}\ ] ] with @xmath138 \left\ { \sin\xi\cos\xi\left[\cos^2\psi + \rho_k\,\sin^2\psi - \iota(\theta)\rho_k\right]\right\},\\ \label{tens_our6 } & & k^a_{\theta r}=-k^a_{r\theta}= k_a \sin\psi.\end{aligned}\ ] ] equations ( [ eq_parker ] , [ eq_parker1 ] ) can be re - expressed in heliocentric spherical coordinates as : @xmath139 where @xmath6 is the number density of gcrs ( sect . [ model ] ) and
@xmath7 is the kinetic energy . in turn , in a 2-d ( radial distance and co - latitude ) approximation , eq .
( [ eq::parkereq_sph ] ) can be re - expressed as : @xmath140 let us define the variable @xmath141 , then we obtain @xmath142 in addition , one can introduce the function @xmath143 using eqs .
( [ eq::parkereq_sph_2d][eq::f_mu ] ) , for a sw radially propagating ( i.e. @xmath144 ) we find : @xmath145 -\frac{\partial}{\partial r}\left[-f \frac{\partial}{\partial \mu}\left(\frac { k^s_{r\theta}\sqrt{1-\mu^2}}{r}\right ) \right ] -\frac{\partial}{\partial r}\left[f(v_{\rm{sw}}+v_{d , r } ) \right ] \nonumber\\ & & -\frac{\partial}{\partial \mu}\left[-\frac{f}{r^2 } \frac{\partial}{\partial r}\left(r k^s_{\theta r}\sqrt{1-\mu^2}\right ) \right ] -\frac{\partial}{\partial \mu}\left\{f \frac{\partial}{\partial \mu}\left [ \frac{k^s_{\theta \theta}(1-\mu^2)}{r^2}\right ] \right\ } \nonumber\\ & & -\frac{\partial}{\partial \mu}\left[-f\frac { v_{d,\theta}\sqrt{1-\mu^2}}{r } \right ] -\frac{\partial}{\partial t}\left[-f\frac{\alpha_{\rm rel } t}{3 r^2}\frac{\partial v_{\rm sw}r^2}{\partial r } \right ] \nonumber\\ & & + \frac{1}{2}\frac{\partial}{\partial r}\frac{\partial}{\partial r}[2k^s_{rr}f ] + \frac{1}{2}\frac{\partial}{\partial r}\frac{\partial}{\partial \mu}\left[- \frac{2 k^s_{r\theta}\sqrt{1-\mu^2}}{r } f\right ] \nonumber\\ & & + \frac{1}{2}\frac{\partial}{\partial \mu}\frac{\partial}{\partial r}\left[- \frac{2 k^s_{\theta r}\sqrt{1-\mu^2}}{r } f\right ] + \frac{1}{2}\frac{\partial}{\partial \mu}\frac{\partial}{\partial \mu}\left [ \frac{2 k^s_{\theta \theta}(1-\mu^2)}{r^2 } f\right].\end{aligned}\ ] ] furthermore , following the treatment discussed in sections 4.34.3.5 of @xcite , one can a ) express the fokker - plank equation involving @xmath146 - which , in turn , is a function of @xmath147 - as : @xmath148 + \frac{1}{2}\sum_{i , j } \frac{\partial}{\partial q_i } \frac{\partial}{\partial q_j } \{[\mathbf{\tilde d}(\mathbf q , t)]_{ij}\,f \}\ ] ] with @xmath149 and b ) obtain the equivalent set of differential equations @xmath150 where @xmath151 accounts for the so - called _ advective processes _
( e.g. , * ? ? ?
* ) , @xmath152 is the _
stochastic term _ containing @xmath153 which is the increment of the so - called _ wiener process _
( e.g. , section 4.3 of * ? ? ?
equations ( [ eq::ito_sde ] ) are termed _
stochastic differential equations _ ( sdes ) .
furthermore , one can note that i ) the first right - hand term of eq .
( [ eq::fokkergeneral ] ) is equal to those included in the first three lines of eq .
( [ eq::parkereq_xsde ] ) and ii ) the second right - hand term of eq .
( [ eq::fokkergeneral ] ) is equal to those included in the fourth and fifth line of eq .
( [ eq::parkereq_xsde ] ) . thus , using eqs .
( [ eq::parkereq_xsde ] , [ eq::fokkergeneral ] ) one derives : @xmath154-\frac{v_{d,\theta}\sqrt{1-\mu^2 } } { r}\\ -\frac{\alpha_{\rm rel } t}{3 r^2}\frac{\partial v_{\rm sw}r^2}{\partial r } \end{array}\right]\end{aligned}\ ] ] and @xmath155.\end{aligned}\ ] ] as discussed by @xcite - see section 2 therein - , the matrix @xmath156 can be downgraded to a two - by - two matrix , because second order acceleration mechanisms are not considered in eqs .
( [ eq_parker ] , [ eq_parker1 ] ) .
as already shown by @xcite - see appendix b therein - , the matrix @xmath157 is not unique .
however , there is only one positive definite , i.e. , @xmath158^{1/2 } & - k^s_{r\theta}\left(\frac{2}{k^s_{\theta\theta}}\right)^{1/2 } \\ 0 & \left[\frac{2 k^s_{\theta \theta}(1-\mu^2)}{r^2 } \right]^{1/2 } \end{array } \right].\ ] ] finally , for a 2-dimensional model ( like that treated in sects .
[ hmf ] , [ antysymmetric_part ] ) , from eqs .
( [ eq::ito_sde ] , [ eq::sde_a_vector ] ) and using eq .
( [ eq::matrixl ] ) one finds the following sdes : @xmath159 { \rm d } t + \nonumber\\ & & + \sqrt{\frac{k^s_{rr}k^s_{\theta\theta}-(k^s_{r\theta})^2}{0.5 k^s_{\theta\theta}}}\ , { \rm d}w_r - k^s_{r\theta}\sqrt{\frac{2}{k^s_{\theta\theta}}\,}\ , { \rm d}w_\mu , \\ \label{eq::app2 } { \rm d } \mu&= & \left\{-\frac{1}{r^2 } \frac{\partial}{\partial r}\left(r k^s_{\theta r}\sqrt{1-\mu^2}\right)+ \frac{\partial}{\partial \mu}\left [ \frac{k^s_{\theta \theta}(1-\mu^2)}{r^2}\right]-\frac{v_{{\rm d},\theta}\sqrt{1-\mu^2}}{r}\right\ } { \rm d } t + \nonumber\\ & & + \sqrt{\frac{2 k^s_{\theta \theta}(1-\mu^2)}{r^2 } \,}\ , { \rm d}w_\mu , \\ \label{eq::app3 } { \rm d } t & = & - \frac{2}{3 } \frac{\alpha_{\rm rel } v_{\rm sw } t}{r } { \rm d } t\end{aligned}\ ] ] with @xmath160 [ @xmath161 the increment of the wiener process .
it has be remarked that the above usage of eq .
( [ eq::matrixl ] ) ensures imaginary terms do not appear in eqs .
( [ eq::app1][eq::app6 ] ) . for an imf described by a standard parker field [ eq .
( [ fun_parkerfield ] ) ] requiring @xmath162 , the above sdes reduce to @xmath163 { \rm d } t + \left ( v_{\rm sw}+ v_{{\rm d},r } \right ) { \rm d } t + \sqrt{2k^s_{rr}\ , } \ , { \rm d}w_r , \\ \label{eq::app5 } { \rm d } \mu & = & \frac{1}{r^2 } \left\{\frac{\partial}{\partial \mu}\left [ ( 1-\mu^2)k^s_{\theta\theta}\right]\right\ } { \rm d } t - \frac{v_{{\rm d},\theta } \,\sqrt{1-\mu^2}}{r}\ , { \rm d } t \nonumber\\ & & + \sqrt { \frac{2\,k^s_{\theta \theta}\ , ( 1-\mu^2)}{r^2 } \ , } \,{\rm d}w_\mu , \\
\label{eq::app6 } { \rm d } t & = & - \frac{2}{3 } \frac{\alpha_{\rm rel } v_{\rm sw } t}{r } { \rm d } t.\end{aligned}\ ] ] gardiner ( 1985 ) - see section 4.3.1 therein - demonstrated that , following a _ euler cauchy procedure _ , the sdes can be approximated in terms of the increments @xmath164 , @xmath165 and @xmath166 occurring after a time @xmath167 has elapsed .
the corresponding increment of the wiener process is given by @xmath168 [ with @xmath161 , where @xmath169 is a random number following a gaussian distribution with a mean of zero and a standard deviation of one ( e.g. , see @xcite and appendix a of @xcite ) .
as a consequence , the sdes [ eqs .
( [ eq::app1][eq::app3 ] ) ] can be approximated by : @xmath170\delta t + \nonumber\\ & & + \omega_r\sqrt{\frac{k^s_{rr}k^s_{\theta\theta}-(k^s_{r\theta})^2}{0.5 k^s_{\theta\theta}}\,\delta t\ , } - \omega_{\mu } k^s_{r\theta}\sqrt{\frac{2}{k^s_{\theta\theta}}\,\delta t\ , } , \label{eq::app7 } \\
\delta \mu & = & \left\{-\frac{1}{r^2 } \frac{\partial}{\partial r}\left(r k^s_{\theta r}\sqrt{1-\mu^2}\right)+ \frac{\partial}{\partial \mu}\left [ \frac{k^s_{\theta \theta}(1-\mu^2)}{r^2}\right]-\frac{v_{{\rm dr},\theta}\sqrt{1-\mu^2}}{r}\right\}\delta t + \nonumber\\ & & + \omega_{\mu } \sqrt{\frac{2 k^s_{\theta \theta}(1-\mu^2)}{r^2}\ , \delta t\ , } , \label{eq::app8}\\ \delta t & = & - \frac{2}{3 } \frac{\alpha_{\rm rel } v_{\rm sw } t}{r}\delta t \label{eq::app9}.\end{aligned}\ ] ]
the authors express recognition for the contribution of francesco noventa to simulations regarding the heliosphere dimensions .
kk wishes to acknowledge vega grant agency project 2/0081/10 for support . finally , we acknowledge the use of nasa / gsfc s space physics data facility s omniweb service , and omni data .
bobik , p. , gervasi , m. , grandi , d. , rancoita , p.g . , & usoskin , i.g .
2003 , in solar variability as an input to the earth s environment .
international solar cycle studies ( iscs ) symposium , ed . : a. wilson ( noordwijk : esa publications division ) , 637640 bobik , p. , boschini , m.j . , gervasi , m. , grandi , d. , & rancoita , p.g .
2008 , proc . of the 10th icatpp , ed .
leroy , c. , rancoita , p.g . ,
barone , m. , gaddi , a. , price , l. & ruchti , r. , ( singapore : world scientific ) , 881885 bobik , p. , boschini , m.j .
, della torre , s. , gervasi , m. , grandi , d. , kudela , k. , & rancoita , p.g .
2010a , proc . of the 11th icatpp ,
leroy , c. , rancoita , p.g . ,
barone , m. , gaddi , a. , price , l. & ruchti , r. , ( singapore : world scientific ) , 210219 bobik , p. , et al .
2011a , proc . of the 12th icatpp conference on cosmic rays for particle and astroparticle physics , ed .
giani , s. leroy , c. & rancoita , p.g .
( singapore : world scientific ) , 360368 and ( arxiv:1012.3086 ) bobik , p. , et al .
2011b , proc . of the icatpp conference on cosmic rays for particle and astroparticle physics , ed . c. leroy , p.g .
rancoita , m. barone , a. gaddi , l. price & r. ruchti ( singapore : world scientific ) , 337342 and ( arxiv:1011.4843 ) braja , r. , vrnak , b. , rudjak , v. , roa , d. , hrina , d. , whl , h. , clette , f. , hochedez , j .- f .
2001 , in recent insights into the physics of the sun and heliosphere , proc . of iau
symposium 203 , edited by p. brekke , b. fleck , & j.b .
gurman ( san francisco , astronomical society of the pacific ) , 377 king , j.h .
& papatashili , n.e . , 2005 , , 110 , a02104 ; solar wind and imf magnitude average data as extracted from nasa / gsfc s omni data set through omniweb at the website : _ http://omniweb.gsfc.nasa.gov/form/dx1.html _ mertsch , p. & sarkar , s. , 2011 , proc . of the 12th icatpp conference on cosmic rays for particle and astroparticle physics , ed .
giani , s. leroy , c. & rancoita , p.g .
( singapore : world scientific ) , 535543 pdb ( 2010 ) : nakamura , k. et al .
( particle data group ) , 2010 .
review of particle physics , j. phys .
g , 37 , 075021 ; the latest versions of tables , reference data and constants are available at the particle data group web site : _ http://pdg.lbl.gov/pdg.html _ | a propagation model of galactic cosmic protons through the heliosphere was implemented using a 2-d monte carlo approach to determine the differential intensities of protons during the solar cycle 23 .
the model includes the effects due to the variation of solar activity during the propagation of cosmic rays from the boundary of the heliopause down to earth s position .
drift effects are also accounted for .
the simulated spectra were found in agreement with those obtained with experimental observations carried out by bess , ams and pamela collaborations .
in addition , the modulated spectrum determined with the present code for the year 1995 exhibits the latitudinal gradient and equatorial southward offset minimum found by ulysses fast scan in 1995 . |
the idea of a `` pulsational pair - instability supernova '' ( ppisn ) was introduced by @xcite and explored in some detail by @xcite . for a range of helium core masses above approximately 35 @xmath0 ,
the production of electron - positron pairs occurs during central neon burning and triggers an instability that leads to rapid contraction of the core and explosive nuclear burning .
if the helium core mass is above about 65 @xmath0 , the pair production instability occurs after carbon ignition and the energy released can completely unbind the star in a single explosive event known as a pair - instability supernova ( pisn ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
if the core mass is above 133 @xmath0 , for reasonable nuclear reaction rates , convection theory , and no rotation , nuclear burning is unable to reverse the collapse and a black hole forms @xcite . for helium core masses from 35 - 65 @xmath0 ,
although explosive burning is violent and energetic , it can not unbind the entire star , and cycles of instability and mass ejection can instead occur .
the energy , duration , and mass ejected by these nuclear - powered pulses increases as the helium core mass rises , and by about 45 @xmath0 they are sufficient to produce sn - like displays @xcite .
if the helium core is capped by a substantial hydrogen envelope , the first strong pulse ejects it . depending on the radius and mass of the envelope , this initial
ejection may give rise to either a faint or rather typical type iip sn .
subsequent pulses later overtake and collide with the first and produce a much brighter type iin sn .
if there is no hydrogen envelope , collisions between helium shells can produce a bright type i sn .
the helium core mass range from 45 to 55 @xmath0 is particularly interesting because the characteristic time scale between pulsations is years and the collisions between shells ejected with speeds @xmath2 1000 km s@xmath3 occur at 10@xmath4 - 10@xmath5 cm , where the collision energy is mostly dissipated by optical emission .
since the collision energy can approach 10@xmath6 erg , a superluminous event can result .
the discovery of both pi and ppi sn candidates in the local universe , and the realization that such events might be visible at high redshifts , have excited interest in this exotic explosion mechanism .
sn 2007bi at @xmath7 0.127 @xcite and sn 2213 - 1745 at @xmath8 2.06 @xcite are pisn candidates , and sn 1000@xmath90216 at @xmath7 3.90 @xcite and perhaps sn 2006 oz at @xmath7 0.376 @xcite may be ppisne .
@xcite suggest that sn 2007bi may be a magnetar spin - down powered event @xcite or it could be due to an interaction between h - poor sn ejecta and a circumstellar medium @xcite .
the nature of these transients is still under debate . @xcite
have now shown that ppisne , like pisne and type iine , may be visible to the _ james webb space telescope _ ( _ jwst _ ) and the wide - field infrared survey telescope ( wfirst ) at @xmath10 20 ( see also * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
these explosions , together with other population iii ( pop iii ) sne @xcite , could probe the properties of the first stars and galaxies @xcite , early cosmological reionization and chemical enrichment @xcite , and the origins of supermassive black holes @xcite .
understanding the observational signatures of ppisne is key to properly identifying them as more of them are discovered by the new sn factories such as the palomar transient factory ( ptf ; * ? ? ?
* ) , the panoramic survey telescope and rapid response system ( pan - starrs ; * ? ? ?
* ) and the large synoptic survey telescope ( lsst ; * ? ? ?
high-@xmath11 sn surveys by _ jwst _ and wfirst may harvest even greater numbers of ppisne if , as many suspect , the pop iii initial mass function ( imf ) is top - heavy .
@xcite modeled a ppisn in one dimension ( 1d ) with the ` kepler ` code @xcite and its light curve with the ` stella ` code @xcite . in these simulations , a large density spike formed during the collision between the second two pulses and the first .
rapid variations in the density of the spike due to a radiative instability @xcite imposed large fluctuations in the bolometric luminosity .
@xcite noted similar features in the near infrared ( nir ) light curves of the same ppisn modeled with the los alamos ` rage ` code .
such spikes would probably not appear in multidimensional flows because hydrodynamic instabilities would likely erase the sharp interface between the pulses .
it is not clear how these processes would change the luminosity of the collision .
radiation hydrodynamical simulations in at least two dimensions ( 2d ) with adaptive mesh refinement ( amr ) are clearly required to resolve the thin radiating region between the shells during the collision and properly model the light curves of ppisne .
but such simulations push the envelope of even state - of - the - art numerical codes . as a first step to this goal
, we have performed the first 2d simulations of a ppisn with the ` castro ` code with hydrodynamics but not radiation transport . in section 2
we describe our ppisn model and how it is evolved in ` castro ` . in section 3
we examine the collision of the shells in 2d in detail , and in section 4 we show how the evolution of the collision in 2d might change the ppisn light curves of previous calculations .
we conclude in section 5 .
we take as a fiducial case the ppisn examined in @xcite , whose progenitor was a solar - metallicity star with an initial main sequence mass of 110 @xmath0 .
the mass loss rate was artificially reduced so that the star had a total mass of 74.6 @xmath0 and a helium core mass of 49.9 @xmath0 when it died .
the star produces three violent outbursts .
the first , p1 , ejects most of the hydrogen envelope , making a faint type ii supernova and leaving a residual of 50.7 @xmath0 , just a bit more than the helium core itself .
after 6.8 yr , the core again contracts and encounters the pair instability , twice in rapid succession .
the total mass of the second and third pulses ( p2 and p3 ) is 5.1 @xmath0 and their kinetic energy is @xmath12 erg .
p3 collides with p2 at large optical depths that are not visible to an external observer .
these combined shells then overtake p1 at @xmath2 10@xmath4 cm and speeds of a few 1000 km s@xmath3 . about 90% of the energy of this collision
is radiated away by 10@xmath5 cm , with peak bolometric luminosities of @xmath13 . `
castro ` is a multi - dimensional amr astrophysical fluid dynamics code @xcite .
it has an unsplit piecewise parabolic method ( ppm ) hydro scheme @xcite and multispecies advection .
the equation of state used in our study was taken from @xcite , and has contributions from relativistic @xmath14 pairs of arbitrary degeneracy , ions , which are treated as an ideal gas , and photons .
densities , velocities , temperatures and mass fractions from ` kepler ` were mapped onto a 2d cylindrical grid in ` castro ` with the conservative scheme of @xcite , which guarantees that quantities such as energy and mass are strictly conserved . because only one quadrant of the star is mapped onto the mesh , outflow and reflecting boundary conditions were set on the upper and lower boundaries in @xmath15 and @xmath11 , respectively .
three levels of adaptive refinement ( for up to 64 times greater resolution along each axis ) were used to resolve the scales on which instabilities form in the flow , and the grid refinement criteria are based on gradients of density , velocity , and pressure .
we use the monopole approximation for self - gravity , in which a 1d gravitational potential is constructed from the radial average of the density and then applied to gravitational force updates everywhere in the amr hierarchy .
this approximation is very efficient , and well - suited to the nearly spherical symmetry of the star and its pulsations .
in principle , partial fallback from one pulse could collide with a subsequent ejection and seed the formation of instabilities and mixing before the pulses themselves collide .
we investigate this process with a 2d simulation in ` castro ` in which we follow fallback from p2 onto p3 .
this simulation must be performed on a much smaller mesh that excludes p1 in order to resolve the length scales of nuclear burning , which powers the pulses , and the onset of any fluid instabilities near the core .
it is initialized with a ` kepler ` output at an intermediate time from @xcite , @xmath2 100 sec before p2 .
our ` castro ` mesh is 5 @xmath16 10@xmath17 cm on a side with @xmath18 zones at the coarsest level . as in ` kepler `
, nuclear burning is calculated with a 19-isotope network that includes species from hydrogen through @xmath19 @xcite .
the star is evolved until the expulsion of p3 .
gas velocities and densities at the end of this run are shown in figure [ fig : ppsn_p2p3 ] .
fallback from p2 does trigger mild perturbations in the velocities near the surface of p3 , on the order of a few percent of the local sound speed , but they do not result in any discernible features in the densities .
this simulation demonstrates that launching collision models in ` castro ` from ` kepler ` snapshots at later times will not exclude serious dynamical instabilities originating from fallback at earlier times .
modeling perturbations in the gas due to fallback from all three pulses at the same time at the required resolution would require a grid that is four orders of magnitude larger than this one , so we approximate these features in our collision models by seeding the grid with random velocity perturbations of order about @xmath20 of local sound speed .
to study the collisions between the three pulses , we initialize ` castro ` with a ` kepler ` profile taken at a later time , after all three eruptions have occurred and when p1 is at 2 @xmath16 10@xmath5 cm , as shown in figure [ fig : init_v ] . since @xcite and @xcite both predict that most of the radiation from the collision between shells p2/p3 and p1 is emitted by the time the shock has reached 10@xmath5 cm , we set the outer boundaries of the grid in @xmath15 and @xmath11 to be 10@xmath5 cm .
we do not include all of p1 in our simulation because the density beyond @xmath21 cm is extremely low and most fluid instabilities forming during collisions have become frozen in mass coordinate before reaching this boundary .
with amr , the effective spatial resolution can be as fine as 10@xmath22 cm , which is sufficient for capturing the fine structure of each pulse , as we show in figure [ fig : init_den ] ) and fluid instabilities in the flow later on .
all three pulses are evolved until the collision shock reaches the simulation boundary , about 260 days after the launch of the run . at the beginning of the simulation ,
the peak velocities of p2 and p3 are 3.9 @xmath16 10@xmath23 cm s@xmath3 and 4.8 @xmath16 10@xmath23 cm s@xmath3 , respectively .
the faster p3 overtakes p2 within 50 days and has completely merged with it at @xmath24 .
the contact discontinuity that forms between p3 and p2 upon collision rapidly destabilizes into the mild , finger - like fluid instabilities that are visible at 2.5 @xmath16 10@xmath25 cm in figure [ fig : col_den ] .
these fingers are 2 - 3@xmath26 in size , with overdensities of 5 - 10 .
the combined pulse has a complex structure with multiple velocity peaks , the highest of which is @xmath27 cm s@xmath3 .
the bright emission from this collision is probably not visible to an external observer because it originates from very optically thick regions that are well below the photosphere at @xmath2 8 @xmath28 .
nevertheless , this radiation may be able affect the dynamics of the ejecta or diffuse out of it at later times .
after p2 and p3 collide , they together begin to overtake p1 , which now has a radius 100 times that of the original star . unlike the first collision ,
the second collision is more like that between supernova ejecta and a dense shell expelled by the star prior to its death ( type iin sne , e.g. , * ? ? ?
* ; * ? ? ?
although this collision is less violent than a type iin sn , the instabilities eventually grow to larger amplitudes than in the first collision , as we show in figure [ fig : shell_rev ] .
the clumpy structures behind the shock grow in size as the flow expands , driving more mixing and dredging up heavier elements from deeper layers .
we plot angle - averaged mass fractions for @xmath29 , @xmath30 , @xmath31 , and @xmath32 in figure [ fig : mixing ] as the shock approaches the photosphere .
most mixing happens by @xmath2 @xmath33 , and elements heavier than @xmath34 remain deeper in the ejecta , so they would not appear in the spectra .
the formation of dynamical instabilities during the collisions can be understood on analytical grounds .
both times , the instabilities appear because the shock decelerates as it plows up gas and forms a reverse shock . for a strong adiabatic shock in a power - law density profile , @xmath35 ,
the flow becomes self - similar , and the physical quantities that describe it can be combined into a single , dimensionless function , @xmath36 , where @xmath37 is the explosion energy and @xmath38 is the time @xcite . the velocity of the shock can then be derived from dimensional analysis : @xmath39 for @xmath40 , the shock decelerates as it plows up material .
this deceleration is communicated to the fluid behind by the shock at the sound speed , and it creates a pressure gradient in the direction of the deceleration . the sound wave generated by the deceleration can steepen this pressure gradient , and a reverse shock forms . if the gas pressure , @xmath41 , and the density , @xmath42 , satisfy the rayleigh - taylor ( rt ) criterion for an fluid @xcite @xmath43 then instabilities will form . whether or not a reverse shock forms can be determined by the @xmath44 profile in the surrounding medium .
if @xmath45 increases with radius , conditions are favorable for the formation of a reverse shock .
we plot this quantity for the time at which our ` castro ` run is launched in figure [ fig : rhor3 ] .
the peaks and valleys at @xmath46 were created by the ejection of p2 and p3 .
the acceleration and deceleration of the shock in this rapidly varying region causes the smaller fluid instabilities in the first collision shown in figure [ fig : col_den ] . the steady increase in @xmath44 at @xmath47
leading up to p1 allows the instabilities to grow to larger amplitudes during the second collision .
when the rt fingers become nonlinear , kelvin - helmholtz ( or shear ) instabilities arise and dominate the mixing , which continues until the collision shock breaks through p1 .
figure [ fig : grid ] shows the amr grid structure when the mixing occurs .
many fine grids are generated to resolve the clumpy structure due to mixing .
most of the luminosity from the second collision is gone before mixing ends .
we approximate the bolometric light curve of this collision by @xcite : @xmath48 where @xmath42 is the gas density just ahead of shock front , @xmath49 is the velocity of the shock , and @xmath15 is the position of the shock .
equation ( [ eq : psn_vs ] ) holds if the region in which the shock propagates is assumed to be optically thin , and is therefore an upper limit in somewhat higher densities . to take into account energy dissipation in the shock by escaping photons , we assume that the luminosity of the shock falls as @xmath50 when the shock front is beyond @xmath51 .
we assume that the temperature at the edge of the shock falls as @xmath52 due to thermal radiation , so the bolometric luminosity @xmath53 we plot this luminosity in figure [ fig : light ] .
the collision produces a very bright transient , with a peak luminosity of @xmath2 @xmath54 and a duration of @xmath2 100 days .
the peak luminosity occurs when p2+p3 collide with p1 .
its magnitude is calculated with equation ( [ eq : psn_vs ] ) , with @xmath55 , @xmath56 , and @xmath57 cm / s .
the peak luminosity and duration are similar to those of the 1d radiation hydrodynamics models of @xcite and @xcite .
but our light curve rises earlier because we assume that the environment of the shock is optically thin . unlike the 1d light curves , which exhibit artificial fluctuations due to a radiative instability , the light curve of our 2d model has a relatively smooth peak and no fluctuations .
our model suggests that radiation hydrodynamical calculations of ppisn light curves in 2d will also be smoother than those in 1d because mixing will dampen the radiative instability , but that they will have similar luminosities .
we calculate 2d light curves with equation ( 3 ) with angle - averaged densities and velocities at radii just ahead of the shock .
this light curve is a rough estimate of the luminosity of the shock and does not include radiation from dense clumps due to shell collisions .
although the 2d light curve has similar luminosities to 1d models , the light curve rising time and fading tail are still quite different between 1d and 2d .
full radiation transport in 2d is needed for more realistic light curves of ppisn models .
we have performed the first 2d hydrodynamical simulations of ppisne with the new ` castro ` code . we find that mild rt instabilities develop first during the collision of p3 with p2 and then become stronger as p2 and p3 interact with p1 .
fallback from one pulsation onto another does not seed strong instabilities but does mildly perturb the flow prior to collisions .
the appearance of dynamical instabilities and mixing has three consequences for the observational signatures of ppisne that are not captured in 1d models .
first , the radiative instabilities that cause light curves to fluctuate in 1d models probably do not happen in actual explosions . mixing between shells
dampens the formation of the density spikes in 1d models that give rise to these features .
second , the formation of dense clumps by rt fingers could trap photons and alter the luminosity of these events .
we neglect radiative cooling by metals and dust in our models so these clumps may catastrophically cool and become much denser ( and more opaque ) in actual collisions .
this in turn could amplify the rt instabilities .
finally , since mixing is strongest in the region of the flow from which most of the luminosity of the collision originates , it will alter the order in which specific absorption and emission lines appear in the spectrum over time .
in particular , our model predicts enhanced spectral lines from @xmath30 and @xmath31 at earlier times in the second collision .
the appearance of such lines may prove to be powerful diagnostics of mixing in ppisne .
since our approximate bolometric luminosities are similar to those calculated with full radiation transport in earlier work , future multidimensional calculations will probably not change the recent result that these explosions will be visible at @xmath7 10 - 20 to _ jwst _ and wfirst .
but they do demonstrate that multidimensional radiation transport will be required to model how radiation is emitted by and escapes from the complex structures formed by dynamical instabilities
. this will be key to obtaining more realistic light curves and spectra for ppisne . in future papers
, we will better survey ppisn explosions by examining more progenitor models at low metallicities with 2d radiation hydrodynamical calculations @xcite .
ppisne will soon open new windows on massive star formation in both the primordial and the local universe .
the authors thank the anonymous referee for reviewing this manuscript and providing valuable comments , and the members of ccse at lbnl for help with ` castro ` .
we also thank volker bromm , dan kasen , lars bildsten , john bell , and adam burrows for many useful discussions .
k.c . was supported by an iau - gruber fellowship , a stanwood johnston fellowship , and a kitp graduate fellowship . s.w .
acknowledges support by doe hep program under contract de - sc0010676 ; the national science foundation ( ast 0909129 ) and the nasa theory program ( nnx14ah34 g ) .
a.h . was supported by a future fellowship from the australian research council ( arc ft 120100363 ) .
djw was supported by the baden - wrttemberg - stiftung by contract research via the programme internationale spitzenforschung ii ( grant p - ls - spii/18 ) .
all numerical simulations were performed at the university of minnesota supercomputing institute and the national energy research scientific computing center .
this work has been supported by the doe grants de - sc0010676 , de - ac02 - 05ch11231 , de - gf02 - 87er40328 , de - fc02 - 09er41618 and by nsf grants ast-1109394 and phy02 - 16783 .
work at lanl was done under the auspices of the national nuclear security administration of the us department of energy at los alamos national laboratory under contract no .
de - ac52 - 06na25396 . ,
n. , et al .
2002 , in society of photo - optical instrumentation engineers ( spie ) conference series , vol .
4836 , society of photo - optical instrumentation engineers ( spie ) conference series , ed . j. a. tyson & s. wolff , 154164 profile at the beginning of the ` castro ` run .
the ejection of p2 and p3 create the spiky pattern at @xmath58 .
the ejection and subsequent expansion of p1 creates an extended and more homogeneous envelope , which is visible as the steady rise of @xmath59 at @xmath47 .
[ fig : rhor3 ] ] | massive stars that end their lives with helium cores in the range of 35 to 65 @xmath0 are known to produce repeated thermonuclear outbursts due to a recurring pair - instability . in some of these events ,
solar masses of material are ejected in repeated outbursts of several times 10@xmath1 erg each .
collisions between these shells can sometimes produce very luminous transients that are visible from the edge of the observable universe .
previous 1d studies of these events produce thin , high - density shells as one ejection plows into another . here , in the first multi - dimensional simulations of these collisions , we show that the development of a rayleigh - taylor instability truncates the growth of the high density spike and drives mixing between the shells .
the progenitor is a 110 @xmath0 solar - metallicity star that was shown in earlier work to produce a superluminous supernova .
the light curve of this more realistic model has a peak luminosity and duration that are similar to those of 1d models but a structure that is smoother . |
in the standard model of particle physics ( sm ) sub - permille accuracy is reached in the prediction of the ratios @xmath9 of decay rates for the pseudoscalar @xmath10 and @xmath11 mesons .
this excellent accuracy ( @xmath12 , @xmath13 ) is due to cancellations of the hadronic uncertainties@xcite and the precise knowledge of the electromagnetic correction @xmath14 .
the factor @xmath15 accounts for helicity suppression and makes @xmath16 very sensitive to sm extensions involving pseudo - scalar currents and non - universal corrections , which are expected to induce variations of @xmath16 of the order @xmath17 .
in particular sm extensions involving lepton flavour violation ( lfv ) are currently not ruled out by experiment .
the current world average @xmath18 ( pdg ) dates back to three experiments performed in the 1970s , and has insufficient precision ( @xmath19 ) for stringent sm tests .
a series of recent preliminary results from na48/2 and kloe represents a significant improvement : combining these results with the pdg value yields a precision of @xmath20@xcite .
on the other hand , for what concerns the future programme , the @xmath7 decay is a flavor changing neutral current process which proceeds through box and electroweak penguin diagrams .
the predicted values @xcite of the branching fraction are of the order of @xmath21 .
the hadronic matrix element can be parametrized in terms of the @xmath22 branching ratio , well known experimentally , reducing the theoretical uncertainty to about 5% and thus making the @xmath7 decay a theoretically clean environment sensitive to new physics .
the existing measurement of the @xmath7 branching ratio , based on 7 observed events @xcite , is compatible with sm within errors , but a 10% accuracy measurement of the branchig ratio is required to provide a decisive test of new physics scenarios .
this is the goal of the proposed na62 experiment at cern sps @xcite , which aims to collect @xmath8 @xmath7 events in two years , keeping background contamination at the level of 10% .
for the first phase of na62 the experimental setup is largely based on na48/2 .
the beam line k12 at cern sps is capable of delivering simultaneous narrow momentum band @xmath23 and @xmath24 beams . a central momentum of 75 gev / c was used during 2007 data taking .
most of the data were taken with the @xmath25 beam only because the muon sweeping system is more effective in rejecting @xmath26 .
conversely , about @xmath27 of the data were recorded with @xmath24 beam only in order to measure the background induced by the beam halo via @xmath28 decays .
the following detectors , located downstream a vacuum decay volume ( @xmath29 in length ) , were exploited : a magnetic spectrometer composed of four drift chambers ( dchs ) and a spectrometric magnet ( providing a track momentum resolution of @xmath30}/p=0.47\% \oplus p\cdot0.02\%$ ] ) ; a liquid krypton electromagnetic calorimeter ( lkr ) used for @xmath31 detection and particle identification ; a scintillator hodoscope ( hod ) used for timing measurement and to produce fast trigger signals .
a minimum bias trigger configuration was employed , resulting in relatively low trigger purities , but high efficiencies for leptons with momentum @xmath32 .
the main data sample was taken during the four months of data taking in 2007 .
about @xmath33 sps spills were recorded , corresponding to 300 tb of raw data .
additional data were collected in 2008 to address in detail some systematic uncertainties .
the analysis strategy is based on counting the numbers of reconstructed @xmath4 and @xmath34 candidates collected simultaneously , consequently the result does not rely on kaon flux measurement and several systematic effects , such as parts of the trigger and detection efficiencies , cancel in the ratio .
the analysis is performed independently in bins of momentum of the charged track ( @xmath35 ) , due to strong dependence of the backgrounds and acceptance on this variable .
the ratio @xmath1 is computed in each momentum bin after having subtracted the background events from the @xmath4 and @xmath34 candidates .
correction factors are then applied in order to account for acceptances and for trigger , reconstruction and particle identification efficiencies .
due to the topological similarity of @xmath4 and @xmath34 decays , a large part of the selection criteria is common for both decays .
this leads to cancellations of the related systematic uncertainties in @xmath1 .
particle identification based on the ration of energy @xmath36 deposited in the calorimeter and the momentum measured by the spectrometer @xmath37 ( @xmath38 for electrons , @xmath39 for muons ) and kinematic identification obtained with the measurement of the missing invariant mass ( @xmath40 for electrons ) are exploited for separating @xmath4 and @xmath34 decays .
the average momentum of the incoming kaons is measured with @xmath41 decays and is used for @xmath6 kinematical reconstruction ( @xmath42 ) ; a narrow kaon momentum spectrum ( @xmath43 ) is used in order to minimize the corresponding contribution to the @xmath42 resolution . after selection , @xmath44 @xmath4 candidates are identified .
the present analysis is based on the study of a sub - sample ( @xmath45 ) of the full data sample .
the study of the background is the main topic of the analysis .
detailed studies with data and montecarlo are performed in order to identify and measure the main sources of background .
the following background contributions have been identified : ( 1 ) @xmath34 events where the @xmath46 is mis - identified as an electron due to `` catastrophic '' bremsstrahlung in the lkr ( the amount of background induced is @xmath47 ) ; ( 2 ) @xmath48 events : by convention , the inner bremsstrahlung ( ib ) part of the radiative @xmath49 process is included into @xmath1 , while the structure dependent ( sd ) @xmath49 process , which can not be accurately computed , is not .
its contribution has to be measured directly and is @xmath50 ; ( 3 ) beam muon halo with @xmath51 decays : the @xmath6 decay signature consists of a single reconstructed track , thus the background in @xmath6 samples induced by the beam halo is an issue ( amount of background induced is @xmath52 ) .
samples of reconstructed @xmath6 candidates , of the charge not present in the beam , provide measurements of the halo background . thus the separate @xmath23 and @xmath24 beam only samples allow for cross - subtraction of the halo background with precisions much better than statistical uncertainty of the measurement ; ( 4 ) minor sources of background are electrons from @xmath53 and @xmath54 decays ( contributions at the level of @xmath55 ) .
for the second phase of na62 , a highly intense ( 800 mhz rate ) 75 gev / c positively charged particle beam will be produced by directing a 400 gev / c primary sps proton beam onto a be target .
the beam is positron free and is composed by 6% of @xmath23 ( unseparated beam ) , which will be tagged by the existing differential cerenkov counter called cedar @xcite .
the average rate seen by the downstream detectors will be 11 mhz , due to kaon decays and accidentals coming from the beam - line .
the beam spectrometer , measuring @xmath56 and the direction of the kaon track inside the decay volume , consists of three stations of a hybrid si pixel detector . a time resolution of at least 200 ps per station is required to provide a suitable tag for the kaon track .
the pion spectrometer , measuring @xmath57 and the pion direction , is made by low mass straw chambers placed in vacuum , in order to reduce multiple scattering .
a 18 m long rich located after the pion spectrometer , and filled with ne at atmospheric pressure , is the core of particle identification : it will identify pions with momentum greater than 15 gev / c and it will provide a time measurement of the downstream track with 100 ps time resolution . a hermetic combination of calorimeters , comprising the existing lkr calorimeter and covering up to 50 mrad , will be used to identify and veto the photons produced in kaon decays .
in addition , a specific detector to veto muons tracks will be used downstream .
the measurements of @xmath1 in bins of track momentum ( @xmath37 ) are found to be compatible and independent of @xmath37 , which indicates good control over the main systematic effects .
the total error is @xmath58 consisting of a statistical error of @xmath59 and systematic uncertainties of @xmath2 mainly due to background subtraction ( @xmath60 , @xmath61 and @xmath55 , respectively , due to @xmath34 , @xmath48 and beam halo background ) .
further studies of second - order effects and minor background sources are currently underway in order to finalize the measurement of @xmath1 with the currently considered partial @xmath45 data sample .
the total uncertainty of the partial result is @xmath62 ( below the @xmath63 level for the first time ) . the whole na62 data sample of @xmath64 millions @xmath4 decay candidates , which is one order of magnitude larger than the existing world sample , allows pushing the statistical uncertainty to a level below @xmath65 .
significant improvements of the @xmath34 background and beam halo systematic uncertainties are possible thanks to the special data sets collected in 2008 and dedicated to the ke2 systematics studies .
the uncertainty due to the @xmath48 ( sd ) background is expected to be decreased by a direct measurement of this process .
the ultimate precision of the measurement is expected to reach the @xmath2 level , meeting the goal declared in the proposal@xcite .
the na62 programme foresees a measurement of br(@xmath7 ) with a @xmath6610% accuracy .
the ultra - rare @xmath7 decay is a unique environment for new physics searches .
the r&d program is well advanced : during the 2007 run the collaboration tested successfully a full length rich counter prototype and a full length straw prototype in the actual vacuum tube .
0 a. masiero , p. paradisi and r. petronzio , phys . rev .
* d74 * ( 2006 ) 011701 .
v. cirigliano and i.rosell , phys . lett .
* 99 * ( 2007 ) 231801 . c. amsler
( pdg ) , phys . lett .
* b667 * ( 2008 ) 1 .
m. antonelli et al . , nucl .
b , proc . suppl .
181 - 182 ( 2008 ) 83 [ arxiv:0801.1817 ] .
na48/2 and p-326 status report , cern - spsc-2006 - 033 .
a. j. buras , m. gorbahn , u. haisch and u. nierste , phys .
lett . * 95 * , 261805 ( 2005 ) ; _ jhep _ * 0611 * , 002 ( 2006 ) .
g. isidori , f. mescia and c. smith , nucl .
b * 718 * , 319 ( 2005 ) .
f. mescia and c. smith , phys .
d * 76 * , 034017 ( 2007 ) .
s. adler _ et al . _ [ e787/949 collaboration ] , phys .
. lett .
* 101 * 191802 ( 2008 ) . g. anelli
_ , cern - spsc-2005 - 013 , spsc - p-326 . g. bovet _ et al .
_ , cern report : cern 82 - 12 ( 1982 ) . | the ratio @xmath0 of leptonic decay rates provides a golden probe for testing the structure of the weak interactions because it can be predicted with high accuracy within the standard model .
the aim of the na62 experimental programme based on the 2007 data set is a measurement of @xmath1 reaching a new accuracy level better than @xmath2 @xcite . to achieve this goal ,
data taking strategy allowing control over the systematic effects and in particular precise background subtraction , was worked out , and a data sample of @xmath3 @xmath4 candidates with just @xmath5 background was collected .
the current status of the @xmath6 analysis based on the dedicated na62 data taking is summarized .
the achieved precision of background subtraction , other systematic uncertainties , and prospects of the analysis are discussed . within the same scientific programme , na62 , in its second phase , will measure the branching ratio of the very rare kaon decay @xmath7 ; the aim is to collect @xmath8 events with 10% background in two years data taking period . the status of the project , the r&d and future perspectives for the experiment are discussed . |
the formation process of massive stars is still a largely unsolved problem .
although it is believed that stars generally originate from the collapse and subsequent accretion of clumps within molecular clouds ( palla & stahler @xcite ) , this model can not explain the formation of stars beyond 10 ( bonnell et al .
the strong radiation pressure of massive stars can halt the infall of matter limiting the mass of the star ( yorke & krgel @xcite , wolfire & cassinelli @xcite , beech & mitalas @xcite ) , while a large fraction of the infalling material may as well be deflected into bipolar outflows by processes which we do not yet know in detail ( churchwell @xcite ) .
moreover , since the evolutionary time scales of massive stars are comparatively short , these stars are believed to enter the main sequence while still embedded in their parent molecular clouds ( yorke & krgel @xcite , shu et al .
@xcite , palla & stahler @xcite , beech & mitalas @xcite , bernasconi & maeder @xcite ) .
this means that massive stars may already experience significant mass loss and subsequent evolution while still accreting mass from the parental cloud . in order to understand the formation of massive stars it is therefore necessary to study them at the earliest phases where they can be reached through the enshrouding material at different wavelengths . while high - resolution radio continuum observations allow the investigation of ultracompact 2 regions ( churchwell @xcite ) formed around newborn massive stars , high angular resolution observations in ultraviolet , visible , and infrared wavelengths ( walborn & fitzpatrick @xcite , walborn et al .
@xcite , schaerer & de koter @xcite , hanson et al .
@xcite ) are necessary to access accurate physical parameters of these stars and then evaluate their states of evolution .
our search for the youngest massive stars in the magellanic clouds started almost two decades ago on the basis of ground - based observations .
this led to the discovery of a distinct and very rare class of 2 regions , that we called high - excitation compact 2 `` blobs '' ( hebs ) .
the blob in n159 , which is the subject of this paper , was the prototype of this category of nebulae ( heydari - malayeri & testor 1982 ) .
so far only four other hebs have been found in the lmc : n160a1 , n160a2 , n83b-1 , and n11a ( heydari - malayeri & testor 1983 , 1985 , 1986 , heydari - malayeri et al.1990 ) and two more in the smc : n88a and n81 ( testor & pakull 1985 , heydari - malayeri et al .
1988 ) . to further improve our understanding of those compact 2 regions and overcome the difficulties related to their small size
, we used the superior resolving power of _ hst _ to image n81 and n88a , in the smc , as well as n159 - 5 in the lmc .
the analysis and discussion for the first two objects was presented by heydari - malayeri et al .
( @xcite,@xcite , hereafter papers i and ii respectively ) . in the present paper
we study our third _ hst _ target , the lmc blob n159 - 5 .
this object lies in the 2 complex n159 ( henize @xcite ) , situated some 30 ( 500 pc ) south of 30 dor .
n159 is associated with one of the most important concentrations of molecular gas in the lmc ( johansson et al . @xcite and references therein ) and contains several signposts of ongoing star formation ( cocoon stars , ir sources , masers ) .
n159 - 5 is the name given by heydari - malayeri & testor ( @xcite ) to a compact 2 region of size 6 ( 1.5 pc ) with high excitation ( /=8 ) and suffering a considerable extinction of @xmath2=5 mag as derived from and radio continuum ( heydari - malayeri & testor @xcite ) .
they also showed that the chemical composition of the object is compatible with that of typical lmc 2 regions .
israel & koornneef ( @xcite ) detected near - ir molecular hydrogen emission towards the object , partly shocked and partly radiatively excited .
they also confirmed the high extinction of the object from a comparison of br@xmath3 and and estimated that n159 - 5 contributes 25% to the total flux of the iras source lmc1518 ( israel & koornneef @xcite ) .
more recently , comern & claes ( @xcite ) used isocam to obtain an image of n159 - 5 at 15@xmath4 m .
similarly , hunt & whiteoak ( @xcite ) used the australia telescope compact array ( atca ) to obtain the highest angular resolution radio continuum observations in existence of n159 - 5 ( beam 8.37.4 ) .
however , none of these observations were able to resolve n159 - 5 .
the observations of n159 - 5 described in this paper were obtained with the wide field planetary camera ( wfpc2 ) on board the _ hst _ on september 5 , 1998 as part of the project go6535 .
we used several wide- and narrow - band filters ( f300w , f467 m , f410 m , f547 m , f469n , f487n , f502n , f656n , f814w ) to image the stellar population as well as the ionized gas .
the observational techniques , exposure times , and reduction procedures are similar to those explained in detail in paper i.
in fig.[mosaic ] we present the wfpc2 image of the eastern part of the giant 2 region n159 .
this image reveals a very turbulent medium in which ionized subarcsecond structures are interwoven with fine absorption features .
a large number of filaments , arcs , ridges , and fronts are clearly visible . in the south - western part of fig.[mosaic ] , we note a relatively large , high excitation ridge bordering a remarkable cavity 25 ( 6pc ) in size .
another conspicuous cavity lies in the northern part of the image .
these are most probably created by strong winds of massive stars . moreover , a salient , dark gulf running westward into n159 cuts the glowing gas in that direction and as it advances takes a filamentary appearance . a comparison with the co map of johansson et al .
( @xcite ) indicates that this absorption is due to the molecular cloud n159-e .
the 2 blob n159 - 5 stands out as a prominent high excitation compact nebula in the center of the wfpc2 field ( fig.[mosaic ] ) , at the edge of two distinct absorption lanes of size 313 ( 0.83.3 pc ) .
the most important result of our wfpc2 observations is shown in the inset of fig.[mosaic ] , namely the n159 - 5 blob resolved for the first time .
in fact n159 - 5 consists of two distinct ionized components separated by a low brightness zone the eastern border of which has a sharp front .
the overall shape of n159 - 5 is reminiscent of a butterfly or _
papillon _ in french .
the centers of the two wings are 2.3 ( 0.6 pc ) apart .
the brightest part of the right wing appears as a `` smoke ring '' or a doughnut with a projected radius of 0.6 ( 0.14 pc ) .
the left wing is characterized by a very bright `` globule '' of radius 0.4 ( 0.1 pc ) to which are linked several bright stripes all parallel and directed towards the central sharp front .
an obvious questions is : where is ( are ) the ionizing star(s ) of n159 - 5 ?
no conspicuous stars can be detected within the papillon itself , although its overall high excitation and morphology require the source of ionization to be very close to the center of this structure .
a faint star of @xmath5=17.9 mag can be seen between the two wings ( fig.[mosaic ] , inset ) and may well be the major source of ionization , heavily obscured by foreground dust .
at least three more stars weaker than @xmath520 mag are detected ( not visible in fig.[mosaic ] ) , two of them lying in the brightest parts of the smoke ring and one towards the other wing , east of the front .
no star is detected towards the globule .
the _ hst _ observations allow us to study the spatial variation of the extinction in the direction of the papillon nebula .
the / ratio is high in both wings ( fig.[rapport]a ) , varying between 5 and 10 , corresponding to a visual extinction @xmath2 between 1.5 and 3.5 mag , if the lmc interstellar reddening law is used ( prvot et al .
the extinction towards the zone separating the two wings also shows comparable ratios .
the / map was used to de - redden the flux on a pixel to pixel basis .
the 5007/ map also displays the butterfly - like structure ( fig.[rapport]b ) with line ratios varying between 3.5 and 8 .
the band separating the wings has overall smaller values .
a comparison of the and images shows that the papillon has the same size and morphology in both filters .
this suggests a hard radiation field in which the high excitation o@xmath6 ions occupy the same zone as the ionized hydrogen . a simple calculation ( paper ii )
shows that almost all oxygen atoms are doubly ionized .
we measure a total flux f ( ) = 2.6810@xmath7 erg 2 above 3@xmath8 level for both wings ( accurate to 3% ) . correcting for the extinction ( sect . 3.2 ) gives @xmath9=5.3510@xmath10 erg 2 .
the flux is not equally distributed between the two wings since 60% is generated by the eastern wing ( globule ) , whereas the western wing ( smoke ring ) contributes by 40% .
a lyman continuum flux of @xmath11=4.1710@xmath12 photons can be derived for the whole n159 - 5 taking @xmath13=10500k ( heydari - malayeri & testor @xcite ) , assuming a distance of 55 kpc , and considering that the 2 region is ionization - bounded .
a single o8v star can account for this ionizing flux ( vacca et al .
@xcite , schaerer & de koter @xcite ) .
however , this should be viewed as a lower limit , since the region is probably not ionization - bounded .
the three magellanic cloud blobs studied so far with _ hst _ ( n81 , n88a , and n159 - 5 ) represent very young massive stars leaving their natal molecular clouds .
while n81 is a rather isolated starburst , n88a and n159 - 5 are formed in richer regions of gas where a preceding generation of massive stars has taken place .
being very young , these two regions are also similar in that they are heavily affected by dust . in spite of these similarities ,
n159 - 5 has a bewildering morphology which is seen neither in the other blobs , nor , generally speaking , in any of the known magellanic cloud 2 regions .
the global morphology of the papillon and more especially the presence of peculiar fine structure features in the wings make this object a unique 2 region in the lmc .
it is also the first bipolar nebula indicating young massive star formation in an outer galaxy . in this respect
the papillon may even represent a new type of very young 2 region in the magellanic clouds overlooked so far because of insufficient spatial resolution . even though high resolution spectroscopy would be necessary for a conclusive picture of the papillon s nature , the parallel ray - like features of the globule as well as the smoke ring strongly suggest a dynamical origin .
several observational facts also advocate a common process for the formation of these two compact features : their proximity ( 2.3 ) , their location in a distinct , more diffuse nebular structure , and the uniqueness of the phenomenon in the large field of the n159 complex . in order to explain the observed morphology of n159 - 5 , two different models can be envisaged .
\1 ) we are looking at a bipolar region produced by strong stellar wind of hot star(s ) hidden behind the central absorption zone .
it has been shown that in hot , rotating stars the mass loss rate is much larger at the poles than at the equator ( maeder 1999 ) .
this model can account for the high excitation of the two wings , as well as the global bipolar morphology .
although we can not yet firmly advocate a bipolar phenomenon , we underline the morphological similarity between the papillon and the high resolution image of kleinmann - low nebula in orion recently obtained with the subaru 8.3 m telescope at 2.12@xmath4 m ( subaru team 1999 ) .
this image shows a butterfly - shaped exploding area produced by the wind of a young cluster of stars , among which irc2 , a particularly active star estimated to have a mass over 30 .
however , this model does not explain the smoke ring neither the parallel stripes .
one may also compare n159 - 5 with the well - studied galactic 2 region sh106 which has a prominent bipolar shape , marvellously shown in an _ hst _ image ( bally et al .
@xcite and references therein ) .
its exciting star is hidden in the absorption region between the two lobes where extinction amounts to 20 mag in the visual .
the southern lobe is much brighter than the northern one because it is blueshifted , while the other is expanding away from the observer . also , both lobes get gradually fainter in their external parts
. however , we do not see such global trends in n159 - 5 and , more importantly , no smoke ring or globule features are present in sh106 .
\2 ) alternatively , n159 - 5 may represent two close but distinct nebulae each with its own massive star providing a high excitation for the ionized gas . in this possibility
, the globule may be a bow shock created by an o star with a powerful stellar wind moving at speeds of 10 km through the molecular cloud ( models of van buren & mac low @xcite and references therein ) .
the presence of the parallel rays and the sharp front make this explanation attractive .
since no stars are detected towards the globule , we may speculate that the star is hidden behind the bow shock / globule that is heading towards the observer .
as for the smoke ring , it may be due to the interaction of a stellar wind from a central star with the surrounding medium creating a bubble structure . if this is the case , the mass loss should be important , comparable to that observed in luminous blue variables .
however , lbvs are evolved massive stars and it is not clear how this phenomenon can occur in such a young region .
another explanation for the smoke ring can be provided by the mentioned bow shock models .
they predict the formation of a stellar wind bubble when the star s motion is subsonic .
the smoke ring can therefore be due to such a slower moving star .
if this picture is correct , we are witnessing a very turbulent star forming site where massive stars formed in group are leaving their birthplace .
one should stress the noteworthy absence of prominent stars towards the papillon . as in the case of the smc n88a ( paper ii ) , this is certainly due to the very young age and the high dust content of this star formation region .
n159 - 5 is just hatching from its natal molecular cloud , and its exciting stars should therefore be buried inside dust / gas concentrations . in order for a star of type o8 with @xmath14=4.66 mag ( vacca et al .
@xcite ) to remain undetected in our strmgren @xmath5 image , we need extinctions larger than @xmath2=6 mag , which is quite possible given the above estimates
. however , n159 - 5 will gradually evolve into a more extended , less dense region exhibiting its exciting stars , like the smc n81 ( paper i ) .
we are grateful to dr .
j. lequeux for a critical reading of the manuscript .
we would like also to thank an unknown referee for suggestions that improved the manuscript .
we are also indebted to dr .
johansson for providing us with the co map of the n159 molecular cloud .
vc would like to acknowledge the financial support from a marie curie fellowship ( tmr grant erbfmbict960967 ) .
bally j. , chu yu k. , rayner j. , zinnecker h. , 1998 , aj 116 , 1868 beech m. , mitalas r. , 1994 , apjs 95 , 517 bernasconi p.a . , maeder a. , 1996 , a&a 307 , 829 bonnell i.a .
, bate m.r . , zinnecker h. , 1998 , mnras 298 , 93 churchwell e. , 1990 , a&a rev . 2 , 79 churchwell e. , 1997 , apj 479 , l59 comern f. , claes p. , 1998
, a&a 335 , l13p hanson m.m . , conti p.s . , rieke m.j
, 1996 , apjs 107 , 281 henize k.g . , 1956 , apjs 2 , 315 heydari - malayeri m. , testor g. , 1982 , a&a 111 , l11 heydari - malayeri m. , testor g. , 1983 , a&a 118 , 116 heydari - malayeri m. , testor g. , 1985 , a&a 144 , 98 heydari - malayeri m. , testor g. , 1986 , a&a 162 , 180 heydari - malayeri m. , le bertre t. , magain p. , 1988
, a&a 195 , 230 heydari - malayeri m. , van drom e. , leisy p. , 1990
, a&a 240 , 481 heydari - malayeri m. , rosa m.r . , zinnecker h. , deharveng l. , charmandaris v. , 1999a , a&a 344 , 848 ( paper i ) heydari - malayeri m. , charmandaris v. , deharveng l. , rosa m.r .
, zinnecker h. , 1999b , a&a 347 , 841 ( paper ii ) hunt m.r .
, whiteoak j.b . , 1994 ,
proc . astr .
11 , 68 israel f. p. , koornneef j. , 1988 , a&a 190 , 21 israel f. p. , koornneef j. , 1991 , a&a 248 , 404 johansson l.b.e . , greve a. , booth r.s .
et al . , 1998 , a&a 331 , 857 maeder a. , 1999 , a&a 347 , 185 palla f. , stahler s.w .
, 1990 , apj l47 palla f. , stahler s.w . , 1993 ,
apj 418 , 414 prvot m.l .
, lequeux j. , maurice e. , prvot l. , rocca - volmerange b. , 1984 , a&a 132 , 389 schaerer d. , de koter a. , 1997 , a&a 322 , 598 shu f.r . , adams f.c . , lizano s. , 1987 , ara&a 25 , 23 subaru team , 1999 , pr990128 , http://www.subaru.naoj.org testor , g. pakull m. , 1985 , a&a 145 , 170 vacca w.d .
, garmany c.d . ,
shull j.m . , 1996 , apj 460 , 914 van buren d. , mac low m .-
, 1992 , apj 394 , 534 walborn n.r .
, fitzpatrick e.l . , 1990 ,
pasp 102 , 379 walborn n.r .
, lennon j.l . , haser s.m . , et al . , 1995 ,
pasp 107 , 104 wolfire m.g .
, cassinelli j.p . , 1987 , apj 319 , 850 yorke h.w . , krgel e. , 1977 , a&a 54 , 183 | we present high spatial resolution _ hst _ imaging of the lmc compact 2 region n159 - 5 .
this high excitation blob is revealed to be a _
`` papillon '' _ or butterfly - shaped ionized nebula with the `` wings '' separated by 2.3 ( 0.6 pc ) .
two subarcsecond features resembling a `` smoke ring '' and a `` globule '' are detected in the wings , the origin of which is briefly discussed .
n159 - 5 may represent a new type of 2 region in the magellanic clouds overlooked so far because of insufficient spatial resolution .
our images also show a strikingly turbulent medium around the _ papillon _ in the giant 2 region n159 , which manifests itself by a large number of subarcsecond filaments , arcs , ridges , and fronts carved in the ionized gas by the stellar winds from massive stars in the n159 complex .
2hii _
@xmath0 2@xmath1 _ |
since the pioneering work of lippmann @xcite and pellat @xcite on the influence of electrostatic potentials on the wetting of substrates by fluids , electrowetting has been simultaneously studied to address fundamental issues of surface science , e.g. , electrocapillarity @xcite , the structure of solid - fluid interfaces @xcite , or the characterization of surface states @xcite , as well as to develop novel applications , e.g. , driving , mixing , or shaping of droplets in lab - on - a - chip devices , optical applications , or microelectromechanical systems @xcite . in the past electrowetting at low voltages
was commonly interpreted as an electrocapillarity effect , i.e. , it is assumed to hinge on the voltage - dependence of the substrate - fluid interfacial tension @xcite . a justification for this approach is frequently given in terms of the vast experimental evidence for systems of uncoated and hydrophobically coated electrodes .
the present work reports on an effort to understand electrowetting in the context of general wetting phenomena @xcite . within classical microscopic
density functional theory one has access to the interfacial structure of fluids so that one can study , e.g. , the contact angle as a function of the electrostatic substrate potential .
however , it turned out that the commonly given derivations of the electrowetting equation @xcite are incorrect in that interfacial properties , e.g. , interfacial tensions or differential capacitances , which describe the contact of _ one _ fluid with a substrate , enter the equation describing the contact of _ two _ fluids with a substrate .
it has been overlooked that the interfacial structure , and thus interfacial quantities , of a fluid can change upon bringing it into contact with another fluid .
although the interfacial structure close of fluids to substrates has been deeply examined in the surface science literature @xcite , its properties seem to be largely ignored in the context of electrowetting so far . by ignoring structural differences which occur at substrate - fluid interfaces upon bringing two fluids in simultaneous contact with a substrate , one can interpret electrowetting as a consequence of voltage - dependent interfacial tensions , which is referred to as the electrocapillarity approach to electrowetting in the following .
it is shown in this work that electrowetting _ can not _ be consistently understood as an electrocapillarity effect .
alternative approaches to interpret electrowetting as a line tension effect have been proposed @xcite but some of the predictions were in disagreement with experimental data @xcite .
the present approach is to study electrowetting in terms of the the effective interface potential @xcite , which is related to the macroscopic contact angle .
the effective interface potential has been analyzed recently for simple models of electrolyte solutions @xcite , showing the general feature of ions inducing wetting transitions of first order .
whereas wetting transitions are concerned with the _ thickness _ of wetting films , the contact angle is related to the _ depth _ of the effective interface potential , which always vanishes continuously at wetting transitions @xcite . in the following the effective interface potential for pellat s classical setup of a vertical parallel plate capacitor @xcite
is determined and used to derive an electrowetting equation ( sec .
[ sec : theory ] ) based on the density functional theory of wetting .
this setup has been chosen because its geometry is precisely defined , an issue which has been recently raised in a critical discussion of the more common setup of a droplet on a substrate with the counter electrode being a thin wire @xcite .
the misconception underlying the classical derivation of the electrowetting equation within the electrocapillarity approach is discussed ( sec .
[ sec : discussion ] ) .
interestingly , the electrocapillarity - based electrowetting equation seems to be a good approximation for systems investigated up to now , i.e. , uncoated or hydrophobically coated electrodes ( sec .
[ sec : discussion ] ) .
however , it is argued that the difference in the predicted electrowetting numbers between the electrocapillarity approach and the present one based on the density function theory of wetting can be expected to occur , e.g. , for hydrophilically coated electrodes or for two immiscible electrolyte solutions as fluids ( sec .
[ sec : discussion ] ) . in view of the conceptional problems of the electrocapillarity approach it
is suggested to rather interpret electrowetting in terms of the density functional theory of wetting ( sec .
[ sec : conclusions ] ) .
moreover , the possibility to obtain microscopic information about solid - fluid interfaces by analyzing electrowetting measurements in terms of the density functional theory of wetting deserves further consideration .
in a vertical parallel plate capacitor of width @xmath0 initially filled with a fluid @xmath1 .
the meniscus height @xmath2 is related to the contact angles @xmath3 and @xmath4 by eq . .
electrowetting corresponds to a dependence of @xmath3 and @xmath4 , and hence of @xmath2 , on the voltage @xmath5 between the plate electrodes .
a closeup of the three - phase contact region marked by the dashed box is depicted in fig .
[ fig:2 ] . ]
consider pellat s classical setup @xcite depicted in fig .
[ fig:1 ] .
a vertical parallel plate capacitor of width @xmath0 is in contact with two immiscible fluids @xmath6 and @xmath1 of mass densities @xmath7 and @xmath8 , respectively .
at least one of the fluids @xmath6 and @xmath1 is assumed to be an electrolyte solution .
it is further assumed that @xmath9 so that both fluids are separated in the gravitational field with @xmath6 being the lower and @xmath1 being the upper phase ( see fig .
[ fig:1 ] ) . provided the capacitor width @xmath0 is smaller than the capillary length @xcite @xmath10 with the @xmath6-@xmath1 interfacial tension @xmath11 and the acceleration due to gravity @xmath12 , the contact angles @xmath3 and @xmath4 of phase @xmath6 are related to the meniscus height @xmath2 by @xcite @xmath13 depending on the interactions of the fluids @xmath14 and the substrates @xmath15 , which are metal electrodes ( represented by the hatched parts in fig .
[ fig:1 ] ) possibly coated with some dielectric ( represented by yellow layers on top of the electrodes in fig .
[ fig:1 ] ) , the respective contact angles @xmath3 and @xmath4 can be smaller or larger than @xmath16 , which corresponds to positive or negative contributions to the meniscus height @xmath2 .
electrowetting can be detected as the dependence of the contact angles @xmath17 and @xmath18 , and in turn , via eq .
, of the meniscus height @xmath19 , on the electrostatic potential difference @xmath20 applied between the electrodes .
( and similarly for the the cathodic substrate @xmath21 ) and two immiscible fluids @xmath6 and @xmath1 marked by the dashed box in fig . [ fig:1 ] .
the fluid which is preferred by the substrates @xmath22 is denoted by @xmath23 , whereas the other , less preferred fluid is called @xmath24 .
panel ( a ) corresponds to the case of an @xmath6-philic ( @xmath25 ) substrate @xmath26 , while panel ( b ) displays the case of an @xmath1-philic ( @xmath27 ) substrate @xmath26 .
here it is assumed that both substrates @xmath26 and @xmath21 prefer the same fluid . the macroscopic contact angle @xmath3 ( @xmath4 ) shown in fig .
[ fig:1 ] describes the asymptotic inclination of the @xmath6-@xmath1 interface far away from the substrate @xmath26 ( @xmath21 ) , whereas close to the substrate @xmath26 ( @xmath21 ) a film of microscopic thickness @xmath28 ( @xmath29 ) of the preferred fluid @xmath23 is formed . ]
the contact angles @xmath3 and @xmath4 in fig .
[ fig:1 ] provide a macroscopic description of the fluid - fluid - substrate three - phase contact region ( highlighted by the dashed box in fig . [ fig:1 ] for the anodic substrate @xmath26 ) . according to the chemical properties of the fluids and the substrates , the contact of substrate @xmath22 with one fluid , henceforth denoted by fluid @xmath23 , is more preferable than with the other fluid , henceforth denoted by fluid @xmath24 .
here it is assumed that both substrates are chemically equal such that either fluid @xmath6 or fluid @xmath1 is preferred by both substrates @xmath26 and @xmath21 .
consequently , if substrate @xmath22 is macroscopically in contact with the bulk of the less preferred fluid @xmath24 and if the thermodynamic state is away from wetting transitions such that the substrate is only partially wet by phase @xmath23 , a film of microscopic extension @xmath30 composed of the preferred fluid @xmath23 forms in between substrate @xmath22 and the bulk of fluid @xmath24 @xcite , where @xmath31 denotes the bulk correlation length , which is of the order of the particle size if the thermodynamic state is away from critical points .
then the fluid structure is similar to that of a composition of an @xmath22-@xmath23 interface at the substrate surface and a free @xmath23-@xmath24 interface at a distance @xmath32 away from the substrate , both being of typical extension @xmath31 @xcite .
this structure leads to a surface contribution @xmath33 to the grand potential of the system , where @xmath34 and @xmath11 denote the @xmath22-@xmath23 and @xmath6-@xmath1 interfacial tensions , respectively , and where @xmath35 is the effective interface potential @xcite .
it is important to distinguish @xmath36 from the interfacial tension @xmath37 of an @xmath22-@xmath24 interface in the _ absence _ of phase @xmath23 .
here @xmath38 because the presence of the preferred phase @xmath23 leads to a structural change , i.e. , the formation of @xmath23-films , as compared to the situation in the absence of phase @xmath23 .
ignoring the difference between @xmath36 and @xmath37 is equivalent to ignoring the formation of @xmath23-films and it is this crucial misconception which underlies the electrocapillarity approach to electrowetting . in contrast , if substrate @xmath22 is in contact with the bulk of the preferred fluid @xmath23 , the fluid is non - uniform only close to the substrate surfaces up to distances @xmath31 , and this interfacial structure is not modified by the presence of fluid @xmath24 , hence @xmath39 . depending on whether the preferred fluid @xmath23 is fluid @xmath6 or fluid @xmath1
the substrates @xmath22 are referred to as @xmath6-philic or @xmath1-philic , respectively . a closeup of the fluid - fluid - substrate three - phase contact region close to substrate @xmath26 marked by the dashed box in fig .
[ fig:1 ] is sketched respectively in figs .
[ fig:2](a ) and ( b ) for an @xmath6-philic ( @xmath25 ) and an @xmath1-philic ( @xmath27 ) substrate .
the macroscopic contact angle @xmath40 ( see fig . [ fig:2 ] ) is related to the surface contributions @xmath41 , @xmath42 and the interfacial tension @xmath11 of the @xmath22-@xmath6 , @xmath22-@xmath1 , and @xmath6-@xmath1 interface , respectively , by young s equation @xcite @xmath43 it is common to assume @xmath44 , but this misconception to ignore the structural differences of a macroscopic @xmath22-@xmath45 contact in the presence and in the absence of additional phases can have significant consequences .
the surface contributions @xmath46 and @xmath42 are related to the depth of the effective interface potential @xmath47 evaluated at the equilibrium film thickness @xmath48 by @xcite @xmath49 for @xmath6-philic substrates @xmath22 ( see above the three - phase contact region in fig .
[ fig:2](a ) ) and by @xmath50 for @xmath1-philic substrates @xmath22 ( see below the three - phase contact region in fig . [ fig:2](b ) ) . hence
, one obtains @xcite @xmath51 where @xmath52 for @xmath6-philic and @xmath53 for @xmath1-philic substrates @xmath22 .
this equation connects the macroscopic contact angle @xmath40 with the microscopic structure represented by the effective interface potential @xmath54 of @xmath23-films at substrate @xmath22 in macroscopic contact with bulk fluid @xmath24 . the next sec .
[ subsec : dft ] describes an approximate calculation of the effective interface potentials @xmath54 for the setting of fig .
[ fig:1 ] . the dependence of @xmath55 on the electrostatic potential difference @xmath5 between the electrodes , together with eq .
, then leads to the electrowetting equations derived in sec .
[ subsec : elweteq ] .
however , already without explicit expressions for the effective interface potentials , one can draw an important conceptual conclusion from eq .
: electrowetting is _ not _ an electrocapillarity effect , since _ no _ @xmath5-dependent substrate - fluid interfacial tensions , which describe the contact of the substrate with _
one _ fluid , occur on the right - hand side . instead
, electrowetting is related to the depth of the effective interface potential , which describes the @xmath5-dependence of the microscopic fluid structure close to the substrate in the presence of _ two _ fluids .
( a ) ) or below ( fig . [ fig:2](b ) ) the three - phase contact regions ( see fig .
[ fig:1 ] ) the dielectric substrates @xmath26 and @xmath21 of thickness @xmath56 and distance @xmath0 are covered by films of the preferred fluid @xmath23 of microscopic thicknesses @xmath32 , which separate the substrates from the bulk of the less preferred fluid @xmath24 .
( b ) since the separation @xmath0 between the substrates @xmath26 and @xmath21 is typically the largest length scale , one can consider the limit @xmath57 , which renders the effective interface potential @xmath54 at substrate @xmath22 that of a semi - infinite system .
a coordinate axis normal to each substrate is introduced with the origin @xmath58 at the surface and the fluids in the range @xmath59 .
the interface between the @xmath23-film and the bulk fluid @xmath24 is located at @xmath60 and the electrode is at position @xmath61 , where the electrostatic potential is @xmath62 and the surface charge density is @xmath63 . ] in order to obtain the effective interface potential @xmath54 of an @xmath23-film of thickness @xmath64 at substrate @xmath22 in fig .
[ fig:1 ] , whose value for the equilibrium film thicknesses @xmath48 is related to the contact angle @xmath40 via eq . , one may represent the structure in fig .
[ fig:1 ] far above ( for an @xmath6-philic substrate @xmath22 , see fig .
[ fig:2](a ) ) or below ( for an @xmath1-philic substrate @xmath22 , see fig . [ fig:2](b ) ) the three - phase contact region by the quasi - one - dimensional slab depicted in fig .
[ fig:3](a ) .
the chemically identical substrates @xmath26 and @xmath21 , which comprise metal electrodes coated with dielectric layers of thickness @xmath56 , are separated by a distance @xmath0 and covered with films of thicknesses @xmath28 and @xmath29 , respectively , of the preferred fluid @xmath23 .
moreover , the electrodes are assumed to be ideally polarized , i.e. , electrochemical reactions do not occur . even under these conditions
the film thicknesses @xmath28 and @xmath29 can differ , if unequal partitioning of ions at the film - bulk fluid interfaces takes place .
this is expected to occur in general due to generic differences in solubility contrasts @xcite .
the macroscopic distance @xmath0 between the substrates is typically the largest length scale such that only the limit @xmath57 is considered in the following .
hence the effective interface potentials @xmath47 and @xmath65 at the substrates @xmath26 and @xmath21 , respectively , are those of semi - infinite systems . for each substrate
@xmath22 a coordinate axis in normal direction with the origin @xmath58 at the substrate surface and the fluid at @xmath59 is introduced ( see fig .
[ fig:3](b ) ) .
the interface between the @xmath23-film and the bulk of phase @xmath24 is located at position @xmath60 and the electrode is at @xmath61 , where the electrostatic potential is @xmath62 and the surface charge density is @xmath63 . considering the two bulk phases of fluids @xmath6 and @xmath1 outside the capacitor in fig . [ fig:1 ] as particle reservoirs , one is naturally led to a grand - canonical description of the thermodynamic state . a starting point for the derivation of the effective interface potential @xmath54 is the grand potential functional per thermal energy @xmath66 and per area @xmath67 of the electrode @xmath68}{\mathcal{a } } & = \frac{\beta\omega_0[\phi]}{\mathcal{a } } + \frac{\beta dd(0)^2}{2\epsilon_\text{vac}\epsilon_s } - \beta \psi_\pm d(0 ) \notag\\ & + \int\limits_0^\infty\!\!\mathrm{d } z\bigg[\sum_{i=\pm}\rho_i(z)\ !
\bigg(\!\ln\frac{\rho_i(z)}{\zeta_i}-1+\beta v_i(\phi(z))\!\bigg ) \notag\\ & \phantom{+\int\limits_0^\infty\!\!\mathrm{d } z\bigg[}+ \frac{\beta d(z)^2}{2\epsilon_\text{vac}\epsilon(\phi(z))}\bigg ] \label{eq : omega1}\end{aligned}\ ] ] in terms of the solvent composition profile @xmath69 and the @xmath70-ion number density profiles @xmath71 . here
the permittivity @xmath72 of the vacuum , the relative permittivity @xmath73 of the substrate @xmath22 , and the fugacities @xmath74 of @xmath70-ions are used .
the density functional @xmath75 $ ] describes the grand potential of the pure , i.e. , salt - free fluids .
the electric displacement @xmath76 is determined by gau s law and the boundary condition of global charge neutrality : @xmath77 here @xmath78 denotes the valency of @xmath70-ions and @xmath79 is the elementary charge .
since the substrate @xmath22 is free of ions the electrode surface charge density is given by @xmath80 ( see fig . [ fig:3](b ) ) . in the absence of specific adsorption of ions at the substrate surface ,
the electric displacement is continuous at @xmath58 , i.e. , @xmath81 , so that @xmath82 .
the electric displacement @xmath76 is related to the electrostatic potential @xmath83 by @xmath84 .
the electrostatic potential of the electrode is given by @xmath85 ( see fig .
[ fig:3](b ) ) . the second term on the right - hand side and the term in the last line of eq .
account for the electrostatic field energy inside the substrate and the fluids , respectively , whereas the last term in the first line represents the internal energy of the voltage source sustaining the electrostatic potential @xmath86 of the electrode ( see fig .
[ fig:3](b ) ) .
the expressions @xmath87 and @xmath88 denote respectively the local solvation free energy of a @xmath70-ion and the relative permittivity at a position @xmath59 where the solvent composition is given by @xmath89 @xcite . finally , the second line of eq
. describes the grand potential of the ions , where the ion number densities are assumed to be sufficiently small such that ions interact with each other only via the electrostatic field . the equilibrium bulk state @xmath90 with the bulk solvent composition @xmath91 and the bulk ionic strength @xmath92 minimizes the bulk grand potential density @xmath93 , which can be derived from eq . by inserting uniform profiles @xmath94 and @xmath95 , omitting all surface terms , and noting @xmath96 in the bulk . the two immiscible fluids @xmath23 and @xmath24 at coexistence in fig .
[ fig:3](a ) correspond to two equilibrium bulk states @xmath97 and @xmath98 , respectively , with equal bulk grand potential density : @xmath99 .
since the present analysis is concerned with films of thicknesses @xmath30 but not with interfacial structures on length scales less than @xmath31 it is natural to approximate the solvent composition profile @xmath69 in eq . within the so - called sharp - kink approximation @xcite @xmath100 furthermore , in view of the small ionic strengths @xmath101 and @xmath102 to be considered here and for sufficiently small surface potential @xmath103 , the term in the second line of eq
. can be expanded up to quadratic order in the ion number density deviations @xmath104 with @xmath105 for @xmath106 and @xmath107 for @xmath108 if @xmath109 .
this leads to an approximate grand potential functional @xmath110)}{\mathcal{a } } = \frac{\beta\omega_0[\phi_\ell]}{\mathcal{a } } + \frac{\beta dd(0)^2}{2\epsilon_\text{vac}\epsilon_s } - \beta \psi_\pm d(0 ) \notag\\ & \phantom{mmm}+\int\limits_0^\ell\!\!\mathrm{d } z \bigg[f_a(\delta\rho_+(z),\delta\rho_-(z ) ) + \frac{\beta d(z)^2}{2\epsilon_\text{vac}\epsilon_a}\bigg ] \notag\\ & \phantom{mmm}+\int\limits_\ell^\infty\!\!\mathrm{d } z \bigg[f_b(\delta\rho_+(z),\delta\rho_-(z ) ) + \frac{\beta d(z)^2}{2\epsilon_\text{vac}\epsilon_b}\bigg ] \label{eq : omega}\end{aligned}\ ] ] with @xmath111 , & i_\alpha>0 \\ 0 , & i_\alpha = 0 .
\end{cases } \notag\end{aligned}\ ] ] by minimizing @xmath112)/\mathcal{a}$ ] in eq . with respect to the profiles
@xmath113 one obtains the equilibrium profiles @xmath114 . inserting @xmath114 into eq . and
subtracting the bulk contribution @xmath115 leads to the surface contribution to the grand potential @xcite @xmath116 ) - \omega_b(\bar\phi_b , i_b)}{\mathcal{a}}. \label{eq : omegas}\end{aligned}\ ] ] finally , the effective interface potential at substrate @xmath22 is given by @xmath117 @xcite , which , in the present context , can be written in the form @xmath118 here @xmath119 denotes the effective interface potential corresponding to the grand potential functional @xmath75 $ ] of the pure fluids , @xmath120 is the galvani potential of phase @xmath121 , and @xmath122 with the length scales @xmath123 , @xmath124 , @xmath125 , and @xmath126 , where @xmath127 is the square of the inverse debye length in the bulk of phase @xmath121 .
in addition , the electrode charge density can be written as @xmath128 with @xmath129 the galvani potential difference ( donnan potential ) between the phases @xmath23 and @xmath24 , @xmath130 , can be inferred from the solubility of the @xmath70-ions in the solvents @xmath23 and @xmath24 @xcite .
moreover , @xmath131 is determined by the potential difference @xmath20 and the fact that no chemical reactions take place at the electrodes so that the total charge of both electrodes vanishes : @xmath132 .
. leads to @xmath133 the equilibrium film thicknesses @xmath28 and @xmath29 are both similar in magnitude ( @xmath134 ) of the order of a few bulk correlation lengths @xmath31 away from wetting transitions . hence , eq . leads to @xmath135 , i.e. , the @xmath5-dependent part of @xmath136 is rather insensitive to a variation of @xmath134 .
moreover , the film thicknesses @xmath28 and @xmath29 are typically much smaller than the debye length @xmath137 in the @xmath23-film .
consequently , the leading @xmath5-dependent contribution @xmath138 to the effective interface potential @xmath54 in eq . , which decays exponentially with @xmath64 on the length scale of half of a debye length , @xmath139 ( see eq . ) , does not significantly shift the equilibrium film thickness @xmath32 but it merely lifts the depth @xmath55 of the effective interface potential .
therefore , in the following , the film thicknesses @xmath32 are considered as independent of the applied voltage @xmath5 . hence , eq . can be written as @xcite @xmath140 with the electrowetting number @xmath141 inserting eq . into eq
. leads to @xmath142 this equation is expected to be valid for sufficiently small voltages @xmath143 such that the quadratic approximation eq .
is applicable .
the electrowetting number @xmath144 in eq . differs from those in the literature @xcite in a number of aspects , as is discussed in the next section .
the most obvious difference is the occurrence of a correction term @xmath145 , which vanishes exactly only if @xmath146 due to a vanishing donnan potential ( galvani potential difference ) @xmath147 . for @xmath148 ,
i.e. , @xmath149 , the electrowetting number @xmath144 in eq . is not minimal at and not symmetric with respect to @xmath150 .
however , for a sufficiently large voltage @xmath143 the subleading term @xmath145 is dominated by the leading term @xmath138 . whereas the full expression for the electrowetting number @xmath144 in eq .
depends on the five possibly largely different length scales @xmath151 , @xmath152 , @xmath153 , @xmath154 , and @xmath155 , the latter two , corresponding to the thicknesses of the @xmath23-films at the substrates @xmath26 and @xmath21 , respectively , are typically of similar magnitude : @xmath134 , i.e. , @xmath156 .
this case @xmath157 is discussed in detail in the next section , for which the electrowetting number @xmath144 in eq . simplifies to @xmath158 moreover , the film thicknesses @xmath28 and @xmath29 are typically smaller than the debye lengths @xmath137 and @xmath159 so that the limiting case @xmath160 is considered throughout , within which @xmath161
before discussing the electrowetting number in eq . obtained within the present density functional analysis , the traditional approach based on the assumption of electrowetting being an electrocapillarity effect @xcite
is repeated . here only the classical method based on lippmann s equation is presented .
however , calculations using alternative methods , e.g. , based on maxwell s stress tensor @xcite , suffer from the same misconceptions .
the starting point is young s equation but with the _ incorrect _ assumption @xmath162 . in order to obtain the @xmath5-dependence of the interfacial tension @xmath163 one considers a semi - infinite fluid @xmath45 bound by a planar substrate @xmath22 . the interfacial tension @xmath163 changes upon changing the electrostatic potential @xmath164 of substrate @xmath22 with respect to that of the bulk of phase @xmath45 according to lippmann
s equation @xcite @xmath165 where @xmath166 is the surface charge density of substrate @xmath22 in contact with phase @xmath45 . describing the @xmath22-@xmath45 interface by means of the potential - independent differential capacitance @xmath167 , which is assumed to not depend on @xmath22 for chemically identical substrates , and integrating twice with respect to the electrostatic substrate potential @xmath164 using lippmann s equation leads to @xmath168 young s equation in conjunction with the assumption @xmath169 reads @xmath170 noting @xmath171 leads to @xmath172 using @xmath173 , one obtains the analog of eq . from @xmath174 and @xmath175 as @xmath176 .
this leads to the commonly used form of the electrowetting equation @xcite @xmath177 with the differential capacitances @xmath178 being those of a substrate in macroscopic contact with only one fluid phase @xmath45 .
these differential capacitances @xmath178 can typically be interpreted as those of a capacitor of capacitance @xmath179 , representing substrate @xmath22 , connected _ in series _ with a capacitor of capacitance @xmath180 , representing fluid @xmath45 : @xmath181 .
if fluid @xmath45 is an electrolyte solution the fluid capacitance is that of the electric double layer in a semi - infinite system , @xmath182 , whereas for a non - conducting dielectric fluid @xmath183 . using the length scales defined after eq .
, this leads to @xmath184 equations and represent the interpretation of electrowetting as an electrocapillarity effect @xcite .
however , the crucial misconception underlying this interpretation is to use the approximation @xmath169 and hence the differential capacitance @xmath178 , which corresponds to a semi - infinite system of a single phase @xmath45 bound by substrate @xmath22 , instead of accounting for the actual fluid structure at the substrate . the interfacial structure , and therefore surface quantities such as the surface contribution to the grand potential as well as the differential capacitance , of substrate @xmath22 in macroscopic contact with the bulk fluid @xmath24 depend significantly on whether the preferred fluid @xmath23 is present or not because an @xmath23-film forms in between the substrate @xmath22 and the bulk fluid @xmath24 in the former case whereas it does not in the latter case . in contrast
, these structural properties are naturally accounted for within the present density functional approach , which relates the contact angle to the effective interface potential ( see eq . ) , a quantity which correctly describes the contact of a substrate with _ both _
fluids @xmath23 and @xmath24 .
the early investigations of electrocapillarity by lippmann @xcite and pellat @xcite have been performed for metal electrodes without any dielectric coating . at that time for some decades mercury electrodes became the experimental standard for investigations of the electric double layer @xcite .
pure metal electrodes can be considered as substrates with thickness @xmath56 being the smallest length scale : @xmath185 . within this limit one obtains @xmath186 from eq . and subsequently from eqs .
@xmath187 for the case
@xmath188 , which is typically the case for hydrophilic substrates , an aqueous electrolyte solution @xmath189 ( i.e. , @xmath52 ) , and an oil @xmath190 , one obtains for the the electrowetting number eq .
@xmath191 hence , if the voltage @xmath143 is much larger than the donnan potential ( galvani potential difference ) @xmath192 , the electrowetting number @xmath144 agrees with that in eq . , where @xmath193 due to eq . .
similarly , for the case @xmath194 , which is typically the case for hydrophobic substrates , an oil @xmath195 ( i.e. , @xmath53 ) , and an aqueous electrolyte solution @xmath196 , one obtains for the electrowetting number eq .
@xmath197 again , if the voltage @xmath143 is much larger than the donnan potential ( galvani potential difference ) @xmath192 , the electrowetting number @xmath144 again agrees with that in eq . , where @xmath198 due to eq . .
therefore , the present formalism ( eqs .
) confirms the electrocapillarity - based form of the electrowetting number for the case of uncoated metal electrodes ( @xmath199 ) , provided the voltage @xmath143 is sufficiently large as compared to the donnan potential ( galvani potential difference ) @xmath192 .
interestingly , for uncoated metal electrodes it is irrelevant whether they are @xmath6-philic ( hydrophilic ) or @xmath1-philic ( hydrophobic ) .
however , a small voltage @xmath200 or @xmath201 , e.g. , for two immiscible electrolyte solutions , leads to electrowetting numbers @xmath202 , in contrast to @xmath203 in eq . due to @xmath204 according to eq . .
since these conditions are rather special , this scenario is not expected to be of practical relevance , but it might provide a test for the present approach . in the last few decades
most electrowetting settings used electrodes coated with an isolating dielectric for technical advantage @xcite .
almost all of these studies used drops of an aqueous electrolyte solution @xmath6 placed on a hydrophobic dielectric and an oil @xmath1 as the environmental fluid in order to achieve large contact angle ranges being covered by electrowetting @xcite . within the present notation this situation corresponds to @xmath27 ( i.e. , @xmath53 ) and @xmath205 . since the thickness @xmath64 of the microscopic oil film on the substrates @xmath22 is typically smaller than the the debye length @xmath159 of the dilute electrolyte solution @xmath205 , which in turn is typically much smaller than the thickness @xmath56 of the dielectric substrates @xmath22 , one identifies the case @xmath206 , where a ( practically ) ion - free oil @xmath27 ( @xmath207 ) is assumed .
for this regime eqs .
read @xmath208 and hence eq .
is given by @xmath209 since @xmath210 , the approximation in the second line of the previous equation almost always applies .
it shows that the electrowetting number @xmath144 for water on a hydrophobic dielectric in oil is also in agreement with @xmath211 in eq . with @xmath212 due to eq . .
replacing the hydrophobic dielectric substrate of the previous sec .
[ sec : phob ] by a hydrophilic one leads to the case @xmath25 ( i.e. , @xmath52 ) , @xmath213 , and @xmath214 , where again a ( practically ) ion - free oil @xmath213 ( @xmath215 ) is assumed .
for this regime eqs .
read @xmath216 which leads to eq . of the form
@xmath217 within the electrocapillarity approach one again expects , as in the previous sec . [ sec : phob ] , an electrowetting number @xmath218 ( see eqs . and ) . however ,
the electrowetting number @xmath144 within the present density functional approach in eq . , for sufficiently large voltage @xmath219 , is actually smaller than @xmath211 by a factor @xmath220 : @xmath221 .
it is apparent from eq . that @xmath222 is _ not _ necessarily close to unity , because the typically small value @xmath223 is multiplied with the typically large value @xmath224 .
assuming typical values of , e.g. , dielectric layers of thicknesses @xmath225 and dielectric constant @xmath226 , a debye length @xmath227 in the aqueous ( @xmath228 ) electrolyte solution @xmath189 , and thicknesses @xmath229 of the electrolyte films on the substrates , eq
. leads to @xmath230 . hence , for this example of electrowetting on a hydrophilic dielectric , the analysis leads to electrowetting numbers @xmath144 which are more than two orders of magnitude smaller than expected within the electrocapillarity approach : @xmath231 .
it appears as if no experimental studies of electrowetting on hydrophilic substrates have been reported so far .
this is remarkable since the preparation of hydrophilic substrates is standard in surface science .
whereas the previous two secs .
[ sec : phob ] and [ sec : phil ] considered an electrolyte solution and an oil , here the case of two immiscible electrolyte solutions is discussed .
this situation is characterized by @xmath232 , which leads to eqs .
of the form @xmath233 if electrolyte solutions @xmath6 and @xmath1 are defined by @xmath234 , i.e. , @xmath235 , the following three cases have to be distinguished : ( i ) @xmath25 ( i.e. , @xmath52 ) and @xmath213 with @xmath188 , ( ii ) @xmath27 ( i.e. , @xmath53 ) and @xmath205 with @xmath194 , and ( iii ) @xmath201 .
case ( i ) leads to the electrowetting number eq .
@xmath236 hence @xmath237 , where , however , the depression factor @xmath220 here is typically much smaller than that of the previous sec . [
sec : phil ] because typically @xmath238 ( see eqs . and ) .
the electrowetting number of case ( ii ) is given by @xmath239 this expression bears some resemblance to eq .
except of the typically very small prefactor @xmath240 here .
therefore , electrowetting is also expected to be strongly suppressed for two immiscible electrolyte solutions with @xmath241 , a condition which is typically fulfilled . for completeness the rather special case ( iii )
is mentioned , for which the electrowetting number reads @xmath242
electrowetting is studied in the present work by analyzing the capillary rise of a fluid in the environment of another fluid , where at least one of the two fluids is an electrolyte solution , for pellat s setup @xcite ( fig .
[ fig:1 ] ) within the density functional theory of wetting . here
, the widely ignored possibility of the formation of films of microscopic thickness on the substrates is taken into account ( fig .
[ fig:2 ] ) . considering the quasi - one - dimensional situation far away from the three - phase contact region ( fig .
[ fig:3](a ) ) allows to transparently derive the electrowetting equation .
the derivation shows that electrowetting is a consequence _ not _ of the voltage - dependence of the substrate - fluid interfacial tensions , i.e. , electrowetting is _ not _ an electrocapillarity effect , _ but _ of the voltage - dependence of the depth of the effective interface potential . the traditional electrocapillarity approach to electrowetting
is shown to be compromised by the reliance on the incorrect assumption that the surface structure of a fluid does not change upon bringing the system into contact with another fluid phase .
the present analysis of pellat s setup for electrowetting studies leads to effectively four length scales corresponding to the debye lengths in both fluids , the thickness of the substrates , and the film thicknesses , the latter being assumed to be approximately equal here , which serve to classify various relevant experimental situations , e.g. , uncoated metal electrodes , hydrophilic or hydrophobic dielectric substrates , or fluids comprising water+oil systems or immiscible electrolyte solutions .
the full dependence of the electrowetting number on these length scales is derived here ( eq . ) , which can be used for an actual experimental system . by considering limiting cases of general interest
it is found that for uncoated metal electrodes ( eqs . and ) and wetting of hydrophobic dielectric substrates by water in an oil environment ( eq . ) the electrowetting number within the present density functional approach agrees with that within the electrocapillarity picture as well as with numerous experimental studies .
however , a significantly reduced tendency of electrowetting is predicted here as compared to the predictions within the electrocapillarity approach for electrowetting on hydrophilic dielectric substrates ( eq . ) or situations with both fluids being immiscible electrolyte solutions ( eqs . ) . due to a lack of experimental data , verification of these predictions
is an open issue .
one can conclude that it is a matter of fortune that the traditional electrocapillarity approach to electrowetting ( eqs . and )
leads to an electrowetting equation , which , although derived by means of physically questionable arguments , turns out to be yet a precise approximation for many practical cases .
however , the present study highlights conditions for which significant deviations from the electrocapillarity picture are expected to be experimentally detectable .
the density functional theory of electrowetting presented here is suggested to be considered as an approach to fundamentally understand as well as to reliably quantify the phenomenon of electrowetting . | the phenomenon of electrowetting , i.e. , the dependence of the macroscopic contact angle of a fluid on the electrostatic potential of the substrate , is analyzed in terms of the density functional theory of wetting .
it is shown that electrowetting is not an electrocapillarity effect , i.e. , it can not be consistently understood in terms of the variation of the substrate - fluid interfacial tension with the electrostatic substrate potential , but it is related to the depth of the effective interface potential .
the key feature , which has been overlooked so far and which occurs naturally in the density functional approach is the structural change of a fluid if it is brought into contact with another fluid .
these structural changes occur in the present context as the formation of finite films of one fluid phase in between the substrate and the bulk of the other fluid phase .
the non - vanishing donnan potentials ( galvani potential differences ) across such film - bulk fluid interfaces , which generically occur due to an unequal partitioning of ions as a result of differences of solubility contrasts , lead to correction terms in the electrowetting equation , which become relevant for sufficiently small substrate potentials .
whereas the present density functional approach confirms the commonly used electrocapillarity - based electrowetting equation as a good approximation for the cases of metallic electrodes or electrodes coated with a hydrophobic dielectric in contact with an electrolyte solution and an ion - free oil , a significantly reduced tendency for electrowetting is predicted for electrodes coated with a dielectric which is hydrophilic or which is in contact with two immiscible electrolyte solutions . |
the risk for space travelers due to solar radiation , as well as galactic cosmic rays , has been studied extensively for nearly four decades , but the predictability and frequency of potentially lethal doses of high - energy solar radiation is still far from understood .
a long series of studies suggest that the most energetic solar proton events ( spes ) @xcite can produce lethal biological doses that require careful consideration of radiation risks for manned space travel ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the august 1972 spe , often taken as the standard high - fluence event for protection studies , was not an isolated anomaly .
the february 1956 , november 1960 , october 1989 , and october - november 2003 events produced solar - particle fluxes sufficiently large that an astronaut on the moon protected by only a spacesuit would likely have perished , and many events approaching these fluences have been recorded @xcite .
the durations of these events ( many hours , see @xcite ) are large enough that evacuation to shelter would prevent serious exposure to an astronaut on an eva .
observations of associated lightspeed precursors ( photon flares and cme eruptions ) , while not always available for the strongest spes @xcite , could allow further time for escape . given the number of studies that have taken on the computational difficulties of the propagation of solar cosmic rays in the inner heliosphere and the resulting biological doses , it is surprising that the simpler study of the effects of ionizing photons , in particular x - rays from solar flares @xcite , have not been considered .
it is easy to show that the fluence from the most energetic hard x - ray flares of the last 50 years would result in an amount of energy absorbed per gram of tissue larger than the lethal dose for an unprotected human at 1 au .
acute doses for various biological endpoints are well studied , especially because of their relevance to oncological research .
the frequency of very energetic flares is similar to that of the most energetic spes ; their durations are so small that evacuation is much more problematic ; and there are no early and reliable precursors that can be used to predict their onset as in the case of spes .
furthermore , the shielding requirements are nontrivial and somewhat different from those for solar energetic particles , preferring material of high atomic number , in contrast to the mostly polymer construction of current spacesuit designs .
bone surrounding blood - forming marrow might mitigate hematological effects , but the risk of carcinogenesis remains significant .
we calculate here the expected doses from energetic solar x - ray flares and use their observed frequency - fluence relation to estimate the probability per unit time of a hazardous flare exposure .
solar flare photons span energies between 10@xmath1 and 10@xmath2 kev @xcite .
rhessi observations ( e.g. * ? ? ?
* ) show that the spectrum of hard x - rays ( hxrs ) from solar flares extends to energies as low as 1025 kev , where `` hard '' refers to the power - law part of the spectrum produced by particle acceleration rather than thermal processes .
the hxr spectrum is usually fit by the form @xmath3 @xcite .
the range of the estimated power - law index ( log - log slope ) @xmath4 is large for hxr flare spectra , with @xmath5 and a median around 3.5 .
additionally , a small correlation between @xmath4 and total x - ray output has been suggested by rhessi @xcite .
x - class solar flares ( the most energetic flares ) have typical durations of about 530 min @xcite , with the more energetic flares tending to last longer . as with spectral slope ,
flare durations vary widely , but even with the longest of these durations the dose should be considered acute ( i.e. effects have rapid onset ) , in contrast to low - level , extended exposure to galactic cosmic rays .
the acute lethal dose is almost always independent of dose rate @xcite , so we assume that the nonlethal acute dose depends on total fluence , not flux , and hence on the flare total energy release .
the total photon energy release @xmath6 in flares is difficult to estimate and varies by at least a factor of 10@xmath7 , but a large number of studies using euv , soft x - ray ( sxr ) , and hxr satellite events roughly agree that the differential distribution @xmath8 of flare energy releases is a power law with index about @xmath9 to @xmath10 over at least six orders of magnitude in total energy release @xmath6 @xcite .
the largest x - ray releases observed among contemporary solar flares is @xmath11 erg @xcite .
eleven x - class flares occurred during the extraordinary solar outbursts between 18 october 2003 and 5 november 2003 @xcite , with sxr releases in the goes 18 ( 210 kev ) band peaking at @xmath12 erg for the largest bursts .
observations using the sorce instrument s total irradiance monitor yielded a total flare energy at all wavelengths for the 28 october flare of @xmath13 erg @xcite . radiation and
charged particles from these flares compressed the earth s van allen belt to within 20,000 km of the surface @xcite , damaged the orbiting mars odyssey communication instruments , and reduced polar ozone levels significantly @xcite .
although most of our calculations are for a 10@xmath14 erg flare , it seems likely that much more energetic flares than considered here have occurred .
such flares have been serendipitously discovered in other solar - like stars with otherwise normal characteristics @xcite .
upper limits on proton fluences inferred from cosmogenic isotopes in lunar samples @xcite , tree ring records of @xmath15c @xcite , and the statistics of impulsive nitrate events @xcite suggest that the frequency - fluence relation steepens for high - energy particle fluences above about 10@xmath16 @xmath17 , probably due to streaming - limited fluxes associated with self - confinement by ion - wave interactions @xcite . in contrast , no such limit , empirical or theoretical , has been established for photon flares , other than an upper limit to avoid divergence of total energy @xcite .
we adopt a mean recurrence time for 10@xmath18 erg flares of 10 yr , agreeing with the estimate of @xcite and broadly consistent with the dozen or so 10@xmath1410@xmath18 erg events that have been observed since goes soft x - ray monitoring began in 1976 . the average hxr frequency - energy release statistic @xmath8 is taken as a power law in @xmath6 , with log - log slope @xmath19 .
thus , we take the mean time between events of energy release @xmath20 ( in units of 10@xmath14 erg ) as @xmath21
for our calculations , solar flare photon number spectra are assumed to be distributed as power laws , @xmath22 , with @xmath23 ( see above ) .
the flare spectrum is assumed to extend from 10 kev to 511 kev .
the 10 kev lower limit is taken to simulate the hxr flares , which often begin to flatten to a thermal form around this energy , while the upper limit is set high enough that a negligible number of incident photons are at higher energies , even for the shallowest spectra ( lowest @xmath4 ) .
henceforth , total energy release quantities refer only to the energy between 10 and 511 kev .
we transport the incident ionizing radiation using a single - scattering approximation .
the primary interaction processes at x - ray energies are photoelectric absorption and compton scattering .
the photoabsorption opacity is much larger , so the radiative transfer can be computed by employing an effective opacity , in which both compton scattering and photoabsorption are treated as absorption processes . by including the compton cross section in the opacity , we in effect assume that the material is optically thin to compton scattering , such that at most a photon will scatter no more than once before exiting .
errors incurred by this approximation are negligible , as we have verified using monte carlo simulations @xcite . the compton - scattering coefficient was computed exactly using the klein - nishina formula , while the photoabsorption cross section was approximated using the empirical form of @xcite : @xmath24 where @xmath25 is the atomic number of the target nucleus and @xmath26 is the energy of the incident photon in units of 511 kev .
we found this form to be an excellent fit to cross section data for light elements and for energies higher than the k edge .
the k edges of carbon and aluminum are at ( respectively ) 284 and 1560 ev @xcite and are well below our spectral cutoff energy of 10 kev , so our calculation is in the energy regime that is consistent with the cross section fit .
the effective opacity of the material is then @xmath27 where @xmath28 is the mean atomic mass in amu of the target material , @xmath29 is the mass of a hydrogen atom , @xmath25 is its atomic number , and @xmath30 is the compton cross section .
the single - scattering approximation results in exponential attenuation of the incident fluence . in terms of the effective opacity @xmath31 and the areal density @xmath32 ( g @xmath17 ) of the shielding material , the attenuated energy fluence @xmath33 is @xmath34.\ ] ] after transport through the shielding material
, the photon fluence is converted to a biological dose by assuming that the remaining radiation is absorbed by pure water .
much of the damage by ionizing radiation is thought to be `` indirect , '' involving chemical reactions initiated by energy deposited in the bulk cell water or first hydration layer , rather than `` direct '' ionization of dna @xcite , although this terminology is now recognized as an oversimplification @xcite .
we then estimate the dose as a function of areal density of the shielding by integrating the attenuated incident photon energy spectrum over the effective opacity of water @xmath35 ( calculated using the single - scattering method above ) : @xmath36 where the attenuated fluence @xmath33 is given above , and the limits of integration are in kev .
this is essentially the skin dose .
to estimate risks to humans exposed to solar flare x - rays , critical levels of acute radiation exposure for various types of health outcomes ( e.g. hematologic damage , organ failure , cancer , and lethality ) are needed .
particularly useful is the minimum dose at which a given enhancement in occurrence of a disease relative to the average population occurs .
we use dosimetry units for which 1 gray = 100 rad = 10@xmath37 erg g@xmath1 absorbed .
for reference , a chest x - ray delivers around 0.1 mgy .
since x - ray photons have small linear energy transfer , @xmath38 , their `` quality factor '' or `` biological effectiveness '' is very close to unity , independent of energy , so dose in gy is approximately the dose in sievert .
the duration of exposures to flare x - rays of large fluence will be relatively short ( 1060 min in most cases ) , so we restrict the discussion to evidence concerning acute human radiation - syndrome data .
at doses above about 0.5 gy , summaries of a number of sources of data relevant to acute radiation syndrome in humans are available @xcite , as well as detailed specialized studies @xcite .
there is general agreement that severe damage , primarily hematologic and without assured recovery , occurs around 1 gy . estimates of the whole - body acute lethal dose vary from 2 to 5 gy @xcite .
an upper limit to the radiation risk dose in an exposed individual is the ratio of the average spontaneous rate of mutations over a large number of genes to the induced mutation rate per gy for low - let ( linear energy transfer ) irradiation , recognizing that the mutation rate is highly variable among loci ; this ratio is the mutation doubling dose .
an estimate using 135 human genes for spontaneous rates and 35 mouse genes for induced rates @xcite gives a doubling dose of 0.8 gy .
this dose is for chronic ( i.e. continual ) irradiation , and a dose rate reduction of a factor of three is usually assumed for acute doses , suggesting a doubling dose of 0.3 gy for acute irradiation .
however this estimate remains uncertain and subject to various definitions @xcite .
most work on x - ray and gamma - ray radiation risk to exposed individuals ( not genetic disease endpoints ) comes from studies of carcinogenesis .
there is little doubt that the incidence of cancer due to radiation - induced genomic instability rises with acute dose above 0.20.3 gy . below this dose
there is continued debate whether there is a `` linear - no - threshold '' relation between risk and ionizing - radiation dose , or whether doses below about 0.1 gy result in `` adaptive response '' causing endogenous dna damage prevention and immune stimulation @xcite .
alternatively , it is also likely that risk at low doses is larger than the linear - no - threshold extrapolation , even increasing with decreasing dose , because of bystander effects , as reviewed in @xcite .
the situation is further complicated by the existence of a significant fraction of humans with predisposing mutations to cancers induced by ionizing radiation ( e.g. * ? ? ?
it is certain that mutations , often at significant loci @xcite , chromosomal abnormalities in blood lymphocytes @xcite , and clustered dna damage @xcite occur at much smaller doses in the range 0.01 to 0.1 gy , but their impact on health risk has been difficult to assess because of the evidence for adaptive response at low doses .
a compelling discussion by @xcite argues that there is good human epidemiological evidence for increased cancer risk at an acute x - ray or gamma - ray dose of 0.05 gy , and reasonable evidence for enhanced risk above 0.01 gy , although the risk enhancement is in the 110% range .
acute critical doses for significantly increased risk for other disease endpoints may also be of order 0.1 gy or less , especially for hematological diseases .
for example , the summary of delayed somatic ( i.e. bodily ) effects due to acute doses by @xcite indicates that gastrointestinal tract syndrome ( leading to loss of digestion ability , bleeding ulcers , and diarrhea ) sets in at 0.1 gy for x - rays .
the risk of a particular _ genetic_-disease endpoint per gy of irradiation is more difficult to estimate , since the disease - specific induced mutation rate varies greatly and most models assume mutation - selection equilibrium
. inspection of risk estimates in humans and animal models for a number of genetic - disease classes that include 26 human disorders encompassing 135 genes @xcite indicates that risks for genetic - disease endpoints at 0.1 gy due to acute x - ray doses are probably much smaller than the risks for somatic disease in an exposed individual discussed earlier .
these estimates include corrections for potential recoverability and concentrate on low - let radiation , x - rays and @xmath39-rays .
the results are given for chronic irradiation , even though they are based on experiments with high dose - rate irradiation .
the `` dose - rate reduction factor '' for acute doses is believed to be roughly a factor of three , a factor which is well established in studies of specific - locus mutations . using this factor to correct the results
, it is found that risks in terms of the excess over an average population for acute doses are typically 1% per gy .
but these results were typically obtained at equivalent acute doses in the range 0.53 gy , so linearity of the above risk with dose can not be assumed .
a review of evidence for enhanced mutation rates in human populations exposed to doses as low as 0.25 gy in the chernobyl accident and nuclear weapons tests in kazakhstan is given in @xcite .
but the main point is that the risks for genetic disease endpoints at 0.1 gy might be smaller than the risk for somatic disease in an exposed individual . given that the linear increase of cancer risk with dose for acute doses above 0.1 gy seems unequivocal and that the threshold for delayed somatic - disease endpoints appears to also be about 0.1 gy , we adopt this biological dose as an upper limit for significant risk increase due to x - rays , with the understanding that enhanced risk for several disease endpoints , but especially cancer @xcite , may still be significant at lower doses .
an important unresolved point is that adoption of a critical dose depends on the risk enhancement that one is willing to accept ; our reading of the literature suggests a 10% enhancement in potentially fatal disease endpoints at 0.1 gy , and less than 1% enhancement for genetic disease endpoints at this dose .
we can now compare our adopted upper dose limit of 0.1 gy with that received behind polymer shielding , which is representative of current spacesuit design , such as the space shuttle extravehicular mobility unit @xcite .
we use pure carbon as a proxy for the mostly polymer construction of spacesuits because only the carbon atoms in polymers significantly absorb x - rays . the results , assuming a 10@xmath14 erg solar flare , are shown in fig . [ fig : eva1 ] .
erg x - ray flare as a function of flare spectral index , @xmath4 , and areal density , @xmath32 .
the dose is roughly independent of @xmath4 for shields with areal densities smaller than about 2 g @xmath17 . for larger shielding columns
, the dose becomes sensitive to the spectral shape because more of the incident spectrum is attenuated , and hence reshaped , before being absorbed by the model water column .
areal densities of polymer in excess of 2 g @xmath17 are needed to reduce the x - ray dose to below our adopted maximum acceptable acute dose of 0.1 gy . ] the typical areal densities of spacesuit components ( @xmath40 0.51.5 g @xmath17 ) provide little protection during a large solar flare .
a relatively common 10@xmath14 erg flare would deliver over 0.2 gy behind the current spacesuit twice our adopted upper limit .
a thicker spacesuit could reduce this dose to 0.1 gy , but larger flares do occur , albeit less often . bolstering spacesuits simply decreases the frequency of dangerous doses and does not eliminate the threat .
what is the likelihood of being exposed to doses above 0.1 gy ? if we use for simplicity a typical flare spectral index of 3.5 , we find that the approximate dose behind polymer shielding of areal density @xmath41 g @xmath17 is @xmath42 where @xmath20 is the flare energy release in units of 10@xmath14 erg , and @xmath32 is the areal density of the shielding material in g @xmath17 . using the flare energy - recurrence frequency relation given in the discussion of flare properties to eliminate @xmath20 , the mean time between flares delivering _ at least _ a given dose @xmath43 behind an polymer shield of areal density @xmath32 is @xmath44 where @xmath43 is the dose in units of 0.1 gy .
if we then assume that time between flare events is typically much larger than the duration of exposure and that flares occur at random , the probability of exposure to a dose of at least the critical dose is @xmath45 per hour of eva .
so in just 100 hours of eva in the current spacesuit , an astronaut would accumulate a 10% risk of a dangerous exposure to solar flare x - rays .
the previous section discusses only the risk due to photons from flares .
the risk due to solar energetic particle events ( spes ) is certainly not negligible , but we argue that the risk during evas is significantly smaller than estimated here for hard x - rays from flares .
the largest spes , such as the 14 july 2000 , feb 1956 , and aug 1972 events , had 1 au fluences in the range 10@xmath1610@xmath46 @xmath17 @xcite , although per - particle energies were 100 times smaller than for galactic cosmic rays .
@xcite estimated the incidence of flares that produce spes with 1 au doses greater than 1 gy as about one per decade , similar to what we estimate for 10@xmath18 erg photon flares that produce 2 gy doses , as shown in the fig .
a similar frequency for spes with fluences above 10@xmath16 @xmath17 can be derived from the recorded number of large events @xcite , nitrate ice core reconstruction covering several centuries @xcite , and probability models @xcite .
in contrast , 10@xmath14 erg x - ray flares that also require substantial shielding are about 50 times more frequent .
spacesuits are the last line of defense until shelter is reached , so exposures during a flare will depend on the ratio of the time to reach shelter to the time to deliver the total fluence .
figure [ fig : eva2 ] shows that the doses behind aluminum shielding are significantly smaller than that received inside spacesuits of the same areal density because of the strong dependence of the photoabsorption cross section on atomic number of the target material . but x - ray flares leave only 1030 min to reach shelter before the total fluence is delivered .
hard x - ray flares are impulsive , with rise times of minutes or less .
the time to withdraw to adequately shielded shelter is very small . a reliable energetic hard x - ray flare precursor signature occurring more than an hour before the flare maximum
would be needed for this purpose .
there are many signatures that have been proposed as flare precursors @xcite , such as uv brightening , soft x - ray enhancements , microwave radio signatures , h@xmath47 filament disturbances , strong magnetic shear , sunspot motions , and the beginnings of chromospheric mass ejections ( cmes ) , whose onset is now known to slightly precede an associated flare on average .
most of these precursors are only observable for less than a few minutes before the onset of the flare , so do not give sufficient warning and would require elaborate monitoring systems .
some radio precursors , such as polarization signatures , are observed up to tens of minutes before a flare , but are not observed in the majority of flares .
similarly , for all precursor signatures there are flares seen without the precursor and observations of `` precursors '' that are not followed by a flare ; no precursors are necessary and sufficient @xcite .
even more seriously , none of these precursors are predictors of the energy release of the flare itself , so given the large frequency of flares that pose no biological hazard , use of these precursors would likely result in a large false - alarm rate .
most probabilistic approaches for flare prediction are based on a combination of historical rate of flaring for a given sunspot classification group and additional information such as shear , magnetic topology , and previous large - flare activity @xcite .
a more recent bayesian method relies only on flare event statistics @xcite .
these methods are most suited for probabilistic prediction of quantities like the number of flares of a certain class in a given year , but not for eva hazard warnings .
for example , the bayesian method would have predicted a 20% probability for an x - class flare on 4 nov 2003 , using data up to one day before , including a highly clustered series of strong and weak flares in the week before , yet actually the most energetic flare in several decades was about to occur @xcite , a flare almost an order of magnitude more energetic than the model flare used in the calculations reported here .
such prediction algorithms might be useful for policies requiring no evas in windows of a week or so , when the probability of a large flare can be somewhat more accurately estimated , but this could greatly curtail manned exploration , depending on where the threshold for significance is placed and the reliability of the prediction . for example , less reliable predictions would require lower thresholds for significance ( and thus higher alarm rates ) to maintain acceptable risk levels .
because of diffusive propagation through the heliosphere , the first particles from spes arrive at 1 au more than an hour after the initiating event ( when one is observed ) and take hours to reach hazardous fluences ( * ? ? ?
2.7 , 12.9 ) @xcite .
this allows a simple and effective retreat strategy based on flare , cme , or even sunspot precursors ( although not all energetic spes are associated with flares ; @xcite ) . consequently ,
x - ray flares are more dangerous during evas
. one simple protection solution would be to include a 23 m@xmath48 shield of high-@xmath25 material , such as aluminum , in eva rover designs .
heavy elements more efficiently stop x - rays than light elements , and no significant secondary radiation is produced by x - rays , so we believe a material such as aluminum to be optimal for this purpose . using the results in fig .
[ fig : eva2 ] , an aluminum shield thick enough to protect against a 10@xmath14 erg flare would have an areal density of at least 7 kg per square meter of shielding ; for a 10@xmath18 erg flare the mass requirement would be 70 kg m@xmath49 . additionally the shield would have to be articulated or detachable , in order to provide protection regardless of the sun s position in the sky .
this shield would be employed during evacuation , since it would be useless or even dangerous if a large spe followed the flare , since the high-@xmath25 composition would enhance the radiation dose from secondary particle production within the shield .
erg x - ray flare as a function of flare spectral index , @xmath4 , and areal density , @xmath32 .
aluminum has a higher atomic number than carbon ( 13 vs. 6 ) and absorbs x - rays much more efficiently . only 2 g @xmath17 of aluminum shielding would be required to reduce the dose to below 0.05 gy for a 10@xmath14 erg flare . ]
the risk to astronauts due to solar flare x - rays was before now mostly unknown , in contrast to the highly studied effects of particle radiation , such as solar protons and galactic cosmic rays .
here we have calculated the transport of ionizing radiation through spacesuit material and present acute biological doses due to x - rays that an astronaut could receive if a large solar flare occurs during an eva .
based on the studies of somatic and genetic disease risks due to acute doses of ionizing radiation , we adopt a minimum hazardous dose of x - rays of 0.1 gy and find that the risk of receiving at least 0.1 gy from an x - ray flare is roughly 10% per 100 hours of accumulated eva .
the onset and duration of x - ray flares is rapid enough and possible precursors are unreliable enough that avoidance would be difficult .
the simplest solution for x - ray protection on rover - based evas could be the inclusion of a mobile body shield to supplement the shielding provided by the spacesuit until shelter can be reached .
dss was supported by the nsf graduate student research fellowship and harrington doctoral fellowship programs .
jms was supported by the nasa exobiology program , grant nng04gk43 g .
this work was carried out as part of the research of the nasa astrobiology institute virtual planetary laboratory lead team , which is supported through the nasa astrobiology institute .
aschwanden , m.j .
( 1999 ) , nonthermal flare emissions , in _ the many faces of the sun : a summary of the results from nasa s solar maximum mission _ , edited by k. t. strong et al . , pp . 273300 ,
springer , ny . , m. j. , t. d. tarbell , r. w. nightingale , c. j. schrijver , a. title , c. c. kankelborg , p. martens , and h. p. warren ( 2000 ) , time variability of the `` quiet '' sun observed with trace .
physical parameters , temperature evolution , and energetics of extreme - ultraviolet nanoflares , _ astrophys .
j. _ , _ 535 _ , 10471065 .
badhwar , g.d .
( 1997 ) , deep space radiation sources , models , and environmental uncertainty , in _ shielding strategies for human space exploration _ , edited by j. w. wilson et al . , pp .
1729 , nasa conf . pub .
baker , d. n. , s. g. kanekal , x. li , s. p. monk , j. goldstein , j. l. burch ( 2004 ) , an extreme distortion of the van allen belt arising from the `` halloween '' solar storm in 2003 , _ nature _ , _ 432 _ , 878881 .
crosby , n. b. , o. h. w. siegmund , p. w. vedder , and j. v. vallerga ( 1993 ) , extreme ultraviolet explorer deep survey observations of a large flare on au microscopii , _ astrophys .
_ , _ 414 _ , l49l52 .
cucinotta , f. a. , w. schimmerling , j. w. wilson , l. e. peterson , g. d. badhwar , p. b. saganti , and j. f. dicello ( 2001 ) , space radiation cancer risks and uncertainties for mars missions , _ radiat .
_ , _ 156 _ , 682688 .
cucinotta , f. a. , w. schimmerling , j. w. wilson , l. e. peterson , p. b. saganti , and j. f. dicello ( 2004 ) , uncertainties in estimates of the risks of late effects from space radiation , _ adv .
_ , _ 34 _ , 13831389 . dubrova , y. e. , v. n. nesterov , n. g. krochinsky , v. a. ostapenko , g. vergnaud , f. giraudeau , j. buard , and a. j. jeffreys ( 1997 ) , further evidence for elevated human minisatellite mutation rate in belarus eight years after the chernobyl accident , _ mutat .
_ , _ 381 _ , 267278 .
getselev , i. , s. rumin , n. sobolevsky , m. ufimtsev , and m. podzolko ( 2004 ) , absorbed dose of secondary neutrons from galactic cosmic rays inside the international space station , _ adv .
res . _ , _ 34 _ , 14291432 . ,
m. , m. audard , v. l. kashyap , j. j. drake , and e. f. guinan ( 2003 ) , are coronae of magnetically active stars heated by flares ?
ii . extreme ultraviolet and x - ray flare statistics and the differential emission measure distribution , _ astrophys . j. _ , _ 582 _ , 423442 . , b. l. , e. m. gullikson , and j. c. davis ( 1993 ) , x - ray interactions : photoabsorption , scattering , transmission , and reflection at @xmath26 = 5030,000 ev , @xmath25 = 192 , _ atomic data and nuclear data tables _ , _ 54 _ , 181342 .
kanbach , g. , d. l. bertsch , c. e. fichtel , and seven others ( 1993 ) , detection of a long - duration solar gamma - ray flare on june 11 , 1991 with egret on compton - gro , _ astron .
_ , _ 97 _ , 349353 . lingenfelter , r. e. , and h. s. hudson ( 1980 ) , solar particle fluxes and the ancient sun , in _ the ancient sun : fossil record in the earth , moon , and meteorites _ , edited by pepin , r. o. , et al . ,
6979 , pergammon press , ny . mccracken , k. g. , g. a. m. dreschhoff , e. j. zeller , d. f. smart , and m. a. shea ( 2001 ) , solar cosmic ray events for the period 1561 - 1994 : 1 .
identification in polar ice , 1561 - 1950 , _ j. geophys .
_ , _ 106 _ , 2158521598 .
reedy , r. c. ( 1996 ) , constraints on solar particle events from comparison of recent events and million year averages , in _ solar drivers of interplanetary and terrestrial disturbances _ , edited by k. s. balasubramaniam et al .
, asp press , san francisco .
sankaranarayanan , k. , and r. chakraborty ( 2000 ) , ionizing radiation and genetic risks .
the doubling dose estimates from the mid-1950s to present and the conceptual change to the use of human data on spontaneous mutation rates and mouse data on induced mutation rates for doubling dose calculations , _ mutat .
_ , _ 453 _ , 107127 . sparrow , a. h. , a.g .
underbrink , and r. c. sparrow ( 1967 ) , chromosomes and cellular radiosensitivity .
i. the relationship of d0 to chromosome volume and complexity in seventy - nine different organisms , _ rad .
_ , _ 31 _ , 915949 .
sutherland , b. m. , p. v. bennett , o. sidorkina , and j. laval ( 2000 ) , clustered dna damages induced in isolated dna in human cells by low doses of ionizing radiation , _ proc .
, _ 97 _ , 103108 .
violot , d. , r. mkacher , e. adjadj , j. dossou , f. de vathaire , and c. parmentier ( 2005 ) , evidence of increased chromosomal abnormalities in french polynesian thyroid cancer patients , _ eur .
_ , _ 32 _ , 174179 .
ward , j. f. ( 1999 ) , ionizing radiation damage to dna : a challenge to repair systems , in _ advances in dna damage and repair : oxygen radical effects , cellular protection , and biological consequences _ , edited by dizdaroglu , m. , and a. e. karakaya , pp .
431439 , kluwer / plenum , ny . woods , t. n. , f. g. eparvier , j. fontenla , j. harder , g. kopp , w. e. mcclintock , g. rottman , b. smiley , and m. snow ( 2004 ) , solar irradiance variability during the october 2003 solar storm period , _ geophys .
_ , _ 31 _ , l1080210806 . | solar hard x - ray flares can expose astronauts on lunar and deep space extravehicular activities ( evas ) to dangerous acute biological doses .
we combine calculations of radiative transfer through shielding materials with subsequent transfer through tissue to show that hazardous doses , taken as @xmath0 gy , should occur with a probability of about 10% per 100 hours of accumulated eva inside current spacesuits .
the rapid onset and short duration of x - ray flares and the lack of viable precursor events require strategies for quick retreat , in contrast to solar proton events , which usually take hours to deliver significant fluence and can often be anticipated by flares or other light - speed precursors .
our results contrast with the view that only particle radiation poses dangers for human space exploration .
heavy - element shields provide the most efficient protection from x - ray flares , since x - rays produce no significant secondary radiation .
we calculate doses due to x - ray flares behind aluminum shields and estimate the required shield masses to accompany eva rovers . |
one of the most serious problems in planet formation is how the dust components grow up to larger objects in protoplanetary disks .
the known key - stage is when the dust particles are meter - size @xcite .
since gas pressure gradient in a protoplanetary disk is negative , gas rotates at sub - keplerian velocity . on the other hand ,
the dust particles tend to rotate at keplerian velocity and therefore they feel headwind of gas . losing their angular momenta , they spiral into a central star quickly on timescales of 100 years .
one possibility to overcome the meter - size barrier is the planetesimal formation via gravitational collapse of a cluster composed with small dust particles @xcite .
the dust would settle to midplane in the absence of turbulence .
early studies of dust settling in turbulent accretion disks were carried out by @xcite , @xcite and @xcite .
they presented that large dust particles settle down toward their equilibrium distribution in a turbulent diffusive time scale while small ones remain mixed throughout the whole gas disk .
if dust density in midplane becomes high enough , kelvin - helmholtz instabilities between a sediment layer of dust and the overlying gas can drive turbulence and mix the dust layer enough to prevent gravitational instability @xcite .
@xcite , however , carried out two - dimensional simulation and found active turbulent concentration .
except the effect of kelvin - helmholtz instabilities , the high dust density in midplane leads to a streaming instability @xcite because the dust is susceptible to preferential radial migration relative to the gas @xcite .
the streaming instability was shown to lead to turbulence and concentrate dust particles locally @xcite .
as one of other mechanisms of turbulent concentration , the effect of the turbulence driven by the magnetorotational instability ( mri ; @xcite ) is recognized .
planetesimal formation by gravitational instability in the mri turbulence is addressed by @xcite and @xcite .
turbulence level is the most fundamental parameter governing whether planetesimals can form by gravitational instability in these models .
the turbulence level of mri depends on the ionization degree of disk gas and the strength of magnetic field .
the linear analysis showed that the growth rate of mri is reduced by ohmic diffusion when collisions are so frequent that not only the ions and neutrals are well coupled but also electrical currents are damped @xcite . at the same time , the growth wavelength becomes longer and transport rate of angular momentum is diminished .
the nonlinear evolution of the mri in nonideal mhd starting from the weak vertical magnetic field has been studied by @xcite .
they found that sustained mhd turbulence requires the magnetic reynolds number @xmath0 , where @xmath1 , @xmath2 and @xmath3 are alfven velocity , magnetic resistivity and keplerian rotation frequency .
the values of @xmath2 are regulated by an ionization structure in a protoplanetary disk .
the inner regions of a disk are thermally ionized @xcite . at greater distances from the disk center , stellar x - rays and diffuse cosmic rays ionize the surface gas layer down to a certain column density @xcite . in moderately distant regions in which disk column density is high enough , midplane layers are not highly ionized and mri could be inactive there .
this region is called `` dead zone '' @xcite .
the characteristics and interesting physics are prospective around dead zones .
the magnetic turbulence outside the dead zone , which transports angular momentum and mass , creates gas density bump at the outer boundary of the dead zone .
this density pile - up triggers the rossby wave @xcite to form long - life anticyclonic vortices .
@xcite suggested that the dust particles are trapped in a center of vortex . by numerical simulation of rossby vortices including solid particles , @xcite confirmed the enhancement of dust density there .
@xcite showed that gravitationally bound embryos can form in the regions of enhanced dust density .
* ; hereafter referred to as paper i ) showed that non - uniform excitation of mri creates angular velocity profile that consists of super and sub - keplerian regions locally in the disk ( for details , see section 2 ) .
it is expected that this radial structure of the angular velocity leads to prevention of spiral - in of dust particles and their accumulation at the boundary of super and sub - keplerian regions , since the super - keplerian flow pushes the dust particles outward while the sub - keplerian flow drags them inward .
this interesting mechanism is based on the co - existence of mri active and inactive regions .
the ionization degree that regulates mri depends on the relative abundance and size distribution of grains @xcite .
it can also be changed by phase chemistry @xcite .
slight changes in these quantities near boundaries at global active / dead zones of mri may create the local inhomogenity in mri growth .
we will address this issue in a subsequent paper .
note that the local model in this paper is not a toy model that mimics the global inhomogeneous structure of dead and active zones . in this paper
, we investigate dust accumulation induced by inhomogeneous mri evolution by three - dimensional non - ideal mhd simulation including dust particles .
the dust particles are represented by lagrangian particles suffering gas drag and are assumed not to affect on gas .
we consider both cases with radially non - uniform magnetic resistivity under uniform magnetic field direction and non - uniform magnetic field with radially uniform magnetic resistivity .
mri develops in low resistivity regions in the former case and in strong vertical magnetic field regions in the latter case .
the latter model is the same as the model adopted in paper i. we briefly summarize the results of paper i in section [ sec : paper1 ] .
we explain our simulation setting in section [ sec : model ] . in section [ sec : result ] , we present that the mri leads to quasi - steady state and particles are locally concentrated as expected .
we also analyze the distributions of particle radial velocity to understand the effect of weak remnant magnetic turbulence .
we examine how high the particle density enhancement is and the possibility of planetesimal formation via gravitational instability .
finally , we summarize our results and discuss their implications in section [ sec : discussion ] .
in paper i , we have studied how the angular velocity profile of gas is modified when mri is excited non - uniformly in a part of a disk .
local two - dimensional simulations ( in the @xmath4-@xmath5 plane where @xmath4 and @xmath5 are radial and vertical coordinates to the disk ) with the shearing box boundary conditions were performed .
we assumed the ionization degree of the gas to be relatively low .
the vertical component to the disk plane ( @xmath5 component ) of magnetic field was changed along the radial direction ( @xmath4 ) while the absolute values of the field is constant . in the part of a disk where the ionization degree is low ( equivalently ,
resistivity is high ) , mri is active only in the regions with relatively large vertical component of the magnetic field .
one of the models used in the present paper has the same setting but with three - dimensional simulations , while the other model in the present paper assumes non - uniform resistivity with uniform magnetic field . the most illustrative result of paper i is shown in figure [ fig : paper1 ] .
time evolution ( @xmath6 ) of vertically averaged pressure and angular velocity of gas is plotted in panels a and b. mri is excited only in the initially unstable region ( @xmath7 ) with large enough initial vertical field and vigorous angular momentum and mass transport occurs within the unstable zone . when the magnetic perturbations are propagated to the initially stable regions ( @xmath8 ) , they are dissipated via resistivity and the large part of the stable region remains undisturbed .
hereafter , we call the initially mri stable and unstable regions simply as `` stable '' and `` unstable '' regions , respectively . after the vigorous angular momentum and mass transfer , a new quasi - steady state with the local rigid rotation feature appears in the unstable region ( figure [ fig : paper1]b )
. the pressure profile of the gas is also modulated ( figure [ fig : paper1]a ) . the rigid rotation feature that appears in the saturated quasi - steady state is sustained by the modified pressure pattern with its two peaks located around the unstable - stable boundary regions and its bottom located in the middle of the unstable zone . the quasi - steady rigid rotation pattern results in super - keplerian angular velocity at the outer part of the unstable region .
since the gas in the stable region is unperturbed and remains sub - keplerian when the global pressure gradient is taken into account , one finds that super- and sub - keplerian zones co - exist in the part of the disk due to the non - uniform excitation of mri .
the co - existence of super- and sub - keplerian zones has significant implication for the dust particle dynamics in a disk .
as mentioned in the introduction , the dust particles feel headwind and migrate inward in the sub - keplerian gas flow . on the other hand ,
the dust particles feel tailwind and migrate outward in the super - keplerian zone . then at a boundary between sub- and super - keplerian zones , with the super - keplerian zone situated on the inner side , which is indeed seen at the outer edge of the unstable region , the inward migration of dust particles through the sub - keplerian zone ( the stable zone )
are halted .
the boundary region potentially acts as a barrier for the dust infall and furthermore as the focal annulus of the dust particle concentration which may eventually lead to planetesimal formation .
the widths of the stable and unstable regions are one of the key parameters that determine the characteristics of the final state .
we have found that the final state is classified by the spatially averaged magnetic reynolds number in the initial condition , @xmath9 where @xmath10 is @xmath5 component of alfven velocity
. if @xmath11 is less than unity , a quasi - steady state with the local super - keplerian rotation appears .
the turbulence level within the stable region and at the boundary varies according to @xmath11 .
the degree of dust concentration at the boundary region would be affected by the level of turbulence . for @xmath12 ,
the magnetic perturbations are not dissipated enough within the stable region and the velocity pattern undergoes perpetual changes without reaching a quasi - steady state .
steady accumulation of dust particles is not expected in this case .
note that our simulations with constant @xmath11 converge with increasing resolution , because the resistivity @xmath2 ( ohmic dissipation ) is explicitly included in our simulations and it is kept constant due to the constancy of @xmath11 . in the present paper ,
we address the following questions that arise from the results of paper i : ( 1 ) do dust particles truly concentrate at the boundary ?
( 2 ) is the emergence of the quasi - steady rigid rotation region preserved in three - dimensional cases ? ( 3 ) how does remaining weak turbulence affect the dust concentration process ? ( 4 ) are dust particles concentrated dense enough on long enough timescales for subsequent gravitational instability to set in ?
here we are concerned with a small region in a protoplanetary disk and study the local dynamics of gas and dust particles .
the gas motion is calculated by resistive mhd simulations .
the dust particles are modeled as test particles that move under the effects of the gravity from the central star and the gas drag .
we adopt a local shearing box that has radial ( @xmath4 ) , azimuthal ( @xmath13 ) and vertical ( @xmath5 ) directions .
the boundary conditions are periodic in the @xmath13 and @xmath5 directions and the shearing box boundary condition is used in the @xmath4 direction @xcite .
we calculate compressible resistive magnetohydrodynamic ( mhd ) equations given by @xmath14 , \label{induction}\\ p&=&c^2_{s}\rho , \label{isothermal}\end{aligned}\ ] ] where @xmath15 is a unit vector in the @xmath4 direction .
we consider only the ohmic dissipation in equation ( [ induction ] ) because influence of ambipolar diffusion is negligible at the density level that is expected in the inner ( @xmath16 ) midplane of a protoplanetary disk @xcite .
the last terms in r.h.s . of equation ( [ motion ] ) allows the effect of the global pressure gradient to be taken into account in our local model .
the parameter @xmath17 is related to the global pressure gradient as @xmath18 where we assume @xmath19 , with constant @xmath20 and an isothermal disk @xmath21 , where @xmath22 is disk scale height .
the value of @xmath23 is treated as a constant parameter in our local model .
we solve equation ( [ induction ] ) with mocct method @xcite and the advection terms in equation ( [ motion ] ) with cip method @xcite .
the dust grains are treated as test particles .
the equation of motion of the @xmath24-th particle is given by @xmath25 where @xmath26 is the gas velocity at the location of the @xmath24-th particle , which is interpolated with the three - dimensional first - order interpolation scheme by values of @xmath27 at the eight grid points surrounding the particle .
the friction force , the last term in equation ( [ particle ] ) , is proportional to the velocity difference between the dust and gas since the dust sizes that we are concerned with are in epstein and stokes drag - law regimes . the characteristic friction ( stopping ) timescale @xmath28 is given by @xmath29 where @xmath30 , @xmath31 and @xmath32 are the dust particle radius , dust particle density and the sound speed , respectively .
@xmath33 is the molecular viscosity and equal to @xmath34 , where @xmath35 is the mean free path of the gas molecules . depending on the particle size relative to the mean free path ( @xmath36 ) the drag law changes from the epstein to the stokes regime .
the scaled friction times of @xmath37 and @xmath38 correspond to the grain sizes of approximately 0.1 and 1 meter , respectively , at the radial location of jupiter in the minimum mass solar nebula model .
all of our simulations start with uniform gas density ( @xmath39 ) and gas pressure ( @xmath40 ) .
the global gas pressure gradient is set to be @xmath41 .
the initial gas angular velocity profile is given by unperturbed keplerian flow ( @xmath42 ) and a small reduction due to global gas pressure gradient ( @xmath43 ) , @xmath44 .
initial disturbances are given to the gas radial velocity with the amplitude @xmath45 .
test particles are initially orbiting at the kepler angular velocity .
the particles move inward in the initial stage because they lose their angular momentum by the headwind that they feel due to the global pressure gradient .
this is the dust infall problem described in introduction .
initially 8 particles per gird are distributed . since typical grid
we use is @xmath46 ( table [ tab:1 ] ) , totally @xmath47 particles are distributed .
the non - uniformity in the mri growth is set either by non - uniform resistivity ( case1 ) or non - uniform vertical ( @xmath5 ) component of the magnetic field ( case2 ) .
we describe these cases in details in the following . in case1 ,
gas ionization degree is set to be radially inhomogeneous under the uniform magnetic field .
the linear analysis for mri caused by vertical magnetic field shows that the larger resistivity makes unstable wavelength longer @xcite .
the upper limit to the wavelength , @xmath48 , is scale height for actual disks while in our simulation , it is the size of the simulation box in the @xmath5 direction ( @xmath49 ) .
the critical value of resistivity with which the modes with wavelength shorter than @xmath48 can not grow is @xmath50 where @xmath51 is the wave number corresponding @xmath48 . here
@xmath52 is the plasma beta . in our case
there is no growing mode if the resistivity makes @xmath53 larger than @xmath49 . in case1 , we set @xmath54 and @xmath55 .
resistivity varies in the radial direction such that mri grows only in a limited zone in the center of the simulation box . the resistivity distribution is not changed throughout runs .
the radial distribution is shown in figure [ fig : ini]a . in this paper ,
the zone at the center where resistivity is initially sub - critical is referred to as the `` unstable '' region .
the radial width of the stable and unstable regions are denoted by @xmath56 and @xmath57 , respectively .
we have performed only one simulation run for case1 , which will be called model-@xmath2 in the later sections .
table [ tab:1 ] lists the parameter values used in the run .
the set - up of case2 is the same as that in paper i except for three - dimensional simulations .
the magnetic resistivity is uniform but the vertical component of magnetic field varies in the radial direction . that is , the initial magnetic field is situated as @xmath58 , where @xmath59 is the angle between the vertical axis and the magnetic field .
the vertical magnetic component , or @xmath60 , determines whether the instability grows or not . for large @xmath60 ,
vertical magnetic field is weak and short vertical wavelength modes grow faster . since
such modes are dissipated by the ohmic dissipation effectively , the vertical field mri does not develop .
the mri caused by the azimuthal magnetic field is known to grow rapidly with the limit of extremely short vertical wavelength in ideal mhd @xcite but the such short vertical wavelength modes tend to be damped resistively .
in fact , @xcite found that the azimuthal field mri growth ceases for the ionization degree higher than that stabilizes the vertical field mri .
we also did not find the development of the azimuthal field mri in our simulation as long as the box size in the @xmath13 direction is expanded by a factor up to 3 from the nominal cases at least for about 20 orbits in which the gas velocity and dust density vary dramatically .
the radial distribution of @xmath60 is shown in figure [ fig : ini]b . here , @xmath61 , and the magnetic resistivity is @xmath62 .
the larger vertical magnetic component in the center of the box makes mri locally unstable . outside the unstable zone
the value of @xmath60 is smaller than the critical one and mri is stabilized .
the two - dimensional version of case2 was studied in paper i , in which we found that if the spatially averaged magnetic reynolds number @xmath11 ( equation ( [ rmave ] ) ) is small enough , the quasi - steady angular velocity profile that is deviated significantly from the keplerian is generated .
since it is expected that this feature is preserved in three dimensional simulations and @xmath11 depends on the radial widths of the unstable region @xmath57 and the stable region @xmath56 , we investigate several cases with various @xmath56 while keeping @xmath63 in case2 : @xmath64(@xmath65 , model - s40 ) , @xmath66(@xmath67 , model - s11 ) , @xmath68(@xmath69 , model - s055 ) .
note that two - dimensional calculations with the same @xmath56 and @xmath57 were performed in paper i. in all of these cases , the friction time is set to be @xmath70 .
to investigate the effects of the friction time , we set it to be @xmath71 in model - t01 , which has @xmath72 and @xmath64 ( the same as model - s40 ) .
the initial settings are summarized in table [ tab:1 ] .
when the condition @xmath73 is met , the non - uniform growth of mri will create a new quasi - steady gas angular flow pattern that should affect the dust particle dynamics substantially ( paper i ) . at the same time
, weak remnant turbulent flow may be superposed upon the quasi - steady pattern and would affect the dust motions , with its degree being dependent on the spatially averaged magnetic reynolds number .
here we evaluate these two effects on the dust accumulation .
we will be focusing on the @xmath4 component of the dust velocities @xmath74 .
if the turbulence is small enough , as we have mentioned in section [ sec : paper1 ] , the velocity difference in the @xmath13 component ( angular velocity ) between dust and gas is the determining factor for the assemblage of dust particles at the boundary between super and sub - keplerian zones . in order to quantitatively analyze the effect of the remnant turbulence , we divide the radial component of the particle velocity into two components : one is that due to the non - keplerian gas motion in the quasi - steady state achieved at the mri saturation and the other is that due to turbulence .
the equation of motion for the dust velocity field at the grid point is @xmath75 we set @xmath76 where @xmath77 is a deviation from kepler flow .
assuming @xmath78 ( @xmath79 ) in @xmath4 and @xmath13 components and eliminating @xmath80 , the @xmath4 component of the particle velocity is given by @xmath81 where @xmath82 is the gas angular velocity relative to keplerian velocity , @xmath83 and @xmath84 is the gas radial velocity due to turbulent viscosity .
the component @xmath85 represents the particle velocity due to non - keplerian gas motion in the quasi - steady state , because it does not vanish even in the non - turbulence case , @xmath86 .
although turbulence also contributes to @xmath85 , the contribution due to turbulence can be neglected compared with that due to the quasi - steady non - keplerian gas flow , except for strong turbulence case where the quasi - steady flow is not achieved .
as we will show below , the component @xmath87 can be regarded as characteristic amplitude of the velocity dragged by turbulence , although @xmath87 changes with time due to the temporal change of @xmath74 and the assumption of @xmath78 is no more valid in the turbulent case . later in this paper , @xmath85 and @xmath87
will be calculated by equation ( [ eq : turb ] ) , using gas velocity obtained by the mhd simulation , and they will be presented as a function of @xmath4 after azimuthal and vertical averaging . the ratio of the absolute values of the two is @xmath88 the turbulence effect is indicated by this ratio .
it is stronger , if i ) turbulent flow is larger than the angular velocity difference @xmath82 , and/or ii ) the dust size ( @xmath89 ) is smaller . in the annulus where dust particles are expected to accumulate , @xmath90 and
the ratio diverges .
however , we will show below that this divergence is restricted only in narrow regions and the effect of turbulence does not prevent dust accumulation .
we have set the global gas pressure gradient to be @xmath41 so that a direct comparison with the result by @xcite could be performed .
this global pressure gradient , however , is less steep than in the minimum mass solar nebular model @xcite .
a steeper global pressure gradient ( @xmath91 ) has at least three effects in the current context : ( 1 ) the baseline of gas velocity becomes lower than the keplerian velocity and it becomes more difficult to create super - keplerian regions whose outer - edges
accumulate the dust particles .
( 2 ) if a super - keplerian region appears , accumulation towards its outer - edge is faster ( @xmath85 is faster ) for steeper pressure gradient .
this has a positive impact on the dust accumulation efficiency .
( 3 ) the outer - edge of super - keplerian region would change its location according to the value of @xmath17 .
the turbulence level at the focal annulus of dust accumulation varies accordingly .
if the turbulence level is higher , it by itself has a negative impact . in case2
, we have also carried out the runs with @xmath92 and found that the trapping of dust particles does occur and its efficiency is slightly higher .
in the model-@xmath2 , the non - uniform excitation of mri is realized by non - uniform resistivity while the magnetic field is set uniform .
the result of model-@xmath2 nicely illustrates our scenario for dust accumulation .
figure [ fig : eta-3d]a shows the evolution of mri .
the black lines depict the magnetic field lines and the gray scale shows the gas radial velocity .
the unstable region lies between two white lines ( @xmath93 ) .
mri is first excited only in the initially unstable region ( see the plot at @xmath94 ) and significant angular momentum and mass are transported there .
the mri turbulence intrudes into the stable region ( @xmath95 ) .
deep inside the stable region , however , is always undisturbed ( @xmath96 ) because of the rapid dissipation by the enhanced resistivity .
figures [ fig : eta - puy]a and [ fig : eta - puy]b show the radial profiles of the pressure and the angular velocity of the gas , respectively .
the quantities have been averaged azimuthally and vertically .
the sampling times are @xmath97 0 , 40 , and 70 .
the zone between the two vertical dotted lines is the unstable region .
the inhomogeneous mri growth creates the rigid rotation pattern in the middle of the simulation box ( @xmath98 ) .
the pressure distribution is considerably modified such that the resultant pressure gradient force balances with the modified coriolis force .
the flattened rotation profile can not sustain the excitation of mri in the unstable zone .
indeed , the turbulence weakens extremely at @xmath99 in figure [ fig : eta-3d]a .
the profiles of pressure or angular velocity in figures [ fig : eta - puy]a and [ fig : eta - puy]b depict the quasi - steady state set up by the non - uniform mri activity .
most of what we see here is quite similar to the results of the two - dimensional simulations described in paper i even though the initial settings for seed magnetic field and resistivity distribution are totally different .
the rigid rotation causes gas to rotate faster than keplerian velocity in @xmath100 ( figure [ fig : eta - puy]b ) .
this can change the particle migration drastically . figure [ fig :
eta-3d]b shows the temporal evolution of the particle density .
the color code is set such that the maximum is ten times the initial value .
after the particles are swept out of the unstable region by the mri flow , they accumulate to the location at @xmath101 ( note that particles leaving the simulation box from the left hand boundary reenter from the right hand boundary after the shearing box correction is taken into account ) .
though not visible in the panels , the particles initially in the stable zone are swept likewise towards the same location .
the accumulation of particles is most clearly shown in figure [ fig : eta - puy]c in which the radial distribution of the number of particles that is averaged azimuthally and vertically and is normalized by the initial value . to analyze the particle concentration dynamics in more details , in figure [ fig : eta - puy]d
, we plot the maximum ( the solid line ) and the minimum ( the dashed line ) values of @xmath85 at a given @xmath4 .
the radial velocities of particles are predicted by equation ( [ eq : turb ] ) using simulated gas velocity after the establishment of the quasi - steady flow ( @xmath102 ) . since the gas velocity is influenced by remnant turbulence , which has variations dependent on @xmath13 and @xmath5 , the predicted radial velocities have the maximums and the minimums
. figure [ fig : eta - puy]d shows that the turbulent fluctuations are sufficiently small compared with the systematic angular velocity change due to the modulation of the pressure profile . in figure [ fig : eta - puy]d , the max / min of radial amplitude of turbulent velocity , @xmath87 , is plotted . the amount of @xmath87 is smaller than those of @xmath85 except in the region @xmath103 .
thus , @xmath85 represents the typical radial velocity of particles that is induced by head / tail angular wind in the absence of turbulence .
also plotted by dots are the actual particle radial velocities in the simulation result , @xmath104 , which are concentrated near the super / sub - keplerian boundary .
the data are taken at @xmath99 , but it stays essentially the same after the mri saturation at @xmath105 . in the radial regions where the maximum @xmath85 is negative , gas rotation is sub - keplerian for all @xmath13 and @xmath5 . in such radial locations ,
all the particles lose their angular momentum by the drag due to the slower angular wind and migrate inward .
conversely , in the regions where the minimum @xmath85 is positive , all the particles migrate outward . in the middle of the stable zone ( @xmath96 ) ,
the gas flow is hardly changed .
the value of @xmath106 corresponds to the infall speed due to global pressure gradient under the initial condition .
the particles are concentrated to the zone where both min[@xmath107 < 0 $ ] and max[@xmath107 > 0 $ ] are satisfied , which is situated around @xmath108 .
the zone satisfying the two inequalities is rather narrow in @xmath4 , and the difference between the maximum and the minimum @xmath85 in this zone is small , resulting in high concentration of the dust particles at the outer - edge of the super - keplerian zone as shown in figure [ fig : eta - puy]c . because @xmath109 in most of the regions except for the dust concentration zone , and the width of the dust concentration zone is small , the effects of turbulence do not significantly expand the dust concentration zone . in figure
[ fig : eta - clump]a , we plot the time variation of the maximum number of particles per grid .
the more or less monotonic increase leads to the peak density of more than 1,000 times the initial density for @xmath110 .
we will show that the same clump grows in density while keeping its identity , which is needed for subsequent gravitational instability .
we pick up a particular time @xmath111 during the period when the number of dust particles in the densest grid in the whole region increase monotonically or is saturated .
then , we search for the cell which has the largest number of particles in it .
this cell is identified as the center of the most prominent clump that is composed of a number of cells .
the i d numbers of the @xmath112 particles in this cell of the highest density at @xmath113 are recorded and their motions are traced in time ( also backward in time if necessary ) . to determine a new position of the center of the same clump at different times , we inspect the cells within 5 grid distance ( @xmath114 ) from the center of mass of the traced particles .
the cell having the largest number of particles among them is defined to be the new position of the center of the clump .
table [ tab:2 ] shows @xmath115 and @xmath116 for the runs described in this paper .
the time evolution of the number of particles in the cell at the center of the traced clump is plotted by the dotted lines in figure [ fig : eta - clump]a .
it agrees with the maximum number of particles over the whole region , indicating that the same clump keeps its identity and stays to be the most prominent clump .
another way to confirm that the clump steadily holds together is to inspect the velocity dispersion of particles .
figure [ fig : eta - clump]b presents the velocity dispersion of particles as a function of distance from the center of the clump at @xmath117 .
all the compositional velocity dispersions are extremely low ( @xmath118 ) , indicating that the particles are not to be diffused significantly .
when the self - gravity is taken into account , the high mass density and the low velocity dispersion should be the preferable condition for the subsequent planetesimal formation via gravitational instability .
when the global pressure gradient is steeper , the effects on dust dynamics described at the final part of section 3 must be considered . for the case shown here , we make discussion by inspecting figure [ fig : eta - puy]d and equation ( [ eq : turb ] ) . from the maximum amplitude of min[@xmath85 ]
change ( @xmath120 ) , we can expect that the value of min[@xmath85 ] remains positive ( vertically and azimuthally uniform super - keplerian region survives ) if @xmath121 .
that is , dust particles are prevented from infall as long as @xmath121 .
this condition would be marginally satisfied at @xmath122au in the mmsn model .
though the dust concentration that occurs at @xmath123 would be shifted inward with the more downward shift of the @xmath85 ( larger negative @xmath17 ) , the turbulence effect on dust concentration would not change because the turbulence level @xmath87 is evenly low .
therefore , with a steeper global pressure within the range of @xmath121 , the dust particles would drift faster towards and accumulate more efficiently at the outer - edge of super - keplerian region . in the models of case2
, the non - uniform excitation of mri is realized by non - uniform distribution of the magnetic field while resistivity is set uniform . in the presence of the substantial dissipation , the variation of the vertical component of the magnetic field
switches the mri condition from stable to unstable . in case2 ,
the magnetic field angle is varied along the @xmath4 ( radial ) direction such that mri unstable region is situated only in the middle of the simulation box .
the overall mhd dynamics varies according to the spatially averaged magnetic reynolds number ( @xmath11 ) which is controlled by the width of the unstable / stable regions in the simulation box , as shown in paper i. here , the rotation pattern of the magnetic field and the width of the unstable region ( @xmath124 ) is fixed . in the model - s40 ,
the radial width of initial stable region @xmath56 is @xmath125 .
the relatively wide stable region results in @xmath126 , which ensures effective enough dissipation to realize the quasi - steady state .
figure [ fig : s40 - 3d]a presents the evolution of mri in model - s40 .
the gray scale and lines are radial velocity of gas and magnetic field lines , respectively .
the boundaries between the stable and unstable regions are depicted by the two white lines ( @xmath127 ) .
it is found that the results of the present three - dimensional run are similar to the corresponding two - dimensional results described in paper i. the instability is excited only inside the unstable region . a part of the stable area is disturbed as well , because the resistivity is not large enough to dissipate the fluctuations immediately ( @xmath128 ) .
they are , however , eventually dissipated ( @xmath129 and @xmath130 ) .
figure [ fig : s40-puy ] presents radial distributions of pressure and angular velocity of gas , and the vertical component of the magnetic field .
all the quantities obtained at @xmath97 0 , 40 and 70 , are averaged both azimuthally and vertically .
the inhomogeneous mri growth creates the near - rigid rotation pattern in the middle ( @xmath131 ) and the modified coriolis force balances with the newly created pressure gradient force .
the small vertical magnetic field together with the flat angular velocity profile suppresses the growth of mri at later times inside the unstable region in which mri was initially excited .
while these results of model - s40 are similar to those of model-@xmath2 ( case1 ) , three differences are pointed out : i ) the pressure and angular velocity are distributed less smoothly in the stable region in figures [ fig : s40-puy]a and [ fig : s40-puy]b , which is attributed to deeper penetration of the mri activity , ii ) super - keplerian region exists at @xmath132 , in addition to the middle region of @xmath133 , and iii ) the weak instability persists at least until @xmath134 in the stable region near the unstable region ( figure [ fig : s40 - 3d]a ) .
the super - keplerian region at @xmath132 arises , because larger stable region makes mass and angular momentum transport less efficient around the middle of the stable region ( @xmath135 ) . as a result ,
the pressure gradient becomes positive at @xmath132 . in model-@xmath2 , the large magnetic resistivity in the stable region
suppresses the instability even in the region of steep radial gradient of angular velocity .
but , since in the models of case2 , constant magnetic resistivity is assumed , the magnetic field is not completely dissipated in the stable region with the steep radial gradient . figure [ fig : s40 - 3d]b presents the temporal evolution of the particle density .
the evolution of the particle density pattern is similar to case1 .
however , unlike case1 , not all the particles are assembled to the focal annulus .
figure [ fig : s40-dust]a shows the azimuthally and vertically averaged value of the number of particles as a function of @xmath4 . as expected , particles are condensed at @xmath136 ( the outer - edge of the super - keplerian region ) .
figure [ fig : s40-dust]a also shows some other peaks in the stable region , which reflects the characteristics i ) and ii ) of the flow pattern described above . figure [ fig : s40-dust]b shows radial velocities of particles predicted by equation ( [ eq : turb ] ) using simulated gas velocity after the establishment of the quasi - steady flow ( @xmath102 ) .
the dots concentrated at @xmath136 in the figure are actual velocities of all the simulated particles .
the actual data of the particle velocities are well reproduced by equation ( [ eq : turb ] ) , since most of the data points are located between the predicted minimum and maximum values .
we further decompose @xmath137 into @xmath85 and @xmath87 according to equation ( [ eq : turb ] ) .
figure [ fig : s40-dust]c presents the radial distribution of the two components .
the maximum and minimum values for each component are plotted .
it is shown that the effect of the remnant turbulence is small and overall features of particle radial migration is represented by @xmath85 . due to the non - smooth pressure and velocity distributions , @xmath138 \simeq { \rm min}[v_{\rm f}]>0 $ ] at @xmath139 and @xmath140 , in addition to @xmath133 .
it means that all particles have positive radial velocity there and accumulate near the outer edges of these regions . on the other hand , @xmath141 \simeq { \rm max}[v_{\rm f}]>0 $ ]
is observed in some areas , for example , around @xmath142 .
a small number of particles are stalled there .
the peaks of particle density at @xmath143 would not grow because there are few particles which can accumulate from outside ( figure [ fig : s40-dust]a ) .
the stagnant areas can not stall all the particles which migrate inward due to @xmath138<0 $ ] ( figure [ fig : s40-dust]b ) .
therefore the peak at @xmath136 is the most prominent concentration zone .
figure [ fig : s40-clump]a shows the time variation of the number of particles in the densest grid in the whole region and at the center of the traced clump .
the agreement between the two indicates that the same clump is growing .
figure [ fig : s40-clump]b is the velocity dispersion of particles within the clump at @xmath144 .
the velocity dispersion is small but is larger than in model-@xmath2 .
while the turbulence does not affect the location of the dust concentration , it does influence the degree of particle accumulation .
when the azimuthal box size is long , shear velocity becomes supersonic near the edge and it may cause artificial density dip @xcite unless a high - order scheme is applied for keplerian advection term . to confirm that the density dip found in our simulations is not caused by such a numerical error , we also carried out a run in which the unstable region is put in the side areas and the stable region is at the center .
the other parameters are the same as model - s40 .
we found that the density dip is formed at the side area and no artificial density dip is created in the central area , which means that our numerical scheme using cip method with the staggard mesh and the short timestep ( @xmath145 ; courant condition is set to be @xmath146 ) is sufficiently high - order scheme .
let us discuss the effects of a steeper global pressure gradient .
following the same argument as in the final part of section 4.1 , from the value of @xmath85 or @xmath147 , a super - keplerian region would remain unless the global prssure gradient is as anomalously large as @xmath148 . to make a more quantitative argument , as was done in @xcite
, we also carried out a run with the pressure gradient of @xmath92 ( model - s40b ) .
the other parameters are the same as model - s40 .
figure [ fig : s40b - drift ] shows the close - up of the distribution of @xmath149 $ ] and @xmath150 $ ] . with the higher negative value of @xmath17 , gas rotates more slowly and the stronger headwind deprives the dust particles of angular momentum more efficiently to accelerate their inward migration in the sub - keplerian region . since @xmath87 does not depend on @xmath17 , the value of @xmath151 exceeds @xmath152 by a larger degree in outer regions .
thus greater amounts of dust particles move to the dust concentration area without becoming locally stagnant due to non - smooth gas pressure and gas velocity fluctuations in the sub - keplerian region . on the other hand , in the dust concentrated area ,
the dust particles are affected by turbulence more severely . with more downward shift of the @xmath85 value in the whole region , the dust concentration that occurs at @xmath123
is shifted inward to @xmath153 from @xmath154 .
the magnetic turbulence is stronger at this new location and it increases the radial width ( @xmath155 ) of the dust accumulated region ( denoted by @xmath141>0 $ ] and @xmath138<0 $ ] : the gray hatched areas in the figure from @xmath156 to @xmath157 and doubles the velocity dispersion of particles there . to see the overall consequences of these both positive and negative effects ,
the number of particles in the most dense grid cell is plotted in figure [ fig : s40-clump]a together with the result of model - s40 . despite the different dust radial velocity and gas turbulence ,
the density increase rate and the peak value are only slightly different from each other , although the new run could show higher dust density on longer timescale .
the effects of turbulence are larger for particles with shorter @xmath158 .
we have run a case with @xmath159 ( model - t01 ) . except for the friction time
, all the settings are the same as model - s40 .
since there is no kick - back from the dust particles to the mhd fluid , mhd results are exactly the same ( figure [ fig : s40 - 3d]a and figure [ fig : s40-puy ] ) .
the difference in the dust concentration pattern is caused by the enhanced gas drag .
figure [ fig : s40 - 3d]c presents the evolution of the particle density .
the difference from the previous case that becomes visible at the later time is that the smaller dust particles are scattered over a broader area .
figure [ fig : t01-dust]a shows the radial profile of the dust number density averaged azimuthally and vertically . since the mhd flow patterns are exactly the same , the prominent density peaks are located at the same position as in model - s40 . however , the peak values are lower and the radial widths are wider . furthermore , more particles stay outside the peaks and are scattered in the stable region .
these differences are caused by the enhanced turbulent diffusion of dust particle motions through enhanced gas drag .
indeed , figure [ fig : t01-dust]c shows that the turbulent scattering @xmath87 dominates over the steady migration @xmath85 in most areas of the simulation box .
figure [ fig : t01-dust]b shows the estimated radial velocity of particles @xmath137 .
the radial velocity of the simulated particles , @xmath104 , are also shown by dots .
although dust particles are swept out of the unstable region , in most parts of the unstable region , max[@xmath160 > 0 $ ] and min[@xmath160 < 0 $ ] are satisfied , so the dust particles are broadly distributed in the stable region with lower concentration .
nevertheless , identity of the clump with the highest dust concentration is still maintained ( figure [ fig : t01-clump]a ) .
indeed the velocity dispersion around center of the traced clump shown is not increased significantly ( figure [ fig : t01-clump]b ) .
we also run a high global pressure gradient model ( @xmath92 ) .
though the inward migration of dust particles becomes faster on average , the maximum density is unchanged because the remnant turbulence that scatters the dust particles in the dust concentrated region is the dominant factor for this @xmath159 case .
recent works ( e.g. , * ? ? ?
* ) proposed that turbulence itself concentrates dust particles in turbulent eddies . in their models ,
lifetime of the eddies has to be longer than the timescale to form dense enough particle clumps for formation of planetesimals . on the other hand
, our model proposes another path to accumulate dust particles that turbulence plays a role in transformation of the nearly keplerian gas flow into the quasi - steady flow with a local rigid rotation region .
because the dust particles accumulate near the outer edge of super - keplerian parts produced by the rigid rotation and the flow pattern is quasi - steady , the timescale problem does not exist in our model .
actually , in the case of model - s40 and model - t01 with @xmath161 , the same clump with the highest density is maintained until the end of simulations ( @xmath162 ) .
however , in our model , as @xmath11 increases to unity , stronger residual turbulence may destroy the clumps , although the clumps are repeatedly created , which may inhibit planetesimal formation . here
, we examine the cases with larger @xmath11 ( but still @xmath73 ) to find the critical value of @xmath11 for persistent clumping . in model - s055 ,
the width of the stable region @xmath56 is one eighth of model - s40 .
accordingly , @xmath69 compared to 0.096 of model - s40 .
figure [ fig : s055 - 3d]a shows that the mri turbulence intrudes into the stable region and covers the whole region without being dissipated .
by @xmath163 , the instability is weakened .
although a clear rigid rotation is not formed inside the unstable region ( figure [ fig : s055-puy]a ) because of the stronger turbulence that remains especially in the stable region , the velocity gradient becomes small inside the unstable region and the super - keplerian zone is formed . as a result ,
the dust particles are swept out of the unstable region and accumulate near the outer stable / unstable boundary ( figure [ fig : s055 - 3d]b ) . however , because the unstable region occupies larger fraction of the simulation box , the resultant magnetic field diffused out of the unstable region is larger than in the previous models .
these lead to stronger remnant turbulence .
figure [ fig : s055-clump]a shows that the number of particles in the densest grid of the whole region does not agree with that at the center of the traced clump .
it means that the clumping is only tentative , although clumps are repeatedly created .
the diffusive nature of the clumps in model - s055 is depicted by the larger velocity dispersion shown in figure [ fig : s055-clump]b .
thin lines in figure [ fig : s055-clump]a shows the clump identity test for model - s11 ( @xmath67 ) .
it shows that the clumps are not persistent as well .
therefore , we conclude that persistent clumps of dust particles are formed for @xmath164 . if the global pressure gradient is steeper , the radial velocity of dust particles is faster and its effect may seem to compete better with that of gas turbulent motion .
however , from figure [ fig : s055-puy]d , the minimum value of @xmath85 becomes negative in the entire regions if @xmath165 and no particles can be accumulated at the boundary of super / sub - keplerian regions .
there is not much parameter space where more dust concentration can be expected .
we have investigated the dust concentration in protoplanetary gas disks that are in the quasi - steady state created by inhomogeneous magnetorotational instability .
we set the inhomogeneous instability with the initial radial non - uniformity of either ionization degree or vertical component of the magnetic field .
the result of our local three - dimensional resistive mhd simulations with lagrangian particles clearly shows that the meter or decimeter sized dust particles are concentrated strongly and steadily in either setting except for the cases with @xmath166 ( where @xmath11 is the averaged magnetic reynolds number averaged over the simulation box ) , which can allow planetesimal formation via gravitational instability .
the mri growth in our model generates the locally confined super - keplerian region and this flow is quasi - stable due to the support of gas pressure gradient which is also formed by the mri . because unperturbed gas flow is sub - keplerian due to global pressure gradient , the particles suffering gas drag are concentrated in the keplerian domain at the outer - edge of the super - keplerian area .
since the flow is in a quasi - steady state , particles are supplied from outer regions by gas drag migration and the particle density increases significantly .
the process of particle concentration and the increasing rate of dust density depend on the particle size , the initial radial widths of unstable / stable regions , and the initial non - uniformity setting as follows : _ the initial non - uniform setting , magnetic field or resistivity _ : : : the non - uniform resistivity model produces more stable and clean state of local rigid rotation than the non - uniform magnetic field model , because the high resistivity dissipates the magnetic perturbations more rapidly in the stable regions . indeed , the velocity dispersion of particles in clumps is smaller in the non - uniform resistivity model . in the non - uniform magnetic field cases ,
the magnetic reconnection and the remnant weak instability create less smooth gas velocity field , so that particles tend to assemble at multiple radial positions . on the other hand , in the non - uniform resistivity case ,
all particles are concentrated at a specific radial position , and the peak values of the particle density in clumps is higher , although the particle density also depends on the size of numerical box . _
the dust size , meter or decimeter _ : : : the meter - size particles are less coupled with gas motion than decimeter - size ones . since their radial migration is faster , the meter - sized particles are more likely to be condensed in the radially narrow area without being scattered by the remnant gas turbulence . the peak dust density becomes @xmath167 times larger than the initial value . even in the decimeter - size particle cases ,
the velocity dispersion of particles in the clump excited by the remnant turbulence is sufficiently smaller than radial drift velocity due to the quasi - steady gas flow and dust concentration is still observed in the simulations , although the enhanced dust density is at most @xmath168 . _ the degree of remnant turbulence _ : : : when the stable region is initially smaller in the non - uniform magnetic field model ( case2 ) , that is , @xmath11 is larger , stronger turbulence tends to remain . for @xmath169
, however , the super - keplerian area continues to exist and clumps with density @xmath168 times larger than initial values are robustly created .
furthermore , for @xmath164 , the clumps are persistent enough for following gravitational instability , while they are tentative and repeatedly produced in the cases of @xmath170 . _ consistent magnetic resistivity _ : : : we need to calculate the value of magnetic resistivity consistently as a time dependent value , because it is sensitive to both gas and dust densities . around strong dust concentration areas ,
gas and dust density become larger and magnetic resistivity becomes higher , which may calm down the remnant turbulence .
_ rossby wave instability ( rwi ) _ : : : rwi can be caused from a pressure bump on the radial - azimuthal plane . if we simulate in an enlarged box , vortex may grow , although in the non - uniform magnetic field cases , the growth of the instability may be reduced by the azimuthal magnetic field . if rwi occurs , it may serve as a mechanism of dust concentration in the azimuthal direction @xcite . _ the vertical structure of a protoplanetary disk _ : : : in the upper disk regions , ionization degree may be so high that the mri to occur @xcite .
the turbulence there may affect the small dust dynamics near the midplane .
_ feedback from dust particles onto gas _ : : : the feedback may be the most important .
a great concentration in our results here will inevitably change gas velocity field and it could affect the condition of dust concentration ( density enhancement , velocity dispersion and so on ) and planetesimal formation .
we have already started simulations with this effect .
_ collisional destruction _ : : : while the velocity dispersions of dust particles in clumps are extremely low in some of our results , the dust particles can be collisionally disrupted into small fragments that are coupled to gas motion and the fragments may be dispersed by even weak turbulence . for clumps
to gravitationally collapse , its timescale needs to be smaller than mean collision time . on the other hand , if the dust grains are also produced by the collisions , they might affect the gas ionization state through recombination of electrons onto their surface .
we need to address these feedback effects by collisional disruption as well .
we thank the referees for insightful and helpful comments .
this work was supported by grant - in - aid for jsps fellows ( 208778 ) .
the simulations presented in this paper were performed by nec sx-6 at isas / jaxa .
adachi , i. , hayashi , c. , & nakazawa , k. 1976 , prog .
phys , 56 , 1756 balbus , s. a. , & hawley , j. f. 1991 , , 376 , 214 balbus , s. a. , & hawley , j. f. 1998 , reviews of modern physics , 70 , 1 balsara , d. s. , tilley , d. a. , rettig , t. , & brittain , s. a. 2009 , , 397 , 24 barge , p. , & sommeria , j. 1996 , , 295 , 1 chiang , e. i. , & murray - cley , r. a. 2007 , nature phys . , 3 , 604 cuzzi , j. n. , dobrovolskis , a. r. , & champney , j. m. 1993 , , 106 , 102 dubrulle , b. , morfill , g. , & sterzik , m. 1995 , , 114 , 237 fleming , t. p. , stone , j. m. , & hawley , j. f. 2000 , , 530 , 464 fleming , t. , & stone , j. m. 2003 , , 585 , 908 fromang , s. , & papaloizou , j. 2006 , , 452 , 751 gammie , c. f. 1996 , , 457 , 355 glassgold , a. e. , najita , j. , & igea , j. 1997 , , 480 , 344 goldreich , p. , & ward , w. r. 1973 , , 183 , 1051 gomez , g. c. , & ostriker , e. c. 2005 , , 630 , 1093 goodman , j. , & pindor , b. 2000 , , 148 , 537 hayashi , c. , 1981 , prog .
70 , 35 hawley , j. f. , gammie , c. f. , & balbus , s. a. 1995 , , 440 , 742 igea , j. , & glassgold , a. e. 1999 , , 518 , 848 ilgner , m. , & nelson , r. p. 2008
, , 483 , 815 inaba , s. , & barge , p. 2006
, , 649,415 ishitsu , n. , & sekiya , m. 2003 , , 148 , 537 jin , l. 1996 , , 457 , 798 johansen , a. , klahr , h. , & henning , th . 2006a , , 636 , 1121 johansen , a. , henning , t. , & klahr , h. 2006b , , 643 , 1219 johansen , a. , & youdin , a. n. 2007 , , 662 , 627 johansen , a. , oishi , j. s. , mac low , m ,- m . ,
klahr , h. , henning , th . , & youdin , a. 2007 , , 448 , 1022 johansen , a. , youdin , a. , & klahr , h. 2009 , , 697 , 1269 johnson , b. m. , guan , x. , & gammie , c. f. 2008 , , 177 , 373 kato , t. m. , nakamura , k. , tandokoro , r. , fujimoto , m. , & ida , s. 2009 , , 691 , 1697 ( paper i ) li , h. , colgate , s. a. , wendroff , b. , & liska , r. 2001 , , 551 , 874 lyra , w. , johansen , a. , klahr , h. , & piskunov , n. 2008 , , 491 , l41 miyake , k. , & nakagawa , y. 1995 , , 441 , 361 nakagawa , y. , sekiya , m. , & hayashi , c. 1986 , , 115 , 304 oishi , j. s. , mac low , m - m .
, & menou , k. 2007 , , 670 , 805 papaloizou , j. c. b. , & terquem , c. 1997 , , 287 , 771 pneuman , g. w. , & mitchell , t. p. 1965
, , 4 , 494 safronov , v. s. 1969 , evolution of the protoplanetary cloud and the planets , nasa tech .
f-677 sano , t. , inutsuka , s. , & miyama , s. m. 1998 , , 506 , l57 sano , t. , & miyama , s .
m. 1999 , , 515 , 776 sano , t. , miyama , s. m. , umebayashi , t. , & nakano , t. 2000 , , 543 , 486 stone , j. m. , & norman , m. l. 1992a , , 80 , 753 , 1992b , , 80 , 791 umebayashi , t. 1983 , prog .
phys . , 69 , 480 varniere , p. , & tagger , m. 2006 , , 446 , l13 wardle , m. , & ng , c. 1999 , , 303 , 239 weidenschilling , s. j. 1977 , , 180 , 57 weidenschilling , s. j. 1980 , , 44 , 172 wisdom , j. , & tremaine , s. 1988 , , 95 , 925 yabe , t. , & aoki , t. 1991 , comput .
comm . , 66
, 219 youdin , a. n. , & goodman , j. 2005 , , 662 , 613 youdin , a. n. , & johansen , a. 2007 , , 662 , 613 .run parameters ( 1 ) : name of run .
( 2 ) : size of simulation box . ( 3 ) :
grid resolution .
( 4 ) : parameter set non - uniformly . ( 5 ) : radial width of unstable region . ( 6 ) : radial width of stable region . ( 7 ) : friction time .
( 8) : number of particles . [ cols="<,^,^,^,<,<,<,<,<",options="header " , ] | how to create planetesimals from tiny dust particles in a proto - planetary disk before the dust particles spiral to the central star is one of the most challenging problems in the theory of planetary system formation . in our previous paper @xcite
, we have shown that a steady angular velocity profile that consists of both super and sub - keplerian regions is created in the disk through non - uniform excitation of magneto - rotational instability ( mri ) .
such non - uniform mri excitation is reasonably expected in a part of disks with relatively low ionization degree . in this paper
, we show through three - dimensional resistive mhd simulations with test particles that this radial structure of the angular velocity indeed leads to prevention of spiral - in of dust particles and furthermore to their accumulation at the boundary of super - keplerian and sub - keplerian regions . treating dust particles as test particles ,
their motions under the influence of the non - uniform mri through gas drag are simulated . in the most favorable cases ( meter - size dust particles in the disk region with a relatively large fraction of mri - stable region ) , we found that the dust concentration is peaked around the super / sub - keplerian flow boundary and the peak dust density is 10,000 times as high as the initial value .
the peak density is high enough for the subsequent gravitational instability to set in , suggesting a possible route to planetesimal formation via non - uniformly excited mri in weakly ionized regions of a disk . |
in many situations there is a need to represent a given quantum state or a probability distribution of states in terms of a simple basis which can be easily handled in further processing , e.g. , for integration in matrix elements .
gaussians form such a basis although it is overcomplete .
gaussians have the additional advantage that they can adopt the form of coherent states in phase space , thereby supplying a direct link to classical mechanics .
hence , the formulation in terms of gaussians is also very useful for a mixed quantum - classical or quasiclassical description of a system . in particular the evolution of systems with many degrees of freedom
can often be described within classical molecular dynamics only . here
, one is interested in obtaining optimal initial conditions by translating the initial quantum state of the system into a suitable classical phase space distribution which can be classically propagated @xcite .
while all quantum information on an amplitude level is lost in this approximation , we will introduce a representation which can preserve basic quantum interferences by construction even in the case of the purely classical propagation .
finally , it is of general and basic interest to reconstruct amplitudes or the density matrix , i.e. , coherent information , from phase space distributions .
this has been done recently in terms of the moments in time of the wigner distribution @xcite .
although there is no principal need to do so we will concentrate on the wigner distribution due to the attention it has received recently in different areas . in several experiments
the wigner distribution of microscopic systems has been measured @xcite . beyond the quantum context , in the time - frequency phase space of signal processing , analysis of the wigner distribution @xcite
is an important tool @xcite , as for optics @xcite , speech analysis @xcite or the monitoring of machinery conditions @xcite .
the decomposition of a given state @xmath1 into coherent states can be regarded as a particular form of phase space representation .
labeling positions by @xmath2 or @xmath3 and momenta by @xmath4 or @xmath5 , we translate the ground state of the harmonic oscillator by @xmath6 to obtain the coherent state in position representation , i.e. its wavefunction is @xcite @xmath7 for simplicity , we chose unit frequency ( @xmath8 ) and mass ( @xmath9 ) for the harmonic oscillator without loss of generality .
( often the complex variable @xmath10 is used and the constant phase @xmath11 is omitted ) . even though the coherent state basis is overcomplete , the exact representation , @xmath12 so that @xmath13 to shorten the notation .
( with the second equation we also define the symplectic matrix @xmath14 ) .
it follows that one can approximate any given state by a finite expansion @xmath15 by minimizing the error , defined in some convenient way .
of course , it is much better to use some features of @xmath1 as a preliminary guide to where the @xmath16 should be located .
an obvious necessary condition for a good fit is that the coherent state intensity @xmath17 in @xmath18 be well approximated by @xmath19 for each phase space point @xmath20 .
but note that , if the diagonal terms in this expression are only appreciable for @xmath21 very close to @xmath20 , a nondiagonal term decays as a product of gaussians centered on a pair of separate phase space vectors .
such a term is negligible for all @xmath20 unless @xmath21 is very close to @xmath22 .
moreover , such nondiagonal terms are not positive definite , which justifies our performing the preliminary fit @xmath23 where @xmath24 the factor @xmath25 has here been introduced so that the left hand side becomes the husimi function @xcite , i.e. , the husimi representation of the state @xmath1 , also known as the @xmath3-function in quantum optics @xcite . here , @xmath26 is then decomposed into gaussians phase space packets ( gpps ) which are husimi representations of the coherent states , specified by @xmath18 .
the husimi function in its own right uniquely defines a quantum state , though part of the quantum phase information is locked up into its delicate analytic properties .
these are hard to translate into other representations , but we propose to sidestep this problem .
our method is to make a preliminary fit of the husimi function @xmath27 , which is real and positive definite , and therefore much simpler than fitting the complex coherent states directly in @xmath28 .
however we can now identify the coefficients of the husimi fitting as being proportional to the square modulus of the coefficients of the superposition of coherent states @xmath28 .
the phases of the coefficients are then determined by relating the husimi function to the coarse - graining of the wigner function , i.e. we bring in yet a third phase space representation of the state @xmath1 .
it should be pointed out that , once the state @xmath1 has been fitted by a finite number of coherent states , it can be viewed in any representation . the special attention given here to the wigner function
is justified by the way that it highlights quantum interferences .
in fact such interferences are sometimes mimicked by extra gaussians in the context of numerical work with classical densities @xcite .
thus , in the next section we present the wigner function , its relation to the husimi function and its specific form for an arbitrary superposition of gaussians .
in addition , we discuss and establish the differences between the representation of quantum states such as the wavefunction , the wigner distribution and the husimi distribution using these gaussians .
the wigner distribution sets the stage for the delicate procedure of fitting the phases , i.e. , to determine the appropriate interference fringes between the gaussians as detailed in section 3 .
in this section we propose a method for determining the phase of the expansion coefficients by fourier analysing the wigner function .
actually , this resulting _ chord function _ or _
characteristic function _ ( in quantum optics @xcite ) can be derived directly from the initial wave function , without obtaining first the wigner function .
the husimi function , on the other hand , supplies a robust overall picture of the quantum states which can be used to place the gaussians .
this will be discussed in section 4 using explicit examples for illustration .
emphasis will be put on general symmetry criteria for placing the gaussians and numerical demonstrations based on monte carlo fitting of sampled wavefunctions . finally in section 5 ,
we semiclassically interpret the results of section 4 , namely the increase of gaussians needed to represent quantum states @xmath29 of a system with increasing quantum number @xmath0 .
a summary in section 6 concludes the paper . for simplicity ,
all our formulae are presented for systems with a single degree of freedom , but they are easily generalized to high dimensional systems .
the state @xmath1 is uniquely described by the real function of phase space points @xmath30 , @xmath31 known as the wigner function @xcite .
inversion of this symmetrized fourier transform leads back to the wavefunction , while its intensity results from the projection @xmath32 , its projection is always positive .
actually , this is just a particular case of the general relation , @xmath33 and @xmath1 .
thus , introducing the wigner function for the coherent state @xmath34 , @xmath35 we obtain the husimi representation of @xmath1 as @xmath36 this shows that the positive definite husimi function results from the gaussian smoothing of the wigner function . inserting the coherent state expansion @xmath1 given by @xmath28 into the wigner
transform ( [ e21 ] ) , we obtain the well known form of the corresponding wigner function , @xmath37 where we define the crossed wigner function ( or moyal bracket @xcite ) @xmath38 here we use the abbreviated relations @xmath39 and @xmath40 , as well as the skew product @xmath41 . comparing @xmath42 with @xmath43 we find the same gaussian form for the diagonal terms of the wigner function as for the husimi function , but the offdiagonal terms are no longer negligible .
for the wigner function , any pair of coherent states determines a third gaussian centered halfway between their two centres , of comparable amplitude to the diagonal contributions of the individual gaussians .
this interference term is only cancelled by the smoothing employed to obtain the husimi function , because of the phase oscillations .
their wavelengths in phase space are @xmath44 , so that the husimi function is increasingly damped with growing separation of these gpps .
this is an excellent illustration of the way that the husimi function highlights the classical structure in a quantum state , while it hides quantum phase information .
this question will be further developed in section 4 .
of course the phase information is still contained in the small offdiagonal terms in @xmath43 , because the husimi function is fully quantum mechanical , but this information is certainly more accessible in the wigner representation .
we can now understand the differences between using the position representation , i.e. the wave function , the wigner representation or the husimi representation , to fit a given state by a set of coherent states .
the position representation has the advantage that there are no interference terms between pairs of coherent states .
however , if the state is highly oscillatory , such as a highly excited eigenstate of a hamiltonian , this needs to be fitted by coherent states with large momenta @xmath5 , which also have narrow oscillations .
clearly the fitting procedure becomes very unstable with respect to small errors in the initial function if large momenta are necessary .
in contrast , the diagonal terms in the expansion of the wigner function are smooth gaussians distributed over phase space , but again we obtain increasingly narrow oscillations for the interferences around the midpoint of each pair . of course
, the oscillations of the wave function and the wigner function are not unrelated , since the latter have the same wavelength as the wave intensity in the case of a pair of gaussians located at ( @xmath45 ) .
finally , the husimi function washes out oscillations , the more effectively the tighter they happen to be .
we shall review in section 4 how the smooth phase space distribution singles out the basic classical structure within the quantum state , which can be stably fitted by real gpps .
the cost is that only the location of the gaussians in phase space and the square modulus of their coefficients are then determined . in the next section we present a method for retrieving the phases of the coefficients by taking the fourier transform of the wigner function
both , the direction and the wavelength of the fringes that modulate the interference peak halfway between two gpps at @xmath21 and @xmath22 in the wigner representation , are entirely determined by @xmath46 .
however , these fringes are translated by changing the phases of the coefficients @xmath47 and @xmath48 , which we wish to determine .
clearly , the best way to recover these phases is to take the fourier transform of the wigner function : @xmath49 it is also possible to express this directly from the wave function @xmath50 which resembles the wigner transform @xmath51 .
this function is known as the woodward ambiguity function in communications theory @xcite or the characteristic function in quantum optics @xcite .
usually the sign of @xmath52 is reversed in the definition of @xmath53 , but here we follow @xcite , where it is simply called the chord function .
a justification for this term appears immediately if we consider the case of a superposition of gaussians that is our concern : @xmath54 where @xmath55 and @xmath56 thus all the diagonal terms collapse onto gaussians centered on @xmath57 , whereas each interference term is a gaussian centered on the chord joining @xmath21 and @xmath22 .
( both directions @xmath58 are present , leading to symmetric contributions in chord space . indeed , we must have @xmath59 ) .
let us now suppose that a given chord @xmath46 is sufficiently far from all the other chords , so that at @xmath60 the chord function is dominated by the single nondiagonal contribution @xmath61 , i.e. @xmath62 then , defining @xmath63 and recalling that we have already fitted @xmath64 from the husimi function , we obtain @xmath65 from @xmath66 of course , the modulus of both sides of @xmath67 should be approximately equal , which provides a check on the previous husimi fitting .
the diagonal sum that is substracted from @xmath68 in @xmath67 decays exponentially with @xmath69 , so it will only affect the phases of the smaller chords .
it might appear to be only consistent with our previous approximation of the husimi function to neglect the diagonal sum in ( [ e36 ] ) , but this would perturb the phases of small chords .
so far , nothing has been said about how to place the gaussians in phase space that are meant to approximate a given husimi function .
an obvious criterion is that any relative maximum of this smooth distribution should also receive a coherent state .
further knowledge of the state should also be used .
for instance , if it is known to be the eigenstate of a given hamiltonian @xmath70 with energy @xmath71 , then the semiclassical considerations in the following section allow the restriction of the gaussian centers to @xmath72 .
of course , the fitting procedure is only useful for cases in which @xmath70 is unknown , or else @xmath1 may itself be a linear superposition of states , such as result from a two slit experiment . here
we immediately feel the advantage that the decomposition @xmath28 is defined in terms of the state itself , rather than the density operator @xmath73 .
thus , the linearity of @xmath74 allows us to fit @xmath75 and @xmath76 independently and then to superpose the fitted state as @xmath77 . a further simplication results from possible symmetries .
for instance , for the eigenstate of an even potential , one should place the gaussians symmetrically at @xmath78 with @xmath79 , so as to guarantee even or odd states .
again , this is a very particular case , but time reversal invariance is much more common .
if the state @xmath1 resulted from initial real wave functions and the evolution proceeded through interactions that preserve time reversal symmetries ( in usual practice , if there are no magnetic fields ) then the final state can also be real . in this case
, one must choose symmetric pairs of gaussians at @xmath80 with equal real coefficients , @xmath81 , so as to ensure that @xmath1 is also real .
this halves the number of independent coherent states to be fitted and reduces the choice of phase to either zero , or @xmath82 .
any further knowledge should generally be used to reduce the randomness of the positions of the coherent states to be fitted to @xmath1 .
the semiclassical considerations in the following section may be a further guideline .
when this knowledge is exhausted , the best course is to optimize random guesses by monte carlo fitting of arbitrary gaussians , within the given constraints .
specifically , in order to measure the quality of the fitting , we define the relative error @xmath83 as the mean square deviation between the known husimi function @xmath84 of a quantum state @xmath85 and its fitting to the superposition of gpps defined in ( [ e19 ] ) .
this error @xmath86 has to be minimized for a set of @xmath87 grid points @xmath88 . to fit this superposition of gpps as closely as possible to @xmath84 ,
we proceed as follows : ( i ) the centers @xmath89 of the coherent states are obtained by monte - carlo sampling the region where the distribution is significant ; that is the region where the distribution exceeds a certain threshold @xmath90 ; ( ii ) the centers must not be too close , this means a minimum distance @xmath91 between centers is fixed a priori , thereby avoiding overlaps and reducing the total number @xmath92 of gpps used for the representation ; ( iii ) starting with @xmath93 we increase n by one and repeat the process until the desired accuracy is reached . since only the coefficients @xmath94 are fitted , minimizing the error @xmath95 reduces to a problem of linear optimization which is equivalent to solving the real matrix equation @xmath96 where the phase space points @xmath97 that define the matrix elements @xmath98 are randomly selected . here
, @xmath99 contains the coefficients to be determined and @xmath100 those of the known husimi function . in general
, the matrix @xmath101 is expected to be sparse and a variety of special algorithms can be used to solve ( [ e42 ] ) efficiently @xcite . in principle one
may also consider to optimize this fitting with respect to the widths of the gpps ; but since the husimi function is always smooth and positive definite , it is convenient to keep them frozen . in
what follows , we have tested our method on the fitting of the eigenstates of the harmonic oscillator ( ho ) and that of the morse oscillator ( mo ) supporting 18 bound states @xcite .
we have not used the knowledge that these states have time reversal symmetry in the way mentioned earlier , so we should obtain results of comparable quality even for states with complex wave functions .
figure 1 shows , on the top , the known husimi function for the 3rd excited state of ho ( left ) and for the 5th excited state of mo ( right ) , with parameters @xmath102 . on the bottom
the corresponding fitted distributions are shown .
we found that @xmath103 is sufficient to reproduce quite well the husimi function of ho whereas 14 coherent states ( gpps ) are needed in the mo example , to achieve a global relative error of @xmath104 .
the respective values for the threshold @xmath90 and the minimum distance @xmath91 are given in the caption of figure 1 .
this preliminary fit on the husimi function provides us with the total number @xmath92 of gpps , their location in phase space @xmath105 and the expansion coefficients @xmath94 .
these coefficients are related to those of the coherent states @xmath48 and their phases @xmath106 by @xmath107 up to this point , the phases @xmath106 are still unknown .
we only need to determine the phases differences @xmath108 in ( [ e26 ] ) which can be easily extracted from ( [ e38 ] ) .
this says that , for @xmath92 gpps , @xmath109 chords ( @xmath108 ) are required to reproduce the wigner function .
these @xmath109 pairs of gaussians ( chords ) are symmetrically spaced about @xmath57 , so that , being more dense , they overlap more than the original set of @xmath92 gaussians in phase space .
figure 2 shows on the top the wigner function of the 3rd excited state of the ho together with its corresponding fit obtained with @xmath110 gpps .
the present fitting reproduces the wigner function everywhere in phase space , even for negative regions which are signatures of quantum interference .
this is clearly illustrated in figure 3 where the deviations of the monte carlo sampled wigner function , less than @xmath111 everywhere in phase space , are depicted for both , the ho ( left , @xmath112 ) and the mo ( right , @xmath113 ) . the global relative error @xmath95 is subsequently found not to exceed @xmath114 .
moreover , we have computed the number @xmath92 of gpps needed to represent the wigner functions of the eigenstates of our test systems , as can be seen in figure 4 .
one sees that @xmath92 grows proportionally to @xmath115 as the vibrational quantum number @xmath0 increases . in the following section
, we provide a semiclassical justification of this result by means of the wkb quantization .
a semiclassical state @xmath116 is supported by a curve in phase space , in the case of one degree of freedom , or in general by a surface with half the phase space dimension @xcite .
if the curve or surface is closed , it must be bohr - quantized and it will be symmetric about @xmath117 in the case of time reversal symmetry . in the case of the position representation , for each branch of the curve ( surface ) , @xmath118 , one defines the action @xmath119 leading to the generalized wkb wavefunction , @xmath120 the amplitudes @xmath121 can also be expressed in terms of the actions , but the important point is that they are purely classical so that the only @xmath122-dependence occurs in the exponential .
thus locally , for any small range of positions , the semiclassical wave function reduces to a superposition of plane waves characterized by the wave - vectors @xmath123 . the semiclassical wigner function is also defined in terms of an action @xmath124 with respect to the classical curve ( surface ) . but instead of the area between the curve and the @xmath2-axis , we are now concerned with the area sandwiched between the curve and one of its chords .
the latter , @xmath125 , is selected by the property that it is centred on the point @xmath126 , as shown in figure 5 .
thus , for one degree of freedom @xcite , @xmath127 with straightforward generalizations @xcite .
the important point is that @xcite @xmath128 where we use the symplectic matrix @xmath129 defined in ( [ e15 ] ) .
thus the semiclassical state is again represented by a superposition of waves , @xmath130 but now these are _ phase space waves_. comparison of this expression with ( [ e36 ] ) immediately reveals that these waves have exactly the same wave vector as a pair of gaussians at the tip of each chord @xmath125 .
it follows that an important feature of a semiclassical state is automatically reproduced by fitting it with gaussians precisely placed on the corresponding classical curve ( surface ) .
notice that , in fact , this is a major a priori obstacle to performing the fitting , because the oscillations near the midpoint of a long chord are very fine for @xmath131 .
one would need very narrow gaussians indeed to fit these phase space waves directly , rather than having them arise naturally as interferences .
all that is left to determine the phase : @xmath132 .
how can we single out the region in which to place the gpps , if we have no a priori knowledge of the classical structures correponding to a quantum state ?
the obvious course is to smooth the wigner function with a gaussian window . in other words
, we initially fit the husimi function . recalling that the gpps have linear width of order @xmath133 , while the wavelength of the wigner oscillations are @xmath134
, it follows that this gaussian window erases effectively a semiclassical wigner function , except in the limit of very small chords .
this is a simple explanation of the well known fact that only in the neighbourhood of the classical curve ( surface ) itself is the semiclassical husimi function appreciable . in the simple case of one degree of freedom , the husimi function is concentrated near the energy shell .
thus , by fitting gpps to a husimi function , one automatically samples its relevant classical mainifold , if it happens to have a ( possibly unknown ) semiclassical structure .
a basic assumption in the above argument is that the gpps along the curve ( surface ) are neither too crowded , to avoid confusing superpositions , nor too sparsely spaced , lest gaps should arise in the fitting .
it is thus easy to estimate the growth of the number of coherent states required to fit the eigenstate @xmath135 with the degree of excitation . since the linear width of the gaussian scales as @xmath136 , the phase space area of the curve grows as @xmath137 and hence its length grows as @xmath138 , the number of equispaced gaussians needed to cover the curve grows as @xmath139
of course , we have assumed here that the shape of the curve does not change with @xmath0 , as for the harmonic oscillator .
if the eigencurve elongates for higher excitation , the number of gpps necessary for a good fit will grow as @xmath140 , with @xmath141 . we can also estimate the growth in time of the number of gpps required to fit an evolving semiclassical state with arbitrary precision . according to the theory of van vleck @xcite ,
it is sufficient to evolve the curve ( surface ) classically and then to reconstruct the wave function .
this principle can also be applied to wigner functions , according to berry and balazs @xcite .
if the driving hamiltonian is chaotic , then the curve will stretch at a rate depending on the lyapunov exponent , @xmath142 .
clearly , this also determines the initial rate of growth of the number of gpps needed for an adequate fit .
ultimately , when the curve covers densely all the available phase space ( the energy shell of the driving hamiltonian ) , a steady state will be reached where the number of gpps saturates .
optimal fitting of quantum states by phase space gaussians is achieved by first fitting the husimi function and then determining the quantum phases from the chord function , i.e. the fourier transform of the wigner function .
it is clear from the discussion in the introduction that the density of gaussians in phase space must be finely adjusted : if ther are too few , essential features of the state will be missed , whereas an excessive number of gaussians would introduce interference terms in the husimi function itself , which could only be accomodated by a much more complex variation of this method .
the numerical examples in section 4 indicate clearly that , as well as achieving excellent overall accuracy , all essential qualitative features of the states are captured by this method , using a basis of gaussians that grows more slowly than the excitation number of the fitted states .
this last result and further insight into the fitting procedure , follows from the semiclassical analysis in the previous section .
this can be generalized readily to states of quantum systems with higher degrees of freedom if these are eigenstates of integrable systems ( i. e. if they are supported by a lagrangean surface in phase space ; see e. g. @xcite ) .
of course , more gaussians will then be needed for the fitting , but the preliminary fit of the husimi function should still provide optimal results .
the potential for this method to deal with the eigenstates of chaotic systems is even more interesting . though the _ chord _ and _ centre _ description of the wigner function still applies to mixtures of eigenstates over narrow energy windows , no classical theory accounts for individual eigenstates at present .
therefore , it will be extremely useful to describe these as interfering superpositions of gaussians placed near the energy shell .
of course , the size of this basis would diverge at the classical limit as @xmath143 , but manageable approximations should be attainable for finite excitations .
ak gratefully acknowledges the financial support by alexander von humboldt ( avh ) foundation / bonn - germany , under the grant of research fellowship no.iv.4-kam 1068533 stp .
amoa thanks the mpipks - dresden for a martin gutzwiller fellowship , during which this work was initiated , and acknowledges support from cnpq and pronex in brazil .
100 horenko i , schmidt b and schuette c 2002 4643 johansen l m 1998 5461 lutterbach l g , davidovich l 1997 2547 1997 wax a , bali s , thomas j s 1999 _ opt .
* 24 * 1188 ; lee k f , reil f , bali s , wax a , thomas j e 1999 _ opt .
* 24 * 1370 ; nogues g , rauschenbeutel a , osnaghi s 2000 a * 62 * 054101 cohen l 1995 time - frequency analysis .
new york : prentice hall , 113 .
bartelt h o , brenner k h and lohmann a w 1980 _ opt .
comm . _ * 32 * 32 riley m and bolton j s 1993 speech time - frequency representations
. kluver academic publishers .
meng q and qu l 1991 mechanical systems and signal proceessing * 5 * 155 cohen tannoudji c , diu b and laloe f 1977 quantum mechanics , wiley , new york husimi k 1940 _ proc .
phys . math.soc.japan_
* 22 * 264 radmore p m and barnett s m 1997 methods in theoretical quantum optics , oxford univ press wolfgang p schleich 2001 quantum optics in phase space , wiley vch , berlin wigner e p 1932 749 moyal j e 1949 _ proc . camb .
_ * 45 * 99 marksii r j , walkup j f and krile t f 1977 _ appl . opt . _
* 16 * 746 ozorio de almeida a m 1998 _ phys
. reports _ * 295 * 265 van vleck j h 1928 _ proc . math
* 14 * 178 ozorio de almeida a m 1988 hamiltonian systems : chaos and quantization cambridge university press , cambridge berry m v 1977 _ phil .
* 287 * 237 ozorio de almeida a m and hannay j h 1982 ann .
* 138 * 115 berry m.v . and
balazs 1979 625 press w h , teukolsky s a , vetterling w t and flanneryt b p 1992 numerical recipes in fortran 22 kenfack a and rost j m 2003 unpuplished | a two - step optimization is proposed to represent an arbitrary quantum state to a desired accuracy with the least number of gaussians in phase space .
the husimi distribution of the quantum state provides the information to determine the modulus of the weight for the gaussians .
then , the phase information contained in the wigner distribution is used to obtain the full complex weights by considering the relative phases for pairs of gaussians , the chords .
the method is exemplified with several excited states @xmath0 of the harmonic and the morse oscillators .
a semiclassical interpretation of the number of gaussians needed as a function of the quantum number @xmath0 is given .
the representation can also be used to characterize wigner and husimi distributions directly which do not originate in a quantum state . |
mankind has admired the sky with its milky way , stars , planets , moons and the sun , the meteors and the northern light . in the age of satellites
we map the sky through all electromagnetic frequencies and also through high energy cosmic particles .
gravitational waves are the last great frontier to be reached and passed for the search for our physical understanding of the universe via the messengers reaching us with the speed of light or very close to it : gravitational waves , these fleeting distortions of space , the signature of accelerating masses in the universe , remain to be discovered .
therefore it is of paramount importance to understand , which objects are about to generate strong gravitational waves . when the black holes residing in the centers of almost all galaxies merge , following the merger of their host galaxies , relatively strong gravitation waves are emitted and various aspects of this process were already investigated @xcite-@xcite .
a recent review of the generic aspects of these galaxies nuclei as sources for ultra high energy cosmic rays is in @xcite .
starting from the broken powerlaw mass distribution of galactic central black holes @xcite-@xcite , with simple assumptions on the cross section and relative velocity of the two galaxies , we have shown @xcite that the most likely mass ratio @xmath1 of the merging black holes is between @xmath2 and @xmath3 , thus a typical value will be at about@xmath4 . within this range of mass ratios the key stages of the evolution leading to and following the merger
can all be treated with post - newtonian ( pn ) approximations .
a numerical analysis of such mass ratios was recently presented in @xcite . in the next section
we describe the sequence of events which will accompany the merger of two supermassive galactic black holes .
then from section 3 we concentrate on the gravitational radiation dominated dissipative regime .
an important quantity in gravitational wave detection is the _ number of cycles spent in the frequency range of the detector_. this quantity is proportional to the _ accumulated orbital phase _ ( the total angle covered by the quasi - circular orbital evolution ) .
the gravitational waves emerging from this process carry the imprint of the physical characteristics ( finite size ) of the black holes , like their spins and quadrupole moments .
spin precessions allow for increased accuracy and degeneracy breaking among the parameters characterizing the source @xcite .
we remark that according to @xcite , the mass quadrupole - mass monopole coupling for maximally rotating black holes ( with quadrupole moment originating entirely in their rotation ) gives rise to contributions in the accumulated orbital phase larger by a half to one order of magnitude than the spin - spin contribution , at least for the typical unequal mass ratios .
therefore for a successful detection this qualifies as the second most important finite mass contribution to be considered , after the spin - orbit effect . in section 3
we discuss the effect of the quasi - precessional evolution of the spins on the accumulated phase of the gravitational waves and we present estimates of the pn order in which various finite size contributions to the change in the relative configurations of the spins and orbital angular momentum contribute .
the cumbersome equations leading to these estimates are presented in the appendix .
sections 4 and 5 explore a convenient analytical way to encompass the effect of spin and quadrupolar quasi - precessions even at a lower level accuracy . the leading order contribution to the spin - spin ( section 4 ) and spin - orbit ( section 5 ) contributions to the accumulated orbital phase
are discussed both for the equal mass and unequal mass cases in distinct subsections .
we summarize our findings in the concluding remarks .
first the dance : two galaxies with central black holes approach each other to within a distance where dynamical friction keeps them bound , spiraling into each other . in the case when both galaxies with their central black holes were radio galaxies ,
their jets get distorted and form the z - shape @xcite : the radio emitting blobs and tails produce the appearance of dancing veils .
second the meeting of the eyes : the central regions in each galaxy begin to act as one unit , in a sea of stars and dark matter of the other galaxy . during this phase ,
the central region can be stirred up , and produce a nuclear starburst .
third the lock : the black holes begin to lose orbital angular momentum due to the interaction with the nearby stars @xcite , and next by gravitational radiation .
the spin axes tumble and precess .
this phase can be identified with the apparent superdisk , as the rapidly precessing jet produces the hydrodynamic equivalent of a powerful wind , by entraining the ambient hot gas , pushing the two radio lobes apart and so giving rise the a broad separation @xcite . due to the combined effect of precessions and orbital angular momentum dissipation by gravitational waves the spin ( and hence jet ) direction of the dominant black hole
is reoriented approximately in the direction of the original orbital angular momentum , leading to a spin - flip @xcite .
fourth the plunge : the two black holes actually merge , with not much angular orbital momentum left to be radiated away , thus the direction of the spin remained basically unchanged ( in the case of the typical mass ratio range we discuss ; for other setups the final spin may vary @xcite-@xcite ) .
the final stage in this merger leads to a rapid increase in the frequency of the gravitational waves , called chirping " , but this chirping will depend on the angles involved .
fifth the rejuvenation : now the newly oriented more massive merged black hole starts its accretion disk and jet anew , boring a new hole for the jets through its environment ( the gigahertz peaked sources ) . in this stage
the newly active jet is boring through the interstellar clouds with a gigantic system of shockwaves , that accelerate protons and other charged particles in a relativistic tennis - game .
these very energetic particles then interact in those same clouds , that slow down the progression of the relativistic jet , and can so give rise to an abundance of neutrinos , tev photons from pion decay , and yet other particles and photons .
these particles and their interactions reach energies far beyond any earth - bound accelerator .
sixth the phoenix : the newly oriented jets begin to show up over some kpc , and this corresponds to the x - shaped radio galaxies , while the old jets are fading but still visible @xcite .
this also explains many of the compact steep spectrum sources , with disjoint directions for the inner and outer jets .
seventh the cocoon : the old jets have faded , and are at most visible in the low radio frequency bubbly structures , such as seen for the virgo cluster region around m87 .
the feeding is slowing down , while a powerful jet is still there , fed from the spin of the black hole .
the wait : the feeding of the black hole is down to catching some gas out of a common red giant star wind as presumably is happening in our galactic center .
this stage seems to exist for all black holes , even at very low levels of activity .
this black hole and its galaxy are ready for the next merger , with the next black hole suitor , which may or may not come .
we predict that the super - disk radio galaxies should have large outer distortions in their radio images , and should be visible in low kev x - rays . a detection of the precessing jet , and its sudden weakening ,
would then immediately precede the actual merger of the two black holes , and so may be a predictor of the gravitational wave signal .
in the rest of the paper we concentrate on the gravitational radiation dominated dissipative regime , when the supermassive black hole binary radiates away gravitational waves . if the binary system evolves unperturbed , the orbit circularizes faster than it shrinks in radius .
therefore we consider circular orbits for which the wave frequency is twice the orbital frequency .
besides the keplerian contribution to the accumulated orbital phase there are general relativistic contributions , which scale with the post - newtonian ( pn ) parameter @xmath5 ( with @xmath6 the total mass and @xmath7 the radius of the reduced mass orbit about the total mass ; we take @xmath8 ) .
it is generally agreed from comparison with numerical evolution , that all contributions up to 3.5 pn orders have to be taken into account . to this order , beside the general relativistic contributions starting at 1pn
, there are various contributions originating in the finite size of the binary components .
these are related either to their rotation ( spin ) or to their irregularities in shape ( like a mass quadrupole moment ) . due to the continued accretion of surrounding matter into the supermassive galactic black holes they spin up considerably . by taking into account only the angular momentum transfer from accreting matter , the dimensionless spin parameters @xmath9 ( with @xmath10 )
grow to the maximally allowed value @xmath11 even for initially non - rotating black holes @xcite . by taking into account the torque produced by the energy input of the in - falling ( horizon - crossing ) photons emitted from the steady - state thin accretion disk , which counteracts the torque due to mass accretion
, the limiting value of the dimensionless spin parameter is slightly reduced to @xmath12 @xcite .
various refinements of this process , with the inclusion of open or closed magnetic field lines @xcite-@xcite , also jets in the magnetosphere of the hole jetdisksymbiosys-@xcite have not changed essentially this prediction .
therefore we assume for the present considerations maximal rotation @xmath13
. due to this rotation the black hole is centrifugally flattened ( it becomes an oblate spheroid ) , a deformation which can be characterized by a mass quadrupole scalar @xmath14 @xcite , or its dimensionless counterpart @xmath15 @xmath16 .
the accumulated orbital phase can be formally given as the pn expansion renormsigma , @xcite , @xcite@xmath17spin - orbit ( so ) contributions appear at 1.5 pn , spin - spin ( ss ; composed of proper s@xmath18s@xmath19 and self - contributions s@xmath20 and s@xmath21 ) and mass quadrupole - mass monopole coupling ( qm ) contributions at 2pn .
they come with various pn corrections ( pnso at 2.5 pn ; pnss , pnqm and so@xmath22 at 3pn ; finally 2pnso at 3.5 pn ; these were not computed yet with the exception of pnso @xcite-@xcite ) . at 3pn
there are additional spin and quadrupole contributions , originating in the quasi - precessional evolution of the spin and orbital angular momentum vectors .
the number of gravitational wave cycles @xmath23 can be computed from the accumulated orbital phase as @xmath24 ( the gravitational wave frequency being twice the orbital frequency for quasi - circular motion ) .
the spin precession equations with so , ss and qm contributions were given by barker and oconnell @xcite .
the relative geometry of the two spins @xmath25 and orbital angular momentum @xmath26 can be best described by the angles @xmath27 and @xmath28 ( an overhat denotes the direction of the respective vector ) .
these angles are related by the the spherical cosine identity@xmath29with @xmath30 the relative azimuthal angle of the spins . due to quasi - precessions ,
@xmath31 and @xmath32 evolve .
supplementing the so and ss contributions already given @xcite with the qm contributions ( with the quadrupole moment arising from pure rotation ) , we give these expressions to second order accuracy:@xmath33with the detailed contributions enlisted in the appendix . the order of magnitude estimates of the terms in eqs .
( [ kappa1_so])-(gamma_qm ) ( cf . a footnote in ref .
@xcite , according to which @xmath34 ) , and assuming maximal rotation lead to@xmath35@xmath36@xmath37 \approx \varepsilon ^{3/2}\ , \nonumber \\
\mathcal{o}\left ( \delta _ { qm}\gamma \right ) & \approx & \varepsilon ^{3/2}\left ( -\nu + \nu \right ) \approx \varepsilon ^{3/2}\nu \ . \label{orders}\end{aligned}\]]thus the angle @xmath32 varies at 1pn , while the angles @xmath31 only at 1.5 pn _ during one orbital period_. however @xmath32 appears only in the 2pn order s@xmath18s@xmath19 contribution to the accumulated orbital phase as opposed to the angles @xmath31 which are present in all finite - size contributions , in particular in the 1.5 pn order so contribution . therefore 3pn order quasi - precessional contributions to the accumulated orbital phase in the comparable mass case @xmath38 arise from @xmath39 , @xmath40 , @xmath41 and @xmath42 ; while in the unequal mass case , found typical for supermassive galactic black hole binaries only from @xmath43 , @xmath44 and @xmath42 .
from among the quasi - precessional evolutions , we have recently integrated @xcite the so evolution ( averaged over one orbit ) of the angle @xmath45 , finding@xmath46here @xmath47 the reduced mass , @xmath48 the radius and @xmath49 . the time - dependent expression for @xmath32
is then given by eq .
( [ gombi ] ) , with @xmath30 from eq .
( [ depsi ] ) .
the variation of both @xmath30 and @xmath32 thus have a periodicity with period@xmath50where @xmath51 for equal mass binaries this time - scale becomes infinite @xmath52 ( see fig1a ) expressing the fact that the time - dependent part of @xmath32 goes to zero . thus @xmath32 is a constant and _ there is no 3pn quasi - precessional contribution to the accumulated orbital phase for equal mass binaries_. for the typical mass ratio the factor @xmath53 is of the order of unity ( see fig 1b ) , thus the variation time - scale of @xmath32 is @xmath54 times larger than the wave period . the contribution of the quasi - precessions can be taken into account then by keeping only the first term in eq .
( [ gombi ] ) when computing the 2pn order s@xmath18s@xmath19 contribution to the accumulated orbital phase .
e. g. by introducing a _
renormalized spin coefficient_. this coefficient arises by modifying the 2pn contribution with the average of some of the 3pn contributions . by using this renormalized coefficient in the context of the 2pn accurate dynamics
, one can bring closer the 2pn analytical prediction to the numerical results @xcite .
, which are still not computed . ]
thus , in the unequal mass case we propose to replace @xmath55with @xmath56 in the respective 2pn contribution to the accumulated orbital phase @xmath57(here @xmath58 is the symmetric mass ratio and @xmath59 a dimensionless time parameter . ) , shown for the whole allowed range of mass ratios , diverges in the equal mass case .
( b ) the factor @xmath60 in the typical mass ratio range @xmath61 found for merging supermassive galactic black hole binaries is of order unity.,title="fig:",height=139 ] 1.3 cm , shown for the whole allowed range of mass ratios , diverges in the equal mass case .
( b ) the factor @xmath60 in the typical mass ratio range @xmath61 found for merging supermassive galactic black hole binaries is of order unity.,title="fig:",height=139 ]
the 1.5 pn contribution to the accumulated orbital phase is @xmath62with the so contribution given by @xmath63 the quasi - precessional equations with the so , ss and qm contributions included were recently integrated by racine for the equal mass case racine . according to his analysis , there is a previously unknown constant of the motion@xmath64as for equal masses @xmath65const .
, _ the quasi - precessional evolution at 3pn does not affect the coefficient _ @xmath66 _ _ up to the 3pn accuracy . _ _ in the unequal mass case the leading order evolutions of the angles @xmath67 are @xmath68 and @xmath44 . averaged over one orbit
they give @xcite@xmath69for unequal masses @xmath70 the estimates ( [ orders ] ) allow to consider @xmath71 as a constant ( in a first approximation ) .
the leading - order time - dependence of @xmath30 is given by eq .
( [ depsi ] ) .
integration gives @xmath72 ~. \label{kappa2}\]](we have employed @xmath73 . )
the time - varying part has the same period @xmath74 given by eq .
( [ t3pns ] ) , which is of the scale of @xmath75 the orbital period . in agreement with this and the estimate ( [ orders ] ) @xmath76as the time varying part averages out , to leading order accuracy we can simply identify the angle @xmath77 with its constant initial value @xmath78 . in conclusion , in the unequal mass case and in the leading order approximation @xmath79const . the time variation due to quasi - precessions
renders @xmath77 to another constant at 3pn accuracy .
_ there is no need to renormalize _
@xmath66 , _ when taking into account the leading order quasi - precessional contributions .
the process of the merger of two supermassive galactic black holes gives rise to exciting signals in all the messengers accessible now and in the near future , and will allow to probe some of the most profound secrets of nature , occurring at energies and circumstances far beyond anything on earth .
we have described the sequence of events during the process of merging of such supermassive galactic black holes with reference to related processes involving accretion , jets , energetic particles and electromagnetic signatures .
we argued that the class of radio galaxies with a super - disk are particularly promising candidates for future powerful low frequency gravitational wave emission .
the accumulated orbital phase is a quantity to be known precisely for successful detection of gravitational waves . besides keplerian ,
first and second pn order general relativistic contributions , the finite size of the black holes also contributes , first at 1.5 pn orders via the spin - orbit interaction , then at 2pn by the spin - spin and quadrupole - monopole coupling .
these contributions are further modulated by quasi - precessional evolutions of the spins .
however this is but a subset of all possible finite size contributions at 3pn , as stressed earlier .
the way we propose to take into account the precessional 3pn contributions is to add their average to the lower order terms , a procedure we call _ renormalization _ of the lower order coefficients .
we have proven here that there are no such quasi - precessional contributions in the equal mass case .
furthermore we have proved for the unequal mass case that renormalizing the 1.5 pn spin - orbit coefficient will not change its value . at 2pn
there are quadrupolar and spin - spin contributions to consider .
as the former can be regarded as constant up to 3pn , the only modification due to quasi - precessions comes from the spin - spin interaction .
the renormalized spin - spin coefficient at 2pn thus is adjusted by the averageof the 3pn quasi - precessional contribution , taken over the period of change induced by these quasi - precessions ( a time - scale one pn order lower than the orbital period ) .
we have checked that the renormalized spin - spin coefficient gives closer results to the numerical estimates for compact binaries with the mass ratio of order @xmath80 @xcite .
our results related to the renormalization of the spin coefficients are significant for supermassive black hole binaries , which are typically lisa sources .
however as we have employed nothing but the mass ratio of the binary in our considerations , they also apply for binaries composed of an intermediate mass and a stellar mass black hole , which qualify as ligo sources .
participation of lg at the gwdaw13 meeting was supported by the polnyi program of the hungarian national office for research and technology ( nkth ) .
lg was also successively supported by a rof award of lsbu and by collegium budapest .
support for work with plb has come from the auger membership and theory grant 05 cu 5pd 1/2 via desy / bmbf and vihkos .
the work of bm was supported by the hungarian scientific research fund ( otka ) grant no .
68228 ; and of zk by the otka grant
the so and ss contributions of the instantaneous variations of the relative spin angles @xcite supplemented with the respective qm contributions ( assuming the quadrupole moment arises from pure rotation ) , to second order accuracy are:@xmath81 \ , \label{kappa1_ss } \\ ( \cos \mathbf{\skew{44}{\dot}{\kappa}}_{1})\!_{\,qm } & = & \mathbf{-}\frac{3\mu m^{3}}{r^{3}}\left [ \frac{p_{1}}{s_{1}}(\mathbf{\hat{r}\cdot \hat{s}}_{% \mathbf{1}})\mathbf{\hat{l}_{n}\cdot } ( \mathbf{\hat{r}}\times \mathbf{\hat{s}% _ { 1})}\right . \nonumber \\ & & \mathbf{+}\left .
\frac{p_{2}}{l_{n}}(\mathbf{\hat{r}\cdot \hat{s}}_{% \mathbf{2}})\mathbf{\hat{r}\cdot ( \hat{s}_{1}\times \hat{s}_{2}})\right ] \ , \label{kappa1_qm}\end{aligned}\]]@xmath82 \ ,
\label{kappa2_ss } \\ ( \cos \mathbf{\skew{44}{\dot}{\kappa}}_{2}\mathbf{)}_{\!qm } & = & -\frac{3\mu m^{3}}{r^{3}}\left [ \frac{p_{2}}{s_{2}}(\mathbf{\hat{r}\cdot \hat{s}_{2}})% \mathbf{\hat{l}_{n}\cdot } ( \mathbf{\hat{r}}\times \mathbf{\hat{s}_{2})}% \right .
\nonumber \\ & & \mathbf{+}\left .
\frac{p_{1}}{l_{n}}\mathbf{(\hat{r}\cdot \hat{s}_{1})\hat{% r}\cdot ( \hat{s}_{1}\times \hat{s}_{2}})\right ] \ , \label{kappa2_qm}\end{aligned}\]]@xmath83 \mathbf{\hat{r}\cdot } ( \mathbf{\hat{s}_{1}}\times \mathbf{\hat{s}_{2})}\ , \label{gamma_ss } \\
( \cos \mathbf{\skew{27}{\dot}{\gamma}}\mathbf{)\!}_{qm } & = & \frac{3\mu m^{3}}{% r^{3}}\left [ \frac{p_{2}}{s_{2}}(\mathbf{\hat{r}\cdot \hat{s}_{2}})-\frac{% p_{1}}{s_{1}}(\mathbf{\hat{r}\cdot \hat{s}_{1}})\right ] \mathbf{\hat{r}\cdot } ( \mathbf{\hat{s}_{1}}\times \mathbf{\hat{s}_{2})}\ .
\label{gamma_qm}\end{aligned}\ ] ]
99 peters p c 1964 _ phys .
rev . _ * 136 * b1224 thorne k s 1979 _ proc .
royal soc . london _ * a 368 * 9 biermann , p.l . ,
chirvasa , m. , falcke , h. , markoff , s. , & zier , ch . , invited review at the paris conference on cosmology , june 2000 , in proceedings , eds . n. sanchez , h. de vega , p. 148
- 164 ( 2005 ) ; astro - ph/0211503 merritt d , ekers r 2002 _ science _ * 297 * 1310 - 1313 gergely l , biermann p l 2009 _ astrophys . j. _ * 697 * 1621 biermann p l , becker j k , caramete a , curutiu l ,
engel r , falcke h , gergely l , isar p g , maris i c , meli a , kampert k - h , stanev t , tascau o , zier c 2009 in proceedings of the cris 2008 cosmic ray international seminar : origin , mass , composition and acceleration mechanisms of uhecrs , malfa , italy 2008 , ed .
a insolia , _ nucl .
* 190 * p. 61 - 78 ; arxiv : 0811.1848v3 [ astro - ph ] gergely l , mikczi b 2009 _ phys .
d _ * 79 * 064023 gopal - krishna , biermann p l , wiita p 2003 _ astrophys .
j. lett . _
* 594 * l103 - l106 zier ch 2007 _ month . not .
soc . _ * 378 * 1309 gopal - krishna , wiita p j , joshi s 2007 _ month . not .
soc . _ * 380 * 703 rezzolla l , barausse e , dorband e n , pollney d , reisswig c , seiler j , husa s 2008 _ phys .
* 78 * 044002 washik m c , healy j , herrmann f , hinder i , shoemaker d m , laguna p , matzner r a 2008 _ phys .
lett . _ * 101 * 061102 tichy w , marronetti p 2008 _ phys
d _ * 78 * 081501 barausse e , rezzolla l 2009 _ predicting the direction of the final spin from the coalescence of two black holes _ , e - print : arxiv:0904.2577 rottmann h. 2001 phd thesis : _ jet - reorientation in x - shaped radio galaxies _ , bonn univ .
: ( http://hss.ulb.uni-bonn.de/diss@xmath84online/math@xmath84nat@xmath84fak/2001/rottmann @xmath84helge / index.htm ) bardeen j m 1970 _ nature _ * 226 * 64 | we analyze galactic black hole mergers and their emitted gravitational waves .
such mergers have typically unequal masses with mass ratio of the order @xmath0 .
the emitted gravitational waves carry the imprint of spins and mass quadrupoles of the binary components . among these contributions ,
we consider here the quasi - precessional evolution of the spins .
a method of taking into account these third post - newtonian ( 3pn ) effects by renormalizing ( redefining ) the 1.5 pn and 2pn accurate spin contributions to the accumulated orbital phase is developed . |
measuring the kinematics of ionized gas disks has become one the most important methods to determine the masses of supermassive black holes ( smbhs ) in the nuclei of nearby galaxies .
the state - of - the - art modeling techniques are refined to a high degree and take into account the major relevant astrophysical processes as well as the main instrumental effects .
applying such modeling techniques to high - quality hst data , a formal smbh mass measurement uncertainty of 25% or better has been claimed with well - behaved disks ( e.g. @xcite , @xcite , @xcite , @xcite ) .
many ionized gas disks contain substantial amounts of interstellar dust . as optical h@xmath0 radiation
is easily absorbed and scattered by interstellar dust grains , these effects might affect the observed rotation curve and hence the smbh mass estimate .
the goal of our investigation is to determine the importance of this effect through detailed monte carlo simulations .
we have set up a detailed radiative transfer model to investigate the effects of neglecting dust absorption and scattering on ionized gas disc rotation curves and the corresponding smbh masses in a typical galactic nucleus .
our model consists of a thin ( but not infinitesimally thin ) axisymmetric double - exponential disc rotating in a galactic nucleus .
the intrinsic rotation velocity of the disc is determined by the combined gravitational potential of a stellar distribution and a central smbh .
we consider various smbh masses , up to @xmath1 .
an additional isotropic bulk velocity dispersion is taken into account .
we assume that a fraction of the disc consists of interstellar dust with typical milky way dust optical properties .
the dust is assumed to be coupled to the ionized gas and has exactly the same spatial and velocity distribution .
the smbh mass @xmath2 , the thickness @xmath3 of the disc and the v - band optical depth @xmath4 , a quantity equivalent to the dust mass , are free parameters in our models .
for each model , we calculate the observed line - of - sight velocity field ( data cubes ) at several inclination angles .
we use the skirt code , a 3d monte carlo radiative transfer code that has been developed explicitly to model the kinematics of dusty galaxies @xcite , @xcite .
the code uses photon packages that carry around information on the line - of - sight velocity of the sources ( in this case the h@xmath0 emitting gas ) that emitted them .
the effects of absorption and multiple anisotropic scattering are properly taken into account . during scattering events ,
the doppler information carried by each photon package is updated , taking into account both the original line - of - sight velocity and the scattering dust grain velocity .
when the photon packages leave the system , their `` observed '' line - of - sight velocity information can be extracted , which is used to build up data cubes . at the end of the simulation , we calculate rotation velocity maps and velocity dispersion maps from the simulated data cubes .
we focus in particular on the major axis rotation curves .
and the disc scale height @xmath3 .
in each panel we plot the rotation curves for different black hole masses ( ranging between 0 and @xmath5 m@xmath6 ) and for different optical depths ( ranging between 0 and 2 ) . for any @xmath7 , @xmath3 and @xmath2
, the slope of the rotation curve slightly decreases with increasing optical depth . ] in figure [ mbaesfig1 ] we plot a set of nuclear rotation curves for our dusty ionized gas discs .
the different panels correspond to different values for @xmath3 and the inclination angle . within each individual panel ,
rotation curves are shown for different values of @xmath2 and @xmath4 . for a fixed spatial resolution and inclination ,
the shape of the rotation curve of dust - free models depends on the smbh mass and disc thickness . in particular
, increasing disc thickness has a significant influence due to the larger integration along the line of sight .
modeling ionized gas disks as infinitesimally thin disks might introduce a bias . to investigate the effect of interstellar dust absorption and scattering
, we must compare models with different values of the optical depth .
the curves corresponding to various @xmath4 basically coincide : apparently absorption and scattering by interstellar dust leaves the shape of the rotation curves qualitatively unaltered .
however , close inspection shows that an increasing amount of interstellar dust slightly decreases the central slope of the rotation curves .
we estimate the bias one makes by neglecting the effect of interstellar dust on smbh mass estimates by translating the observed rotation curve central slopes into smbh masses .
figure [ mbaesfig2 ] shows the effect of increasing dust mass on the `` observed '' smbh mass and on the ratio between the `` observed '' and the intrinsic smbh mass .
it clearly demonstrates that , as the optical depth increases , the observed smbh mass decreases , i.e. the true smbh mass is more and more underestimated .
this means that neglecting interstellar dust when determining smbh masses from ionized gas kinematics systematically leads to an underestimate of the black hole mass .
for @xmath8 the smbh is underestimated by about 10% , for @xmath9 the black hole mass is underestimated by 10 to 20% .
this effect is more or less independent of the smbh mass and the inclination of the disc .
we find that absorption and scattering by interstellar dust leaves the shape of the rotation curves basically unaltered , but slightly decreases the central slope of the rotation curves .
since the central slope can be used as a proxy for the smbh mass , this result means that `` observed '' black hole masses are systematically underestimated by some 10 to 20% for realistic optical depths .
our study demonstrates that the systematic effect of dust attenuation should be taken into account in smbh demographics studies . | using detailed monte carlo radiative transfer simulations in realistic models for galactic nuclei , we investigate the influence of interstellar dust in ionized gas discs on the rotation curves and the resulting black hole mass measurements .
we find that absorption and scattering by interstellar dust leaves the shape of the rotation curves basically unaltered , but slightly decreases the central slope of the rotation curves . as a result ,
the `` observed '' black hole masses are systematically underestimated by some 10 to 20% for realistic optical depths .
we therefore argue that the systematic effect of dust attenuation should be taken into account when estimating smbh masses using ionized gas kinematics .
address = sterrenkundig observatorium , universiteit gent , krijgslaan 281 s9 , b-9000 gent , belgium |
the numerical study of the binary black hole problem has recently made significant progress @xcite .
it is now possible to calculate the evolution of a wide variety of astrophysical scenarios . in order to do this ,
appropriate initial data for the einstein equations are necessary , such that they describe the initial conditions of these scenarios .
a natural physical requirement for binary black hole initial data is that in the limit of large separation distance each of the black holes approximates an stationary , isolated , black hole ( in vacuum , this implies that the black hole approximates the kerr black hole ) .
otherwise , the data will contain spurious radiation which can in principle contaminate the final waveform ( see @xcite ) .
the nonlinearity of the constraint equations makes the problem of constructing initial data satisfying such far limit a nontrivial one . in @xcite
such data for two kerr black holes with far limit has been constructed .
the procedure uses the conformal method for solving the constraint equations ( see the review article @xcite and references therein ) . in this construction , one starts by constructing a superposition of two conformal kerr metrics to get a new conformal metric , which is then used as a ` seed ' metric for the conformal method .
then , a conformal factor is calculated such that the corresponding rescaled metric satisfies the constraint equations . the existence and uniqueness of that conformal factor can be proved provided the ` seed ' metric satisfies an important requirement : its yamabe invariant must be positive ( see @xcite for a discussion about the yamabe invariant in the context of the constraint equations ) . in order to ensure this condition in @xcite , it has been assumed that the individual angular momentum parameters of the black holes are small with respect to the masses .
this assumption is , of course , an undesirable restriction . highly spinning black holes are relevant in many astrophysical situations .
a remarkable example of this are the large merger recoil kicks ( see @xcite ) .
the purpose of this article is to overcome this restriction for the axially symmetric case .
we provide an explicit metric , having three ` ends ' , which is constructed as a superposition of two kerr black holes metrics ( with aligned or anti aligned angular momentum ) .
when the mass parameters of one black hole is zero then the metric reduces exactly to the kerr metric . in this sense
, we say that the metric satisfies the far limit .
the important property of this metric is that when the separation distance between the black holes is large enough its ricci scalar is positive .
the parameters of the individual kerr black holes are only restricted by the kerr inequality @xmath0 , where @xmath1 is the angular momentum per mass unit and @xmath2 is the mass of the individual black hole .
that is , the angular momentum is not assumed to be small .
instead , we assumed the physical reasonable condition that the separation distance between them is large .
this constitutes the main result of the article . when the three ` ends ' are asymptotically flat ( the non - extreme cases ) the positivity of the ricci scalar implies the positivity of the corresponding yamabe invariant .
and hence the metric can be used as a seed metric for the construction mentioned above . in this way
axially symmetric initial data can be constructed for two kerr - like black holes with large angular momentum .
however , we emphasize that our result covers also the extreme limit @xmath3 of any of the black holes .
we believe that this can be used to construct ( the still unknown ) initial data for two extreme kerr black holes .
consider the kerr black hole ( i.e. the kerr metric such that @xmath4 ) in boyer - lindquist coordinates .
take any constant @xmath5 slice in these coordinates and denote by @xmath6 the intrinsic 3-dimensional metric of the slice .
there exist spatial coordinates @xmath7 on the slice such that the metric @xmath6 has the following form @xmath8 where @xmath9 and the functions @xmath10 and @xmath11 ( which do not depend on the coordinate @xmath12 ) are explicitly given by equations ( [ eq:51 ] ) in the appendix [ sec : kerr - data ] .
the slice is a 3-dimensional manifold @xmath13 which has the topology @xmath14 , where @xmath15 denotes the origin in the coordinates @xmath7 . for the results presented here a relevant property of the kerr metric
is that @xmath16 ( see lemma [ t : psi ] ) .
we construct a new manifold @xmath17 by removing two points from @xmath18 denoted by @xmath19 and @xmath20 , namely @xmath21 .
let @xmath7 be cylindrical coordinates on @xmath17 such that the two points @xmath19 and @xmath20 are located at the axis @xmath22 , separated by coordinate distance @xmath23 .
this requirement does not completely fix the coordinate system , since there still is a translation freedom in the @xmath24 coordinate @xmath25 where @xmath26 is any arbitrary constant . later on
we will make use off this freedom to simplify the computations .
however , for the formulation of the results there is no loss off generality if we chose the coordinate system such that @xmath19 is located at @xmath27 and @xmath20 at @xmath28 on the @xmath24 coordinate axis . given the functions @xmath29 and @xmath30 of the kerr initial data , with parameters @xmath31 , we define the new functions @xmath32 by @xmath33 in the same way , for functions @xmath34 and @xmath35 we define @xmath36 by @xmath37 with this set of functions constructed from the kerr metric , we define a metric on @xmath17 , depending on the parameters @xmath38 , by @xmath39 where @xmath40 since @xmath41 and @xmath42 are positive functions , @xmath43 is well defined as a metric on @xmath17 for any choice of the parameter @xmath23 . for
the non - extreme case ( i.e @xmath44 , @xmath45 ) one can check that the metric @xmath43 is asymptotically flat with three asymptotic ends .
these ends are denoted by @xmath46 , which correspond to the limit @xmath47 ( @xmath48 denotes the euclidean distance to the origin ) and @xmath19 , @xmath20 , which correspond to the limits @xmath49 and @xmath50 respectively ( @xmath51 and @xmath52 denote the euclidean distances with respect to the points @xmath19 and @xmath20 respectively ) .
the total mass of the metric ( [ brillsuma ] ) at the end @xmath46 is given by @xmath53 .
if one of the ends is extreme ( that is @xmath54 or @xmath55 ) , then the metric will still be asymptotically flat at the end @xmath46 but will not be asymptotically flat at the extreme end .
denote by @xmath56 the ricci scalar of the metric ( [ brillsuma ] ) .
the following theorem constitutes the main result of this article .
[ t : main ] there exists a constant @xmath57 such that if @xmath58 then @xmath59 on @xmath17 . moreover , if @xmath60 or @xmath61 is different from zero , then @xmath62 on @xmath17 .
for @xmath63 we have @xmath64 . in this limit
the metric reduces to the brill - lindquist metric @xcite .
if we set to zero the mass of one of the black holes ( this implies that the corresponding angular momentum is also zero ) then the metric ( [ brillsuma ] ) reduces to the kerr metric ( [ eq : kerr - metric ] ) .
consider the yamabe invariant defined on the manifold @xmath17 for the metric @xmath43 @xmath65 where @xmath66 denotes the set of smooth functions with compact support in @xmath17 , @xmath67 , @xmath68 is the volume element with respect to @xmath43 , and @xmath69 means that @xmath70 can not be identically zero everywhere . for the non - extreme cases ,
theorem [ t : main ] implies that @xmath71 ( it is obvious that @xmath72 , to prove that it is strictly positive we use lemma 4.1 in @xcite ) .
the main application of theorem [ t : main ] is the construction of a two ( non - extreme ) kerr black holes data .
the results proved so far assumed that the angular momentum parameter @xmath1 is small with respect to the mass ( see @xcite ) .
theorem [ t : main ] allows to generalize these results for any @xmath1 such that @xmath73 , but only for the axially symmetric case ( that is , when the angular momentum of the black holes are aligned or anti aligned ) . using the conformal method and the metric ( [ brillsuma ] ) as a ` seed ' conformal metric
the complete data can be obtained as follows . since
this procedure involves standard applications of known results we only sketch the proof .
first we superpose two extrinsic curvatures from two kerr initial data .
this can be done in general ( see @xcite ) , however in axial symmetry the procedure is much simpler and can be done explicitly ( see , for example , @xcite , @xcite ) .
then , we use existence theorems for the lichnerowicz equation ( see for example , @xcite , @xcite , @xcite ) to prove that there exists a new conformal factor such that a conformal rescaling of the metric @xmath74 satisfies the vacuum constraint equations with the above constructed second fundamental form ( see @xcite for details ) .
it is in this last step when theorem [ t : main ] plays an important role . in order to apply these existence theorems
we need to ensure that the conformal metric has positive yamabe invariant . using theorem [ t : main ]
we only require that the separation distance between the black holes is large .
we also note that although this procedure is identical to the one proposed in @xcite , the ` seed ' conformal metric is different .
the above construction applies to the non - extreme case .
remarkably , theorem [ t : main ] is valid also for the extreme case @xmath75 .
it is very likely that this theorem can also be used to construct a superposition of two extreme kerr black holes . but this remains to be seen .
the existence results for the lichnerowicz equations proved so far in the literature applies only to asymptotically flat manifolds and not to manifolds with cylindrical ends like extreme kerr .
in this section we establish the three key properties of kerr intrinsic metric which allow us to prove theorem [ t : main ] in the next section .
these properties are collected in lemmas [ t : psi ] , [ lemmartildepositivo ] and [ lemmarmenoscompacto ] .
we consider the kerr metric @xmath6 given by ( [ eq : kerr - metric ] ) , on the manifold @xmath14 , where the functions @xmath11 and @xmath10 are given by ( [ eq:51 ] )
. the ricci scalar of @xmath6 is denoted by @xmath76 and the ricci scalar of the the conformal metric @xmath77 ( defined by ( [ brilldef ] ) ) is denoted by @xmath78 .
we emphasize that all the functions involved are smooth on @xmath13 .
[ t : psi ] assume @xmath0 and @xmath79 .
then , we have that @xmath80 everywhere on @xmath13 .
we prove this by explicitly showing that @xmath81 if @xmath79 , where @xmath82 . using ( [ eq:51 ] ) and
( [ eq : x2 ] ) we obtain @xmath83 we use equation ( [ eq : rtilde ] ) for @xmath84 to obtain @xmath85 using ( [ eq : psix ] ) and ( [ eq:5 ] ) we get our final inequality @xmath86 .
we also mention the following upper bound for @xmath87 that can be obtained from equation ( [ eq : x1 ] ) @xmath88 for the extreme case @xmath75 this reduces to @xmath89 these bounds are sharp at infinity , in the sense that the right hand side of ( [ eq:8 ] ) and ( [ eq:23 ] ) goes to @xmath90 as @xmath91 .
[ lemmartildepositivo ] assume @xmath0 and @xmath92 .
then , @xmath93 on @xmath13 , and @xmath94 only at @xmath22 .
the first statement is immediate if one considers the hamiltonian constraint equation in the given slice and the fact that these slices are maximal ( i.e. the trace of the second fundamental form is zero ) . in effect , since @xmath95 ( where @xmath96 is the second fundamental form of the slice ) the hamiltonian constraint is given by @xmath97 to prove the second part of the theorem we compute the right hand side of ( [ eq:3 ] ) .
the kerr second fundamental form @xmath96 can be expressed in terms of derivatives of a potential @xmath98 ( see , for example , @xcite@xcite ) . in particular , for the square of @xmath99 we have @xmath100 the explicit expression for @xmath98 is given in equation ( [ eq : omega ] ) .
we calculate @xmath101 where @xmath102 we have @xmath103 using ( [ eq : kk ] ) and ( [ eq : oo ] ) and the explicit expressions ( [ eq:51 ] ) we obtain the following lower bound @xmath104 which proves the theorem because @xmath105 .
note that for the extreme case this bound reduces to @xmath106 as a side remark we point out that the bounds ( [ eq:11 ] ) and ( [ eq:12 ] ) are sharp in the limit @xmath107 and @xmath47 .
contains the support of @xmath108 ( shaded region).,height=151 ] for the next lemma we define the function @xmath108 to be the negative part of @xmath78 @xmath109 [ lemmarmenoscompacto ] assume @xmath110
. then the function @xmath108 has compact support in @xmath13 .
moreover , the support of @xmath108 does not intersect the @xmath22 axis ( see figure [ fignoext ] ) take a small ball centered at the origin .
the leading order term in the asymptotic expansion of @xmath78 for @xmath107 ( see equation ( [ eq : tnoea ] ) ) is positive .
hence , there exists a constant @xmath111 such that the function @xmath78 is positive in small ball of radius @xmath111 centered at the origin .
we also have that @xmath78 is positive evaluated at the axis @xmath22 ( see equation ( [ eq : posr ] ) ) .
then , there exist a small cylinder of radius @xmath112 such that @xmath78 is positive outside this cylinder . in a similar fashion ,
the leading order term in the asymptotic expansions of @xmath78 for @xmath113 ( see equation ( [ eq : tnoe ] ) ) is positive .
hence , there exist a constant @xmath114 such that @xmath78 is positive outside a ball of radius @xmath114 . bringing these together ,
we conclude that the support of @xmath108 is contained in the region @xmath115 given by @xmath116 the region @xmath115 is showed in fig .
[ fignoext ] , where we also show how the actual support of @xmath108 looks like when obtained by explicit computation for some values of @xmath1 and @xmath2 .
the region @xmath115 defined above will be used in the next section in the following way .
one point , let say @xmath19 , is located at the origin of @xmath115 .
the other point @xmath20 is chosen to be outside @xmath115 , that is @xmath117 .
let @xmath118 the conformal factor of a kerr metric with respect to @xmath20 .
then , the upper bounds ( [ eq:8])([eq:23 ] ) imply that @xmath118 is bounded in @xmath115 ( note that @xmath118 is not bounded at @xmath20 ) . for points in the interior of @xmath115 we have @xmath119 and
then the upper bound ( [ eq:8 ] ) in @xmath115 gives the following bound which depends only on @xmath23 and @xmath114 @xmath120 for the extreme case this bound reduces to @xmath121 the important point is that in the limit @xmath122 we have @xmath123 in @xmath115 ( and hence @xmath124 in @xmath115 ) .
the statement of lemma [ lemmarmenoscompacto ] is false for the extreme case @xmath75 .
the reason for this is that @xmath78 has a different behavior at the origin in that case ( see equation ( [ eq : te ] ) ) , which reflects the change in this limit from an asymptotically flat end to a cylindrical one . in figure [ fig - ext ]
we show the support of @xmath108 in the extreme case .
we see that the support of @xmath108 ` touches ' the origin and hence lies outside of the region @xmath115 .
this is the main technical difficulty to prove theorem [ t : main ] in the extreme limit .
to analyze that case we compute the following formula for @xmath78 .
[ rre ] let @xmath75 .
then , the scalar curvature @xmath78 is given by @xmath125 define @xmath126 by @xmath127 . the metrics @xmath6 and @xmath77
are related by the conformal transformation @xmath128 , hence the relation between the corresponding scalars curvatures @xmath76 and @xmath78 is given by @xmath129 where @xmath130 is the flat laplacian in 3-dimensions .
the important step is that the kerr metric satisfies the stationary and axially symmetric equations . for the extreme case ,
these equations imply ( see @xcite ) @xmath131 we emphasize that in our coordinates equation ( [ eq:17 ] ) is valid only for the extreme case ( see the discussion in @xcite ) . combining ( [ eq:16 ] ) , ( [ eq:17 ] ) and
the expression for @xmath132 which comes from the hamiltonian constraint equation ( [ eq:3 ] ) and ( [ eq : kk ] ) the result follows .
( shaded region ) is not contained in the region @xmath115.,height=151 ]
the important feature of the metric ( [ brillsuma ] ) is that its ricci scalar has a simple decomposition in terms of the ricci scalars of each individual kerr metric , namely @xmath133 where @xmath134 , @xmath135 and @xmath136 .
note that the two terms in ( [ rtot ] ) are symmetric with respect to the points labels @xmath90 and @xmath137 . for the non - extreme case the formula ( [ rtot ] ) , together with previous lemmas , gives the desired result .
[ teoremanoextremo ] assume that @xmath138 , @xmath139 .
then there exists a constant @xmath57 such that if @xmath140 , then @xmath141 .
moreover , if @xmath60 or @xmath61 is different from zero then we have @xmath142 on @xmath17 for @xmath140 .
to analyze the sign of @xmath56 we study each of the two terms in ( [ rtot ] ) separately .
since they are symmetric in @xmath90 and @xmath137 it is enough to analyze only one of them . if @xmath63 we have @xmath64 ( see equation ( [ eq : schw ] ) ) and then the conclusion of the theorem follows .
hence , in the following we will assume that either @xmath60 or @xmath61 is different from zero .
without loss of generality we take @xmath143 and analyze the first term in ( [ rtot ] ) . by lemma [ t : psi ] the factor which multiplies
the parenthesis on that term is positive definite .
then , we only need to analyze the sign of @xmath144 take a fixed coordinate system @xmath7 centered at the point @xmath19 . in these coordinates
we have @xmath20 located at @xmath145 .
note that , in these coordinates , the only function in ( [ eq : k1 ] ) which depends on @xmath23 is @xmath42 .
using lemma [ t : psi ] and [ lemmartildepositivo ] we known that the only function which is not positive in the definition of @xmath146 is @xmath147 .
let @xmath115 be the region defined in lemma [ lemmarmenoscompacto ] centered at @xmath19 , with constants @xmath148 .
take @xmath23 such that @xmath117 .
that is , the point @xmath20 is not in @xmath115 .
since @xmath147 is positive outside @xmath115 , it follows that @xmath146 is positive outside @xmath115 .
note that the constants @xmath148 in ( [ e ] ) do not depend on @xmath23 .
we now turn our attention to the interior of @xmath115 . by lemma [ lemmartildepositivo ] we know that @xmath149 is strictly positive in @xmath115 .
we use the upper bound ( [ eq:7 ] ) for @xmath42 in the interior of @xmath115 .
we have that @xmath124 in @xmath115 as @xmath122 .
since @xmath149 and @xmath147 do not depend on @xmath23 it follows that there exists a constant @xmath150 such that @xmath146 is positive in the interior of @xmath115 for all @xmath151 .
for @xmath152 we repeat this analysis to obtain a constant @xmath153 such that @xmath154 is positive for all @xmath155 .
we chose @xmath57 to be the maximum between @xmath150 and @xmath153 , from which the conclusion of the theorem follows .
the proof in the extreme case is more delicate because lemma [ lemmartildepositivo ] is no longer valid .
however , a more detailed analysis of the asymptotic expansion at the origin of the function @xmath146 , using lemma [ rre ] , allows the desired extension .
[ teoremaextremo ] assume that either @xmath54 , @xmath156 or @xmath55 , @xmath157 or @xmath54 , @xmath158
. then there exists a constant @xmath57 such that if @xmath159 then @xmath160 everywhere on @xmath17 .
moreover , if @xmath60 or @xmath61 are different from zero , then @xmath161 on @xmath17 for @xmath159 . if @xmath63 then the result is trivial since @xmath64 in that case
we assume that @xmath54 and @xmath143 .
we make no restriction on the parameter @xmath61 , that is @xmath162 . by the symmetry in @xmath90 and @xmath137 of ( [ rtot ] )
this assumption will cover all cases .
as in the non - extreme case , we use a coordinate system centered at @xmath19 .
we analyze @xmath146 directly .
we have that @xmath147 is positive outside some ball of large radius @xmath114 ( because the leading term in the asymptotic expansion ( [ eq : te ] ) is positive ) . as in the previous theorem we chose @xmath23 such that @xmath163 . from equation ( [ eq : k1 ] )
it follows that @xmath146 is positive outside the ball @xmath164 of radius @xmath114 centered at @xmath19 . to analyze the behavior of @xmath146 in the interior of @xmath164 we use the explicit expression for @xmath147 given by lemma [ rre ] and the explicit form of @xmath149 given by ( [ eq:3 ] ) and ( [ eq : kk ] ) to obtain @xmath165
this expression is clearly positive if @xmath166 by ( [ eq:13 ] ) we have that @xmath42 is bounded in @xmath164 and it goes to zero as @xmath122 . then , condition ( [ eq:27 ] ) can always be achieved for sufficiently large @xmath23 .
note that using the upper bound ( [ eq:13 ] ) the constant @xmath57 can be explicitly calculated .
we have proved that @xmath167 . to prove that it is strictly positive
, we use that @xmath168 only vanishes at the axis ( see equation ( [ eq : oo ] ) , ( [ eq:14 ] ) and ( [ eq:15 ] ) ) and at the axis we can explicitly compute the term @xmath169 to see that it is strictly positive .
we have constructed our metric as a sum of two kerr black hole metrics .
it is straightforward to generalize this result to include more than two black holes , that is to superpose more @xmath170 and @xmath171 in the definition of the conformal metric ( [ brillsuma ] ) .
we have assumed that the conformal metric is axially symmetric .
due to the particular simpler expression of the scalar curvature of this metric ( see equation ( [ rtot ] ) ) , this assumption represents a mayor simplification with respect to the case where the spins point in arbitrary directions .
we expect that a result similar to theorem [ t : main ] remains true in the general case
. however , proving such a result requires estimations for many new terms in the scalar curvature and it is not clear how to do this in an efficient way . even in the axially symmetric case ,
it is interesting to see what kind of properties of the individual metrics are necessary to make the superposition .
we have provided sufficient conditions which are satisfied by the kerr metric .
we do not known if these properties are also necessary .
our first attempt was to consider the superposition of two arbitrary brill metrics ( i.e metrics of the form ( [ brilldef ] ) ) with positive yamabe invariant .
but we were unable to prove that an analog of theorem [ t : main ] holds under these weak assumptions .
inspired by the kerr data , our next attempt was to impose that the negative part of the ricci scalar @xmath108 has compact support .
this seems to be a natural assumption , because the support of @xmath108 seems to play the role of the ` body ' and the separation can be the taken as the distance between these sets .
furthermore , for brill metrics the support of @xmath108 is always non trivial because the integral of the ricci scalar over @xmath18 is zero ( see @xcite ) .
however , for the proof we have also used another particular property of the kerr metric , namely @xmath80 .
it is not clear to us whether this condition is actually necessary .
it is a pleasure to thank robert geroch for discussions .
s. dain is supported by conicet ( argentina ) .
g. avila was partially supported by a ` conciencias ' fellowship , from the agencia crdoba ciencia ( argentina ) .
this work was supported in part by grant pip 6354/05 of conicet ( argentina ) , grant 05/b270 of secyt - unc ( argentina ) and the partner group grant of the max planck institute for gravitational physics , albert - einstein - institute ( germany ) .
in this section we study the behavior of kerr initial data ( with parameters @xmath2 and @xmath1 ) in the limits @xmath107 , @xmath172 and @xmath173 .
the explicit expressions for the functions @xmath10 and @xmath11 , which characterize the kerr intrinsic metric ( for boyer - lindquist slices ) @xmath6 , in terms of the coordinates @xmath174 are given by ( see , for example,@xcite ) @xmath175 where @xmath176 and @xmath177\sin^2\theta , \label{eq : x1}\\ & = \left ( \tilde r^2+a^2+\frac{2m\tilde r a^2 \sin^2 \theta}{\sigma}\right ) \sin^2 \theta .
\label{eq : x2}\end{aligned}\ ] ] where have defined @xmath178 , @xmath179 and @xmath180 finally , we analyze the behavior of @xmath78 near the axis @xmath22 . for a brill metric like ( [ brilldef ] ) ,
the ricci scalar is given by @xcite @xmath191 where @xmath192 is the flat laplacian in two dimensions , which in coordinates @xmath193 is given by @xmath194 the function @xmath10 defined by ( [ eq:51 ] ) vanishes at the axis @xmath195 this can be of course explicitly seen from ( [ eq:51 ] ) .
it is also a consequence of the regularity of the metric ( [ brilldef ] ) at the axis ( see @xcite ) . as a consequence of ( [ eq:1 ] )
, we have that the radial derivatives of @xmath10 evaluated at @xmath22 ( which are equivalent to @xmath24 derivatives evaluated at @xmath196 , and hence tangential to the axis where @xmath10 is constant ) vanish .
then , in order to calculate @xmath78 using the formula ( [ eq : rb ] ) we need to compute only the derivatives in @xmath197 . to calculate them ,
it is convenient to make the change of variable @xmath198 .
we have @xmath199 and @xmath200
. then we obtain @xmath201 where the following are @xmath202 , @xmath203 and @xmath204 are functions of @xmath205 given by @xmath206 calculating the required derivatives we get @xmath207.\ ] ] and @xmath208 + \sin^2{\varepsilon}\left[\frac{2}{\alpha+\cos{\varepsilon}}-\frac{1}{\alpha+\cos{\varepsilon } } \right]^2- \\ -\sin^2{\varepsilon}\left[\frac{2}{(\alpha+\cos{\varepsilon})^2}-\frac{1}{(\alpha+\cos{\varepsilon})^2 } \right]^2\end{gathered}\ ] ] using ( [ eq : sda ] ) we finally obtain @xmath209 + \left[\frac{1+\beta\cos{\varepsilon}}{(\beta+\cos{\varepsilon})^2}\right].\ ] ] using ( [ eq : rb ] ) and ( [ eq : qtt ] ) evaluating at the axis @xmath210 we get @xmath211.\ ] ] to prove that this expression is positive , consider the following inequality @xmath212 obtained by replacing the denominator @xmath213 by @xmath214 .
note that we assume @xmath215 in order to get the strict inequality .
this inequality implies @xmath216 and therefore we get our final result @xmath217 we emphasize that ( [ eq : posr ] ) is valid for both the non - extreme and extreme cases .
r. bartnik and j. isenberg . the constraint equations . in p.
t. chruciel and h. friedrich , editors , _ the einstein equations and large scale behavior of gravitational fields _ , pages 138 .
birhuser verlag , basel boston berlin , 2004 , gr - qc/0405092 . | an explicit 3-dimensional riemannian metric is constructed which can be interpreted as the ( conformal ) sum of two kerr black holes with aligned angular momentum .
when the separation distance between them is large we prove that this metric has positive ricci scalar and hence positive yamabe invariant .
this metric can be used to construct axially symmetric initial data for two kerr black holes with large angular momentum . |
strong - coupled relativistic quantum field theories ( rqft s ) play an essential role in any modern understanding of subatomic phenomena from the higgs mechanism to fundamental theories of nucleon and nuclear phenomena .
there are important theoretical issues , such as spontaneous symmetry breaking , which can be studied with relatively simple scalar field theories .
unfortunately , even the simplest interacting relativistic quantum field theories are notoriously difficult to solve .
a classic example is the issue of triviality in 4-dimensional @xmath0 scalar field theory which has been tackled over the years using nearly the full complement of weapons in the theorists arsenal ( loop expansion , strong coupling- and @xmath5-expansions , truncated systems of schwinger - dyson ( sd ) equations , renormalization group ( rg ) techniques and monte carlo studies ) , _ without _ yielding a definitive answer .
e.g. _ , ref.@xcite . )
it is self - evident that accurate and relatively simple approximate methods for solving rqft s would be of enormous value . in the present paper
we discuss approximations to @xmath0 theories emphasizing throughout comparison with lattice monte carlo calculations which , in a sense to be made explicit below , we consider exact . in particular
we will focus on the uniform properties which can be expressed in terms of `` effective potentials '' ( or more precisely , `` constraint effective potentials '' as defined , _
e.g. _ , in ref.@xcite ) from which , for example , zero - momentum @xmath6-point couplings may be extracted .
the present results may be summarized most succinctly by stating that the renormalization group ( rg ) effective potential
evaluated so as to account for lattice artifacts provides a spectacularly accurate description of its lattice counterpart with little calculational complexity for a wide variety of scalar field theories including symmetric and broken phases in 2 , 3 , and 4-d with o(1 ) and o(2 ) symmetries .
in contrast , other methods such as the use of truncated sd equations@xcite yield only qualitative agreement at best . moreover ,
as will be emphasized below , the simple rg effective potential used here is known to be the first term in a momentum expansion meaning that systematic improvements in the approximation are possible .
perhaps even more important is the evidence that in contrast to , _
e.g. _ , the hierarchy of sd equations this expansion appears to be strongly convergent for cases with strong coupling .
indeed , estimates of next - to - leading order correction to be made below are quite small but consistently improve the already extraordinary agreement with lattice results . finally , as will also be addressed below , the rg approach is ideally suited to the construction of _ effective _ low energy theories which can economically account for the long - wavelength behavior of the full theories .
we will also discuss some anticipated exploitations of this possibility .
in section ii we review briefly the history of the version of the rg effective potential used in the present work providing highlights and references to several other recent studies . in section iii
we provide a heuristic derivation of the rg equation for the effective potential and compare with truncated schwinger - dyson . in section
iv we describe the methods we use in extracting the effective potential from monte carlo simulations and in solving the rg equations numerically . in section v we present the results of our calculations comparing monte carlo and rg descriptions for broken and symmetric phases of o(1 ) and o(2 ) @xmath0 field theories in 3- and 4-dimensions .
we now trace the development of the form of the rg equations used in the present study .
the original work by wilson on the renormalization group@xcite was in essence a discussion of how to obtain the effective action at long wavelengths from a bare action at short wavelengths .
he and his coworkers discussed a variety of approximations to the full ( and intractable ) rg equations . shortly thereafter , wegner and houghton@xcite derived an exact rg equation for the effective potential by integrating over thin shells in momentum space , a technique referred to as the `` sharp cutoff method . ''
( see ref.@xcite for an early discussion of this and alternative `` smooth cutoff '' methods . also see ref.@xcite for early discussions of many rg concepts .
ref.@xcite provides a later discussion of some of these ideas in the course of constructing a novel and intuitively appealing rg - based proof of the renormalizability of @xmath0 field theories . ) a seminal early application of the wegner and houghton method to @xmath0 theories was presented by hasenfratz and hasenfratz@xcite .
( see also ref.@xcite for an illuminating comparison of analyses of the cutoff dependence of the higgs mass using the wilson - based and wegner / houghton - based rg methods . )
these authors obtained a tractable approximation to the exact wegner and houghton equation by projecting the full effective action onto an effective potential for which all @xmath6-point couplings are momentum independent .
they used the resulting rg equation to extract information about rg flows in the vicinity of wilson and gaussian fixed points in @xmath1 and to establish the absence of a wilson fixed point in @xmath2 .
they also extracted various critical exponents for the @xmath1 case and found them to be in good agreement with values determined by a variety of other techniques .
there has been a recent resurgence of interest in rg effective potentials .
wetterich and coworkers@xcite , for example , have extended the approach of hasenfratz and hasenfratz using the so - called `` smooth cutoff '' method in deriving their rg equations .
their technique permits evaluation of , _
e.g. _ , the anomalous dimension , @xmath7 , which is implicitly zero in the hasenfratz and hasenfratz approximation and yields values of critical exponents which are in impressive agreement with those determined by other methods .
( see ref.@xcite for an early calculation of @xmath7 using an extension of one of wilson s original approximate rg equations . )
this increase in calculational power comes , however , at the cost of a substantial increase in numerical complexity . in other work
, morris@xcite gives an in - depth discussion of the rg effective potential as derived using both sharp and smooth cutoffs .
he also@xcite developes a version of the systematic momentum expansion alluded to above and presents an analysis of @xmath0 in @xmath1 using the first correction to what is in effect the hasenfratz and hasenfratz method ( which appears as the zero - momentum limit in both the sharp- and smooth - cutoff derivations ) . without attempting to be comprehensive , we mention other related papers which appear here as refs.@xcite , @xcite and @xcite .
derivations of rg equations for the effective action or their reduction to expressions governing the effective potential appear in refs.@xcite and will not be reproduced here .
instead we provide a heuristic and , we hope , intuitive development of the rg equation in the zero - momentum limit which has as its point of departure the familiar expression for the one - loop contribution to the effective action of a uniform system . in @xmath8
euclidean dimensions , the bare action for a one - component @xmath0 theory is given by @xmath9 \label{ca}\ ] ] where @xmath10 the one - loop perturbative correction to the bare potential for a uniform system corresponds to the familiar expression ( primes denote differentiation with respect to @xmath11 ) @xmath12 \label{cc}\ ] ] which can be interpreted physically as the sum of zero - point energies of vacuum modes for a free scalar meson theory with meson mass - squared @xmath13 .
this quantity is divergent and we regularize it by adopting a uv cutoff @xmath14 : @xmath15 \label{cd}\ ] ] we note that the lowest order truncated schwinger - dyson equation can be obtained from the following expression for what we will call the schwinger - dyson effective potential , @xmath16 , in which the 1-loop contribution @xmath17 as given above is self - consistent in the following sense : @xmath18 .
\label{ce}\ ] ] the rg equation for the effective potential @xmath19 may be obtained heuristically by starting with the loop expansion of the effective action , the first term of which is given in eq .
assume we know @xmath20 , the effective potential at momentum scale @xmath21 .
we then have the following expression for @xmath22 , the effective potential at a slightly smaller momentum scale : @xmath23 \cr & = & u(\phi,\lambda)+{a_d\over 2 } \lambda^{d-1 } \ln\big[\lambda^2+u''(\phi,\lambda)\big]\ d\lambda + o(d\lambda^2)\end{aligned}\ ] ] where @xmath24 and the higher order loops contribute at @xmath25 .
taylor expanding the left hand side and taking the limit @xmath26 gives the rg equation : @xmath27 .
\label{ch}\ ] ] the essential point is that where the eq . [ ce ] requires @xmath16 be made self - consistent across all momentum scales at once , eq . [ ch ] ensures that @xmath19 is self - consistent at every scale .
put another way , the evolution ( or `` flow '' ) of @xmath19 at some scale @xmath21 is determined self - consistently by @xmath19 at that same scale .
we make think of @xmath16 as the result of the crudest possible discretization of the flow which yields @xmath19 , namely one in which the integration of the differential equation in eq . [ ch ] is carried out in a single step .
the extension to @xmath28 is straight - forward . defining @xmath29 and with primes denoting differentiation with respect to @xmath30
, we find @xmath31 + ( n-1 ) \ln\bigl[\lambda^2+u_{rg}^{'}(\rho,\lambda)/\rho\bigr]\biggr\ } .
\label{ci}\ ] ] to make contact with , _
e.g. _ , hasenfratz and hasenfratz@xcite , we scale all dimensionful quantities by the appropriate power of @xmath21 so as to render them dimensionless . for example , @xmath32 ( where in this approximation the anomalous dimension , @xmath7 , vanishes ) . designating the rescaled potential as @xmath33 where @xmath34 , the one - component equation ( eq . [ ch ] ) becomes , after dropping irrelevant constants , becomes @xmath35 + ( 1-d/2)\ x\ \tilde u_{rg}^{'}(x , t)+d\ \tilde u_{rg}(x , t ) \label{cj}\end{aligned}\ ] ] where primes now denote differentiation with respect to @xmath36 .
this is eq .
22 of ref .
@xcite ( with @xmath37 ) . finally , defining @xmath38 , eq .
[ cj ] becomes @xmath39 where the `` dot '' means differentiation with respect to @xmath40 .
this is eq .
9 of ref .
@xcite and is a form of the rg equation for the effective potential frequently encountered in the literature .
these rg equations may be integrated to find the effective potential at momentum scale @xmath21 subject to the _ boundary condition _ that the bare and effective potentials coincide at the uv cutoff scale @xmath14 . in so doing , quantum corrections associated with single meson modes with momenta between @xmath21 and @xmath14
are ( approximately ) included in the effective potential while contributions from modes at lower frequencies must still be treated explicitly .
as touched upon above , the power of the rg approach derives from its self - consistency _ at each momentum scale_. this feature is evident in eq .
[ ch ] where it is clear that the evolution of @xmath19 at momentum scale @xmath21 depends on a quantum correction which is consistent with the potential itself at that particular momentum scale .
another important and perhaps surprising aspect of the rg equations is that while they have the _ form _ of a truncated loop expansion no such truncation has been made .
as originally shown by wegner and houghton@xcite higher - order loops are suppressed by powers of the differential @xmath41 and their equation therefore has the form of a one - loop result but is in fact _ exact_. thus the rg equation for the effective potential , obtained by hasenfratz and hasenfratz as an approximation to the wegner and houghton expression has nothing to do with explicitly dropping high order quantum corrections but instead depends on projecting the full effective action onto a form with independent @xmath6-particle couplings .
( this is what is meant by the `` zero - momentum '' rg effective potential . ) of course , just as is the case when solving truncated sd equations@xcite , quantum corrections of arbitrarily high order in a perturbative loop expansion are automatically generated in the course of integrating the rg equation .
( the relationship between the rg equation and the loop expansion was touched upon by polchinski@xcite and addressed in more detail by morris@xcite who also made a direct comparisons with the sd expression given below in eq .
[ ceb ] . )
once the effective potential is obtained , the renormalized mass and 4-point coupling are given by : @xmath42 @xmath43 where @xmath44 is the value of the field at the minimum of @xmath45 .
for example in the case of the schwinger - dyson equation , eq .
[ ce ] , in the symmetric phase where @xmath46 , we have @xmath47 where we have used @xmath48 and @xmath49 . approximating @xmath50 by @xmath51 yields the familiar ( see , _
e.g. _ , eq . 2.17 of ref.@xcite ) sd self - consistency relation for the `` dressed '' mass : @xmath52 where the self - consistent nature of the sd method is apparent .
( we note that the approximation defined by this equation for @xmath53 is also known@xcite as the `` cactus '' approximation . ) likewise , within the same approximation , the 4-point is @xmath54 .\label{cec}\ ] ] examination of rg renormalized masses and 4-point couplings following from eq .
[ ch ] ( or the equivalent expressions given subsequently ) requires finding numerical solutions to the non - linear partial differential rg equations . in the next section
we discuss our numerical methods for solving the rg equations and for determining the `` exact '' effective potentials from our monte carlo calculations .
we have carried out rg calculations for the effective constraint potential in two complementary ways . in the first , appropriate to studying the continuum limit
, we treat @xmath21 as a continuous variable . in the second approach ,
appropriate to direct comparison with our monte carlo calculations which of course refer to a finite lattice , we treat the lattice normal modes explicitly .
we now give detailed descriptions of both methods . in the first approach
, we solve eq .
[ ch ] by treating @xmath21 as a continuous variable .
( we have also evaluated the equivalent expressions , eq . [ cj ] or eq . [ ck ] .
although one version is often better suited to a given calculation than the others , we found them to be entirely equivalent to one another as expected . )
we begin by expanding @xmath55 in a power series in @xmath11 truncated at @xmath56 : @xmath57 the boundary condition referred to in section iii is imposed by setting @xmath58 the range of the @xmath21 integration is divided into @xmath59 intervals of equal length @xmath60 : @xmath61 with @xmath62 where @xmath63 is the infrared scale appropriate to a lattice of @xmath64 sites and @xmath14 is chosen to yield the same phase space volume as the finite sum - over - modes .
a similar decomposition is made of a range of @xmath11 , the other independent variable : @xmath65 each step in the integration is accomplished by first making a least squares fit of the polynomial on the r.h.s of eq . [ da ] to @xmath66 in order to determine the @xmath67 and then computing @xmath68 \label{de}\ ] ] for each @xmath69 where @xmath70 is computed analytically from the polynomial fit . starting with @xmath71
, this procedure is iterated until ultimately the @xmath72 values of @xmath73 are found .
we refer to this as the `` continuum rg effective potential '' ( c - rg ) .
obviously this method yields directly the @xmath74-point couplings , @xmath75 , as well .
the calculation can sometimes be unstable in the vicinity of the @xmath11-endpoint , @xmath76 , and some care is required in choosing values of @xmath77 , @xmath59 and @xmath78 to ensure stable integration . in the calculations to be presented below ,
we typically used @xmath79 in our second approach , an alternate rg calculation was formulated to facilitate comparison with the lattice results to be described below .
the derivative terms in the action ( eq .
[ ca ] ) are approximated on the lattice by a finite difference formula . for a uniform system on a @xmath64 lattice with periodic boundary conditions , the vacuum modes are momentum eigenstates with wavenumbers in each direction given by @xmath80 , @xmath81 where we take the lattice constant to be unity , @xmath82 .
we make the following identification : @xmath83 where @xmath84 .
now define the @xmath64 scalar quantities @xmath85 then sort them by magnitude and relabel : @xmath86 where @xmath87 and @xmath88 .
our `` latticized '' version of the rg equation for the effective potential , eq .
[ de ] , is @xmath89 \label{di}\ ] ] with the boundary condition @xmath90 we will refer to this as the `` lattice rg effective potential '' ( l - rg ) .
the lattice monte carlo calculations to be presented below employed @xmath8-dimensional cubic lattices with periodic boundary conditions in conjunction with the metropolis algorithm to sample the @xmath11-configurations with weights given by the exponentiated action .
the monte carlo observable of primary interest here is the `` constraint effective potential''@xcite which we refer to as the `` mc effective potential '' and designate by @xmath45 .
formally , this quantity is defined via @xmath91= \int{\cal d}\{\phi\}\ \delta(\bar\phi-\sum_i\phi_i/ n_{lat}^d)\ \exp\bigl[-s[\phi]\bigr ] \label{dk}\ ] ] where @xmath92 means a sum over all lattice sites and @xmath93 is the _ average _ field on the lattice for a given field configuration in the path integral .
since the monte carlo method generates sequences of field configurations , @xmath94_n$ ] , weighted by @xmath95_n\bigr]$ ] , @xmath96 is extracted from the lattice monte carlo calculation from a histogram of the number of configurations generated _ vs _ the average value of the field for those configurations .
let @xmath97 be the number of configurations with average field values in an interval @xmath98 about @xmath93 . then @xmath99\ d\bar\phi \label{dla}\ ] ] and , to within an overall constant , @xmath100 for @xmath28 theories this expression is modified slightly .
let @xmath101 be the average value of @xmath102 on the lattice and let @xmath103 be the number of configurations in an interval @xmath104 about @xmath101 . then , since @xmath105\ d\bar\rho , \label{dm}\ ] ] we have @xmath106 .
\label{dn}\ ] ] of course the most probable average field values are near the minimum of @xmath45 .
in order to determine @xmath45 away from its minimum ; _ i.e. _ , to sample a range of relatively improbable values of @xmath93 ( or @xmath101 ) , it is often necessary to add a constraint to the updating algorithm which rejects `` moves '' which take the average field value outside a preset `` window''@xcite . by matching @xmath45 s for a number of adjacent windows , it is possible to detemine the lattice effective potential over a wide range of average field values .
( see ref .
@xcite for an interesting alternative method . )
finally we comment on the `` convexity '' of @xmath45 .
it can be shown by very general arguments@xcite that the condition @xmath107 ( primes denote differentiation with respect to @xmath93 ) must be satisfied for all @xmath93 .
presently , we will show @xmath45 s for `` broken '' symmetry phases of @xmath0 theories which _ do not _ satisfy the convexity condition .
the apparent contradiction is resolved by noting that convexity is required only in the thermodynamic limit ; _ i.e. _ , as @xmath108 . for the finite lattice results to be presented below ,
convexity is not a fundamental property of @xmath45 .
we now present mc effective potentials as defined in eqs .
[ dl ] or [ dn ] for a variety of cases .
we also show the corresponding continuum and lattice rg effective potentials of eqs . [ de ] and [ di ] , respectively . in computing the former rg potential
, we always set the ir cutoff ; _ i.e. _ , @xmath109 in eq .
[ dc ] , to @xmath110 . fixing the uv cutoff , @xmath14 , is ambiguous .
we adopt the prescription that @xmath14 should be chosen so that the phase space volumes integrated over should be the same for continuum and lattice calculations .
this correspondance is made quantitative by setting @xmath111 in eq .
[ df ] which implies , in effect , @xmath112 then , _ e.g. _ , @xmath113 in the calculations shown in this work the bare coupling was fixed at @xmath114 while the bare mass @xmath115 was adjusted to give a desired renormalized mass .
the mc effective potentials were fit with a polynomial of the form @xmath116 to determine the renormalized mc masses and 4-point couplings . in our rg calculations
we also kept the 4-point coupling fixed at @xmath114 .
we now describe our procedure for determining the bare mass in the rg calculations . for symmetric phases ,
the rg bare masses were adjusted to yield @xmath117 . for broken phases ,
the rg bare masses were adjusted so as to give the best fit `` by eye '' to @xmath45 in the vicinity of its minimum ( see discussion below ) .
since the mc and rg effective potentials in the symmetric phase behave identically at small @xmath11 _ by construction _ , we typically compare for both symmetric and broken phases
mc and rg results for the subtracted quantity @xmath118 rather than results for the @xmath119 s themselves .
table 1 summarizes results of @xmath3 and @xmath4 calculations in both symmetric ( s ) and broken ( b ) phases in @xmath1 and @xmath2 . for the rg results , @xmath120 and @xmath121 .
the table shows that our bare masses squared are typically @xmath122 while the renormalized masses squared are @xmath123 to @xmath124 with renormalized couplings of @xmath125 to @xmath126 .
clearly quantum corrections are large and simple perturbation theory ( _ e.g. _ , the loop expansion ) can not account for them . for comparison purposes we examine results from the simplest schwinger - dyson ( sd ) equations@xcite .
consider , for example , a `` latticized '' @xmath1 , @xmath3 version of eq .
[ ceb ] : @xmath127 where the @xmath128 s are defined in eq .
[ dh ] . in this expression ,
a value of @xmath129 is required to give @xmath130 for @xmath114 and @xmath131 while the corresponding mc calculation yielding @xmath132 requires @xmath133 . the latticized expression for the sd 4-point coupling , eq .
[ cec ] , is @xmath134 , \label{ed}\ ] ] and gives @xmath135 as compared to the mc value of @xmath136 . as another example , we find that , for @xmath2 , @xmath3 , the mc - determined renormalized mass is @xmath137 using @xmath138 compared to the latticized sd value @xmath139 needed to yield the same renormalized mass .
similarly , the latticized sd equation for the 4-point coupling gives @xmath140 _ vs _ @xmath141 . not surprisingly , these simple sd equations do not in general provide a quantitatively accurate picture of the quantum effects . although we do not display sd results for the effective potential ,
it is apparent from the 4-point couplings that the sd - derived effective potential will yield at best only a qualitative description of the mc effective potential .
we turn now to the monte carlo calculations of the effective potential and comparisons with the two rg effective potentials , c - rg and l - rg , discussed in the previous section . as mentioned previously , since we adjust the bare mass in the rg calculations to give the mc fit mass , @xmath142 , in most of the figures that follow we display the subtracted quantity , @xmath118 , to emphasize the predictive power of our calculations .
figure [ figb ] displays the subtracted ( _ i.e. _ , @xmath118 ) mc effective potential for @xmath1 , @xmath3 in the symmetric phase on a @xmath143 lattice . a mc bare mass squared of @xmath144 yields @xmath132 and @xmath145 , as mentioned above and shown in table 1 . note that we have used the `` windowing '' technique described in section iv to extend the range of @xmath11 at which the mc effective potential is determined from a maximum value of about @xmath146 which is feasible without windowing to a maximum value of @xmath147 . ( the relative normalizations of adjacent windows were determined by matching intermediate values well determined in both windows . )
however , the values of @xmath142 and @xmath148 were determined by fits to the mc values of @xmath45 for @xmath149 only . to reproduce this value of @xmath142 , a c - rg bare mass squared of @xmath150 was required .
of course this numerical value depends on our prescription for @xmath14 which was discussed above .
( however , the c - rg effective potential is only weakly dependent on @xmath14 provided the rg bare mass is adjusted to give a particular value of the renormalized mass , @xmath74 . )
the c - rg value for the effective 4-point coupling is @xmath151 and it is evident from figure [ figb ] that the c - rg calculation ( solid line ) does _ not _ accurately reproduce the quartic and higher @xmath11-dependence of the mc effective potential . by comparison
, a bare mass squared of @xmath152 was required in the l - rg calculation to give @xmath153 .
that the l - rg bare mass differs from the mc value by less than 1% is especially gratifying ; in contrast , the c - rg bare mass differs by about 12% while the sd difference is also about 12% but in the opposite direction .
the l - rg value of @xmath154 is also quite close to @xmath145 .
however , inspection of figure [ figb ] where the agreement between mc and l - rg results over the wide range of @xmath11 spanned using the windowing technique is spectacular suggests that even this small discrepancy is probably due to the fact that the lattice @xmath19 contains terms of order @xmath155 and higher while our fit to @xmath45 does not .
we note that our limited mc `` data '' are not good enough to extract the @xmath155 or higher terms reliably . also appearing in figure [ figb ] as the dashed line is the l - rg result omitting all contributions of order @xmath155 and higher .
comparison of this curve with the mc results show that these high order contributions are important for large values of @xmath11 and that the l - rg accounts for them quite well . to summarize , the l - rg calculation which includes lattice effects in the integration of the differential equation for @xmath19 yields _ excellent quantitative _ agreement with mc effective potentials even on a relatively small lattice provided a very small adjustment in the rg bare mass is permitted . for the c - rg calculation ,
comparison with the mc results is not straightforward because of the ambiguity in choosing @xmath14 which then results in ambigiuties in the bare mass .
futhermore , even when the bare mass is allowed to vary freely , agreement is poorer than for the l - rg .
unsubtracted effective potentials for the @xmath1 , @xmath3 case in the _ broken _ phase appear in figure [ figca ] .
as mentioned above , the bare masses were adjusted in the rg calculations to yield the correct location of the minimum . from figure [ figca ]
we see again that the description of the effective potential provided by the l - rg calculation is excellent .
the c - rg calculation , however , is incapable of simultaneously reproducing the location in @xmath11 and the depth of the minimum in the mc effective potential .
the fit shown is a ( non - unique ) compromise .
the subtracted potentials for this case are shown in figure [ figc ] and again demonstrate the superior description of @xmath155 and higher contributions to the effective potential provided by the l - rg .
indeed , the description of the mc results by the l - rg is again spectacular .
the bare mass squared used in this calculation is @xmath156 which differs from the mc value of @xmath157 by less than 1% .
( the high quality of this fit again leads us to attribute the differences between @xmath148 and @xmath158 from the l - rg calculation 1.48 and 0.52 , respectively to the absence of @xmath155 and higher terms in the fitting function rather than to shortcomings of the l - rg method . )
similar findings were obtained for @xmath2 , @xmath3 as well as @xmath4 cases in @xmath1 and @xmath2 in both symmetric and broken phases .
values of the bare and renormalized masses and 4-point couplings are reported in table i. the l - rg description of the subtracted mc effective potential is nearly perfect in all cases and is distinctly superior to that provided by the c - rg .
we note that this is true even when the correlation length ( _ i.e. _ , @xmath159 ) is large compared to the overall lattice size . for our @xmath2 ,
@xmath4 case , a small mc mass of @xmath160 was found ( _ i.e. _ , the correlation length was @xmath161 ) on an @xmath162 lattice ( ! ) and , yet , as shown in figure [ figh ] , the l - rg calculation agrees almost prefectly with the subtracted mc effective potential .
furthermore , the l - rg bare masses never differ by more than 1.5% from the mc values and are typically much closer .
overall , we conclude that the l - rg method provides an extraordinarily accurate , semianalytic description of lattice mc calculations of the effective potential at a tiny fraction of the computational cost .
we have presented a discussion of the renormalization group ( rg ) effective potential for @xmath0 scalar field theories which stresses its relationship to the loop expansion and the schwinger - dyson ( sd ) method .
we show in a simple way that the unique feature of the rg method which is presumably the origin of its ability to describe the full theory as well as it does is its self - consistency at all momentum scales . we have formulated a version of the rg equation for the effective potential which explicitly treats finite lattice effects so that rg calculations may be compared directly with lattice monte carlo ( mc ) calculations without extrapolations of the latter to infinite lattice size and zero lattice spacing ( see ref .
@xcite and references contained therein for some examples of such extrapolations ) .
we compare such `` latticized '' or l - rg potentials along with standard continuum or c - rg potentials@xcite to constraint effective potentials obtained via mc methods .
the l - rg effective potentials are in excellent quantitative agreement with the mc results for @xmath1 and @xmath2 , @xmath3 and @xmath4 cases in both symmetric and broken phases .
more specifically , we find that , upon adjusting the l - rg bare masses so as to yield renormalized masses equal to those found in the mc calculations , the l - rg bare masses never differ by more than 1.5% from the mc values and are typically much closer .
the renormalized 4-point couplings found by solving the l - rg equations are also typically in close agreement with the mc values . in at least one case for which the mc effective potential was determined over a wide range of @xmath11 using windowing techniques , there is strong evidence that @xmath155 and higher contributions are also well accounted for in the l - rg approach . in another case , we showed that the l - rg approach works very well even when the correlation length is appreciably larger than the overall lattice size ( @xmath163 ) .
comparison of c - rg calculations with the mc results was found to be ambiguous ; however , even when ambiguous parameters were tuned to give optimal agreement with the mc quantities , this accord was found to be inferior to that achieved using the l - rg without tuning .
finally , latticized sd equations were seen to yield only a rough qualitative description of of mc masses and 4-point couplings .
we conclude that the l - rg method provides an very accurate , relatively simple description of lattice mc calculations .
it may prove to be a useful guide in extrapolating finite lattice results to the continuum limit since the relation between the l - rg and the c - rg is simple and well - defined .
we may also be able to apply it to some specific problems of current interest such as accounting for the `` universal effective potential '' for @xmath0 field theories in @xmath1 recently obtained via mc techniques by tsypin@xcite .
the success of the method encourages us to make various extensions of it .
for example , we plan to go beyond the `` zero - momentum '' approximation so as to be able to compute wavefunction renormalizations@xcite .
we have also begun to examine including yukawa - coupled fermions in the approach which would be a crucial step toward our long term goal of constructing effective theories of low - energy nuclear phenomena whose field theoretic structure is fully understood and for which we have an accurate and relatively simple approximation scheme of proven reliability .
the authors gratefully acknowledge valuable discussions with steven pollock and anna hasenfratz .
this work supported in part by the u.s.d.o.e .
.input and ouput quantities for monte carlo ( mc ) as well as continuum ( c - rg ) and lattice ( l - rg ) renormalization group calculations for @xmath1 and @xmath2 , @xmath3 and @xmath4 cases in both symmetric ( s ) and broken ( b ) phases .
the bare masses used are denoted by @xmath115 while the mc renormalized masses and 4-point couplings are @xmath142 and @xmath148 , respectively , as defined in the text and their rg counterparts are @xmath74 and @xmath158 .
the mc and l - rg calculations use the same value of @xmath164 . [ cols="^,^,^,^,^,^,^,^,^,^,^,^,^,^,^ " , ] | we study constraint effective potentials for various strongly interacting @xmath0 theories .
renormalization group ( rg ) equations for these quantities are discussed and a heuristic development of a commonly used rg approximation is presented which stresses the relationships among the loop expansion , the schwinger - dyson method and the renormalization group approach .
we extend the standard rg treatment to account explicitly for finite lattice effects .
constraint effective potentials are then evaluated using monte carlo ( mc ) techniques and careful comparisons are made with rg calculations .
explicit treatment of finite lattice effects is found to be essential in achieving quantitative agreement with the mc effective potentials .
excellent agreement is demonstrated for @xmath1 and @xmath2 , @xmath3 and @xmath4 cases in both symmetric and broken phases .
= 10000 = 10000 monte carlo and renormalization group effective potentials in scalar field theories department of physics , university of colorado , boulder co 80309 - 0446 department of physics , colorado school of mines , golden co 80401 |
motion of a test particle around a black hole is a very old topic of investigation to know the behavior of the gravitational field around black hole(for example [ 1 - 3 ] and any other book on general relativity ) .
such investigation was started long back when within a year of publication of general theory of relativity by einstein(1915)[4 - 7 ] , schwarzchild [ 8 ] gave a vacuum solution to the einstein field equations .
the solution describes the geometry of vacuum space - time outside a spherical massive body and is known today as schwarzchild black hole solution . at present in all text books on general relativity
there is an exhaustive study of the motion of a test particle around the schwarzchild black hole(see for example [ 1 - 3 ] ) . in the present work
, an investigation of the motion of a test particle is done around a general static non - rotating black hole .
a general formula for determining bending of light is evaluated and is tested for schwarzchild black hole.gravitational field outside the black hole is approximated by pseudo - newtonian(pn ) gravitational potential and is compared with the corresponding effective potential for test particle motion .
finally all results are verified for reissner - nordstrm black hole solution .
+ the paper is organized as follows : section ii deals with the motion of a massive test particle around a general spherically symmetric non - rotating black hole . also effective potential , energy and condition for circular orbit are determined for the black hole .
further some comments are presented point wise . in section iii trajectory of a photon and bending of light is studied .
pseudo - newtonian gravitational potential is determined and it is compared with effective potential in section iv . as an example
all results are deduced for reissner - nordstrm black hole in section v. the paper ends with a short conclusion in section vi .
the line element of a static spherically symmetric space time which describes a black hole can be written as , @xmath0 where @xmath1 is at least a @xmath2 function and it should satisfy the following conditions so that line element ( eq.1 ) describes a black hole solution : i ) @xmath1 must have a zero at some positive r ( say @xmath3 ) so that time dilation is infinite at @xmath3 , ii ) the kretschmann scalar ( @xmath4 ) should be finite at @xmath5 but it diverges at @xmath6 i.e. the space time described by equation ( 1 ) has curvature singularity only at @xmath6 . here @xmath7is the metric on unit two sphere .
suppose we consider the motion of a test particle of rest mass m around the black hole .
the corresponding lagrangian will be @xmath8 where @xmath9 is any affine parameter .
+ as the lagrangian has two cyclic co - ordinates @xmath10 and @xmath11 so the corresponding momenta must be constant .
this leads to @xmath12 a constant and @xmath13 is also a constant .
as the space - time is spherically symmetric so the motion is always confined to a plane which for convenience chosen to be the equatorial plane @xmath14 .
the explicit form of the momentum components are @xmath15 @xmath16 @xmath17 @xmath18 using the above expressions for momentum components in the energy momentum conservation relation @xmath19 we obtain @xmath20 where @xmath21 is called the ( square of the ) effective potential .
+ now differentiating both sides of equation ( 7 ) we have @xmath22 also from ( 5 ) the momentum in the @xmath11 direction gives @xmath23 so eliminating the affine parameter between @xmath24 and @xmath25 the differential equation of the trajectory of the particle in the equatorial plane is given by @xmath26\ ] ] which can be written as @xmath27 where @xmath28 we can make the following conclusions on the trajectory of the particle : + @xmath29 the energy of the particle should not be less than the potential @xmath30 i.e. for a given @xmath31 the trajectory should be such that the radial range is restricted to those radii for which @xmath32 is smaller than @xmath31 .
+ @xmath29 if @xmath33 for all values of @xmath34 then the particle comes from infinity and moves directly to the origin .
this is called the terminating escape orbit .
+ @xmath29 if @xmath35 has one positive zero then the particle either starts from finite distance moves directly to the origin ( known as terminating bound orbit ) or it may move on an escape orbit with a finite impact parameter @xmath36 .
if @xmath35 has two positive zeros then we have two possible cases : i. if @xmath33 between the two zeros then the trajectory is called periodic bound orbit like planetary orbit or ii . if @xmath37 between these two zeros then the trajectory is either an escape orbit or a terminating bound orbit .
+ @xmath29 the points where @xmath38 are known as turning points of the trajectory i.e. the value of r which satisfies @xmath39 are turning points and eq.(13 ) determines the potential curves .
+ @xmath29for circular orbit @xmath40 we have from ( 9)@xmath41 i.e. circular orbits are possible for those radial co ordinates which correspond to maximum ( unstable ) or minimum(stable ) of the potential .
thus for circular orbit we must have @xmath42
to determine the photon trajectory we shall proceed as before . here from the energy momentum conservation relation we have @xmath43 i.e. @xmath44 is the ( square of the ) effective potential . so differentiating both sides we get@xmath45 thus the differential path of a light ray is given by @xmath46}}\ ] ] where @xmath47 .
now for photon circular orbit we have @xmath48 i.e. the radius of the circular orbit satisfies @xmath49 one may note that the radius of photon circular orbit is independent of the angular momentum of photon .
in particular , for a schwarzchild black hole the radius of photon circular orbit is @xmath50 .for an ingoing photon choosing @xmath51 we have @xmath52}}\ ] ] where @xmath53 .
+ note that if @xmath54 is a constant then @xmath55 has the solution @xmath56 ( choosing the constant to be unity ) a straight line .
this is expected as f(r)=constant means the space time is minkowskian ( having no gravitational effect ) and photon trajectory will be straight line .
further we see that at large distance ( small @xmath57 ) the gravitational field due to the black hole will be negligible so we may expand @xmath54 in a power series of u i.e. @xmath58 so keeping up to first order in @xmath57 we have from @xmath55@xmath59}}\ ] ] let @xmath60 then @xmath61 .
then the above differential equation becomes @xmath62 so on integration , @xmath63 if we choose @xmath64 as the initial incoming direction of light i.e. @xmath65 as @xmath66 and as in the approximation @xmath67 corresponds to the smallest @xmath34 that photon can travel then @xmath68 so the angle of deflection is @xmath69 , as the photon comes from infinity to the point of closest approach .
hence the total deflection would be @xmath70 .
therefore , considering the straight line path the net amount of deflection will be @xmath71 if we take schwarzschild solution then we have @xmath72 and @xmath73 .
further , if it so happen that @xmath74 then we choose @xmath75 . using the transformation @xmath76 the net amount of deflection will be given by @xmath77
based on a general heuristic method([9 ] ) , the pn gravitational potential can be defined as @xmath78 where , r is the usual radial co - ordinate and @xmath79 is the general relativistic specific angular momentum i.e. @xmath80 is the ratio of the conserved angular momentum and energy per particle mass , related to the circular geodesic in the equatorial plane .
+ in newtonian theory , the gravitational potential @xmath81 is given by @xmath82 with @xmath83 , the newtonian angular momentum per mass of the particle moving in the circular orbit .
the motivation of choosing pn potential ( 27 ) is to match the newtonian angular momentum per particle mass on a circular orbit with the general relativistic angular momentum . in the present study , the general
relativistic conserved angular momentum and energy per particle mass for circular orbit are given by ( from equations ( 13 ) and ( 14 ) ) @xmath84^{\frac{1}{2}}\ ] ] and @xmath85}}\ ] ] i.e. @xmath86 hence from ( 27 ) the pn gravitational potential is @xmath87 where the constant of integration c is determined from the known result of the schwarzchild black hole as follows:(note that c has no physical meaning ) + for schwarzchild black hole , the well known paczyski - witta gravitational potential([10 ] ) is @xmath88 substituting in eq.(31 ) we get @xmath89 and hence the pn gravitational potential for a general spherically symmetric black hole described by eq.(1 ) is @xmath90\ ] ] as for static radius @xmath91 gravitational potential should be zero so from ( 33 ) we have @xmath92 .
hence from equations ( 28),(29 ) the circular orbit of the test particle exist for radius @xmath93 in the range([11 ] ) @xmath94 where @xmath95 satisfies @xmath96 i.e. @xmath95 is the photon circular orbit .
thus all circular orbits of the test particle are lower bounded by the photon circular orbit and are extended upto the static radius .
+ further , from eq.(33 ) we see that the pn potential diverges at the event horizon ( i.e.@xmath97 ) reaches its maximum value at @xmath98 ( where@xmath99 ) and then decreases for @xmath100 i.e. the gravitational field corresponding to pn potential become repulsive for @xmath100 . also if the metric(1 ) becomes asymptotically flat(i.e .
@xmath101 as @xmath102 ) then @xmath103 asymptotically([11],[12 ] ) .
moreover , for central gravitational fields , if we assume that the motion of the test particle is confined to the equatorial plane then for keplarian motion along the radial direction gives @xmath104 where e stands for total pn energy per particle mass and @xmath105 stands for pn effective potential per particle mass having explicit form ( [ 13 ] ) @xmath106 where @xmath107 is the pn gravitational potential given by eq.(33 ) and @xmath108 is the pn angular momentum per particle mass .
thus circular keplarian orbits are characterized by the extrema of @xmath105(i.e.@xmath109 ) and we have @xmath110^{2}}\ ] ] which can be written as ( using eq.s ( 28 ) and ( 29 ) ) @xmath111 the corresponding expression for energy is @xmath112\ ] ] using @xmath113 it can be written as @xmath114\ ] ] it is to be noted that angular momentum per particle mass is same for both general relativity as well as for pn effective potential theory .
+ we shall now examine the stability of circular orbits studied above . as stability criteria is determined by the extrema of the effective potential @xmath105 , so we have @xmath115 for stable circular orbit and @xmath116 for unstable circular orbit . due to identical nature of @xmath79 both gr and
pn potential theory have same criteria for stability
. further , inner and outer marginally stable circular orbits corresponds to extrema of @xmath117 and we have @xmath118^{2}=f(r)\left[3f'(r)+rf''(r)\right]\ ] ] hence for stability one should have @xmath119^{2}<f(r ) \left[3f'(r)+rf''(r ) \right]\ ] ] finally , from the above expressions ( eq.s ( 36 ) and ( 38 ) ) of @xmath117 and @xmath120 we have the following observations : + @xmath121 at the event horizon(@xmath97 ) both @xmath117 and @xmath120 diverge while @xmath113 and @xmath122 exist if @xmath123 $ ] is positive definite .
for example in case of schwarzchild - de sitter space time([11 ] ) they become finite while for our reissner - nordstrm black hole ( see next section ) they do not exist .
+ @xmath121 at the static radius ( @xmath124 ) both @xmath117 and @xmath120 vanish while @xmath113 and @xmath122 depend on choice of @xmath1 .
+ @xmath125 at the photon circular orbit ( @xmath126 ) both @xmath122 and @xmath113 diverge while @xmath127 and @xmath79 is finite there .
+ thus we conclude that both in pn approach and in relativistic approach the circular orbits are bounded from above by static radius while bound from below by event horizon in pn approach and that by the radius of photon orbit in relativistic approach .
this is due to the fact that we do not get photon circular orbit in pn approach([11 ] ) .
for reissner - nordstrm(r - n ) black hole solution we have @xmath128 where m being the mass and q the charge of black hole .
so for horizon we have @xmath97 i.e. @xmath129 black hole solution exists for @xmath130 and we have event horizon at @xmath131 and @xmath132 corresponds to black hole cauchy horizon .
when @xmath133 then both the horizons coincide and it is the case of extremal black hole .
the static radius is given by ( @xmath124 ) @xmath134 the radius at which @xmath135gives the solution@xmath136 one may note that @xmath137 so both the static radius and @xmath138 lie inside the event horizon . hence they have no physical significance for r - n black hole .
also the radius of photon circular orbits are given by @xmath139\ ] ] * relativistic theory * : + for circular orbit the conserved angular momentum and energy per particle mass are @xmath140 variation of @xmath122 with respect to the variation of @xmath141 and @xmath34 for @xmath142 has been shown in figure 1 [ variables in the figures are in any standard units ] .
figure 1 figure 1 : the figure shows the variation of relativistic conserved angular momentum @xmath122 with respect to variation of q and r for the choice @xmath142 .
@xmath143}{\sqrt{r^{2}-3mr+2q^{2}}}\ ] ] @xmath144 with effective potential @xmath145}\ ] ] * pseudo - newtonian theory * : + pn gravitational potential and effective potentials are @xmath146}\ ] ] the graph of @xmath107 for variation of both q and r with @xmath142 is presented in figure 2 . also in figure 3
, we have shown the dependence of @xmath107 by drawing the graphs of @xmath107 for three different values of q. figure 2 figure 2 : here variation of pn gravitational potential @xmath107 with variation of both q and r for @xmath142 is shown .
figure 3 figure 3:here pn gravitational potential @xmath107 is drawn for three different values of q. the upper one for @xmath147 , middle one for @xmath148 and the lower one for @xmath149 .
@xmath150}+\frac{l^{2}}{2r^{2}}\ ] ] also we have @xmath151 the graphical presentation of @xmath79 for the variation of both q and r with @xmath142 has been shown in figure 4 .
also graphically a comparative study of @xmath122(given by eq.(46 ) ) and @xmath79(given by eq.(52 ) ) has been presented in figure 5(a ) and 5(b ) for two different values of q. figure 4 figure 4:the figure shows the graphical representation of pn angular momentum per particle mass @xmath79 for variation of both q and r with @xmath142 figure 5(a ) figure 5(b ) figure 5(a ) and 5(b ) show the comparative study of the variation of relativistic angular momentum @xmath122(given by eq.(46 ) ) and pn angular momentum per particle mass @xmath79(given by eq.(52 ) ) for @xmath148 and @xmath152 respectively.where upper one gives @xmath122 and lower one @xmath79 @xmath153^{2}}\ ] ] the marginally stable circular orbits are given by the positive root of the equation @xmath154 if the cubic eq has three positive real roots(@xmath155)then @xmath156 and @xmath157 are respectively the radii of the outer and inner marginally stable circular orbits .
finally , for stable circular orbit we have @xmath158
in this work we give a general formulation of the trajectory of a test particle ( or a photon ) around any spherically symmetric black hole in four dimensional space time .
we also classify the trajectories by studying the possible positive zeros of the function @xmath28 so once @xmath1 is given for a given black hole we can immediately tell the trajectories of a test particle or a photon around it . in the pn approach physical parameters for circular orbits are evaluated and are compared with the corresponding quantities in relativistic treatments .
stability condition for the circular orbits are determined and bounds of the marginally stable circular orbits are compared in both formalism . as an example
we have applied our results for r - n black hole solution .
+ it is to be noted that the above analysis of the trajectory is not restricted to einstein gravity , it can also be applied to any black hole solution in a modified gravity theory .
further our analysis can be extended to any higher dimension .
for example , the metric ansatz for a @xmath159 dimensional black hole is written in the form @xmath160 where @xmath161 is the metric on unit @xmath162 sphere and is given by @xmath163 @xmath164 then due to spherical symmetry of the space time the motion of a test particle can be restricted to the equatorial plane defined by @xmath165 . as before energy @xmath31 and
the angular momentum @xmath166 are two conserved quantities and the differential equation for the path of the test particle becomes @xmath167\ ] ] where @xmath168 or @xmath169 for photon or massive particle .
finally , one may note that throughout our calculations we have not used any specific gravity theory .
so if we have a black hole solution given by equation ( 1 ) not only in einstein gravity but also in any other gravity theory , the above analysis of the trajectory of a test particle is valid .
moreover from equations ( 56 ) and ( 57 ) , we may conclude that above analysis of particle trajectory can be extended to any dimension of space time .
+ for future work , an extension of the above approach to non - spherical systems ( particularly axi - symmetric ) would be interesting .
+ * acknowledgement : * + the first author is thankful to dr.prabir kr .
mukherjee , department of physics , presidency college , kolkata , for valuable help in preparing the manuscript.the first author is also thankful to dst , govt.of india for awarding kvpy fellowship.the authors greatly acknowledge the warm hospitality at iucaa , pune where apart of the work has been done .
+ + * references : * + @xmath170 $ ] . narlikar.j.v _ lectures on general relativity and cosmology _ ( the macmillan company of india ) ( 1978 ) + @xmath171 $ ] schutz.bernerd _ a first course in general relativity _ ( cambridge university press)(1995 ) + @xmath172 $ ] ray.d.inverno _ introducing einstein s general theory of relativity _
( clarenden press , oxford)(2003 ) + @xmath173 $ ] einstein.a _ preuss , akad.wiss.berlin , sitzber , _ * 778*,*799*,*831 * and * 844 * ( 1915 ) + @xmath174 $ ] einstein.a _ preuss , akad.wiss.berlin , sitzber , _ * 142 * ( 1917 ) + @xmath175 $ ] einstein.a _ the meaning of relativity_(methuen , london ) ( 1951 ) + @xmath176 $ ] einstein .
relativity : the special and general theory _
( methuen , london ) ( 1920 ) + @xmath177 $ ] schwarzchild.k _ sitzber , deut.akad.wiss.berlin.ki.math-phys .
* 189 * ( 1923 ) + @xmath178 $ ] mukhopadhyay.b _ astrophysical journal _ * 581 * ( 2002 ) 427 .
+ @xmath179 $ ] paczyski.b and witta.p _ astron.astrophys . _
* 88 * ( 1980 ) 23 .
+ @xmath180 $ ] stuchlk.z and kov.j _ int.j.mod.phys.d _ * 17 * ( 2008 ) 2089 .
+ @xmath181 $ ] stuchlk.z , slan and kov.j _ class.quantum.grav _ * 26 * ( 2009 ) 215013 .
+ @xmath182 $ ] minser.c.w , thorne.k.s and wheeler.j.a , _ gravitation _ ( freeman , san francisco , 1973 ) . | trajectory of a test particle or a photon around a general spherical black hole is studied and bending of light trajectory is investigated .
pseudo - newtonian gravitational potential describing the gravitational field of the black hole is determined and is compared with the related effective potential for test particle motion . as an example , results are presented for reissner - nordstrm black hole . |
this paper deals with a species of a class of models on topological studies of triangulations .
such models have been studied in several contexts 2-d gravitation , froth , ( * ? ? ? * and references therein ) .
the variant we use here was introduced in @xcite , but it turned out that a very similar study was initiated earlier by aste and sherrington @xcite .
so , we hope that david will accept this paper as a small sign of recognition .
we reconsider here the model which was inspired by @xcite and introduced in @xcite .
for completeness we repeat the definition of the model : we fix a ( large ) number @xmath0 of nodes , half of which are red , and the other half blue .
these nodes are the nodes of a topological triangulation @xmath2 of the sphere @xmath3 .
the set of all possible such labelled triangulations will be denoted @xmath4 .
we define a dynamics on @xmath4 by the following metropolis algorithm whose elementary steps are flips ( t1 moves ) : a link is chosen uniformly at random ( among the @xmath5 links ) . in , if the link ab was chosen then the flip consists in replacing it by the link cd .
this move is not admissible if the link cd already exists before the move .
otherwise it is admissible .
note that the number of nodes , @xmath0 , does not change in this model .
however , we will be interested in the behavior for @xmath6 .
the metropolis algorithm is based on the energy function @xmath7 on @xmath4 which , for any triangulation @xmath8 , is defined as e(t)= _ i
( d_i-7)^2 + _ i ( d_i-5)^2 , where @xmath9 is the degree ( number of links ) of the node @xmath10
. thus , this energy favors 7 links for the blue nodes and 5 for the red ones .
_ mutatis mutandis _ , the detailed definition of the energy is not important for the discussion of the model , and we will stick to this particular form of the energy . given an admissible flip , compute the energy of the triangulation before and after the flip ; this defines = e _ -e _ .
an admissible flip is performed if either @xmath11 or , when @xmath12 , with probability @xmath13 , where @xmath14 is the inverse temperature of the system .
several properties of this model were discussed in @xcite , but here we study in more detail the dynamical properties of the model . in particular , we introduce a `` charge '' defined as follows : the _ charge _ of a red node is defined by @xmath15 and the charge of a blue node is defined by @xmath16 .
we will say the charge is a _ defect _ if it is @xmath17 and if it is @xmath18 . in general , the color of the charge will not matter and will not be mentioned . in principle
, all charges between @xmath19 and @xmath20 can occur , since @xmath21 , but , obviously , at low temperatures mostly the charges , , and will come into play .
the dynamics of the model is given by the metropolis algorithm . in it ,
a link is chosen uniformly at random among all possible links .
the change of energy induced by the flipping of this link is called @xmath22 . if @xmath23 the flip is performed , if @xmath12 the flip is performed with probability @xmath24 .
this process satisfies detailed balance , and most of the paper is dealing with the equilibrium properties of this process at low temperatures . because of the detailed balance , the equilibrium measure @xmath25 has the property that the probability to see a given state whose energy is @xmath7 is proportional to @xmath26 .
we use this elementary observation to argue that at low temperature there are only few defects , by which we mean that there are few red nodes whose degree is not 5 and also few blue ones whose degree is not 7 .
given that there are few of these `` defects '' , we further assume that the `` positions '' of these defects are random in the sense that there are no strong conditional expectations : for example , having a defect + 1 does not say that there is a defect -1 close - by .
the upshot of this way of reasoning , which we corroborate by numerical studies , is that one can approximate the dynamics by just looking at defects .
indeed , the full dynamics must be described by the evolution of correlation functions .
it would have to take into account correlation functions between the charges ( and the colors ) of , say , the 4 nodes on a pair of triangles sharing an edge . then , flipping that edge , the correlations of many neighboring triangles would be changed simultaneously , and this would necessitate considering a full hierarchy of correlations ( like bbgky ) .
what we will see is that in this model , these higher order correlation functions do not influence our basic understanding of what is going on .
in contrast , the euler relations play a small but not totally negligible role for the sizes of the systems we consider .
it will be useful to define throughout the paper the natural parameter e^- .
we are interested in a regime where the density @xmath27 of charges ( which equals @xmath28 ) is low but also , where the number @xmath29 of charges is large , so that good statistics and a certain independence of the euler relations is attained .
more precisely , we fix @xmath30 and @xmath31 , and require @xmath32 and @xmath33 .
we furthermore consider the limit of large @xmath0 .
the main result of this section is summarized in the following proposition : consider an equilibrium state at temperature @xmath34 satisfying the above conditions on @xmath0 and @xmath35 . 1 . at first order in @xmath35 ,
the only charges present in the system are simple defects @xmath36 .
their density is @xmath37 .
the distribution of the colors ( red or blue ) is independent in the limit @xmath38 .
the distribution of the charges is independent in the limit @xmath39 .
the meaning of @xmath39 above is that the quantities become more and more decorrelated as @xmath38 while still maintaining the inequalities @xmath32 and @xmath33 . in this paragraph
, we will calculate the energy of the stationary state in the limit specified above , as a function of the temperature .
[ est:1 ] consider the region @xmath40 and @xmath41 .
for sufficiently large @xmath42 and sufficiently small @xmath43 the density of charges @xmath27 is ce / n=2+(^2 ) . assuming equilibrium , by detailed balance , the probability to see a defect of charge @xmath44 is @xmath45 , while the probability to see higher charges is @xmath46 , by the assumption of equilibrium and the form of the hamiltonian , since , if @xmath47 then it is at least 4 .
so it remains to estimate the coefficient in front of the factor @xmath48 .
there are 4 cases to consider : the number of red nodes with degree 4 or 6 , resp . the number of blue nodes with degree 6 or 8 .
all these cases cost energy 1 per instance , and thus these 4 numbers are equal by the virial theorem .
we also need to estimate the cases with 0 charge , , blue nodes with 7 neighbors and red nodes with 5 neighbors , which appear again equally often , by the virial theorem .
since there are @xmath49 nodes of each color , and each of the colors has 2 states of defect 1 ( namely @xmath44 ) , we conclude that the expected total number of defects is [ e : twonepsilon ] 22 ( n/2 ) = 2n+(^2 ) .
we next calculate the probabilities that a randomly chosen link connects 2 red ( blue ) nodes .
we denote these probabilities by @xmath50 for red - red , @xmath51 for red - blue and so on .
if there are no defects , , at order @xmath52 , all red nodes have @xmath53 neighbors and all blue nodes have @xmath54 .
this leads to the following relations : 2p _ + p _ & = 5/6 , + 2p _ +
p _ & = 7/6 . assuming that the positions of the colors are uncorrelated , we find that the relative probabilities to find a red - red , resp
. blue - blue pair are p _ /
p _ = 25/49 .
this leads to @xmath55 , @xmath56 , and @xmath57 . in
we show that numerical simulations confirm this simple approximation to a very high degree of fidelity .
we adopt an approach similar to .
we use the hypothesis that the charges are randomly distributed over the nodes to calculate the probability of finding a link with a given neighborhood of charges and compare it to simulation results . in this case
however , given a link @xmath58 , the neighborhood we consider is the ordered set of all 4 nodes involved in its flipping .
for example in , this set would be @xmath59 where @xmath60 is the charge of the node @xmath61
. this choice will be very useful for to study the dynamics later on since it determines the energy cost of flipping a given link : ( ) & = _ n\{a , b } ( c(n)-1 ) ^2 - ( c(n ) ) ^2 + _ n\{c , d } ( c(n)+1 ) ^2 - ( c(n ) ) ^2 + & = 4 + 2 ( c(c)+c(d)-c(a)-c(b ) ) it is easy to enumerate all the various cases and the energy cost associated with each of them .
we restrict the discussion to those situations where the charges take values in @xmath62 . in principle , there are @xmath63 configurations , which are reduced to @xmath64 , by symmetry . they are summarized in table [ table:1 ] ( symmetrical cases omitted ) .
note that if the defects of the original configuration are bounded by @xmath44 , then @xmath22 varies between @xmath19 and @xmath65 .
[ table:1 ] we assume throughout that the number of red ( blue ) nodes is @xmath66 ( @xmath67 ) and that @xmath68 .
we denote by @xmath69 the probabilities to find charges @xmath44 , respectively . assuming that there are no other charges ( except 0 ) , we can write n ( p_- + p _ ) & = e , + n ( p_- - p _ ) & = 12 - , where the second equation follows from the euler formula
. in equilibrium , @xmath70 , by , and therefore we get [ e : pplusminus ] p_= 6/n /(2n ) + ( ^2 ) .
we will assume that @xmath71 so that the second term in can be neglected .
in a similar way , one can show that p_2=^4 + ( ^5 ) , and combining these we find that the probability of nodes with charge @xmath72 is p_0= 1 - 2+(^2 ) .
we next consider in more detail what happens in those pairs of triangles where a flip leads to @xmath73 .
looking again at we see that the case @xmath73 appears in 3 cases : case @xmath74 : one of a or b has charge and c or d has charge ( and the others , charge 0).case @xmath75 : a and b charge + , c and d charge 0.case @xmath76 : a and b charge 0 , c and d charge . continuing with the independence assumption , we now look at the probability to find a configuration of type @xmath75 , @xmath74 , and @xmath77 .
note that there are @xmath78 half - links emanating from the nodes , and we are to pair them up randomly .
note that if a site is red , it has @xmath79 , @xmath53 , @xmath80 outgoing links , depending on whether its charge is @xmath81 , @xmath72 , @xmath82 , respectively .
similarly , the numbers for a blue node are @xmath80 , @xmath54 , @xmath83 .
therefore , given that there are on average @xmath84 defects of type red-@xmath79 , red-@xmath80 , blue-@xmath80 , blue-@xmath83 , there will be @xmath85 links from the red-@xmath79 , @xmath86 from red-@xmath80 and blue-@xmath80 , and @xmath87 from blue-@xmath83 .
the blue-@xmath54 and red-@xmath53 occur with probability almost 1 and have therefore respectively @xmath88 and @xmath89 dangling edges ( with a correction factor @xmath90 which we omit throughout .
the probabilities to see such dangling edges are the quantities above , divided by @xmath78 , the total number of dangling edges . we get , omitting higher order terms : q _ = & ( 7p_+/6 ) ^2 p_0 ^ 2=49 ^ 2/36 , + q _
= & ( 5p_-/6 ) ^2 p_0 ^ 2=25 ^ 2/36 , + q _
= & 4 ( 5p_-/6 ) ( 7p_+/6 ) p_0 ^ 2=140 ^ 2/36 .
+ we also get , by looking at : p_=0 & = q _ + q _
+ q_=214 ^ 2/36 , + p_=2 & = 2 ( 7p_+/6 + 5p_-/6 ) p_0 ^3=4 , + p_=4 & = p_0 ^ 4=1- ( ) , the discussion of the other values of @xmath22 shows the limitations due to our closing assumptions : by the virial theorem , in total independence , we would simply have [ e : pde2 ] p_=0=p_=8p_=2=p_=6 .
but we could also have computed the probabilities as above , with the result : [ e : pde3 ] p_=-2 & = 2 ( 7p_+/6 ) ^2 ( 5p_-/6 ) p_0 + 2 ( 5p_-/6 ) ^2 ( 7p_+/6 ) p_0 + & 3.89 ^3 instead of @xmath91 given by the stationarity assumption , which proves that the distribution of defects is not completely uncorrelated .
we will say that the correlation is bounded by @xmath92 , and can thus be neglected in the limit @xmath38 . in figs .
[ f : energy ] and [ f : fitde0 ] we show with 2 examples that the numerical simulations confirm these simple approximations to a very high degree of fidelity . note that in @xcite , the _ uniform _ measure on @xmath4 was considered , and even this leads to correlations of degrees of neighboring nodes .
we can use the results of the previous section to estimate the dynamics of the system under the metropolis algorithm .
if a flip leads to an energy change @xmath22 then it is accepted in the metropolis algorithm with probability [ e : accept ] p_acceptance= ^(0 , ) .
on the other hand , the probabilities to pick a link with fixed @xmath22 are given by and .
multiplying these numbers with the probabilities of leads to an estimate of the probability that the flip in question actually happens .
the results are summarized in ( calculated this time with the method of ) .
[ table:2 ] * discussion * : inspection of shows that the events with the highest transition rate are those which cost no energy , followed by those which have an energy cost of @xmath93 .
also note that the probability to find a link which will lead to a given @xmath22 is equal to the quantity in the table times @xmath94 since then we neglect the metropolis factor .
this leads to a table with the same prefactors , but with a power @xmath95 .
in particular , _ in the steady state , the local landscape is given by the 3rd column of table [ table:2 ] : it is symmetric around @xmath96 . _
henceforth , we will only consider the 3 most frequent types of flips ( the others are an order @xmath48 less probable ) : 1 .
flips which change from 1 defect to 3 of them and which raise the energy by 2 .
these flips will be called _
creation events_. 2 . flips which start from 3 defects and end with 1 defect and which decrease the energy by 2 .
these flips will be called _
annihilation events_. creation and annihilation events are obviously dual to each other and equiprobable in the stationary state .
flips which do not change the energy , and in which a pair , , or is involved .
these flips are by far the most probable .
we will discuss below in more details the 3 configurations which lead to @xmath73 . as stated above , if @xmath97 , then over @xmath98 of the flips ( which are accepted by the metropolis algorithm ) do not change the energy .
it is clear that , in order to understand the dynamics of the system , one should start by studying these flips .
looking at we see that there are 3 candidates for @xmath73 and they all involve only 2 defects .
we will now show that the cases of and are quite different from that of ( and its 3 other variants ` -`,@xmath99 ) . in the first case , , which is similar to the case
, the local neighborhood looks like in . in this case
, what happens is a flipping back and forth between the 2 states , with probability @xmath100 ( the probability to choose the colored link ) .
the case is illustrated in . here a new , and important phenomenon appears : the pattern , which we will call a _ pair _
, is recreated , but _ at a new position _ a distance 1 away from the old one .
we will also say that the pair _ walks _ one step .
the more important observation is that the pairs of defects must walk along a _ predefined , 1-dimensional path _ as shown in .
this means that _ the @xmath101 motion of pairs is a * one*-dimensional random walk in the current triangulation @xmath2 .
this random walk ( flipping back and forth on the predefined path ) will continue until some other type of event happens .
_ as we have seen , a pair of opposite charges can move through the system without energy cost .
its motion is a 1d random walk along a fixed 1-dimensional path .
edges are still chosen randomly and will be flipped if possible and if the metropolis condition is met in case @xmath12 .
here we ask about the relative probabilities that a pair disappears , and we will show that _
predominantly a pair will die when it collides with a defect_. we need to compare 3 possibilities of which the first will be seen to be the most probable : 1 .
the random walk reaches another defect . 2 .
the pair is destroyed because a creation event involving 1 or 2 of its 2 defects occurs .
3 . two independent random walks meet .
our earlier discussion says that the concentration of pairs @xmath102 is much smaller than the concentration of defects @xmath103 , implying that the probability of 2 pairs meeting is insignificant when compared to the probability of a pair meeting a defect .
we next estimate the probability of destroying a pair as in case ( 2 ) . on average
, there are 7 links in the neighborhood of a given pair which increase @xmath7 , and flipping such a link has an energy cost of @xmath104 .
the probability of this to happen is @xmath105 .
since the pair moves every @xmath20 attempted flips , we conclude that , on average , a pair will do @xmath106 steps before it is destroyed as in case ( 2 ) .
the number of steps needed for case ( 1 ) to happen depends obviously on the density of defects .
we let @xmath107 denote the average distance between defects ( counted in number of links ) .
since the number of defects is @xmath108 and the system is 2-dimensional , we conclude that = ( ^-1/2 ) .
as long as the pair is not destroyed by the mechanism leading to case ( 2 ) it can thus do @xmath109 steps by which time it can visit @xmath110 defects .
this terminates the comparison of the 3 cases , and shows that a pair has the time to visit a very large number of defects before it is destroyed by the 2 other mechanisms .
in this section we consider the collisions between a pair and a defect .
the discussion is really in two parts : on one hand , we must consider the probability that a collision between a pair and a defect is initiated .
this depends on the density of the defects , and hence on @xmath48 .
but , once a collision is initiated , we can ask what the effect of the collision is going to be .
the next proposition shows that this effect is purely geometrical and independent of @xmath48 .
[ p : pairdefect ] there are 9 topologically different possibilities @xmath111 for a collision to be initiated . for each of them , there are 2 purely geometrical constants @xmath112 and @xmath113 ( depending on @xmath10 ) which tell us the probability that a collision leads to a move ( @xmath114 ) of a defect ( by 1 or 2 links ) resp . the deletion of the pair ( @xmath115 ) . the remainder of the section deals with the proof of .
we will study in detail how collisions move defects .
first of all , we will define what we mean by a collision . assuming that the density of defects is very small , the only collisions we will consider are those involving @xmath116 defects , namely , the pair which will collide with a defect or .
[ def : collision ] consider some configuration @xmath2 .
three defects @xmath117 of @xmath2 are said to be _ in a collision _ if there are @xmath118 flips ( @xmath119 links of @xmath2 ) that do not increase the energy such that 1 .
the only defects involved in these @xmath119 flips are @xmath120 .
all @xmath116 defects are involved in these @xmath119 flips .
at least one of these @xmath119 flips will move a pair ( the others might be any of the 4 cases which do not increase the energy ) . in this section
, we will describe all possible configurations of a collision and we will show that the probability of any such configuration depends solely on the topology ( and not on the temperature ) .
the third condition of the definition of a collision states that we can always identify a pair ; as a result , the set of all possible configurations of a collision can be obtained by taking a pair and placing either a defect or a defect in any position where it can interact with one of the pair s 2 defects . as seen in the previous section , a defect can interact with any defect if and only if both defects are at distance one .
two defects can interact if and only if they are on opposite corners of 2 adjacent triangles .
the last ingredient is that defects can have a degree of 6 or 8 whereas defects have a degree of 4 or 6 .
this yields a systematic method of constructing all possible configurations of a collision : consider a pair and let @xmath121 be the set of all empty sites ( charge @xmath72 ) which are at distance 1 of any of the pair s 2 defects and @xmath122 be the set of all empty sites which are opposite to the defect of the pair .
the set of all possible configurations of a collision is obtained by placing a defect in any of @xmath121 s sites or a defect in any of @xmath122 s sites , as shown in in the case where the defect is red and the defect is blue .
all in all , we get 9 different configurations of a pair and a defect ( symmetrical case omitted ) . assuming that the defects are randomly distributed , it is clear from that the probability of a collision to be of some type @xmath123 is a temperature independent constant that can be calculated . to prove one must study in detail each of the 9 cases .
we will study in particular : * what are the possible outcomes of each collision type and what is the ( conditional on having initiated the collision @xmath124 ) probability of each outcome ? * what is the probability ( conditional on having initiated the collision @xmath124 ) that a pair pushes a defect ?
we can summarize the answers as follows : * there will always be a defect left over at the end of the collision . * finding a pair and a defect at the end of the collision is possible in all 9 cases .
* an annihilation of the pair is possible in 2 of the 9 cases .
* it is possible that the defect is pushed in 8 cases .
a defect can be pushed by more than 1 step .
* it is possible that the defect remains in its initial position in all 9 cases .
the relative probabilities of any of the above outcomes only depend on the local geometry .
while all the cases have been worked out in detail , we illustrate the discussion for just 2 of them , and this will complete the proof of .
there are 2 cases where an annihilation might occur .
we consider here the case of .
a pair collides with a defect . for simplicity , assume that the pair came from the left .
once the pair and the defect are in collision , there are 3 links whose flipping leads to @xmath125 .
two of these links ( the green ones ) allow the pair to walk away from ( or enter ) the collision . flipping the red link on the other hand causes an annihilation
: the pair is destroyed and the defect is pushed by one step .
we clearly see that there are 3 possible outcomes : * the pair exits the collision through the same way it entered ( in our case , on the left ) .
the defect remains in its initial position .
* the pair exits the collision through the other green link .
the defect moves 2 steps . *
an annihilation event occurs .
the pair disappears and the defect moves 1 step .
the ( conditional ) probability of each outcome is 1/3 and the ( conditional ) probability that the defect will have moved at the end of the collision is 2/3 . here , we look at the collision case of . + + + + no annihilation is possible here and the outcome of the collision is always one pair and one defect . the only relevant question is what is the probability that the defect will have moved at the end of the collision .
but the combinatorics is more involved .
the pair enters and may exit the collision through a green link . flipping a red link on the other hand will not end the collision .
notice that the fifth diagram contains 4 red links and no green ones . moving
a red link will visit the 6 figures sequentially . but moving the two lower red links in the lowest left figure will lead to another circle of five configurations , which is not shown in the figure .
this collision case can be represented by a `` state diagram '' as in , where each node represents a state and each link represents the effect of flipping one of the colored links in .
the pair enters the collision through a dangling link @xmath126 .
it can wander around the vertices of the state diagram before exiting through a dangling link @xmath127 .
if @xmath128 , then it is as if the collision never occurred . in particular , the defect does not move . furthermore ,
if @xmath126 and @xmath127 are of the same color , then the defect will remain in its initial position at the end of the collision . using this remark and the diagrams of , one can explicitly compute the ( conditional ) probability that a pair pushes a defect if the collision is of the above type
. this probability will be temperature independent .
the other 7 cases are treated similarly , and this completes the proof of . note that the proof means that collisions lead , on average to a _ positive _ probability of moving a defect .
_ this mechanism is the basic reason for the diffusive wandering of the defects in the triangulations .
it is mediated by the collision of pairs with the defects . clearly , if there are no pairs , the defects can not move by this mechanism , but only through much less probable events . _
in , we have seen that a pair lives long enough to explore its 1d path , before being destroyed by other mechanisms .
we now analyze in detail what can happen during this exploration phase .
when a pair is created , it is one step away from its birthplace .
it will then perform a random walk on its predefined 1-dimensional path .
each time it comes back to its birthplace , it can die with probability @xmath129 as shown in .
_ if this happens , the triangulation will not have changed .
_ we will call this an ineffective pair .
the probability @xmath130 can be estimated as follows : assume that a defect @xmath131 is at a distance @xmath107 from the birthplace of the pair .
then , by extending slightly the gambler s ruin principle @xcite , _ the probability @xmath132 that the pair actually can reach @xmath131 is @xmath133_. this implies that the probability for any event implying @xmath131 when starting from the birthplace depends on @xmath107 , and in the case of many defects , on their average distance ( which we call again @xmath107 )
. thus , [ e : pi ] p_i = 1-(1/ ) , p_r=(1/ ) .
here , we estimate the rate of change of triangulations ( as a function of time ) . since our triangulations are purely topological , we need to define what we mean by the distance between 2 triangulations @xmath134 and @xmath135 in @xmath4 ( the space of triangulations of the sphere with @xmath0 labeled nodes ) .
there are many possible choices , see , @xcite many of which lead to equivalent metrics .
the one defined below is convenient for our purpose .
let @xmath136 .
consider a node @xmath137 of @xmath134 .
the flower @xmath138 of @xmath137 is defined as the ordered cyclic set of all neighbors of @xmath137 in @xmath134 .
two flowers are then said to be equal if one can be obtained from the other by a cyclic rotation .
we can now define the following metric on @xmath139 : d(t_1,t_2)&=_n=1^nd_n(t_1,t_2 ) + d_n(t_1,t_2)&= [ cols= " < , < " , ] . using this metric , we define the following correlation function : c()=1- where @xmath140 is the system state at time @xmath141 .
our result for the decay of this function at equilibrium , , when @xmath142 , is as follows : [ decay ] the correlation function @xmath143 decays like [ e : cor ] c()=e^-/_r with a relaxation time @xmath144 of the form [ e : taur ] _
r=(e^3)=(^-3 ) the correlation function @xmath145 is nothing but the fraction of nodes whose flower is unchanged after @xmath146 time units . at equilibrium ,
the number of pairs @xmath147 was established in to be @xmath148 . on the other hand ,
the density of defects in equilibrium is @xmath149 and hence , their average distance @xmath107 equals @xmath150 . by the estimates of this
means that the effective number of pairs which change the configuration in a permanent way is @xmath151 .
we further saw in that the number of collisions a relevant pair will undergo is a temperature independent constant @xmath152 .
if @xmath107 is the average distance between 2 defects , then , on average , this pair will change , on its way , the flowers of @xmath153 nodes . at time @xmath146 , each of these flowers is still unchanged with probability @xmath145 .
since the pair makes a 1d random walk , all this happens within an average time interval @xmath154 .
this in turn leads to c(+ ) = c()-2pp_rc ( ) in the limit @xmath155 , we find ( ) = -4c ( ) and this leads to with [ e : taur2 ] _
r = finally , using [ e : s ] p_r ( ) = ( 1/ ) = ( ^1/2 ) follows from equations ( [ e : taur2 ] ) and ( [ e : s ] ) .
by the aging process , we mean the approach of the energy to its equilibrium value .
since the energy is by and large just the density @xmath156 of defects we can formulate the result as under the assumptions @xmath40 and @xmath41 one has for the density @xmath157 of defects : [ e : bound ] d(t ) = ( ( ^2 t ) ^-2/5 ) note that this result differs from that proposed in @xcite , where the decay rate was given as @xmath158 .
this difference is caused by our observation that the diffusion constant of the defects actually depends on their density , because , if they are rarer , the pairs , which are the only ones able to move them around , need longer to find them .
power decay rates are extremely hard to distinguish , but we have performed some tests which are illustrated in .
they give a slight advantage to a decay of @xmath159 as compared to @xmath160 .
we study the aging process by assuming that , in approach to equilibrium , the system is in a quasistationary state , with charge density @xmath161 . here , and in the sequel
, time will be in units of @xmath162 .
let @xmath156 and @xmath163 be the density of defects and pairs respectively .
then , up to terms of order @xmath164 one has @xmath165 .
as we will see in this section , the quasistationarity assumption simply means that the relaxation of the energy is a consequence of the _ annihilation of colliding defects_. the number of pairs is , up to fluctuations , essentially unchanged during the process we consider .
we saw that a fraction @xmath166 of all occurring flips in the system do not change the energy , and are either motions of pairs or blinkers .
of those , the only relevant ones are the wandering pairs , which induce diffusion of the defects as we have seen in .
the discussion of the equilibrium probabilities apply also to states close to equilibrium , which is the regime we want to consider now .
the pair dynamics happens on the time scale @xmath167 and it conserves both the number of pairs @xmath163 and the number of defects @xmath156 . the next slower time scale concerns creation and annihilation of pairs .
even though this changes @xmath163 , it conserves @xmath156 .
whenever one of these events happens , defects are pushed around by the pairs with some geometrically defined probability , and this leads to a diffusion , whose constant @xmath168 measures this second time scale @xmath169 .
the third time scale @xmath170 is related to collision rate @xmath171 of defects ; @xmath172 .
they undergo a 2d random walk .
sooner or later , 2 defects of opposite charges will meet and will form a new pair which will run on timescale @xmath173 until it annihilates . in the regime
we consider , _ only _ this sequence of events ( collision and running pair ) of the dynamics destroys 2 defects and , as a consequence , is responsible for the relaxation of the energy .
given the 3 time scales , the derivation of the decay rate is now rather straightforward . by the previous discussion , _
meeting(t ) _ diffusion(t)_pair(t)==1 .
the orders of magnitude of these quantities near equilibrium are _
meeting(t ) = ( ^-2 d^-7/2 ) , _
diffusion(t)=(^-2d^-1/2 ) , _ pair(t)=1 . consider a system for which , at time 0 , @xmath174 and @xmath175 .
it is clear that the relaxation of pairs is much faster than that of defects .
we will assume that pairs are always at equilibrium density , , that creation and destruction rates of pairs are equal and @xmath163 is independent of @xmath141 .
the above discussion implies that @xmath163 is constant over time intervals of order @xmath176 .
in fact , both creation and annihilation events necessitate the presence of defects so that the creation and destruction rates of pairs will be linear in @xmath156 at low density .
this implies that @xmath147 depends on @xmath141 only through the value of @xmath156 . by abuse of notation
, we will write @xmath177 instead of @xmath178 .
the creation rate of pairs is @xmath179 and the destruction rate is simply @xmath180 .
therefore , by balancing the rates , we find : [ e : p ] p(d)=12d_lifetime^2 since a pair needs to diffuse from one defect to the other in order to annihilate , we estimate that @xmath181 .
this implies that the density of pairs is @xmath182 . repeating the arguments of the average number of collisions @xmath183 and the average number of moved defects @xmath184 are temperature independent constants
the diffusion constant of a defect is simply the probability that a given defect moves by one space unit during one time unit and it is given by dd & = + d(d ) & =( p(d ) p_r((d^1/2 ) ) ) using equations ( [ e : p ] ) and ( [ e : s ] ) , this leads to d(d ) =( ^2 d^1/2 ) the annihilation of 2 diffusive particles @xmath185 has been studied in depth in @xcite . here
, we use the mean field argument of @xcite , to deduce the collision rates .
however , there will also be particle creation .
on the other hand , , in @xcite creation is indeed considered , but the study is for a fixed substrate , namely the lattice @xmath186 , while our study is on a more floppy domain . given a 2d gas of 2 particles @xmath61 and @xmath187 of equal densities @xmath188 such that the diffusion constants @xmath189 , it can be deduced from @xcite that the collision rate @xmath190 is \(d ) = ( d d^3 ) extending this identity to a varying diffusion constant , we end up with = -2(d)=- ( ^2 d^7/2 ) where we assumed that we are far enough from equilibrium to neglect the creation rate of defects . note
that this result differs from that proposed in @xcite , where the decay rate was given as @xmath158 .
this difference is caused by our observation that the diffusion constant of the defects actually depends on their density , because , if they are rarer , the pairs , which are the only ones able to move them around need longer to find them .
this research was partially supported by the fonds national suisse . | in this paper we continue the study of a topological glassy system .
the state space of the model is given by all triangulations of a sphere with @xmath0 nodes , half of which are red and half are blue .
red nodes want to have 5 neighbors while blue ones want 7 .
energies of nodes with other numbers of neighbors are supposed to be positive .
the dynamics is that of flipping the diagonal between two adjacent triangles , with a temperature dependent probability .
we consider the system at very low temperatures .
we concentrate on several new aspects of this model : starting from a detailed description of the stationary state , we conclude that pairs of defects ( nodes with the `` wrong '' degree ) move with very high mobility along 1-dimensional paths . as they wander around , they encounter single defects , which they then move `` sideways '' with a geometrically defined probability .
this induces a diffusive motion of the single defects .
if they meet , they annihilate , lowering the energy of the system .
we both estimate the decay of energy to equilibrium , as well as the correlations .
in particular , we find a decay like @xmath1 .
dedicated to david sherrington , with admiration and best wishes |
the search for the higgs boson continues to be of paramount importance to complete our understanding of the standard model . in the last few years
the cdf and d0 collaborations at the tevatron have ruled out a mass window for the sm higgs boson in the range @xmath0 gev @xcite .
the lhc has now joined the search for the higgs boson , and relying on the @xmath1 mode , the atlas and cms collaborations expect to either discover or exclude a standard model with three generations ( sm3 ) higgs boson in a wide mass range . up to now , for sm3 , a higgs boson with a mass in the ranges @xmath2 and @xmath3 gev has been ruled out at 95% c.l . by atlas @xcite , and in the ranges @xmath4 gev and @xmath5 gev by cms @xcite . with a fourth generation ( sm4 )
, cms has ruled out the higgs boson mass in the range of @xcite @xmath6 gev at the 95% c.l .
the dominant production mechanism for the sm3 higgs boson at the lhc is gluon fusion @xcite through a top - quark loop .
this mechanism is very sensitive to new physics : for example , with a fourth generation ( sm4 ) , the amplitude is roughly 3 times larger as it simply counts the number of heavy flavors in the loop .
this results in a significantly larger cross - section for higgs boson production , about 9 times larger than in sm3 @xcite .
this observation , combined with the higgs search results from the lhc this year and expected from future , places a considerable strain on the possibility of a fourth generation .
if the higgs boson is indeed discovered with a production cross - section roughly in agreement with the sm , one would have to explain why the factor of 9 enhancement present in sm4 is not there . on the other hand , if the higgs boson is not observed , its exclusion is even more significant in sm4 , and its allowed mass is pushed towards the unitarity bound
. a possible way out of this predicament that has been discussed for sm4 is that it may be more natural to have a heavy higgs if indeed there are four generations @xcite . in this note
we consider the possibility of extended scalar sectors that could remove the tension between higgs physics at the lhc and a heavy fourth generation .
we discuss two possibilities : a scalar sector extended with a color octet , electroweak doublet @xcite ; and a variation of the two - higgs doublet model `` for the top '' @xcite . in the first case
we argue that it is possible to suppress the higgs production rate in gluon fusion with suitable additional particles . in the second case
we argue that it is possible to make the neutral scalar with sm - higgs - like couplings heavy , or to suppress its production cross - section relative to sm4 .
similar arguments have been used in ref .
@xcite to constrain a possible fourth generation in two - higgs doublet models of type ii .
another possibility recently discussed occurs in models where the higgs boson has a larger invisible decay width than that in the sm @xcite .
we now consider the case of a scalar sector that has been extended with a color octet , electroweak doublet and hypercharge @xmath7 scalar @xmath8 .
this particular choice is motivated by the requirement of minimal flavor violation and has been recently elaborated in ref .
it was noted in that paper that in minimal flavor violation only scalars with the same gauge quantum numbers as the sm higgs doublet @xmath9 or color octet scalars with the same weak quantum numbers as the higgs doublet @xmath10 can couple to quarks , a scenario with many interesting collider and flavor physics consequences .
this color octet can be written in a conventionally normalized component form with color index @xmath11 as @xmath12 .
since the new scalars carry color , they may have a significant effect on the two gluon coupling to the sm higgs boson @xmath13 , and thus affect its production at lhc .
the contribution of @xmath14 to the @xmath15 coupling can not happen at the tree level , but occurs at the one loop level by having the color octet @xmath14 in the loop and allowing @xmath13 to couple to @xmath14 from the interactions in the scalar potential .
the most general scalar potential with @xmath16 and @xmath14 is ref.@xcite , @xmath17\nonumber\\ & + & \lambda_6 { \rm tr } s^{\dagger i } s_i s^{\dagger j}s_j + \lambda_7 { \rm tr } s^{\dagger i } s_j s^{\dagger j}s_i + \lambda_8 { \rm tr } s^{\dagger i } s_i{\rm tr } s^{\dagger j}s_j \nonumber\\ & + & \lambda_9 { \rm tr } s^{\dagger i } s_j { \rm tr } s^{\dagger j}s_i + \lambda_{10 } { \rm tr } s_i s_j s^{\dagger i}s^{\dagger j } + \lambda_{11 } { \rm tr } s_i s_j s^{\dagger j}s^{\dagger i}\;.\end{aligned}\ ] ] the additional contribution to the higgs boson production due to the octet - scalar loops has been calculated in ref .
@xcite . in the limit of very heavy quarks and color scalars in the loop
, the @xmath15 coupling can be obtained form the effective lagrangian@xmath18 where @xmath19 is the vacuum expectation value ( vev ) of the usual sm higgs doublet @xmath20 .
@xmath21 and @xmath22 are the would - be goldstone bosons `` eaten '' by @xmath23 and @xmath24 , and @xmath13 is the physics higgs field .
@xmath25 is the number of heavy quark flavors , one in the case of sm3 and three in the case of sm4 .
it is clear from the above expression for the @xmath26 coupling that the addition of the color octet scalar doublet has the potential to significantly alter the higgs production cross - section at lhc .
indeed , it was already pointed out in ref
. @xcite that this cross section could double for @xmath27 , which are consistent with electroweak constraints .
it is equally possible to cancel an enhancement from sm4 in the higgs production cross - section , @xmath28 , with appropriate parameters in the color octet scalar potential . for this cancellation to be possible , we need to check that the required values of @xmath29 are not in conflict with anything else .
there is no constraint from the electroweak parameter @xmath30 ( equivalently from custodial @xmath31 ) on the combination @xmath32 , as long as a third parameter in the potential , @xmath33 , takes the value @xmath34 .
the contribution from @xmath35 to the electroweak parameter @xmath14 is given by @xcite @xmath36 the current best fit value for s parameter and its allowed range , for @xmath37 gev is , @xmath38 @xcite . assuming that @xmath14 is saturated by @xmath35 , the implied range for the combination appearing in eq .
[ effhgg ] is @xmath39 this is enough for a large suppression of sm3 at one sigma level . to sufficiently suppress sm4 , an additional negative contribution from @xmath40 , which is not constrained by @xmath14 ,
is required .
for example , with @xmath41 , a @xmath42 would halve the sm4 coupling .
of course , the color octet scalar can also enhance the higgs production cross section , exacerbating any potential conflict .
another simple extension of the scalar sector of the sm consists of adding a second higgs doublet @xcite . many variations on this theme exist in the literature , but one stands out motivated by the fact that the fourth generation quarks are necessarily much heavier than the lightest five quarks
this suggests that a second higgs doublet is responsible for the masses of the fourth generation quarks ( and possibly the top - quark as well ) , in a variation of the `` for the top - quark '' two higgs doublet model of ref .
the phenomenology of these models and their significance for a heavy fourth generation has been recently emphasized in ref .
@xcite . to implement this model as a solution to the sm4 tension with higgs physics at the lhc , we do not focus on reducing the @xmath26 effective coupling .
we consider instead the observability of @xmath13 including its production and its couplings to @xmath43 pairs as well as its mass .
in essence , we require the higgs with relatively large couplings to @xmath43-pairs either to be outside the accessible mass range , or to have a smaller enhancement over sm3 than what is found in sm4 . in a generic two higgs doublet model with scalar fields @xmath44 and @xmath45 each has a vev @xmath46 , @xmath47 . assuming that @xmath44 couples to the first three generations and @xmath45 couples to the fourth generation , the yukawa couplings are given by ( we call this model i ) @xmath48 the model can be modified so that the top - quark also couples to @xmath45 , we call this model ii . after spontaneous symmetry breaking there remain two physical neutral scalars @xmath13 and @xmath16 , a pseudo - scalar @xmath11 , and a charged higgs @xmath21 .
the physical neutral scalars are the ones relevant to our discussion here . in general , the higgs potential parameters mix the neutral real components of the doublets @xmath49 to form the physical neutral scalars @xmath13 and @xmath16 .
we parameterize this mixing by @xmath50 at this point it is worth noting that the parameter space in the higgs potential can accommodate either @xmath13 or @xmath16 as the heavier of the two .
the couplings of @xmath13 and @xmath16 to @xmath43 and @xmath24 bosons are given by @xmath51 where @xmath52 is defined by @xmath53 and @xmath54 , and @xmath55 is presumably large as @xmath47 gives mass to the heavy fourth generation .
the yukawa couplings of @xmath13 and @xmath16 are given by ( @xmath56 ) @xmath57 the lhc search strategy for the higgs masses in the ( 120 - 600 ) gev range mentioned before , corresponds to the process @xmath58 followed by @xmath59 or @xmath60 @xcite . with the couplings of model i described above ,
the cross - sections for the overall processes , as compared to the sm3 higgs , become ( in the infinite heavy quark mass limit ) are the same in the 2hdm type ii , although the origin of the separate terms is different : the term with the factor of 2 arising from two up - type quarks @xmath61 and the second term arising from the @xmath62 quark . ]
@xmath63 ^ 2 \nonumber \\
\frac{\sigma_h}{\sigma_{sm3 } } \equiv \frac{\sigma(pp\to h \to vv)}{\sigma(pp\to
h_{sm } \to vv ) } & \sim & \left[\left(2\frac{\cos\alpha}{\sin\beta}-\frac{\sin\alpha}{\cos\beta}\right)\sin(\beta-\alpha)\right]^2 \label{ratios}\end{aligned}\ ] ] where @xmath64 . if we choose to couple the top - quark to @xmath45 instead of @xmath44 as in model ii , these ratios become @xmath65 ^ 2 \nonumber \\ \frac{\sigma_h}{\sigma_{sm3 } } \equiv \frac{\sigma(pp\to h \to vv)}{\sigma(pp\to h_{sm } \to vv ) } & \sim & \left[\left(3\frac{\cos\alpha}{\sin\beta}\right)\sin(\beta-\alpha)\right]^2 . \label{ratios2}\end{aligned}\ ] ] in figure [ fig : ratios ] we plot the ratios in eq .
[ ratios ] ( model i ) and eq .
[ ratios2 ] ( model ii ) for a region of parameter space .
we use @xmath66 for model i and @xmath67 for model ii which make the @xmath13 coupling to @xmath43 pairs very similar to the sm higgs coupling to @xmath43 pairs .
we plot ranges for @xmath55 which are similar to those considered in ref .
a full analysis of the parameter space of these models is beyond the scope of this note , but our figure illustrates the salient features that can alleviate the tension with sm4 .
( left plot ) and for model ii in eq .
[ ratios2 ] ( right plot ) as a function of @xmath55 .
the dashed lines mark the points @xmath68 , where the two ratios are equal ; and @xmath69 , where the cross - section is half the sm3 value .
, title="fig : " ] ( left plot ) and for model ii in eq .
[ ratios2 ] ( right plot ) as a function of @xmath55 .
the dashed lines mark the points @xmath68 , where the two ratios are equal ; and @xmath69 , where the cross - section is half the sm3 value .
, title="fig : " ] one possibility is to make the sm - like higgs boson heavy .
the main difference between this scenario and a heavy higgs in sm4 is that the second , lighter , higgs increases the mass bound allowed by unitarity for the heavy one @xcite .
for example , consider a case in which the @xmath13 is the sm - like higgs boson such that its coupling to @xmath43-pairs is close to that in the sm . in this limit ,
the @xmath13 couplings to the fermions are also the same as those for the sm4 and @xmath70 .
it follows that if @xmath13 is also the lightest neutral scalar eigenstate , this model is identical to and has the same tension as sm4
. however , in this 2hdm , the mass of @xmath13 can be higher than the mass of the higgs boson of the sm4 .
in particular , generic unitarity bounds for two - higgs doublet models suggest it can be as heavy as @xmath71 gev @xcite and outside the range of current searches . in this scenario , @xmath16 is the lighter higgs boson and can be produced at the lhc , but the search for this object requires a different strategy as it does not couple to @xmath43-pairs , @xmath72 . in figure [ fig : ratios ] we illustrate this case for model i , eq . [ ratios ] , on the left plot .
the figure shows that there is a value of @xmath55 where @xmath73 is the same for both higgs bosons , and only 2.25 times larger than it is in sm3 .
as we increase @xmath55 from that point , the ratio @xmath74 increases until it reaches the value of 9 as in sm4 . at the same time the ratio @xmath75 is decreasing being less than half the sm3 value for most of the range .
this illustrates a second way to alleviate the tension , in which the sm - like higgs boson is still enhanced relative to sm3 but by a smaller factor than it is in sm4 . in the right plot of figure
[ fig : ratios ] we show what happens in model ii ( where the top - quark also couples to @xmath45 ) for @xmath67 . notice that it is not possible to find values of @xmath76 and @xmath52 that simultaneously suppress @xmath75 and @xmath74 in eq .
[ ratios2 ] below one .
however , the enhancement over sm3 can be made as small as 2.25 simultaneously for both neutral higgs bosons .
the figure also shows that one can suppress the lighter neutral higgs below the sm3 value while enhancing the heavier one by factors around 5 , which is still below sm4 . in summary , in a generic two - higgs doublet model ,
the process @xmath77 will not be able to conclusively rule out a heavy fourth generation unless the search is extended to masses reaching the unitarity bound and the sensitivity to the cross - section reaches values that are close to twice the sm3 cross - section .
the search for the higgs boson at the lhc during the upcoming year will be at odds with the predictions of a heavy fourth generation if the higgs boson is found with a cross - section consistent with sm3 or if the higgs boson is not found at all . in this note
we have examined the possibility of ameliorating this conflict with extended scalar sectors .
there are two possible paths that we illustrate with two examples . in the first example
, an additional color - octet electroweak - doublet can suppress the higgs production cross - section in both sm3 and sm4 .
the details , of course , depend on the parameters in the scalar potential but we have seen that the numbers needed are allowed by other constraints . in the second example we have shown that in a two - higgs doublet model there is sufficient freedom to suppress the higgs cross - section for the lighter neutral higgs , while at the same time the unitarity bounds allow the mass of the heavier higgs boson to be larger than in sm4 , possibly outside the current lhc search range . this work was supported in part by doe under contract number de - fg02 - 01er41155 .
thanks the physics department at the national taiwan university , taipei , taiwan for their hospitality while this work was completed .
xgh was partially supported by nsc and ncts od roc , and sjtu 985 grant of prc .
t. aaltonen _ et al . _
[ the cdf collaboration ] , phys .
lett . * 104 * , 061803 ( 2010 ) .
[ arxiv:1001.4468 [ hep - ex ] ] ; v. m. abazov _ et al . _ [ the d0 collaboration ] , phys .
lett . *
104 * , 061804 ( 2010 ) .
[ arxiv:1001.4481 [ hep - ex ] ] ; t. aaltonen _ et al . _ [ cdf and d0 collaboration ] , arxiv:1103.3233 [ hep - ex ] .
r. casalbuoni , d. dominici , f. feruglio , r. gatto , nucl .
* b299 * , 117 ( 1988 ) ; s. kanemura , t. kubota , e. takasugi , phys .
* b313 * , 155 - 160 ( 1993 ) .
[ hep - ph/9303263 ] ; a. g. akeroyd , a. arhrib , e. -m .
naimi , phys .
lett . * b490 * , 119 - 124 ( 2000 ) .
[ hep - ph/0006035 ] ; j. horejsi , m. kladiva , eur .
j. * c46 * , 81 - 91 ( 2006 ) .
[ hep - ph/0510154 ] . | it is expected that the lhc will soon discover the higgs boson , or that failure to find it will severely constrain its production cross - section over a large mass range .
either one of these results spells trouble for a fourth generation that significantly enhances the higgs production cross - section at lhc .
in fact the lhc has already ruled out a sm higgs mass in the range of 120 gev to 600 gev with a fourth generation at the 95% c.l . in this paper
we explore options within extended scalar sectors to maintain the viability of a heavy fourth generation if an enhanced ( relative to the standard model ) higgs production cross - section is not observed . |
in weakly chaotic hamiltonian systems regions of regular ( periodic and quasi - periodic ) and chaotic motion typically coexist in the phase - space @xcite . in high dimensions , due to arnold diffusion
, all initial conditions leading to chaotic motion are connected in the phase - space building a single chaotic component @xcite . even if the volume of the regular regions becomes vanishingly small , as expected for high - dimensional nonlinear systems , the dynamics inside the chaotic component of the phase - space
is strongly affected by such regions .
this happens because trajectories approaching non - hyperbolic regions or regular motion remain a long time close to them before visiting again other parts of the chaotic component of the phase - space .
this signature of weak mixing ( or weak chaos ) is known as stickiness @xcite . since chirikov - shepelyansky @xcite
, the main quantification of stickiness in hamiltonian systems has been through the fat - tail distribution of poincar recurrence times ( see , _ e.g. _ , @xcite ) .
an alternative approach is to use finite - time lyapunov exponents ( ftles ) @xcite , with recent applications using large deviation techniques @xcite and the cumulants @xcite of the ftle distribution . in area - preserving maps ,
stickiness generically occurs at the border of @xmath0-dimensional kolmogorov - arnold - moser ( kam ) island @xcite ( _ i.e. _ , at @xmath1-dimensional tori ) .
the recurrence time is a measure of the time the trajectory spends around such structures before returning to the chaotic sea ( stickiness happens also to one - parameter families of parabolic orbits @xcite and even to isolated parabolic fixed points @xcite ) . near the non - hyperbolic structures , the local instability of chaotic trajectories
is reduced so that ftles can be used to characterize phase - space regions of interest @xcite .
stickiness has been studied also in higher - dimensional systems @xcite , long recurrence times can be due to different non - hyperbolic regions and tori of different dimensionalities @xcite .
an improved characterization of stickiness events ( long recurrence time ) requires thus to measure the number of stable and unstable directions in the trajectory during this event .
froeschl conjectured that lower - dimensional tori could not exist @xcite .
in early studies in the 80 s such events of stickiness to lower dimensional tori were reported in some systems @xcite but were not found in other examples @xcite .
even if invariant tori do not exist , small local lyapunov exponent could effectively act as a lower - dimensional trap .
this is similar to almost invariant sets @xcite , which are regions in phase - space where typical trajectories stay ( on average ) for long periods of time . in this paper
we introduce a methodology that uses time - series of local lyapunov exponents to define regimes of ordered , semi - ordered and totally chaotic motion and obtain an improved characterization of stickiness in high - dimensional hamiltonian systems .
we illustrate this general procedure in a chain of coupled standard maps and confirm that stickiness events of different times length are dominated by trajectories with different ftles . a significant improvement of the characterization of sticky motion in high - dimensional systems is found .
we also characterize the ftles for small couplings and compare them to expected universal properties in fully chaotic systems @xcite .
the method proposed here is general and can be used to investigate hamiltonian systems in any dimension .
the paper is divided as follows . in sec .
[ mod ] we describe the hamiltonian model we use to illustrate our method . in sec .
[ methods ] we introduce our method to compute and analyze time series of local lyapunov exponents .
this methodology is then applied to the symplectic model of coupled standard maps in sec .
[ nr ] . section [ cc ] summarizes the main results of the paper .
we use a time - discrete @xmath2-dimensional hamiltonian system obtained as the composition @xmath3 of independent one - step iteration of @xmath4 symplectic @xmath0-dimensional maps @xmath5 and a symplectic coupling @xmath6 . as a representative example of @xmath0-dimensional maps we choose for our numerical investigation the standard map : @xmath7 and for the coupling @xmath8 \\ x_i \\
\end{array } \right),\ ] ] with @xmath9 ( all - to - all coupling ) .
the motivation for working with this system is that in the limit of small coupling @xmath10 it can be understood looking at the dynamics of the @xmath4 uncoupled maps .
this system was studied in refs .
@xcite using recurrence time distribution .
this allow us to critically compare the benefits of our methodology .
in all numerical simulations we used @xmath11 for the map @xmath12 and @xmath13 for the maps @xmath14 and @xmath15 .
in this section we describe the method proposed in this work . to be illustrative , we present numerical simulations for the system defined in sec .
[ mod ] . consider a chaotic trajectory in a closed hamiltonian system which , after reducing the phase - space dimension due to global invariant of motion , has @xmath4 degree of freedoms . for long times
@xmath16 the trajectory ergodically fills the whole chaotic component of the phase - space which is characterized by a spectrum of @xmath4 lyapunov exponents @xmath17 , where @xmath18 .
largest lyapunov exponents because due to the symplectic character of hamiltonian systems the others @xmath4 exponents are simply @xmath19 . ]
the central ingredient of our analysis is the spectrum of ftles computed along a trajectory during a window of size @xmath20 where we obtain a time dependent spectrum @xmath21 .
the window size @xmath20 has to be sufficiently small to guarantee a good resolution of the temporal variation of the @xmath22 s , but sufficiently large in order to have a reliable estimation ( see refs .
the probability density function of @xmath23 has been extensively studied @xcite . here
we go beyond the study of the probability density function and explore temporal properties in the time series of @xmath24 .
figures [ lyapsloct](a ) and ( b ) , for @xmath25 and @xmath26 respectively , show the time series of @xmath22 , ( @xmath27 ) .
the sharp transitions towards @xmath28 motivates the classification in regimes of motion @xcite as ( a ) _ ordered _ ( @xmath29 ) ; ( b ) _ semi - ordered or semi - chaotic _ ( @xmath30 ) ; and ( c ) _ chaotic _ ( @xmath31 ) . for a system with @xmath4 degrees of freedom
we will say that the trajectory is in a regime of type @xmath32 if it has @xmath33 local lyapunov exponents @xmath34 , where @xmath35 are the small thresholds .
this means that @xmath36 and @xmath37 are ordered and chaotic regimes respectively .
whenever there is no ambiguity , we will drop the superscript @xmath38 to have a simpler notation .
practical implementations of the general method described above require the choice of a few parameters and conventions .
first of all , the window size @xmath20 and the threshold @xmath39 directly affect the classification in regimes . they can be thought as the phase - space resolution of the analysis and should be chosen so that it provides maximal information about the regions of interest . unless stated otherwise , we use @xmath40 and @xmath41 , where @xmath42 denotes average over @xmath16 , where @xmath43 ( even though the _ classification _ in regimes is strongly @xmath20-dependent , our _ conclusions _ are not sensitively affected by variations around the chosen values ) . another important choice is the method for computation of the ftles .
we use benettin s algorithm @xcite , which includes the gram schmidt re - orthonormalization procedure .
the decreasing order of @xmath23 is valid on average , but inversions of the order ( @xmath44 ) may happen for some times @xmath16 and we have chosen to impose the order of @xmath23 for all @xmath16 .
finally , it is possible to decide how to sample the time series @xmath23 . while the ftles are defined for all @xmath16 , there is a trivial correlation between the values of ftles inside a window of size @xmath20 because they are computed using the same points of the trajectory . in order to avoid this trivial correlation
the series of @xmath23 can be computed using non - overlapping windows , _
i.e. _ plotting @xmath23 only every @xmath20 time steps ( a choice we adopt in our simulations ) . in order to understand the properties of the time series @xmath45 it is useful to consider the phase - space regions associated to each regime @xmath46 .
we denote by @xmath47 the phase - space volume ( liouville measure ) of region @xmath48 in the bounded phase - space @xmath49 , _ i.e. _ @xmath50 .
the most important distinction is between the regions of regular @xmath51 and chaotic @xmath52 motion . in hamiltonian systems , typically @xmath53 and @xmath54 . in principle
, the regular region @xmath55 can be subdivided according to the dimensionality of the tori . however , according to froeschl s conjecture , in a @xmath2-dimensional phase - space , tori with dimension @xmath4 have positive measure and thus @xmath56 @xcite . for @xmath57 , the chaotic region @xmath52
is expected to build a single ergodic component because tori of @xmath4 dimension do not partition the @xmath2-dimensional phase - space in different regions and therefore any chaotic trajectories eventually explores ( through arnold diffusion ) the whole @xmath52 .
our interest is not to test the froeschl conjecture or arnold diffusion , but to show the insights about the chaotic dynamics we can obtain using the time series of @xmath45 together with the definition of the regimes @xmath46 .
one application is to use the regimes @xmath46 to split the chaotic component of the phase - space in meaningful components .
this is done by considering the set of points @xmath58 in the phase - space leading to each regime @xmath46 as @xmath59 where @xmath60 is the total length of the trajectory and @xmath61 indicates that at time @xmath16 the trajectory at @xmath62 had @xmath63 .
figure [ ps ] shows numerical estimates of the phase - space regions obtained for each regime @xmath46 in the chain of coupled maps defined in sec .
[ mod ] . the regime @xmath64 ( or the ordered regime )
is associated to region localized close to the border of the kam island of the uncoupled case ( compare to fig .
points which belong to the regime @xmath65 are closer to the center of the torus from the uncoupled case .
this suggests that when trajectories are inside the region related to regime @xmath65 , they more likely penetrate inside the torus from the uncoupled case . in the chaotic sea
both regimes @xmath65 and @xmath66 are visible .
these results are naturally understood in the perturbative limit ( small coupling @xmath67 ) .
the regime @xmath64 corresponds to @xmath68 for every @xmath27 , which is expected when the trajectory is stuck close to the @xmath4-dimensional tori built as the product of the @xmath1-dimensional tori of the uncoupled maps . in contrast , @xmath46 for @xmath69 implies that at least one ftle @xmath70 and therefore the trajectory projected in one map can be both in the chaotic and regular regions ( _ e.g. _ , @xmath65 for @xmath25 can be obtained from @xmath71 or from @xmath72 ) .
altogether , these observations confirm that our method allows for a meaningful division of the chaotic component of the phase - space and can thus be used to identify regions of interesting dynamics . in the case
partial barriers exist inside the chaotic component such as in area - preserving maps with mixed phase space @xcite we expect the regions obtained through our method to depend weakly on @xmath20 and to coincide with those obtained from the partial barriers .
in this section we apply the lyapunov time - series methodology described in sec .
[ methods ] to the @xmath2-dimensional system defined in sec . [ mod ] .
we compute and interpret four basic properties of the method : the total time spent in each regime ( residence time ) , the transition between regimes , the consecutive time in each regime , and the scaling of lyapunov exponents .
the first and most basic quantity we measure is the probability @xmath73 of finding the trajectory in each regime , defined as the fraction of the total time @xmath60 that @xmath74 ( _ i.e. _ @xmath75 , where @xmath76 if @xmath77 and @xmath78 otherwise ) . .
( a ) @xmath25 with @xmath79 and @xmath80 .
( b ) @xmath81 with @xmath79 , @xmath82 and @xmath83 . in ( a ) the values obtained with @xmath40 are compared ( gray curves ) with results for @xmath84 and @xmath85 . only for the case
@xmath86 the gray curves ( right for @xmath84 and left for @xmath85 ) show a shift in the @xmath87-axis @xmath88 .
estimations for each @xmath88 are based on a trajectory with length @xmath89 . ]
figure [ commot ] shows the probabilities @xmath73 for the map with @xmath90 as a function of the coupling strength @xmath88 .
we now explain the behavior of @xmath73 with @xmath88 by discussing the effect of coupling @xmath88 on the phase - space regions associated to @xmath46 , as defined in sec .
[ ph ] . by the ergodicity of @xmath52
, @xmath73 corresponds to the ( normalized ) volume of the region related to regime @xmath46 in the phase - space @xmath91 the results of fig .
[ commot ] show that the chaotic region is the largest region in phase - space for any coupling , while the region associated to @xmath92 has a larger volume than @xmath93 for couplings @xmath94 . for larger @xmath88
we see oscillations with a local maximum close to @xmath95 for the cases @xmath96 and @xmath86 .
we now interpret the @xmath88 dependence observed in fig .
[ commot ] by arguing how the different terms in eq .
( [ eq.smu ] ) vary with @xmath88 .
we denote by @xmath97 the measure of tori for the @xmath98-th map with control parameter @xmath99 in the uncoupled case @xmath100 ( which we assume to be approximately equal to the measure of the kam islands ) . for small coupling @xmath101
we expect that most tori of the uncoupled maps to survive and therefore : * @xmath102 , which in the simple case of @xmath103 for all @xmath98 reduces to @xmath104 . * @xmath105 corresponds to a small volume around @xmath106 , _ i.e. _ @xmath107 .
* for @xmath108 we have that @xmath109 maps are in their corresponding kam island ( with probability @xmath97 ) and @xmath33 maps in the chaotic area ( with probability @xmath110 ) .
for example , for @xmath81 and @xmath111 we have that @xmath112[1-\mu(u_3)]\cr & + & \mu(u_2)[1-\mu(u_1)][1-\mu(u_3)]\cr & + & \mu(u_3)[1-\mu(u_1)][1-\mu(u_2 ) ] .
\nonumber\end{aligned}\ ] ] in general this leads to @xmath113 where the last product is over all @xmath114 except @xmath115 . in the simple case of @xmath103
, it reduces to @xmath116 we now consider the effect of growing @xmath88 . in the spirit of the kam theorem ,
the tori of the coupled maps ( generated as the product of the @xmath4 maps ) are expected to be robust to small couplings @xmath88 , which act as a perturbation .
this explains why the curves in fig .
[ commot ] are essentially flat for small @xmath88 . increasing @xmath88
even further , the nonlinearity of the system increases and therefore @xmath117 is expected to decrease ( @xmath118 for @xmath119 ) .
this reduction of the tori leads to an increase in the denominator of eq .
( [ eq.smu ] ) and explains the observed tendency of reduction of @xmath73 for all regions related to stickiness ( @xmath120 ) .
indeed , for @xmath121 no signature of tori or stickiness was detected numerically and @xmath122 .
the nontrivial dependencies of @xmath123 in fig .
[ commot ] appear at @xmath95 values , close to the values of @xmath88 for which the last tori disappear ( see also fig .
@xmath124 in ref .
@xcite ) . in this regime
the volume of the tori is already negligible @xmath125 but stickiness is still effective ( notice that even zero measure non - hyperbolic sets can lead to stickiness @xcite ) .
the denominator in eq .
( [ eq.smu ] ) is therefore @xmath126 , not significantly affected by further increases of @xmath88 , and therefore not driving the reduction of @xmath123 .
small variation of a control parameter of the system ( in this case @xmath88 ) are known to lead to sensitive creation and destruction of tori , with non - trivial dependencies on the stickiness @xcite .
we can thus expect that
close to the disappearance of the tori the small volume of stickiness regions @xmath127 to fluctuate with @xmath88 leading even to an increase with @xmath88 .
it is interesting to note that this non - trivial increase with @xmath88 appears for @xmath128 in fig .
[ commot ] precisely when the curves @xmath129 show a sharp decreasing fluctuation .
this suggests an exchange between measure of different sticky regions associated to regimes @xmath130 , without interference of the much larger fully chaotic component @xmath131 . as a function of the coupling strength @xmath88 .
( a ) transition probability @xmath132 ; ( b - d ) conditional probability @xmath133 defined in eq .
( [ nullmodel ] ) of moving to @xmath134 given that the trajectory was at @xmath33 .
estimations for each @xmath88 based on a single trajectory with length @xmath89 in the case of @xmath25 coupled maps and @xmath79 and @xmath80 . ]
we now focus on the transition between regimes .
the simplest analysis correspond to the two - time ( joint ) probability @xmath132 , computed as the fraction of the total trajectory time @xmath60 that @xmath135 and @xmath136 . the probabilities considered in the previous section can be obtained as @xmath137 and @xmath138 .
figure [ prob1](a ) shows the dependence of @xmath132 on @xmath88 for our model .
we notice that @xmath139 is equal to @xmath140 .
this is expected considering that the system is ergodic , volume preserving , and time - reversible .
the dependence of @xmath132 on @xmath88 follows a similar pattern observed for @xmath73 in fig .
[ commot ] .
more information is obtained from the conditional probability @xmath141 which quantifies the probability that trajectories at @xmath46 will move to @xmath142 .
the results shown in fig . [ prob1](b - d ) show for all @xmath46 that ( i ) persistence in the same @xmath46 ( @xmath143 ) is dominant and ( ii ) the most likely transitions occur between neighboring regimes ( _ e.g. _ , @xmath144 ) .
the only ( slight ) deviations of this picture happen for large values of @xmath88 , close to the disappearance of the kam island .
altogether , these results confirm that in the perturbative regime ( @xmath145 ) stickiness happens approaching the region of regular motion of different maps one after the other ( in opposite to a direct approach from @xmath64 to @xmath131 ) .
the results of the previous section confirm that residence in the same regime is the dominant behavior .
this motivates us to study the time @xmath146 spent consecutively in a regime @xmath46 ( _ i.e. _ , @xmath146 is the time between two consecutive transitions between different regimes , the first to @xmath46 and the second out of @xmath46 ) . in a trajectory of length
we collect a series of @xmath146 .
we are mainly interested in the probability distribution @xmath147 ( or , equivalently its cumulative @xmath148 ) for different @xmath46 s in the limit @xmath149 .
these distributions should be compared to the distribution of recurrence times @xmath150 , defined as the time between two successive entries to a pre - defined recurrence region ( usually taken in the fully chaotic component of the phase - space ) .
events in the tails of @xmath151 are associated to times for which the trajectory is stuck to the non - hyperbolic components of the phase - space and @xmath151 is the traditional method to quantify stickiness in hamiltonian systems @xcite . of times
@xmath146 is shown for each regime @xmath130 for @xmath40 and ( a ) @xmath25 and ( b ) @xmath81 . in ( a )
the gray curves show results for @xmath84 and @xmath85 . only for the case
@xmath86 the gray curves ( left for @xmath84 and right for @xmath85 ) show a shift in the @xmath87-axis @xmath88 .
the cumulative distribution @xmath152 for recurrence times @xmath150 to a region in the chaotic component of the phase - space ( in @xmath131 ) for ( c ) @xmath25 and ( d ) @xmath81 . for comparison , in panels ( c )
we show the results obtained combining the normalized curves for @xmath153 ( blue dotted line : divided by @xmath154 for convenience of scale ) of panel ( a ) , and in ( d ) the normalized curves for @xmath155 ( blue dotted line : divided by @xmath156 ) of panel ( b ) .
results obtained using maps of sec . [ mod ] with @xmath157 , @xmath79 and @xmath80 for the case @xmath25 and @xmath79 , @xmath158 and @xmath83 for the case @xmath81.,title="fig : " ] of times @xmath146 is shown for each regime @xmath130 for @xmath40 and ( a ) @xmath25 and ( b ) @xmath81 . in ( a )
the gray curves show results for @xmath84 and @xmath85 . only for the case
@xmath86 the gray curves ( left for @xmath84 and right for @xmath85 ) show a shift in the @xmath87-axis @xmath88 . the cumulative distribution @xmath152 for recurrence times @xmath150 to a region in the chaotic component of the phase - space ( in @xmath131 ) for ( c ) @xmath25 and ( d ) @xmath81 .
for comparison , in panels ( c ) we show the results obtained combining the normalized curves for @xmath153 ( blue dotted line : divided by @xmath154 for convenience of scale ) of panel ( a ) , and in ( d ) the normalized curves for @xmath155 ( blue dotted line : divided by @xmath156 ) of panel ( b ) .
results obtained using maps of sec .
[ mod ] with @xmath157 , @xmath79 and @xmath80 for the case @xmath25 and @xmath79 , @xmath158 and @xmath83 for the case @xmath81.,title="fig : " ] the numerical simulations in fig .
[ pcum ] confirm that the distribution obtained summing @xmath159 for ordered and semi - ordered regimes ( or @xmath130 ) is equivalent to cumulative distribution @xmath160 obtained using recurrences .
this is in agreement with the association of long consecutive times in regimes of ordered and semi - ordered motion to long recurrence times .
looking at the individual distributions @xmath159 provide valuable additional information on the sticky motion . for semi - ordered motion ( when @xmath161 ) we observe an exponential tail after an intermediate decay with scaling @xmath162 .
this behavior confirms the interpretation given in ref .
more interestingly , the @xmath96 case shows an asymptotic algebraic decay which characterize stickiness .
while the scaling is compatible with the results obtained using recurrence time , @xmath163 obtained in our methodology provides a better characterization of the scaling ( over several orders of magnitudes ) and allows for an independent analysis of the different regimes .
these properties are essential when dealing with high - dimensional systems ( which may contain different pre - asymptotic regimes ) and for an accurate estimation of the stickiness exponent @xmath164 . finally , panel ( a ) in fig .
[ pcum ] shows that all decays discussed above remain ( qualitatively ) the same for different choices of @xmath20 , with the curve for @xmath86 showing the largest sensitivity on @xmath20 ( as in fig .
[ commot](a ) .
so far we have focused at the temporal properties of the time series of ftles @xmath23 and how they change with the coupling strength @xmath88 .
we now consider how the value of the lyapunovs respond to an external perturbation , which in our case is the coupling to the other maps .
it is known that the largest exponent @xmath165 is extremely sensitive to perturbation .
more specifically , daido s relation @xcite states that for small couplings @xmath88 to another chaotic system , a universal logarithmic singularity is observed , @xmath166 where @xmath167 are the unperturbed lyapunov exponents and @xmath168 is a constant and @xmath27 .
this relation is valid for totally chaotic systems and a small mismatch between lyapunov exponents of the uncoupled systems compared to their fluctuations @xcite .
here we investigate the relation @xmath169 as a function of @xmath88 , for distinct values of @xmath20 and different regimes @xmath46 . to this end
we compute the temporal averages of the ftles @xmath170 for times @xmath16 such that @xmath171 . ) and uncoupled ( @xmath172 ) maps as a function of @xmath173 , where @xmath88 is the coupling strength .
results are shown for @xmath25 ( @xmath174 ) and different time windows @xmath20 ( a)-(b ) @xmath175 , ( c)-(d ) @xmath176 , and ( e)-(f ) @xmath177 .
black dashed lines in ( c)-(f ) are the expected linear behaviour , consistent with eq .
( [ daido ] ) .
panels in the left column ( ace ) were computed for the full time series , while on the right column ( bdf ) only ftles in the regime @xmath66 were used .
the different colors correspond to different choices of threshold imposed to define the ftle : @xmath178 uses @xmath179 while @xmath180 uses @xmath181 , where @xmath182 is computed over the full time series ( left column ) . ]
our numerical simulations reported in fig . [ scal ] show that small values of @xmath20 lead to a situation in which @xmath183 at a finite value of @xmath88 ( figs .
[ scal](b ) and ( d ) ) , while larger values of @xmath20 lead to situations in which @xmath184 for any @xmath88 .
these results depend crucially on our choice to impose the order of @xmath23 for all @xmath16 , as discussed in sec .
[ regions ] .
this makes the average over the trajectory time @xmath185 to be @xmath20-dependent and different from the average over the lyapunov time @xmath186 . applying the analysis without the division in regimes @xmath46 leads to strongly fluctuating results ( figs .
[ scal](a , c , e ) ) . much smoother results ( figs .
[ scal](b , d , f ) ) are obtained when we apply our method and compute @xmath187 only for @xmath16 in the fully chaotic regime @xmath188 .
looking at these smoother results we observe that the difference in lyapunovs scales as @xmath189 , but that even for @xmath190 the sticky motion leads to a deviation from daido s relation ( [ daido ] ) ( curves are shifted vertically ) .
in summary , we have proposed a method to characterize the dynamics of hamiltonian systems with mixed phase - space based on time series of finite - time lyapunov exponents . using this method it is possible to define and study with high accuracy the time evolution of regimes of ordered , semi - ordered , and totally chaotic motion .
this allows for an individualized characterization of the different stickiness mechanisms , improving alternative methods based on the statistics of recurrence times or on the distribution of finite - time lyapunov exponents .
we applied our method to a chain of coupled standard maps and showed how the frequency of different regimes and the transition probabilities between them are related to the volume of different phase - space regions . using the consecutive time in distinct regimes we have reproduced previous results obtained using recurrence times and showed that our method allows for a significant improvement in the characterization of the sticky motion ( _ e.g. _ , in the determination of the scaling exponents ) .
this indicates that our method can be used to characterize stickiness in general high - dimensional systems and is particularly suited for cases in which different regions of sticky motion coexist .
we have also shown that the dependence on the coupling strength of the largest lyapunov exponents , after conveniently using our procedure , tend to follow only the qualitative universal properties of fully chaotic system .
results obtained in a simple chain of standard maps confirm that our methodology can be applied to high - dimensional systems and problems of current interest , such as controlling fermi acceleration @xcite , galactic models @xcite , and plasma physics @xcite .
another example of application is to associate each regime @xmath46 with effective hamiltonian functions , a procedure used to reproduce the complicated dynamics of kicking electrons @xcite or the high harmonic generation in laser - assisted collisions @xcite .
cm and rms thank cnpq , capes and fapesc and mwb thanks cnpq for financial support and mpipks in the framework of the advanced study group on optical rare events .
cm also thanks eduardo g. altmann for the financial support and hospitality at the mpipks .
ega thanks d. paz for suggesting the analysis performed in sec . | we investigate chaos in mixed - phase - space hamiltonian systems using time series of the finite - time lyapunov exponents .
the methodology we propose uses the number of lyapunov exponents close to zero to define regimes of ordered ( stickiness ) , semi - ordered ( or semi - chaotic ) , and strongly chaotic motion .
the dynamics is then investigated looking at the consecutive time spent in each regime , the transition between different regimes , and the regions in the phase - space associated to them . applying our methodology to a chain of coupled standard maps
we obtain : ( i ) that it allows for an improved numerical characterization of stickiness in high - dimensional hamiltonian systems , when compared to the previous analyses based on the distribution of recurrence times ; ( ii ) that the transition probabilities between different regimes are determined by the phase - space volume associated to the corresponding regions ; ( iii ) the dependence of the lyapunov exponents with the coupling strength . |
the instability of the stably stratified shear flow is one of main problems in fluid dynamics , astrophysical fluid dynamics , oceanography , meteorology , etc .
although both pure shear instability without stratification and statical stratification instability without shear have been well studied , the instability of the stably stratified shear flow is still mystery . on the one hand ,
the shear instability is known as the instability of vorticity maximum , after a long way of investigations @xcite .
it is recognized that the resonant waves with special velocity of the concentrated vortex interact with flow for the shear instability @xcite .
other velocity profiles are stable in homogeneous fluid without stratification . on the other hand , @xcite proved out that buoyancy is a stabilizing effect in the statical case .
thus , it is naturally believed that the stable stratification do favor the stability ( see , e.g. * ? ? ?
* ; * ? ? ?
* ) , which finally results in the well known miles - howard theorem @xcite . according to this theorem , the flow is stable to perturbations when the richardson number @xmath8 ( ratio of stratification to shear ) exceeds a critical value @xmath9 everywhere . in three - dimensional stratified flow , the corresponding criterion is @xmath10 @xcite
however , the stabilization effect of buoyancy is a illusion . in a less known paper ,
@xcite had shown with several special examples that stratification effects can be destabilizing due to the vorticity generated by non - homogeneity , and the instability depends on the details of the velocity and density profiles .
one instability is called as holmboe instability @xcite .
then @xcite stated three main points from the examples without any further proof .
( a ) stratification may shift the band of unstable wave numbers so that some which are stable at homogeneous cases become unstable .
( b ) conditions ensuring stability in homogeneous flow ( such as the absence of a vorticity maximum ) do not necessarily carry over to the stratified case , so that static stability can destabilize .
( c ) new physical mechanisms brought in by the stratification may lead to instability in the form of a pair of growing and propagating waves where in the homogeneous case one had a stationary wave .
recall the points by @xcite , and that there is a big gap between rayleigh s criterion and miles - howard criterion , @xcite even wrote
miles criterion for stability is not the nature generalization of rayleigh s well - known sufficient condition for the stability of a homogeneous fluid in shear flow " .
the mystery of the instability is still cover for us .
following the frame work of @xcite , this study is an attempt to clear the confusion in theories .
we find that the flow instability is due to the competition of the kinetic energy with the potential energy , which is dominated by the total froude number @xmath1 . and
the unexpected assumption in miles - howard theorem leads the contradiction to other theories .
the miles - howard criterion is one of the most important theorem for stably stratified shear flow . according to this theorem , the flow is stable to perturbations when the richardson number @xmath8 exceeds a critical value @xmath9 @xcite .
this criterion is widely used in fluid dynamics , astrophysical fluid dynamics , oceanography , meteorology , etc .
specifically , it is the most important criterion for turbulence genesis in ocean modelling @xcite .
using arnold s method @xcite , the the corresponding criterion for three - dimensional stratified flow is @xmath10 @xcite .
consequently , it is widely believed that the stable stratification do favor the stability , and that all perturbations should decay when @xmath11 ( see , e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
however , the experiments indicate otherwise .
in fact , the flow might be unstable at very large @xmath8 @xcite . thus , to dispel the contradiction between the experiments and the miles - howard criterion , different explanations were postulated in literature : this interval , @xmath12 , separates two different turbulent regimes : strong mixing and weak mixing rather than the turbulent and the laminar regimes , as the classical concept states " @xcite , or , the richardson number criteria is not , in general , a necessary and sufficient condition " @xcite .
in fact , the miles - howard criterion is also in contradiction to other criteria for neutral stratification ( homogeneous ) fluid , e.g. , rayleigh - kuo s criterion ( e.g. * ? ? ? * ; * ? ? ? * ) , fjortoft s criterion @xcite , arnold s criteria @xcite and sun s criterion @xcite .
this contradiction was also noted long time ago by @xcite , who noticed that
miles criterion for stability is not the nature generalization of rayleigh s well - known sufficient condition for the stability of a homogeneous fluid in shear flow " .
he also also made an attempt to make a generalization @xcite .
following the frame work of @xcite , this study is an attempt to clear the confusion in theories and to build a bridge between the lab experiments and theories .
the taylor - goldstein equation for the stratified inviscid flow is employed @xcite , which is the vorticity equation of the disturbance @xcite . considering the flow with velocity profile @xmath13 and the density field @xmath14 , and the corresponding stability parameter @xmath15 ( the brunt - vaisala frequency ) , @xmath16 where @xmath17 is the acceleration of gravity , the single prime @xmath18 denotes @xmath19 , and @xmath20 denotes a stable stratification .
the vorticity is conserved along pathlines .
the streamfunction perturbation @xmath21 satisfies @xmath22\phi=0 , \label{eq : stable_stratifiedflow_taylorgoldsteineq}\ ] ] where @xmath23 is the real wavenumber and @xmath24 is the complex phase speed and double prime @xmath25 denotes @xmath26 . for @xmath23 is real , the problem is called temporal stability problem .
the real part of complex phase speed @xmath27 is the wave phase speed , and @xmath28 is the growth rate of the wave .
this equation is subject to homogeneous boundary conditions @xmath29 it is obvious that the criterion for stability is @xmath30 ( @xmath31 ) , for that the complex conjugate quantities @xmath32 and @xmath33 are also physical solutions of eq.([eq : stable_stratifiedflow_taylorgoldsteineq ] ) and eq.([eq : stable_parallelflow_rayleighbc ] ) . multiplying eq.([eq : stable_stratifiedflow_taylorgoldsteineq ] ) by the complex conjugate @xmath34 and integrating over the domain @xmath35 , we get the following equations @xmath36\ , dy = \int_{a}^{b } \frac{(u - c_r)^2-c_i^2}{|u - c|^4}n^2|\phi|^2\ , dy .
\label{eq : stable_stratifiedflow_taylorgoldsteineq_int_rea}\ ] ] and @xmath37|\phi|^2\,dy=0 .
\label{eq : stable_stratifiedflow_taylorgoldsteineq_int_img } \ ] ] in the case of @xmath38 , @xcite used eq.([eq : stable_stratifiedflow_taylorgoldsteineq_int_img ] ) to prove that a necessary condition for inviscid instability is @xmath39 , where @xmath40 is the inflection point and @xmath41 is the velocity at @xmath40 .
using eq.([eq : stable_stratifiedflow_taylorgoldsteineq_int_img ] ) , @xcite also pointed out that a necessary condition for instability is that @xmath42 should change sign .
but such condition is useless as there are two unknown parameters @xmath27 and @xmath43 . as a first step in our investigation
, we need to estimate the ratio of @xmath44 to @xmath45 .
this is known as the poincar s problem : @xmath46 where the eigenvalue @xmath47 is positively definite for any @xmath48 .
the smallest eigenvalue value , namely @xmath49 , can be estimated as @xmath50 @xcite . in departure from previous investigations
, we shall investigate the stability of the flow by using eq.([eq : stable_stratifiedflow_taylorgoldsteineq_int_rea ] ) and eq.([eq : stable_parallelflow_poincare ] ) .
as @xmath47 is estimated with boundary so the criterion is global .
we will also adapt a different methodology .
if the velocity profile is unstable ( @xmath51 ) , then the equations with the hypothesis of @xmath31 should result in contradictions in some cases .
following this , a sufficient condition for instability can be obtained .
firstly , substituting eq.([eq : stable_parallelflow_poincare ] ) into eq.([eq : stable_stratifiedflow_taylorgoldsteineq_int_rea ] ) , we have @xmath52 where g(y)&=&+k^2 + , and + h(y)&=&(+k^2)(u - c_r)^2+u(u - c_r)-n^2 .
[ eq : stable_stratifiedflow_hygy ] it is noted that @xmath53 for @xmath0 .
then @xmath54 if @xmath55 throughout the domain @xmath35 for a proper @xmath27 and @xmath23 . obviously , @xmath56 is a monotone function of @xmath23 : the smaller @xmath23 is , the smaller @xmath56 is . when @xmath57 , @xmath56 has the smallest value .
\label{eq : stable_stratifiedflow_hyfroude}\ ] ] if we define shear , parallel and rossby froude numbers @xmath59 , @xmath60 and @xmath61 as @xmath62 where the shear froude number @xmath60 is a dimensionless ratio of kinetic energy to potential energy . as @xmath63 plays the same role of @xmath64 effect in the rossby wave @xcite ,
the shear froude number @xmath61 is a dimensionless ratio of rossby wave kinetic energy to potential energy .
then @xmath65 equals to @xmath66 .
thus a general theorem for instability can be obtained from the above notations .
theorem 1 : if velocity @xmath67 and stable stratification @xmath68 satisfy @xmath55 or @xmath66 throughout the domain for a certain @xmath27 , the flow is unstable with a @xmath69 .
physically , @xmath2 implies that the total kinetic energy is smaller than the total potential energy .
so the potential energy might transfer to the kinetic energy after being disturbed , and the flow becomes unstable . on the other hand , when the potential energy is smaller than the kinetic energy ( @xmath70 ) , the flow is stable because no potential energy could transfer to the kinetic energy .
mathematically , we need derive some useful formula for applications , since there is still unknown @xmath27 in above equations . to this aim ,
we rewrite eq.([eq : stable_stratifiedflow_hyfroude ] ) as @xmath71 assume that the minimum and maximum value of @xmath72 within @xmath73 is respectively @xmath74 and @xmath75 .
it is from eq.([eq : stable_stratifiedflow_hy ] ) that @xmath76 for the smallest value of @xmath56 .
thus a general theorem for instability can be obtained from the above notations .
theorem 2 : if velocity @xmath67 and stable stratification @xmath68 satisfy @xmath55 throughout the domain for a certain @xmath77 , there must be a @xmath69 and the flow is unstable . it is from eq.([eq : stable_stratifiedflow_hy ] ) that @xmath55 requires @xmath78 less than @xmath79 .
the bigger @xmath68 is , the smaller @xmath56 is .
so the stable stratification has a destabilization mechanism in shear flow .
this conclusion is new as former theoretic studies always took the static stable stratification as the stable effects for shear flows .
according to eq.([eq : stable_stratifiedflow_hy ] ) , the bigger @xmath80 is , the more stable the flow is .
it is obvious that @xmath80 is bigger for @xmath4 than that for @xmath5 .
so the flow is more stable with the velocity profile @xmath4 .
although theorem 1 gives a sufficient unstable condition for instability , the complicated expression makes it difficult for application . in the following section we will derive simple and useful criteria
: ( a ) for @xmath81 , ( b ) for @xmath82 . , title="fig:",width=226 ] : ( a ) for @xmath81 , ( b ) for @xmath82 . , title="fig:",width=226 ] the simplest flow is the inviscid shear flow with @xmath38 .
the sufficient condition for instability is @xmath55 . to find such condition ,
we rewrite @xmath56 in eq.([eq : stable_stratifiedflow_hy ] ) as @xmath83 where @xmath84 and @xmath85
. then there may be three cases .
two of them have @xmath86 intersecting with @xmath87 at @xmath88 ( fig.[fig : hy ] ) .
the first case is that @xmath81 ; thus , @xmath89 always holds at @xmath90 as shown in fig.[fig : hy]a .
the second case is that @xmath82 ; thus , @xmath91 always holds in the whole domain , as shown in fig.[fig : hy]b . in this case , the flow might be unstable .
the sufficient condition for instability can be found from eq.([eq : rayleigh - hy ] ) as shown in fig.[fig : hy]b . given @xmath90 , eq.([eq : rayleigh - hy ] ) becomes @xmath92\ ] ] if @xmath93 is always satisfied , @xmath91 holds within the domain .
corollary 1.1 : if the velocity profile satisfies @xmath94 within the domain , the flow is unstable . since @xcite obtained a sufficient condition for stability , i.e. @xmath95 within the domain .
the above condition for instability is nearly marginal @xcite .
the last case is that @xmath96 throughout the domain ; thus , @xmath89 always exists somewhere within the domain , as shown in fig.[fig : tg - hy]a . for @xmath97 in case 3 and case 4.,title="fig:",width=226 ] for @xmath97 in case 3 and case 4.,title="fig:",width=226 ] if the static stratification is stable ( @xmath20 ) , then @xmath98 is positive .
the flow is unstable if @xmath56 is negatively defined within @xmath35 at @xmath57 .
we rewrite @xmath56 as @xmath99\\ & \displaystyle \times [ u+\frac{1}{2\mu}(u''+\sqrt{u''^2 + 4\mu n^2}\,)-c_r ] .
\end{array}\label{eq : taylorgoldsteineq - hy}\ ] ] the value of @xmath56 can be classified into 4 cases .
the first and the second ones ( @xmath88 and @xmath100 at @xmath101 ) are similar to discussed above and shown in fig[fig : hy]a and fig[fig : hy]b . for such cases , we have a sufficient condition for instability , @xmath102 this can be derived directly from eq.([eq : stable_stratifiedflow_sun_int_rea ] ) , similar to corollary 1.1 .
the first sufficient condition for instability is due to the shear instability , and the unstable criterion is eq.([eq : taylorgoldsteineq - sic1 ] ) .
corollary 1.2 : if the velocity profile satisfies @xmath103 within the domain , the flow is unstable .
the third case ( @xmath96 ) is also similar to the case in fig[fig : tg - hy]a , and the flow is stable .
the last one is unstable flow shown in fig.[fig : tg - hy]b , where @xmath96 and @xmath91 throughout . in the last case , the maximum of @xmath86 must be smaller than the minimum of @xmath87 so that a proper @xmath27 within the @xmath86 and @xmath87 could be used for the unstable waves . although the exact criterion can not be obtained as the required maximum and minimum can not be explicitly given , the approach is very straightforward .
nevertheless , we can also obtain some approximate criterion for the fourth case .
it is from eq.([eq : stable_stratifiedflow_hy ] ) that @xmath55 if the minimax of @xmath78 is less than the minimum of @xmath79 . as the minimax value of @xmath78 is @xmath104 when @xmath105 , we obtained a new criterion according to eq.([eq : stable_stratifiedflow_froude ] ) .
\label{eq : stable_stratifiedflow - fs2}\ ] ] thus a sufficient ( but not necessary ) condition for @xmath91 is that the following equation holds for @xmath73 .
@xmath107 from the above corollaries , the flow might be unstable if the static stable stratification is strong enough .
the stably stratification destabilize the flow , which is a new unstable mechanism .
the above corollary contradicts the previous results @xcite , but it agrees well with the recent theory @xcite , experiments @xcite and simulations @xcite .
again , we point out here that the flow is unstable due to potential energy transfer to kinetic energy under the condition of @xmath2 .
this conclusion is new because it is quite different from previous theorems in which the static stable stratification plays the role as a stabilizing factor for shear flows .
in the above investigation , it was found that stable stratification is a destabilization mechanism for the flow .
such finding is not surprising if one notes the terms in eq.([eq : stable_stratifiedflow_taylorgoldsteineq ] ) .
mathematically the sum of terms in square brackets should be negative for the wave solution .
thus both @xmath108 and @xmath109 do favor this condition .
this is why the unstable solutions always occur at @xmath110 in shear flow . and @xmath20 here might lead to @xmath54 .
physically , the perturbation waves are truncated in the neutral stratified flow .
but the stable stratification allows wide range of waves in the perturbation .
such waves might interact with each other like what was illustrated in @xcite . as theorem 1 is the only sufficient condition ,
it is hypothesized that the criterion is not only the sufficient but also the necessary condition for instability in stably stratified flow .
this hypothesis might be criticized in that the flow might be unstable ( @xmath54 ) if @xmath56 changes sign within the interval ( fig[fig : tg - hy]a ) , where a proper chosen @xmath21 would let the right hand of eq.([eq : stable_stratifiedflow_sun_int_rea ] ) become negative . however , this criticism is not valid for the case in fig[fig : tg - hy]a .
it is from the well - known criteria ( e.g. rayleigh s inflexion point theorem ) that the proper chosen @xmath21 always let the right hand of eq.([eq : stable_stratifiedflow_sun_int_rea ] ) vanish .
it seems that the flow tends to be stable , or the perturbations have a prior policy to let @xmath31 .
the flow become unstable unless any choice of @xmath21 would let the right hand of eq.([eq : stable_stratifiedflow_sun_int_rea ] ) be negative .
in this situation , we hypothesize that theorem 1 fully solves the stability problem . in inviscid shear flows
, it has been recognized that very short - wave perturbations are dynamically stable under neutral stratification , and the dynamic instability is due to the larger wavelengths @xcite .
it should be noted that rayleigh s case is reduced to the kelvein - helmholtz vortex sheet model under the long - wave limit @xmath111 @xcite . we have shown that this can be extended to shear flows , and that the growth rate @xmath112 , is proportional to @xmath113 @xcite .
such conclusion can be simply generated to the stratified shear flows , which can be seen from eq.([eq : stable_stratifiedflow_hygy ] ) .
if @xmath23 is larger than a critical value @xmath114 , the sufficient condition in theorem 1 can not be satisfied and the flow is stable . for shortwave ( @xmath115 ) ,
@xmath56 is always larger than that for long - wave @xmath111 .
the long - wave instability in the stratified shear flow was also noted by @xcite and @xcite , who showed a likelihood of @xmath116 at @xmath117 .
the long - wave instability theory can explain the results in numerical simulations @xcite , where the unstable perturbations are long - wave . in the above investigations ,
an parameter @xmath47 is used , which represents the ratio of two integrations with boundaries .
so the criteria are global . on the other hand
, we can also investigate the local balance without boundary conditions .
for example , consider the flow within a thick layer @xmath118 .
the velocity is @xmath119 , and the kinetic energy is @xmath120 .
the stratification is @xmath68 , and the potential energy is @xmath121 , where @xmath122 is the thickness of the layer .
the froude number is @xmath123 for @xmath124 .
the instability criterion in eq.([eq : stable_stratifiedflow - criterion ] ) becomes @xmath125 if local gradient richardson number exceeds @xmath126 , the local disturbances is unstable .
however , the flow might be stable as the globe total froude number @xmath3 .
this criterion is opposite to miles - howard theorem , we will show why miles - howard theorem is not correct from their derivations . in the inviscid
shear flow , the linear theories , e.g. , rayleigh - kuo cirterion @xcite , fjrtoft criterion @xcite and sun s criterion @xcite , are equal to arnold s nonlinear stability criterion @xcite .
arnold s first stability theorem corresponds to fjrtoft s criterion @xcite , and arnold s second nonlinear theorem corresponds to sun s criterion @xcite .
it is obvious that the present theory , especially corollary 1.1 is a natural generalization of inviscid theories . in the stratified flow , @xcite and
@xcite applied a transform @xmath7 to eq.([eq : stable_stratifiedflow_taylorgoldsteineq ] ) , which allows different kind of perturbations .
thus @xmath127 gives miles s theory and @xmath128 gives howard s semicircle theorem .
considering that @xmath128 and @xmath38 @xcite , eq.([eq : stable_stratifiedflow_taylorgoldsteineq_int_rea ] ) becomes @xmath129 dy=0 .
\label{eq : stable_stratifiedflow_taylorgoldsteineq_miles_rea}\ ] ] it is from eq.([eq : stable_stratifiedflow_taylorgoldsteineq_miles_rea ] ) that all the inviscid flows ( no mater what the velocity profile @xmath13 is ) must be temporal unstable ( @xmath23 is real ) .
this contradicts the criteria ( both linear and nonlinear ones ) for inviscid shear flow .
so the wavenumber @xmath23 in eq.([eq : stable_stratifiedflow_taylorgoldsteineq_miles_rea ] ) should be complex @xmath130 .
besides , from eq.([eq : stable_stratifiedflow_hy ] ) , eq.([eq : taylorgoldsteineq - hy ] ) and fig.[fig : tg - hy]b , the unstable @xmath27 might be either within the value of @xmath67 or beyond the value of @xmath67 .
this also contradicts the howard s semicircle theorem for the stratified flow .
it implies that the transform @xmath131 is not suitable for temporal stability problem .
taking @xmath127 , howard extracted a new equation from taylor - goldstein equation , @xmath132'-[k^2(u - c)+\frac{u''}{2}+(\frac{1}{4}u'^2-n^2)/(u - c)]f=0 \label{eq : stable_stratifiedflow_taylorgoldsteineq_howard}\ ] ] after multiplying above equation by the complex conjugate of @xmath131 and integrating over the flow regime , then the imaginary part of the expression is @xmath133 miles - howard theorem concludes that if @xmath134 , then @xmath135 for instability .
however , the transform @xmath136 requires a complex function @xmath131 , even though both @xmath21 and @xmath137 are real . in that @xmath138 might be complex somewhere as @xmath139 .
consequently , the wave number @xmath23 in eq.([eq : stable_stratifiedflow_taylorgoldsteineq_howard ] ) is a complex number but no longer a real number as that assumed in taylor - goldstein equation .
the complex wavenumber @xmath23 leads to spatial stability problem but temporal stability problem investigated in this study .
the assumption of @xmath31 with @xmath140 implies the flow is unstable with @xmath141 . however , howard ignored this in his derivations .
that is why miles - howard theorem leads contradictions to the present studies .
although the transform @xmath7 leads some contradictions with rayleigh criterion and present results , it might be useful for the viscous flows . in these flows ,
the spatial but temporal stability problem is dominated , and @xmath130 is the complex wavenumber .
it is well known that the plane couette flow is viscously unstable for reynolds number @xmath142 from the experiments but viscously stable from the orr - sommerfeld equation @xcite .
if applying the transform in eq.([eq : stable_stratifiedflow_taylorgoldsteineq_miles_rea ] ) all the inviscid flows must be unstable .
thus the plane couette flow might be stable only for @xmath143 due to the stabilization of the viscosity .
it is argued that the taylor - goldstein equation represents temporal instability , the transform represents spatial instability @xcite . in that
the perturbation is seen along with the flow at the speed of @xmath144 in @xcite .
the transform @xmath7 also turns real wavenumber @xmath23 into complex number , @xmath31 implies @xmath145 .
the assumption of real @xmath23 after transform will leads to contradictions with the results derived from the taylor - goldstein equation .
so the previous investigators can hardly generalize their results from homogeneous fluids to stratified fluids .
in summary , the stably stratification is a destabilization mechanism , and the flow instability is due to the competition of the kinetic energy with the potential energy .
globally , the flow is always unstable when the total froude number @xmath66 , where the larger potential energy might transfer to the kinetic energy after being disturbed .
locally , the flow is unstable as the gradient richardson number @xmath6 .
the approach is very straightforward and can be used for similar analysis . in the inviscid stratified flow
, the unstable perturbation must be long - wave scale .
this result extends the rayleigh s , fjrtoft s , sun s and arnold s criteria for the inviscid homogenous fluid , but contradicts the well - known miles and howard theorems .
it is argued here that the transform @xmath7 is not suitable for temporal stability problem , and that it will leads to contradictions with the results derived from taylor - goldstein equation .
the author thanks dr .
yue p - t at virginia tech , prof .
yin x - y at ustc , prof .
wang w. at ouc and prof .
huang r - x at whoi for their encouragements .
this work is supported by the national basic research program of china ( no .
2012cb417402 ) , and the knowledge innovation program of the chinese academy of sciences ( no .
kzcx2-yw - qn514 ) . | the temporal instability of stably stratified flow was investigated by analyzing the taylor - goldstein equation theoretically . according to this analysis
, the stable stratification @xmath0 has a destabilization mechanism , and the flow instability is due to the competition of the kinetic energy with the potential energy , which is dominated by the total froude number @xmath1 .
globally , @xmath2 implies that the total kinetic energy is smaller than the total potential energy .
so the potential energy might transfer to the kinetic energy after being disturbed , and the flow becomes unstable . on the other hand ,
when the potential energy is smaller than the kinetic energy ( @xmath3 ) , the flow is stable because no potential energy could transfer to the kinetic energy .
the flow is more stable with the velocity profile @xmath4 than that with @xmath5 . besides
, the unstable perturbation must be long - wave scale .
locally , the flow is unstable as the gradient richardson number @xmath6 .
these results extend the rayleigh s , fjrtoft s , sun s and arnold s criteria for the inviscid homogenous fluid , but they contradict the well - known miles - howard theorem .
it is argued here that the transform @xmath7 is not suitable for temporal stability problem , and that it will lead to contradictions with the results derived from the taylor - goldstein equation .
however , such transform might be useful for the study of the orr - sommerfeld equation in viscous flows . |
the principle of linear superposition is central to quantum mechanics .
however , it is difficult to create and observe superposed states because the fragile coherence of these states can be easily spoiled by the environment .
typical examples are the schrdinger cat states ( scss ) @xcite .
many theoretical schemes @xcite have been proposed to generate scss and superpositions of macroscopic states ( smss ) in optical systems .
also , much experimental progress has been made to demonstrate smss and scss : in superconducting systems ( e.g. ref .
@xcite ) , laser trapped ions @xcite , optical systems constructed by rydberg atoms , and superconducting cavity in the microwave regime @xcite .
the smss , which are formed by two optical coherent states , e.g. in ref .
@xcite , have been investigated for applications in quantum information processing @xcite .
these states can be used as a robust qubit encoding for a single bosonic mode subject to amplitude damping .
they can also be used to study both the measurement process and decoherence by coupling the system to the external environment @xcite .
thus , generating and measuring smss and scss are not only important to understand fundamental physics , but also to explore potential applications . superconducting quantum devices @xcite allow to perform quantum state engineering including the demonstration of scss and smss .
theoretical schemes to generate superpositions that are different from the above experiments @xcite have also been proposed @xcite in superconducting quantum devices . for example , the scheme in ref .
@xcite generates a superpositions of bloch states for the current of a josephson junction .
marquardt and bruder @xcite proposed ways to create smss for a harmonic oscillator approximated by a large superconducting island capacitively coupled to a smaller cooper - pair box .
et al . _
@xcite proposed a similar scheme as in ref .
@xcite but using a micromechanical resonator as the harmonic oscillator .
a review paper on micromechanical resonators @xcite can be found in ref .
@xcite . in ref .
@xcite , a scheme was proposed to generate smss and squeezed states for a superconducting quantum interference device ( squid ) ring modelled as an oscillator . since then , several proposals have been made which focus on superconducting qubits interacting with the nonclassical electromagnetic field @xcite .
optical states allow a fast and convenient optical transmission of the quantum information which is stored in charge qubits .
compared with the harmonic system @xcite formed by the large superconducting junction and the micromechanical resonator , optical qubits can easily fly relatively long distances between superconducting charge qubits .
moreover , the qubit formed by smss enables a more efficient error correction than that formed by the single photon and vacuum states , and the generation and detection of coherent light are easy to implement .
in contrast to @xcite , here we aim at generating scss in the interaction system between a single - mode microwave cavity field and a squid - based charge qubit , and then creating smss by virtue of the measurements of the charge states .
the generation of such states has been studied theoretically @xcite and demonstrated in optical cavity qed experiments @xcite . however , in these cases : i ) several operations are needed because atoms must pass through three cavities , and ii ) the interaction times are tuned by the controlling velocity of the atoms flying through the cavity . in our proposal , we need only one cavity , and interaction times are controlled by changing the external magnetic field . although our scheme is similar to that proposed in ref .
@xcite , the interaction between the box and the resonator in ref .
@xcite is not switchable . due to the fixed coupling in ref .
@xcite , the transfer of information between the micromechanical resonator and the box still exists even after the scss or smss are produced . in our proposal ,
the interaction between the cavity field and the squid can be switched off by a classical magnetic field after the scss or smss are generated .
furthermore , three operations , with different approximations made in every operation , are required in ref .
in addition , in order to minimize the environmental effect on the prepared state , the number of operations and instruments should be as small as possible ; _ one _ operation is enough to generate scss or smss .
thus our proposed scheme offers significant advantages over the pioneering proposals in refs .
@xcite and @xcite .
we consider a squid - type qubit superconducting box with @xmath0 excess cooper - pair charges connected to a superconducting loop via two identical josephson junctions with capacitors @xmath1 and coupling energies @xmath2 . a controllable gate voltage @xmath3 is coupled to the box via the gate capacitor @xmath4 with dimensionless gate charge @xmath5 .
the qubit is assumed to work in the charge regime with @xmath6 , where @xmath7 , @xmath8 , @xmath9 , and @xmath10 are the boltzmann constant , temperature , charge and superconducting gap energies , respectively . for known charge qubit experiments , e.g. in ref .
@xcite , @xmath11 mk which means @xmath12ev , @xmath13ev , @xmath14ev , and @xmath15ev , so the above inequalities are experimentally achievable .
we consider a gate voltage range near a degeneracy point @xmath16 , where only two charge states , called @xmath17 and @xmath18 , play a leading role .
the other charge states with a much higher energy can be neglected , which implies that the superconducting box can be reduced to a two - level system or qubit @xcite .
this superconducting two - level system can be represented by a spin-@xmath19 notation such that the charge states @xmath17 and @xmath18 correspond to eigenstates @xmath20 and @xmath21 of the spin operator @xmath22 , respectively .
if such a qubit is placed into a single - mode superconducting cavity , the hamiltonian can be written as @xcite @xmath23 where the first term represents the free hamiltonian of the single - mode cavity field with frequency @xmath24 , and @xmath25 with the single - electron charging energy @xmath26 . here
@xmath27 is the flux quantum , @xmath28 is the flux generated by the classical magnetic field through the squid , and @xmath29 is the identity operator .
the last term in eq .
( [ eq:1 ] ) is the nonlinear photon - qubit interaction .
the parameter @xmath30 has units of magnetic flux and its absolute value represents the strength of the quantum flux inside the cavity .
we will later on assume this
quantum magnetic flux " @xmath30 to be small , becoming our perturbation parameter .
the parameter @xmath30 can be written as @xmath31 where @xmath32 is the mode function of the single - mode cavity field , with annihilation ( creation ) operators @xmath33 , and @xmath34 is the surface defined by the contour of the squid . for convenience ,
hereafter , we denote @xmath35 and @xmath36 by @xmath37 and @xmath38 , respectively .
the cosine in eq .
( [ eq:1 ] ) can be decomposed into classical and quantized parts ; eq . ( [ eq:1 ]
) can then be expressed as @xmath39 the factors @xmath40 $ ] and @xmath41 $ ] can be further expanded as a power series in @xmath33 . for the single photon transition between the states
@xmath42 and @xmath43 , if the condition @xmath44 is satisfied , all higher orders of @xmath45 can be neglected in the expansion of eq .
( [ eq:2 ] ) . to estimate the interaction coupling between the cavity field and the qubit
, we assume that the single - mode cavity field is in a standing - wave form @xmath46 where @xmath47 , @xmath48 , @xmath49 , and @xmath50 are the volume of the cavity , permittivity of the vacuum , light speed and wave vector of the cavity mode , respectively .
because the superconducting microwave cavity is assumed to only contain a single mode of the magnetic field , the wave vector @xmath51 is a constant for the given cavity .
the magnetic field is assumed to propagate along the @xmath52 direction and the polarization of the magnetic field is along the normal direction of the surface area of the squid .
if the area of the squid is , e.g. , of the order of @xmath53 , then its linear dimension , e.g. , approximately of the order of 10@xmath54 m , should be much less than the microwave wavelength of the cavity mode .
thus , the mode function @xmath55 can be considered to be approximately independent of the integral area and the factor @xmath56 only depends on the position @xmath57 where the qubit is located .
so the parameter @xmath30 can be expressed as @xmath58 which shows that the parameter @xmath59 depends on the area @xmath34 and the position @xmath57 of the squid , the wavelength @xmath60 of cavity field , and the volume @xmath47 of the cavity .
it is obvious that a larger @xmath34 for the squid corresponds to a larger @xmath59 . if the squid is placed in the middle of a cavity with full wavelength , that is , @xmath61 .
then @xmath62 , the interaction between the cavity field and the qubit reaches its maximum , and @xmath63 in the microwave region with @xmath64 cm @xmath65 cm . for a half- or quarter - wavelength cavity , the condition @xmath66 ) can be safely made in the microwave regime , and then eq . ( [ eq:2 ] ) can be further simplified ( up to first order in @xmath67 ) as @xmath68 where @xmath69 is a dimensionless complex number with its absolute value equal to the dimensionless quantum magnetic flux , and it is defined by @xmath70
we assume that the qubit is initially in the ground state @xmath71 where @xmath72 is eigenstate of the pauli operator @xmath73 with the eigenvalue @xmath74 .
the cavity field is assumed initially in the vacuum state @xmath75 .
now let us adjust the gate voltage @xmath76 and classical magnetic field such that @xmath16 and @xmath77 , and then let the whole system evolve a time interval @xmath78 .
the state of the qubit - photon system evolves into @xmath79\tau\}|0\rangle|g\rangle\nonumber\\ & = & \frac{1}{2}[a(\omega)|0\rangle|+\rangle+a(-\omega)|0\rangle|-\rangle]\nonumber\\ & = & \frac{1}{2 } ( |\alpha\rangle+|-\alpha\rangle)|g\rangle\nonumber\\ & + & \frac{1}{2}(|\alpha\rangle-|-\alpha\rangle)|e\rangle,\end{aligned}\ ] ] where the complex rabi frequency @xmath80 , @xmath81\tau\}$ ] , and a global phase factor @xmath82 $ ] has been neglected .
@xmath83 denotes coherent state @xmath84 with @xmath85 in the derivation of eq .
( [ eq:4 ] ) , we use the formula @xmath86
= \exp[f_{1}a^{\dagger}]\exp[f_{2}a^{\dagger}a]\exp[f_{3}a]\exp[f_{4}]$ ] with the relations @xmath87 , @xmath88 , @xmath89 , and @xmath90 .
after the time interval @xmath78 , we impose @xmath91 by adjusting the classical magnetic field , thus the interaction between the charge qubit and the cavity field is _ switched off _ ( e.g. , the last term in eq .
( [ eq:3 ] ) vanishes ) .
( [ eq:4 ] ) shows that entanglement of the qubit and the microwave cavity field can be prepared for an evolution time @xmath92 with the integer number @xmath93 , then schdinger cat states can be created @xcite . if the condition @xmath94 is satisfied in eq .
( [ eq:4 ] ) , the smss of the cavity field denoted by @xmath95 @xmath96 can be obtained by measuring the charge state @xmath37 or @xmath38 , by using , for example , a single - electron transistor ( set ) . if we initially inject a coherent light @xmath97 , then by using the same method as in the derivation of eq .
( [ eq:4 ] ) , we can also obtain the entanglement of two different optical coherent states @xmath98 and qubit states with the evolution time @xmath99 : @xmath100 where @xmath101\ ] ] and @xmath102\ ] ] with @xmath103 after a time interval @xmath99 , we can switch off the interactions between the charge qubit and the cavity field by setting @xmath91 and @xmath104 . measuring the charge states , we can obtain another smss denoted by @xmath105 @xmath106 with normalized constant @xmath107 where @xmath108 can be easily obtained @xcite by the above expression of @xmath109 , for example , @xmath110-i2\kappa\alpha^{\prime}\sin(\omega \tau_{1})\right\},\ ] ] here we assume that the injected coherent field has a real amplitude @xmath111 . in eq .
( [ eq:5 ] ) , we entangle two different superpositions of coherent states with the ground and excited states of the qubit . we can also entangle two different coherent states @xmath98 with the qubit states by applying a classical flux such that @xmath112 . then with the time evolution @xmath113 , we have @xmath114 it should be noticed that a global phase factor @xmath82 $ ] has been neglected in eqs .
( [ eq:5 ] ) and ( [ eq:6 ] ) . from a theoretical point of view , if we can keep the expansion terms in eq .
( [ eq:2 ] ) up to second order in @xmath115 , we can also _ prepare a superposition of two squeezed coherent states , which could be used to encode an optical qubit _
@xcite . to obtain this superposition of two squeezed coherent states , we can set @xmath16 and @xmath91 , and derive the hamiltonian from eq .
( [ eq:2 ] ) to get ( up to second order in @xmath69 ) @xmath116 if the system is initially in the coherent state @xmath117 , and if the charge qubit is in the ground state @xmath38 , we can entangle qubit states with superpositions of two different squeezed coherent states with an evolution time @xmath118 as @xmath119|g\rangle \nonumber\\ & & + \frac{1}{2}\left[e^{-i\theta t}|\gamma , -i\frac{\xi^{*2}e_{\rm j}}{\hbar } t\rangle - e^{i\theta t}|\gamma , i\frac{\xi^{*2}e_{\rm j}}{\hbar } t\rangle \right]|e\rangle,\end{aligned}\ ] ] where @xmath120 and @xmath121 here , @xmath122 denote squeezed coherent states , and the degree of squeezing @xcite is determined by the time - dependent parameter @xmath123 .
a superposition of two squeezed coherent states can be obtained by the measurement on the charge qubit .
however , we should note that if we keep to first order in @xmath124 the expansions of eq .
( [ eq:2 ] ) , the interaction between the cavity field and the charge qubit is switchable ( e.g. , the last term in eq . ( [ eq:3 ] ) vanishes for @xmath91 ) .
but if we keep terms up to second order in @xmath125 for the expansions of eq .
( [ eq:2 ] ) , then the qubit - field coupling is not switchable .
our analytical expressions show how to prepare the schrdinger states for the system of the microwave cavity field and the squid - based charge qubit , we further show that the superpositions of two macroscopically distinguishable states can also be created by measuring the charge states .
however , similarly to the optical cavity qed @xcite , prepared superpositions of states are limited by the following physical quantities : the rabi frequency @xmath126 ( which determines the quantum operation time @xmath127 of two charge qubit states through the cavity field ) , the lifetime @xmath128 of the cavity field , the lifetime @xmath129 and dephasing time @xmath130 of the charge qubit , as well as the measurement time @xmath131 on the charge qubit .
we now estimate the rabi frequency @xmath132 in the microwave regime for a standing - wave field in the cavity . a squid with an area of about @xmath133 is assumed to be placed in the middle of the cavity . in the microwave regime with different ratios of @xmath134 , we provide a numerical estimate of @xmath135 for @xmath136 in a full - wavelength cavity , shown in fig .
[ fig1](a ) , and a quarter - wavelength cavity , shown in fig . [ fig1](b ) .
the results reveal that a shorter wavelength of cavity field corresponds to a larger rabi frequency @xmath132 . for example , in the full - wavelength cavity and the case of the ratio @xmath137 , @xmath135 with microwave length @xmath138 cm is of the order of @xmath139 hz , and yet it is about @xmath140 hz for a microwave wavelength of @xmath141 cm . in both cases ,
the transition times from @xmath142 to @xmath143 are about @xmath144 s and @xmath138 s respectively , where @xmath75 ( @xmath145 ) is the vacuum ( single - photon ) state . the experiment for this scheme should be easier for shorter wavelengths than for long wavelengths .
since the cavity field has higher energy for the shorter wavelength , so it is better to choose the material with a larger superconducting energy gap to make the josephson junction for the experiment in the region of the shorter microwave wavelengths . for a fixed wavelength ,
the effect of the ratios @xmath134 on the coupling between the cavity field and the charge qubit is not so large .
however , decreasing the volume @xmath47 of the cavity can also increase the coupling . in order to obtain a sms , the readout time @xmath146 of the charge qubit should be less than the dephasing time @xmath130 of the charge qubit ( because the relaxation time @xmath129 of the charge qubit is longer than its dephasing time @xmath130 ) and the lifetime time @xmath147 of the cavity field .
for example , in ref .
@xcite with a set of given parameters , the estimated time , @xmath148 ns , is less than @xmath149 ns @xcite . for a good cavity @xcite
, the quality factor @xmath150 can reach very high values , such as @xmath151 , and then the lifetimes of the microwave field would be in the range @xmath152 s @xmath153 s , which implies @xmath154 .
so the readout is possible within current technology .
it is easier to prepare a sms in such a system even when the coupling between the charge qubit and the cavity field is weak because , in principle , two different coherent states could be obtained with a very short time @xmath127 such that @xmath155 . versus the microwave wavelength @xmath60 for a full - wavelength cavity ( a ) and a quarter - wavelength cavity ( b ) with ratios @xmath156 ( top solid line ) , 7 ( dashed line ) , 10 ( dashed - dotted line ) , 15 ( bottom dotted line ) , respectively.,title="fig:",width=151 ] versus the microwave wavelength @xmath60 for a full - wavelength cavity ( a ) and a quarter - wavelength cavity ( b ) with ratios @xmath156 ( top solid line ) , 7 ( dashed line ) , 10 ( dashed - dotted line ) , 15 ( bottom dotted line ) , respectively.,title="fig:",width=151 ]
in conclusion , we have analyzed the generation of schrdinger cat states via a controllable squid - type charge qubit . based on our scheme ,
the smss can be created by using one quantum operation together with the quantum measurements on the charge qubit .
after the scss or smss are created , the coupling between the charge qubit and the cavity field can be switched off , in principle . because all interaction terms of higher order in @xmath115 are negligible for the coupling constant @xmath157 .
this results in a switchable qubit - field interaction .
this means _ sudden switching _ of the flux on time scales of the inverse josephson energy(@xmath158ghz ) . at present
this is difficult but could be realized in the future .
we have also proposed a scheme to generate superpositions of two squeezed coherent states if we can keep the expansion terms in eq .
( [ eq:2 ] ) up to second order in @xmath115 .
however , in this case the interaction between the cavity field and the charge qubit can not be switched off . by using the same method employed for trapped ions @xcite , we can measure the decay rate of the smss and obtain the change of the @xmath150 value due to the presence of the squid .
also , the generated smss can be used as a source of optical qubits .
our suggestion is that the first experiment for generating nonclassical states via the interaction with the charge qubit should be the generation of superpositions of two macroscopically distinct coherent states .
it needs only one quantum operation , and the condition for the coupling between the cavity field and the charge qubit can be slightly relaxed .
this proposal should be experimentally accessible in the near future .
we acknowledge comments on the manuscript from x. hu , s. c. bernstein , and c.p .
this work was supported in part by the national security agency ( nsa ) , advanced research and development activity ( arda ) under air force office of research ( afosr ) contract number f49620 - 02 - 1 - 0334 , and by the national science foundation grant no .
eia-0130383 .
see , e.g. , b. yurke and d. stoler , phys .
lett . * 57 * , 13 ( 1986 ) ; b. yurke , w. schleich , and d.f .
walls , phys .
a * 42 * , 1703 ( 1990 ) ; v. buzek , h. moya - cessa , p. l. knight , s. j. d. phoenix , _ ibid . _
* 45 * , 8190 ( 1992 ) ; m.o .
caceres , _ ibid .
_ * 33 * , 647 ( 1986 ) ; t. ogawa , m. ueda , and n. imoto , _ ibid . _ * 43 * , r6458 ( 1991 ) ; a. miranowicz , r. tanas , and s. kielich , quantum opt .
* 2 * , 253 ( 1990 ) .
van der wal , a.c.j .
ter haar , f.k .
wilhelm , r.n .
schouten , c.j.p.m .
harmans , t.p .
orlando , s. lloyd , j. e. mooij , science * 290 * , 773 ( 2000 ) ; s. han , y. yu , x. chu , s. chu , and z. wang , _ ibid . _ * 293 * 1457 ( 2001 ) ; j.r .
friedman , v. patel , w. chen , s.k .
tolpygo , and j. e. lukens , nature * 406 * , 43 ( 2000 ) ; j.e .
mooij , t.p .
orlando , l. levitov , l. tian , c.h .
van der wal , and s. lloyd , science * 285 * , 1036 ( 1999 ) ; y. yu , s. han , xi chu , s.i .
chu , and z. wang , _ ibid .
_ * 296 * , 889 ( 2002 ) ; i. chiorescu , y. nakamura , c.j.p.m .
harmans , and j.e .
mooij , _ ibid . _ * 299 * , 1869 ( 2003 ) ; a.j .
berkley , h. xu , r.c .
ramos , m.a .
gubrud , f.w .
strauch , p.r .
johnson , j.r .
anderson , a.j .
dragt , c.j .
lobb , and f.c .
wellstood , _ ibid . _
* 300 * , 1548 ( 2003 ) .
you , j.s .
tsai , and f. nori , phys .
lett . * 89 * , 197902 ( 2002 ) .
a longer version of it is available in cond - mat/0306203 ; see also _ new directions in mesoscopic physics _ , edited by r. fazio , v.f .
gantmakher , and y. imry ( kluwer academic publishers , 2003 ) , 351 . m.d .
lahaye , o. buu , b. camarota , and k.c .
schwab , science , * 304 * , 74 ( 2004 ) ; e.k .
irish and k. schwab , phys . rev .
b * 68 * , 155311 ( 2003 ) ; a. hopkins , k. jacobs , s. habib , and k. schwab , phys .
b * 68 * , 235328 ( 2003 ) j. krause , m.o.scully , and h. walther , phys . rev .
a * 36 * , r4547 ( 1987 ) ; j. krause , m.o.scully , t. walther , and h. walther , _ ibid . _ * 39 * , 1915 ( 1989 ) ; s. brattke , b.t.h .
varcoe , and h. walther , phys .
lett . * 86 * , 3534 ( 2001 ) ; j.m .
raimond , m. brune , and s. haroche , rev .
phys . * 73 * , 565 ( 2001 ) . | we propose how to generate schrdinger cat states using a microwave cavity containing a squid - based charge qubit . based on the measurement of charge states ,
we show that the superpositions of two macroscopically distinguishable coherent states of a single - mode cavity field can be generated by a controllable interaction between a cavity field and a charge qubit .
after such superpositions of the cavity field are created , the interaction can be switched off by the classical magnetic field through the squid , and there is no information transfer between the cavity field and the charge qubit .
we also discuss the generation of superpositions of two squeezed coherent states . |
the logical propositions that represent physical questions in quantum mechanics are of the form ` the state of the system is an eigenvector of the operator @xmath0 with eigenvalue @xmath1 ' , for an observable @xmath0 .
the non commutativity of the observable operators gives place to a non boolean logic for the propositions , in contrast to the case of the propositions about the phase space of a system in classical physics .
this is mathematically described by the orthomodular lattice structure of projectors in hilbert space , which is often called quantum logic @xcite .
the algebra of propositions about the space in classical physics is boolean .
consider for example a particle , and the set of propositions of the form ` the particle is in the set @xmath2 ' .
the opposite in this logic to the proposition given by the subset @xmath2 of the space is given by @xmath3 , the set of all points that do not belong to @xmath2 .
a positive answer to @xmath2 implies a negative one to @xmath3 .
the intersection and union of sets corresponds to the conjunction and disjunction of logical propositions about the physical system .
the algebra of propositions becomes the same as the boolean algebra of subsets of the space in set theory .
this structure is inherent to the space , because it will be the same for any system .
however , as we will see in the following , the situation when including time is different . in the relativistic context
the kinematics imposed by the causal structure will induce a non boolean character for the algebra of propositions given by certain space - time subsets ( see @xcite for the case of galilean physics ) .
a particle can not be concentrated at a given point @xmath4 in space - time because it existed in the past and will exist in the future of @xmath4 , but we can think that the answer to the proposition given by a point is true if the particle passes through @xmath4 and false if not .
the causal structure gives a definite prescription of what the opposite means in this case , because if the particle passes through @xmath4 it can not pass through any point spatially separated from @xmath4 .
all the points non related by a time - like curve to a given set @xmath2 will form the causal and logical complement of @xmath2 .
a surprising fact about the algebra of propositions that arises along this line of thought from the causal structure of space - time alone , independently of the propagating physical system , is that it is also an orthomodular lattice .
though , we think , not widely known , this very interesting point was shown for minkowski space some time ago in ref.@xcite . in this work we will show how to define the lattice of causally closed sets @xmath5 for a general space - time @xmath6 , even non causally well behaved , and prove that it is a complete orthomodular lattice .
in addition , we will show that the lattice is atomic , and find its atoms .
they are the points for causally well behaved space - times , but in general they can be of three different kinds .
we will also find the necessary and sufficient conditions for this lattice to be irreducible .
the lattice @xmath7 can have interest in two senses .
we think that this lattice structure can be more fundamental than the causal structure of the space - time , much in the spirit of the studies of the postulates of quantum mechanics from the lattice perspective @xcite . in this sense
this paper can be consider as a step in the study of the causal structure of space - times from an algebraic point of view .
in fact , we find that certain undesirable aspects of the causal structure of general space - times as closed time - like curves can be eliminated or localized in the structure of @xmath7 , this lattice being largely independent of the details of the metric in the bad sectors of space - time . two space - times related by a conformal transformation will have the same causal structure and lattice @xmath7 , however the converse can be false .
the other aspect of interest is the possible relation of @xmath7 with quantum mechanics and quantum field theory , motivated by the intriguing fact that causal structure and hilbert space share an important piece of its mathematical structure .
we will not develop in this sense here .
we note , however , that the causally closed sets in @xmath8 appear naturally in the algebraic approach to quantum field theory . in that context
, we have a set of @xmath9-algebras associated to space - time sets .
to a set in @xmath6 and its causal complement correspond two @xmath9-algebras that commute
. however , the whole relation between the lattice structure and the operations among local algebras seems not to be completely clear at the moment @xcite .
we will say a bit more in section iii . to make this article as much self contained as possible , in the following section we include a brief introduction to orthomodular lattices and some relevant mathematical preliminaries .
further details can be found in refs .
@xcite . in section iii
we define the orthocomplemented complete lattice @xmath7 and discuss several related possibilities for constructing causal set lattices . in section iv we show that @xmath10 is orthomodular . in section v we show that it is atomic and find the atoms , which are an important piece in the understanding of the general structure of the lattice that will allows us to characterize irreducibility . in the last section we make some comments and discuss directions open to future work .
the family of subsets @xmath11 of any set @xmath12 forms a boolean algebra when considered with the relation of inclusion , and the operations of intersection , union , and complement ( denoted @xmath13 , @xmath14 , @xmath15 and @xmath16 respectively ) .
this is at the base of the relation between set theory and classical logic , where the propositions can be restated in terms of the subsets of a given set @xmath12 in set theory . to a given proposition
@xmath17 we assign a subset @xmath2 with the interpretation that an element belongs to @xmath2 if and only if it satisfies the proposition @xmath17 .
thus , the operations of implication , conjunction , disjunction , and negation between propositions are interpreted as the relation of inclusion , and the operations of intersection , union , and complement between subsets respectively . in quantum mechanics the space of individuals is given by a hilbert space @xmath18 . the corresponding propositions or questions that are answered with a yes or a no ,
are given by projection operators @xmath19 , @xmath20 . these operators must be observables , that is , auto - adjoint operators , @xmath21 , such that their domain is all @xmath22 , and in consequence they are also bounded .
the projection operators are in one to one correspondence with the closed subspaces @xmath23 of @xmath22 , such that @xmath24 , and @xmath25 . here
we use the notation @xmath26 for the subspace orthogonal to @xmath27 , which is always closed .
the interpretation is that the proposition @xmath19 is true for the elements of @xmath23 and false for the elements of @xmath28 .
here is possible to talk in terms of propositions ( projection operators ) or in terms of closed linear subspaces indistinctly , as it happens in the case of classical logic , where it is possible to talk in terms of propositions or in terms of sets .
the set @xmath29 of closed subspaces of a hilbert space can be furnished with a structure that resembles the one in @xmath11 , where the relation of inclusion and the operation of intersection between linear subspaces are the set inclusion and intersection , but where the set complement is replaced by the operation of taking the orthogonal subspace , and the union of two sets by the sum of subspaces .
the resulting structure is called an orthomodular lattice . in
what follows we will describe this structure in detail .
the inclusion between subsets is an order relation , that makes @xmath30 ( and @xmath31 ) a _ partially order set _ ( poset ) , @xmath32 in @xmath29 or @xmath11 there always exist elements that are the greatest lower bound ( g.l.b . ) and the lowest upper bound ( l.u.b ) of two given elements with respect to the order relation given by @xmath33 .
in @xmath11 they are given by the set intersection and union while in @xmath34 by the set intersection and sum of linear subspaces .
we will denote generically the g.l.b . and l.u.b .
of two elements @xmath35 and @xmath36 as @xmath37 , and @xmath38 respectively , and call them the meet and the join of @xmath35 and @xmath36 .
the meet and the join are associative and symmetric .
a poset where the join and the meet of any two elements always exist is called a _
lattice_. a lattice is in addition _ complete _ if the meet and the join exist for arbitrary families of elements .
a lattice @xmath39 have a _ unit _
@xmath40 and a _ zero _ element @xmath41 if @xmath42 and @xmath43 for every element @xmath44 .
an element @xmath35 in a lattice has a _
@xmath36 if it is @xmath45 and @xmath46 .
the lattice is called _ orthocomplemented _ ( or _ ortholattice _ )
if there exist a unary operation @xmath47 that assigns to every element @xmath35 a complement @xmath48 , and in addition @xmath49 in an orthocomplemented lattice it can be deduced that @xmath50 and @xmath51 , what shows that the negation is a dual map with respect to the operations @xmath52 and @xmath53 .
both , @xmath11 and @xmath29 are complete and orthocomplemented . however , for the lattice @xmath11 holds the distributivity law @xmath54 while it is easy to see that it does not hold for @xmath29 .
an orthocomplemented distributive lattice is called a _
boolean algebra_. however , in the lattice @xmath29 holds a relaxation of the distributivity , called _
orthomodularity is equivalent to the simpler relation @xmath56 for finite dimensional hilbert spaces a stronger form holds , the _ modular _ law , @xmath57 whose definition , as the one of distributivity , makes no use of the orthocomplementation operation .
distributivity implies modularity and , for an orthocomplemented lattice , modularity implies orthomodularity .
we see that the orthomodular law ( [ ortho ] ) means that distributivity holds under special circumstances for the elements involved . in the lattice of closed linear subspaces of the hilbert space , @xmath29 , when using the representation of projection operators , these conditions can be rewritten as that the projector @xmath58 commutes with @xmath59 and @xmath60 . in a general orthocomplemented lattice two elements @xmath35 and
@xmath36 are said to _ commute _ if @xmath61 the relation of commutativity is symmetric in orthomodular lattices , and the boolean subalgebras are formed by mutually commuting elements .
we will need a few additional definitions . with two lattices @xmath62 and
@xmath63 we can construct the _ direct product _ of lattices @xmath64 defined in the cartesian product of sets , and where the operations are done component by component .
the product of orthomodular lattices is orthomodular .
a lattice is _ irreducible _ if it can not be written as a direct product of lattices .
an orthocomplemented lattice is irreducible if and only if its _ centre _ , that is , the set of elements that commute with all other elements , is equal to @xmath65 .
an _ atom _ of a lattice is a non zero element @xmath35 such that @xmath35 and @xmath41 are the only elements included in @xmath35 .
a lattice is _ atomic _ if every non zero element contains an atom .
an orthomodular atomic lattice is also _
, that is , every non zero element is the join of the atoms that contains .
both , @xmath66 and @xmath29 are atomic , its atoms being the points in @xmath12 and the unidimensional vector spaces in @xmath22 .
we will now describe a fundamental method for constructing complete orthocomplemented lattices that will be useful later @xcite .
the orthocomplemented lattice will be constructed with a subset ( not a sublattice ) of a given complete lattice @xmath39 ( not necessarily orthocomplemented ) .
the idea is to start with a would be complement operation @xmath67 and then to require the elements in the new lattice to be the @xmath44 that satisfy @xmath68 .
this is imposed by eq.([negacion ] ) .
this is similar to the case of a hilbert space , where , if we start with @xmath69 and propose the orthogonal to be the orthocomplement operation , the resulting lattice is @xmath29 , because the orthogonal of every set of vectors is a closed linear subspace . given a complete lattice @xmath39 a _ galois connection _ in @xmath39 is a unary operation @xmath70 in @xmath39 such that @xmath71 the operation @xmath72 is called a _ closure operation _
, and the elements @xmath35 of @xmath39 that satisfy @xmath73 are called _ closed _ with respect to the closure operation @xmath74 .
the closure @xmath75 has the following properties @xmath76 if we want the galois connection @xmath70 to be an orthocomplementation we must pick up the elements of the lattice such that the closure operation is the identity operator in them , as is enforced by eq.([negacion ] ) .
indeed , if we further have @xmath77 the set of closed elements @xmath78 of @xmath39 with respect to @xmath74 , forms a complete orthocomplemented lattice where the order and the meet are the order and the meet in @xmath39 , the complement is given by @xmath70 and the join is the join in @xmath39 followed by the operation of closure @xmath74 .
finally , we add that there is a very natural way to construct a galois connection in the lattice @xmath11 .
let @xmath79 be a symmetric reflexive relation between points in the set @xmath12 , with no further requirements .
thus @xmath80 and @xmath81 .
the operation @xmath82 given by @xmath83 that is , the set of all points not related by @xmath79 with no point of @xmath84 , is a galois connection that satisfies eqs.([g1 ] , [ g2 ] ) and ( [ x1 ] , [ x2 ] ) .
the space - time we will consider is a ( hausdorff ) manifold @xmath6 of dimension @xmath85 with a pseudometric tensor @xmath86 with signature @xmath87 , smooth and non singular everywhere . we do not ask the space - time @xmath6 to be time orientable nor to satisfy any causality condition ( for example to be free from closed time - like curves ) . for a review of causal structure see refs .
@xcite . given a definite naming of the local light cones of a point @xmath4 as future and past light cones , its causal future and past , @xmath88 and @xmath89 , are defined as the set of points that can be reached by time - like curves starting at @xmath4 with tangent vector in the future and past light cone of @xmath4 respectively .
when referring to time - like curves we always mean differentiable curves with continuous tangent vector except a finite number of points where the tangent can have a discontinuity , but keeping always in the same local light cone .
these discontinuities can always be removed by modifications in arbitrarily small neighborhoods .
the past and future of @xmath4 are open sets .
we will use the symbol @xmath90 for the open set @xmath91 that always makes sense , even in non time - orientable space - times .
the natural orhtocomplement operation for sets in @xmath6 based on the causal structure identifies all that is not time - like related to a given set with its complement . to be explicit
, we can define in @xmath6 a symmetric reflexive relation that relies only on the causal structure .
given two points @xmath4 and @xmath92 of @xmath6 we write @xmath93 , and say that @xmath4 and @xmath92 are time - connected or just connected , if there is time - like curve that passes through @xmath4 and @xmath92 .
thus , @xmath93 is equivalent to @xmath94 .
all points that are not time - connected to @xmath4 must be in the complement of @xmath4 .
starting from this complement operation , as it was described in the previous section , we can construct a complete orthocomplemented lattice coming from the galois connection in @xmath95 given by eq.([cual ] ) .
we call it the lattice of the causally closed sets @xmath7
. the orthocomplement is given by @xmath96 and , @xmath97 if and only if @xmath98 .
1 shows an example of @xmath99 and @xmath100 for a set @xmath101 . an important property to keep in mind
is that , for every set @xmath2 , the opposite @xmath99 is already an element of @xmath102 , because @xmath103 .
a typical element @xmath2 of @xmath104 in minkowski space is a diamond shaped set , with the upper and lower cone being null surfaces included in @xmath2 , while the points in the spatial corner may or may not be in @xmath2 .
however , in minkowski space many other sets , including lower dimensional objects , are in the lattice .
the bounded null surfaces of dimension @xmath105 or lower are sets in @xmath102 while space - like surfaces of dimension @xmath106 or less are in @xmath104 , while spatial surfaces of dimension @xmath105 are not . on the other hand
every set with at least two different points time - like connected will generate a set that contains an open set in @xmath6 .
the sets generated by subsets in achronal surfaces form boolean subalgebras .
the meet in @xmath7 is just the set intersection .
the join of a family of elements is given by the set union followed by the causal closure ( the double orthogonal ) @xmath107 the join of two sets is exemplified in fig .
2 . for two
spatially separated sets @xmath35 and @xmath36 the set @xmath108 will be just the set union @xmath109 , while otherwise @xmath38 will contain also at least all the time - like curves that connect points from @xmath35 to @xmath36 .
in general , for a set @xmath35 in @xmath6 , the causal closure @xmath110 is bigger than the domain of dependence of @xmath35 , taking this latter as the set of points such that all past ( future ) inextendible time - like curves from the point intersects the given set @xcite ( see fig .
the lattice @xmath7 is not distributive nor modular in the general case , as it is shown for minkowski space in fig .
3(a ) and fig.3(b ) , while it is orthomodular as will be shown in the next section and is illustrated in fig .
as mentioned in the introduction , the lattice @xmath102 has a logical interpretation in terms of propositions for classical particles .
the proposition corresponding to a space - time subset @xmath2 is given by `` the particle passes through s '' .
however , for an arbitrarily chosen set @xmath2 the logical opposite of that proposition is `` the particle does not passes through @xmath2 '' . in general
this is not of the form `` the particle passes through @xmath111 '' for some space - time subset @xmath111 .
the prescription of taking causally closed sets is just what is needed to have a closed algebra of propositions coming from space - time subsets in the above sense .
in contrast to the case of the boolean logic of space subsets , here only some subsets of space - time have an opposite in terms of space - time subsets , and the resulting algebra is non boolean .
the logical interpretation of the operation @xmath112 between two propositions corresponding to the sets @xmath35 and @xmath36 is given by the maximal space - time proposition that implies `` the particle passes through @xmath35 and @xmath36 '' , while the proposition corresponding to @xmath38 is given by the minimal space - time proposition implied by `` the particle passes through @xmath35 or @xmath36 '' .
it is immediate that when the manifold @xmath6 is non connected the lattice @xmath7 is the direct product of the lattices @xmath113 of each connected component @xmath114 of @xmath6 . in this case
the lattice is reducible .
we will see a sufficient condition for irreducibility in section v. one can try to construct other lattices based on the causal structure of @xmath115 . for example , the basic symmetric and reflexive relation between points used to define the orthocomplement can be taken as the relation @xmath116 , that holds if there is a space - like curve passing through @xmath4 and @xmath92 , or @xmath117 , that holds if there is a causal curve ( i.e. a curve with time - like or null tangent ) passing through @xmath4 and @xmath92 , or @xmath118 , that holds if there is null curve passing through @xmath4 and @xmath92 , or several logical combinations between them ( no discontinuities are allowed in the tangent to null and causal curves ) . accordingly , three orthocomplemented and complete lattices can be constructed , @xmath119 , @xmath120 and @xmath121 respectively .
the lattice @xmath119 is trivial when @xmath122 because a space - like curve from a point can get inside the light cone , while in two dimensions it does not add anything new to @xmath123 .
the same triviality will occur in @xmath7 if the metric would have more than one time directions .
it is easy to see that @xmath124 and @xmath125 fail to be orthomodular in the simplest case of minkowski space .
it seems that in the case of @xmath125 it is not possible to change slightly the definition of the lattice to make it orthomodular .
the lattice @xmath120 differs from @xmath8 fundamentally in details of the borders of the sets , what is crucial for orthomodularity .
other logical combinations seem to lead to the same lattices or to uninteresting and trivial cases .
one may wonder if there is another way for constructing a causal lattice that yielding open sets or sets with better topological properties .
a lattice of open sets can be constructed using the complete lattice @xmath126 of open sets in @xmath6 as a base where to define the galois connection @xcite .
the join in @xmath127 is the usual set union and the meet of arbitrary families of open sets is given by the set intersection followed by the operation of interior .
then we define @xmath128 for an open @xmath2 as @xmath129 where here @xmath130 means the usual topological closure of @xmath35 .
this satisfies eqs.([g1 ] , [ g2 ] ) and ( [ x1 ] , [ x2 ] ) , and lead to a complete orthocomplemented lattice @xmath131 of open sets . however , again due to details in the borders of the sets , the lattice is not orthomodular , as is shown in fig .
4 . it can be shown , at least for globally hyperbolic space - times , that distributivity in @xmath132 holds under a little more stringent conditions than the required by orthomodularity .
in fact , instead of eq.([or ] ) we have @xmath133 it seems essential for orthomodularity that the lattice should contain along with the diamond shaped sets at least the null surfaces .
once we have null surfaces in the lattice its intersection will generate lower dimensional sets as lines and points .
this makes difficult a correspondence between the elements in an orthomodular lattice of causal sets with @xmath9-algebras in the algebraic approach to quantum field theory , because the points should be assigned non trivial algebras ( two points can generate a set that includes an open set ) ( see the discussion in @xcite ) .
the lattice @xmath134 would be a better candidate , but it is only approximately orthomodular in the sense of eq.([apro ] ) ( weaker conditions for distributivity can also be found ) .
form now on we will only refer to @xmath7 . from its definition
we see that it does not change with conformal transformations , and thus it is a property of the conformal structure .
we will now prove that @xmath7 is orthomodular in any space - time .
first , we will prove the following previous result .
* lemma 1 * : given a time - like curve @xmath135 parametrized by a real variable @xmath136 in a closed interval @xmath137 $ ] and a set @xmath138 , the intersection @xmath139 ) \cap s$ ] is empty or it is a closed segment of @xmath140 , corresponding to @xmath141\subseteq [ a , b]$ ] .
suppose @xmath142 $ ] and @xmath143 $ ] , with @xmath144 are such @xmath145 and @xmath146 , then it is @xmath147)\subseteq s$ ] . otherwise ,
if @xmath148 $ ] is such that @xmath149 is not in @xmath2 , it can be connected by a time - like curve to @xmath128 ( if not it would be in @xmath150 ) , and thus either @xmath151 or @xmath152 is connected by a time - like curve to @xmath99 , contradicting the assumptions .
then , if the intersection @xmath139 ) \cap s$ ] is non empty , let @xmath74 and @xmath153 be the infimum and supremum of the points @xmath136 in @xmath154 $ ] such that @xmath155 .
the point @xmath156 is connected with @xmath2 so @xmath157 @xmath99 , then if @xmath158 it is connected also with @xmath99 , and thus belong to the open set @xmath159 .
a neighborhood of @xmath160 will belong to @xmath161 and will not be in @xmath162 , what is not possible , so @xmath163 .
similarly we have @xmath164 .
therefore , a set @xmath2 in @xmath7 contains all time - like segments between points in @xmath2 and also the time - like border of @xmath2 , that is , the border of @xmath2 accessible from @xmath2 by time - like curves .
the points in the space - like border of @xmath2 may or may not belong to @xmath2 . using lemma 1
we prove * theorem 1 : * @xmath7 is orthomodular orthomodularity is the implication @xmath165 for any @xmath35 and @xmath36 in @xmath7 .
if the set @xmath36 is included in @xmath35 it is also included in @xmath166
. then we have to show that @xmath167 .
to start we assume that @xmath168 .
let @xmath169 , so @xmath170 and @xmath171 . therefore no time - like curve connects @xmath4 with @xmath172 nor with @xmath173
we have to prove that @xmath174 .
then , let us suppose that @xmath175 and show that it leads to a contradiction .
this assumption implies that there is a time - like curve @xmath140 that connects @xmath4 with a point @xmath176 .
as @xmath4 is not connected with @xmath48 nor with @xmath177 it is @xmath178 and @xmath179 .
a segment of @xmath180 with end points @xmath4 and @xmath92 intersects @xmath181 and @xmath35 in non empty sets , which according to lemma 1 are closed segments .
these are disjoint , because otherwise @xmath4 could be connected with a point in @xmath173 .
thus , there is a point @xmath182 in the segment of @xmath140 between @xmath4 and @xmath92 that is not in @xmath183 nor in @xmath35 . as @xmath184
, @xmath182 can be connected with a point @xmath185 .
thus , as @xmath182 is in the segment of @xmath186 between @xmath4 and @xmath92 it is either @xmath92 connected with @xmath187 or @xmath4 connected with @xmath187 .
however , the point @xmath188 can not be connected with @xmath185 .
then @xmath170 is connected with @xmath189 by a curve passing through @xmath182 .
then , by lemma 1 , it is @xmath190 , what shows a contradiction . therefore @xmath174 and orthomodularity holds .
of course , there are orthomodular lattices that are not of the form @xmath191 for a space - time @xmath6 . in the following section we go a step further in the characterization of causal structures from the lattice theoretical point of view by showing that @xmath7 is atomic , and finding its atoms .
in the remaining part of this section we will extract what are the essential properties of the causal structure used in the proof of orthomodularity ( see ref.@xcite for a related discussion ) . for constructing an orthocomplemented lattice we used a symmetric reflexive relation between points in the space - time with no further properties .
now we can read from the previous theorem what additional structure of the causal relation @xmath192 that we used in the definition of @xmath7 is central in the proof of orthomodularity .
basically , what we need is a set of curves in a set @xmath193 , locally one to one functions of ( possible infinite ) open intervals of real numbers to @xmath6 , that we can call time - like curves . given a point
@xmath4 and any curve @xmath140 with @xmath194 , the points @xmath195 with @xmath196 belong to two sets , non necessarily disjoint , that we can call the past and future sets of @xmath197 .
we have either all @xmath149 for @xmath198 belong to the future of @xmath4 and all @xmath149 for @xmath199 belong to the past of @xmath4 or vice versa . if a point @xmath92 is in the future set of @xmath4 , then either all the future set of @xmath92 or the past set of @xmath92 are included in the future set of @xmath4 , and , in addition , there is a neighborhood of @xmath92 ( in the sense induced by the real numbers ) in all time - like curves passing through @xmath92 that belongs to the future set of @xmath4 .
the same can be said regarding the past set of @xmath4 .
these geometrical properties , that are immediate for the time - like curves in a general space - time @xmath6 , are sufficient in order to have an orthomodular lattice generated by the relation @xmath192 .
we see that certain sort of transitivity condition is essential for the causal relation but , for non orientable space - times no global transitive relation can be defined .
there is also a condition of continuity .
none of these is respected by the relation @xmath200 defined in the preceding section , while the relation @xmath201 does not respect the continuity condition .
an atom @xmath2 in @xmath202 must be equal to the set @xmath203 generated by any point @xmath204 , because @xmath205 is a non empty element of @xmath7 included in @xmath2 .
therefore , all atoms are of the form @xmath205 for some point @xmath4 , what shows what are the sets among which to look for atoms .
however , the element @xmath203 need not be an atom .
by contrast , if @xmath4 as a subset of @xmath115 is an element of @xmath202 , @xmath206 , then it is an atom .
this is the case for all points in minkowski space .
we will now analyze which of the points of @xmath6 are atoms in the general case . as we noted before ,
if two points @xmath4 and @xmath92 belong to @xmath2 in @xmath207 , then all points in time - like curves connecting @xmath4 and @xmath92 are also in @xmath2 .
if there is a time - like segment @xmath140 connecting a point @xmath4 with itself , then @xmath4 can not be an atom because @xmath208 .
there are two types of such curves , as shown in fig .
one , the _ closed time - like curves _ , that are totally included in the intersection @xmath209 , where a segment of the curve between @xmath4 and @xmath4 start and end with its tangent vector in the same light cone of @xmath4 .
the other type , that we will call _ vertex curves _
, where a segment between @xmath210 and @xmath4 start and end with tangent vectors at @xmath4 in different light cones of @xmath211 in this later case we will call @xmath4 a _
vertex_. for the existence of these types of curves the manifold must not be simply connected @xcite . in the vertex curve case
we see that the space - time is not time orientable . however , there are time non orientable space - times without vertex curves .
the vertex curves mark space - times where the notion of future and past have no sense for a single observer .
the points @xmath4 that belong to closed time - like curves or are vertex points can be characterized as all points that belong to @xmath212 for some @xmath92 of @xmath6 ( any @xmath92 of @xmath140 in the open segment between @xmath4 and @xmath4 for example )
. we will call it the set @xmath213 of bad points of @xmath6 , @xmath213 @xmath214 . as union of open sets , @xmath215 is also open .
the set of points in closed time - like curves @xmath216 , and the set of vertices @xmath217 , are also separately open , and by definition @xmath218 . to see this ,
let @xmath219 , where @xmath186 is a closed time - like curve , and take a sufficiently small , open , time orientable , normal neighborhood @xmath12 of @xmath4 @xcite .
let @xmath92 and @xmath182 be two points of an interval of @xmath186 included in @xmath12 that contains @xmath4 , respectively in the future @xmath220 and past @xmath221 of @xmath4 , where @xmath222 and @xmath223 are the notions of future and past in the submanifold @xmath12 . doing a composition with the curve @xmath186 , we have that all points in the open neighborhood @xmath224 of @xmath4 are connected by a closed time - like curve to every point in @xmath186 .
thus , the set of points connected by closed time - like curves to a point @xmath4 form an open subset @xmath225 of @xmath216 .
it is immediate that the sets of the form @xmath225 are either disjoint or identical , so @xmath216 is the union of disjoint sets of the form @xmath225 . note that for any time - like curve @xmath226 that passes through @xmath4 we have that there is an open interval around @xmath4 in @xmath226 that is included in a closed time - like curve trough @xmath4 .
the proof that @xmath217 is open can be done similarly .
the following lemma solves the question of when a point in @xmath6 is an element , and therefore an atom , in @xmath7 * lemma 2 : * the set formed by a point @xmath4 in @xmath6 is an element and an atom in @xmath7 if and only if @xmath227 as we have seen if @xmath228 the set formed by @xmath4 is not an element of @xmath7 and can not be an atom .
we will show that if @xmath4 is not an atom then @xmath228 . let us suppose @xmath229 .
thus , there is a point @xmath230 with @xmath231 .
the point @xmath92 must be time connected with @xmath4 otherwise it would be @xmath232 .
therefore @xmath233 and a connected , time orientable , open neighborhood @xmath12 of @xmath4 is included in @xmath234 .
thus , as every point in @xmath12 is connected with @xmath230 , it must be that @xmath235 is empty .
in other words @xmath236 is separated from @xmath4
. then all points @xmath182 in the open connected set @xmath237 are in @xmath88 or in @xmath89 . with a given continuous time orientation in @xmath12 we have for every @xmath238 that @xmath4 is in @xmath239 or in @xmath240 , and the sets @xmath241 and @xmath242 are open .
they cover all the connected set @xmath243 , and thus must have a non empty intersection , with a point @xmath187 .
therefore @xmath4 belongs to @xmath244 and thus @xmath228 .
we have shown that in the closed set @xmath245 all points are atoms .
next we will show that a different kind of atoms can exist in space - times that have both types of defects , closed time - like and vertex curves .
consider the time - like curve @xmath246 of fig .
. we will call such a curve a _ bridge_. it is a closed time - like curve but also a past and future vertex curve for all points in the curve .
the essential property of a bridge is that for any point @xmath4 in it one can construct a time - like curve that passes through @xmath4 with tangent in any light cone and passes through any point @xmath92 in the bridge with tangent in the any light cone of @xmath92 .
thus any two points time connected with two points in the bridge @xmath247 are also connected to each other .
then we have the following * lemma 3 : * the set @xmath248 generated by any bridge @xmath249 is an atom let @xmath250 ( see fig .
6 ) . if @xmath230 then @xmath92 is connected with @xmath4 .
the set @xmath251 is included in @xmath252 and if they are different it must be @xmath253 . in that case
@xmath4 must be connected with @xmath254 .
but that is not possible because as @xmath4 belongs to a bridge it would lead to @xmath92 connected with @xmath255 .
thus @xmath256 for every @xmath257 and @xmath205 is an atom that includes @xmath247 .
we will call the set @xmath258 generated by a bridge @xmath246 a _ bridge atom_. any bridge atom contains an open set .
we also have that two atoms can not intersect .
therefore in a paracompact manifold there can be at most a numerable amount of bridge atoms .
the bridge atoms must be totally included in the set of bad points @xmath213 .
we have shown what happens to the sets @xmath205 for @xmath259 or @xmath4 in @xmath248 where @xmath247 is a bridge . in both cases @xmath205 is an atom .
we will see that this is not always the case .
first consider the case where @xmath4 is a vertex .
we have * lemma 4 : * the set @xmath205 where @xmath4 is a vertex contains a point in @xmath245 or a bridge atom let @xmath260 , @xmath261 $ ] , and @xmath262 be a vertex curve of the vertex @xmath4 . let us suppose that every point in @xmath140 is in @xmath263 . for every value of the parameter
@xmath261 $ ] we define the future light cone of @xmath260 at @xmath264 as the light cone of @xmath260 given by the tangent vector of @xmath186 at @xmath264 .
thus , the direction of the tangent vector at the point @xmath265 will mark the future light cone of @xmath4 at @xmath266 , but it will be a different light cone of the point @xmath4 than the future light cone at @xmath267 . for any @xmath264 such that the point @xmath268 is a vertex , we can define it as a future or a past vertex according the vertex curve based at @xmath92 starts with tangent in the future or the past light cone of @xmath92 at @xmath264 .
let us call @xmath269 $ ] and @xmath270 $ ] the set of values of the parameter that have future and past vertices respectively .
these sets are non empty as @xmath271 has a future vertex and @xmath272 has a past vertex .
if @xmath273 has a past vertex and @xmath274 has a future vertex and @xmath275 , then both vertex curves together with @xmath276)$ ] will form a bridge .
let the infimum of @xmath277 be @xmath278 and the supremum of @xmath279 be @xmath280 , and assume @xmath281 .
the points in @xmath282)$ ] must belong to closed time - like curves , @xmath282)\subseteq { \cal c}$ ] ( the points @xmath280 and @xmath283 can not be vertices as @xmath217 is open ) .
but as @xmath216 is open there must be @xmath284 and @xmath285 with @xmath286 and such @xmath287)\subseteq { \cal c}$ ] .
for each @xmath288 $ ] let @xmath289 be an open interval of @xmath186 around the point @xmath149 that is included in a closed time - like curve through @xmath149 .
the sets @xmath290\cap \gamma ^{-1}(i_{\gamma ( y)})$ ] are open in the topology of the interval , and cover the compact @xmath290 $ ] . after extracting a finite covering and gluing together a finite number of closed curves , we have that @xmath284 and @xmath285 are connected by a closed time - like curve @xmath291 that includes @xmath287)$ ] .
then it is easy to see that the union of @xmath292)$ ] with the vertices at @xmath284 and @xmath285 is a bridge .
then let us consider now the case of points in closed time - like curves .
let @xmath4 be a point in @xmath216 and @xmath225 be the connected component of @xmath216 that includes @xmath4 .
it is @xmath293 .
if @xmath294 is a connected component of @xmath6 then @xmath294 will be an atom .
if this is not the case , let @xmath92 be a point in the border of @xmath295 .
the point @xmath92 can only be a vertex or a point in @xmath296 .
therefore if @xmath297 contains a point in the border of @xmath225 it will contain an atom .
but any point in the border of @xmath225 that can be reached by a time - like curve passing through a point in @xmath297 must belong to @xmath297 because of lemma 1 .
thus , if there are no point or bridge atoms in @xmath297 we have @xmath298 .
a point in the set @xmath299 can not be joined by a time - like curve to @xmath225 , and , as every point is in the future or the past of a different point , it is @xmath300 .
therefore @xmath225 and @xmath299 are open disjoint sets that cover @xmath6 , and @xmath294 is a connected component of @xmath6 . resuming
, we have shown * theorem 2 * : @xmath7 is an atomic lattice .
its atoms are all the points in @xmath245 , the bridge atoms , and the connected components of @xmath6 included in @xmath216 taking into account that @xmath7 is orthomodular we also have , * corollary : * @xmath7 is atomistic that is , every element is the join of the atoms that includes .
thus , from the algebraic point of view all the bad points of the space - time disappear from the algebra , except the discrete bridge atoms and connected components included in @xmath216 ( see fig
these later obviously commute with all @xmath7 and then belong to the centre .
we can see that the same is true for the bridge atoms as follows .
* lemma 5 : * the bridge atoms belong to the centre of @xmath7 if an atom does not belong to @xmath301 for a bridge @xmath247 it must be connected with @xmath247 . if the atom is a point @xmath4 then as @xmath4 is connected with @xmath247 it implies that @xmath4 is a vertex , what is not consistent with @xmath4 being an atom . then let the bridge @xmath302 be connected with @xmath247 , it is easy to see that both will generate the same set , @xmath303 ( they form a unique bridge curve ) . thus an atom is either equal to @xmath248 or is included in @xmath301 , and therefore commutes with @xmath248 .
as @xmath7 is atomistic , and every element is the join of its atoms , @xmath304 will commute with all the elements of the lattice @xcite .
we have shown that if the manifold is not connected or if it contains at least two atoms and a bridge then @xmath7 is reducible .
we will complete the following * theorem 3 : * @xmath7 is trivial ( equal to @xmath305 ) if and only if @xmath306 for a bridge @xmath247 or @xmath115 is connected and @xmath307 . if non trivial , @xmath8 is irreducible if and only if @xmath6 is connected and does not contain any bridge curves assume that @xmath6 does not have bridge curves , nor connected components includes in @xmath308 . then , its only atoms are the points in @xmath245
. we can identify all the sets in @xmath7 with the set of its atoms , @xmath309 .
let now @xmath2 be a proper element in @xmath310 .
if there is a point @xmath4 in @xmath311 it will be an atom that does not commute with @xmath2 ( see eq.([commu ] ) ) .
thus , if @xmath2 is in the centre of @xmath102 , it must be @xmath312 , and @xmath313 .
let @xmath92 be a point in @xmath6 , then , as the atoms that generate @xmath251 are in @xmath2 or in @xmath99 , we have @xmath314 , the last equality coming from the form of the join for orthogonal sets .
thus , @xmath92 belongs to @xmath315 , @xmath316 , @xmath317 .
then , all future and past of all points of @xmath2 must be in @xmath2 , otherwise @xmath2 would be connected with @xmath99 , and , as for a point @xmath182 @xmath318 there is always a different point @xmath187 in a neighborhood that is in @xmath319 , and so @xmath182 @xmath320 with @xmath321 .
therefore @xmath322 is an open set .
the same can be said for @xmath323 , @xmath324 .
therefore , @xmath325 is non connected .
the case where @xmath7 is nontrivial and irreducible is the most interesting for physical space - times . in that case
there is a non countably number of atoms formed by the points in @xmath326 , that can be consider the relevant space - time from the algebraic point of view .
we have already mentioned that the lattice @xmath29 of closed linear subspaces of the hilbert space is complete , atomic , irreducible and orthomodular .
it has a further property called the _ covering law_. an element @xmath35 in a lattice is said to _ cover _
@xmath36 if @xmath168 and if for every element @xmath201 such that @xmath327 it is @xmath328 or @xmath329 .
the covering law means that given an atom @xmath84 and an element @xmath35 with @xmath330 , then @xmath331 covers @xmath35 .
there exist reconstruction theorems stating that a complete , atomic , irreducible , orthomodular lattice , satisfying the covering law , can be represented as a lattice of closed linear subspaces of a vector space with a hermitian form @xcite . under minor assumptions regarding the field of scalars for the vector space
it is the lattice of a hilbert space .
thus , a lattice with this set of properties characterize almost uniquely hilbert spaces .
the lattice @xmath7 is also complete , atomic and orthomodular , and , in most of the physically interesting cases , irreducible .
however , the covering law does not apply . for example , the join of two points connected by a time - like curve in minkowski space does not cover any of them .
this is the essential difference with @xmath29 , that obstructs the immersion of the lattice in a linear structure @xcite .
it would be very interesting to explore what additional algebraic properties does the lattice @xmath7 have , and , if possible , to find a set of properties that characterize the lattices of causally closed sets in the sense of a reconstruction theorem as the already mentioned . such a possibility is suggested by the fact that the causal structure for stably causal space - times determines not only its conformal structure , but also the topological and differential structure of the manifold @xcite .
the next step in the construction of a dictionary between the causal structure and the algebraic setting is to translate into the lattice theoretical language the different causality conditions such as strong causality and stable causality .
taub - nut like space - times or the space - time of fig . 8
are not strongly causal and the structure of the resulting lattices is that of a _ horizontal sum _ of several lattices . given two lattices @xmath332 and @xmath333 , their horizontal sum is the lattice formed by the set union of @xmath332 and @xmath333 , where the unit and zero elements are identified ( see fig . 8) , so that the meet and join of an element of @xmath332 with an element of @xmath333 are @xmath41 and @xmath40 respectively .
as a different kind of example we note that the galilean space - time gives place to a causal lattice formed by the horizontal sum of the boolean algebras corresponding to the space at different times .
all this suggest that the appearance of horizontal sums is characteristic of non causally well behaved space - times .
we postpone the general analysis for a future paper @xcite .
another application of the lattice framework could be found in the study of the asymptotic infinity .
the spatial corner have a very simple expression in the algebraic context , being simply the image of @xmath7 under an endomorphism of orthomodular lattices @xcite .
the causal structure is more fundamental in a logical sense than other aspects of space - time such as the metric .
as we have already recalled , it also determines the topological and differential structure in well behaved space - times . in part motivated by this fact , there are in the literature approaches to quantum gravity that use a poset or lattice structure representing a causal order in space - time as a fundamental object .
one is based on an axiomatic generalization of the history approach to quantum mechanics to a situation where the time evolution is less rigid than in ordinary quantum theory @xcite .
another one , parts from discrete posets that would represent the causal structure at the planck scale , and looks at the possible dynamics and the large scale structure that emerges @xcite .
it would be very interesting to see if the discussion in this paper , also focused in a lattice structure coming from the causal order , could add to these various invertigations .
i would like to thank c. rovelli for useful discussions .
this work was supported by conicet , argentina . ,
p.mittelstaedt ( d. reidel publishing company , dordrecht , 1978 ) . , e.g. beltrametti and g. cassinelli ( addison wesley publishing company , reading , 1981 ) . w. cegla and a.z .
jadczyk , rep .
* 9 * , 377 ( 1976 ) .
w. cegla and a.z .
jadczyk , comm .
phys . * 57 * , 213 ( 1977 ) .
_ orthomodularity of causal logics _ , w. cegla and j. florek , preprint 471 , univ .
wroclaw pp .
262 , 265 ( 1979 ) . , r. haag ( springer - verlag , berlin , 1992 ) ( see specially the discussion in pages 143 - 147 ) . for a recent review and furhter references
see d. buchholz , _ algebraic quantum field theory : a status report _ , math - ph/0011044 .
g.birkhoff , _ lattice theory _
( american mathematical society colloquium publications , vol xxv , providence , 1967 ) . , g. kalmbach ( academic press , london , 1983 ) .
r.m.wald , _ general relativity _
( the university of chicago press , chicago , 1984 ) . ,
s.w . hawking and g.f.r .
ellis ( cambridge university press , cambridge , 1973 ) . ,
r.geroch and g.t .
horowitz ( in _ general relativity , an einstein centenary survey _ , eds .
s.w hawking and w. israel , cambridge university press , 1979 ) .
w. cegla and j. florek , _ causal logic with physical interpretation _ , univ .
wroclaw pp . 524
kronheimer and r. penrose , proc . cam .
. soc . * 63 * , 481 ( 1967 ) .
s.w . hawking , ph.d .
thesis , cambridge university ( 1966 ) .
h. casini , in preparation .
isham , j. math .
35 * , 2157 ( 1994 ) .
isham and j. butterfield , found .
of phys . *
30 * , 1707 ( 2000 ) .
l. bombelli , j. lee , d. meyer , and r.d .
sorkin , phys .
lett . * 59 * , 521 ( 1987 ) ; phys .
60 * , 656 ( 1988 ) .
see for a recent work d. dou , ph.d . thesis , trieste , sissa ( 1999 ) , gr - qc/0106024 , and references therein . | the causal structure of space - time offers a natural notion of an opposite or orthogonal in the logical sense , where the opposite of a set is formed by all points non time - like related with it .
we show that for a general space - time the algebra of subsets that arises from this negation operation is a complete orthomodular lattice , and thus has several of the properties characterizing the algebra of physical propositions in quantum mechanics .
we think this lattice could be used to investigate the causal structure in an algebraic context . as a first step in this direction
we show that the causal lattice is in addition atomic , find its atoms , and give necessary and sufficient conditions for ireducibility . |
many problems of interest in quantum computing can be reduced to an instance of the _ hidden subgroup problem _ ( hsp ) . we are given a group @xmath0 and a function @xmath1 with the promise that , for some subgroup @xmath2 , @xmath1 is invariant precisely under translation by @xmath3 : that is , @xmath1 is constant on the cosets of @xmath3 and takes distinct values on distinct cosets .
we then wish to determine the subgroup @xmath3 by querying @xmath1 .
for example , in simon s problem @xcite , @xmath4 and @xmath1 is an oracle such that , for some @xmath5 , @xmath6 for all @xmath7 ; in this case @xmath8 and we wish to identify @xmath5 . in shor s factoring algorithm @xcite @xmath0 is the group @xmath9 where @xmath10 is the number we wish to factor , @xmath11 for a random @xmath12 , and @xmath3 is the subgroup of @xmath9 whose index is the multiplicative order of @xmath13 . both simon s and shor s algorithms
use the following approach , referred to as the _ standard method _ or _
fourier sampling _ @xcite .
: : prepare two registers , the first in a uniform superposition over the elements of @xmath0 and the second with the value zero , yielding the state @xmath14 step 2 .
: : query ( or calculate ) the function @xmath1 defined on @xmath0 and xor it with the second register .
this entangles the two registers and results in the state @xmath15 step 3 .
: : measure the second register .
this puts the first register in a uniform superposition over one of @xmath1 s level sets , i.e. , one of the cosets of @xmath3 , and disentangles it from the second register .
if we observe the value @xmath16 , we have the state @xmath17 where @xmath18 alternately , we can view the first register as being in a mixed state with density matrix @xmath19 where the sum includes one representative @xmath20 for each of @xmath3 s cosets .
: : carry out the quantum fourier transform on @xmath21 and measure the result .
( note that in shor s algorithm , since @xmath22 is unknown , the fourier transform is performed over @xmath23 for some @xmath24 ; see @xcite or @xcite . ) in both simon s and shor s algorithms , the group @xmath0 is abelian ; it is not hard to see that , in this abelian case , a polynomial number ( i.e. , polynomial in @xmath25 ) of experiments of this type determine @xmath3 .
in essence , each experiment yields a random element of the dual space @xmath26 perpendicular to @xmath3 s characteristic function , and as soon as these elements span @xmath26 we are done .
while the _ nonabelian _ hidden subgroup problem appears to be much more difficult , it has very attractive applications .
in particular , solving the hsp for the symmetric group @xmath27 would provide an efficient quantum algorithm for the graph automorphism and graph isomorphism problems ( see e.g. jozsa @xcite for a review ) .
another important motivation is the relationship between the hsp over the dihedral group with hidden shift problems @xcite and cryptographically important cases of the shortest lattice vector problem @xcite .
so far , algorithms for the hsp are only known for a few families of nonabelian groups , including wreath products @xmath28 @xcite ; more generally , semidirect products @xmath29 where @xmath30 is of polynomial size , and groups whose commutator subgroup is of polynomial size @xcite ; `` smoothly solvable '' groups @xcite ; and some semidirect products of cyclic groups @xcite .
ettinger and hyer @xcite provided another type of result , by showing that fourier sampling can solve the hsp for the dihedral groups @xmath31 in an _ information - theoretic _ sense .
that is , a polynomial number of experiments gives enough information to reconstruct the subgroup , though it is unfortunately unknown how to determine @xmath3 from this information in polynomial time . to discuss fourier sampling for a nonabelian group @xmath0
, one needs to develop the fourier transform over @xmath0 which relies on the group s _ linear representations_. for abelian groups , the fourier basis functions are homomorphisms @xmath32 such as the familiar exponential function @xmath33 for the cyclic group @xmath34 . in the nonabelian case , there are not enough such homomorphisms to span the space of all @xmath35-valued functions on @xmath0 ; to complete the picture , one introduces _ representations _ of the group , namely homomorphisms @xmath36 where @xmath37 is the group of unitary matrices acting on some @xmath35-vector space @xmath38 of dimension @xmath39 .
it suffices to consider _ irreducible _ representations , namely those for which no nontrivial subspace of @xmath38 is fixed by the various operators @xmath40 .
once a basis for each irreducible @xmath41 is chosen , the matrix elements @xmath42 provide an orthogonal basis for the vector space of all @xmath35-valued functions on @xmath0 .
the quantum fourier transform then consists of transforming ( unit - length ) vectors in @xmath43}= \
{ \sum_{g \in g } \alpha_g { \left| g \right\rangle } \mid \alpha_g \in { \mathbb{c}}\}$ ] from the basis @xmath44 to the basis @xmath45 where @xmath41 is the name of an irreducible representation and @xmath46 index a row and column ( in a chosen basis for @xmath38 ) . indeed , this transformation can be carried out efficiently for a wide variety of groups @xcite .
note , however , that a nonabelian group @xmath0 does not distinguish any specific basis for its irreducible representations which necessitates a rather dramatic choice on the part of the transform designer .
indeed , careful basis selection appears to be critical for obtaining efficient fourier transforms for the groups mentioned above .
perhaps the most fundamental question concerning the hidden subgroup problem is whether there is always a basis for the representations of @xmath0 such that measuring in this basis ( in step 4 , above ) provides enough information to determine the subgroup @xmath3 .
this framework is known as _
strong fourier sampling_. in this article , we answer this question in the negative , showing that natural subgroups of @xmath27 can not be determined by this process ; in fact , we show this for an even more general model , where we perform an arbitrary positive operator - valued measurement ( povm ) on coset states @xmath47 .
we emphasize that the subgroups on which we focus are among the most important special cases of the hsp , as they are those to which graph isomorphism naturally reduces .
[ [ related - work . ] ] related work .
+ + + + + + + + + + + + + the terminology `` strong fourier sampling ''
@xcite was invented to distinguish this approach from the natural variant , called _ weak fourier sampling _
, where one only measures the name of the representation @xmath41 , and ignores the row and column information .
weak fourier sampling is basis - independent , making it attractive from the standpoint of analysis ; however , it can not distinguish conjugate subgroups from each other , and hallgren , russell and ta - shma @xcite showed that it can not distinguish the trivial subgroup from an order-2 subgroup consisting of @xmath48 disjoint transpositions .
specifically , they used character bounds to show that the probability distribution obtained on representation names for the trivial and order-2 subgroups are exponentially close in total variation distance : it requires an exponential number of such experiments to distinguish them .
kempe and shalev @xcite have generalized this result to other conjugacy classes , and conjecture that one can do no better than classical computation with this approach . in an effort to shed light on the power of strong fourier sampling , grigni , schulman , vazirani and vazirani @xcite showed that , for groups such as @xmath27 , measuring in a _ random _ basis yields an exponentially small amount of information .
this can be explained , roughly , by the fact that projecting a vector into a sufficiently high - dimensional random subspace results in tightly concentrated length .
on the other hand , moore , rockmore , russell and schulman @xcite showed that for the affine and @xmath49-hedral groups , measuring in a well - chosen basis can solve the hsp ( at least information - theoretically ) in cases where random bases can not .
[ [ our - contribution . ] ] our contribution .
+ + + + + + + + + + + + + + + + + in this paper we show that strong fourier sampling , in an arbitrary basis of the algorithm designer s choice , can not solve the hsp for @xmath27 .
as in @xcite we focus on order-2 subgroups of the form @xmath50 where @xmath51 is an involution consisting of @xmath48 disjoint transpositions ; we remark that if we fix two rigid connected graphs of size @xmath48 and consider permutations of their disjoint union , then the hidden subgroup is of this form if the graphs are isomorphic and trivial if they are not .
then we show that strong fourier sampling and more generally , arbitrary measurements of coset states cannot distinguish these subgroups from each other , or from the trivial subgroup , without an exponential number of experiments .
we remark that our results do not preclude the existence of an efficient quantum algorithm for the hsp on @xmath27 .
rather , they force us to consider _ multi - register _ algorithms ,
in which we prepare multiple coset states and subject them to entangled measurements .
ettinger , hyer and knill @xcite showed that the hsp on arbitrary groups can be solved information - theoretically with a polynomial number of registers , and the authors have shown how to carry out such a measurement in the fourier basis @xcite .
kuperberg @xcite devised a subexponential ( @xmath52 ) algorithm for the hsp on the dihedral group @xmath31 that works by performing entangled measurements on two registers at a time , bacon , childs , and van dam @xcite have determined the optimal multiregister measurement for the dihedral group , and the present authors have generalized this to all cases where the @xmath3 and @xmath0 form a gelfand pair @xcite . whether a similar approach can be taken to the symmetric group
is a major open question . in a companion paper @xcite
, the first two authors take a step towards answering this question by showing that if we perform arbitrary entangled measurements over pairs of registers , distinguishing @xmath53 from the trivial group in @xmath27 requires a superpolynomial number ( specifically , @xmath54 ) of experiments .
we briefly discuss the elements of the representation theory of finite groups .
our treatment is primarily for the purposes of setting down notation ; we refer the reader to @xcite for complete accounts .
let @xmath0 be a finite group .
a _ representation _
@xmath41 of @xmath0 is a homomorphism @xmath55 , where @xmath38 is a finite dimensional hilbert space and @xmath37 is the group of unitary operators on @xmath38 .
the _ dimension _ of @xmath41 , denoted @xmath39 , is the dimension of the vector space @xmath38 . by choosing a basis for @xmath38 ,
then , each @xmath40 is associated with a unitary matrix @xmath56 $ ] so that for every @xmath57 , @xmath58 = [ \rho(g ) ] \cdot [ \rho(h)]$ ] where @xmath59 denotes matrix multiplication . fixing a representation @xmath55
, we say that a subspace @xmath60 is _
( left)-invariant _ if @xmath61 for all @xmath62 ; observe that in this case the restriction @xmath63 , given by restricting each @xmath40 to @xmath64 , is also a representation .
when @xmath41 has no invariant subspaces other than the trivial space @xmath65 and @xmath38 , @xmath41 is said to be _
irreducible_. when @xmath41 is irreducible , _
schur s lemma _ asserts that the centralizer of the subgroup @xmath66that is , the set of @xmath67 such that @xmath68 for all @xmath62consists only of the scalar matrices @xmath69 .
we use this fact below . in the case
when @xmath41 is _ not _ irreducible , then , there is a nontrivial invariant subspace @xmath70 and , as the inner product @xmath71 is invariant under the unitary maps @xmath40 , it is immediate that the dual subspace @xmath72 is also invariant
. associated with the decomposition @xmath73 is the natural decomposition of the operators @xmath74 . by repeating this process
, any representation @xmath75 may be decomposed into a direct sum of irreducible representations : we write @xmath76 and , for the @xmath77 appearing at least once in this decomposition , @xmath78 . in general , a given @xmath79 can appear multiply in this decomposition , in the sense that @xmath41 may have an invariant subspace isomorphic to the direct sum of @xmath80 copies of @xmath79 . in this case
@xmath80 is called the _ multiplicity _ of @xmath77 in @xmath41 , and we write @xmath81 where @xmath82 .
if two representations @xmath41 and @xmath79 are the same up to a unitary change of basis , we say that they are _ equivalent_. it is a fact that any finite group @xmath0 has a finite number of distinct irreducible representations up to equivalence and , for a group @xmath0 , we let @xmath83 denote a set of representations containing exactly one from each equivalence class .
the irreducible representations of @xmath0 give rise to the fourier transform .
specifically , for a function @xmath84 and an element @xmath85 , define the _ fourier transform of @xmath1 at @xmath41 _ to be @xmath86 the leading coefficients are chosen to make the transform unitary , so that it preserves inner products : @xmath87 there is a natural product operation on representations : if @xmath88 and @xmath89 are representations of @xmath0 , we may define a new representation @xmath90 by extending the rule @xmath91 .
in general , the representation @xmath92 is not irreducible , even when both @xmath41 and @xmath79 are .
this leads to the _ clebsch - gordan problem _ , that of decomposing @xmath92 into irreducibles . for a representation @xmath41
we define the _ character _ of @xmath41 , denoted @xmath93 , to be the function @xmath94 given by @xmath95 . as the trace of a linear operator is invariant under conjugation ,
characters are constant on the conjugacy classes of @xmath0 ; for a conjugacy class @xmath96 , we define @xmath97 .
characters are a powerful tool for reasoning about the decomposition of reducible representations .
in particular , when @xmath98 we have @xmath99 and , moreover , for @xmath100 , we have the orthogonality conditions @xmath101 then for an irreducible representation @xmath41 and representation @xmath79 , @xmath102 is equal to the multiplicity with which @xmath41 appears in the decomposition of @xmath79 .
for example , since @xmath103 , the multiplicity of @xmath104 in @xmath92 is @xmath105 in general , we can consider subspaces of @xmath43}$ ] that are invariant under left multiplication , right multiplication , or both ; these subspaces are called _ left- _ , _
right- _ , or _
bi - invariant _ respectively .
each @xmath85 corresponds to a @xmath106-dimensional bi - invariant subspace of @xmath43}$ ] , which can be broken up further into @xmath39 @xmath39-dimensional left - invariant subspaces , or ( transversely ) @xmath39 @xmath39-dimensional right - invariant subspaces .
however , this decomposition is not unique .
if @xmath41 acts on a vector space @xmath38 , then choosing an orthonormal basis for @xmath38 allows us to view @xmath40 as a @xmath107 matrix .
then @xmath41 acts on the @xmath106-dimensional space of such matrices by left or right multiplication , and the columns and rows correspond to left- and right - invariant spaces respectively .
more generally , each left - invariant subspace isomorphic to @xmath41 corresponds to a unit vector @xmath108 .
in this section we show that starting with a random coset state , the optimal one - register measurement for the hidden subgroup problem is precisely an instance of strong fourier sampling ( possibly in an over - complete basis ) .
this has been pointed out several times in the past , at varying levels of explicitness @xcite ; we state it here for completeness .
everything we say in this section is true for the hidden subgroup problem in general .
however , for simplicity we focus on the special case of the hidden subgroup problem called the _ hidden conjugate problem _ in @xcite : there is a ( non - normal ) subgroup @xmath3 , and we are promised that the hidden subgroup is one of its conjugates , @xmath109 for some @xmath62 .
we may treat the states arising after step 3 of the procedure above as elements of the group algebra @xmath43}$ ] .
we use the notation @xmath110}$ ] so that the vectors @xmath111 form an orthonormal basis for @xmath43}$ ] . given a set @xmath112
, @xmath113 denotes a uniform superposition over the elements of @xmath114 , @xmath115 .
the most general type of measurement allowed in quantum mechanics is a _ positive operator - valued measurement _ ( povm ) .
a povm with a set of possible outcomes @xmath116 consists of a set of positive operators @xmath117 subject to the completeness condition , @xmath118 since positive operators are self - adjoint , they can be orthogonally diagonalized , and since their eigenvalues are positive , they may be written as a positive linear combination of projection operators ( see e.g. @xcite ) .
any povm may thus be refined so that each @xmath119 where @xmath120 is a projection operator and @xmath121 is positive and real .
the result of this measurement on the state @xmath122 is a random variable , taking values in @xmath116 , that is equal to @xmath123 with probability @xmath124 .
note that outcomes @xmath125 need not correspond to subgroups directly ; the algorithm designer is free to carry out a polynomial number @xmath126 of experiments , observing outcomes @xmath127 , and then apply some additional computation to find the most likely subgroup given these observations .
if @xmath128 is chosen from @xmath0 uniformly so that the hidden subgroup is a uniformly random conjugate of @xmath3 , we wish to find a povm that maximizes the probability of correctly identifying @xmath128 from the coset state @xmath129 .
( of course , to identify a conjugate @xmath130 , we only need to specify @xmath128 up to an element of the normalizer of @xmath3 . ) since a random left coset of @xmath130 can be written @xmath131 for a random @xmath132 , the probability we observe outcome @xmath125 is @xmath133 ip @xcite observed that in the special case that each outcome @xmath125 corresponds to a subgroup , maximizing the probability that @xmath125 is correct subject to the constraint gives a semi - definite program . since
such programs are convex , the optimum is unique and is a fixed point of any symmetries possessed by the problem .
however , our proof relies on an elementary `` symmetrization '' argument . given a group element @xmath134 ,
let @xmath135 denote the unitary matrix corresponding to left group multiplication by @xmath7 .
in particular , applying @xmath136 maps one left coset onto another : @xmath137 . writing @xmath138 we conclude that replacing @xmath120 for each @xmath125 with the symmetrization @xmath139 does not change the resulting probability distribution @xmath140 .
since @xmath141 commutes with @xmath136 for every @xmath134 and provides exactly the same information as the original @xmath120 , we may assume without loss of generality that the optimal povm commutes with @xmath136 for every @xmath134 .
it is easy to see that any projection operator that commutes with left multiplication projects onto a left - invariant subspace of @xmath43}$ ] , and we can further refine the povm so that each @xmath120 projects onto an
_ irreducible _ left - invariant subspace .
each such space is contained in the bi - invariant subspace corresponding to some irreducible representation @xmath41 , in which case we write @xmath142 . as discussed in section [ sec : representation - theory ]
, a given irreducible left - invariant subspace corresponds to some unit vector @xmath143 in the vector space @xmath38 on which @xmath41 acts .
thus we can write @xmath144 where @xmath145 acts within the left - invariant subspace . let @xmath146 ; then implies a completeness condition for each @xmath85 , @xmath147 and so @xmath148 is a ( possibly over - complete ) basis for @xmath38 . in other words ,
the optimal povm consists of first measuring the representation name @xmath41 , and then performing a povm on the vector space @xmath38 with possible outcomes @xmath148 .
another way to see this is to regard the choice of coset as a mixed state ; then its density matrix is block - diagonal in the fourier basis , and so as kuperberg puts it @xcite measuring the representation name `` sacrifices no entropy . ''
we note that in the special case that this povm is a von neumann measurement that is , when the @xmath148 is an orthonormal basis for @xmath38then it corresponds to measuring the column of @xmath41 in that basis , which is how strong fourier sampling is usually defined .
( as pointed out in @xcite , nothing is gained by measuring the row , since we have a random left coset @xmath149 and left - multiplying by a random element @xmath20 in an irreducible representation completely mixes the probability across the rows in each column . here
this is reflected by the fact that @xmath120 is a scalar in each left - invariant subspace . )
however , in general the optimal measurement might consist of an over - complete basis , or _ frame _
, in each @xmath41 , consisting of the vectors @xmath150 and the weights @xmath121 .
now that we know @xmath120 takes this form , let us change notation .
given @xmath85 acting on a vector space @xmath38 and a unit vector @xmath108 , let @xmath151 denote the projection operator onto the left - invariant subspace corresponding to @xmath143 .
then @xmath152 , and becomes @xmath153 we can write this as the product of the probability @xmath154 that we observe @xmath41 , times the conditional probability @xmath155 that we observe @xmath156 . note that by , @xmath157 is the projection operator onto the bi - invariant subspace corresponding to @xmath41
. then @xmath158 where @xmath159 note that @xmath155 depends on @xmath128 but @xmath154 does not , which is why weak sampling is incapable of distinguishing conjugate subgroups .
now let us use the fact that @xmath160 is a superposition over a subgroup , and calculate @xmath154 and @xmath155 as defined in and .
this will set the stage for asking whether we can distinguish different conjugates of @xmath3 from each other or even the trivial subgroup .
fix an irreducible representation @xmath41 that acts on a vector space @xmath38 .
then fourier transforming the state @xmath161 yields the coefficient @xmath162 where @xmath163 is a projection operator onto a subspace of @xmath38 .
the probability that we observe @xmath41 is then the norm squared of this coefficient , @xmath164 and , as stated above , this is the same for all conjugates @xmath130 .
the conditional probability that we observe the vector @xmath156 , given that we observe @xmath41 , is then @xmath165 in the case where @xmath3 is the trivial subgroup , @xmath166 and @xmath155 is given by @xmath167 we call this the _ natural distribution _ on the frame @xmath168 . in the case that @xmath148 is an orthonormal basis , @xmath169 and this is simply the uniform distribution .
this probability distribution over @xmath148 changes for a conjugate @xmath130 in the following way .
again ignoring left multiplication since the columns are left - invariant , the fourier transform becomes @xmath170 and we have @xmath171 where we write @xmath172 for @xmath173 .
it is not hard to show that , for any @xmath108 , the _ expected _ value of @xmath174 , over the choice of @xmath62 , is @xmath175 .
our primary technical contribution is a method for establishing concentration results for this random variable .
in this section we focus on the case where @xmath53 for an element @xmath51 chosen uniformly at random from a conjugacy class @xmath176 $ ] of involutions .
( observe that order is preserved under conjugation so that if @xmath51 is an involution , then so are all elements of @xmath176 $ ] . )
given an irreducible representation @xmath55 and a vector @xmath108 , we bound the variance , over the choice of @xmath51 , of the probability @xmath177 that @xmath143 is observed given that we observed @xmath41 .
our key insight is that this variance depends on how the tensor product representation @xmath178 decomposes into irreducible representations @xmath79 , and how the vector @xmath179 projects into the constituent orthogonal irreducibles .
recall that , if a representation @xmath41 is reducible , it can be written as an orthogonal direct sum of irreducibles @xmath180 where @xmath80 is the multiplicity of @xmath79 .
we let @xmath181 denote the projection operator whose image is @xmath182 , that is , the span of all the irreducible subspaces isomorphic to @xmath79 .
[ lem : exp ] let @xmath41 be a representation of a group @xmath0 acting on a space @xmath38 and let @xmath108 .
let @xmath51 be an element chosen uniformly from a conjugacy class @xmath176 $ ] of involutions .
if @xmath41 is irreducible , then @xmath183)}{\dim \rho } { \left\| { { \mathbf{b}}}\right\|}^2 \enspace .\ ] ] if @xmath41 is reducible , then @xmath184)}{\dim \sigma } { \left\| \pi^\rho_\sigma { { \mathbf{b}}}\right\|}^2 \enspace .\ ] ] fix a particular element @xmath185 $ ] .
let @xmath186)$ ] denote the average of @xmath187 over @xmath176 $ ]
; this is @xmath188 ) = \frac{1}{|[m]| } \sum_{m \in [ m ] } \rho(m ) = \frac{1}{|g| } \sum_{g \in g } \rho(g^{-1 } \mu g ) = \frac{1}{|g| } \sum_{g \in g } \rho(g)^\dagger
\rho(\mu ) \,\rho(g ) \enspace .\ ] ] observe that @xmath186)$ ] commutes with @xmath40 for all @xmath62 and hence , by schur s lemma , its action on any irreducible subspace is multiplication by a scalar . note that for the scalar linear operator @xmath189 acting on a space of dimension @xmath190 , we have @xmath191
. in particular , if @xmath41 is irreducible then @xmath188 ) = \frac{\chi_\rho([m])}{\dim \rho } \,{\varmathbb 1}_{d_\rho}\ ] ] and so @xmath192 ) { { \mathbf{b}}}\rangle = \frac{\chi_\rho([m])}{\dim \rho } { \left\| { { \mathbf{b}}}\right\|}^2 \enspace .\ ] ] if @xmath41 is reducible , these same considerations apply to each irreducible subspace : @xmath188 ) = \sum_{\sigma \prec \rho } \frac{\chi_\sigma([m])}{\dim \sigma } \pi^\rho_\sigma\ ] ] and so @xmath192 ) { { \mathbf{b}}}\rangle = \sum_{\sigma \prec \rho } \frac{\chi_\sigma([m])}{\dim \sigma } \langle { { \mathbf{b } } } , \pi^\rho_\sigma { { \mathbf{b}}}\rangle = \sum_{\sigma \prec \rho } \frac{\chi_\sigma([m])}{\dim \sigma } { \left\| \pi^\rho_\sigma { { \mathbf{b}}}\right\|}^2 \enspace .\ ] ] turning now to the second moment of @xmath193 , we observe that @xmath194 where the action of @xmath51 on the vector @xmath179 is precisely given by the action of @xmath0 in the representations @xmath178 .
this will allow us to express the second moment of the inner product @xmath195 in terms of the projections of @xmath179 into the irreducible constituents of the tensor product representation @xmath196 .
[ lem : second ] let @xmath41 be a representation of a group @xmath0 acting on a space @xmath38 and let @xmath108 .
let @xmath51 be an element chosen uniformly at random from a conjugacy class @xmath176 $ ] of involutions
. then @xmath197)}{\dim \sigma } { \left\| \pi^{\rho \otimes \rho^*}_\sigma ( { { \mathbf{b}}}\otimes { { \mathbf{b}}}^ * ) \right\|}^2 \enspace .\ ] ] we write the second moment as a first moment over the product representation @xmath178 : as above , @xmath198 , so that @xmath199 and applying lemma [ lem : exp ] completes the proof .
now let @xmath200 denote the projection operator given by @xmath201 for a given vector @xmath202 , we will focus on the expectation and variance of @xmath203 .
these are given by the following lemma .
[ lem : var ] let @xmath41 be an irreducible representation acting on a space @xmath38 and let @xmath108 .
let @xmath51 be an element chosen uniformly at random from a conjugacy class @xmath176 $ ] of involutions .
then @xmath204)}{\dim \rho } \right ) \label{eq : exp } \\
\operatorname{var}_{m } { \left\| \pi_m { { \mathbf{b}}}\right\|}^2 & \leq & \frac{1}{4 } \sum_{\sigma \prec \rho \otimes \rho^ * } \frac{\chi_\sigma([m])}{\dim \sigma } { \left\| \pi^{\rho \otimes \rho^*}_\sigma({{\mathbf{b}}}\otimes { { \mathbf{b}}}^ * ) \right\|}^2 \enspace .
\label{eq : var } \end{aligned}\ ] ] for the expectation , @xmath205)}{\dim \rho } \right ) \end{aligned}\ ] ] where the last equality follows from lemma [ lem : exp ] . for the variance
, we first calculate the second moment , @xmath206)}{\dim \rho } + \sum_{\sigma \prec \rho \otimes \rho^ * } \frac{\chi_\sigma([m])}{\dim \sigma } { \left\| \pi^{\rho \otimes \rho^*}_\sigma({{\mathbf{b}}}\otimes { { \mathbf{b}}}^ * ) \right\|}^2 \right)\end{aligned}\ ] ] where in the last line we applied lemmas [ lem : exp ] and [ lem : second ] and the fact that any character evaluated at an involution is real
. then @xmath207)}{\dim \rho } { \left\| \pi^{\rho \otimes \rho^*}_\sigma ( { { \mathbf{b}}}\otimes { { \mathbf{b}}}^ * ) \right\|}^2 - { \left\| { { \mathbf{b}}}\right\|}^4 \left(\frac{\chi_\rho([m])}{\dim \rho } \right)^{\!2\ , } \right ] \label{eq : var1 } \enspace .\end{aligned}\ ] ] ignoring the second term , which is negative , gives the stated result .
finally , we point out that since @xmath208 we have @xmath209)}{\dim \rho } \right ) \enspace , \ ] ] a fact which we will use below .
in this section we record the particular properties of @xmath27 and its representation theory applied in the proofs of the main results .
the irreducible representations of @xmath27 are labeled by young diagrams , or equivalently by integer partitions of @xmath10 , @xmath210 where @xmath211 and @xmath212 for all @xmath213 .
the _ conjugate _ young diagram @xmath214 is obtained by flipping @xmath215 about the diagonal : @xmath216 where @xmath217 .
in particular , @xmath218 . the number of such diagrams , equal to the number of conjugacy classes in @xmath27 , is the partition number @xmath219 , which obeys @xmath220 we denote these irreducibles @xmath221 , their characters @xmath222 , and their dimensions @xmath223 . in particular
, @xmath224 is the trivial or parity representation if @xmath215 is a single row @xmath225 or a single column @xmath226 respectively , and @xmath227 is the ( tensor ) product of @xmath221 with the parity representation .
the dimensions of the representations of the symmetric group are given by the remarkable _ hook length formula _
: @xmath228 where this product runs over all cells of the young diagram associated with @xmath215 and @xmath229 is the number of cells appearing in either the same column or row as @xmath20 , excluding those that are above or to the left of @xmath20 .
for example , the partition @xmath230 is associated with the diagram shown in figure [ fig : young ] below .
the hook associated with the cell @xmath231 in this diagram appears in figure [ fig : hook ] ; it has length 6 . a hook of length 6 . ] a hook of length 6 . ]
the symmetric groups have the property that every representation @xmath221 possesses a basis in which its matrix elements are real , and so all its characters are real . however , in a given basis @xmath221 might be complex , so we will refer below to the complex conjugate representation @xmath232 ( not to be confused with @xmath227 ) .
the study of the asymptotic properties of the representations of @xmath27 typically focuses on the _ plancherel _ distribution .
( see , e.g. , kerov s monograph @xcite for further discussion . ) for a general group @xmath0 , this is the probability distribution obtained on @xmath83 by assigning @xmath41 the probability density latexmath:[$d_\rho^2
/ @xmath41 is proportional to its contribution , dimensionwise , to the group algebra @xmath43}$ ] ; this will allow us to reduce structural questions about representations chosen according to this distribution to questions about the regular and conjugation representation .
note that in the context of the hidden subgroup problem , the plancherel distribution is exactly the one obtained by performing weak fourier sampling on the trivial hidden subgroup . in the symmetric groups a fair amount is known about representations chosen according to the plancherel distribution .
in particular , vershik and kerov have shown that with high probability they have dimension equal to @xmath234 .
[ thm : plancherel ] let @xmath221 be chosen from @xmath235 according to the plancherel distribution . then there exist positive constants @xmath236 and @xmath237 for which @xmath238 = 1\enspace.\ ] ] vershik and kerov have also obtained estimates for the _ maximum _ dimension of a representation in @xmath235 : [ thm : max - dimension ] there exist positive constants @xmath239 and @xmath240 such that for all @xmath241 , @xmath242 along with these estimates , we shall require some ( one - sided ) large - deviation versions of theorem [ thm : plancherel ] , recorded below . [
lem : lower ] let @xmath221 be chosen according to the plancherel distribution on @xmath235 . 1 .
[ item : lower - sqrt ] let @xmath243 .
then for sufficiently large @xmath10 ,
@xmath244 < e^{-\delta\sqrt{n}}$ ] .
[ item : lower - linear ] let @xmath245
. then @xmath246 = n^{-\omega(n)}$ ] .
for the first bound , setting @xmath247 and using @xmath248 , we have @xmath249 for the second bound , recalling sterling s approximation @xmath250 , we have @xmath251 finally , we will also apply roichman s @xcite estimates for the characters of the symmetric group : for a permutation @xmath252 , define the _ support _ of @xmath253 , denoted @xmath254 , to be the cardinality of the set @xmath255 \mid \pi(k ) \neq k \}$ ] .
[ thm : roichman ] there exist constants @xmath256 and @xmath257 so that for @xmath258 , for every conjugacy class @xmath259 of @xmath27 , and every irreducible representation @xmath221 of @xmath27 , @xmath260 where @xmath261 for any @xmath262 . in our application , we take @xmath10 to be even and consider involutions @xmath51 in the conjugacy class of elements consisting of @xmath48 disjoint transpositions , @xmath263 .
note that each @xmath264 is associated with one of the @xmath265 perfect matchings of @xmath10 things , and that @xmath266 .
we consider the hidden subgroup @xmath267 , where @xmath51 is chosen uniformly from @xmath268 , the conjugacy class @xmath269 we assume throughout that @xmath10 is even .
we start by measuring the name of an irreducible representation , yielding @xmath221 for a diagram @xmath215 .
we remark that hallgren , russell , and ta - shma @xcite established that the probability distribution on @xmath215 is exponentially close to the plancherel distribution in total variation .
we allow the algorithm designer to choose an arbitrary povm , with a frame @xmath270 and weights @xmath271 obeying the completeness condition .
we will show that with high probability ( over @xmath51 and @xmath215 ) , the conditional distribution induced on the vectors @xmath148 is exponentially close to the natural distribution on @xmath148 .
it will follow by the triangle inequality that it requires an exponential number of single - register experiments to distinguish two involutions from each other or , in fact , distinguish @xmath3 from the trivial subgroup . for simplicity , and to illustrate our techniques ,
we first prove this for a von neumann measurement , i.e. , where @xmath148 is an orthonormal basis for @xmath221 . in this case , we show that the probability distribution on @xmath148 ( or equivalently , on the columns of @xmath221 ) is exponentially close to the uniform distribution .
[ thm : one ] let @xmath272 be an orthonormal basis for an irreducible representation @xmath221 . given the hidden subgroup @xmath53 where @xmath51 is chosen uniformly at random from @xmath273 , let @xmath274 be the probability that we observe the vector @xmath143 conditioned on having observed the representation name @xmath221 , and let @xmath275 be the uniform distribution on @xmath148 .
then there is a constant @xmath276 such that for sufficiently large @xmath10 , with probability at least @xmath277 in @xmath51 and @xmath215 , we have @xmath278 first , recall from in section [ sec : optimal ] that the conditional distribution on @xmath148 is given by ( since @xmath169 ) @xmath279 our strategy will be to bound @xmath280 using lemma [ lem : var ] , and apply chebyshev s inequality to conclude that it is almost certainly close to its expectation .
recall , however , that our bounds on the variance of @xmath203 depend on the decomposition of @xmath281 is into irreducibles and , furthermore , on the projection of @xmath179 into these irreducible subspaces .
matters are somewhat complicated by the fact that certain @xmath282 appearing in @xmath281 may contribute more to the variance than others . while theorem [ thm : roichman ] allows us to bound the contribution of those constituent irreducible representations @xmath282 for which @xmath283 and @xmath284 are much smaller than @xmath10 , those which violate this condition could conceivably contribute large terms to the variance estimates .
fortunately , we will see that the total fraction of the space @xmath281 , dimensionwise , consisting of such @xmath282 is small with overwhelming probability .
despite this , we can not preclude the possibility that for a _ specific _ vector @xmath143 , the quantity @xmath285 is large , as @xmath143 may project solely into spaces of the type described above . on the other hand , as these troublesome spaces amount to a small fraction of @xmath281 , only a few @xmath143 can have this property , and we will see that this suffices to control the distance in total variation to the uniform distribution .
specifically , let @xmath286 be a constant , and let @xmath287 denote the collection of young diagrams @xmath288 with the property that either @xmath289 or @xmath290 .
we have the following upper bounds on the cardinality of @xmath291 and the dimension of any @xmath282 with @xmath292 : [ lem : lambda ] let @xmath219 denote the number of integer partitions of @xmath10
. then @xmath293 , and @xmath294 for any @xmath292 .
for the first statement , note that removing the top row of a young diagram @xmath288 with @xmath295 gives a young diagram of size @xmath296 .
the number of these is at most @xmath297 , and summing over all such @xmath283 gives @xmath298 .
the case @xmath299 is similar , and summing the two gives @xmath293 .
now let @xmath292 with @xmath295 .
by the hook - length formula , since the @xmath213th cell from the right in the top row has @xmath300 , @xmath301 .
the case @xmath299 is similar . as a result ,
the representations associated with diagrams in @xmath291 constitute a negligible fraction of @xmath235 ; specifically , from lemma [ lem : lower ] , part [ item : lower - linear ] , the probability that a @xmath215 drawn according to the plancherel distribution falls into @xmath291 is @xmath302 .
the following lemma shows that this is also true for the distribution @xmath154 induced on @xmath235 by weak fourier sampling the coset state @xmath160 .
[ lem : notinlambda ] let @xmath303 be a constant and let @xmath10 be sufficiently large .
then there is a constant @xmath304 such that we observe a representation @xmath221 with @xmath305 with probability at least @xmath306 .
the proof is nearly identical to that of lemma [ lem : lower ] , and follows an argument from @xcite .
recall from in section [ sec : optimal ] that we observe an irreducible @xmath41 with probability @xmath307 . given @xmath308 , @xmath309 , and @xmath310
, we have @xmath311 .
since the total number of irreducibles of @xmath27 is @xmath219 and @xmath312 , we have @xmath313 and setting @xmath314 completes the proof .
on the other hand , for a representation @xmath282 with @xmath315 , theorem [ thm : roichman ] implies that @xmath316 for a constant @xmath317 .
thus the contribution of such an irreducible to the variance estimate of lemma [ lem : var ] is exponentially small .
in addition , let @xmath318 be the event that @xmath319 as in lemma [ lem : notinlambda ] ; then conditioning on @xmath318 and setting @xmath190 such that @xmath320 , lemma [ lem : lambda ] implies that @xmath321 , and equations and imply @xmath322 we turn now to the problem of bounding the multiplicities with which representations @xmath282 , for @xmath292 , can appear in @xmath323 . while no explicit decomposition is known for @xmath323 , the _ endomorphism representations _ of @xmath27 , we record a coarse bound below which will suffice for our purposes .
recall that character of @xmath281 is @xmath324 as characters of @xmath27 are real .
the multiplicity of the representation @xmath282 in @xmath281 is @xmath325 .
however , this is equal to @xmath326 , the multiplicity of @xmath221 in the representation @xmath327 .
counting dimensions , this is clearly no more than @xmath328 .
hence the multiplicity of @xmath282 in @xmath281 is never more than @xmath329 ; we have @xmath330 let @xmath331 be the subspace consisting of copies of representations @xmath282 with @xmath292 , and let @xmath332 be the projection operator onto this subspace . by lemma [ lem : lambda ] , we have @xmath333 note by lemma [ lem : notinlambda ] with @xmath334 , @xmath335 is a vanishingly small fraction of @xmath223 . as @xmath148 is an orthonormal basis for @xmath221 , the vectors @xmath336 are mutually orthogonal in @xmath281
therefore , @xmath337 in particular , if @xmath338 denotes the set of @xmath143 with the property that @xmath339 then , conditioning on @xmath318 , @xmath340 can not be larger than @xmath341 .
for a basis vector @xmath342 , we apply lemma [ lem : var ] to bound the variance as follows : assuming pessimistically that @xmath343 for all @xmath292 , and applying , gives @xmath344 \\ & \leq & \frac{1}{4 } \left [ e^{-\alpha n }
+ e^{-\alpha n } \sum_{s^\mu : \mu \notin \lambda } { \left\| \pi^\lambda_\mu({{\mathbf{b}}}\otimes { { \mathbf{b}}}^ * ) \right\|}^2 \right ] \;\ ; \leq \;\ ; \frac{1}{2 } \,e^{-\alpha n } \enspace . \end{aligned}\ ] ] in this case , by chebyshev s inequality , @xmath345 \leq e^{-\alpha n/3 } \enspace .\ ] ] we say that a basis vector is _ bad _ if this bound is violated , i.e. , @xmath346 let @xmath347 denote the subset of @xmath148 consisting of bad basis vectors ( observe that while @xmath348 depends only on the choice of @xmath215 , @xmath347 depends also on @xmath51 ) .
then implies @xmath349 now let @xmath350 be the event that @xmath351 then by markov s inequality , @xmath350 occurs with probability at least @xmath352 .
now let us separate @xmath353 into contributions from basis vectors outside and inside @xmath354 : @xmath355 the first sum is taken only over vectors @xmath143 for which @xmath356 conditioning on @xmath318 and recalling that @xmath357 , the rank estimate of gives @xmath358 it follows that @xmath359 is at most @xmath360 .
therefore , since conditioning on @xmath318 and @xmath350 we have @xmath361 , the second sum in is at most @xmath362 then combining , and , @xmath363 with probability at least @xmath364 \geq 1 - n^{-\gamma n } - e^{-\alpha n/6 }
\geq 1 - 2e^{-\alpha n/6}$ ] .
we complete the proof by setting @xmath365 .
we now generalize the proof of theorem [ thm : one ] to the case where the algorithm designer is allowed to choose an arbitrary finite frame @xmath366 of unit length vectors in @xmath221 , with a family of positive real weights @xmath367 , that satisfy the completeness condition @xmath368 ( note that this is simply where we have written @xmath143 and @xmath367 instead of @xmath156 and @xmath121 . )
[ thm : one - frames ] let @xmath272 be a frame with weights @xmath369 satisfying the completeness condition for an irreducible representation @xmath221 .
given the hidden subgroup @xmath53 where @xmath51 is chosen uniformly at random from @xmath273 , let @xmath274 be the probability that we observe the vector @xmath143 conditioned on having observed the representation name @xmath221 , and let @xmath370 be the natural distribution on @xmath148 . then there is a constant @xmath276 such that for sufficiently large @xmath10 , with probability at least @xmath277 in @xmath51 and @xmath215 , we have @xmath371 recall from in section [ sec : optimal ] that the conditional distribution on @xmath148 is given by @xmath372 and the natural distribution is given by @xmath373 .
the proof of theorem [ thm : one ] goes through with a few modifications .
first , let us change some semantics : given a subset @xmath374 , we let @xmath375 denote the weighted size of @xmath376 , @xmath377 with this definition , the total probability that falls in @xmath376 under the natural distribution is @xmath378 .
then we will use the following lemma : [ lem : framesum ] let @xmath379 be a subspace of @xmath281 and let @xmath332 be the projection operator onto @xmath379 . then @xmath380 first note that a vector @xmath381 has entries @xmath382 for @xmath383 .
there is a unique linear operator @xmath384 on @xmath221 whose matrix entries are @xmath385 , and the inner product @xmath386 in @xmath281 can then be written as the bilinear form @xmath387 in @xmath221 .
the frobenius norm of @xmath384 is @xmath388 .
now let @xmath389 be an orthonormal basis for @xmath379 and let @xmath390 be the operator corresponding to @xmath391 .
then @xmath392 = { \textbf{tr}\,}e_i^\dagger e_i = { \left\| { { \mathbf{e}}}_i \right\|}^2 = 1 \end{aligned}\ ] ] where we used the cauchy - schwartz inequality in the second line and completeness in the second . summing over the @xmath335 basis vectors @xmath391 then gives .
we define @xmath291 and @xmath318 as before , and lemmas [ lem : lambda ] and [ lem : notinlambda ] still apply . as
before , let @xmath331 be the subspace consisting of copies of representations @xmath282 with @xmath292 , and let @xmath338 denote the set of @xmath143 with the property that @xmath393 then lemma [ lem : framesum ] implies that @xmath394 where @xmath340 is defined as above .
we again define @xmath347 as the set of @xmath395 such that @xmath396 and chebyshev s and markov s inequalities imply that the event @xmath350 , namely @xmath397 occurs with probability at least @xmath352 .
we separate @xmath398 as we did @xmath353 before : @xmath399 since @xmath400 for all @xmath401 , and since @xmath402 , conditioning on @xmath318 and using bounds the first sum as follows , @xmath403 it follows that @xmath359 is at most @xmath404 . conditioning on @xmath318 and @xmath350
, we have @xmath405 .
thus the second sum in is at most @xmath406 then combining , and , @xmath407 with probability at least @xmath364 \geq 1 - n^{-\gamma n } - e^{-\alpha n/6 }
\geq 1 - 2e^{-\alpha n/6}$ ] .
we complete the proof by setting @xmath365 as before .
the preceding development focuses on the case where the hidden subgroup is distributed uniformly among the conjugates of the subgroup @xmath408 . as such
, this shows that the canonical reduction of graph automorphism ( the problem of determining whether a given graph has a non - trivial automorphism ) to the hidden subgroup problem does not , in general , give rise to an efficient quantum algorithm via fourier sampling .
however , the canonical reduction of graph isomorphism to the hidden subgroup problem induces a more structured set of involutions . as referred to in the introduction , fixing two rigid graphs @xmath409 and @xmath410 , each with @xmath10 vertices , the automorphism group of their disjoint union @xmath411 is nontrivial exactly when they are isomorphic , in which case it is generated by an involution @xmath51 with full support such that @xmath412 . identifying @xmath413 and @xmath414 with the sets @xmath415 and @xmath416 respectively , and letting @xmath417
denote the involution @xmath418 , the standard reduction to the hidden subgroup problem in @xmath419 then results in a hidden subgroup @xmath420 where @xmath51 is a conjugate involution @xmath421 .
however , rather than @xmath422 being drawn from all of @xmath419 , it is an element of the young subgroup @xmath423 which fixes @xmath413 and @xmath414 : @xmath424 in other words , rather than considering all conjugates of @xmath51 in @xmath419 , it suffices just to consider conjugates in @xmath423 . _ a priori _
, it seems that this smaller set of possible hidden subgroups might be easier to identify .
moreover , let @xmath30 be the subgroup generated by @xmath423 and @xmath417 : this is the wreath product @xmath425 , which can also be written as a semidirect product @xmath426 .
then each such @xmath3 is contained in @xmath30 , and it seems that it might be more intelligent to fourier sample over @xmath30 rather than over all of @xmath419 .
however , we can show that nothing is gained by this approach .
first , note that the involutions described above form the ( @xmath30-)conjugacy class @xmath427 we remark the that the development of section [ sec : optimal ] is unchanged , and that the optimal measurement to find a hidden conjugate again consists of strong fourier sampling . now note that fourier sampling over @xmath419 and over @xmath30 are equivalent for the following reason : suppose we are trying to distinguish a set of hidden subgroups @xmath428 , all of which are contained in a subgroup @xmath429 .
let @xmath430 be a set of representatives for the cosets of @xmath30 .
then a random left coset of @xmath431 in @xmath0 is the product of a random left coset of @xmath431 in @xmath30 with a random element of @xmath430 .
thus the mixed state describing a uniformly random coset of @xmath431 in @xmath0 can be written as the corresponding state over @xmath30 with the completely mixed state over @xmath430 .
since this completely mixed state ( whose density matrix is the identity ) contains no information , nothing is gained ( or lost ) by sampling over all of @xmath0 rather than over @xmath30 .
finally , we observe that one can obtain easy character bounds for this conjugacy class of @xmath30 .
one determines @xmath30 s irreducible representations and their characters as follows . for two irreducible representations @xmath41 and @xmath79 of @xmath27 , let @xmath432 denote their tensor product as a representation of @xmath433 .
we consider the _ induced representation _
@xmath434 ( see @xcite ) and denote its character @xmath435 .
it is easy to see that @xmath436 as the notation suggests , this depends only on the multiset @xmath437 . observe that if @xmath376 and @xmath148 are two distinct multisets of representations of @xmath27 ( each of cardinality two ) then the characters @xmath438 and @xmath439 are orthogonal .
furthermore , an easy computation shows that @xmath440 and hence that @xmath441 is irreducible if @xmath442 , while if @xmath443 then the induced representation is the direct sum of two irreducible representations , @xmath444 where @xmath445 and @xmath253 are the trivial and sign representations , respectively , of @xmath446 .
each of these irreducible representations acts on @xmath447 , the vector space supporting the action of @xmath448 .
both of them realize the element @xmath449 as the linear map @xmath450 , while @xmath451 and @xmath452 realize the element @xmath453 as the maps which send @xmath454 to @xmath455 and @xmath456 respectively .
the characters of these representations can be determined by inspection ; in particular , @xmath457 where @xmath458 and @xmath459 denote the characters of @xmath451 and @xmath452 respectively . in light of this description of the irreducible characters of @xmath30 , one finds that @xmath460 when @xmath461 and that @xmath462 for @xmath463 . in particular , the absolute value of the normalized character at @xmath464 is @xmath465 .
as before , the dimensions @xmath39 of the diagrams with large normalized character sum to very little , and the analysis of section [ sec : single ] can be undertaken _
mutatis mutandis_. remarkably , this can be carried out without use of roichman s estimates !
this work was supported by nsf grants ccr-0093065 , phy-0200909 , eia-0218443 , eia-0218563 , ccr-0220070 , and ccr-0220264 , and aro grant w911nf-04-r-0009 .
we are grateful to denis thrien , mcgill university , and bellairs research institute for organizing a workshop at which this work began ; to dorit aharonov , daniel rockmore , and umesh vazirani for helpful conversations ; to chris lomont for pointing out several typos ; and to tracy conrad and sally milius for their support and tolerance .
also thanks rosemary moore for providing a larger perspective .
finally , we thank gorjan alagic for his comments on the structured involutions material .
michelangelo grigni , leonard j. schulman , monica vazirani , and umesh vazirani .
quantum mechanical algorithms for the nonabelian hidden subgroup problem .
33rd acm symposium on theory of computing _ , pages 6874 , 2001 .
sean hallgren , alexander russell , and amnon ta - shma .
normal subgroup reconstruction and quantum computation using group representations .
32nd acm symposium on theory of computing _ , pages 627635 , 2000 .
_ asymptotic representation theory of the symmetric group and its applications in analysis_. translated by n. v. tsilevich .
volume 219 in translations of mathematical monographs .
american mathematical society , 2003 .
cristopher moore , daniel rockmore , alexander russell , and leonard schulman . the value of basis selection in fourier sampling : hidden subgroup problems for affine groups . _ proc .
15th annual acm - siam symposium on discrete algorithms _ , pages 11131122 , 2004 .
vershik and s. v. kerov .
asymptotic behavior of the maximum and generic dimensions of irreducible representations of the symmetric group . funk .
i prolizhen , 19(1):2536 , 1985 ; english translation , funct .
appl . , 19:2131 , 1989 . | we resolve the question of whether fourier sampling can efficiently solve the hidden subgroup problem . specifically , we show that the hidden subgroup problem over the symmetric group can not be efficiently solved by strong fourier sampling , even if one may perform an arbitrary povm on the coset state .
these results apply to the special case relevant to the graph isomorphism problem . |
[ cols="<,^,^,^,^,^ " , ] [ tab : xmmspectralfits ] a fit with an absorbed blackbody model yields @xmath7 for 2282 degrees of freedom ( d.o.f . ) , with a hydrogen column density @xmath8 @xmath9 , temperature in the 0.760.90 kev range , and emitting radius in the 1.22.5 km interval . when an absorbed power law model is used ( @xmath10 for 2282 d.o.f . ) , the photon index is within the range 4.24.9 , much larger than what is usually observed for this class of sources ( see mereghetti 2008 ; rea & esposito 2011 for reviews ) .
moreover , a larger absorption value is obtained ( @xmath11 @xmath9 ) .
the large values for the photon index and the absorption are likely not intrinsic to the source , but rather an artifact of the fitting process which tends to increase the absorption to compensate for the large flux at low energies defined by the power law .
the addition of a second component to the blackbody , i.e. another blackbody or a power law , is not statistically required ( @xmath7 for 2238 d.o.f . in both cases ) .
we thus conclude that a single absorbed blackbody provides the best modelling of the source spectrum in the 0.38 kev energy range ( see table [ tab : spectralfits ] ) . taking the absorbed blackbody as a baseline
, we tried to model all the spectra tying either the radius or the temperature to be the same for all spectra .
we found @xmath12 for 2303 d.o.f . when the radii are tied , with @xmath13 @xmath9 , @xmath14 km and temperatures in the 0.660.97 kev range .
we found instead @xmath15 for 2303 d.o.f . when the temperatures are tied , with @xmath16 @xmath9 , @xmath17 kev and radii spanning from @xmath41.1 to @xmath43 km .
the goodness of fit of the latter model improves considerably if the temperatures are left free to vary as well ( @xmath18-test probability of @xmath19 ; fitting the temperature evolution with a constant yields a poor @xmath20 for 24 d.o.f . in this case ) .
we conclude that both the temperature and the size of the blackbody emitting region are varying .
zeroth - order spectral data of the three grating observations were fitted together and indipendently with this model , without including the pile - up component and fixing @xmath21 to that obtained in non - grating fit : @xmath22 @xmath9 ( see table [ tab : spectralfits ] and fig .
[ fig : parameters ] ) . for all the _ xmm
newton _ observations , we extracted the source counts from a circular region of radius 15 arcsec centred on the source psf , and the background counts through the same circle at the same position in an archival ( 2011 ) _ xmm newton _ observation of the galactic centre ( obs .
i d 0694640301 ) , when the magnetar was not detected and no transient events were identified within the source psf .
we built the light curves for the source and background event files to visually inspect and filter for high particle background flaring in the selected regions .
we checked for the potential impact of pile - up using the epatplot task of sas : the observed pattern distributions for both single and double events are consistent with the expected ones ( at a 1@xmath23 confidence level ) for all the three cameras , proving that the _ xmm newton _ data are unaffected by pile - up .
we restricted our spectral analysis to photons having flag = 0 and pattern @xmath24 for the pn ( moss ) data and created spectral redistribution matrices and ancillary response files .
we co - added the spectral files of consecutive observations ( obs .
i d 0743630201 - 301 and 0743630401 - 501 ; see table [ tab : log ] ) to improve the fit statistics and reduce the background contamination .
we then grouped the source spectral channels to have at least 200 counts per bin and fitted the spectra in the 212 kev range , given the high background contamination within the source psf at lower energies .
the spectral data extracted from the two mos cameras gave values for the parameters and fluxes consistent with those obtained from the pn camera . to minimize the systematic errors
introduced when using different instruments , we considered only the pn data , which provide the spectra with the highest statistics . due to the large psf of _ xmm newton _
, it is not possible to completely remove the contamination of both the galactic centre soft x - ray diffuse emission and the emission lines from the supernova remnant sgr a east , including in particular the iron line ( fe xxv ; rest energy of 6.7 kev ) and the sulfur line ( s xv ;
rest energy of 2.46 kev ) ( see e.g. maeda et al .
2002 ; sakano et al . 2004
; ponti et al .
2010 , 2013 ; heard & warwick 2013 ) .
these features were clearly visible especially in the spectra of the last observations , when the flux is lower , and prevented us from obtaining a good spectral modelling in xspec .
we thus decided to discard the energy interval comprising the fe xxv line ( 6.47.1 kev ) for all the spectra , as well as that associated with the s xv line ( 2.32.7 kev ) for the spectrum of the last observations ( obs .
i d 0690441801 , 0743630201 - 301 , 0743630401 - 501 ) , involving a loss of @xmath25 per cent in the total number of spectral bins .
based on the results of the _ chandra _ spectral analysis , we fitted the data first with an absorbed blackbody model .
the hydrogen column density was consistent with being constant at a 90 per cent confidence level among all observations and was tied to be the same in the spectral fitting .
we obtained @xmath26 for 636 d.o.f .
, with large residuals at high energies .
the latter disappear if an absorbed power law component is added and the fit improves considerably ( @xmath27 for 624 d.o.f . ; see left panel of fig .
[ fig : pnspectra ] ) .
a fit with a two - blackbody model is statistical acceptable as well ( @xmath27 for 624 d.o.f . ) and yields temperatures of @xmath28 kev and emitting radii of @xmath29 km for the second blackbody .
however , this model would be physically hard to justify , since it is unlikely that these large temperatures can be maintained on a neutron star surface for such a long time . as an alternative to these fits
, we applied a 3d resonant cyclotron scattering model ( ntz : nobili et al .
2008a , b ; zane et al . 2009 ) , obtaining @xmath30 for 624 d.o.f .
( see right panel of fig .
[ fig : pnspectra ] ) .
the hydrogen column densities and fluxes inferred both from the bb+pl and the ntz models are consistent with each other within the errors ( see table [ tab : xmmspectralfits ] ) . to test the robustness of our results ,
we compared the inferred parameters with those derived by fitting the spectra without filtering for the spectral channels and applying the varabs model for the absorption , which allows the chemical abundances of different elements to vary ( only the sulfur and iron abundances were allowed to vary for the present purpose ) .
we found consistent values over the two methods . we conclude that both models successfully reproduce the soft x - ray part of the spectra up to @xmath412 kev , implying that , similar to other magnetars , the reprocessing of the thermal emission by a dense , twisted magnetosphere produces a non - thermal component .
the power law detected by _ xmm
newton _ is consistent with that observed by _
( kaspi et al . 2014 ) , and its very low contribution below 8kev is consistent with its non - detection in our _ chandra _ data . to search for spectral variability as a function of rotational phase and time , we first extracted all the spectra of the _ chandra _ observations selecting three pulse phase intervals ( see fig .
[ fig : pulseprofiles ] ) : peak ( @xmath31=0.50.9 ) , minimum ( @xmath31=0.20.5 ) , and secondary peak ( @xmath31=0.91.2 ) .
we adopted the same extraction regions and performed the same data analysis as for the phase - averaged spectroscopy . for each of the three different phase intervals , we fitted the spectra of all _ chandra _ observations jointly in the 0.38 kev energy band with an absorbed blackbody model and tying the hydrogen column density to be the same in all the observations ( the pile - up model was included ) . since the values of the column density are consistent with being the same at a 90 per cent confidence level ( @xmath32 @xmath9 , @xmath33 @xmath9 , and @xmath34 @xmath9 for the peak , the secondary peak and the minimum , respectively ) , we fixed @xmath21 to @xmath22 @xmath9 , i.e. to the best - fit value determined with the phase - averaged spectroscopy ( see table [ tab : spectralfits ] ) .
we obtained a good fit in all cases , with @xmath35 for 1005 d.o.f .
for the peak , @xmath36 for 635 d.o.f .
for the secondary peak , and @xmath37 for 713 d.o.f .
for the pulse minimum .
the fit residuals were not optimal for energies @xmath38 kev for the peak spectra , due to the larger pile - up fraction .
we extracted the source counts excluding the central piled up photons ( within a radial distance of 0.7 arcsec from the source position ) , and repeated the analysis for the peak spectra : the residuals are now well shaped , and the inferred values for the spectral parameters did not change significantly . the temporal evolution of the blackbody temperature and radius for both the peak and the pulse minimum are shown in fig .
[ fig : pps ] .
no particular trend is observed for the inferred temperatures , whereas the size of the emitting region is systematically lower for the pulse minimum .
this is consistent with a viewing geometry that allows us to observe the hot spot responsible for the thermal emission almost entirely at the peak of the pulse profile , and only for a small fraction at the minimum of the pulsation .
-0.1truecm the higher statistics of the _ xmm newton_/epic - pn data allowed us to put more stringent constraints on the variations of the x - ray spectral parameters along the spin phase .
we extracted the background - subtracted spectra in six different phase intervals for each observation , as shown in fig .
[ fig : parametersvsphase ] .
we fitted all spectra with a bb+pl model , adopting the same prescriptions used for the phase - averaged spectroscopy in the filtering of the spectral channels .
we tied the hydrogen column density and the power law photon indices to the best - fit values determined with the phase - averaged analysis ( see table [ tab : xmmspectralfits ] ) .
we obtained statistically acceptable results in all cases .
the evolutions of the blackbody temperature and emitting radius as a function of the rotational phase for all the observations are shown in fig .
[ fig : parametersvsphase ] .
variability of both the parameters along the rotational phase is more significant during the first observation ( a fit with a constant yields @xmath39 for 5 d.o.f . in both cases ) than in the following observations ( @xmath40 for 5 d.o.f . in all cases ) . to search for possible phase - dependent absorption features in the x - ray spectra of ( similarly to the one detected in ; tiengo et al .
2013 ) , we produced images of energy versus phase for each of the eight epic - pn observations .
we investigated different energy and phase binnings . in fig .
[ fig : evsphase ] we show the image for the observation with the highest number of counts ( obs .
i d 0724210201 ) , produced by binning the source counts into 100 phase bins and 100-ev wide energy channels .
the spin period modulation is clearly visible , as well as the large photoelectric absorption below 2kev . for all observations we then divided these values first by the average number of counts in the same energy bin and then by the corresponding 0.310 kev count rate in the same phase interval .
no prominent features can be seen in any of the images . -0.1truecm
-0.1truecm for all the _ chandra _ observations , we used the chandra ray tracer ( chart ; carter et al .
2003 ) to simulate the best available psf for , setting the exposure time of each simulation equal to the exposure time of the corresponding observation . for the input spectrum in chart we employed the blackbody spectrum of table [ tab :
spectralfits ] , accounting for the pile - up .
we then projected the psf rays on to the detector plane via the model of axaf response to x - rays software ( marx , version 4.5.0 ; wise et al .
we extracted the counts of both the simulated psfs and the acis event files through 50 concentric annular regions centred on the source position and extending from 1 to 30 pixels ( 1 acis - s pixel corresponds to 0.492 arcsec ) .
we then generated the x - ray brightness radial profiles and normalized the nominal one ( plus a constant background ) to match the observed one at a radial distance of 4 pixels , i.e. at a distance at which pile - up effects are negligible .
a plot of the observed and simulated surface brightness fluxes ( in units of counts @xmath41 pixel@xmath42 ) versus radial distance from the position of is shown in fig . [
fig : psf_profile ] for the observation with the highest number of counts ( obs .
i d 15041 ) .
extended emission around is clearly detected in all the observations , and it is likely dominated by the intense galactic centre diffuse emission .
a detailed analysis of the diffuse emission , including its spatial extension and spectral properties , is beyond the scope of this paper , and will be published in a subsequent work .
the past decade has seen a great success in detecting magnetar outbursts , mainly thanks to the prompt response and monitoring of the _ swift _ mission , and to the dedicated follow - up programs of _ chandra _ , _ xmm newton _ , and more recently , _
nustar_. the detailed study of about ten outbursts has shown many common characteristics ( see rea & esposito 2011 for a review ; see also fig.[fig : outbursts ] ) , although the precise triggering mechanism of these outbursts , as well as the energy reservoir responsible for sustaining the emission over many months , remains uncertain .
all the outbursts that have been monitored with sufficient detail are compatible with a rapid ( @xmath43days ) increase in luminosity up to a maximum of a few @xmath44 erg s@xmath3 and a thermally dominated x - ray spectrum which softens during the decay . in the case of and
, a non - thermal component extending up to 100200 kev appears at the beginning of the outburst , and becomes undetectable after weeks / months ( rea et al .
2009 ; bernardini et al . 2011 ; kuiper et al .
2012 ) .
the initial behavior of the 2013 outburst decay of was compatible with those observed in other magnetars .
the outburst peak , the thermal emission peaked at about 1 kev , the small radiating surface ( about 2 km in radius ) , and the overall evolution in the first few months , were consistent with the behavior observed in other outbursts .
however , after an additional year of x - ray monitoring , it became clear that the subsequent evolution of showed distinct characteristics .
the source flux decay appears extremely slow : it is the first time that we observe a magnetar with a quiescent luminosity @xmath45 erg s@xmath3 remaining at a luminosity @xmath46 erg s@xmath3 for more than one year , and with a temperature decreasing by less than 10% from the initial @xmath47kev .
a further interesting feature of this source is that the non - thermal component ( as detected by _
xmm newton _ ) persisted on a very long temporal baseline during the outburst evolution .
the flux due to the power law component does not change significantly in time and , as a result , its fractional contribution to the total flux is larger at late times : @xmath48 d after the outburst onset , @xmath49 per cent of the 110 kev absorbed flux is due to the non - thermal component .
we first modelled the decay empirically to gauge the characteristic decay timescales .
we adopted three different functions to model the blackbody temperature , radius and 110 kev absorbed flux temporal evolutions ( see fig .
[ fig : parameters ] ) : ( i ) a linear model ; ( ii ) a power law : @xmath50 ; ( iii ) an exponential : @xmath51 $ ] , where @xmath52 is the epoch of the first burst detected ( which we fixed to 2013 april 24 in all cases ) and @xmath53 is the e - folding time . the temporal evolution of the magnetar temperature is well represented by a linear model ( @xmath54 for 23 d.o.f . ) , with initial temperature @xmath55 kev and slope ( @xmath56 ) @xmath41 10@xmath57 .
the hot spot shrinking is best modeled by an exponential ( @xmath58 for 23 d.o.f . ) .
best - fit parameters are @xmath59 d and initial radius @xmath60 km
. the shape of the flux decline appears to change in time and in fact none of these models can accurately describe the magnetar flux overall decay .
the flux decay during the first 100days since the outburst onset is well modelled by a linear plus exponential model with @xmath61 d ( @xmath62 = 1.5 for 4 d.o.f . ) .
after @xmath63 days , the best fitting model turns out to be an exponential with @xmath64 d ( @xmath62 = 1.4 for 15 d.o.f . ) .
we applied the crustal cooling model ( see e.g. pons & rea 2012 ) to the data collected during the 1.5-yr outburst of .
although this model was successful in explaining several other magnetar outbursts ( rea et al .
2012 , 2013b ) , in this case we could not reproduce the very slow cooling and high luminosity observed for this source .
we ran several models varying the total injected energy , the angular size , and the depth of the region where the energy is released , but we could not find any set of parameters that fit the data . in the framework of the starquake model , the maximum temperature reached in the region where the energy is released is limited by neutrino emission processes .
this internal temperature determines the maximum surface temperature and therefore the luminosity at which the outburst peaks during the first few days . for injected energies
@xmath65erg , there is no significant increase in the peak luminosity because the crustal temperature saturates ( at about @xmath66k ) due to the efficient neutrino processes . after reaching the maximum luminosity ( between 1 hour and 1 day depending on the depth and injection rate )
, the cooling curve tracks the thermal relaxation of the crust .
independent of the initial injected energy and surface temperature , the luminosity is expected to drop below @xmath44 erg s@xmath3 after @xmath4320 - 30 days ( see e.g. fig.1 in pons & rea 2012 ) , due to neutrino emission processes in the crust ( mainly plasmon decay , and probably neutrino synchrotron for magnetar field strengths ) . in fig.[fig : cooling ] ( left panel , lower curves ) we show an example of the expected cooling curve of a magnetar with the same characteristics of .
we assume that a sudden large energy release , @xmath67 erg , heats up a layer of the outer crust up to @xmath68k .
we also assume that the event affects the entire magnetar surface , to create the most favorable scenario ( the luminosity simply scales with the area of the emitting region ) , and that the layer where the energy is injected extends from an external boundary at @xmath69 g@xmath70 , to an inner boundary at @xmath71 and @xmath72 g@xmath70 ( we show these two cases in the two lower curves of fig.[fig : cooling ] , left panel ) .
it is clear that , even in this most favorable case , the high luminosities observed at late times are difficult to reconcile with any cooling model . in particular , injecting more energy or changing @xmath73 will only affect the peak luminosity during the first days or weeks . on the other hand , injecting energy deeper into the crust ( i.e. at higher @xmath74 ) ,
is expected to change the late time evolution only slightly .
this can be seen by comparing the solid and dashed lines in the left panel of fig .
[ fig : cooling ] , which correspond to @xmath75 and 4 @xmath76 g@xmath70 , respectively . for illustrative purposes
, we also show the cooling curves obtained when plasmon and synchrotron neutrino processes are switched off ( see the upper curves in the left panel of fig.[fig : cooling ] ) .
these provide a much closer match to the data ; however , there is no clear reason why these neutrino processes should not operate in these conditions .
this example is only meant to highlight the relevance of understanding neutrino processes in the crust , especially under the presence of strong fields .
another possibility to fit the data is to tune the energy injection , which must be maintained during the first @xmath77days , resulting in a higher luminosity at late times .
if we assume that only a region 5 km in radius is affected ( this is closer to the @xmath78 km emitting region observed ) , we need a continuous injection of at least @xmath79 erg s@xmath3 ( per day ) for about 200days , which results in a total energy of a few @xmath80erg .
while this energy budget may not be unrealistic , a physical mechanism that can operate for such a long timescale is not known .
a possibility might be a continuous injection of energy to keep the surface at high temperatures for so long , although in this latter case we should possibly expect more sgr - like bursts during the first hundreds days . in this section
we discuss the possibility that the prolonged high luminosity of is in part due to external particle bombardment as a consequence of the existence of a twisted magnetic field bundle .
a valid alternative model to the crustal cooling scenario invokes the presence of magnetospheric currents flowing along a gradually shrinking magnetic bundle , and heating the surface from outside . according to beloborodov ( 2007 , 2013 ) ,
this bundle can untwist on different timescales : i ) in the equatorial regions of the magnetosphere , where the magnetic field reaches a few stellar radii , currents are dissipated after weeks or months , while ii ) at higher latitudes ( close to the poles ) , a bundle may untwist more slowly , possibly in one to ten years . here , particles can reach lorentz factors of a few tens ( beloborodov 2007 ) . in this scenario ,
a quasi steady - state outflow of electrons and positrons is maintained thanks to magnetic pair production close to the surface .
the non - negligible electric voltage along the magnetic field lines and the radiative force due to compton scattering regulate the streams of positrons and electrons along the field line .
[ fig : lumradius ] the presence of a non - thermal component observed by _ nustar _
( mori et al . 2013 ; kaspi et al . 2014 ) , and confirmed also by our _ xmm newton _
observations on a much longer temporal baseline , is suggestive of a large density of magnetospheric particles which boost thermal photons emitted from the surface via resonant compton scattering , providing the power law component . in this context , the observed @xmath42 km size of the emitting blackbody is consistent with a relatively small j - bundle . in the scenario in which the outburst evolution is dominated by an untwisting bundle and the poloidal magnetic field
has a dipole geometry , the luminosity is expected to decrease with the square of the blackbody area ( @xmath81 ; beloborodov 2007 , 2009 ) .
a flatter dependence may arise from a more complex field geometry . in fig.[fig :
lumradius ] we show the fits of the bolometric luminosity as a function of @xmath82 with two different models , a quadratic function @xmath83 ( black line ; @xmath84 for 23 d.o.f . ) and a power law @xmath85 ( red line ; @xmath86 for 23 d.o.f . ) . for the latter model
we find @xmath87 .
interestingly , a similar relation was observed also for the outburst decay of ( rea et al . 2013b ) and ( an et al . 2013 ) . in the following
we will assess , using first order approximations , whether the particle density needed to keep the footprint of the bundle at a temperature of @xmath41kev for the first hundreds of days after the outburst onset is consistent with the particle density in the bundle responsible for the non - thermal power law tail .
the power of the infalling particles is @xmath88 , where @xmath89 is the kinetic energy of a single particle at the surface and @xmath90 is the total number of infalling particles per unit time . if this kinetic energy is transferred by the infalling particles to the footprint of the bundle , and produces thermal luminosity from the footprint surface , then : @xmath91 where @xmath92 is the area of the footprint surface , @xmath93 is the spot temperature , @xmath94 is the density of the infalling particles ( assumed to be electrons and/or positrons , created by means of pair production ) , and @xmath95 is the lorentz factor .
we calculated the density of the infalling particles by considering the kinetic energy they need to heat the base of the bundle spot . for a given temperature , one can estimate @xmath94 as @xmath96 ^ 4}{\gamma } { \rm cm^{-3}}~.\ ] ] on the other hand , we can estimate the density of the particles responsible for the resonant compton scattering which produces the x - ray tail as @xmath97 where @xmath98 is the conduction current , @xmath99 is the local magnetic field , and @xmath100 is the length - scale over which @xmath99 varies ( @xmath101 is the star radius ) . in the magnetosphere of a magnetar the real current is always very close to @xmath102 and it is mostly conducted by e@xmath103 pairs ( beloborodov 2007 ) . the abundance of pairs is accounted for by the multiplicity factor @xmath104 which is the ratio between the actual charge density ( including pairs ) and the minimum density needed to sustain @xmath102 ; the latter corresponds to a charge separated flow in which the current is carried only by electrons ( and ions ) . if the same charge population is responsible for both resonant compton scattering and surface heating , the densities given by eqs .
( [ eq : n_bomb ] ) and ( [ eq : n_rcs ] ) should be equal .
this implies @xmath105 according to beloborodov ( 2013 ) , both the lorentz factor and the pair multiplicity change along the magnetic field lines , with typical values of @xmath106 ( i.e. efficient pair creation ) , @xmath107 in the largest magnetic field loops , and @xmath108 ( i.e. charge - separated plasma ) , @xmath109 in the inner part of the magnetosphere
. the previous equality can not be satisfied for a typical temperature of @xmath110 kev , unless the magnetic field changes over an exceedingly small length - scale , a few meters at most .
it appears , therefore , very unlikely that a single flow can explain both surface heating and resonant up - scattering .
the spectacular angular resolution of _ chandra _ and the large effective area of _ xmm newton _ , together with an intense monitoring of the galactic centre region , has allowed us to collect an unprecedented dataset covering the outburst of , with very little background contamination ( which can be very severe in this region of the milky way ) .
the analysis of the evolution of the spin period allowed us to find three different timing solutions between 2013 april 29 and 2014 august 30 , which show that the source period derivative has changed at least twice , from 6.6@xmath5 s s@xmath3 in 2013 april at the outburst onset , to 3.3@xmath6 s s@xmath3 in 2014 august . while the first @xmath111 change could be related with the occurrence of an sgr - like burst ( kaspi et al .
2014 ) , no burst has been detected from the source close in time to the second @xmath111 variation ( although we can not exclude it was missed by current instruments ) .
this further change in the rotational evolution of the source might be related with the timing anomaly observed in the radio band around the end of 2013 ( lynch et al .
2015 ) , unfortunately during our observing gap .
the 0.38kev source spectrum is perfectly modelled by a single blackbody with temperature cooling from @xmath40.9 to 0.75kev in about 1.5 years .
a faint non - thermal component is observed with _ xmm
newton_. it dominates the flux at energies @xmath1128kev at all the stages of the outburst decay , with a power law photon index ranging from @xmath113 to @xmath114 .
it is most probably due to resonant compton scattering onto non - relativistic electrons in the magnetosphere . modelling the outburst evolution with crustal cooling models
has difficulty in explaining the high luminosity of this outburst and its extremely slow flux decay .
if the outburst evolution is indeed due to crustal cooling , then magnetic energy injection needs to be continuous over at least the first @xmath4200 days .
the presence of a small twisted bundle sustaining currents bombarding the surface region at the base of the bundle , and keeping the outburst luminosity so high , appears a viable scenario to explain this particular outburst . however , detailed numerical simulations are needed to confirm this possibility .
this source is rather unique , given its proximity to sgr a@xmath0 .
in particular , it has a @xmath115 per cent probability of being in a bound orbit around sgr a@xmath0 according to our previous n - body simulations ( rea et al .
2013a ) , and the recent estimates inferred from its proper motion ( bower et al .
we will continue monitoring the source with _
chandra _ and _ xmm newton _ for the coming year .
fcz and nr are supported by an nwo vidi grant ( pi : rea ) and by the european cost action mp1304 ( newcompstar ) .
nr , ap , dv and dft acknowledge support by grants aya2012 - 39303 and sgr2014 - 1073 .
ap is supported by a juan de la cierva fellowship .
jap acknowledges support by grant aya 2013 - 42184-p .
pe acknowledges a fulbright research scholar grant administered by the u.s.italy fulbright commission and is grateful to the harvard
smithsonian center for astrophysics for hosting him during his fulbright exchange .
dh acknowledges support from _ chandra x - ray observatory _ ( cxo ) award number go3 - 14121x , operated by the smithsonian astrophysical observatory for and on behalf of nasa under contract nas8 - 03060 , and also by nasa _ swift _
grant nnx14ac30 g .
gp acknowledges support via an eu marie curie intra - european fellowship under contract no .
fp - people-2012-ief- 331095 and the bundesministerium fr wirtschaft und technologie / deutsches zentrum fr luft - und raumfahrt ( bmwi / dlr , fkz 50 or 1408 ) and the max planck society .
rp acknowledges partial support by _ chandra _ grants ( awarded by sao ) g03 - 13068a and g04 - 15068x .
rpm acknowledges funding from the european commission seventh framework programme ( fp7/2007 - 2013 ) under grant agreement n. 267251 .
fcz acknowledges csic - ieec for very kind hospitality during part of the work and geoffrey bower for helpful discussions .
the scientific results reported in this article are based on observations obtained with the _ chandra x - ray observatory _ and _ xmm newton _ , an esa science mission with instruments and contributions directly funded by esa member states and nasa .
this research has made use of software provided by the _
chandra _ x - ray center ( cxc ) in the application package ciao , and of softwares and tools provided by the high energy astrophysics science archive research center ( heasarc ) , which is a service of the astrophysics science division at nasa / gsfc and the high energy astrophysics division of the smithsonian astrophysical observatory .
fruscione a. et al . , 2006 , in silva d. r. , doxsey r. e. , eds , observatory operations : strategies , processes , and systems vol .
6270 of spie conference series , ciao : chandra s data analysis system .
spie , bellingham garmire g. p. , bautz m. w. , ford p. g. , nousek j. a. , ricker jr .
g. r. , 2003 , in truemper j. e. , tananbaum h. d. , eds , x - ray and gamma - ray telescopes and instruments for astronomy .
4851 of proceedings of the spie .
, advanced ccd imaging spectrometer ( acis ) instrument on the chandra x - ray observatory .
spie , bellingham , pp 2844 | in 2013 april a new magnetar , , was discovered as it entered an outburst , at only 2.4 arcsec angular distance from the supermassive black hole at the centre of the milky way , sagittarius a@xmath0 .
has a surface dipolar magnetic field of @xmath1 g , and it is the neutron star closest to a black hole ever observed .
the new source was detected both in the radio and x - ray bands , with a peak x - ray luminosity @xmath2erg s@xmath3 .
here we report on the long - term _ chandra _ ( 25 observations ) and _ xmm newton _ ( 8 observations ) x - ray monitoring campaign of from the onset of the outburst in april 2013 until september 2014 .
this unprecedented dataset allows us to refine the timing properties of the source , as well as to study the outburst spectral evolution as a function of time and rotational phase .
our timing analysis confirms the increase in the spin period derivative by a factor of @xmath42 around june 2013 , and reveals that a further increase occurred between 2013 oct 30 and 2014 feb 21 .
we find that the period derivative changed from 6.6@xmath5 s s@xmath3 to 3.3@xmath6 s s@xmath3 in 1.5 yr . on the other hand ,
this magnetar shows a slow flux decay compared to other magnetars and a rather inefficient surface cooling .
in particular , starquake - induced crustal cooling models alone have difficulty in explaining the high luminosity of the source for the first @xmath4200days of its outburst , and additional heating of the star surface from currents flowing in a twisted magnetic bundle is probably playing an important role in the outburst evolution .
galaxy : centre stars : magnetars x - rays : individual : sgrj1745 - 2900 . |
the molecular interstellar medium plays a critical role in the evolution of galaxies since it provides the material from which stars form .
there are observations @xcite of rotational transitions of co within @xmath2 , mostly in quasars and early universe molecular emission line galaxies ( hereinafter qso s and emgs , respectively ) which demonstrate that molecular clouds exist already at the epoch of the galaxy formation . however , the current sample is scarce for @xmath3 , where the study of the star and galaxy formation is specially interesting .
alma is the instrument that may solve this problem . in our chemical evolution models
the gas phase has two components , the diffuse and the molecular phases of gas , which we treat separately .
this way , it is possible to follow the evolution of the expected density of h@xmath4 with redshift .
we will show here our predictions about the @xmath0 density at high redshift , and we will analyze if they might be detectable with alma .
we have computed a wide grid of chemical evolution models for a large set of theoretical spiral and irregular galaxies of different total mass @xcite by using the multiphase chemical evolution model from @xcite .
we have simulated theoretical models with 44 different radial distributions of total mass and 10 star - forming efficiencies , with values in the range [ 10 ] , which we denoted by @xmath5 from 1 to 10 , for each one of them .
the 440 resulting models provide the time evolution for specific radial regions within each galaxy modeled , with the corresponding mass in each gaseous or stellar phase , as well as their elemental abundances .
this type of model was already applied to several individual spiral galaxies in @xcite , where it was shown that the model is adequately calibrated and that it may be used for galaxies of different morphological types and different total masses .
the code assumes that a total mass , @xmath6 , initially as diffuse gas , is located within a protogalaxy or halo .
the gas falls onto the equatorial plane forming out the spiral disk at a rate defined by a collapse time - scale which depends on @xmath6 . in each zone ,
halo or disk , several radial regions are defined .
we follow separately the evolution in each region .
the assumed star formation in the halo is a power law of the diffuse gas . in the disk , however , we assume that the star formation occurs in two steps : first , molecular clouds ( @xmath0 ) form from the diffuse gas ( hi ) also following a power law .
then , stars form by two means : 1 ) cloud - cloud collisions , and 2 ) by the interaction of the massive stars with the surrounding molecular clouds : this way , in our models the gas has two components : hi and @xmath0 .
since we have computed separately the diffuse gas evolution and the molecular clouds evolution , we obtain the radial distributions of the molecular gas mass , which is represented in fig .
there we show the radial distributions of the molecular gas density at the present - time for 4 values of the rotation velocity , or total mass of the galaxy .
thus , panel a ) shows the radial distributions produced in a low mass galaxy while d ) represents a massive galaxy .
panel c ) is the modeled galaxy most similar to our milky way galaxy and panel d ) is an intermediate mass galaxy . in each one of these panels models for different star formation efficiencies , named by @xmath5 ,
are represented .
the lines in color correspond to the lowest star - forming efficiencies ( high @xmath5 values ) , which maintain a higher density than in cases where stars are formed more efficiently ( black lines ) .
the distributions are close to exponential in the external regions but are flatter at the inner regions , mostly when the n is low , as is in fact observed in nature @xcite .
the evolution of the total mass of @xmath0 in galaxies is represented in fig [ fig2 ] .
the magenta open stars are estimates from @xcite for nearby qsos . the predicted present time molecular gas masses are similar to those ones of local objects with a maximum for @xmath7 in agreement with the observational limit @xmath8 , shown by a green solid line , obtained for those authors .
the red full dots are the estimates from @xcite for emgs .
the observed molecular gas masses are decreasing for decreasing @xmath9 in agreement with the most of our model results . only the models with low star - forming efficiencies ( colored lines ) are still producing @xmath0 . on the other hand , following that work @xcite , and taking into account the estimates for the star formation rate and molecular gas total mass , these objects might be bulges or starbursts in centrally concentrated disks , but not elliptical galaxies as it was initially assumed .
our models are also in agreement with this statement : the total mass of @xmath0 for the most massive spiral galaxies ( with @xmath10 in the range 100 - 300 km.@xmath11 ) and high star formation efficiencies ( @xmath12 ) are always below of the emgs data region since in those cases the molecular gas is very quickly consumed . only the models for massive galaxies -panels c ) and d ) with the lowest efficiencies show values closer to the observations of emg s . at this high redshift ,
these particular models show star formation only at the central regions and therefore they would appear as concentrated disks , with a large quantity of @xmath0 .
observation of emission from co rotational transitions is the usual means of detecting interstellar molecular clouds .
( half optical radius ) for three redshifts as labelled.,scaledwidth=45.0% ] co is a very stable molecule and the most abundant after h@xmath4 . following @xcite levels @xmath13
are the best ones to detect molecular gas at high redshifts . in order to verify if our densities of molecular gas may be detected with alma , we need to transform the limiting flux for this instrument to a limit in gas density .
the flux limit depends on the observed frequency .
we take the highest value from the report `` science with alma '' , and we use eq . 1 from @xcite , to calculate the minimum l@xmath14 necessary for detection .
this luminosity is transformed into gas mass by using , as those authors , a constant conversion factor @xmath15 ( which assumes independence of redshift @xmath9 ) .
the result is shown as a thick line in fig [ fig3 ] , where we show the expected radial distributions of molecular gas density normalized to a characteristic radius @xmath16 , taken as the half optical radius , for each total mass radial distribution .
we show in top panel a ) the results for @xmath7 ( present time ) while panels b ) and c ) represents the same models for the higher redshifts @xmath17 ( @xmath18 gyr ) and @xmath19 ( @xmath20 gyr ) , respectively . for @xmath7 most modeled galaxies
show distributions with values higher than the limiting value which may be detected with alma down to the optical radius , while for high redshifts only our inner disks , mainly of low efficiencies ( colored ) models , have a flux density sufficiently high to be detected .
therefore , if the red points of fig . 2 would correspond to spiral galaxies , they must represent , effectively , their central parts . in fig .
[ fig4 ] we represent the expected evolution of the flux density of co lines as a function of @xmath9 , for an integrated galaxy , computed from the molecular gas density @xmath21 ) obtained from our models , transformed to fluxes by using the method outlined .
the most massive galaxies -panel d- and those with high star formation efficiencies -black lines maintain an almost constant @xmath0 density and flux density up to @xmath22 , with values just over the detection limits .
thus , such as we have before claimed , the most efficient galaxies in forming stars , which consume very quickly the molecular gas , are not the most probable objects to be detected .
those with low rotation velocities @xmath23 km.s-1 , on the other hand , show increasing densities for decreasing z , having values lower than the limiting flux for @xmath24 . only the massive galaxies @xmath25 km.s@xmath26 with @xmath27 have fluxes above the observational limit even at @xmath28 .
therefore , the most plausible objects to be observed with alma will be the centers of low - intermediate efficiencies massive galaxies .
* at redshift @xmath29 only the central regions of galaxies might be observed in co with alma * emgs are not necessarily the largest or the most massive stellar disks . | we have computed a grid of chemical evolution models for a large set of spiral and irregular theoretical galaxies of different total mass . in our models ,
the gas phase has two components , the diffuse and the molecular one ( @xmath0 ) .
it is possible , therefore , to follow the time ( or redshift ) evolution of the expected density of the @xmath1 phase .
we will show the predictions of this gas density at hight redshift , which might be detected with alma , in this type of galaxies . |
in the standard model ( sm ) of particle physics , cp violation is due to a complex parameter in the cabibbo - kobayashi - maskawa ( ckm ) matrix .
extensions of the sm typically include new sources of cp violation . to understand the origin of cp violation it is important to investigate as many systems as possible .
one area that deserves further investigation is the @xmath8 system .
we consider a few hadronic decay modes of the @xmath8 . in the sm
, there is virtually no cp violation in these decay modes .
thus , any observation of cp violation for these decays would be a clear indication of new physics ( np ) .
cp - odd observables require at least two contributing amplitudes in order to be non - zero .
in addition to the usual w - exchange amplitude that is present in the sm , we assume that there is also a contribution that arises due to the exchange of a charged higgs boson
see fig .
[ fig : feyndiag ] .
decays showing the @xmath9 and @xmath10 contributions.,width=151 ] many extensions of the sm include extra higgs bosons .
the charged higgs couplings to light quarks in these models are often proportional to the quark masses , and are thus very small .
we investigate @xmath8 decays that contain light quarks in the final state @xcite .. copyright ( 2007 ) by the american physical society .
the reader is referred to refs .
@xcite for more details . ]
if cp violation is to be large in these decays , the charged higgs couplings must be large .
thus , these decays probe `` non - standard '' np cp violation .
suppose two amplitudes contribute to a particular process
. then the total amplitude may be written as @xmath11 where @xmath12 and @xmath13 are the relative weak ( cp - odd ) and strong ( cp - even ) phases , respectively .
the full rate for the process is obtained by calculating @xmath14 , summing / averaging over spins , and integrating over phase space .
the regular rate asymmetry for a particular decay mode is proportional to the difference between the rate for the process and that for its associated anti - process .
the result is proportional to @xmath15 and thus requires both a weak phase difference ( @xmath12 ) and a strong phase difference ( @xmath13 ) between the contributing amplitudes .
the rate asymmetry can be altered in various ways .
for example , if some of the spins are measured , one does not sum over them .
one can also integrate asymmetrically over phase space ( or use a more general weighting function ) to isolate certain terms in the differential width . in some cases this leads to asymmetries that are proportional to @xmath16 , similar to the regular rate asymmetry .
one particular class of asymmetries involves triple products , which have the form @xmath17 , where @xmath18 , @xmath19 and @xmath20 are momenta and/or spins .
cp asymmetries constructed from triple products have the form @xmath21 and thus require a weak phase difference , but not a strong phase difference .
consider the decay @xmath23 , where @xmath24 represents a vector or axial vector meson . the general structure for the sm hadronic current for this decay
is given by @xcite , @xmath25 where @xmath26 and where @xmath27 denotes the polarization tensor of the @xmath24 .
experimentally , the form factor @xmath28 dominates for @xmath29 and @xmath30 and/or @xmath31 are expected to dominate for @xmath32 .
the scalar form factor , @xmath33 , is expected to be very small within the sm .
the np charged higgs effect may be parameterized in terms of new scalar and pseudo - scalar quark - level operators .
the resulting current is given by , @xmath34 where @xmath35 is a form factor for the quark - level operators and @xmath36 and @xmath37 arise from the coupling of the charged higgs boson to the quarks and leptons .
the charged higgs effect can be incorporated into the expression for the hadronic current [ eq . ( [ eq : wamp ] ) ] by the replacement @xmath38 , where @xmath39 the parameters @xmath36 and @xmath37 can be complex ; that is , they can contain a ( cp - odd ) weak phase .
the various form factors are potential sources of strong phases .
the regular rate asymmetry is proportional to @xmath40 for the @xmath29 case ( integrated over phase space ) , where the @xmath13 s are the strong phases associated with the form factors and @xmath41 is the ( cp - odd ) phase of the complex higgs coupling @xmath37 .
an analogous expression ( with @xmath37 replaced by @xmath36 ) holds for the case @xmath32 .
since @xmath33 is expected to be very small , we conclude that the rate asymmetry is unlikely to be measureable , even in the presence of np . weighting the differential width by @xmath42 while performing the integration over phase space ( @xmath43 is a particular kinematical angle
see ref .
@xcite ) extracts terms containing the combinations @xmath44 and @xmath45 .
such terms could be non - vanishing for the decay @xmath2 ( since @xmath30 and/or @xmath31 are expected to be the dominant sm form factors in this case ) . constructing a cp asymmetry from this quantity yields an expression that depends on the polarization of the @xmath8 and that contains pieces such as @xmath46 .
this asymmetry requires a strong phase difference between @xmath30 ( or @xmath31 ) and @xmath35 .
numerical estimates ( see ref .
@xcite for details ) indicate that asymmetries of order 15% ( 7.5% ) could be possible , given the uncertainties in the experimental measurement of the branching ratio .
the first estimate assumes that only @xmath31 contributes to the sm amplitude ; the second that only @xmath30 does . a triple product ( tp ) can be constructed using the polarization tensor of the vector meson .
the tp contains the combination of form factors @xmath47 , and could thus be non - zero for the decay @xmath5 ( for which @xmath28 dominates the sm hadronic current ) . constructing a cp asymmetry from the tp yields an expression that contains @xmath48 ; thus this cp asymmetry does not require the presence of a relative strong phase between the sm and np amplitudes . a numerical estimate performed in ref .
@xcite indicates that the tp asymmetry could be as large as 30% multiplied by @xmath49 , where @xmath50 and @xmath51 are particular direction vectors in the hadronic rest frame and @xmath52 is the polarization vector of the @xmath53 .
the previous approach can be generalized to the decay @xmath55 .
the quark - level process is the same as that shown in fig .
[ fig : feyndiag ] , but with @xmath56 .
the hadronic current is given by @xcite , @xmath57t^{\mu\nu } \nonumber \\ & & + i f_3(s_1,s_2,q^2 ) \epsilon^{\mu\nu\rho\sigma } p_{1\nu}p_{2\rho}p_{3\sigma } \nonumber \\ & & + f_4(s_1,s_2,q^2 ) q^\mu\ , , \label{eq : jmu_sm}\end{aligned}\ ] ] where @xmath58 , @xmath59 , @xmath60 and @xmath61 .
several decay chains contribute to the form factors within the sm .
the dominant form factors @xmath30 and @xmath31 have been studied experimentally by cleo @xcite .
these form factors receive contributions due to @xmath62 and @xmath63 , respectively .
the subdominant processes @xmath64 and @xmath65 are also possible , and could contribute to @xmath28 .
the scalar form factor , @xmath33 , is expected to be small within the sm .
the np charged higgs contribution may be taken into account by the replacement , @xmath66 where @xmath35 is the pseudoscalar form factor and @xmath67 and @xmath68 are ( possibly complex ) np parameters defined somewhat similarly to @xmath36 and @xmath37 in the @xmath69 case .
we again consider several possible cp asymmetries .
each asymmetry is proportional to @xmath70 ( integrated over phase space ) .
the regular rate asymmetry contains the combination @xmath71 and is thus expected to be very small .
it is also possible to construct other asymmetries by employing various weighting functions when performing the integration over phase space ( see , for example , refs .
@xcite ) . in this manner
we have constructed two modified rate asymmetries , as well as one tp asymmetry .
the modified rate asymmetries contain the sm hadronic currents @xmath30 and @xmath31 and they require a non - zero relative strong phase between the sm and np amplitudes in order to be non - zero .
the tp asymmetry depends on @xmath28 and does not require a relative strong phase between the interfering amplitudes .
a preliminary numerical analysis of the modified and tp asymmetries indicates that some amount of cancellation tends to occur as one performs the integrations over phase space .
these cancellations would tend to make the asymmetries quite small .
larger asymmetries are possible if one includes extra weighting factors to offset the cancellations .
alternatively , experimentalists could perform fits to differential asymmetries ( such as @xmath72 , for example ) .
our initial study indicates that asymmetries of up to the order of a percent might be possible if the only assumption regarding the np scalar contribution is that it is `` hidden '' in the current experimental uncertainty in the branching ratio . incorporating the cleo bound on the scalar coupling coming from @xmath73 @xcite and making a reasonable estimate for the scalar form factor , we find that the cp asymmetries are likely to be smaller than this . further analysis and refinement of the numerical estimates
will be provided in ref .
@xcite .
[ * note added : * as noted in ref .
@xcite ( and contrary to the statement made above ) , the cleo bound does not actually place a direct constraint on the np coupling considered here . in the notation of ref .
@xcite , the cleo experiment placed a bound on @xmath74 , while @xmath75 probes @xmath76 . for more details
, please see ref .
we have considered cp violation in certain @xmath69 and @xmath77 @xmath8 decays . in both cases
cp violation requires the interference of ( at least ) two amplitudes that have a differing weak phase .
one amplitude is provided by the usual sm w - exchange diagram .
the other amplitude is assumed to be due to a np higgs - exchange diagram .
while the regular rate asymmetries are expected to be very small in these decays , larger asymmetries can be obtained by forming triple products or by considering other modifications to the usual rate asymmetries .
it is a pleasure to thank those with whom i have collaborated on this work
a. datta , k. little , d. london , m. nagashima , p.j .
odonnell and a. szynkman .
i also thank the conference organizers for an excellent conference .
this work was supported in part by the u.s .
national science foundation under grants no .
phy-0301964 and no .
phy-0601103 .
k. kiers , k. little , a. datta , d. london , m. nagashima and a. szynkman , arxiv:0808.1707 [ hep - ph ] .
r. decker , z. phys .
c * 36 * , 487 ( 1987 ) .
r. decker and e. mirkes , z. phys .
c * 57 * , 495 ( 1993 ) .
j. h. kuhn and e. mirkes , z. phys .
c * 56 * , 661 ( 1992 ) [ erratum - ibid .
c * 67 * , 364 ( 1995 ) ] .
r. decker , e. mirkes , r. sauer and z. was , z. phys .
c * 58 * , 445 ( 1993 ) .
d. m. asner _
[ cleo collaboration ] , phys .
d * 62 * , 072006 ( 2000 ) [ arxiv : hep - ex/0004002 ] .
u. kilian , j. g. korner , k. schilcher and y. l. wu , z. phys .
c * 62 * , 413 ( 1994 ) .
g. bonvicini _ et al . _
[ cleo collaboration ] , phys .
lett . * 88 * , 111803 ( 2002 ) [ arxiv : hep - ex/0111095 ] . | we examine cp violation in the @xmath0 decays @xmath1 and @xmath2 and the @xmath3 decay @xmath4 .
we assume that the new physics is a charged higgs .
we show that sizeable cp - violating effects are possible in @xmath2 ( polarization - dependent rate asymmetry ) and @xmath5 ( triple - product asymmetry ) .
the @xmath6 decay @xmath7 can proceed via several resonances .
we construct two modified rate asymmetries and a triple product asymmetry for this decay and discuss the potential sensitivities of these asymmetries . |
effective equations are useful tools to analyze quantum theories by adding suitable correction terms to classical equations , for instance by using equations of motion for expectation values of operators of interest . just like expectation values , effective equations that describe their evolution
depend on the states or class of states whose evolution is being approximated .
the low - energy effective action often used in particle physics , as perhaps the best - known example , describes quantum corrections for states near the vacuum of the interacting theory . for other states ,
different quantum corrections arise .
effective equations are state - independent only in rare cases such as free theories or the harmonic oscillator , when no quantum back - reaction and therefore no dynamical quantum corrections occur
. such systems also exist in ( loop ) quantum cosmology , where they play a similar role as the base order for perturbation theories when interactions are present @xcite . as an example for potentially large perturbative effects in quantum cosmology
, it has been suggested that classically ever - expanding models could eventually collapse due to quantum corrections @xcite .
such a feature is unexpected because the more the universe expands , the more classical it is supposed to become , with quantum corrections playing smaller and smaller roles .
however , with long evolution involved , small quantum effects could potentially add up and eventually drive the universe into collapse even in semiclassical regimes .
as shown in @xcite , the outcome depends on what semiclassical quantum state is realized and how its fluctuations of different variables change in time . to test the proposed scenario
, one needs information about dynamical semiclassical states over long evolution times , an issue that requires good control on quantum evolution and , when addressed with effective equations , a systematic scheme of going beyond the leading classical order .
the effective equations used in @xcite ( as well as @xcite ) were an extension of the classical limit obtained from methods provided in @xcite .
this scheme of going beyond classical order , however , is not entirely systematic because assumptions about the evolving state must be made ; they have not been self - consistently derived within the scheme .
for this reason , the suggested collapse remained a possibility but could not be demonstrated firmly .
( note that going to the other extreme , the high densities of the big bang , requires even better control of evolving states no longer required to be semiclassical .
for this reason , the high - curvature regime of quantum cosmology remains poorly understood , even though rather definite - looking statements are sometimes attempted . ) in this article , we use a systematic perturbative framework to derive information about effective equations and dynamical semiclassical states and see whether the suggestions of @xcite can be realized .
we will be led to a more systematic look on factor - ordering ambiguities than provided so far .
although the low - curvature regime can be formulated in wheeler
dewitt quantum cosmology @xcite and does not require additional corrections and modifications from extensions such as loop quantum cosmology @xcite , we will follow @xcite and use the latter for a general discussion
. this choice also helps us keep our notation close to that of @xcite . in brief terms ,
loop quantum cosmology of spatially flat isotropic models proceeds by formulating the friedmann equation in canonical variables of the homogeneous region considered for a minisuperspace model , normalized to @xmath1 .
classically , @xmath2 can be changed ( non - canonically ) without affecting the dynamics , but this feature is violated after quantization , where no unitary transformation exists to change @xmath2 @xcite .
this issue is a minisuperspace limitation which we can ignore in this paper because we will be interested only in qualitative effects .
note that no minisuperspace quantization can overcome this limitation , although , compared to loop quantum cosmology , it is less severe in wheeler dewitt models with continuous minisuperspace . ] @xmath3 and @xmath4 , @xmath5 and modifying it by using periodic functions of @xmath6 ( or , as seen in more detail below , @xmath7 with some function @xmath8 ) instead of @xmath6 . in this way
, one mimics the fact that the full theory of loop quantum gravity @xcite provides operators only for holonomies of the connection corresponding to @xmath6 , not for the su(2)-connection itself @xcite .
the ( almost ) periodic nature of functions substituting @xmath6 is motivated by a reduction of the internal gauge freedom from su(2 ) to u(1 ) when isotropy is imposed @xcite .
recent attempts to bring about a closer relation between minisuperspace models of loop quantum cosmology and the full theory have questioned the almost periodic nature of isotropic holonomies as a reliable feature of loop quantum gravity @xcite , as a consequence of crucial differences between abelian and non - abelian treatments @xcite . in this paper
, we aim to investigate a question that can be posed within pure minisuperspace models , whose form and quantum modifications we will take for granted .
we are not claiming that the features analyzed are in any way related to the full theory of loop quantum gravity .
such a modification to @xmath9 is most relevant in regimes where @xmath7 is large ( at large curvature ) .
if one expands the sine function , terms beyond the classical ones may be interpreted as contributions to higher - curvature terms .
( they are not complete as higher - curvature corrections , however , because higher time derivatives have not been included . the latter would result from quantum back - reaction @xcite . ) from equations of motion generated by ( [ modconstraint ] ) used as a hamiltonian constraint in proper - time gauge , one finds that the modified friedmann equation in terms of @xmath10 or @xmath11 takes the form @xmath12 with @xmath13 @xcite .
if one assumes @xmath14 as in @xcite , @xmath15 is a constant and planckian for @xmath16 .
this form makes it especially clear that the modification is relevant only at large density .
loop quantum cosmology represents the modified constraint ( [ modconstraint ] ) as a difference equation @xcite , obtained from shift operators @xmath17 acting on a wave function in the triad representation . from this difference equation or the underlying hamiltonian - constraint operator , ( [ modconstraint ] )
can be reproduced as the `` classical '' limit by computing the expectation value in gaussian states and ignoring all fluctuation - dependent terms @xcite . from the point of view of quantum dynamics , the result is the classical limit because fluctuations and their back - reaction on expectation values are ignored .
still , the result does not agree with the classical equation ( [ friedmann ] ) because loop quantum gravity implies quantum - geometry corrections in addition to the usual dynamical quantum corrections , which give rise to the modifiction ( [ modconstraint ] ) of ( [ friedmann ] ) .
only a combined classical and low - curvature limit @xcite brings loop quantum cosmology fully back to the classical friedmann equation ( [ friedmann ] ) .
( see also @xcite . ) the modification in ( [ modconstraint ] ) alone does not affect late - time evolution much .
however , quantum evolution implies additional corrections , which in canonical models can be computed by quantum back - reaction of fluctuations and higher moments of a state on expectation values @xcite .
the leading - order expectation value that gives rise to the classical limit must be expanded systematically to include these terms .
such terms were computed in @xcite , with potentially significant low - curvature implications .
in addition to the usual quantum corrections on small distance scales , higher - order corrections may allow for the possibility of an expanding universe to collapse at large size when the energy density is suitably small , even in the absence of spatial curvature a significant departure from classical theory .
( the possibility of a recollapse was also conjectured in @xcite by an alternative quantum hamitonian scheme , as well as in @xcite by a coherent state functional integral approach . in the latter cases , assumptions on the asymptopic semiclassical behavior of coherent states
were more restrictive . ) in @xcite , lacking a derivation of properties of dynamical semiclassical states , a suitable coherent state for large volume and late times had to be chosen , assumed , as in most cases , to be an uncorrelated gaussian . with this choice
, the classical - limit scheme of @xcite was developed further to calculate fluctuations and expectation values .
the end result is a modified friedmann equation , which we write here as @xmath18 \,.\ ] ] in this expression , @xmath19 is the quantum fluctuation of curvature or the hubble parameter @xmath20 , while @xmath21 is the relative volume fluctuation , using @xmath22 .
( unlike @xmath23 , the latter is not invariant under a change of spatial coordinates or of @xmath2 . ) as mentioned , eq .
( [ ym friedmann ] ) was derived with a gaussian ansatz for the wave function with zero correlations of the canonical pair @xmath24 , for which @xmath25 , the inverse curvature fluctuation , is proportional to the volume fluctuation @xmath26 thanks to the uncertainty relation .
the last term in ( [ ym friedmann ] ) is especially important for the question posed here , because it has the lowest power of the energy density @xmath27 .
the equation has the correct classical form when all fluctuations are sent to zero first , and then @xmath15 to infinity .
the semiclassical behavior , in which fluctuations are small but not zero , agrees with the classical one provided @xmath19 goes sufficiently fast to zero as the energy density decreases in an expanding universe . for a gaussian state
, @xmath28 would increase accordingly , but in ( [ ym friedmann ] ) it is always suppressed by factors of @xmath29 in @xmath23 .
indeed , one can solve the quantum model in a specific factor ordering and find properties of dynamical coherent states @xcite . in this harmonic quantum model , with the same classical dynamics , @xmath19 must decrease at low curvature and can not remain constant , while @xmath23 is asymptotically constant for any state @xcite .
this behavior may , however , change when the model is quantized with a different choice of factor ordering .
the question is whether it is then possible for @xmath19 to decrease sufficiently slowly that the last term in eq .
( [ ym friedmann ] ) can cancel the others and enable a recollapse , where @xmath30 at large @xmath31 . in the present context , the important part to note from eq .
( [ ym friedmann ] ) is that at late time and large volume , for an expanding universe with suitably low energy density ( @xmath32 ) , the universe could perhaps collapse .
an inverse-@xmath27 quantum correction associated with @xmath19 in ( [ ym friedmann ] ) may seem surprising .
but one way of looking at this is by distributing the @xmath27 in front to see that the final term is independent of the energy density and only depends on the quantum fluctuations .
it functions as a negative cosmological - constant term if @xmath19 varies slowly .
when the energy density is of the order of those fluctuations , the final term dominates and could lead to collapse depending on the exact quantum dynamics . from the perspective of effective equations , the classical model quantized in @xcite
a spatially flat isotropic model sourced by a free , massless scalar is identical to a harmonic one , without any quantum back - reaction @xcite .
the only corrections to classical evolution are due to quantum geometry ; no quantum back - reaction should occur . however , compared to @xcite , @xcite used a different factor ordering of the quantum constraint as well as quantum - geometry modifications of inverse - triad type , not just holonomy corrections as in ( [ modconstraint ] ) .
these variations imply quantum back - reaction , which , thanks to the proximity to a harmonic model , should not be large but could still be significant after long evolution . in this paper , we explore the possibility of such a term in the modified friedmann equation found by embarking on a systematic method of deriving effective equations , including the underlying properties of dynamical semiclassical states @xcite . by way of the example of collapse solutions
, we will therefore study several implications of terms in effective equations , including the role of factor ordering .
we will provide a general parameterization of these ambiguities in an expansion of the quantum - corrected hamiltonian constraint by powers of @xmath20 , which turn out to require inverse negative powers as well .
these methods will require a shift in viewpoint , focusing on the form of the last term in ( [ ym friedmann ] ) instead of the possibility of a recollapse which is difficult to assess in an expansion with negative powers of @xmath20 .
this fact highlights how subtle the semiclassical limit of quantum cosmology can be . in particular , it is not clear whether ( loop ) quantum cosmology has the standard semiclassical behavior even in isotropic models .
in spatially flat isotropic models , the basic pair of canonical variables is the extrinsic curvature @xmath33 and its conjugate @xmath34 where @xmath35 is the barbero - immirzi parameter @xcite of loop quantum gravity , @xmath36 is the scale factor , and @xmath37 is proper time . the momentum @xmath38 is derived from an isotropic densitized triad and so can take both signs due to the triad orientation .
the fundamental poisson bracket is @xmath39 where @xmath40 is newton s constant as before .
( as already mentioned , we assume the coordinate volume of our homogeneous region to equal one . )
beginning with the friedmann equation for an isotropic and homogeneous background , we can write the hamiltonian constraint in these canonical variables as ( [ friedmann ] ) with @xmath41 for a free , massless scalar @xmath42 where @xmath43 is the momentum conjugate to @xmath42 such that @xmath44 . as this work is motivated by @xcite , we will use a similar notation for a direct comparison , and so by a canonical transformation we introduce new conjugate variables : @xmath45 here , @xmath46 is a length parameter that determines the size of holonomy corrections . with
@xmath47 referring , as in @xcite , to the smallest non - zero area eigenvalue in loop quantum gravity .
owing to the lack of contact between loop quantum cosmology and the area spectrum in loop quantum gravity , we refrain from fixing @xmath48 and rather keep it as a free quantum parameter .
constructions of inhomogeneous models @xcite suggest that @xmath48 is not related to the general area spectrum but instead to detailed properties of the quantum state underlying quantum space - time . ]
looking at large - scale evolution of large @xmath49 , we are free to fix the sign of @xmath38 or @xmath31 , from now on assumed positive .
( these variables are a specific case of the general parameterization of @xcite , providing different cases of lattice refinement . while the harmonic model has a dynamics independent of the specific choice of basic variables where @xmath31 could be any power of @xmath38 , quantum corrections in models that differ for instance by factor ordering depend on such parameters . here , however , we confine ourselves to the choice made in @xcite because our aim is to explore the scenario suggested there . ) in the new variables , the classical hamiltonian constraint is @xmath50 we use the free scalar @xmath51 as an internal time variable ; the monotonicity of @xmath52 is guaranteed for the free case , and so evolution for @xmath53 and @xmath54 is governed by the hamiltonian @xmath55 which is found as a phase - space function by solving for @xmath43 in eq .
( [ constraint ] ) .
doing this , we have @xmath56 .
. [ factor ordering ] we discuss ambiguities related to the choice of internal time . ) the momentum @xmath57 of internal time , as a function of the remaining degrees of freedom , acts as the hamiltonian generating evolution .
since it is quadratic , the system is equivalent to a harmonic one , which , like the harmonic oscillator , can be quantized to a model without quantum back - reaction . as a hamiltonian is @xmath42-independent and thus preserved .
if one makes sure that an initial state is supported only on the positive part of the spectrum of @xmath58 , it will remain so supported . on the evolved state
the action of the quantized @xmath58 then agrees with that of the quantized @xmath59 .
] however , in a model of quantum gravity there may be quantum - geometry corrections as well , changing the hamiltonian not just by quantum back - reaction terms but also by more - severe modifications . in loop quantum cosmology , as one popular example , holonomy and inverse - triad corrections occur .
holonomy corrections replace the @xmath60 in @xmath57 by the non - linear @xmath61 , making the hamiltonian non - quadratic in canonical variables .
nevertheless , as shown in @xcite , surprisingly , they do not remove harmonicity even upon quantization provided a suitable factor ordering is chosen : if we define @xmath62 , the new set @xmath63 of ( non - canonical ) basic variables satisfies a linear algebra under poisson brackets , in which the _ linear _ hamiltonian @xmath64 is just the right holonomy - modified version of the internal - time hamiltonian @xmath65 .
these properties are preserved when the model is quantized with hamiltonian @xmath66 , in this specific ordering when seen as a function of @xmath60 and @xmath31 . with a linear hamiltonian in variables obeying a linear algebra
, no back - reaction occurs .
( again , the absolute value turns out not to spoil harmonicity features . ) unlike the harmonic oscillator in quantum mechanics , the quadratic hamiltonian here is subject to factor ordering choices , and only one of them , as remarked in @xcite , results in an exactly solvable quantum system as just described . at the quantum level , @xmath67 in the specific form @xmath68 .
( in the wheeler
dewitt limit @xmath69 in which holonomy corrections disappear , the ordering reduces to the standard symmetric one , @xmath70 . )
any other choice gives rise to quantum back - reaction , as spelled out in more detail below .
moreover , there is a second type of quantum - geometry modifications inverse - triad corrections @xcite which implies quantum back - reaction even if the original ordering is chosen in which holonomy corrections alone would not spoil harmonicity .
( in this case , the @xmath31- rather than @xmath60-dependence of the classical friedmann equation is modified . ) to derive effective equations from our classical hamiltonian above , we follow the general procedure of @xcite . for semiclassical ( or even more general ) states , the method of effective equations allows us to avoid the technical difficulties of working directly with wave functions and operators .
also the hard problem of deriving physical hilbert spaces and explicit representations for constrained systems is avoided , and yet physical normalization is implemented by reality conditions @xcite .
large classes of states can be dealt with by suitable parameters , such as expectation values and fluctuations , of direct statistical significance for physical properties .
effective equations therefore allow for much larger generality , and thereby more reliable conclusions , than traditional methods used in quantum cosmology .
our analysis of factor - ordering ambiguities provides a further example . instead of working with wave functions
, we have a framework in which states are represented by the expectation values and moments they imply for basic operators . as quantum variables in addition to expectation values
, we use the moments which describe a general quantum state : @xmath71 where @xmath72 are positive integers such that @xmath73 and the subscript weyl denotes totally symmetric ordering of the operators .
the moments and expectation values define a phase space with poisson bracket @xmath74\rangle}{i\hbar}\ ] ] in terms of the commutator , extended by the leibniz rule to arbitrary polynomials of the expectation values as they occur in moments .
semiclassical regimes can be defined very generally to any integer order @xmath75 , by keeping moments only up to order @xmath76 .
indeed , in a gaussian state , the prime example of a semiclassical one , the moments behave as @xmath77 however , our semiclassical regimes defined by the order of moments are much more general than the 1- or at most 2-parameter families of gaussians , and thus avoid possible artifacts due to the selection of states . the basic identity @xmath78\rangle / i\hbar$ ] of quantum mechanics , used for instance to derive ehrenfest s equations ,
can then be written as a classical - type hamiltonian flow generated by the quantum hamiltonian @xmath79 where the expectation value is computed in a state specified by expectation values @xmath80 and @xmath81 and all its moments @xmath82 .
such expectation values of interacting hamiltonians may be difficult to compute , but in semiclassical ( or other ) regimes in which only finitely many moments matter , they can be derived perturbatively by expansions such as @xmath83 ( we define @xmath84 and @xmath85 as a short cut used from now on . ) in this particular expression , we assume the hamiltonian operator @xmath86 to be weyl ordered just as the moments .
if there are reasons for working with a different ordering , as explored below , it differs from the weyl - ordered one by re - ordering terms which can be expressed by the moments as well .
accordingly , there will be additional quantum corrections in ( [ hq ] ) .
for instance , if @xmath87 , the weyl - ordered quantization would be @xmath88 .
another symmetric ordering is @xmath89 .
using @xmath90 ^ 2 $ ] and @xmath91 ^ 2 $ ] for canonical commutators of @xmath92 and @xmath93 , we have @xmath94 ^ 2 $ ] . the effective hamiltonians @xmath95 and
@xmath96 differ just by constants , but for higher polynomials moment - dependent factor - ordering terms can result . we will present a more detailed example below . in general , any ordering of a term @xmath97 in a polynomial ( or series - expanded )
@xmath86 can be written as @xmath98 by a finite number of applications of @xmath99=i\hbar$ ] .
effective hamiltonians of differently ordered hamiltonian operators therefore differ only in terms containing explicit factors of @xmath100 .
only even powers of @xmath100 occur since @xmath99=i\hbar$ ] must be applied an even number of times to relate symmetric orderings , ensuring that the associated effective hamiltonians are real . quantum evolution
determined from commutators with the hamiltonian operator is the hamiltonian flow of @xmath101 , generating equations of motion for expectation values and quantum moments : @xmath102 using ( [ poisson ] ) . unless the classical hamiltonian is quadratic in canonical variables , the last term in ( [ hq ] ) contains products of expectation values and moments .
their equations of motion are then coupled , with moments back - reacting on the dynamics of expectation values as a source of quantum corrections .
re - ordering terms added to ( [ hq ] ) for a non - weyl ordered hamiltonian thereby affect the effective dynamics . in this paper , we will work to leading order @xmath103 , and ignore higher moments as they are subdominant for semiclassical states .
( for analyses of cosmological quantum back - reaction using this scheme with moments of different orders , see @xcite . )
moreover , as we will see below , they appear with factors which are suppressed by inverse powers of @xmath31 and become even more suppressed at large volume .
still , to leading order there is quantum back - reaction of fluctuations and correlations in expectation values . to compare evolution equations with a single friedmann equation of the usual form
, we should solve for all moments included and insert the solutions in equations of motion for expectation values .
but as we are , for now , looking at particular terms in a general formulation , we do not have an explicit system of equations to solve .
we are extending the typical analysis by including three forms of corrections the holonomy replacement , factor ordering ambiguities , and inverse triad corrections which we will now define and summarize .
the holonomy replacement is motivated from the full theory of loop quantum gravity @xcite , as already discussed .
furthermore , holonomy corrections are of interest here because we are looking to reproduce the @xmath104 term of eq .
( [ ym friedmann ] ) . in our framework , @xmath105 is the squared variance of @xmath60 .
thus , if we are looking for a @xmath106 dependence in our corrections to the friedmann equation , we require @xmath20 to have a non - zero third derivative with respect to @xmath60 .
( from eq .
( [ hq ] ) we see that to obtain a @xmath106 term we need at least a non - zero second derivative , and as we will see in more detail , we need an additional derivative of @xmath60 to get @xmath106 in the equations of motion from the poisson bracket . )
the holonomy replacement easily satisfies this requirement , but less obviously it may be realized also in wheeler
dewitt quantum cosmology with a non - standard factor ordering .
inverse - triad corrections modify the @xmath31-dependence of the hamiltonian . the hamiltonian constraint ( [ friedmann ] ) to be quantized contains inverse powers of @xmath31 . also here , as with holonomy corrections , the full theory suggests characteristic modifications by elementary properties of the form of quantum geometry realized .
the variable @xmath31 , or rather @xmath38 as the isotropic component of the densitized triad used as a basic field in the full theory , is quantized to flux operators with discrete spectra containing zero .
such operators have no densely defined inverse , and therefore one must use more indirect quantization techniques , provided in @xcite to represent inverses of @xmath31 or @xmath38 in the hamiltonian constraint . in isotropic models ,
the resulting correction functions can be computed explicitly @xcite ; they appear in quantized hamiltonian constraints and determine , for instance , the coefficients of dynamical difference equations . for our present purposes ,
we take these effects into account by writing the @xmath31-factor of @xmath65 as a general function @xmath107 which , in the large volume regime considered here , has the form @xmath108 .
a commonly used parameterization , for instance in phenomenological analysis , is @xmath109 with @xmath6 real valued and @xmath110 a positive integer . summarizing holonomy and inverse - triad corrections , we find our hamiltonian with corrections to be @xmath111 the next step is to use this example hamiltonian to evaluate the dynamics of our system and arrive at an effective friedmann equation .
following the effective - equation prescription , we find our quantum hamiltonian , up to second order in moments , to be @xmath112\ ] ] where the prime represents a derivative with respect to @xmath31 .
as mentioned in the general discussion of effective equations , the specific terms written here assume that the hamiltonian operator is weyl ordered in @xmath92 and @xmath93 .
at this stage , factor - ordering choices may introduce additional quantum corrections which also depend on the moments , as discussed in more detail below . from ( [ hq ] ) , we can find the relational equations of motion for our expectation values via poisson brackets .
for instance , we have the expectation value of the volume in terms of expectation values and moments : @xmath113 .
\end{aligned}\ ] ] here our moments back - react on the classical trajectory .
the moments are not constant and subject to their own equations of motion , but we will not need the latter for the subsequent analysis . now let us find the quantum corrections to the friedmann equation for our example . from eq .
( [ b and v ] ) and the definition of @xmath34 , we have @xmath114 where the proper - time derivative @xmath115 is generated by the original hamiltonian constraint .
( at low curvature , we need not worry about signature change @xcite which could eliminate time . ) rearranging and substituting , we find @xmath116 where we used eqs .
( [ nudot ] ) and ( [ hq ] ) , now with @xmath117 , to arrive at the second line and dropped squared moments as well as the @xmath118 and @xmath119 terms for simplicity . note
that these derivative terms are not neglected in the actual analysis to follow . in terms of energy density
we have @xmath120 where we again used eq .
( [ hq ] ) for @xmath121 . rearranging to solve for @xmath122
, we can substitute into eq .
( [ friedmann ] ) to find @xmath123 the fluctuation added to @xmath27 in the combination @xmath124 is , within the approximations used here , analogous to the correction found in @xcite ( see also @xcite ) .
in addition to the holonomy replacement and inverse - triad corrections , we include the possibility of a factor - ordering ambiguity as the main ingredient in this paper .
none of these corrections are expected to be very large in low - curvature semiclassical regimes , but they could play a role in long - term evolution . to show that quantum cosmology has the correct large - scale behavior , all these terms must be studied carefully , but no such analysis has been completed yet . here , we present systematic expansions by which the issue of long - term behavior can be addressed , focusing specifically on the question of whether semiclassical effects can be so strong that they trigger a recollapse of a classically ever - expanding universe . the hamiltonian constraint of ( loop ) quantum cosmology ( or loop quantum gravity in general ) is far from being unique , owing for instance to choices in the representation of holonomies and factor orderings .
although uniqueness claims seem to exist in the literature , they are based on certain mathematical features posed ad - hoc , rather than physical considerations ; they can not be used when addressing the semiclassical limit , long - term evolution , or concrete physical effects .
in addition to ambiguities in the formulation of the hamiltonian - constraint operator , further choices are usually made when one addresses the problem of time .
one chooses or even introduces simple matter degrees of freedom , here and elsewhere assumed as a free and massless scalar field , whose momentum is exactly constant and , in particular , never vanishing .
the scalar degree of freedom therefore has no turning points when solved for as a function of proper time , and can be used as a mathematical evolution parameter itself .
simple and manageable equations at the classical or quantum levels usually result .
however , the question of whether different choices of internal times provide the same physics remains complicated to address ; see e.g.@xcite .
( an exception is given by semiclassical regimes , where effective methods allow one to change time by gauge transformations of moments @xcite . ) if independence of one s choice of internal time can be achieved at all , as some kind of space - time anomaly freedom , it likely requires detailed prescriptions of factor orderings and other quantization ambiguities .
the hamiltonian one may choose in a deparameterizable version in rather simple terms may not be what is required for the correct physics .
therefore , a wider view on factor ordering and other choices is preferable for reliable statements , implemented here by general parameterizations . instead of directly quantizing @xmath125 with rather controlled factor ordering options , or in the harmonic formulation the even more unique - looking @xmath126
, we should start with the hamiltonian constraint ( [ modconstraint ] ) and its terms @xmath127 and @xmath128 .
the types of factor - ordering ambiguities differ in those two cases , showing the restricted nature of deparameterized models regarding factor ordering .
it is not easy to parameterize all possible ordering effects , and therefore we will present the main implications by key examples . in the former case , looking at possible orderings of the @xmath42-hamiltonian @xmath125
, we could take a symmetric ordering such as @xmath129 the weyl ordering of operators @xmath92 and @xmath130 but also @xmath131 or a more complicated re - ordering .
the first two differ by @xmath132 + [ \sqrt{\hat{\nu}},\widehat{\sin(b)}]\sqrt{\hat{\nu}})= \frac{1}{2}[\sqrt{\hat{\nu}},[\widehat{\sin(b)},\sqrt{\hat{\nu}}]]$ ] .
such terms , in a low - curvature expansion , just add positive powers of @xmath60 with ordering - dependent coefficients .
the classical correspondence of the ordering term just derived , for instance , is @xmath133 as in this example , iterated commutators , or poisson brackets in the classical correspondence , amount to terms of the order @xmath134 to some positive power , by which various orderings differ .
( a single commutator contributes a factor of @xmath135 .
however , there must always be an even number of commutators if the difference of two symmetric orderings is considered to avoid factors of @xmath136 . )
such terms can be interpreted as changing the form of inverse - triad corrections in @xmath107 or the choice of holonomy corrections initially implemented by the sine function , and therefore do not differ much from the usual ambiguities of holonomy corrections .
factor - ordering ambiguities can , however , be more radical if we start with the original hamiltonian constraint rather than the deparameterized @xmath42-hamiltonian .
one could first take the square root of the quantized @xmath137 , for instance of @xmath138 , and then multiply with @xmath139 to obtain a version of @xmath140 .
factor - ordering choices now include expressions such as @xmath141 which , compared with the previous ones , differ by terms that require some quantization of the inverse @xmath142 .
it is more diffcult to compute commutators of square - root operators to see what factor - ordering terms now arise .
however , we can estimate contributions in an @xmath100-expansion , suitable for effective equations , by noting that arguments of the square root such as @xmath143 differ from squares of our previous hamiltonians by double commutators as in ( [ reord ] ) , but involving a power of @xmath31 and @xmath144 ( not @xmath145 as before ) .
for instance , comparing the two orderings @xmath138 and @xmath146 , we have @xmath147\right]\nonumber\\ & \sim & -\frac{\hbar^2}{2\nu } ( \sin(b)^2-\cos(b)^2)= -\frac{\hbar^2}{2\nu}\cos(2b)\ , .
\label{double}\end{aligned}\ ] ] in the semiclassical correspondence , such a re - ordering term added to @xmath148 under the square root implies a leading @xmath100-correction of @xmath149 which is not analytic at @xmath150 .
( semiclassical physics follows from an expansion in @xmath100 , assumed to be small .
but when @xmath60 is small too , the square - root is evaluated near its non - analytic point . )
such inverse - power terms will give us more options for moments in an effective hamiltonian possibly resembling the one used in @xcite .
if we include inverse - triad corrections by a generic @xmath107 instead of @xmath31 , we will have double commutators of the form @xmath151 $ ] .
the @xmath29 in ( [ double ] ) will then be replaced by @xmath152 .
taking the square root and expanding as in ( [ root ] ) will produce a series @xmath153 this series would be used as the gravitational contribution to the effective hamiltonian constraint , which is to equal the effective matter contribution @xmath127 . also here , in the matter contribution , inverse - triad corrections @xmath154 may result , in general by a function @xmath155 that differs from @xmath156 in higher - order terms . assuming the ordering just discussed , the effective @xmath65 generating @xmath42-evolution is then @xmath157 the types of ordering just presented are more involved than simple ones when one directly quantizes @xmath125 .
however , they are no less natural ; in fact , they could be argued to be more natural from the point of view that the terms in the original hamiltonian constraint , @xmath29 and @xmath128 , should first be quantized , with a square root for @xmath65 taken afterwards at the quantum level . for full generality and to guarantee independence of one s results from special features available only in deparameterized systems but not otherwise one must take the properties of the orderings shown here into account .
this realization makes the low - curvature regime of quantum cosmology much more complicated than it may naively seem . in a low - curvature expansion of expressions such as ( [ root ] ) , even inverse powers of @xmath60 can appear due to factor ordering .
this feature may sound surprising because such terms seem unduly large in semiclassical regimes .
however , they are always accompanied with factors of @xmath100 and , in some cases , moments such as the curvature fluctuation which decrease as @xmath60 gets small .
the fate of long - term evolution now crucially depends on properties of state parameters and the detailed form of quantum corrections and effective equations , which requires a dedicated analysis on which we now embark .
so far , we have kept all terms expected from loop quantum cosmology , but they will not be fully required in the semiclassical regimes we are interested in . in the small - curvature regime , we can expand the hamiltonian in powers of @xmath60 : @xmath158 where the @xmath159 are functions of @xmath31 only .
such an expansion allows us to consider generic orderings .
we include inverse powers of @xmath60 as they may arise from factor orderings ; these terms therefore all include factors of @xmath100 .
( after expanding all trigonometric functions in ( [ hexp ] ) , also the positive - power terms in the @xmath60-expansion will be corrected by reordering terms with explicit factors of @xmath100 . )
they will appear explicitly in the effective hamiltonian , added to the original ( [ hq ] ) which is based on an operator in which all contributions are weyl ordered .
the effective hamiltonian will contain moments for all terms in the curvature expansion , including the inverse - power ones . at small curvature
, these latter terms will become progressively more important .
we will illustrate our main analysis based on the truncation shown in ( [ general h ] ) , which will not allow us to address the recollapse ( where @xmath150 )
. however , our approach will be sufficient to investigate the possible terms that can appear in a general effective friedmann equation for a model without ad - hoc assumptions . in ( [ general h ] ) ,
the linear term @xmath160 is what is expected classically .
the cubic term is the first higher - order quantum correction due to the holonomy modification ; even higher orders are not relevant at low curvature .
for very small and very large curvature , respectively , we should expand to higher orders of positive and negative powers in @xmath60 , but as we will discuss in the analysis , this does not lead to categorically different terms . from ( [ general h ] ) , we can then solve for the equation of motion for @xmath31 @xmath161 to second order in moments , which we expand term by term to find @xmath162 where @xmath163 , @xmath164 , and we define @xmath165 in the first equation .
now we have @xmath166 expanded in powers of @xmath60 .
recall however , that for the friedmann equation , we are interested in @xmath167 which requires us to look at the expansion of @xmath168 . for terms of interest ,
those that depend on the second - order moments but are independent of @xmath27 , we can limit our search to terms with @xmath169 and @xmath170 which would resemble the term responsible for a potential recollapse .
( this term , specifically , contains @xmath106 but no factor of @xmath31 or @xmath60 .
there may be a small dependence on @xmath31 , but it can not be strong at large volume because : if there is a power - law dependence with positive exponent , the term would violate the semiclassical limit ; if the exponent is negative , the term will be suppressed at large volume . )
it turns out that the dependence on expectation values is already a restrictive condition , irrespective of the dependence on moments . from now on
, we can ignore the factor of @xmath171 in ( [ hexp ] ) because ( in quadruplicate ) it simply multiplies the right - hand side of ( [ friedmann2 ] ) without changing the leading - order dependence on @xmath31 .
the simplest procedure is to look at all @xmath60-independent terms in @xmath172 and then restrict to all remaining ones proportional to @xmath173 to cancel the explicit @xmath31-dependence in ( [ friedmann2 ] ) . since both @xmath174 and @xmath175 itself are expanded as in ( [ general h ] ) , all expansion terms in the product are quartic expressions in the @xmath176 .
terms independent of @xmath60 but containing @xmath106 include , for instance , @xmath177 in the first example , the contribution @xmath178 comes from the last term in ( [ general nudot ] ) , @xmath179 from the second term in the same equation ( both factors appearing in one term of @xmath180 ) , while @xmath181 is one term in @xmath182 that does not depend on @xmath60 .
now we are interested in what is necessary such that the combination of @xmath183 gives us terms with @xmath184 , thus satisfying the @xmath27 independence , which could dominate the friedmann equation in a regime with low energy density .
with that in mind , and looking at ( [ hexp ] ) , let us categorize the @xmath31-dependence in the @xmath176 terms as @xmath185 where these associations come from the power expansion of our original @xmath20 : @xmath186 are the coefficients of @xmath187 , @xmath188 and @xmath189 arise from the factor ordering implementation , each additional inverse power of @xmath60 requiring another commutator of @xmath190 with @xmath191 as in ( [ rootexpand ] ) . recall that @xmath192 is due to the inverse triad corrections , @xmath193 .
although we have an expectation for the form of @xmath107 we do not presume it here ; rather , we will solve for the necessary values that provide @xmath27 independence and compare for consistency .
with these classifications , we can then introduce a simple expression for the combinations @xmath194 depending only on the integer @xmath195 for all possible values of @xmath165 , @xmath196 , @xmath197 and @xmath198 .
( note that from above @xmath186 , a relation which extends to all higher - curvature terms @xmath199 with @xmath200 obtained from expanding @xmath61 .
for this reason , we use the minima of the order and one to compute the power of @xmath201 .
each @xmath176 then contributes @xmath202 factors of @xmath107 and @xmath203 factors of @xmath201 . ) the maximum of @xmath204 according to its definition is @xmath205 which would be attained if and only if @xmath165 , @xmath196 , @xmath197 and @xmath198 are all positive .
however , this case can not eliminate all factors of @xmath60 that come along with the @xmath176 .
there must be at least one negative coefficient in allowed terms , lowering the maximum to @xmath206 .
this upper bound will play a crucial role in our subsequent arguments .
we are then interested in @xmath207 such that we have a @xmath208 dependence which cancels in eq .
( [ friedmann2 ] ) .
thus we can solve for @xmath107 with respect to @xmath31 to see what is required for a @xmath31-independent term in the effective friedmann equation .
we need @xmath209 or , ignoring for simplicity higher - order terms in a general power law @xmath210 , we must have @xmath211 or @xmath212 .
since @xmath204 is an integer at most 2 inclusive due to the possible combinations of our terms , the smallest possible exponent is @xmath213 increasing by @xmath214 as @xmath204 decreases in integer steps , reaching a maximum of @xmath215 for @xmath216 ( for the truncation used here ) . allowing for higher orders in the @xmath217-expansion ( [ general h ] ) will only increase the power of @xmath31 in @xmath107 .
none of these behaviors is compatible with the classical limit @xmath218 .
our expectation of @xmath107 due to inverse triad corrections is only compatible with @xmath219 in the case that there are no corrections or when @xmath31 is extremely large .
however , the @xmath220 case is never realized as it would require a four - term combination of only @xmath221 and @xmath222 that still gives @xmath169 in the effective friedmann equation .
we do have terms with only these two factors in @xmath182 but they correspond to positive , even powers of @xmath60 ( in fact , @xmath223 corresponds to terms with @xmath224 ) .
similarly , terms with only these two factors from @xmath180 also correspond to positive or zero powers of @xmath60 and thus the only way we could achieve a @xmath31-independent correction would require a @xmath225 or higher dependence .
for all other values of @xmath204 , we would require some @xmath226 with @xmath227 which , as mentioned above , is not compatible with our requirements on @xmath107 .
now let us return to the inclusion of higher derivative terms of @xmath107 and their effects on our analysis .
returning to our previous notation ( and using @xmath186 again ) we have : @xmath228 with the benefit of foresight , it is now prudent to extend eq .
( [ bbh ] ) to allow for more than four terms by defining @xmath229 where @xmath204 is the sum of the labels as defined before and @xmath75 determines how many @xmath176 terms are multiplied together .
while we will continue to focus on the @xmath230 case corresponding to our possible contributions to the effective friedmann equation , this parameter is important for categorizing derivatives . in fact , with this definition , we can then write the solution to terms which have derivatives as @xmath231 where the @xmath232 superscript represents that there is a @xmath233 term in the combination and we see that @xmath75 is altered in terms which have higher derivatives .
we could also write out a general formula for combinations with more than one derivative term ; however , terms which include two derivative factors are included ( up to numerical coefficients ) , and combinations that contain more derivative terms lead to inconsistent conditions on @xmath107 , as we will see in eq .
( [ generalized ansatz ] ) .
we can also express terms with second derivatives : @xmath234 for @xmath235 .
now that we have these expressions , we once again must solve for @xmath107 such that we have a @xmath31-independent term in the effective friedmann equation .
these solutions are obtained from differential equations with non - linear products of varying orders of derivatives .
we can generalize these differential equations from our terms as : @xmath236 assuming @xmath237 ( this describes all possible contribution terms ) and , without loss of generality , @xmath107 a power law , we have solutions given by @xmath238 with the added caveat that for @xmath239 .
all @xmath31-independent contributions to the corrected friedmann equation can be written as this type of differential equation and thus can only allow solutions given by eq .
( [ generalized ansatz ] ) .
in fact , as hinted at previously , the general form includes expansions in higher positive and negative powers of @xmath60 in eq .
( [ general h ] ) because the higher powers create more terms , but never qualitatively different results .
for instance , higher positive powers in @xmath60 are due to the power expansion of @xmath145 which will have the same @xmath31-dependence , that is @xmath240 for non - negative integer @xmath241 .
higher negative powers come from expansions in the factor ordering which leads to varying powers of @xmath242 , but this is included in the analysis above with an appropriate choice of @xmath204 and @xmath75 .
what then is the lowest possible exponent for @xmath31 ? for @xmath226
where @xmath243 we have : @xmath244 however , as discussed above , the @xmath220 case does not allow @xmath60-independence in the corrected friedmann equation .
additionally , @xmath245 and @xmath246 since they are formed from our higher derivative terms which always lead to positive powers .
thus , we see that all possible cases that could lead to a term independent of @xmath31 and @xmath60 in the effective friedmann equation require @xmath226 with @xmath247 which is incompatible with our expectation of @xmath107 within our framework for an isotropic , homogeneous universe with a free , massless scalar field .
we did not find a term independent of @xmath31 and @xmath60 even in the very generalized case of factor orderings , which would have allowed us to provide an explicit example for the effects suggested in @xcite .
however , with generic factor orderings considered , the large - volume regime in quantum cosmology turns out to be much more subtle than it appears in models with fixed ( and simple ) ordering .
in particular , the semiclassical expansion by powers of @xmath100 leads to terms not analytic in the curvature or hubble parameter , so that the low - curvature regime is not obvious .
while we were able to provide ( non - affirmative ) indications regarding the terms suggested in @xcite , the non - analytic nature of our equations makes it difficult to obtain rigid statements about the possibility or impossibility of a recollapse . in order to show some of the possible behavior in the approach to low curvature
, we now provide numerical solutions of our previous equations with specific choices of coefficients of non - analytic terms .
formally , these equations with a truncated @xmath248-expansion allow possible recollapses , but near the recollapse with small @xmath60 any such truncation breaks down .
only a non - truncated effective hamiltonian could provide a firm answer ; however , without expanding the square root it is more difficult to parameterize generic ordering ambiguities . to provide an example
, we will include only @xmath221 and @xmath188 terms and continue to focus on the regime of large volume and small curvature .
our expanded hamiltonian from eq .
( [ general h ] ) then becomes @xmath249 encapsulating only inverse - triad corrections and factor ordering , but no holonomy modifications .
( note that non - analytic terms are mainly due to taking a square root of a reordered hamiltonian ; they do not require the use of holonomies and can therefore be realized in a wheeler
dewitt quantization . ) the quantum hamiltonian to second order in moments is @xmath250 we solve for the full system of equations of motion by poisson brackets of our variables with the quantum hamiltonian , where the dot represents derivatives with respect to @xmath42 as before : @xmath251 as a further simplification , we truncate higher derivatives of @xmath176 as those that correspond to increasing inverse powers of @xmath31 . keeping @xmath252 and @xmath253 to first order
we are left with @xmath254 where the @xmath255 comes from @xmath188 and the particular choice in factor ordering .
our previous example had a negative sign , but in our general discussion the sign or even numerical factors did not play a role .
we can solve this system of equations analytically .
equations ( [ bneg ] ) and ( [ gbbneg ] ) are independent of the other variables , which greatly simplify the task .
we easily obtain @xmath256 and with this continue to solve for @xmath257 where our coefficients can be chosen to meet our initial conditions accordingly ( and to respect the uncertainty relation for moments ) .
note that depending on the sign of @xmath188 we get @xmath255 in the equations for @xmath258 and @xmath259 as well as slightly different entries for @xmath260 labeled by the @xmath255 subscript .
looking at eq .
( [ nuneg1 ] ) , we see that there is a possibility of negative @xmath42 within the squareroot , an indication that there could be turnover in the volume as @xmath42 shrinks or grows ( depending on the sign of the factor ordering correction of @xmath188 ) .
we also see that the constants under the square root will play an important role , with @xmath261 and @xmath262 .
is plotted with a @xmath263 axis .
the volume is exponentially increasing with @xmath264 over a large range of values until it begins a sharp descent when the square - root term of eqn .
( [ nuneg1 ] ) approaches zero at @xmath265 .
[ nucollapsehneg1],scaledwidth=80.0% ] is plotted with a @xmath263 axis .
it is well behaved over a large range of values but does increase sharply at @xmath266 and is increasingly larger for @xmath267 where @xmath258 becomes small .
[ gvvrelhneg1],scaledwidth=80.0% ] is plotted with a @xmath263 axis .
it is reasonably well behaved over a large range of values , though it increases unboundedly at large @xmath263 , which is inconsistent with our semiclassical approximations in that region .
[ gbvrelhneg1],scaledwidth=80.0% ] let us then explore this system numerically for @xmath188 positive , @xmath268 , noting that our initial values satisfy the uncertainty relation ( where we set @xmath269 for our numerics ) @xmath270 using our truncated equations of motion , eqs .
( [ bneg ] ) ( [ nunegend ] ) , we can easily verify that the uncertainty relation is preserved , that is @xmath271 and so will remain satisifed for all @xmath42 . in figure
[ nucollapsehneg1 ] , we see that our values do indeed lead to a collapse at late negative time @xmath42 ( plotted with @xmath42 increasingly negative to the right for convenience , as we will do with all of this section s numerics )
. however , for this plot to be meaningful , we must verify the consistency of our assumptions within this region . in figure [ gvvrelhneg1 ]
we see that the relative volume fluctuation @xmath272 is well behaved for most of @xmath42 , only increasing to the upper limits of our assumption of small relative fluctuations when @xmath258 becomes exponentially small for @xmath273 .
in fact , from our analytic solutions given by eqs .
( [ gnunu+ ] ) and ( [ nuneg1 ] ) , we see that the relative volume fluctuations @xmath274 asymptotically approaches a local maximum for @xmath275 : @xmath276 which is @xmath277 for the numerical example used for our graphs ( however , graphing the large @xmath42 regime numerically would require a very large number of integration steps ) . near the turning point of @xmath31 at @xmath278 , the relative volume fluctuation begins to increase , quickly surpassing our assumption of small relative values . from eqs .
( [ bneg1 ] ) we see that relative curvature fluctuations @xmath279 will remain constant according to where we set them by our constants @xmath280 and @xmath281 . in figure [ gbvrelhneg1 ]
we see that the relative covariance @xmath282 also surpasses our assumption of relatively small moments as it approaches the turning point of @xmath31 , becoming unboundedly large as @xmath31 collapses at large @xmath283 . in addition , higher inverse powers of @xmath60 contributing to the hamiltonian with generic factor ordering will become important near the potential recollapse .
thus , while we see collapse in figure [ nucollapsehneg1 ] , it coincides with a breakdown of our assumptions in the regime where @xmath60 becomes exponentially small : a regime where higher inverse powers of curvature can become important , even when paired with inverse - volume terms .
( we can also explore the opposite sign choice for @xmath188 negative which is what our example from subsection [ factor ordering ] had . in that case , the squareroot in @xmath258 of eq .
( [ nucollapsehneg1 ] ) goes to zero as @xmath284 .
however , rather than insinuating a recollapse , this is the regime where @xmath285 and @xmath286 gets exponentially large , indicating a singularity .
regardless of sign choice , in this regime we would require a higher - curvature expansion by positive powers of @xmath60 for holonomy ( or other ) corrections and quantum back - reaction before any physical meaning can be drawn from such analysis . )
this article discusses the influence of factor - ordering choices on effective equations in quantum cosmology , with possible modifications from loop quantization .
as we have emphasized , the quantization of the hamiltonian constraint is far from being unique even in the most reduced models . while deparameterized models may offer simple quantization choices , they are not the most general or natural ones . quantum corrections and an analysis of semiclassical physics
must therefore take these ambiguities into account .
we have implemented such an analysis to study the question of whether quantum effects in long - term semiclassical evolution could lead to significant departures from classical behavior , for instance a recollapse of spatially flat isotropic models .
since the model we analyzed allows a quantization free of quantum back - reaction , low - curvature effects could only come from factor - ordering corrections .
we developed a suitable parameterization of factor - ordering ambiguities in effective equations , paying special attention to deparameterization choices .
the generality of effective equations indeed allows us to draw conclusions , indicating that too - drastic effects do not occur .
our results therefore do not lead to concrete reasons for unexpected late - time effects in semiclassical ( loop ) quantum cosmology .
but they provide strong caution against too - quick positive assurances of a straightforward classical limit based on the analysis of a few simple models and orderings .
in particular , it does not appear possible to have quantum corrections of the modified friedmann equation independent of the energy density in this formulation .
thus , quantum collapse at low energy densities or large scales does not seem feasible by this mechanism given our corrections and assumptions .
the only possibility , according to our analysis , would be to have inverse - triad corrections with a function @xmath107 increasing more strongly than linearly , but this is not acceptable by the classical limit of the hamiltonian constraint .
furthermore , we explored the effects of factor ordering choices at first semiclassical order .
a general tractable parameterization of quantization ambiguities in this regime surprisingly turns out to require a second expansion by the inverse hubble parameter .
although the solutions to these effective equations hint at the possibility of a recollapse , the expansion by the inverse hubble parameter breaks down before a recollapse could be reached .
interestingly , also some relative fluctuations grow too large to satisfy our approximations for all @xmath42 of interest .
extending such an analysis to higher quantum moments could be fruitful ; for instance , unlimited semiclassical evolution may be possible within our expansion scheme if there are corrections to @xmath286 such that it does not asymptotically approach zero , thus preventing factor ordering terms @xmath248 to approach infinity .
for now , however , the semiclassical behavior of generic quantum cosmology remains incompletely understood .
we are grateful to yongge ma for several discussions and for hospitality at beijing normal university during an early stage of this project .
this research was supported in part by the nsf east asia and pacific summer institute fellowship to ds , by nsf grants phy-0748336 and phy-1307408 , and by perimeter institute for theoretical physics during a visit of mb .
research at perimeter institute is supported by the government of canada through industry canada and by the province of ontario through the ministry of economic development & innovation .
d. l. wiltshire , an introduction to quantum cosmology , in b. robson , n. visvanathan , and w. s. woolcock , editors , _ cosmology : the physics of the universe _ , pages 473531 .
world scientific , singapore , 1996 , [ gr - qc/0101003 ] m. bojowald , d. brizuela , h. h. hernandez , m. j. koop , and h. a. morales - tcotl , high - order quantum back - reaction and quantum cosmology with a positive cosmological constant , 84 ( 2011 ) 043514 , [ arxiv:1011.3022 ] | quantum cosmology implies corrections to the classical equations of motion which may lead to significant departures from the classical trajectory , especially at high curvature near the big - bang singularity .
corrections could in principle be significant even in certain low - curvature regimes , provided that they add up during long cosmic evolution .
the analysis of such terms is therefore an important problem to make sure that the theory shows acceptable semiclassical behavior .
this paper presents a general search for terms of this type as corrections in effective equations for a @xmath0 isotropic quantum cosmological model with a free , massless scalar field .
specifically , the question of whether such models can show a collapse by quantum effects is studied , and it turns out that factor - ordering choices in the hamiltonian constraint are especially relevant in this regard . a systematic analysis of factor - ordering ambiguities in effective equations
is therefore developed .
factor ordering and large - volume dynamics + in quantum cosmology + martin bojowald and david simpson + institute for gravitation and the cosmos , the pennsylvania state university , + 104 davey lab , university park , pa 16802 , usa + |
planck hatte sich schon in jungen jahren ein ehrgeiziges forschungsprogramm zurechtgelegt .
er wollte den sogenannten 2.hauptsatz ) .
der 2.hauptsatz besagt in dieser interpretation , da ' 's die entropie _
i m zeitmittel _ nicht abnimmt ( statistische schwankungen , in denen die entropie vor``ubergehend kurz abnimmt , sind also erlaubt ) .
der 2.hauptsatz regelt die irreversibilit''at gewisser prozesse .
das sind dann solche , bei denen die entropie zunimmt . ]
der thermodynamik mit hilfe der theorie elektromagnetischer vorg``ange streng begr''unden .
dies geschah aus einer opposition zu den vertretern des atomismus , die in den gesetzen der thermodynamik lediglich statistische gesetzm``a''sigkeiten einer sonst regellosen bewegung sehr vieler molek``ule sehen wollten , w''ahrend planck fest an eine strenge gesetzlichkeit ohne statistische ausnahmen glaubte . in einer jugendarbeit aus dem jahre
1884 schreibt der 24-j``ahrige selbstbewu''st ( @xcite , bandi , dokument nr.4 , pp.162 - 163 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` der zweite hauptsatz der mechanischen w``armetheorie consequent durchgef''uhrt , ist unvertr``aglich mit der annahme endlicher atome .
es ist daher vorauszusehen , da ' 's es i m laufe der weiteren entwicklung der theorie zu einem kampfe zwischen diesen beiden theorien kommen wird , der einer von ihnen das leben kostet . ''
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ zwei zeilen weiter l``a''st er wenig zweifel dar``uber , welche der theorien seiner meinung und hoffnung nach das leben wird lassen m''ussen : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` ... indessen scheinen mir augenblicklich verschiedenartige anzeichen darauf hinzudeuten , da``s man trotz der bisherigen erfolge der atomistischen theorie sich schlie''slich doch einmal zu einer aufgabe derselben und zur annahme einer continuierlichen materie wird entschlie``sen m''ussen . ''
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ zu dieser zeit war der junge planck ein erkl``arter anti - atomist .
sein plan war , zu versuchen , die thermodynamischen gesetze nicht ' ' uber eine mechanik elementarer konstituenten ( atome , molek``ule ) zu begr''unden , sondern mit hilfe der gesetze der elektrodynamik , die mit rein kontinuierlichen , i m raum verteilten gr``o''sen operiert . in seiner antrittsrede
anl``a''slich seiner aufnahme in die preu``sische akademie der wissenschaften i m jahre 1894 erkl''arte er ( @xcite , band iii , dokument nr.122 , p.3 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` es hat sich neuerdings in der physikalischen forschung auch das bestreben bahn gebrochen , den zusammenhang der erscheinungen `` uberhaupt gar nicht in der mechanik zu suchen [ .. ] .
ebenso steht zu hoffen , da ' 's wir auch `` uber diejenigen elektrodynamischen prozesse , welche direkt durch die temperatur bedingt sind , wie sie sich namentlich in der w''armestrahlung `` au''sern , n``ahere aufkl''arung erfahren k``onnen , ohne erst den m''uhsamen umweg durch die mechanische deutung der elektrizit``at nehmen zu m''ussen . ''
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ planck glaubte also an die m``oglichkeit , die gesetze der thermodynamik , namentlich den 2.hauptsatz , als strenge folge bekannter elektromagnetischer gesetze zu verstehen .
dieser sollte aus allgemeinsten prinzipien ableitbar sein , entsprechend seiner wissenschaftlichen disposition , die er in seinem sp''aten , pers``onlich gehaltenen artikel ' ' zur geschichte der auffindung des physikalischen wirkungsquantums" aus dem jahre 1943 so charakterisierte ( @xcite , bandiii , dokument141 , p.255 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ " ` was mich in der physik von jeher vor allem interessierte , waren die gro``sen allgemeinen gesetze , die f''ur s``amtliche naturvorg''ange bedeutung besitzen , unabh``angig von den eigenschaften der an den vorg''angen beteiligten k``orper.'' _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
man denke sich einen hohlraum , der vollst``andig durch w''ande umschlossen ist , etwa das innere eines ofens .
bringt man die w``ande auf eine konstante temperatur , d.h .
@xmath0 grad celsius entsprechen @xmath1 grad kelvin .
null grad kelvin , also @xmath2 grad celsius , bildet eine absolute untere grenze f''ur alle erreichbaren temperaturen , die unter keinen umst"anden unterschritten werden kann . ]
@xmath3 , so wird sich nach einiger zeit i m hohlraum eine bestimmte konfiguration elektromagnetischer strahlung einstellen , die sogenannte w''armestrahlung .
diese wird aus elektromagnetischen wellen aller frequenzen mit unterschiedlichen intensit``aten bestehen .
zwischen strahlung und der die w''ande bildenden materie wird nach einiger zeit ein thermodynamisches gleichgewicht bestehen .
einzig wesentliche voraussetzung f``ur die existenz eines
stabilen gleichgewichtszustandes ist die annahme , da ' 's die materie ( oder zumindest anteile davon ) in _ allen _ frequenzbereichen mit der strahlung wechselwirkt , also strahlung aller frequenzen emittieren und absorbieren kann .
mit hilfe dieser annahme folgerte gustav kirchhoff bereits 1859 die existenz einer _ universellen _
funktion @xmath4 f``ur die spektrale energieverteilung der strahlung .
diese gibt an , wieviel energie in form von elektromagnetischen wellen der frequenz @xmath5 ( genauer : in einem kleinen frequenzintervall @xmath6 $ ] um den wert @xmath5 ) in einem einheitsvolumen ( z.b .
kubikzentimeter ) des hohlraumes enthalten ist , wenn die w''ande auf die temperatur @xmath3 aufgeheizt wurden . da``s diese funktion ' ' universell`` ist , bedeutet , da ' 's sie _ nicht _ von der genaueren beschaffenheit der w``ande abh''angt , also nicht von ihrer form oder ihrem material .
egal , ob die w``ande aus kupfer , uran , keramik oder sonstwas bestehen , immer wird sich bei vorgegebener temperatur ein und dieselbe spektrale energieverteilung von w''armestrahlung einstellen .
darin liegt die nichttriviale einsicht kirchhoffs .
daraus entsteht nun die theoretische aufgabe , diese universelle funktion aus den bekannten gesetzen der thermodynamik und elektrodynamik zu bestimmen .
man beachte , da``s diese aufgabe nur wegen der universalit''at l``osbar erscheint , da dadurch die kenntnis komplizierter materialeigenschaften sowie deren ( zum damaligen zeitpunkt gr''o``stenteils unbekannter ) einfl''usse auf die wechselwirkung zwischen material und strahlung nicht vorausgesetzt werden m"ussen .
durch weitere , raffiniertere thermodynamische `` uberlegungen konnte wilhelm wien 1893 zeigen , da ' 's die funktion @xmath4 aus dem produkt der dritten potenz der frequenz @xmath5 und einer funktion @xmath7 bestehen mu``s , die jetzt nur noch von _ einer _ variablen abh''angt , n``amlich dem quotienten der frequenz und der temperatur auf , aber nur als quotient @xmath8 .
dieser quotient ist die _ eine _ variable , von der @xmath7 alleine abh"angt . ] .
es mu ' 's also gelten : @xmath9 der fortschritt dieser einsicht wiens besteht also in der reduktion des problems auf die bestimmung einer funktion mit nur _ einer _ anstatt zwei unabh``angigen variablen .
bestimmt man @xmath7 , so ist damit nach ( [ eq : wien1 ] ) auch @xmath4 bekannt .
folgen aus ( [ eq : wien1 ] ) auch ohne kenntnis der funktion @xmath7 bereits erste , experimentell pr``ufbare konsequenzen , die gl''anzend best``atigt wurden .
so ergibt sich einerseits das sogenannte wiensche verschiebungsgesetz , welches besagt , da ' 's die frequenz , bei der die spektrale energieverteilung ihr maximum hat , proportional mit der temperatur w``achst .
ebenso ergibt sich , da ' 's die gesamte , `` uber alle frequenzen summierte energieabstrahlung mit der vierten potenz der temperatur anw''achst . dies bezeichnet man als das stefan - boltzmannsche gesetz .
wie gesagt , bestand die eigentliche aufgabe nun in der bestimmung der einen funktion @xmath7 .
durch weitere anwendung fundamentaler prinzipien sollte dies schlie``slich ohne allzu gro''sen aufwand gelingen
so dachten die physiker zwischen 1893 und 1900 .
doch erwies sich diese aufgabe `` uberraschenderweise als fast unl''osbar .
r"uckschauend aus dem jahre 1913 charakterisierte einstein die situation
so ( @xcite , band4 , dokument nr.23 , p.562 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ " ` es w``are erhebend , wenn wir die gehirnsubstanz auf eine waage legen k''onnten , die von den theoretischen physikern auf dem altar dieser universellen funktion @xmath7 hingeopfert wurde ; und es ist diesen grausamen opfers kein ende abzusehen ! noch mehr : auch die klassische mechanik fiel ihr zum opfer , und es ist nicht abzusehen , ob maxwells gleichungen der elektrodynamik die krisis `` uberdauern werden , welche diese funktion @xmath7 mit sich gebracht hat.'' _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ doch zur``uck zum geschehen . aus ' ' uberlegungen , die man eher als `` educated guessing '' bezeichnen kann , schl``agt wien eine einfache exponentialfunktion f''ur @xmath7 vor , die dann i m verbund mit ( [ eq : wien1 ] ) zum sogenannten _
wienschen strahlungsgesetz _ f"uhrt ( @xmath10 bezeichnet i m folgenden die exponentialfunktion ) : @xmath11\,,\ ] ] wobei @xmath12 und @xmath13 konstanten sind , die es noch zu bestimmen gilt .
zahlreiche experimente schienen ausnahmslos diese form der spektralen energieverteilung zu best``atigen ( dies blieb der fall bis etwa mitte 1900 ) . ' ' uberzeugt
von seiner richtigkeit stellt sich daher planck die aufgabe , das wiensche strahlungsgesetz aus ersten prinzipien theoretisch abzuleiten .
als `` prinzipienlieferant '' akzeptiert er vornehmlich die elektrodynamik und die thermodynamik und hier an erster stelle den 2.hauptsatz `` uber die zunahme der entropie .
nach langen m''uhen gelingt ihm schlie``slich i m jahre 1899 eine ableitung von ( [ eq : wien2 ] ) .
er res''umiert stolz ( @xcite , bandi , dokument nr.34 , p.597 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ " ` ich glaube hieraus schlie``sen zu m''ussen , da``s die gegebene definition der strahlungsentropie und damit auch das wiensche energieverteilungsgesetz eine notwendige folge der anwendung des principes
der vermehrung der entropie auf die elektromagnetische strahlungstheorie ist und da ' 's daher die grenzen der g``ultigkeit dieses gesetzes , falls solche ' ' uberhaupt existieren , mit denen des zweiten hauptsatzes der w``armetheorie zusammenfallen.'' _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ironischerweise sind es experimentalphysiker ( lummer und pringsheim ) , die den theoretiker planck in einer ver``offentlichung des gleichen jahres , die der experimentellen ' ' uberpr``ufung des
wienschen strahlungsgesetzes gewidmet ist , h''oflich darauf hinweisen , da``s hier ein logisch unzul''assiger umkehrschlu"s vorliegt ( @xcite , p.225 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` herr planck spricht es aus , da``s dieses [ d.h .
( [ eq : wien2 ] ) ] gesetz eine nothwendige folge der anwendung des principes der vermehrung der entropie auf die elektromagnetische strahlung ist , und da ' 's daher die grenzen seiner g``ultigkeit , falls solche ' ' uberhaupt existieren , mit denen des zweiten hauptsatzes der w``armetheorie zusammenfallen .
soviel uns scheint , w''are die plancksche theorie erst zwingend , wenn wirklich nachgewiesen werden kann , da``s _ jede _ von obiger gleichung abweichende form zu einem ausdruck der entropie f''uhrt , der dem entropiegesetz widerspricht . ''
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ planck hatte n``amlich keineswegs gezeigt , da ' 's das wiensche gesetz eine logische folge des 2.hauptsatzes der thermodynamik ist , sondern nur , da``s es dem 2.hauptsatz nicht widerspricht .
trotz dieses logischen lapsus ist die von planck verwendete methode bemerkenswert . da
sie charakteristisch f''ur das vorgehen eines theoretischen physikers ist , soll sie hier etwas ausf"uhrlicher beschrieben werden .
planck st``utzt sich auf kirchhoff , der ja einwandfrei argumentiert hatte , da ' 's i m thermodynamischen gleichgewicht die spektrale energieverteilung @xmath4 eine _ universelle _ funktion ist , also von der form des hohlraums und der beschaffenheit der
w``ande g''anzlich unabh``angig ist .
die geniale , aber in den meisten darstellungen wenig hervorgehobene idee plancks ist nun folgende ( vorgetragen in @xcite , bandi , dokument nr.34 , pp.592 - 593 ) : wegen der unabh''angigkeit der spektralen energieverteilung von der physikalischen beschaffenheit der wand darf man sich _
zum zwecke der theoretischen bestimmung _ der funktion @xmath4 die wand auch aus einem hypothetischen , der theoretischen beschreibung leicht zug``anglichen material ersetzt denken .
dabei ist es ganz unwesentlich , ob dies hypothetische medium in der realen welt tats''achlich existiert , sondern wesentlich ist nur , da``s es den bekannten gesetzen der physik gen''ugt , also in diesem sinne existieren _
k"onnte_. die kirchhoffsche `` uberlegung versichert dann , da ' 's die spektrale energieverteilung , die sich ( theoretisch ) i m hohlraum des hypothetischen mediums einstellt , dieselbe ist wie die i m hohlraum eines tats"achlich existierenden materials .
planck w``ahlt als hypothetisches medium eine art gitter von kleinen elektrischen ladungen , die mit einer kleinen feder elastisch an eine ruhelage befestigt sind .
planck nennt diese gebilde ' ' resonatoren`` , denn sie sollen f''ahig sein , kleine schwingungen mit einer festen frequenz @xmath5 ( der sogenannten `` eigenfrequenz '' ) auszuf``uhren , wenn sie von einer elektromagnetischen welle dieser frequenz getroffen werden .
dieses sehr vereinfachte modell einer ' '
wand`` ist nun durch die damals bekannten gesetze der elektrodynamik und mechanik vollst''andig zu erfassen ganz i m gegensatz zu einer realistischen wand , deren mikroskopischer aufbau und vor allem deren komplizierte wechselwirkung mit auftreffenden lichtstrahlen zum damaligen zeitpunkt noch ganz unverstanden waren .
aus der selbstverst``andlichen bedingung , da ' 's i m thermodynamischen gleichgewicht jeder dieser elementaren resonatoren genauso viel elektromagnetische energie emittiert wie absorbiert , leitete planck die folgende bedingung zwischen spektraler energiedichte @xmath4 und mittlerer energie @xmath14 eines einzelnen resonators der schwingungsfrequenz @xmath5 bei der temperatur @xmath3 ab : @xmath15 es mu``s hier nochmals betont werden , da ' 's diese gleichung eine unzweideutige folge der gesetze der klassischen physik ( mechanik und elektrodynamik ) ist .
planck die damals bereits von seinem wissenschaftlichen widersacher ludwig boltzmann ( 1844 - 1906 ) ausgearbeitete statistische mechanik akzeptiert , so h''atte er sofort einen ausdruck f``ur @xmath14 angeben k''onnen .
aus dem sogenannten `` aquipartitionsgesetz der statistischen mechanik folgt n''amlich , da``s @xmath16 wobei @xmath17 die sogenannte universelle gaskonstante ist ( durch messungen gut bekannt ) und @xmath18 die avogadro - zahl , also die zahl der in einem mol gas enthaltenen molek''ule .
er w``are damit zum sogenannten rayleigh - jeans - gesetz gelangt : @xmath19 das obwohl eine ebenso unzweideutige folge der klassischen physik
ganz unsinnige aussagen macht .
zum beispiel besagt es , da ' 's bei fester temperatur @xmath3 die in elektromagnetischen wellen der frequenz @xmath5 abgestrahlte energie quadratisch in @xmath5 w``achst , insgesamt also unendlich viel energie abgestrahlt wird , wenn man ' ' uber alle frequenzen summiert .
auch hinsichtlich der abh``angigkeit von @xmath3 geht der ausdruck ( [ eq : rayleigh - jeans ] ) v''ollig fehl .
das direkt proportionale ansteigen der strahlungsenergie mit der temperatur h``atte zum beispiel zur folge , da ' 's bei jeder frequenz die energieabstrahlung bei raumtemperatur @xmath3 etwa gleich 290grad kelvin
immerhin noch ein sechstel der abstrahlung bei der temperatur von 1700 grad kelvin w``are .
letztere entspricht etwa der temperatur schmelzenden ( d.h .
wei''sgl``uhenden ) stahls . dies ist offensichtlich eine groteske ' ' ubersch``atzung der abstrahlung bei raumtemperatur .
doch planck erw''ahnt diese katastrophale folge mit keinem wort .
erst einstein wird in seiner nobelpreisarbeit von 1905 darauf beharren , da``s die klassische physik notwendig zum inakzeptablen rayleigh - jeans - gesetz f''uhrt und deswegen fundamental nicht richtig sein kann .
planck geht v``ollig andere , recht seltsame wege , um die jetzt noch fehlende funktion @xmath4 zu bestimmen . in der annahme der richtigkeit des wienschen gesetzes kennt er das ziel und wei ' 's daher , welchen ausdruck f``ur @xmath4 er ' ' herbeiargumentieren`` mu ' 's , um ( [ eq : wien2 ] ) aus ( [ eq : planck1 ] ) folgen zu lassen .
an dieser stelle bringt er nun den 2.hauptsatz der thermodynamik ins spiel : statt die energie @xmath14 des einzelnen resonators zu bestimmen
wof``ur er keine direkte methode hat , geht er den umweg ' ' uber die entropie @xmath20 , denn diese sollte sich aus den forderungen des 2.hauptsatzes ergeben .
aus einer allgemein g``ultigen thermodynamischen relation , nach der die ableitung der entropie nach der energie die inverse temperatur ist ( siehe gleichung ( [ eq : entropie - energie - temp ] ) i m anhang[sec : anhangplanck ] ) , w''urde sich dann auch die funktion @xmath14 ergeben .
planck gibt dann tats``achlich einen entropieausdruck an , von dem er zeigen kann , da ' 's er allen anforderungen des 2.hauptsatzes gen``ugt und der direkt zum wienschen gesetz f''uhrt .
entgegen seiner obigen aussagen zeigt er aber nicht , da``s dieser ausdruck eindeutig ist .
es k''onnte also durchaus andere , ebenfalls mit dem 2.hauptsatz formal vertr``agliche strahlungsgesetze geben ( was sich sp''ater auch als tats"achlich gegeben herausstellt ) .
experimentelle messungen an der physikalisch - technischen reichsanstalt in berlin i m jahre 1899 ergaben systematische abweichungen vom wienschen strahlungsgesetz i m bereich niederer frequenzen ( d.h .
gro``ser wellenl''angen ) @xcite .
die gemessenen energien lagen bei kleinen frequenzen systematisch oberhalb der wienschen kurve .
dazu m``u''sten erst neue me``smethoden entwickelt werden , um den niederfrequenten anteil des spektrums m''oglichst sauber zu isolieren .
diese `` divergenzen von erheblicher natur '' ( planck ) wurden in der sitzung der deutschen physikalischen gesellschaft am 19.oktober mitgeteilt .
es ist bekannt , da``s planck bereits am 7.oktober einem sonntag
von heinrich rubens , einem der experimentatoren , aufgesucht und von den neuen experimentellen befunden unterrichtet wurde .
noch am gleichen abend fand planck durch geschicktes probieren ( i m anhang[sec : anhangenergiefluktuationen ] erl''autert ) eine neue , von der wienschen leicht abweichende strahlungsformel , die die neuen resultate befriedigend wiederzugeben vermochte .
diese teilte er dann ebenfalls am 19.oktober i m anschlu``s an das referat des experimentalphysikers kurlbaum der deutschen physikalischen gesellschaft mit .
damit war die plancksche strahlungsformel geboren : @xmath21 sie unterscheidet sich von der wienschen formel ( [ eq : wien2 ] ) lediglich durch die -1 i m nenner , so da ' 's f``ur hohe frequenzen und / oder kleine temperaturen beide ausdr''ucke approximativ gleich sind .
f``ur kleine verh''altnisse @xmath8 verl``auft die plancksche kurve aber systematisch _
oberhalb _ der wienschen
, sagt also bei gegebener temperatur eine merklich h''ohere energiedichte i m bereich kleiner frequenzen ( d.h .
gr``o''serer wellenl"angen ) voraus . dies ist in abbildung[fig : planckwien ] dargestellt .
( 6,9)@xmath5 ( -250,165)@xmath4 ( -212,80)@xmath22wien ( -236,130)planck@xmath22 experimentell wurden diese abweichungen von der wienschen formel bei gro``sen wellenl''angen zuerst von otto lummer und ernst pringsheim @xcite gemessen und sofort darauf von heinrich rubens und ferdinand kurlbaum @xcite noch eindr``ucklicher best''atigt .
beide gruppen arbeiteten zu dieser zeit an der physikalisch - technischen reichsanstalt in berlin - charlottenburg .
die me``skurve aus der originalver''offentlichung von lummer und pringsheim ist als abbildung[fig : speklp ] i m anhang[sec : anhanglp ] wiedergegeben .
zum schlu``s dieses abschnitts erw''ahnen wir noch , da``s in moderner schreibweise die konstanten @xmath12 und @xmath13 in ( [ eq : planck ] ) durch andere konstanten ausgedr''uckt werden , n``amlich die lichtgeschwindigkeit @xmath23 , die boltzmann - konstante @xmath24 und das plancksche wirkungsquantum @xmath25 : @xmath26 dieser zusammenhang wird allerdings erst durch die theoretische begr''undung der planckschen strahlungsformel verst"andlich werden .
zu beginn seiner ber``uhmten arbeit ' ' uber lichtquanten aus dem jahre 1905 ( @xcite , band2 , dokument14 ) macht einstein eine wichtige bemerkung , die man etwa so zusammenfassen kann : fordert man , da``s das gesetz ( [ eq : rayleigh - jeans ] ) , was eine notwendige folge der klassischen physik ist , als grenzgesetz in der als
ph''anomenologisch g``ultig angesehenen planckschen formel enthalten ist , so ergibt sich eine von jeder _
theoretischen begr"undung _ der planckschen formel _ unabh"angige _ methode zur bestimmung der avogadro - zahl @xmath18 . entwickelt man die exponentialfunktion i m nenner von ( [ eq : planck ] ) bis zu linearer ordnung in @xmath27 , so ergibt sich das gesetz ( [ eq : rayleigh - jeans ] ) genau dann , wenn die avogadro - zahl @xmath18 mit den konstanten @xmath28 des planckschen gesetzes in folgender beziehung
steht : @xmath29 da @xmath17 und @xmath23 gut bekannt sind , liefert jede bestimmung von @xmath12 und @xmath13 durch strahlungsmessungen auch einen wert f''ur @xmath18 .
einstein erhielt so den wert @xmath30 .
zu dieser zeit war dies der mit abstand genaueste wert der avogadro - zahl ( vgl .
kapitel 5 in @xcite ) .
aber man konnte noch weiter schlie``sen : aus der kenntnis der faradaykonstante ( elektrische ladung eines mols einwertiger ionen ) , die aus elektrolysedaten gut bekannt war , erh''alt man nach division durch
@xmath18 den wert der elektrischen elementarladung @xmath31 .
die elementarladung ( betrag der ladung eines elektrons ) lie``s sich also aus strahlungsmessungen mit hilfe der planckschen formel gewinnen , wobei sich ein weit besserer wert als jemals zuvor ergab .
dies geht eindr''ucklich aus folgendem vergleich der damals diskutierten werte mit dem heutigen wert der `` particle data group '' ( pdg ) hervor ( in einheiten von @xmath32 , wo @xmath33 die ladungseinheit `` electrostatic units '' bezeichnet ) : [ cols= " > , < " , ] tats``achlich hatte bereits planck 1901 die hier angegebenen werte f''ur die avogadro - zahl und die elementarladung mit hilfe seiner formel und den ergebnissen von strahlungsmessungen ausgerechnet ( @xcite , bandi , dokument nr.44 , pp.717 - 727 ) .
aber erst einstein sah , da``s dieses vorgehen weitgehend unabh''angig von plancks theoretischer begr``undung seiner strahlungsformel gerechtfertigt werden kann , wenn man nur die forderung nach dem klassischen limes stellt .
die ironie dieser episode ist , da ' 's diese pr``azisionsbestimmung einer fundamental atomistischen gr''o``se ausgerechnet durch den damaligen anti - atomisten planck erm''oglicht wurde .
wie sollte nun planck nach all seinen m``uhen , das wiensche gesetz theoretisch zu begr''unden , eine ableitung des neuen gesetzes ( [ eq : planck ] ) herzaubern ?
hatte er nicht noch gerade argumentiert , da``s der 2.hauptsatz notwendig zum wienschen gesetz f''uhre ?
immerhin blieb er seiner `` klassischen '' formel ( [ eq : planck1 ] ) treu und seiner strategie , die mittlere resonatorenergie @xmath14 aus der entropie zu bestimmen .
er erkannte jetzt endg``ultig , da ' 's der ausdruck f``ur letztere , den er vorher nach vielen m''uhen erhalten hatte und der ihm scheinbar unausweichlich zum wienschen gesetz f``uhrte , nicht der formal einzig m''ogliche sein konnte .
so sehr sich planck aber auch abm``uhte , eine begr''undung des erforderlichen neuen ausdrucks zu liefern , es wollte ihm einfach nicht gelingen . in seinem
ringen um das auffinden allgemeiner methoden , die es erlauben w``urden , die entropie eines resonators i m strahlungsfeld zu berechnen , verfiel er schlie''slich auf den verzweifelten ausweg , ausgerechnet die von ihm bisher vehement bek``ampfte methode der statistischen interpretation der entropie seines widersachers boltzmann zu verwenden .
danach ist die entropie eine rein kombinatorische gr''o``se , die bekannt ist , wenn man die anzahl der m''oglichkeiten kennt , eine feste energiemenge auf eine feste anzahl von resonatoren zu verteilen .
diese anzahl w``are unendlich
und damit die entropie unbestimmt , wenn jeder resonator energie in kontinuierlichen mengen aufnehmen k''onnte .
damit die entropie endlich herauskommt , mu``s planck annehmen , da ' 's die gesamtenergie nur in ganzzahligen vielfachen einer bestimmten grundeinheit `` uber die resonatoren verteilt werden kann .
aus dem allgemeinen gesetz ( [ eq : wien1 ] ) ergibt sich , da ' 's diese grundeinheit proportional zur eigenfrequenz @xmath5 des resonators sein mu``s .
diese proportionalit''atskonstante nennt man heute das plancksche wirkungsquantum @xmath25 .
f``ur die energie - grundeinheit @xmath34 gilt also die plancksche formel @xmath35 f''ur planck war dies eine rein formale annahme von h``ochstens heuristischer bedeutung , die er hoffte , sp''ater durch ein physikalisches argument eliminieren zu k``onnen .
immerhin f''uhrte sie ihn zu einer ableitung , die er der deutschen physikalischen gesellschaft in der sitzung am 14.dezember des jahres 1900 mitteilte .
dieses datum gilt bis heute als die geburtsstunde der quantentheorie .
`` uber die ihm so seltsam aufgezwungene annahme der energiequantelung schrieb planck r''uckschauend in einem brief aus dem jahre 1931 ( @xcite ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` das war eine rein formale annahme [ energiequantelung ] , und ich dachte mir eigentlich nicht viel dabei , sondern eben nur das , da``s ich unter allen umst''anden , koste es , was es wolle , ein positives resultat herbeif``uhren m''u``ste .
[ ... ] kurz zusammengefa''st kann ich die ganze tat als einen akt der verzweiflung bezeichnen .
denn von natur bin ich friedlich und bedenklichen abenteuern abgeneigt . ''
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ i m anhang[sec : anhangplanck ] ist plancks `` akt der verzweiflung '' nochmals etwas genauer beschrieben .
einstein war mit plancks theoretischer begr"undung der strahlungsformel ( [ eq : planck ] ) zutiefst unzufrieden .
er i m wesentlichen zwei hauptkritikpunkte vor : * planck benutzt wesentlich gleichung ( [ eq : planck1 ] ) , die mit hilfe der maxwellschen theorie abgeleitet ist und voraussetzt , da``s der energieaustausch zwischen planckschen resonatoren und strahlungsfeld kontinuierlich verl''auft , _ i m gegensatz _ zur quantisierungsannahme ( [ eq : hnu ] ) .
zwar kann man zun``achst argumentieren , da ' 's ( [ eq : planck1 ] ) ja nur f``ur den statistischen mittelwert der resonatorenergie g''ultigkeit beansprucht und somit vielleicht auch unter einem gequantelten energieaustausch , zumindest in guter n``aherung , g''ultig bleibt .
doch w``are das nur dann zu erwarten , wenn die mittlere energie des einzelnen resonators @xmath14 sehr gro ' 's gegen die energieportionen ( [ eq : hnu ] ) ist .
aus ( [ eq : planck1 ] ) und der planckschen formel ( [ eq : planck ] ) kann man aber @xmath14 direkt ablesen .
mit den bezeichnungen ( [ eq : konstanterel ] ) ergibt sich @xmath36 demnach ist sogar umgekehrt @xmath14 sehr viel kleiner als @xmath37 , falls @xmath37 viel gr``o''ser als @xmath38 ist . dies ist genau i m geltungsbereich des wienschen gesetzes der fall , in dem die annahme von ( [ eq : planck1 ] ) mit der quantisierungsvorschrift ( @xmath39 ) also unvertr"aglich zu sein scheint . *
zur berechnung von @xmath14 `` uber die entropie verwendet planck die boltzmannsche entropiedefinition ( [ eq : boltzmann - entropie ] ) .
die zun''achst nur durch formales
abz``ahlen bestimmte mikroskopische multiplizit''at eines makroskopischen zustands ist jedoch nur dann proportional seiner physikalischen wahrscheinlichkeit , wenn die mikrozust``ande i m sinne der tats''achlich gegebenen dynamik des systems auch _ physikalisch gleich wahrscheinlich _
sind , soll hei``sen : i m laufe einer langen zeit mit gleichen relativen zeitdauern eingenommen werden . da planck
f''ur den resonator die klassische dynamik als richtig annimmt ( z.b . in der ableitung der gleichung ( [ eq : planck1 ] ) ) , w``urde die boltzmannsche entropiedefinition korrekt angewendet notwendig zur rayleigh - jeans - formel ( [ eq : rayleigh - jeans ] ) f''uhren .
diesbez"uglich kommentiert einstein 1909 in der ihm eigenen charmant - frechen weise ( @xcite , band2 , dokument56 , pp.544 - 545 ) : + _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ " ` so sehr sich jeder physiker dar``uber freuen mu ' 's , da``s sich herr planck in so gl''ucklicher weise `` uber diese forderung hinwegsetzte , so wenig w''are es angebracht , zu vergessen , da``s die plancksche strahlungsformel mit der theoretischen grundlage , von welcher herr planck ausgegangen ist , unvereinbar ist.'' _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ diese bedenken tr``agt einstein u.a .
auch in seinem umfassenden bericht ' ' `` uber die entwicklung unserer anschauungen ' ' uber das wesen und die konstitution der strahlung`` auf der 81 .
versammlung deutscher naturforscher und ' ' arzte 1909 in salzburg vor , wo der 30-j``ahrige seinen ersten gr''o``seren ' ' offentlichen auftritt hatte . in der sich
anschlie``senden diskussion erl''autert planck nochmals seine sichtweise der quantisierungsannahme ( [ eq : hnu ] ) , die er dezidiert aufgefa``st wissen wollte als ausdruck eines noch unverstandenen mechanismus , der lediglich die _ wechselwirkung _ von strahlung und materie betraf .
materie war eben nur in der lage , energie in gewissen endlichen portionen an das strahlungsfeld anzugeben oder aus dem strahlungsfeld aufzunehmen .
weder hatte planck i m sinn , damit eine grunds''atzliche modifikation der dynamik des resonators auszusprechen und schon gar nicht eine quantisierung des strahlungsfeldes selbst zu postulieren .
er hoffte auf jeden fall , die maxwellsche theorie des elektromagnetismus , die durchweg von der vorstellung kontinuierlicher prozesse in raum und zeit ausgeht , zumindest i m wechselwirkungsfreien fall beizubehalten .
w"ortlich sagte planck : ( @xcite , band2 , dokument nr.61 , pp.585 - 586 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ " ` jedenfalls meine ich , man m``u''ste zun``achst versuchen , die ganze schwierigkeit der quantentheorie zu verlegen in das gebiet der _ wechselwirkung _ zwischen der materie und der strahlenden energie ; die vorg''ange i m reinen vakuum k``onnte man dann vorl''aufig noch mit den maxwellschen gleichungen erkl``aren.'' _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ dem gegen"uber steht einsteins resumee seines vortrages : ( @xcite , band2 , dokument nr.60 , pp.576 - 577 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ " ` die plancksche theorie annehmen hei``st nach meiner meinung geradezu die grundlagen unserer strahlungstheorie verwerfen.'' _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ w``ahrend plancks ablehnende haltung gegen''uber einer modifikation der maxwellschen theorie des freien strahlungsfeldes aus seinen schriften ganz offenbar wird ( was sich auch in seiner kritik der lichtquantenhypothese `` au''sert ) , ist seine haltung gegen``uber einer modifikation der mechanischen gesetze , hier i m zusammenhang mit den resonatoren , etwas umstritten .
diesbez''uglich hat sich in j``ungerer zeit sogar ein sogenannter ' ' historikerstreit`` entz''undet ( vgl .
@xcite ) , der mir aber etwas `` ubertrieben scheint . in seiner
urspr''unglichen ableitung macht planck in der tat die formale annahme ( [ eq : hnu ] ) , ohne eine modifikation der mechanischen gesetze zu erw``ahnen . 1906 gibt einstein eine ableitung des planckschen gesetzes mit hilfe der von ihm selbst entwickelten allgemeinen methoden der statistischen mechanik ( @xcite , band2 , dokument34 ) .
dort zeigt er , da ' 's man konsistent ( d.h .
unter vermeidung der oben unter punkt2 ge``au''serten kritik ) zu ( [ eq : resenergiemittel ] ) gelangt , wenn man annimmt , da``s die resonatoren selbst nur ganzzahlige vielfache der energie @xmath37 annehmen k''onnen und formuliert dies als eigentliche , der planckschen ableitung zugrundeliegende annahme .
auf die verallgemeinerung dieser annahme auf jedes schwingungsf``ahige gebilde in einem festk''orper st``utzt einstein kurz darauf seine quantentheorie der spezifischen w''arme ( @xcite , band2 , dokument38 ) .
planck ist damit aber nicht einverstanden und versucht sp``ater ( 1911 - 12 ) sogar , eine ableitung seines strahlungsgesetzes zu geben , in der nur der proze ' 's der emission , nicht jedoch der proze``s der absorption ' ' gequantelt`` ist ( @xcite , band2 , dokumente73,74,75 ) . )
f``ur die mittlere energie eines resonators um einen additiven term @xmath40 .
dies markiert das erste auftreten der heute aus der quantenmechanik wohlbekannten ' ' nullpunktsenergie". ] die unterscheidung dieser ' ' neuen strahlungshypothese`` plancks von der urspr''unglichen ist aber nur dann sinnvoll , wenn man annimmt , da``s die resonatorenergien grunds''atzlich kontinuierliche werte annehmen k``onnen .
daraus mu ' 's man m.e .
schlie``sen , da ' 's zwar einstein , aber nicht planck die gleichung ( [ eq : hnu ] ) i m heutigen quantenmechanischen sinne verstanden haben wollte , n``amlich als allgemeine quantisierungsbedingung materieller schwingungsf''ahiger systeme . in seiner
`` lichtquantenhypothese '' erweiterte einstein diese quantisierungsbedingung dann auch auf das freie strahlungsfeld , was der heutigen sichtweise der quantenelektrodynamik entspricht .
in den uns vorliegenden schriftlichen dokumenten einsteins kennzeichnet er nur eine einzige seiner wissenschaftlichen ideen als " ` sehr revolution``ar'' ( @xcite , band5 , dokument nr.27 , p.31 ) die vierte arbeit liegt erst i m konzept vor und ist eine elektrodynamik bewegter k''orper unter ben``utzung einer modifikation der lehre von raum und zeit ; der rein kinematische teil dieser arbeit wird sie interessieren ' '' .
die zuletzt , eher lapidar angek``undigte arbeit , ist die spezielle relativit''atstheorie .
] , n``amlich die lichtquantenhypothese .
diese ver''offentlicht er i m jahre 1905 , 26-j``ahrig , i m gleichen zeitschriftenband wie seine spezielle relativit''atstheorie und die theorie der brownschen bewegung .
die arbeit tr``agt den titel ' ' `` ` uber einen die erzeugung und verwandlung des lichtes betreffenden heuristischen gesichtspunkt ' '' .
dieser `` heuristische gesichtspunkt '' besteht in einer v``ollig anderen interpretation der planckschen quantisierungsbedingung ( [ eq : hnu ] ) , n''amlich als eigenschaft des strahlungsfeldes selbst .
er schreibt ( @xcite , band2 , dokument nr.14 , p.151 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ " ` nach der hier ins auge zu fassenden annahme ist bei ausbreitung eines von einem punkte ausgehenden lichtstrahles die energie nicht kontinuierlich auf gr``o''ser und gr``o''ser werdende verteilt , sondern es besteht dieselbe aus einer endlichen zahl von in raumpunkten lokalisierten energiequanten , welche sich bewegen , ohne sich zu teilen und nur als ganze absorbiert und erzeugt werden k``onnen.'' _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ diese scheinbare r``uckkehr zur l''angst `` uberkommenen partikelvorstellung des lichts , die zwar noch newton vertreten hatte , die aber dann i m fr''uhen 19.jahrhundert durch den siegeszug der wellentheorie geradezu hinweggefegt wurde , mu``ste auf die zeitgenossen einsteins als eine mischung aus naiv und provokant gewirkt haben , eben geradezu unverfroren . ' ' au``serungen dazu werden uns weiter unten begegnen .
und doch war einsteins sichtweise , wie die seiner gegner , nicht unbegr''undet .
durch eine scharfe analyse des strahlungsgesetzes , insbesondere des ihm immer suspekt erschienenen wienschen bereichs , zeigt er , da"s ( @xcite , band2 , dokument nr.14 , p.161 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ " ` monochromatische strahlung von geringer dichte ( innerhalb des g``ultigkeitsbereiches der wienschen strahlungsformel ) verh''alt sich in w``armetheoretischer beziehung
so , wie wenn sie aus voneinander unabh''angigen energiequanten von der gr``o''se @xmath37 best``unde''.[zit : einsteinlq ] _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ die genauere argumentation einsteins ist in anhang[sec : anhangeinstein ] erl"autert .
einstein ist klar , da``s sich diese vorstellung auch an der erkl''arung bekannter ph``anomene wird behaupten m''ussen , namentlich solcher , die die noch unverstandenen prozesse bei der wechselwirkung von licht mit materie betreffen .
einer dieser prozesse ist der sogenannte `` photoelektrische effekt '' , bei dem durch bestrahlung einer metallplatte mit licht elektronen aus dem material herausgel``ost werden .
die energie des ankommenden
lichtes wird also durch irgendeinen proze ' 's dazu verwandt , das elektron aus dem atomverband herauszul``osen , wozu eine nur vom material abh''angige energie @xmath41 aufzuwenden ist .
die `` ubersch''ussige energie des ankommenden lichtes wird dann in die bewegungsenergie @xmath42 des austretenden elektrons investiert .
gem``a ' 's der traditionellen wellentheorie des lichtes erfolgt dessen ausbreitung stetig `` uber alle raumbereiche .
da die energie des lichtes dann proportional zu seiner intensit''at ist , m``u''ste z.b .
die energie der herausgel``osten elektronen mit dem abstand der lichtquelle von der metallplatte fallen , da mit dem abstand auch die intensit''at abnimmt .
was aber durch den experimentalphysiker philipp lenard ( 1862 - 1947 , nobelpreis 1905 ) i m jahre 1900 tats``achlich beobachtet wurde , ist , da ' 's zwar die anzahl der herausgel``osten elektronen mit fallender intensit''at abnimmt , nicht aber deren individuelle energien , die sich als _
von der intensit``at des eingestrahlten lichtes unabh''angig _ ergaben .
auf das einzelne elektron wird also eine immer gleiche energie `` ubertragen .
dieser tatbestand pa''st nun `` uberhaupt nicht zur wellentheorie des lichtes , wird aber sofort plausibel bei zugrundelegung der lichtquantenhypothese .
nach dieser wird jedes der einzelnen elektronen durch ein ganzes , unteilbares lichtquant der energie @xmath37 herausgel''ost und mit einer bewegungsenergie @xmath42 heraustreten , die der differenz der energie des lichtquants zur abl``osungsenergie @xmath41 entspricht : @xmath43 diese sogenannte ' ' einsteinsche gleichung`` zum photoeffekt wurde teilweise durch lenard und sp''ater vor allem durch den amerikanischen experimentalphysiker robert millikan ( 1868 - 1953 , nobelpreis 1923 ) vollauf best``atigt , was sogar mit ein grund f''ur die
vergabe des nobelpreises war : `` for his work on the elementary charge of electricity and on the photoelectric effect '' .
somit schien der photoeffekt mit einem schlag eine v``ollig nat''urliche erkl"arung zu finden vorausgesetzt , man akzeptierte die lichtquantenhypothese !
trotzdem war aber allen beteiligten klar , da``s die annahme der einsteinschen vorstellung der lichtquanten v''ollig unvereinbar sein w``urde mit der g''angigen ( maxwellschen ) theorie des elektromagnetismus , die wenige jahre zuvor durch die aufsehen erregenden versuche heinrich hertz scheinbar so gl``anzend best''atigt wurde und auf die planck die ableitung seiner ausgangsgleichung ( [ eq : planck1 ] ) wesentlich gest``utzt hatte .
die aus der planckschen strahlungsformel in gewisser weise ableitbare lichtquantenhypothese anzunehmen , hie''se dann also gleichzeitig , der theoretischen begr``undung dieser formel den boden zu entziehen .
das genau war die kritik einsteins , die er ' ' uber viele jahre hinweg in mannigfacher variation immer wieder vorbrachte .
wenig `` uberraschend ist es daher , da ' 's die einsteinsche lichtquantenhypothese vor allem bei planck , aber auch bei anderen physikern auf starke ablehnung stie``s , darunter auch solche , die einstein wissenschaftlich und pers''onlich sehr nahe standen ( worunter man sonst auch planck z``ahlen mu ' 's , aber eben mit ausnahme dieses einen punktes betreffend die lichtquantenhypothese ) .
so beginnt z.b .
der theoretiker arnold sommerfeld , einer der besten kenner der materie und von einstein sehr geachtet , i m jahre 1911 seinen l``angeren vortrag auf der 83 .
versammlung der gesellschaft deutscher naturforscher und ' ' arzte
so ( @xcite , p.31 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ " ` als der wissenschaftliche ausschu``s unserer gesellschaft an mich die aufforderung richtete , dieser versammlung einen bericht ' ' uber die relativit``atstheorie zu erstatten , erlaubte ich mir dagegen geltend zu machen , da ' 's das relativit``atsprinzip kaum mehr zu den eigentlich aktuellen fragen der physik geh''ore .
obwohl erst 6 jahre alt
einsteins arbeit erschien 1905 scheint es schon in den gesicherten besitz der physik `` ubergegangen zu sein .
ganz anders aktuell und problematisch ist die theorie der _ energiequanten _ [ ... ] .
hier sind die grundbegriffe noch i m flu ' 's und die probleme ungez``ahlt.'' _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ und f"ahrt kurz darauf fort : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` einstein zog aus der planckschen entdeckung die weitestgehenden folgen [ ... ] und `` ubertrug das quantenhafte von dem emissions- und absorptionsvorgang auf die struktur der lichtenergie i m raume , ohne , wie ich glaube , seinen damaligen standpunkt heute noch in seiner ganzen k''uhnheit aufrecht zu erhalten . ''
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ und selbst der bereits erw``ahnte gro''se amerikanische experimentalphysiker robert millikan , der 10 jahre seines forscherlebens der experimentellen `` uberpr''ufung der einsteinschen formel ( [ eq : photoeffekt ] ) f``ur den photoeffekt widmete und dadurch auch die ersten pr''azisionsmessungen des planckschen wirkungsquantums @xmath25 realisierte ( siehe anhang[sec : anhangmillikan ] ) , schrieb 1916 in einem langen , zusammenfassenden artikel `` uber die gerade von ihm so gl''anzend best"atigte einsteinsche formel ( @xcite , p.384 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` despite then the apparently complete success of the einstein equation , the physical theory of which it was designed to be the symbolic expression is found so untenable that einstein himself , i believe , no longer holds to it . ''
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ die lichtquantenhypothese i m allgemeinen kommentiert millikan bereits auf der ersten seite seines artikels wie folgt ( @xcite , p.355 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` this hypothesis may well be called reckless , first because an electromagnetic disturbance which remains localized in space seems a violation of the very conception of an electromagnetic disturbance , and second because it flies in the face of the thoroughly established facts of interference . ''
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ kurz zuvor , i m jahre 1913 , als einstein die ehre zuteil wird , in die preu"sische akademie der wissenschaften aufgenommen zu werden , verfassen planck , nernst , rubens und warburg ein empfehlungsschreiben , das mit folgenden worten endet ( @xcite , band5 , dokument nr . 445 , p.527 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ " ` zusammenfassend kann man sagen , da``s es unter den gro''sen problemen , an denen die moderne physik so reich ist , kaum eines gibt , zu dem nicht einstein in bemerkenswerter weise stellung genommen h``atte .
da ' 's er in seinen spekulationen gelegentlich auch einmal `` uber das ziel hinausgeschossen haben mag , wie z.b . in seiner hypothese der lichtquanten , wird man ihm nicht allzuschwer anrechnen d''urfen ; denn ohne ein risiko zu wagen , l``a''st sich auch in der exaktesten naturwissenschaft keinerlei wirkliche neuerung einf``uhren.'' _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ acht jahre sp``ater , 1921 , wird einstein f''ur die erkl``arung des photoelektrischen effektes mit hilfe der lichtquantenhypothese der nobelpreis f''ur physik zuerkannt .
aber auch danach verklingen die zweifel noch nicht .
ein jahr nach einstein bekommt niels bohr den nobelpreis . in seiner
nobel - vorlesung mit dem titel `` the structure of the atom '' schl``agt bohr ganz ' ' ahnliche t``one an wie sechs jahre zuvor millikan .
unter anderem findet sich in der niederschrift von bohrs vorlesung folgender eindr''ucklicher passus @xcite : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` this phenomenon [ des photoelektrischen effekts ] , which had been entirely unexplainable on the classical theory , was thereby placed in quite a different light , and the predictions of einstein s theory have received such exact experimental confirmation in recent years , that perhaps the most exact determination of planck s constant is afforded by measurements on the photoelectric effect . in spite of this heuristic value , however , the hypothesis of light - quanta , which is irreconcilable with so - called interference phenomena , is not able to throw light on the nature of radiation .
i need only recall that these interference phenomena constitute our only means of investigating the properties of radiation and therefore of assigning any closer meaning to the frequency which in einstein s theory fixes the magnitude of the light - quantum . ''
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ wie bereits erw``ahnt , folgte auf bohr millikan als physik - nobelpreistr''ager des jahres 1923 . in seiner
nobel - vorlesung mit dem titel `` the electron and the light - quant from the experimental point of view '' `` au''sert auch er sich nochmals kritisch , wenn auch mittlerweile in etwas abgeschw"achter form ( die hervorhebungen sind seine ) @xcite : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` in view of this methods and experiments the general validity of einstein s equation [ gemeint ist gleichung ( [ eq : photoeffekt ] ) ] is , i think , now universally conceded , and _ to that extent the reality of einstein s light - quanta may be considered as experimentally established . _ but the conception of _ localized _ light - quanta out of which einstein got his equation must still be regarded as far from being established . whether the mechanism of interaction between ether waves and electrons has its seat in the unknown conditions and laws existing within the atom , or is to be looked for primarily in the essentially corpuscular thomas - planck - einstein conception as to the nature of radiant energy is the all - absorbing uncertainty upon the frontiers of modern physics . ''
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ein letztes dramatisches aufb``aumen der kritiker ' ' au``serte sich 1924 in einer damals sehr viel aufsehen erregenden arbeit von bohr , kramers und slater @xcite , in der eine statistische theorie der wechselwirkung zwischen strahlung und materie formuliert wird mit dem erkl''arten ziel , g``anzlich ohne die lichtquanten auszukommen .
als preis daf''ur sollte man hinnehmen , da``s die erhaltungss''atze von energie und impuls zwar i m statistischen mittel , nicht jedoch f``ur den individuellen elementarproze ' 's g``ultig seien .
dabei hatte gerade ein jahr zuvor arthur compton ( 1892 - 1962 , geteilter nobelpreis 1927 ) die klassisch unverst''andlichen eigenschaften der streuung von r``ontgenstrahlen an materie mit der annahme erkl''art , da``s es sich dabei um individuelle st''o``se von lichtquanten mit einzelnen elektronen handle , wobei f''ur jeden sto``s individuell energie- und impulserhaltung gelten @xcite ( sogenannter comptoneffekt ) .
dies mag andeuten , wie verzweifelt der vorschlag von bohr , kramers und slater damals war , die nun argumentierten mu''sten , da``s die ph''anomene auch mit einer nur i m statistischen mittel g``ultigen energieerhaltung vertr''aglich w``aren , was aber schon kurz darauf durch zahlreichen experimente widerlegt wurde ( z.b . auch wieder durch compton ; siehe @xcite ) .
erst ab 1925 , das auch das geburtsjahr der quantenmechanik ist , kann man also davon sprechen , da ' 's sich einsteins lichtquantenhypothese in den ma"sgebenden fachkreisen wirklich durchgesetzt hatte .
plancks gr``o''ste wissenschaftliche leistung ist auf ironisch und fast tragische weise erkauft mit dem scheitern seines gro``s angelegten planes , dessen ziel es war , den 2.hauptsatz als streng kausales gesetz aus den gesetzen der elektrodynamik zu begr''unden .
auf seinem weg dorthin findet er stattdessen ein neues , experimentell gl``anzend best''atigtes strahlungsgesetz unter zugrundelegung der von ihm sonst vehement bek``ampften statistischen entropiedefinition .
die theoretischen implikationen dieses gesetzes , namentlich die lichtquantenhypothese einsteins , entziehen planck geradezu die gesamte grundlage , von der aus er urspr''unglich startete .
wie kein anderer f``orderte einstein in dieser zeit durch hartn''ackiges hinterfragen der grundlagen der planckschen strahlungstheorie den endg``ultigen bruch mit der klassischen physik .
so wurde planck _ durch einstein _
zum revolution''ar wider willen .
aber auch einstein selbst bleibt dieses schicksal nicht erspart .
noch 1916 gibt er eine wunderbar einfache ableitung der planckschen strahlungsformel , die nun vollst``andig auf den gebrauch der beziehung ( [ eq : planck1 ] ) verzichtet .. dazu betrachtet einstein die absorption und emission von licht als statistische prozesse , m''oglicherweise in der hoffnung , sie sp``ater doch noch deterministisch verstehen zu k''onnen .
interessanterweise mu``s er , um zur planckschen formel zu gelangen , neben dem proze ' 's der _ spontanen _ emission auch einen bis dahin unbekannten proze``s der _ induzierten _ emission postulieren ( ohne ihn w''are er formal zum wienschen gesetz gelangt ) ,
der sp``ater die grundlage des funktionsprinzips des lasers werden sollte .
bez''uglich der statistischen natur dieser prozesse schreibt er am ende dieser arbeit ( @xcite , band6 , dokument nr.38 , p.396 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ " ` die schw``ache der theorie liegt einerseits darin , da ' 's sie uns dem anschlu``s an die undulationstheorie [ d.h .
wellentheorie ] nicht n''aher bringt , andererseits darin , da``s sie zeit und richtung der elementarprozesse [ der lichtabsorption und emission ] dem ` zufall ' ' ' uberl``a''st ; trotzdem hege ich das volle vertrauen in die zuverl``assigkeit des eingeschlagenen weges''. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ doch f``uhrte eben dieser eingeschlagene weg nach weiteren 10 jahren geradewegs zur heutigen quantenmechanik ( 1925 - 26 ) und quantenelektrodynamik ( 1928 ) , die einstein mit seinen wissenschaftlichen grund''uberzeugungen nicht in einklang bringen konnte insbesondere deshalb , weil in ihnen der zufall als irreduzibler bestandteil der naturerkl``arung auftritt .
doch das genauer zu erl''autern , bed"urfte eines weiteren vortrags .
betrachtet man r``uckblickend die fr''uhe entstehungsgeschichte der quantentheorie , so h``atte sie beim besten willen ironischer nicht sein k''onnen .
wir erinnern uns , da``s ihr ausgangspunkt die gemessenen abweichungen vom wienschen gesetz waren , welches zu diesem zeitpunkt ( f''alschlicherweise ) als strenge konsequenz der klassischen physik angenommen wurde , ma``sgeblich durch die arbeiten von planck .
wie einstein in seiner lichtquantenarbeit aber gezeigt hatte , repr''asentiert das wiensche gesetz gerade den typisch quantentheoretischen teilchenaspekt der strahlung .
die von lummer und pringsheim gemessenen abweichungen vom wienschen gesetz liegen i m langwelligen bereich , in dem das rayleigh - jeans - gesetz ann``ahernd g''ultig ist , das nun tats"achlich eine unabweisbare konsequenz der klassischen physik ist , wie einstein ebenfalls zeigte , und dem wellenbild der strahlung entspricht .
_ etwas `` uberspitzt kann man i m nachhinein also sagen , da ' 's die quantentheorie aus messungen klassischer korrekturen an einem gl``ucklich erratenen quantengesetz entstand , das irrt''umlich f"ur ein gesetz der klassischen physik gehalten wurde .
_ als ausblick sei zum schlu``s noch erw''ahnt , da``s nicht nur in der wissenschaft vom kleinsten , sondern auch in den gr''o``sten uns heute zug''anglichen dimensionen , in der kosmologie , die plancksche strahlungsformel eine zentrale rolle spielt .
so ist ja unser gesamtes universum ein einziger strahlungshohlraum , der erf``ullt ist von einer elektromagnetischen strahlung der temperatur von etwas unter 3 grad kelvin ( etwa -270 grad celsius ) .
diese strahlung entstand etwa 30000 jahre nach dem urknall , als sich aus zun''achst gegenseitig ungebundenen elektronen und atomkernen stabile atome bildeten .
zu diesem zeitpunkt betrug die temperatur etwa @xmath44 grad kelvin .
wegen der best``andigen ausdehnung des universums k''uhlt sich die strahlung stetig ab und hat zur gegenw"artigen epoche den eben genannten wert .
seit einigen jahren werden charakteristische eigenschaften dieser sogenannten `` kosmischen hintergrundstrahlung '' durch satelliten vermessen , da diese eine reiche f``ulle von informationen ' ' uber entwicklung und zusammensetzung unseres physikalischen universums verraten .
nat``urlich wurde dabei auch die spektrale energieverteilung gemessen und mit der planckschen formel verglichen .
das resultat ist in abbildung[fig : cmb ] dargestellt , in der die fehlerbalken auf unnat''urliche 400 ( ! )
standardabweichungen vergr``o''sert wurden , damit sie `` uberhaupt sichtbar sind .
normale fehlerbalken von wenigen standardabweichungen w''aren weniger hoch als die strichdicke der kurve .
damit ist dies die pr"azisest vermessene planckkurve bis zum heutigen tag .
* anh"ange *
es wurde beschrieben , da``s planck bei der theoretischen begr''undung sowohl des wienschen als auch seines eigenen strahlungsgesetzes stets von der beziehung ( [ eq : planck1 ] ) ausging und da``s er die darin auftretende funktion @xmath14 , die die mittlere energie eines resonators der eigenfrequenz @xmath5 i m strahlungsfeld der temperatur @xmath3 angibt , durch die entropiefunktion @xmath20 dieses resonators zu bestimmen suchte .
so ging er auch bei der theoretischen begr''undung seines gesetzes ( [ eq : planck ] ) am 19.dezember 1900 vor .
dazu wandte er die statistische definition der entropie von ludwig boltzmann an .
diese besagt , da``s die entropie eines systems proportional zum nat''urlichen logarithmus des statistischen gewichtes dieses zustandes ist .
letzteres ist definiert als die anzahl @xmath45 der m``oglichkeiten , den ( makroskopisch definierten ) zustand auf verschiedene mikroskopische arten zu realisieren .
dies dr''uckt folgende formel aus ( genannt die `` boltzmannsche '' , die aber erst planck so hinschrieb ) , die man noch heute auf der grabplatte boltzmanns auf dem wiener zentralfriedhof bewundern kann : @xmath46 dabei ist eben @xmath47 die proportionalit``atskonstante zwischen entropie und logarithmus des statistischen gewichtes .
man kann zeigen , da ' 's diese konstante gerade gleich ist dem quotienten aus zwei uns bereits bekannten gr``o''sen , n"amlich der universellen gaskonstante @xmath17 und der avogadro - zahl @xmath18 . in seiner verzweiflung , endlich eine theoretische begr``undung seiner bisher nur gl''ucklich
erratenen strahlungsformel ( [ eq : planck ] ) liefern zu m``ussen , verfiel planck auf den ausweg , die boltzmannsche gleichung ( [ eq : boltzmann - entropie ] ) als definition der entropie zu akzeptieren und sie zur berechnung der entropie eines resonators i m strahlungsfeld zu verwenden .
dazu ging planck so vor : angenommen , es gibt @xmath48 resonatoren der eigenfrequenz @xmath5 , die zusammengenommen in einem zustand der energie @xmath49 sind .
dann ist das statistische gewicht @xmath45 dieses zustandes definiert durch die anzahl der m''oglichkeiten , die energie @xmath49 auf die @xmath48 resonatoren zu verteilen .
physikalisch geht hier die oft nicht explizit genannte , aber dennoch sehr wichtige hypothese ein , da``s jede dieser verteilungen gem''a``s der dynamik des systems i m laufe der zeit gleich h''aufig vorkommt .
w``aren die resonatoren in der lage , kontinuierliche mengen von energie aufzunehmen
und abzugeben , so w''are das statistische gewicht unendlich und formel ( [ eq : boltzmann - entropie ] ) erg``abe ebenfalls keinen endlichen wert .
diesen schlu ' 's kann man durch die annahme umgehen , da``s jeder der resonatoren seine energie nur portionsweise in einheiten einer festen grundmenge aufnehmen und abgeben kann .
sei diese grundmenge @xmath34 , so gibt es also insgesamt @xmath50 energieportionen zu verteilen .
es ist nun eine elementare kombinatorische aufgabe , zu berechnen , wie viele m''oglichkeiten es gibt , @xmath48 portionen energie auf @xmath51 resonatoren zu verteilen .
die antwort ist @xmath52 daraus erh``alt man mit ( [ eq : boltzmann - entropie ] ) die entropie des zustandes aller resonatoren der eigenfrequenz @xmath5 und nach weiterer division durch die anzahl @xmath51 dieser resonatoren die gesuchte entropie eines einzelnen resonators der eigenfrequenz @xmath5 .
das ergebnis kann man ausdr''ucken , die f''ur gro``se
@xmath51 g''ultig ist .
nach planck werden an dieser stelle sowohl @xmath51 als auch @xmath48 als gro``s angenommen .
letzteres ist tats''achlich nicht immer korrekt , was einstein planck sp"ater vorwirft . ]
durch die mittlere energie @xmath53 eines resonators und des noch unbekannten `` energiequantums '' @xmath34 : @xmath54 damit ist die aufgabe fast gel``ost .
denn es gilt in der thermodynamik immer ( unabh''angig davon , ob man die statistische interpretation der entropie zugrundelegt ) , da``s die ableitung der entropie nach der energie gleich dem kehrwert der temperatur ist : @xmath55 wendet man dies auf ( [ eq : entropie - osz ] ) an , so kann man sofort @xmath56 als funktion von @xmath34 und @xmath3 berechnen , was dann eingesetzt in ( [ eq : planck1 ] ) f''ur das strahlungsgesetz liefert : @xmath57 damit dies dann dem planckschen strahlungsgesetz ( [ eq : planck ] ) gleicht , mu``s man eine annahme ' ' uber die tats``achliche gr''o``se der ' ' energiequanten`` @xmath34 machen , was ja bisher noch nicht geschehen ist .
schon aus einem direkten vergleich von ( [ eq : osz - energie ] ) mit der allgemein g''ultigen gleichung ( [ eq : wien1 ] ) ergibt sich , da``s @xmath34 proportional zu @xmath5 sein mu ' 's .
nennt man die proportionalit``atskonstante @xmath25 , die die dimension einer wirkung haben mu ' 's , so hat man gerade ( [ eq : hnu ] ) , und es ergibt sich die plancksche formel .
denkt man sich die planckschen energieportionen als lichtquanten , d.h .
i m raum lokalisierte energiepakete , so entspricht die durch ( [ eq : kombinatorik ] ) ausgedr``uckte abz''ahlung der sogenannten _ bose - einstein - statistik_. an dieser ist bemerkenswert , da``s die lichtquanten als _ ununterscheidbare _
entit''aten behandelt werden , d.h .
es ist egal , welche der @xmath51 lichtquanten den individuellen resonator besetzen , wichtig ist nur ihre anzahl .
f``ur planck war ( [ eq : kombinatorik ] ) jedoch nicht ausdruck einer irgendwie ungew''ohnlichen statistik , da er nicht i m bild der lichtquanten argumentierte .
so bekommt man etwa ( [ eq : kombinatorik ] ) auch als antwort auf die frage , wieviel m"oglichkeiten es gibt , @xmath51 kellen suppe auf @xmath48 ( unterscheidbare ) teller zu verteilen .
der planckschen quantisierungsannahme entspricht hier lediglich die regel , immer nur ganze kellen an suppe zu verteilen .
in diesem anhang wollen wir etwas n``aher ausf''uhren , durch welche mathematische schlu``skette einstein zu seiner lichtquantenhypothese gef''uhrt wurde .
grundlage ist wieder das boltzmannsche prinzip ( [ eq : boltzmann - entropie ] ) . in diesem
d``urfen wir @xmath45 auch durch die wahrscheinlichkeit des makrozustands ersetzen , denn diese ist proportional zur anzahl @xmath45 seiner mikroskopischen realisierungen .
das ersetzen von @xmath45 durch einen dazu proportionalen ausdruck unter dem logarithmus f''uhrt aber zu einer additiven konstanten zur entropie , die in den nachfolgenden `` uberlegungen herausf''allt , da stets nur entropie__differenzen _ _ eine rolle spielen .
als vorbereitung betrachte man ein gas aus @xmath51 atomen in einem volumen @xmath58 bei fester temperatur @xmath3 .
hinsichtlich der dynamik der atome wird nur vorausgesetzt , da``s ihre aufenthaltswahrscheinlichkeit i m volumen konstant ist , da ' 's also teilvolumina gleichen inhalts auch mit gleicher wahrscheinlichkeit von einem atom besetzt werden .
die wahrscheinlichkeit daf``ur , da ' 's sich alle atome in einem teilvolumen @xmath59 befinden , ist dann gegeben durch @xmath60 .
entsprechend hat dieser zustand eine um einen betrag @xmath61 geringere entropie als der `` uber ganz @xmath58 gleichverteilte zustand , wobei @xmath62 eine analoge ' ' uberlegung stellt einstein nun mit w``armestrahlung an , ebenfalls i m volumen @xmath58 bei der temperatur @xmath3 .
dazu mu ' 's er aber den ausdruck f``ur die strahlungsentropie berechnen .
diesen erh''alt er
so : sei @xmath4 die spektrale dichte der energie ( hier als bekannt vorausgesetzt ) und @xmath63 der ( zu bestimmende ) ausdruck f``ur die spektrale dichte der entropie .
das hei''st , da``s der auf das volumen @xmath58 und das frequenzintervall @xmath64 $ ] entfallende anteil der strahlungsenergie durch @xmath65 und der anteil der strahlungsentropie durch @xmath66 gegeben ist .
ganz allgemein gilt in der thermodynamik , da ' 's die ableitung der entropie nach der energie das inverse der absoluten temperatur @xmath3 ist .
das gilt auch f``ur die spektralen verteilungen .
also hat man @xmath67 kennt man das strahlungsgesetz , d.h .
die funktion @xmath4 , so kann man damit auf der rechten seite @xmath68 als funktion von @xmath5 und @xmath69 ausdr''ucken und die gleichung integrieren , wodurch man @xmath70 als funktion von @xmath5 und @xmath69 erh"alt .
einstein benutzt nun nicht das plancksche , sondern das wiensche gesetz ( vgl .
( [ eq : wien2 ] ) , das sich i m grenzfall hoher frequenzen und / oder kleiner temperaturen aus ersterem ergibt .
l``ost man dieses nach @xmath68 auf , setzt es auf der rechten seite von ( [ eq : einlicht1 ] ) ein und integriert einmal nach @xmath69 , so erh''alt man @xmath71+konst.\ ] ] setzt man f``ur die i m volumen @xmath58 und frequenzintervall @xmath64 $ ] enthaltene energie @xmath72 und entropie @xmath73 , so kann man dies auch so schreiben : @xmath74+konst.\ ] ] betrachtet man bei konstantem @xmath56 und @xmath5 die differenz der entropien der strahlung , einmal i m volumen @xmath58 und einmal i m volumen @xmath75 ,
so erh''alt man @xmath76 dies vergleicht einstein mit ( [ eq : gasentropie ] ) und kommt zu dem schlu``s , da ' 's sich w``armestrahlung i m g''ultigkeitsbereich des wienschen strahlungsgesetzes entropisch gesehen so verh``alt , wie ein gas aus @xmath77 atomen ( siehe das zitat einsteins auf seite ) . die ' ' atome`` des lichts hei''sen _ lichtquanten_. sie sind ( i m g``ultigkeitsbereich des wienschen gesetzes ! ) als r''aumlich lokalisiert zu denken und haben die energie @xmath78 wobei wir noch ausgenutzt haben , da``s die konstante @xmath13 des wienschen gesetzes mit der planckschen konstante @xmath25 ' ' uber die boltzmann - konstante @xmath47 gem``a ' 's @xmath79 verbunden ist .
als planck seine formel zum ersten mal niederschrieb , geschah dies ohne kenntnis der erst sp``ater von ihm erdachten ableitung .
vielmehr erhielt er sie , indem er eine gewisse thermodynamische gr''o``se f''ur die wiensche und die rayleigh - jeanssche formel einfach addierte .
eine physikalische interpretation dieses formalen vorgehens hatte planck nicht
wie er sp``ater selber zugab .
erst einstein hat diese interpretation sp''ater geliefert , die einen interessanten aspekt des `` welle - teichen - dualismus '' darstellt .
setzen wir zur abk``urzung @xmath80 und sei @xmath81 wieder die mittlere energie eines resonators ,
jetzt aufgefa''st als funktion von @xmath82 ( statt @xmath3 ) und @xmath5 , so ist die von planck betrachtete gr``o''se gegeben durch die ableitung @xmath83 .
was bedeutet sie ?
bevor wir dies kl``aren , wollen wir sie f''ur die aus den drei strahlungsgesetzen ( rayleigh - jeans , wien , planck ) folgenden ausdr``ucke f''ur @xmath81 berechnen .
es ist @xmath84 also gilt @xmath85 somit ist in der tat f``ur das plancksche gesetz diese gr''o``se ( als funktion von @xmath81 ) additiv aus den entsprechenden ausdr''ucken des rayleigh - jeansschen und wienschen gesetzes zusammengesetzt .
einsteins `` uberlegungen sind nun statistischer natur genauer gesagt betrachtet er statistische fluktuationen der energie des strahlungsfeldes , was man wegen ( [ eq : planck1 ] ) auch auf die resonatoren ' ' ubertragen kann .
dabei ist seine zentrale idee , die boltzmannsche gleichung ( [ eq : boltzmann - entropie ] ) umgekehrt zu lesen , d.h .
das statistische gewicht als funktion der entropie auszudr``ucken .
dr''uckt man die entropie ( bei fester temperatur ) als funktion der energie aus und entwickelt um das dem gleichgewicht entsprechende lokale maximum bei @xmath86 , so kann man daraus in quadratischer ordnung die normierte wahrscheinlichkeitsverteilung f``ur eine energiefluktuation @xmath87 ableiten : @xmath88 hier ist @xmath89 wobei die zweite gleichheit aus der allgemein g''ultigen thermodynamischen relation ( [ eq : entropie - energie - temp ] ) folgt .
also ist das mittlere schwankungsquadrat der energie gegeben durch : @xmath90 damit ist die gr``o''se , die planck seiner formalen interpolation zugrundelegte , als das mittlere schwankungsquadrat der energie erkannt .
dieses verh``alt sich bei der planckschen formel
so , als ob es zwei statistisch unabh''angige ursachen h``atte : die energieschwankungen der rayleigh - jeans - formel , die man mit dem klassischen wellenbild erkl''aren kann , und die der wienschen formel , die dem teilchenbild der lichtquanten entspricht . in dieser hinsicht vereinigt
die plancksche formel beide aspekte in gleichberechtigter weise .
stellt man das gesagte konsequent i m teilchenbild ( lichtquanten ) dar , so kann man statt von resonatorenergien von besetzungszahlen @xmath48 sprechen , indem man jede energie durch @xmath37 dividiert .
aus der letzten zeile in ( [ eq : flukt2 ] ) erh``alt man dann einen einfachen ausdruck f''ur das schwankungsquadrat der besetzungszahl : @xmath91 der erste term w``are alleine vorhanden , wenn es sich um klassisch unabh''angige teilchen handelte , wie man leicht nachpr"uft .
unterscheidbaren teilchen irgendwelche @xmath48 in einem zustand der wahrscheinlichkeit @xmath92 anzutreffen ( z.b .
i m teilvolumen
@xmath59 zu sein , wobei @xmath93 ) , ist @xmath94 . man berechnet nun leicht @xmath95 und @xmath96 .
aus beiden zusammen ergibt sich @xmath97 , was f``ur kleine @xmath92 in @xmath98 ' ' ubergeht . ]
99 niels bohr : `` the structure of the atom '' .
nobel - vorlesung vom 11 .
dezember 1922 ( nobelpreis 1922 ) .
+ online unter @xmath99nobelprize.org / physics / laureates/1922/bohr - lecture.html@xmath100 .
niels bohr , hendrik kramers und j. slater : " `` ` uber die quantentheorie der strahlung''. zeitschrift f``ur physik , 24 ( 1924 ) pp.69 - 87 . arthur h. compton : ' '
a quantum theory of the scattering of x - rays by light elements.`` physical review 21 ( 1923 ) pp.483 - 502 .
arthur h. compton : ' ' directed quanta of scattered x - rays``. physical review 26 ( 1925 ) pp.289 - 299 .
oliver darrigol : ' ' the historians disagreement over the meaning of planck s quantum``. centaurus , 43 ( 2001 ) 219 - 239 .
+ online unter @xmath99www.mpiwg - berlin.mpg.de / de / forschung / preprints.html@xmath100 als preprint nr.150 verf''ugbar .
albert einstein : collected works ( princeton university press ) .
siehe auch die internetseite des `` einstein papers project '' : @xmath99www.einstein.caltech.edu@xmath100 .
gerald grawert : `` quantenmechanik '' ( akademische verlagsgesellschaft wiesbaden 1977 ) .
otto lummer und ernst pringsheim : " ` die vertheilung der energie i m spectrum des schwarzen k``orpers und des blanken platins''. verhandlungen der deutschen physikalischen gesellschaft i m jahre 1899 , erster jahrgang , pp.23 - 41 .
herausgegeben von arthur k``onig , verlag von johann ambrosius barth , leipzig 1899 .
robert millikan : ' ' a direct photoelectric determination of planck s ` @xmath25'.`` physical review , vol.7 ( 1916 ) , pp.355 - 388 .
robert millikan : ' ' the electron and the light - quant from the experimental point of view``. nobel - vorlesung vom 23 .
mai 1924 ( nobelpreis 1923 )
. + online unter @xmath99nobelprize.org / physics / laureates/1923/millikan - lecture.pdf@xmath100 .
abraham pais : ' ' raffiniert ist der herrgott ... `` albert einstein .
eine wissenschaftliche biographie ( vieweg & sohn , braunschweig , 1986 ) .
max planck : physikalische abhandlungen und vortr''age , bd .
i - iii ( vieweg & sohn , braunschweig , 1958 ) .
max planck : brief an robert williams wood von 1931 .
wiedergegeben in " ` fr``uhgeschichte der quantentheorie'' , p. 31 , von a. hermann ( physik verlag , mosbach 1969 ) .
ernst pringsheim : `` einfache herleitung des kirchhoffschen gesetzes '' .
verhandlungen der deutschen physikalischen gesellschaft in jahre 1901 , dritter jahrgang , pp.81 - 84 .
herausgegeben von arthur k``onig , verlag von johann ambrosius barth , leipzig 1901 .
heinrich rubens und ferdinand kurlbaum : ' ' `` ` uber die emission langwelliger w''armestrahlen durch den schwarzen k``orper bei verschiedenen temperaturen ' '' .
sitzungsberichte der preu``sischen akademie der wissenschaften 1900 , gesamtsitzung vom 25.oktober , pp.929 - 941 .
arnold sommerfeld : ' ' das plancksche wirkungsquantum und seine allgemeine bedeutung f``ur die molek''ulphysik``. verhandlungen der gesellschaft deutscher naturforscher und ' ' arzte , 83
. versammlung zu karlsruhe 1911 , zweiter teil , pp.31 - 50 . | am 14.dezember des jahres 1900 berichtete max planck der deutschen physikalischen gesellschaft `` uber seine physikalische interpretation einer harmlos aussehenden , von ihn selbst zuvor aufgestellten formel , die das spektrale verhalten der sogenannten w''armestrahlung beschreibt .
ma``sgeblich durch das eingreifen albert einsteins entwickelte sich daraus i m folgenden vierteljahrhundert eine fundamentale krise der physik , die dann in einer wissenschaftlichen revolution gr''o``sten ausma''ses m"undete : der quantentheorie .
die quantentheorie entwickelte sich von anfang an diametral gegen die intentionen ihrer sch``opfer . f''ur
planck bedeutete sie
trotz gr``o''ster `` au''serer anerkennungen
das vollst``andige scheitern eines langj''ahrigen forschungsprogramms , f``ur einstein letztlich eine absage an seine wissenschaftlichen grund''uberzeugungen .
wir schildern die hintergr``unde dieser seltsamen entwicklung und beleuchten damit die begriffliche seite physikalischer ( und allgemein naturwissenschaftlicher ) forschung , die gemeinhin stark untersch''atzt wird .
um die rolle zu verstehen , die albert einstein bei der entwicklung der quantentheorie gespielt hat , m``ussen wir uns zun''achst die vorangegangenen leistungen plancks vergegenw``artigen , die ihn zur aufstellung seiner ber''uhmten strahlungsformel gef``uhrt haben .
mit dieser gelang ihm die vollst''andige _ quantitative _ aufkl``arung des ph''anomens der _ w"armestrahlung _ , die ihm den nobelpreis des jahres 1918 einbrachte : `` als anerkennung des verdienstes , das er sich durch seine quantentheorie um die entwicklung der physik erworben hat '' .
zu diesem zeitpunkt lag die eigentliche tat schon mehr als 17 jahre zur"uck .
genauer ist sie auf den 14.dezember des jahres 1900 zu datieren .
davon wird weiter unten die rede sein .
etwas weniger bekannt ist die tatsache , da``s diese wissenschaftliche gro''stat plancks gleichzeitig auch die restlose zerschlagung seines langj``ahrigen , akribisch vorbereiteten und meisterhaft durchgef''uhrten forschungsprogramms bedeutete , das in einer tief anti - atomistischen , an absoluten gesetzm``a''sigkeiten sich orientierenden naturauffassung wurzelt .
in der verfolgung dieser ideale legt planck den grundstein zur quantentheorie , die dem konsequenten atomismus zum endg``ultigen durchbruch verhilft und dem element des zufalls eine fundamentale bedeutung innerhalb des gef''uges physikalischer gesetzm``a''sigkeiten zuweist .
hauptmotor dieser entwicklung , die den planckschen vorstellungen diametral entgegenlief , war albert einstein .
und mitunter unverfroren bestand er auf der restlosen kl''arung der begrifflichen grundlagen und konsequenzen der planckschen theorie .
mit seiner lichtquantenhypothese erkl``arte er nicht nur den photoelektrischen effekt , sondern legte den eigentlich revolution''aren kern dieser theorie frei und provozierte so ma``sgeblich eine tiefe krise , die 20 jahre sp''ater in der formulierung der quantenmechanik m``undete .
etwas ' ' ubertreibend , aber i m kern doch zutreffend , kann man sagen , da``s einstein der einzige war , der die plancksche theorie wirklich ernst nahm so ernst , da ' 's die konsequenzen sich schlie``slich auch gegen seine grund''uberzeugungen richteten . |
accretion onto a black hole is a fundamental astrophysical process capable of producing copious amounts of energy from radio @xmath7-ray frequencies .
this radiation in turn will have a profound impact on its surroundings . in the local universe
this is illustrated by the discovery of large jet blown cavities around stellar mass black holes @xcite . when integrated over the lifetime of the black hole
, this energy will have a significant impact on its environment , e.g. , feedback from supermassive black holes ( smbh ) in the quiescent state ( @xmath8 ) ] is required to reproduce the observed properties of galaxies @xcite .
the observation of tight correlations between the smbh and properties of the host galaxy are further evidence for the importance of feedback from the black hole , and understanding this will provide insight into the process of galaxy formation and evolution @xcite . at higher redshifts still
, the accretion power of early hmxbs could have an impact on the formation of the earliest stars and galaxies @xcite .
there are over 500 accreting binaries currently known in the milky way , though this sample is dominated by systems containing neutron star accretors .
observations of other galaxies have also revealed large populations of accreting objects @xcite , which can be used to reveal information on the star formation history @xcite and stellar content of these galaxies @xcite .
the x - ray luminosity function of the milky way can be used to constrain the population and evolution of x - ray emitting sources in the galaxy @xcite , which is found to be dominated by accreting binary systems with compact object accretors ( bh , ns & wd , @xcite ) .
detailed study of the accretion flow in galactic x - ray binaries can provide constraints on the nature of accretion in the presence of a large gravitational potential , and the theory of general relativity itself @xcite .
in particular , they provide a nearby window on black hole spin @xcite , the production mechanism of the ubiquitous relativistic jet emission @xcite , and even the nature of supernovae ( sne ) and gamma - ray bursts ( grbs ; @xcite ) . to date ,
there are 20 confirmed stellar mass black holes ( i.e. , with dynamical constraints on the mass of the bh ) and greater than 20 strong candidate systems , e.g. , see @xcite . bh x - ray binaries ( xrbs ) are typically transient systems that are observed to brighten from @xmath9 on timescales of weeks to months before returning to a low luminosity state .
the galactic black hole binaries have been classified based on their behavior in the soft x - ray band ( 0.1 10 kev ) .
this resulted in the definition of a number of apparently discrete spectral states @xcite .
when a black hole is accreting at an appreciable percentage of its eddington luminosity ( @xmath10 ) , 2 primary spectral forms are observed .
the first of these is the low - hard state , characterized by a power - law continuum in the 2 10 kev band with a spectral index @xmath11 , and a spectral cut - off at energies @xmath1 100 kev .
the rms spectral variability is observed to be high in this state .
the second state is the so - called high - soft state .
the x - ray spectrum is dominated by emission in the soft band ( @xmath12 ) consistent with a thermal blackbody component , a softer low luminosity power - law component at higher energies ( @xmath13 ) , and low rms noise @xcite .
the launch of the _ rxte _ mission @xcite resulted in a refinement in our understanding of these spectral states , and the transitions between them , which have resulted in the definition of a number of intermediate spectral states based on both the spectral state and the timing properties of the source @xcite .
the broadband emission from accreting black holes is correlated with the behavior at x - ray wavelengths , as clearly illustrated by the observed phenomenology of the radio jet @xcite .
studies of jets from accreting black holes on all mass scales have led to the realization that the accretion disk ( inflow ) and the jet ( outflow ) are closely related , i.e. , ` fundamental plane of black hole activity ' @xcite and ` disk - jet coupling ' @xcite . in particular it has been noted that ( i ) the low - hard state ( @xmath14 ) is typically associated with the presence of a compact quasi - steady jet ( ii ) radio / jet emission appears to cease when the system enters the high - soft state ( @xmath15 ) .
this suggests that the accretion disk inflow is intimately involved in the process of launching a jet , e.g. , @xcite . at higher energies ( @xmath16 ) , the large archive of high energy observations by _ rxte _
@xcite , observations by _ integral _ through the bulge monitoring program @xcite , and pointed observations are helping to elucidate the nature of the high energy x - ray component .
_ suzaku _ observations have demonstrated the importance of having broadband spectral coverage , where the broad bandpass ( 0.6 300 kev ) has proven to be crucial to simultaneously constrain the soft accretion disk component , the hard x - ray emission and the reflection features , i.e. , fe k line , compton hump e.g. , see @xcite for high energy spectra and @xcite for constraints on the reflection component .
70% of the observations having an exposure time less than 2 ks and 90% less than 4 ks .
the total exposure time is 940 ks .
the hatched histogram shows the total number of counts for the spectra in our sample ( see x2 axis ) .
the peak is at @xmath1 60k counts , while 70% and 90% of the spectra had greater than 3.5k and 500 total counts respectively .
the number of counts in the background relative to the source is negligible in all cases . ]
understanding the relationship between the accretion disk inflow and the jet outflow is one of the outstanding goals of modern astrophysics .
the anti - correlation between the accretion state of the black hole and the presence or absence of a relativistic jet indicates that understanding the geometry of the accretion flow in each of the primary accretion states will be key , i.e. , the low - hard state and the high - soft state .
the current paradigm for the configuration of these accretion states is consistent with a hot geometrically thin optically thick multi - temperature blackbody accretion disk @xcite extending to the innermost stable circular orbit ( isco ) in the soft state . as the accretion rate and hence disk temperature drop ,
the inner edge of the accretion disk recedes from the isco and the inner region is filled with a geometrically thick adaf ( advection dominated accretion flow , @xcite ) , which produces the observed hard spectral component that dominates the low - hard state x - ray spectrum @xcite .
however , a number of recent observations contradict this picture and reveal evidence for the persistent of the accretion disk at the isco in the low - hard state , e.g. , @xcite .
these observations suggest that the accretion disk is not the primary driver of the observed low - hard @xmath17 high - soft state transition .
indeed , the latest observations suggest that the disk does not recede from the isco until luminosities of @xmath18 @xcite .
a number of alternative scenarios are consistent with the accretion disk remaining at the isco , e.g. , the hard x - ray flux may originate in the relativistic jet @xcite or from a hot corona @xcite , and it is the turning on / off of the jet or evolution of the corona that drives the observed state transitions . questions also persist at longer wavelengths .
recent observations have highlighted a problem in our understanding of the uv emission mechanism in x - ray binaries .
standard theory explains the uv emission as direct emission from the accretion disk and observations of the classical @xmath19 spectrum in early iue observations supported this picture @xcite .
it was soon realized that the large flux of x - rays present in an x - ray binary should modify the standard disk spectrum , with the x - ray emission irradiating the disk , where it is subsequently reprocessed disk reprocessing ( e.g. @xcite ) .
a standard steady state accretion disk will be observed to exhibit a radial temperature dependence of the form @xmath20 @xcite .
it is inherently difficult to constrain the temperature profile of the accretion disk at x - ray wavelengths as the primary modification to the accretion disk spectrum occurs at wavelengths to the red of the typical x - ray detector , i.e. , uv / optical or @xmath21 .
this is illustrated in @xcite who studied a sample of _ asca _ observations of black holes in outburst and found the temperature profile of the measured spectra to be consistent with the standard @xmath22 accretion disk at all times , though there was evidence , of low statistical significance , for a deviation towards a more irradiated disk in some of the observations .
in contrast , observations at optical and uv wavelengths have revealed clear signatures for disk irradiation @xcite .
indeed , the addition of x - ray irradiation of the accretion disk to the disk instability model is required in order to reproduce the observed outburst lightcurves of black holes xrbs @xcite . over the last decade
the picture has been further complicated by the realization that the radio jet may contribute significantly across the entire spectral energy distribution , e.g. , @xcite .
furthermore , simultaneous observations at optical and x - ray wavelengths have revealed complex correlation functions , hinting at the presence of a non - thermal component in addition to the accretion disk emission , e.g. , @xcite .
@xcite has recently carried out a comprehensive study of black hole x - ray binaries using over 13 years of archival _ rxte _ observations . in @xcite ,
hardness intensity ( hid ) and disk fraction luminosity diagrams ( dfld ) were constructed for 25 black holes .
@xcite focused on the behavior of the accretion disk component in each of the outbursts studied in the previous paper .
in particular , they analyzed the observed distribution of temperature versus luminosity for each outburst and found it to be consistent with @xmath23 , and with previous studies , which found the inner radius of the accretion disk to be constant in disk dominated states .
they do note a deviation from this relation at the entry to / exit from the disk dominated state , which they suggest is consistent with a varying color correction factor ( e.g. , @xcite , see [ accretion_disk]iii for further details ) .
however , due to the 3 kev lower energy limit of _ rxte _ , this study was necessarily restricted to relatively hot accretion disks ( @xmath24 ) . despite the significant progress made over recent years ,
a number of unanswered questions remain . in this study , we focus our attention on the following issues : a. does the accretion disk recede from the isco during the soft to hard state transition ?
b. what is the nature of the accretion geometry in the low hard state , i.e. , is the accretion disk truncated ?
what is the source of the observed hard x - ray flux , a jet or a corona ?
c. what produces the uv emission ,
does it originate from the accretion disk or is there a non - thermal contribution ?
the _ swift _ observatory was designed to discover and provide rapid multi - wavelength follow - up of gamma ray bursts ( grbs ) .
these same attributes allow detailed study of many other time variable phenomena , in particular accreting stellar mass black holes . due to the rapid pointing capability and
low pointing overhead _ swift _ provides the first detailed ccd resolution monitoring of black hole binaries in outburst . combined with the strictly simultaneous observations at optical / uv wavelengths provided by uvot ,
_ swift _ is the ideal platform to study the accretion process in detail . in comparison to _ rxte _ , the low energy x - ray cutoff allows us to study the accretion disk across the entire outburst cycle , whereas previously the cool disk would have been outside the rxte low energy cutoff ( @xmath1 3 kev ) . in this paper
, we describe observations of black hole binaries undertaken with the _ swift _ observatory prior to june 2010 . in 2 , we describe the observations and extraction of source spectra and lightcurves .
we proceed to analyze the data in 3 and the results are presented in 4 . in particular , we find that there is no evidence for significant truncation of the accretion disk , at least to luminosities @xmath25 . in the broadband data , we find evidence for an irradiated accretion disk at all luminosities probed , but increasing as we enter the low - hard state .
these results are discussed in the context of models for the accretion flow in 5 , and finally our conclusions are presented in 6 .
our data sample consists of _ swift _ observations of all known and suspected stellar mass black hole candidates observed through august 2010 ( see table [ bhc_table ] ) .
data for each system was obtained via the _ heasarc _ archive service .
the _ swift _ observatory @xcite is a highly versatile multi - wavelength platform , providing rapid pointing capabilities and simultaneous broad wavelength coverage from optical to hard x - ray energies .
the _ swift _ observatory contains three instruments : _ ( i ) burst alert telescope ( bat , @xcite ) : _ a coded mask hard x - ray ( @xmath16 ) imaging telescope , covering the energy band from 15 150 kev .
low resolution spectra are also provided at higher fluxes . _
( ii ) x - ray telescope ( xrt , @xcite ) : _ a 3.5 m focal length wolter i type grazing incidence x - ray telescope , containing a 600 pix@xmath26 ccd ( @xmath1 2.35@xmath27 pix@xmath28 , i.e. , 23 arcmin@xmath26 ) , sensitive over the energy range 0.2 10 kev at the telescope focus , with an effective area of @xmath1 135 , 65 , 20 @xmath29 at 1.5 , 6.4 and 8 kev respectively .
bright sources are typically problematic to observe with ccd based detectors ( see [ pile_up ] ) .
the xrt uses a number of different read - out modes to avoid / mitigate this problem , facilitating observations of sources with fluxes in excess of @xmath1 1 crab ( @xmath30 ) .
these may be divided into windowed timing ( wt ) and photon counting modes . in windowed timing mode
only spectral information is provided as the central 8 arc - minutes of the ccd is collapsed into a single spatial direction , providing 1.8 ms timing resolution .
this is the primary mode used at high fluxes , with the observations presented herein primarily utilizing wtw2 mode .
photon counting mode ( pc ) is the standard method of reading out a ccd providing both spectral and 2 dimensional spatial information but with a limited time resolution of only 2.5 s. the majority of photon counting mode observations herein are in pcw2 mode . _ ( iii ) ultra - violet / optical telescope ( uvot , @xcite ) : _ simultaneous observations are obtained at optical / uv energies with the uvot instrument ( 30 cm ritchie chretien telescope ) . the detector is a complex system consisting of a photodiode , micro - channel plates and a ccd ( see the instrument paper for details ) .
the ccd provides a 17 arcmin@xmath26 fov and a frame time of 10.8 ms .
coincidence loss occurs at high count rates , typically those in excess of 20 ct s@xmath28 , but this is not an issue for the optical / uv counterparts to the black hole binaries in our sample .
observations are normally obtained in only an single filter ; however , there are some sources with multi - wavelength lightcurves .
the systems in our sample contain data in one or more of the 6 uvot photometric filters v , b , u , w1 , m2 , w2 ( @xmath31 5468 , 4392 , 3465 , 2600 , 2246 , 1928 ) .
cc|cc + & + + + 0.5 1 ct s@xmath28 & 2 pix & 150 200 ct s@xmath28 & 3 pix + 1 3 ct s@xmath28 & 5 pix & 200 300 ct s@xmath28 & 5 pix + 3 6 ct s@xmath28 & 7.5 pix & 300 500 ct s@xmath28 & 10 pix + 6 9 ct s@xmath28 & 10 pix & @xmath32 500 ct s@xmath28 & 15 pix + [ 0.5ex ] + our sample comprises 27 stellar mass black holes .
these range from those systems with dynamical constraints on the mass of the compact object , i.e. , @xmath33 ( e.g. , cyg x-1 , gro j1655 - 40 ) , to a large number of systems that are more correctly labeled as candidate black holes
. these are systems that have displayed spectral / photometric attributes consistent with those of the known black hole systems , e.g. , soft spectrum , timing properties , lack of pulsations or x - ray bursts , but lack a dynamical constraint on the mass of the compact object , e.g. , 4u 1957 + 11 , swift j1753.5 - 0127 .
the results presented herein are necessarily biased by those objects with the majority of observations , i.e. , cyg x-1 , gx 339 - 4 , xte j1817 - 330 , xte j1752 - 223 and swift j1753.5 - 0127 .
however , these object are distributed relatively uniformly over the relevant binary parameters , e.g. , hmxb vs lmxb , @xmath34 etc . , and
hence do not overly bias our results .
the known system parameters for each of the black hole binaries in our sample are listed in table [ bhc_table ] .
a typical _ swift _ pointing is short with an exposure time less than 2 ks .
for the purposes of this study , we ignore all pointings for which the total number of counts detected by the xrt is less than 100 .
thus our sample consists of the following : * 476 observations of 27 bhs with @xmath35 100 counts in the 0.6 10 kev band . * 255 observations for which we have simultaneous detections at optical / uv wavelengths , in at least one band ( v , b , u , w1 , w2 , m2 ) .
the lower number of optical / uv observations reflects the high column density through which some of these systems are observed , with many systems having @xmath36 . in fig .
[ histogram_expos ] , we plot the histogram of exposure times and total number of detected counts for all of the observations considered herein .
the exposure time distribution peaks at an exposure time of @xmath1 1 ks , with @xmath32 70% and 90% of the exposures being less than 2 ks and 4 ks in length respectively . the total counts distribution peaks at @xmath1 60k counts , with 70% and 90% of the spectra containing greater than 3.5k and 500 total counts respectively . in fig .
[ bh_all_systems ] , we plot the absorbed flux vs time for all of the systems in our sample .
we are sensitive to fluxes over a dynamic range of @xmath37 .
we aim to use the data sample described herein ( see [ sample_characteristics ] ) to probe the physics of accretion onto black holes . to begin , we outline the steps taken to extract the final spectral and photometric data products .
all data reduction and analysis was performed using the heasoft 6.9 suite , which includes ftools 6.9 , swift 3.5 and xspec 12.6.0k .
the latest versions of the relevant _ swift _ caldb files are also used .
data were downloaded from the heasarc archive using the coordinates listed in table [ bhc_table ] , where the search radius was restricted to 10@xmath38 .
all raw - data were reprocessed using the ` xrtpipeline ` to ensure usage of the latest _ swift _ caldb when producing the relevant observation event files .
spectra were then extracted using ` xselect ` from circular and annular regions in the photon counting mode data , while rectangular regions were used to extract spectra from the windowed timing mode data .
exposure maps were generated for each observation using the ` xrtexpomap ` task .
these are particularly important due to the presence of a number of dead columns near the center of the xrt ccd due to a likely micro - meteorite impact in may 2005 @xcite .
background spectra were subsequently extracted from neighbouring regions .
ancillary response function files ( .arf ) were created with the ` xrtmkarf ` task , while the appropriate response matrix ( .rmf ) was sourced from the calibration database .
these files were then grouped together with ` grppha ` and exported to xspec for spectral fitting . for our best fit ` pha(diskbb+po ) ` model .
the dashed histogram is the model containing the power - law component alone , while the blue histogram denotes the best fit model after the addition of the ` diskbb ` component .
an accretion disk is required at greater than the 5@xmath0 confidence level in @xmath32 61% of the observations .
fits with alternative continuum components reveal similar results , see table [ chi2_table ] . ]
due to the photon counting nature of modern x - ray ccds , they are strongly susceptible to a phenomenon known as ` pile - up ' .
this occurs at high fluxes , where multiple photons can impact a detector pixel at approximately the same time ( i.e. , less than the detector frame time ) , and be recorded as a single higher energy event .
this will produce an excess of hard photons in the measured spectrum , which as a result will appear harder than is actually the case .
proper interpretation of the observed spectra requires a strategy to mitigate the adverse pile - up related effects . for further details
see @xcite for a detailed discussion of pile - up in x - ray ccds , including _
chandra , xmm - newton and suzaku _ in addition to the _ swift _ xrt . in our analysis , pile - up
is corrected in the standard manner , i.e. , by the extraction of annular regions centered on the source position .
the excluded region radius was chosen to be conservative and agrees with those used by other groups , e.g. , @xcite .
the exclusion regions utilized are listed in table [ pile_up_table ] along with the corresponding count rates . in order to confirm the effectiveness of this strategy
, we ensured that the chosen exclusion radius was such that the best fit spectral parameters did not exhibit significant variance in comparison to the next largest exclusion region .
the sample we present here is dominated by observations of black holes in the spectrally hard state , which typically results in lower observed fluxes in comparison to the soft state . in total , approximately 20% of our spectra are effected by pile - up of some form when classified using the count rates in table [ pile_up_table ] .
the majority of this occurs at the lower count rate levels , e.g. , only 2.5% & 4.5% of our observations required the largest and second largest exclusion regions respectively .
the final spectra that we choose to analyze are those where the quantity of interest is robustly on the conservative side , for example , @xcite exclude only the inner pixel for count rates in the range 100 300 @xmath39 and 2 pixels for 300 400 @xmath40 and 4 pixels for counts rates @xmath41 .
the optical & uv photometry required minimal additional processing @xcite .
we begin by utilizing the level ii pipeline processed image file ( sk.img ) .
images are aspect corrected via the ` uvotskycorr ` tool , where the images are registered to the usno - b1 catalogue @xcite . source flux is extracted from an aperture of 5@xmath27 radius and the background is extracted from a neighbouring source free sky position .
the counterparts to the x - ray sources are easily identified , primarily due to the transient nature of the optical / uv counterpart in combination with the accurate x - ray position provided by the xrt .
this also allows us to detect any contamination from possible line of sight stars contributing to the 5@xmath27 source aperture .
source flux and magnitudes are calculated using ` uvotmaghist ` tool .
these fluxes are converted to xspec readable ` .pha ' files with ` uvot2pha ` , which are then exported for spectral fitting .
the are a large number of models available in xspec for modelling the x - ray emission from black hole binary systems .
as outlined in the introduction , the standard model for black hole x - ray binaries involves the combination of soft emission ( @xmath42 2 kev ) from a thermal accretion disk and hard x - ray ( @xmath32 2 kev ) emission from power - law like component , typically attributed to compton scattering of the thermal photons from the accretion disk to higher energies @xcite , or perhaps emission from a jet @xcite . in an effort to model the data in a manner that minimizes bias , we choose two complimentary approaches : \(i ) phenomenological here we aim to determine the empirical relationships displayed in the data , which can be used to inform / constrain theoretical models and characterise source types .
for example , is the system spectrally hard or soft ?
a very soft x - ray spectrum is one of the characteristic signatures of an accreting black hole in the soft state .
additionally , we can create color - color and hardness intensity diagrams to search for characteristic variability and search for correlation at other frequencies , i.e. , optical / uv .
\(ii ) physical using these models we will attempt to constrain the intrinsic properties of the accretion flow , e.g. , accretion disk temperature and inner radius . for the soft component , we assume a standard steady state , geometrically thin , optically thick accretion disk @xcite , described by the ` diskbb ` model @xcite .
extensions to this base model are also considered , e.g. , corrections for gr effects in the inner disk region close to the black hole ( @xcite , @xcite , @xcite , or via the ` diskpn ` model @xcite ) and accounting for possible changes in the temperature profile of the disk due to , for example , irradiation by the corona or flux generated otherwise close to the black hole ( ` diskpbb ` , @xcite ) .
these models have the advantage of simplicity while incorporating the important physics , and a large archive of published observations with which the resulted presented herein may be readily compared .
detailed state of the art models of accretion disks have been developed , e.g. , ` bhspec ` @xcite and ` kerrbb ` @xcite ; however , systematic uncertainties remain ( e.g. , @xcite ) and a detailed investigation of these issues is beyond the scope of the present work .
the hard x - ray component , i.e. , that which typically dominates above @xmath1 2 kev , is modelled using 2 distinct prescriptions . the most basic model for the hard x - ray flux is a simple power - law , characterized by a spectral index , @xmath43 .
while this model offers the simplest method to characterize the hard x - ray flux , at low energy it does not exhibit a cut - off as one would expect if the hard x - rays are generated via compton scattering of the soft x - ray photons from the accretion disk .
this is an important issue in the current analysis due to the sensitivity of the _ swift_/xrt to energies as low as 0.5 kev . in this case
, the hard component may be modelled assuming it originates via scattering of the low temperature seed photons from the accretion disk by a high temperature corona .
detailed comptonization models including reflection ` compps ` @xcite and non - thermal electron contributions ` eqpair ` @xcite are available ; however , given the upper energy bound of the xrt ( 10 kev ) , we choose to use the most computationally efficient thermal comptonization model ( ` comptt ` , @xcite ) . in the interest of completeness , we also consider the ` simpl ` comptonization model of @xcite .
this model differs from the ` comptt ` model in that it allows one to specifically choose the input model for the seed photons , whereas the seed photons are assumed to originate from the wein tail of a blackbody in the ` comptt ` case . where simultaneous uvot / xrt data are available , joint fits are undertaken in an effort to constrain the nature of the broadband x - ray to opt / uv spectrum .
the extinction in this case is modelled using the xspec ` phabs ` & ` redden ` models for the x - ray and opt / uv data respectively .
the standard galactic extinction curve was assumed @xcite along with the gas to dust ratio from @xcite , i.e. , the reddening was set to @xmath44 .
we note that in sightlines towards the galactic center the column density is highly variable and as such the standard extinction curve may not be appropriate for all sources in our sample .
are assumed .
the unabsorbed luminosity is consistent with a normal distribution peaking at @xmath1 1% eddington . ] for our baseline model , we utilize a blackbody accretion disk in addition to a power - law to model the x - ray spectrum ( i.e. , ` diskbb+po ` ) .
this model will also aid in the comparison of our results to the significant archive of _ rxte _ observations , e.g. , @xcite .
to model the absorption by intervening neutral hydrogen , we use ` phabs ` , where throughout this paper the abundances and cross - sections assumed are ` bcmc ` @xcite and ` angr ` @xcite respectively .
the column density is held fixed at a value consistent with previous observations , see table [ bhc_table ] for details .
this is in agreement with the observed absence of significant local absorption in these systems @xcite , with obvious exceptions for the case of the wind - accreting hmxbs .
spectra with less than 100 total counts were ignored in all further spectral fitting .
such spectra are typically those obtained with the xrt operating in auto - mode , i.e. , the first observation takes place in windowed timing mode and if the count rate is safe the observation is stopped and switched to photon counting mode , sometimes resulting in an under - exposed or piled - up spectrum . the best fit statistic is chosen based on the total number of counts in the spectrum .
standard chi - squared fitting is used for observations with more than 500 counts .
all spectra with sufficient counts to meet our @xmath45 fitting criteria were binned so as to have a minimum of 20 counts in each bin via the ` grppha ` task thus ensuring that use of this statistic is valid .
the cash - statistic was utilized for those observations with less than 500 total counts @xcite .
finally , the spectral fitting is restricted to the energy range 0.6 10 kev . are assumed .
although the overall number is small , this distribution appears to peak above 10% eddington with a significant tail to lower luminosities consistent with previous observations @xcite . ]
lcc + [ -2.0ex ] model & @xmath46 & @xmath47 + [ 0.5ex ] + [ -2.0ex ] ` pha*(po ) ` & 0.47 & 0.58 + [ 0.5ex ] ` pha*(comptt ) ` & 0.53 & 0.68 + [ 0.5ex ] ` pha*(diskbb+comptt ) ` & 0.83 & 0.90 + [ 0.5ex ] ` pha*(diskbb+po ) ` & 0.88 & 0.93 + [ 0.5ex ] ` pha*(diskpn+po ) ` & 0.85 & 0.94 + [ 0.5ex ] lcccc + [ -2.0ex ] system & @xmath48 & @xmath49 & @xmath50 & comment + [ 0.5ex ] + [ -2.0ex ] gro j1655 - 40 & 115 & 20 & 6 & largest amplitude + [ 0.5ex ] swift j1842.5 - 1124 & & 30 & & slow decay + [ 0.5ex ] xte j1817 - 330 & 119 & 32 & 5 & brightens @xmath1 60 days + [ 0.5ex ] & & & & const after @xmath1 130 days + [ 0.5ex ] xte j1752 - 223 & & 20 & & highly variable + [ 0.5ex ] h1743 - 322 & & & 4 & fast decay + [ 0.5ex ] swift j1539.2 - 6277 & 26 & 35 & 7.5 & slow decay for @xmath1 20 days + [ 0.5ex ] gx 339 - 4 & 15 & 35 & 7.5 & flattens at @xmath1 30 days + [ 0.5ex ] cyg x-1 & 5 & 5 & 120 & const after @xmath1 5 days + [ 0.5ex ] swift j1753.5 - 0127 & 28 & 20 & 40 & const after @xmath1 80 days + [ 0.5ex ] grs 1915 + 105 & & & 1 & very fast decay + [ 0.5ex ] we then proceed to fit each spectrum as follows :
( i ) the spectrum is loaded in xspec , the number of counts are determined and if this is greater than 100 the fitting statistic is then chosen .
( ii ) the initial continuum model is fit to the data assuming a constant value for the interstellar column density as discussed previously , e.g. , a power - law ` pha*po ` .
best fit parameters and the associated errors are determined in addition to the model flux ( iii ) an accretion disk component is then added to the model in the previous step , the best fit parameters of this new model are determined , e.g. , ` pha*(diskbb+po ) ` .
( iv ) an f - test is carried out to determine if the new 2 component model is a statistically significant improvement on the previous single component model , if the second component is required error scans are undertaken and fluxes are calculated .
the above fitting procedure was then repeated , replacing the continuum components above with those indicated earlier , i.e. , ` diskpn ` or ` diskpbb ` in place of the ` diskbb ` and/or ` comptt ` to replace the power - law component .
the seed photon temperature for the ` comptt ` component is set to equal the temperature of the accretion disk , i.e. , @xmath51 .
a subset of these fits were also repeated with the column density as a free parameter to check the sensitivity of our fitting methodology to this parameter . as the xrt bandpass is limited to energies below 10 kev , we are insensitive to the presence of a spectral cut - off at higher energies . as such , we are unable to simultaneously constrain both the electron temperature ( @xmath52 ) and optical depth ( @xmath53 ) of the ` comptt ` model .
we therefore choose to freeze the electron temperature to 50 kev in all fits using the ` comptt ` component unless otherwise indicated .
this will result in the production of a power - law like spectrum in the 2 10 kev range .
additional fits were repeated at other electron temperatures to confirm the robustness of the fits to changes in this parameter . in the case of the opt / uv
+ x - ray fits , the basic procedure was similar .
the x - ray and uvot data are loaded into xspec and the model is initialized at the relevant best fit model from the x - ray fits alone .
the extinction at optical and ultra - violet wavelengths is modelled using the ` redden ` component , i.e , ` pha(po)+redden(po ) ` , and we relate this to the x - ray column density following @xcite , i.e. , @xmath44 .
all of the above models were defined and fit using custom ` tcl ` scripts .
errors where quoted are calculated via the ` error ` command and are equivalent to the 90% confidence interval unless otherwise explicitly stated . has been set to the time of maximum flux indicated by the shaded region ( see also fig . [ bh_all_systems ] ) .
each system has been offset by an arbitrary amount for clarity .
the symbol / color are as defined in fig .
[ bh_all_systems ] . ]
the primary results from this study relate to the analysis of the x - ray and uv spectra of each of the black hole systems in our sample .
we briefly discuss the results of the x - ray spectral fits here .
the spectra were fit with 2 families of models , one with a power - law representing the hard spectral component and the other with the power - law component replaced with a comptonization model . in fig .
[ chi_diskbb_po ] , we plot the @xmath54 distribution for the spectra when fit with a power - law only , and after the addition of a disk component , i.e. , ` po ` @xmath55 ` diskbb+po ` .
it is immediately apparent that a disk component is statistically required in a large number of observations . in total
, we had 476 observations with @xmath35 100 counts that were fit with the above model , and of these 445/476 ( @xmath1 94% ) returned a best fit @xmath56 .
an accretion disk is required at greater than the 5@xmath0 confidence level , as measured by an ` ftest ` , in 282/445 ( 63% ) of the observations with a best fit @xmath57 .
a further 30 observations also required a disk , but in this case the `` best '' fit is still poor , i.e. , @xmath58 . in table
[ chi2_table ] , we list the percentage of spectra below best - fit cut - offs of @xmath59 1.5 and 2 for our sample of spectral models .
we find that fits with the hard component represented by the ` comptt ` spectral model return best fit values consistent with that returned by those modelled with a ` po ` component .
this is to be expected , as we are unable to place constraints on the electron temperature of the comptonizing corona due to the lack of data above 10 kev .
nonetheless , the accretion disk component is still strongly required in this case , though the disk properties are slightly different to those found with the power - law component ( see [ accretion_disk ] ) .
likewise , replacing the ` diskbb ` component above with a ` diskpn ` model does not provide a statistically significant improvement in the best fit results . before proceeding to a detailed discussion of the results of our spectral analysis
, we first present the photometric properties of our sample .
the morphology of a black hole binary x - ray lightcurve contains information which may be used to constrain the outburst , e.g. , outburst energy , @xmath60 etc . in fig .
[ bh_all_systems ] , we plot the absorbed flux for each of the 476 pointed observations in our sample containing greater than 100 total counts .
we are sensitive to outbursts over a dynamic range of greater than 4 orders of magnitude , and including the extra - galactic systems ( ngc 300 x-1 , ic10 x-1 , m33 x-7 ) , we are sensitive to accreting stellar mass black holes over 6 orders of magnitude in flux . in fig .
[ histogram_luminosity ] the 0.6 10 kev luminosity histogram of all observations in our sample is plotted , in both absorbed and unabsorbed form .
there is a clear peak in the distribution at a luminosity of @xmath1 1% eddington .
our limiting luminosity is @xmath61 , though there are a number of isolated observations an order of magnitude fainter ( due to gro j1655 - 40 , which experienced a large amplitude outburst , see fig .
[ bh_all_systems ] ) .
the unabsorbed luminosity equals or marginally exceeds the eddington limit for a small sample of observations , which correspond to the known eddington limited persistent source grs 1915 + 105 @xcite , while the 2005 outburst from the transient black hole gro j1655 - 40 also approached this limit . in terms of flux
, the distribution is broader owing to the varying distances of the detected sources .
this peak corresponds to an unabsorbed flux of @xmath62 .
the distribution of the maximum luminosity observed during each outburst is displayed in fig .
[ histogram_peak_luminosity ] .
a peak is observed at a luminosity of @xmath1 20% @xmath63 , this is consistent with the earlier observations ( @xmath64 @xcite , @xmath65 @xcite ) ; however , small number statistics preclude any deeper analysis . in fig .
[ lc_plot ] , we plot the lightcurves for 11 outbursts in our sample , where we have set @xmath66 to be the time of observed outburst maximum . due to the variable and transient nature of the black hole binaries in our sample
, we note that our estimate of @xmath66 is inherently uncertain for those sources without observations prior to the peak , i.e. , xte j1817 - 330 , h1743 - 322 , grs 1915 + 105 .
in addition , due to uncertainty in assignment of @xmath66 , our interpretation of the lightcurves is also biased by the sampling and the duration of the outburst for which the system continued to be observed .
this is particularly noteworthy for the cases of swift j1842.5 - 1124 , swift j1539.2 - 6277 , h1743 - 322 , grs1915 + 105 , and to a lesser extent grs j1655 - 40 .
the lightcurves are characterized by fitting with a model assuming an exponential decay with timescale @xmath49 @xcite .
a break is added at late times ( @xmath67 ) where necessary , i.e. , @xmath68 the results of these fits are listed in table [ lc_morph ] .
our sample of lightcurves includes 2 persistent systems that display plateau type lightcurves ( cyg x-1 , swift j1753.5 - 0127 ) .
three systems display an exponential decay type lightcurve followed by a break a later times ( gro j1655 - 40 , xte j1817 - 330 , xte j1752 - 223 ) .
swift j1842.5 - 1124 & swift j1539.2 - 6277 also display exponential decay type lightcurves but with a characteristic decay timescale much shorter than the 3 previous systems .
h1743 - 322 was likely observed late in the decay phase of the outburst as the lightcurve is consistent with that of gro j1655 - 40 after the late time break . finally , we display 2 ` outbursts ' from grs1915 + 105 .
these events display a much shorter timescale than that observed from the other systems , although the event is roughly consistent with that seen from cyg x-1 during its transition to the soft state .
perhaps such flaring type variability is a characteristic of persistent accretion flows .
the characteristic exponential decay timescales is similar across a number of systems , lying in the 20 40 day range .
this is consistent with findings of the study of xrb lightcurves by @xcite .
the typical outburst durations of @xmath1 50 100 days ( fig .
[ lc_plot ] ) , imply a total outburst fluence in the range @xmath69 for the galactic stellar mass black hole binary systems . assuming the standard accretion efficiency ( @xmath70 ) , we find the typical amount of matter accreted per outburst to be @xmath71 . though the actual amount may be lower than this as the accretion efficiency is expected to be lower than 0.1 in the hard state , e.g. , due to the presence of outflow / jets and/or an adaf . taking this into account
, the measurement of an accretion rate onto the outer disk in quiescent black holes of @xmath72 would support this @xcite . in this section
, we create a number of diagrams that provide significant insight into the behavior of the accreting black holes in our sample .
we begin by considering the source evolution in color - color diagrams . _
i. color color diagrams_[cc_diagram ] + color - color diagrams @xcite are a useful means by which to characterise the spectral behavior and evolution of accreting black holes and neutron stars in a model independent manner .
though , we note that the results are detector and bandpass dependent , which demands care when making inter - mission comparisons , e.g. , _ swift _ vs _
rxte_. .
] in fig .
[ color_color_plot ] , we plot color - color diagrams for the systems in our sample , based on the results of the ` diskbb+po ` model and the ` diskbb+comptt ` model .
the soft colour is defined as ( 2.5 kev
4.5 kev)/(0.6 kev 2.5 kev ) , while the hard color is defined as ( 4.5 kev 10.0 kev)/(2.5 kev 4.5 kev ) .
we plot unabsorbed fluxes as we want to avoid the detector dependent characteristics introduced by using the measured count rate .
a number of our spectra contain some pile - up ( [ pile_up ] ) , and correcting for this to obtain the true count rate is not straightforward .
the colors are also impacted by the column density in the direction of each object , which varies considerably for the sources in out sample , i.e. , @xmath73 . the expected behavior for a number of different continuum components is plotted in the center panel , e.g. , power - law , accretion disk etc .
the large number of low - hard state observations in our data sample is immediately apparent with the power - law dominated spine dominating the color - color diagrams .
these _ swift _ diagrams are qualitatively similar to those constructed by @xcite for a sample of black holes observed by _ rxte_. . ] _ ii .
hardness luminosity diagram_[hld_diagram ] + the hardness luminosity diagram ( hld ; @xcite ) for the sources in our sample is plotted in fig .
this diagram is limited due to its reliance on count rates to calculate the hardness ratio .
there will additionally be errors due to uncertainty in the column density , black hole mass , and distance estimates ( @xmath74 ) , which make it difficult to separate the different spectral states .
if simultaneous timing information is available , then the hld is very useful to characterise spectral evolution during an outburst .
nonetheless , we include it for comparison with previous work , e.g. , @xcite .
as we utilize a ccd detector , we choose not to calculate a hardness ratio from the spectral counts due to possible pile - up related effects .
hence , we define the hardness ratio as the ratio of the hard to the soft flux , i.e. , @xmath75 . while the overall shape of the pattern traced by the objects is consistent with the ` q'-pattern observed with _
, there is considerable scatter .
in particular , the low - hard state dominated vertical branch see in _ rxte _ hlds is not present here due to the superior ability of _ swift _ to detect low temperature accretion disks .
these differences are to be expected given the caveats outlined above and the differing bandpasses between the 2 instruments .
we can also ask if defining the hardness ratio in terms of energy ranges will mask changes in the source properties .
this is most likely an issue in the energy range 0.6 3 kev , where there will be significant overlap between the hard ( ` po , comptt ` ) & soft ( ` diskbb ` ) spectral components .
this has been investigated by comparing the number of soft / hard sources when we define the hardness ratio as the ratio of the flux in each spectral component ( @xmath76 ) vs the the flux in the spectral bands above and below 2 kev .
we find that for both definitions , the percentage of sources for which the above ratio is less than 1 is approximately 70% , i.e. , both methods find approximately 30% of the observations to be dominated by the soft component .
disk fraction luminosity diagram_[dfld_diagram ] + in fig .
[ dfld ] , we display the disk fraction luminosity diagram ( dfld ) @xcite for the black holes in our _ swift _ sample .
the plot is generated from the best fit model ` diskbb+po ` over the spectral range 0.6 10.0 kev . creating this plot with the comptonization model
does not appreciably modify the resulting pattern traced by the observations in the dfld .
the sample used by @xcite is biased to higher luminosities in comparison to the current sample presented herein , which is dominated by hard state spectra .
@xcite calculated the luminosity in the 1 100 kev and 0.001 100 kev bands for the power - law and disk components respectively .
this requires an extrapolation to energies well below the _ rxte / pca _ bandpass ( i.e. , below 3 kev ) , which will lead to errors as the accretion disk cools .
nonetheless , the dfld presented herein is qualitatively similar to that measured by _
rxte _ for a similar sample of black hole binaries @xcite . assuming a 10 m@xmath77 black hole .
the symbol / color are as defined in fig .
[ bh_all_systems ] .
the lone agn at a luminosity of @xmath78 is the low luminosity agn ngc 4395 .
the black dashed line denotes the best fit relationship between the corona and the disk ( @xmath79 ) for the agn sample , while the magenta line plots the inverse relationship corrected for malmquist bias ( @xmath80 ) , see equation # 15 & # b4 in @xcite . ] in this section , we focus on the 0.6 10 kev x - ray properties of the accretion disk as measured in our sample , before also considering the available uv data .
as outlined earlier ( [ results ] ) , an accretion disk is statistically required at greater than the 5@xmath0 level in @xmath1 56% of our observations .
these disks are detected at varying levels of dominance ( relative to the hard component ) , which can be seen in the disk fraction luminosity diagram ( fig .
[ dfld ] ) . specifically , in the broad 0.6 10 kev bandpass , 95% ( 68% ) of the detected disks contribute greater than 7% ( 17% ) of the total unabsorbed flux .
if we consider only the soft bandpass ( 0.6 2.0 kev ) , then 95% ( 68% ) of the detected accretion disks contribute greater than 16% ( 40% ) of the total unabsorbed flux .
these disk fraction percentages correspond to the spectra when the hard component is modelled using a comptonization model . utilizing a power - law model results in slightly lower disk fractions
as expected , i.e. , 95% ( 68% ) of the disks contribute greater than 14% ( 34% ) of the total unabsorbed flux in the 0.6 2.0 kev band . _ i. accretion disk
corona relationship_[disk_vs_corona ] + in fig .
[ ldisk_lcomptt ] , we plot the relationship between the unabsorbed luminosity in the corona versus that generated by the accretion disk , where the fluxes have been calculated from the ` diskbb+comptt ` model .
we note that when the hard component in our sample is instead modelled with a simple powerlaw component , the distribution of sources in this plot is similar albeit with somewhat larger scatter .
@xcite have investigated this relationship in a sample of 61 hard x - ray selected luminous seyfert galaxies , where it was found that the disk flux is approximately proportional to the coronal flux .
these systems are also plotted in fig .
[ ldisk_lcomptt ] along with the best fitting relationship from @xcite .
the black dashed line denotes the relation between the disk flux as a function of the coronal flux ( @xmath81 ) , while the magenta line denotes the inverse relation corrected for malmquist bias ( @xmath82 ) , see the discussion in @xcite for details .
it is immediately apparent that an extrapolation of these relations to the stellar mass systems reveals an intriguing correspondence . on
the distance introduces a spurious correlation . a partial correlation ( pc )
test can be used to reveal this effect , e.g. , @xcite . for the data presented in fig .
[ ldisk_lcomptt ] , a pc test reveals this correlation to be robust and not simply a spurious distance driven correlation .
this is true for the entire sample and for the agn / stellar systems when considered on their own .
we defer a detailed discussion of this subject to a future publication . ] in contrast to agn where the accretion disk emission is difficult to constrain , for example , requiring the use of mid - ir emission as a proxy in the @xcite study , the accretion disk flux in stellar mass systems peaks in the standard x - ray bandpass ( 0.2 10.0 kev ) allowing simultaneous constraints to be placed on both the disk and coronal emission .
it is apparent from the stellar systems that the use of a single relation is unlikely to provide an accurate description of the observed disk
coronal behavior for the entire sample .
this is not surprising given that the stellar mass sample studied herein contains a large number of transient systems , dominated by accretion rates varying by many orders of magnitude , in contrast to the relatively constant accretion rates exhibited by the agn studied in @xcite .
in addition , the stellar mass systems display apparently discreet spectral states , while seyfert galaxies are more likely to correspond to stellar mass systems in a cyg x-1-like soft state , e.g. , @xcite .
although there is considerable scatter , extrapolation of the seyfert galaxy disk
corona relationship to the stellar mass black holes suggests that at least a subset of the stellar mass bhs exhibit a disk
corona relationship that is consistent with that of their super - massive counterparts ( e.g. , see fig .
[ ldisk_lcomptt ] ) . further study of this relation is clearly warranted , as it would add evidence pointing to the scale invariant nature of the accretion process , e.g. , fundamental plane @xcite .
currently , the fundamental plane relates radio flux to hard x - ray flux and the black hole mass . while the 2 10 kev flux may be a good proxy for the true disk luminosity , it is necessarily an indirect measure .
the relation presented in fig .
[ ldisk_lcomptt ] may provide a means to finally link the true disk and jet emission across the mass scale for accreting black holes .
we defer a detailed consideration of this subject to a future publication .
lccc + [ -2.0ex ] radius [ @xmath83 ] & @xmath84 & @xmath85 & @xmath86 + [ 0.5ex ] + [ -2.0ex ] 6 & 0.57 ( 0.36 ) & 0.43 ( 0.21 ) & 0.74 ( 0.59 ) + [ 0.5ex ] 10 & 0.78 ( 0.59 ) & 0.57 ( 0.36 ) & 0.91 ( 0.81 ) + [ 0.5ex ] 33 & 0.97 ( 0.97 ) & 0.95 ( 0.97 ) & 0.98 ( 0.97 ) + [ 0.5ex ] _ ii .
disk temperature & radius _ + the primary observational quantities providing us with information on the configuration of the accretion disk are the inner disk color temperature ( @xmath87 ) and the disk normalization , which is proportional to the distance and the disk color radius , i.e. , @xmath88 .
the best fit values for the parameters are dependent on how the spectra are modelled .
in particular , the choice of model for the hard x - ray component can modify the measured disk flux , e.g. , using a power - law ( ` po ` ) or thermal comptonization ( ` comptt `
@xcite ) because the comptonization cuts off at low energies consistent with the input seed photon temperature .
we compare the 2 models above with the goal of determining , which model is a better description of the spectra .
this will also allow us to probe the effect of incorrect modelling on the accretion paradigm ( see [ spec_issues ] ) . in the appendix ( a[appendix_cyg ] ) , we discuss the evolution of the persistent bh cyg x-1 & the transient system gx 339 - 4 in detail , and compare the behavior of the key observational quantities as a function of time for both the ` diskbb+po ` and ` diskbb+comptt ` models . in fig .
[ disk_temp_histogram ] ( left panel ) , we plot the measured accretion disk temperature histogram for both of the ` diskbb+po ` and ` diskbb+comptt ` continuum models .
the primary difference between these models is that fewer observations require the addition of an accretion disk component at greater than the 5@xmath0 significance level , i.e. , 282 and 254 observations require a disk at the 5@xmath0 level for the ` diskbb+po ` and ` diskbb+comptt ` respectively .
this is due to the differing nature of the hard components at low energies , which also modifies the disks parameters , i.e. , the accretion disks detected when the hard component is modelled with ` comptt ` are slightly cooler and have a slightly larger inner radius as is evident in the figure .
the inner radius of the accretion disk may be determined from the normalization of the ` diskbb ` model , i.e. , @xmath89/d [ 10 kpc])^2 cos\theta$ ] .
the disk radius thus measured is subject to a number of additional uncertainties , which bias the measured radius .
the most important are ( i ) to account for spectral hardening , a multiplicative correction factor of 1.7 is typically assumed ( @xcite , though see below ) , and ( ii ) to account for the fact that the disk temperature does not peak at the inner radius @xcite , which combined give us the corrected inner radius @xmath90 \simeq \frac{1.18\sqrt{norm}}{\sqrt{cos\theta}}d_{10 kpc}\ ] ] in fig .
[ disk_temp_histogram ] ( right panel ) , we plot the inner disk radius calculated for each of the accretion disks in the sample , assuming the distances and black hole masses listed in table [ bhc_table ] ( see appendix a[appendix_param ] for a discussion of the system parameters and their related uncertainties ) .
the distribution peaks at small inner radii , @xmath91 , with a tail to larger radii ] .
of particular note , the largest radii measured in our sample are @xmath42 40 @xmath83 . in table
[ disk_rin_table ] , we list the cumulative number of accretion disk with radii below 6 , 10 , 33 @xmath83 respectively and these values taking into account uncertainties in the mass and distance to each system respectively .
the maximum inner radius is determined to be robustly @xmath42 40 @xmath83 . in fig .
[ disk_temp_histogram ] , those spectra for which the source luminosity is less than 1% eddington are plotted in green .
it is clear that at low luminosities , the disk temperature is low @xmath1 0.2 kev , but the inner radius of the accretion disk also remains low , exhibiting a clear peak at a radius of @xmath1 10 r@xmath92 .
examination of the temperature distribution in fig .
[ disk_temp_histogram ] clearly reveals 2 distinct regimes , i.e. , those at @xmath1 0.2 kev and those at a temperature of @xmath32 0.5 kev .
these are consistent with the primary active accretion states , the hard state and the soft state respectively .
we note that there is a deficit of systems in the temperature range between these two .
the so called intermediate states are expected to reside here , i.e. , those states during which the system transitions from the hard state to the soft state and vice versa @xcite .
in contrast , the inner disk radius distribution does not display obvious evidence for 2 or more distinct spectral states , instead being consistent with a single continuous distribution of low inner radii with a tail to higher radii , @xmath93 . in the `
diskbb+po ` fits there are a number of high temperature disks that are not required when the hard component is modelled using a comptonization component .
these high temperature disks are seen from grs 1915 + 105 , gx 339 - 4 and xte j1752 - 223 . in all of these cases ,
the spectrum is dominated by the hard component , i.e. , these spectra appear to be consistent with the systems being in the very high state at the time of the observation @xcite .
a number of high temperature disks also persist at a luminosity less than 1% eddington ( see fig . [ disk_temp_histogram ] ) .
these spectra belong to cyg x-1 , igr j17091 - 3624 and igr j17098 - 3628 .
the igr j17098 - 3628 spectra are only marginally below the nominal 1% selection cut , i.e. , @xmath94 . in contrast , the spectra of the other two systems are closer to 0.001 l@xmath95 .
the igr j17091 - 3624 observation is consistent with a soft state disk dominated spectrum , whereas the cyg x-1 spectrum is dominated by power - law emission and is more consistent with a very - high state like spectrum .
the system parameters for both of the igr sources are poorly constrained , unlike cyg x-1 , where both the mass and distance are accurately known @xcite .
however , we note that the simple spectral modelling used herein may be a poor representation of the actual spectral form for these systems , e.g. , @xcite . in fig .
[ lx_t4 ] , the relationship between the disk luminosity and temperature is plotted for both of the continuum models . if the emitting surface area of the accretion disk is constant ( i.e. , constant inner radius ) , then we expect the systems to follow a relation of the form @xmath96 . at higher temperatures and luminosities , a relation of this form ( as indicated by the dashed lines ) is observed ; however , at the lowest disk temperatures ( indicated by the shaded region ) , this no longer holds and instead the disk temperature is observed to remain approximately constant while the luminosity changes by orders of magnitude . modelling the accretion disk emission with the `
diskpn ` model @xcite , returns similar results .
the deviation of the spectra at low temperature from the expected relation implies a change in the conditions of the accretion flow , for example , a changing inner accretion disk radius in the low - hard state @xcite .
the behavior of the hard x - ray components and their relationship to the observed behavior of the accretion disk was also examined , i.e. , the power - law characterized by spectral index , @xmath43 , and the comptonizing component characterized by the optical depth , @xmath53 , see fig .
[ gamma_tau_plot ] . focusing on the results of the ` diskbb+po ` fit , we find from left to right ( a ) disk temperature : a clear dichotomy exists where we see a population of low temperature disks consistent with the hard state , and then a second population of hotter disks
. unfortunately due to our relatively narrow x - ray bandpass , the power - law is not well constrained in those spectra where the disk is dominant , i.e. , @xmath97 .
( b ) the spectral index is observed to increase ( i.e. , soften ) as the radius of the disk increases until a spectral index of approximately 2 is reached at which point the radius tends to be constant .
the decrease of the accretion disk radius as the spectrum hardens points towards a changing spectral hardening factor , see
[ spec_hard]iii .
( c ) at the lowest luminosities @xmath98 , the spectral index is observed to remain constant , @xmath99 . at luminosities above this the spectral index increases , though the rate at which it increases appears to vary from system to system .
additionally , the spectral index may soften at the lowest luminosities ( e.g. , @xcite ) ; however , the number of spectra at these luminosities is too few to make any definitive statement .
as noted previously , the ` diskbb+comptt ` model exhibits similar behavior to the ` diskbb+po ` model with the caveat that the optical depth , @xmath53 , trends in the opposite sense to the power - law spectral index .
this effect is strikingly apparent in fig .
[ gamma_tau_plot ] . in the lower panels
, we see that ( d ) the optical depth varies significantly with high temperature disks mainly requiring optically thin coronae whereas those low temperature disks in the hard state require higher optical depths . in the hard state
it is noticeable that while the disk temperature remains approximately constant ( @xmath1 0.2 kev ) , the optical depth varies significantly suggestive of coronally driven spectral evolution at this disk temperature / luminosity .
( e ) the inner radius of the accretion disk increases as the optical depth decreases , and ( f ) the optical depth of the corona decreases as the 0.6 10 kev luminosity increases , i.e. , as the spectrum becomes dominated by the soft thermal emission from the accretion disk . similar behavior is observed when the hard component is modeled via the ` simpl ` model , where the scattering fraction can be understood as a proxy for the optical depth of the corona .
spectral hardening_[spec_hard ] + in the previous section , the accretion disk emission is modelled as a standard multi - color blackbody accretion disk using the ` diskbb ` model . however , the emission from the accretion disk is not a true blackbody and instead will be modified as it scatters through the disk atmosphere .
this emission may nonetheless be approximated by a blackbody .
the temperature returned by the ` diskbb ` model , @xmath100 , actually represents the temperature of this modified blackbody which approximates the emergent disk spectrum .
this temperature is related to the true effective disk temperature through the relation @xmath101 where @xmath102 is the spectral hardening factor @xcite . in the standard scenario @xmath102
is assumed to be approximately 1.7 based on the work of @xcite .
@xcite revisited this problem and found @xmath103 , with the spectral hardening factor increasing as the spectrum became less disk dominated .
further theoretical study of the behavior of @xmath102 in the disk dominated state predicts @xmath104 @xcite and @xmath105 for disk temperatures greater than 1 kev @xcite . with the exception of the study by @xcite , each of the other investigations focused on the soft state behavior of @xmath102 . in fig .
[ rin_histogram ] ( left ) , we plot the relationship between the hardness ratio and the inner disk radius , as calculated from the best - fit ` diskbb+po ` model . for the softer spectra ( hr @xmath32 1 )
, the radius is observed to be approximately constant for each source , whereas for the harder spectra ( hr @xmath42 1 ) , the radius is observed to decrease .
this is consistent with the behavior predicted by the @xcite spectral hardening model for coronally dominated spectra . in fig .
[ rin_histogram ] ( right ) , we plot the histogram showing the inner disk radius measured for each spectrum , normalised to the smallest radius measured from that system .
the relative radius change is observed to peak at @xmath42 20x and 10x for the ` diskbb+po ` and ` diskbb+comptt ` models respectively .
the apparent changes in the disk radius implied by the previous figures suggests that spectral hardening may play an important role . in order to investigate this ,
the previous continuum fits were repeated but the accretion disk model ` diskbb ` is replaced with ` diskpn ` @xcite . in this model ,
the inner radius is a free parameter ( which we fix at @xmath106 ) and the normalization explicitly includes the spectral hardening factor , i.e. , @xmath107 the results of these spectral fits , where a constant inner radius is explicitly assumed , are found to be consistent those using the variable radius ` diskbb ` model , see table [ chi2_table ] . restricting ourselves to the predicted values for the spectral hardening factor
mentioned earlier ( @xmath108 ) , we see that a change in the disk normalization by a factor of 10 20 is achievable through variations in the spectral hardening alone . assuming that the radius variations observed in fig .
[ rin_histogram ] are solely due to variations in the spectral hardening factor , the ` diskpn ` model fits may be used to calculate @xmath102 . to do this , we assume that the minimum measured accretion disk normalization corresponds to the canonical spectral hardening factor , i.e. , @xmath102 = 1.7 @xcite .
this minimum normalization corresponds to an accretion disk dominated observation , and as the disk temperature / luminosity decreases we expect @xmath102 to increase as the continuum becomes dominated by emission from the hard x - ray component @xcite . as the normalization @xmath109
, we scale the other observations relative to this , hence allowing us to estimate the spectral hardening factor for the observations in our sample . in fig [ fcol_histogram ]
, we plot the histogram of calculated spectral hardening factors .
the vast majority of observations are consistent with the theoretically expected bounds on @xmath102 ( @xmath110 ) , with a tail of observations requiring a spectral hardening factor outside this range , i.e. , 76% ( 88% ) and 60% ( 73% ) of the observations require @xmath111 for the ` comptt ` and ` po ` models respectively . in cases where the default value at @xmath112 of @xmath113 is incorrect , for example , in systems such as swift j1753.5 - 0127 that remained in the hard state at all times
, a larger value of the spectral hardening factor may be warranted .
this will not affect the observed pattern , which remains approximately the same ; however , the amplitude will be reduced .
therefore our choice of @xmath113 at @xmath112 is conservative , and the derived maximum hardening factors should be considered as upper limits in this case . in fig .
[ fcol_plot ] , we illustrate the behavior of the spectral hardening factor as a function of the disk temperature ( left ) and 0.6 10 kev luminosity ( middle ) . finally , in the right hand panel we plot the corrected disk luminosity versus disk effective temperature , i.e. , @xmath114 .
an important question when modelling the broadband emission from an xrb is to determine whether the emission is direct viscous dissipation from the accretion disk , reprocessed hard x - ray emission , or if it is a separate component entirely , i.e. , a jet .
previous work has demonstrated that irradiation of the disk is important during the outburst of a black hole binary , e.g. , @xcite .
@xcite , using a large sample of black holes binaries , showed that the opt / nir vs hard x - ray relationship was consistent with either an irradiated disk or jet origin .
+ _ i. x - ray uv / optical flux relations _ + the uvot obtained simultaneous observations in the optical / uv spectral range ( 5000 2000 ) for over half of our sample in at least a single filter ( v , b , u , uvw1 , uvm2 , uvw2 ) .
the lightcurves are morphologically similar to those observed at x - ray wavelengths . unfortunately due to the typical
_ swift _ daily observing cadence , and exposure time of @xmath1 100s
1000s , we are unable to search for correlated delays of the optical / uv emission with respect to that observed at x - ray energies .
such features can be used to probe the response of the outer to the inner flow , e.g. , @xcite . broadly characterizing the opt / uv photometry
, we measure the ratio of the flux at x - ray ( 0.6 10 kev ) to optical / uv - wavelengths ( @xmath115 ) to be @xmath116 in the v - band .
the uv flux ratio is similar , though we caution that the red leak in the uvw1 filter @xcite may prohibit us from constraining the true uv flux . .
] the relationship between the flux observed in the optical / uv versus that in the x - ray can be used to constrain the physical processes contributing to the observed spectrum , for example , a correlation of the form @xmath117 is expected from the reprocessing of x - rays by the accretion disk @xcite . in fig .
[ optuv_vs_xray_flux ] , we plot the relationship between the total unabsorbed x - ray luminosity in the 0.6 10 kev band and that detected in the optical / uv bandpasses .
the differing number of points in each plot reflects the differing observing strategies undertaken for each system , i.e. , one or multiple uvot filters .
the optical emission is observed to be consistent with viscous emission from the accretion disk ( @xmath118 , e.g. , see @xcite ) , while in the uv we find evidence for both viscous emission and irradiated emission from the accretion disk ( @xmath119 ) . in the case of gx 339 - 4 ( lightblue )
, we find an approximately flat relation between the uv and the x - ray fluxes , particularly apparent for luminosities less than @xmath120 .
this hints at the presence of an additional source of flux during these observations , which we note are primarily from when the x - ray spectrum is power - law dominated , i.e. , in the low - hard state . in the optical ,
two distinct tracks are observed likely due the nature of the observed emission ( purely viscous or contaminated ) and to the varying characteristics of each system ( e.g. , disk size @xmath121 ) .
neither the optical or the uv flux is consistent with a significant contribution from the jet , which would be expected to produce a relation @xmath122 @xcite . previous observations of a number of black hole binaries have revealed convincing evidence for the presence of a significant contribution from the jet at nir wavelengths , with a smaller contribution also in the optical , e.g. , @xcite .
however the analysis presented herein is necessarily limited by the bandpass of our observations , i.e. , @xmath123 , which is at a wavelength shorter than that at which the jet component has been observed to dominate . _
broadband spectroscopy_[bband_spec ] + in order to investigate the relations observed above in more detail , the uvot data were converted into xspec readable files and both the optical / uv and the x - ray data were fit simultaneously . to investigate the origin of the uv emission ,
three models were considered ( i ) a standard steady state disk plus a power - law component , i.e. , ` diskbb+po ` , ( ii ) variable temperature profile accretion disk plus a power - law , i.e. , ` diskpbb+po ` , and ( iii ) variable temperature profile accretion disk plus comptonization , i.e. , ` diskpbb+comptt ` . in all cases , we utilized ` phabs ` for the x - ray column & ` redden ` for the optical / uv column assuming the standard galactic extinction curve @xcite and the dust to gas ratio from @xcite .
the column density was fixed at the standard x - ray value ( table [ bhc_table ] , see appendix a[appendix_param ] for a discussion of the uncertainties in the column density ) . in those observation where we had multiple opt / uv datapoints , it is possible to constrain the column density .
the typical best fit values were found to be consistent with the assumed x - ray column density within the errors in all cases .
the results of the x - ray plus uv spectral fits may be summarised as follows : model ( i ) ` diskbb+po ` : 267 broadband spectra in total , and 196 of these are found to require an accretion disk at greater than the 5@xmath0 confidence level .
the number of broadband spectra which do not require a disk component are 71 , 60 and 46 at the 5@xmath0 , 4@xmath0 , and 3@xmath0 levels respectively . model ( ii ) ` diskpbb+po ` : 264 broadband spectra in total . model ( iii ) ` diskpbb+comptt ` : 257 broadband spectra in total
. a similar number of models as in ( i ) above were found not to require an accretion disk for model ( ii ) .
these spectra are discussed in detail later .
a disk component is required in all of the ` diskpbb+comptt ` models as the corona can not provide significant flux below the temperature of the input seed photons . when considering the quality of the model fits to the sample as a whole , they are found to be comparable , i.e. , as measured by the respective @xmath124 distributions .
however , the physical differences in the model components result in differing contributions from the accretion disk relative to the power - law , for example , as a function of wavelength . in the appendix a[appendix_bband ] , we examine the uv spectral models via a detailed look at a hard & soft state observation of the transient system gx 339 - 4 .
the basic properties of the accretion disks ( @xmath125 ) in these fits are consistent with those measured in fits made to the x - ray data alone .
however , it is important to note that it is impossible to constrain the temperature profile using the x - ray data alone , as the data will invariably be consistent with a steady state disk within the statistical errors , e.g. , @xcite . for the variable temperature profile model ( @xmath126 ) the distribution of the index differs between the comptonization and power - law models reflecting the contribution by the hard component , in the power - law model case , to the optical / uv emission . the average temperature profile ( @xmath127 ) we measure is @xmath128 for the ` diskpbb+comptt ` model and @xmath129 for the ` diskpbb+po ` model ( see fig .
[ histogram_tprofile ] ) , though in both cases we are biased by the large number of hard state observations in our sample as may be seen in fig .
[ disk_temp_histogram ] .
we emphasise , that irradiated disks are seen at all luminosities , with hard state observations typically requiring a more irradiated disk in comparison to soft state observations .
there is no evidence for a dominant jet contribution at uv wavelengths .
the power - law component is not dominant for any of the models in the hard state , and while it can dominate in the case of the ` diskbb+po ` model in the soft state , we do not consider this model to be valid , i.e. , the power - law index would have to be much harder than is typically observed in a soft state ( e.g. , see fig .
[ gx339_bband_hss ] & [ gx339_bband_lhs ] in the appendix ) .
this is consistent with the known absence of detectable jet emission in the soft state @xcite . in all other cases the emission from the irradiated accretion disk dominates at uv wavelengths at all times . though it is important to emphasise the fundamental difference implied by the two hard component choices .
when the hard x - ray emission is most successfully modelled by a comptonization component ( i.e. , soft state ) , then the uv emission is dominated by the accretion disc .
in contrast , during harder states , where both the comptonization and power - law components provide fits of comparable statistical quality , two different physical scenarios are suggested . for the comptonization model ,
again all of the uv emission originates in the irradiated accretion disk , whereas for the power - law model the uv is dominated by emission from the irradiated accretion disk _ but _ with a contribution from the power - law component that may be as large as 10% of the total continuum flux in the uv .
such a model may be consistent with timing observations that have found evidence for non - thermal flux at uv wavelengths , e.g. , @xcite .
ideally we would have had access to simultaneous nir observations that would allow us to break this degeneracy and accurately constrain the possible power - law ( jet ) contribution at optical / uv / x - ray wavelengths , e.g. , @xcite .
as stated earlier , there are a number of the broadband spectra that are consistent with a power - law alone , i.e , the addition of a disk component does not significantly improve the fit . in the case of the ` diskpbb+po ` model 50/264 of the spectra are consistent with a power - law alone when we only mark an accretion disk as detected if it improves the fit statistic by greater than the 5@xmath0 level as measured by an ` ftest ` .
if instead we lower the criteria to a 3@xmath0 improvement only 28 spectra are consistent with a power - law alone .
the number of spectra consistent with a power - law alone is larger for the ` diskbb+po ` model ; however , the excess relative to the ` diskpbb ` model is due to the inability of a standard steady state accretion disk to adequately reproduce the observed data . to examine these power - law spectra in more detail ,
we restrict ourselves to spectra with greater than 1000 counts and a best fit @xmath130 , leaving us with 24 spectra distributed across 6 systems ( gro j1655 - 40 , gx 339 - 4 , swift j1753.5 - 1027 , swift j1842.5 - 1124 , xte j1752 - 223 , xte j1818 - 245 ) .
these systems broadly represent the contents of our sample , with no obvious biases towards system parameters , e.g. , @xmath131 .
though our knowledge of these parameters is limited in most of these systems , their observed behavior to date provides no evidence of any fundamental difference in comparison to the well known systems , i.e. , gro j1655 - 40 . the average spectral index for this sample is found to be @xmath132 , while the average luminosity is @xmath133 , both consistent with expectations for emission from a black hole in the low - hard state .
the absence of a detectable accretion disk in these spectra may point toward a truncated accretion disk or conversely a disk that has cooled below our detection limit .
an alternative possibility , suggests that these spectra are dominated by emission form the broadband jet , e.g. , @xcite .
we have presented a comprehensive systematic analysis of all of the available _ swift _ data on accreting stellar mass black hole binaries as of summer 2010 .
this is the most detailed study of accretion onto stellar mass black holes at luminosities of @xmath25 . to summarize : * we analysed 476 ccd spectra of 27 systems , in the 0.6 10 kev band
. approximately half of these had simultaneous coverage in the optical / uv band . *
an accretion disk is required to successfully model the data in over half of the x - ray observations .
the excellent sensitivity at low energies allows us to follow the evolution of the accretion disk to temperatures @xmath1 0.2 kev .
this is the largest sample of such cool accretion disks collected to date . *
the observed accretion disk components are consistent with maintaining inner radii @xmath134 at all times .
more severe disk truncation must take place at luminosities @xmath135 . *
the evolution of the inner disk radius as a function of spectral hardness , i.e. , a decreasing radius as the source hardens , suggests a contribution from spectral hardening to the observed disk radius variations . *
the simultaneous x - ray plus optical / uv data allow us to constrain the temperature profile of the accretion disk for a large sample of observations .
irradiation of the accretion disk is important at all times , likely even at x - ray wavelengths , i.e. , @xmath136 .
* in the soft state , the broadband spectra are consistent with a spectrum dominated by emission from the accretion disk and an additional hard component consistent with compton scattering of seed disk photons . *
while an irradiated disk is preferred in the hard state , the data are unable to statistically distinguish between power - law and comptonization models for the hard flux .
* the relationship between the flux emitted by the accretion disk and that emitted by the corona is in broad agreement with the observed behavior in seyfert galaxies , suggesting a scale invariant coupling between the accretion disk and the corona .
we discuss these results and their implications in detail below .
the largest radii measured in our sample are @xmath1 40 @xmath83 , which rules out any large scale truncation of the accretion disk , for the luminosities probed in this survey .
the observed distribution of disk temperature and radius for those disks detected at a luminosity @xmath137 serves to emphasise this point ( fig .
[ disk_temp_histogram ] ) .
there is currently no conclusive observational evidence that the accretion disk is truncated in the hard state , at least at the luminosities probed in this work ( @xmath138 ) .
this has long been known to be the case at the higher luminosities observed in the soft state , where the inner edge of the accretion disk is expected to lie at the isco , e.g. , @xcite .
it is this constancy of radius that allows one to constrain the black hole spin via the accretion disk continuum emission , e.g. , @xcite . in recent years evidence that the accretion disk is not truncated in the hard state
has also been accumulating , e.g. , @xcite .
however , we acknowledge that the disk inner radius distribution measured herein can be interpreted as evidence for small scale truncation of the inner disk at lower luminosities . clearly , further observational / theoretical tests are required to break this degeneracy .
if the accretion disk does not truncate ( e.g. , @xcite ) , then we must re - consider the driving mechanism behind state transitions , in particular the hard and soft states , and how this is related to the production of a relativistic jet .
the most promising means for powering these relativistic jets is the ` blandford - znajek ' mechanism , where the luminosity of the jet is @xmath139 @xcite .
if the inner disk is present at the isco at all times as the above data implies , then this suggests that the accretion disk itself is not required to launch the jet .
indeed simulations have demonstrated the formation of jets in the absence of a disk .
for example , @xcite analyze the formation of a jet launched from a black hole accreting from the stellar wind of a companion star .
the driving mechanism behind the state transitions is less clear , though it is clear that a more detailed understanding of the evolution of , and interaction between , the hard component and the accretion disk is key .
the evidence presented herein points towards a non - varying disk geometry during the hard to soft state transition , as such , attention must turn to the corona and its coupling to the accretion disk as a function of mass accretion rate . .
an irradiated temperature profile is required in a large majority of the observations . ] as discussed in [ spec_hard ] , the accretion disk temperature measured from the observed x - ray spectrum is not the true effective temperature of the disk , but is instead the temperature of the disk modified as the emitted disk photons pass through the disk atmosphere .
this effect must be accounted for if we are to correctly interpret the data .
this is done via the spectral hardening factor , @xmath140 , i.e. , @xmath141 .
theory predicts the correction factor to be in the range @xmath142 @xcite . in fig .
[ rin_histogram ] , it is clear that the behavior of the apparent disk radius is consistent with spectral hardening : the apparent radius decreases as the source hardens ( e.g. , @xcite ) .
this runs counter to the prediction of adaf models , which suggest that the disk should gradually recede in the hard state .
indeed , fits with the `` diskpn ` ' model explicitly show that the spectra are consistent with a constant radius while the correction factor changes in the manner predicted by @xcite .
this finding does not rule out adaf models in a broad sense ; disk truncation at much lower fractions of the eddington limit is expected and may be partially confirmed by recent data , e.g. , @xcite on gx 339 - 4 .
for the first time , we present a large sample of empirically determined color correction factors for a large sample of systems , see fig .
[ fcol_histogram ] & [ fcol_plot ] .
the color correction factor is found to be consistent with the canonical value ( @xmath143 ) for the majority of our observations , with a tail extending to larger values , i.e. , 76% ( 88% ) and 60% ( 73% ) of the observations require @xmath144 for the ` comptt ` and ` po ` models respectively .
when the hard x - ray component is modelled using the ` comptt ` model , the maximum value of the color correction factor is @xmath145 5 .
figure [ fcol_plot]c plots the corrected disk temperature as a function of luminosity . at larger temperatures ( larger disk fractions )
the standard @xmath146 is recovered ; however , as we move toward lower disk temperatures ( dominant hard component ) a deviation is observed .
we have demonstrated that a significant fraction of this deviation can be accounted for with plausible variations in the spectral hardening factor alone . though the high @xmath140 tail present in fig .
[ fcol_histogram ] does leave room for an additional process , i.e. , is the disk finally beginning to recede as we approach these luminosities ( @xmath147 , also see @xcite ) or are the limitations of our spectral models , in the presence of a dominant hard component , being exposed , see below .
spectral hardening appears to contribute to the observed disk properties at _ both _ high and low disk temperatures / luminosities . at the highest luminosities , we are likely beginning to see the re - emergence of the hard component as the system moves to the very high state where in contrast to the lhs the disk is now hot , i.e. , @xmath148 @xcite
this is clearly seen in the study of accreting black holes by @xcite , which due to the low energy cut - off of _ rxte _ ( @xmath149 )
was necessarily limited to studying the brighter accretion phases . at low luminosities ,
a dominant power - law component modifies the emitted disk spectrum . in both cases ,
the color correction factor naturally increases as the non - thermal emission begins to dominate ( e.g. , @xcite ) . at the lowest measured disk temperatures
, we begin to approach the lower energy limit of the xrt used herein , i.e. , 0.6 kev ( remember that the disk spectrum peaks at @xmath150 ) . in the limit of low s / n some of the low temperature disks we detect could be systematics caused by an excess of flux in some bins at the low energy end of our bandpass , which may mimic a low temperature disk component .
our decision to categorize a disk as detected _ only _ if it improves the fit by greater than 5@xmath0 as measured by an ` ftest ` , should eliminate such errors in this sample of low temperature disks .
this has been verified by re - fitting a sample of the data assuming a lower energy bound of 0.4 , 0.5 kev respectively , where we find disk components entirely consistent with those measured assuming a lower bound of 0.6 kev .
similarly , we have no reason to believe there is an issue with the xrt calibration ; nonetheless , this is an area worthy of further detailed consideration , which we defer to a future study .
previous theoretical efforts to model the behavior of the spectral hardening factor have primarily been focused on brighter more disk dominated states @xcite .
the study by @xcite is the only one to consider a situation in which the disk does not dominate the x - ray emission .
these simulations pointed towards an increase in @xmath140 as the disk fraction decreased , in agreement with the observed behavior of the accretion disks studied herein ( see fig .
[ rin_histogram ] ) . however significant issues remained , including the use of a number of simplifying assumptions regarding the structure of the accretion disk , and the failure to account for the effect of the returning coronal radiation on the disk itself @xcite .
subsequent work clearly illustrated how the assumption of a constant density for the vertical structure of the accretion disk is inappropriate for a cold gas pressure dominated disk in the presence of a powerful corona @xcite . while the study of @xcite included a detailed treatment of the vertical structure of the accretion disk in addition to opacity related effects , it focused on soft - state accretion disks , i.e. , kt @xmath32 0.5 kev , and as such , ignored the effect of the returning radiation from the corona on the spectrum from the accretion disk , which should be minimal in this regime .
this reflected radiation can significantly modify the observed spectral shape @xcite and dominate the spectrum in x - ray bandpass , e.g. , see @xcite for a detailed study of cyg x-1 .
finally , we note that even in the limit of a disk dominated spectrum such as that observed from lmc x-3 , significant model uncertainties remain @xcite . given the importance of this issue , more theoretical investigations , particularly in states where the hard component is dominant over the emission from the disk , are clearly warranted .
the empirical distribution of the color correction factor provided herein ( see fig .
[ fcol_histogram ] & [ fcol_plot ] ) will provide a valuable observational constraint for these studies . in this work ,
we have utilized the combination of simultaneous x - ray and uv / optical data provided by _
swift _ to investigate the irradiation phenomenon in black holes binaries . in soft state observation
, we find evidence for a mildly irradiated disk , i.e. , @xmath151 , whereas hard - state observations are consistent with an increasingly irradiated disk , i.e. , @xmath152 .
the change in the temperature profile relative to the soft state is consistent with an increase in the irradiating hard x - ray flux , as expected in the hard state . in fig .
[ histogram_tprofile ] , we plot the distribution of accretion disk temperature profiles obtained from the ` diskpbb ` broadband spectral fits . there is a strong preference for a non - standard temperature profile irrespective of how the hard spectral component is modelled , i.e. , ` po ` or ` comptt ` . in the appendix ( a[appendix_bband ] )
, we discuss specific examples of broadband fits to an observation of gx 339 - 4 in both the hard and soft states . in the hard state
both the ` comptt`/corona and ` po`/jet provide acceptable fits to the data .
fundamentally , we are unable to differentiate between these 2 physical scenarios with the available data , though appealing to previous timing observations ( e.g. , @xcite ) leads us to favor the power - law model .
in addition , correlated multi - wavelength studies ( radio x - ray ) have revealed clear evidence for a jet in the hard state @xcite ; however , these studies suggest that it is unlikely to contribute significantly at uv wavelengths , consistent with the analysis herein .
the soft state spectra appear to rule out a power - law origin for the observed hard x - ray component as this results in a much harder spectral index than is normally measured in the soft state .
this is in agreement with the well known behavior of relativistic jets , whereby they are observed to be present in the hard state , but appear to be absent in the soft state @xcite . in light of these results , it may be necessary to re - consider the effect of the assumed temperature profile on the measured black hole spin , as even though the x - ray spectrum may be consistent with a standard steady state disk , when one considers the broadband spectrum deviations from this profile become apparent .
this may materially impact the derived disk parameters .
for example , a single observation of the black hole binary 4u 1957 + 11 is included in this study .
when the broadband data is fit with the variable temperature profile disk model , the best fit temperature profile is consistent with an irradiated accretion disk with a lower disk temperature in comparison to the steady state case .
recently , @xcite have measured the spin of the black hole in this system via the continuum method using _ suzaku _ spectra , where the spin was consistent with a near maximal value .
clearly , failure to correctly account for disk irradiation could bias the measured spin .
we defer detailed consideration of this to a future study .
we have presented a comprehensive analysis of all of the data on stellar mass black binaries in the _ swift _ public archive as of june 2010 .
this is the largest collection of ccd spectra of accreting stellar mass black holes published to date , amounting to 476 spectra of 27 binary systems .
the excellent sensitivity and broad soft x - ray bandpass of _ swift _ allows us to study the evolution of the accretion disk down to temperatures as low as @xmath1 0.2 kev for the first time .
approximately half of the spectra are found to require the addition of an accretion disk component at greater than the 5@xmath0 .
there is no compelling evidence in the data presented herein , which samples the accretion disk evolution down to luminosities of @xmath153 , for a large scale change in the inner disk radius during the transition from the hard to the soft state , i.e. , @xmath154 at all times .
consideration of the evolution of the inner radius versus spectral hardness points towards spectral hardening being a factor in the observed disk radius evolution .
clearly , future observations probing luminosities below @xmath155 are necessary to provide further constraints on the evolution of the accretion flow geometry .
irradiation is found to be present in the majority of observed spectra where simultaneous optical / uv and x - ray data are available .
this irradiated disk component is found to dominate the optical / uv emission in both hard and soft states .
the soft state spectrum is found to be consistent with a combination of an irradiated accretion disk plus corona , whereas for the hard state a significant contribution from a power - law component ( i.e. , a jet ) can not be ruled out at optical / uv energies .
we find the relationship between the flux emitted by the accretion disk and that emitted by the corona observed herein is in broad agreement with the observed behavior in seyfert galaxies , suggesting a scale invariant coupling between the accretion disk and the corona .
though , further study is warranted . the data and analysis presented herein clearly demonstrates the value of the broadband observational capabilities of the _ swift _ observatory , in the study of accretion physics .
an analysis of the neutron star xrbs observed by _ swift _ , and a comparison with the black hole binary systems presented herein will be presented in a companion paper ( reynolds et al . , 2012 in prep .
finally , we would like to re - emphasise the importance of the broadband coverage provided by _ swift _ ( @xmath156 ) in constraining the properties of the observed accretion disk emission , which is key if we are to understand the relationship between the accretion disk inflow and the jet outflow .
it is a pleasure to acknowledge the excellent contribution made to the study of accretion physics by the _ swift _ observatory , we extend our thanks to all those on the _ swift _ team who made this possible .
we thank the referee andrea merloni for an insightful report , which improved the content and clarity of this paper .
we also thank rubens reis , dipankar maitra and david russell for useful discussions .
we acknowledge the use of public data from the swift data archive .
this research made extensive use of data obtained from the high energy astrophysics science archive research center ( heasarc ) , provided by nasa s goddard space flight center , of the _ simbad _ database , operated at cds , strasbourg , france and nasa s astrophysics data system .
miller j.m .
, da a. , bautz m.w . , bhattacharyya s. , burrows d.n . , cackett e.m .
, fabian a.c . , freyberg m.j . , haberl f. , kennea j. , nowak m.a . , reis r.c . ,
strohmayer t.e . , tsujimoto m. , 2010 , apj , 724 , 1441 zurita c. , sanchez - fernandez c. , casares j. , charles p.a . , abbott t.m .
, hakala p. , rodrguez - gil p. , bernabei s. , piccioni a. , guarnieri a. , bartolini c. , masetti n. , shahbaz t. , castro - tirado a. , henden a. , 2002 , mnras , 334 , 999 in fig . [ disk_properties ] , we plot the primary disk properties as a function of time for the best fit ` diskbb+po ` models in the case of the hmxb cyg x-1 and the lmxb gx 339 - 4 .
these systems are chosen as examples of the accretion flow behaviors observed from a persistent and transient black hole binary system respectively .
cyg x-1 is a persistent x - ray source where the @xmath1 14.8 @xmath157 black hole accretes matter from the o - type secondary star @xcite .
it is normally detected at a flux of @xmath1 0.2 @xmath158 by swift bat in the 15 50 kev band , while it is in the spectrally hard state . in fig .
[ disk_properties ] , we see that cyg x-1 was observed sparingly prior to the spectral softening which occurred in late 2010 .
the disk normalization ( where the color radius , @xmath159 ) and spectral index are observed to increase as the spectrum softens while the temperature of the disk component remains fairly constant .
for the comptonization model the optical depth is observed to behave in the opposite sense to the power - law spectral index .
we remind the reader that we have fixed the electron temperature to be 50 kev due to the lack of data above 10 kev . even though cyg x-1 is a hmxb with a massive o - type donor , in contrast to the the low mass main sequence stars typically detected in the transient black hole binaries , the basic x - ray continuum behavior described above
is also observed from the persistent lmxb swift j1753.5 - 0127 .
this suggests that the behavior we observe is driven by the inner accretion flow and not by the nature of the mass transfer , i.e. , wind vs roche lobe overflow .
in contrast to the persistent systems , the transients as represented by gx 339 - 4 display discrete changes in measured disk temperature , consistent with the well known transitions between the low temperature hard state and the high temperature soft state , after brightening from the quiescent state @xcite . since the launch of the _ swift _ mission there have been 4 outbursts detected from gx 339 - 4 ( fig . [ disk_properties ] right ) .
two of these outbursts are consistent with the system remaining in the hard - state for the duration of the outburst . in fact , a statistically significant ( 5@xmath0 ) accretion disk component was not detected during the second outburst .
such hard - state outbursts have been detected from numerous black hole systems , e.g. , gro j0422 + 32 , xte j1118 + 480 , gs 1354 - 64 among others @xcite .
the other two outbursts are observed to show a clear transition to the accretion disk dominated soft - state .
here , we see that the normalization ( color radius ) of the accretion disk increases as the disk temperature decreases , though this effect essentially disappears when the hard component is modelled using the comptonization model .
the power - law index is observed to decrease in concert with the decreasing disk temperature , while again the optical depth behaves in the opposite sense , i.e. , @xmath53 increases as the disk temperature decreases .
finally , we note that this change in the disk temperature and any related changes in the structure / geometry of the accretion flow takes place on timescales of days ( e.g. , see fig .
[ disk_properties ] ) , consistent with evolution of the inner accretion disk on the viscous timescale .
the analysis presented in the preceding sections is dependent on a number of assumptions regarding the properties of the individual x - ray binary systems , see table [ bhc_table ] . in particular , the inner radius of the accretion disk depends on the line of sight column density and the assumed distance and inclination of the source ( @xmath160 , see [ accretion_disk ] ) .
likewise , the broadband spectral fitting ( see [ bband_spec ] ) will be sensitive to the column density as it is the data at optical and uv wavelengths that provides the constraints on deviations of the accretion disk temperature profile from that expected for a steady state disk .
we discuss these points in further detail below .
+ _ 1 ) column density
@xmath161 : _ + there are a number of different issues regarding the column density , which affect different areas of the paper .
the assumed value of the column density will effect the low energy region of the x - ray spectrum , and hence the normalization of the measured accretion disk component .
the assumed abundance model will alter the magnitude of flux absorbed by the intervening gas for a given column density . a different gas / dust ratio will effect the disk temperature profile for the joint x - ray / uv fits .
@xcite have demonstrated , via _ chandra _ ccd grating observations of a sample of black hole binaries , that the column density is consistent with remaining constant for those binary systems dominated by disk accretion .
this is in contrast to the hmxbs ( e.g. , cyg x-1 , lmc x-1 ) , where a large stellar wind may be present .
we emphasise here that the winds detected from accretion disks during certain stages of an xrb outburst @xcite are different to the absorbing line of sight column , i.e. , density / temperature , making it possible to separate the local / non - local absorption component .
the column density is held fixed at a value consistent with previous observations , see table [ bhc_table ] for details .
this is in agreement with the observed absence of significant local absorption in these systems @xcite , with obvious exceptions for the case of the wind - accreting hmxbs .
care must also be taken with the chosen abundance model , for example , the metal abundances in the latest solar model differs from the @xcite abundances used in the absorption model herein ( ` phabs ` ) .
updated models reflecting our improved knowledge of the solar abundance contain lower metal abundances , which imply decreased absorption for a constant column density , i.e. , inferring larger disk components via the extra soft excess .
such models are available in xspec for usage with the ` phabs ` model via the wilms " or aspl " abundance models @xcite .
we have repeated the x - ray fits assuming the aspl " abundances and find that while the measured disk properties may change during individual observations at the 5%
10% level , with disk temperature increasing and inner radius decreasing , the disk temperature and inner radius distributions remain consistent with those plotted in fig .
[ disk_temp_histogram ] .
the accretion disk contribution to the optical / uv portion of the spectrum is especially susceptible to the chosen gas to dust ratio . throughout the broadband fitting a value of @xmath162 was assumed to link the optical / uv ( ` redden ` ) to the x - ray ( ` phabs ` ) absorption @xcite . however , different values have been measured previously e.g. , @xmath163 @xcite , @xmath164 @xcite .
the broadband fits in
[ bband_spec ] were repeated using both gas / dust ratios above in order to determine if the observed accretion disk temperature profile behavior is real or simply an artifact due to the assumed law . as noted earlier for the disk temperature and radius ,
the temperature profile is observed to vary with respect to the values determined in [ bband_spec ] for individual observations .
these deviations are typically at less than the 5% level , but the overall distribution is similar to that presented in fig .
[ histogram_tprofile ] , e.g. , 68% of the observation have p @xmath42 0.64 and 90% have p @xmath42 0.68 when using the @xcite gas / dust ratio in comparison to 62% and 66% for the @xcite model respectively .
+ _ 2 ) mass , distance & inclination @xmath165 : _ + less than half the systems in our sample have dynamical constraints on the black hole mass . for the other systems , we assume a fiducial mass of @xmath166 , in agreement with the peak in the mass distribution for the stellar mass black holes with dynamical mass estimates @xcite
. this mass will be a good estimate for the majority of the systems , though we might expect systems such as igr j17091 - 3624 , grs 1758 - 258 & 1e 1740.7 - 2942 to be outliers in the mass distribution in analogy with grs 1915 + 105 .
however , the mass of the black hole is only important in those plots where luminosities are plotted in eddington units .
the measured accretion disk properties are independent of mass , at least as characterized by the models used herein .
the distances to the majority of these systems are unknown , and are instead estimates based on a number of factors ( e.g. , the observed column density , observed outburst flux , magnitude of the optical counterpart in outburst / quiescence ) , and comparison with those systems whose properties have been constrained via detailed observations , see table [ bhc_table ] .
the distance enters the inner radius calculation as a linear factor and it is highly unlikely that any of the distances are over / under estimated by a factor of @xmath1 2 .
crucially , we have no reason to believe that we are systematically over or underestimating the distance and as such these errors should tend towards canceling each other out , for the sample as a whole . as with distance
, the inner radius scales proportional to @xmath167 .
this amounts to a factor of 1.7 for an inclination of 70@xmath168 . to date , no eclipses have been observed from any of these systems indicating that typical inclinations are less than 70@xmath168 .
for more typical inclinations of 30@xmath168 60@xmath168 ( see @xcite ) , the inclination will increase the inner radius by between 7% 40% percent .
bearing these caveats in mind , we believe that the distributions of accretion disk properties presented herein radii ( e.g. , @xmath169 , see fig .
[ disk_temp_histogram ] ) are valid to within the limitations of the currently available data .
as an illustrative test case for the broadband fits , we plot the results of these fits to representative xrt+uvot spectra of gx 339 - 4 . in fig .
[ gx339_bband_hss ] , we plot fits to a soft state observation ( obsid : 00030919005 ) .
in fact , this observation is more accurately categorized as being a soft intermediate state , see @xcite .
the unabsorbed model components are indicated by the thick grey dashed lines .
all 3 models return a similar temperature for the accretion disk of @xmath170 , and disk normalization of @xmath42 1000 , i.e , @xmath171 .
as can be seen in the figure , there is a clear difference between the model with the steady state accretion disk where the optical emission is dominated by the power - law component and those models with an irradiated disk , where the optical emission is dominated by the disk component .
however , the power - law index is poorly constrained by the x - ray data , which is dominated by the disk emission .
both of the variable temperature profile models return non - standard profiles , i.e. , @xmath172 ; however , the standard disk model provides a statistically comparable fit to the data .
we must now consider the returned parameters and their implications .
in particular , we note that both of the models that utilize a power - law to generate the hard x - ray emission , return hard photon indices ( @xmath173 ) .
such hard power - law emission is in conflict with the well known soft power - laws normally observed in the soft disk dominated states , i.e. , @xmath13 @xcite .
hence for the soft state the data actually favour a model consisting of an irradiated accretion disk and a corona , i.e. , ` diskpbb+comptt ` . in fig .
[ gx339_bband_lhs ] , we plot fits to a hard state observation of gx 339 - 4 ( obsid : 00030943002 ) , where the unabsorbed model components are again indicated by the thick grey dashed lines .
in contrast to the soft state observation above , here the ` diskbb+po ` model returns a significantly poorer fit in comparison to the ` diskpbb ` models , i.e. , @xmath174 .
this is due to an inability of this model to re - produce the observed optical / uv slope as may be clearly seen in the residuals ( middle panel ) .
the models with the ` diskpbb ` component are both observed to provide fits of a statistically similar quality .
the returned accretion disk parameters are also similar , e.g. , @xmath175 . crucially in both models , a non steady state accretion disk is required
. the measured temperature profile for the accretion disk is @xmath176 .
this temperature profile is consistent with that expected from an accretion disk that is being irradiated .
such irradiation is expected to occur in the hard state , when the x - ray emission is dominated by the hard x - ray component , i.e. , @xmath177 .
finally , we note the crucial role played by the broadband coverage provided by the xrt plus uvot combination .
in particular , the coverage from optical to uv wavelengths is crucial in allowing us to constrain the temperature profile of the accretion disk , which in this case is consistent with an irradiated accretion disk , i.e. , @xmath178 .
lcccccccccc + [ -2.0ex ] source & @xmath179 & @xmath180 & p@xmath181 & spec & d & f(m ) & m@xmath182 & @xmath183 & ref + [ 0.5ex ] & & & [ hrs ] & & [ kpc ] & [ @xmath157 ] & [ @xmath157 ] & [ @xmath184 & + [ 0.5ex ] + [ -2.0ex ] ngc 300 x-1 & 00:55:09.00 & -37:42:12.16 & 32.8 & wr & 1800 & 2.6 @xmath185 0.3 & 12 24 & 0.6 & @xmath186 + [ 0.5ex ] ic 10 x-1 & 00:20:29.05 & + 59:16:52.3 & 35.6 & wr & 660 & 7.4 @xmath185 1.25 & @xmath35 23.1 & 12.5 & @xmath187 + [ 0.5ex ] lmc x-3 & 05:38:56.63 & -64:05:03.2 & 40.9 & b3v & 50.0 @xmath185 2.3 & 2.3 @xmath185 0.3 & 5.9 9.2 & 0.6 & @xmath188 + [ 0.5ex ] gx 339 - 4 & 17:02:49.5 & -48:47:23 & 42.1 & & 4.0 @xmath185 2.0 & 5.8 @xmath185 0.5 & & 6.0 & @xmath189 + [ 0.5ex ] gro j1655 - 40 & 16:54:00.14 & -39:50:44.9 & 62.9 & f3 f5iv & 3.2 @xmath185 0.2 & 2.73 @xmath185 0.09 & 6.0 6.6 & 7.4 & @xmath190 + [ 0.5ex ] sax j1819.3 - 2525 & 18:19:21.58 & -25:24:25.1 & 67.6 & b9iii & 9.8 @xmath185 2.5 & 3.13 @xmath185 0.13 & 6.8 7.4 & 1.7 & @xmath191 + [ 0.5ex ] m33 x-7 & 01:33:34.12 & + 30:32:11.6 & 82.8 & o7 o8iii & 840 & 0.48 @xmath185 0.08 & 15.65 @xmath185 1.45 & 1.0 & @xmath192 + [ 0.5ex ] lmc x-1 & 05:39:38.7 & -69:44:36 & 101.5 & o7iii & 50.0 @xmath185 2.3 & 0.14 @xmath185 0.05 & 4.0 10.0 & 11.0 & @xmath193 + [ 0.5ex ] cyg x-1 & 19:58:21.68 & + 35:12:05.78 & 134.4 & o9.7iab & 2.1 @xmath185 0.1 & 0.244 @xmath185 0.005 & 14.8 @xmath185 1.0 & 6.6 & @xmath194 + [ 0.5ex ] grs 1915 + 105 & 19:15:11.55 & + 10:56:44.8 & 804.0 & k miii & 11.5 @xmath185 0.5 & 9.5 @xmath185 3.0 & 10.1 13.4 & 43.5 & @xmath195 + [ 0.5ex ] & & & & & & & & + [ 0.5ex ] 1e 1740.7 - 2942 & 17:43:54.88 & -29:44:42.5 & 12.7 & & 8.5 & & & 100 & @xmath196 + [ 0.5ex ] h1743 - 322 & 17:46:15.61 & -32:14:00.6 & & & 10 & & & 18 & @xmath197 + [ 0.5ex ] 4u 1630 - 47 & 16:34:01.61 & -47:23:34.8 & & & 10 & & & 82 & @xmath198 + [ 0.5ex ] grs 1758 - 258 & 18:01:12.67 & -25:44:26.7 & 18.5 & & 8.5 & & & 15 & @xmath199 + [ 0.5ex ] 4u 1957 + 11 & 19:59:24.0 & + 11:42:30 & 9.3 & & 10 & & & 1.5 & @xmath200 + [ 0.5ex ] slx 1746 - 331 & 17:49:48.94 & -33:12:11.6 & & & 7.0 & & & 7.0 & @xmath201 + [ 0.5ex ] & & & & & & & & + [ 0.5ex ] igr j17091 - 3624 & 17:09:07.6 & -36:24:24.9 & & & 10 & & & 8 & @xmath202 + [ 0.5ex ] swift j1539.2 - 6277 & 15:39:11.96 & -62:28:02.3 & & & 10 & & & 4.5 & @xmath203 + [ 0.5ex ] xte j1652 - 453 & 16:52:20.33 & -45:20:39.6 & & & 10 & & & 49 & @xmath204 + [ 0.5ex ] igr j17098 - 3628 & 17:09:45.93 & -36:27:57.30 & & & 10 & & & 9.0 & @xmath205 + [ 0.5ex ] igr j17497 - 2821 & 17:49:38.1 & -28:21:17 & & & 10 & & & 50 & @xmath206 + [ 0.5ex ] xte j1752 - 223 & 17:52:15.095 & -22:20:32.782 & & & 8.5 & & & 4.7 & @xmath207 + [ 0.5ex ] swift j1753.5 - 0127 & 17:53:28.29 & -01:27:06.22 & & & 6.0 & & & 2.0 & @xmath208 + [ 0.5ex ] xte j1817 - 330 & 18:17:43.526 & -33:01:07.47 & & & 8.5 & & & 1.5 & @xmath209 + [ 0.5ex ] xte j1818 - 245 & 18:18:24.43 & -24:32:17.96 & & & 3.5 & & & 5.4 & @xmath210 + [ 0.5ex ] swift j1842.5 - 1124 & 18:42:17.44 & -11:25:03.95 & & & 8.5 & & & 3.9 & @xmath211 + [ 0.5ex ] xte j1856 + 053 & 18:56:42.92 & + 05:18:34.3 & & & 10 & & & 44 & @xmath212 + [ 0.5ex ] + [ -2.0ex ] | we present a systemic analysis of all of the stellar mass black hole binaries ( confirmed & candidate ) observed by the _ swift _
observatory up to june 2010 .
the broad _ swift _ bandpass enables a trace of disk evolution over an unprecedented range in flux and temperature .
the final data sample consists of 476 x - ray spectra containing greater than 100 counts , in the 0.6 10 kev band .
this is the largest sample of high quality ccd spectra of accreting black holes published to date .
in addition , strictly simultaneous data at optical / uv wavelengths are available for 255 ( 54% ) of these observations .
the data are modelled with a combination of an accretion disk and a hard spectral component .
for the hard component we consider both a simple power - law and a thermal comptonization model .
an accretion disk is detected at greater than the 5@xmath0 confidence level in 61% of the observations .
lightcurves and color - color diagrams are constructed for each system .
hardness luminosity and disk fraction luminosity diagrams are constructed and are observed to be consistent with those typically observed by _ rxte _ , noting the sensitivity below 2 kev provided by _
swift_. the observed spectra have an average luminosity of @xmath1 1% eddington , though we are sensitive to accretion disks down to a luminosity of @xmath2 .
thus , this is also the largest sample of such cool accretion disks studied to date .
the accretion disk temperature distribution displays two peaks consistent with the classical hard and soft spectral states , with a smaller number of disks distributed between these .
the distribution of inner disk radii is observed to be continuous regardless of which model is used to fit the hard continua .
there is no evidence for large scale truncation of the accretion disk in the hard state ( @xmath3 ) , with all of the accretion disks having radii @xmath4 .
plots of the accretion disk inner radius versus hardness ratio reveal the disk radius to be decreasing as the spectrum hardens , i.e. , enters the hard state .
this is in contrast to expectations from the standard disk truncation paradigm and points towards a contribution from spectral hardening .
the availability of simultaneous x - ray and optical / uv data for a subset of observations , facilitates a critical examination of the role of disk irradiation via a modified disk model with a variable emissivity profile ( i.e. , @xmath5 ) .
the broadband spectra ( x - ray
opt / uv ) reveal irradiation of the accretion disk to be an important effect at all luminosities sampled herein , i.e. , @xmath6 .
the accretion disk is found to dominate the uv emission irrespective of the assumed hard spectral component .
overall , we find the broadband soft state spectra to be consistent with an irradiated accretion disk plus a corona , but we are unable to make conclusive statements regarding the nature of the hard state accretion flow ( e.g. , adaf / corona vs jet ) . finally , the _ swift _ data reveal a relation between the flux emitted by the accretion disk and that emitted by the corona , for this sample of stellar mass black holes , that is found to be in broad agreement with the observed disk
corona relationship in seyfert galaxies , suggesting a scale invariant coupling between the accretion disk and the corona . |
soit @xmath9 un corps .
@xmath9-algbre magmatique @xmath10 est un espace vectoriel @xmath11 muni dune application bilinaire qui chaque bipoint @xmath12 associe @xmath13 .
pour tout @xmath14 on note @xmath15 lapplication linaire
de @xmath11 dfinie par : @xmath16 on munit @xmath11 du crochet suivant : @xmath17=x*y - y*x.\ ] ] soit @xmath18 on munit @xmath19 du crochet de lie habituel : @xmath20=f\circ g - g\circ f,~\text{pour tout } f , g\in\mathcal{l}(e).\ ] ] @xmath10 est dite algbre pr - lie gauche @xcite si : @xmath21}=[l_{x},l_{y}],~\text{pour tout}~x , y\in e,\ ] ] i.e. @xmath22 ainsi une algbre pr - lie gauche @xcite est un @xmath9-espace vectoriel @xmath23 muni dune application bilinaire @xmath24 vrifiant la relation suivante : pour tout @xmath25 et @xmath26 dans @xmath23 : @xmath27 cest - - dire si on note par @xmath28 lexpression @xmath29 la relation ( [ eqprelie ] ) scrit : @xmath30 de la mme manire si on a : @xmath31 on dira que @xmath23 est une algbre pr - lie droite . en particulier , si @xmath32 est une algbre pr - lie gauche alors @xmath33 est une algbre pr - lie droite o @xmath34 est dfini par : @xmath35 une algbre pr - lie @xmath32 munie du crochet : @xmath17=x*y - y*x~~ \text{pour tout}~x , y\in l,\ ] ] est une algbre de lie .
soit @xmath36 lespace vectoriel engendr par les arbres enracins @xmath37 sommets .
@xmath38 lespace vectoriel gradu @xmath39 .
lalgbre de hopf de connes - kreimer @xcite est dfinie comme lalgbre symtrique de @xmath38 , i.e : @xmath40 . on note @xmath41 lunit de @xmath5 identifie larbre vide .
le coproduit @xmath42 @xcite est dfini par : @xmath43 o @xmath44 dsigne lensemble des coupes admissibles @xmath45 de larbre @xmath46 , o @xmath47 correspond au branchage et @xmath48 correspond au tronc .
on rappelle quune coupe est un ensemble non vide dartes .
une coupe est dite non admissible sil existe un sommet de @xmath46 tel que le trajet de la racine ce dernier rencontre au moins @xmath49 artes de la coupe .
une coupe lmentaire est une coupe contenant une seule arte de @xmath46 . si @xmath50 est une fort , on a : @xmath51 @xmath5 est une algbre de hopf gradue , la graduation est suivant le nombre de sommets .
@xmath52 soit @xmath54 le dual gradu de @xmath5 : @xmath55 o @xmath56 dsigne lespace vectoriel engendr par les forts @xmath37 sommets .
alors @xmath54 est une algbre de hopf @xcite . on note
@xmath57 la base duale dans @xmath58 de la base des forts de degr @xmath37 dans @xmath56 .
la correspondance @xmath59 dfinit un isomorphisme linaire de @xmath60 sur son dual gradu @xmath61 .
( voir @xcite ) pour tout @xmath46 arbre dans @xmath38
llment @xmath62 est un caractre infinitsimal ( i.e. @xmath63 pour tout @xmath25 arbres dans @xmath38 , o @xmath64 est lelment neutre pour le produit de convolution @xmath65 ) .
lespace des caractres infinitsimaux est une algbre de lie pour le crochet : @xmath66=\delta_{t}\ast\delta_{t'}-\delta_{t'}\ast\delta_{t}.\ ] ] @xmath62 et @xmath67 sont primitifs dans @xmath54 donc leur crochet de lie aussi .
+ soient @xmath68 et @xmath69 des arbres dans @xmath38 , on dfinit : @xmath70 o @xmath71 dsigne lensemble des coupes lmentaires de @xmath69 .
ainsi @xmath72 est le nombre de coupes lmentaires de @xmath69 tels que @xmath73 et @xmath74 par suite : @xmath75 o @xmath76 est le nombre de coupes lmentaires de @xmath69 tels que @xmath73 et @xmath77 .
on obtient : @xmath66=\delta_{t\rightarrow t ' -t'\rightarrow t}.\ ] ] @xmath78 pour un arbre @xmath46 on note @xmath79 son facteur de symtrie , cest - - dire le nombre dautomorphismes de @xmath46 .
on dfinit une autre loi @xmath80 lie la prcdente comme suit : @xmath81 o @xmath82 est le nombre de manires de greffer @xmath46 sur @xmath83 pour obtenir @xmath69 .
@xmath84 on a de faon vidente : @xmath85 o @xmath86 est lisomorphisme linaire de @xmath87 dans @xmath87 dfini par @xmath88 pour tout arbre @xmath46 .
[ greffe ] les lois de greffe @xmath89 et @xmath90 sont des lois pr - lie gauche .
soient @xmath68 et @xmath91 des arbres dans @xmath38 , on a : @xmath92 o @xmath93 est donn par la greffe de @xmath83 sur le sommet @xmath94 de @xmath91 .
par suite : + @xmath95 + de mme on a : @xmath96 ainsi @xmath97 .
on voit que le membre de droite est symtrique en @xmath46 et @xmath83 , ce qui montre que @xmath90 est pr - lie gauche .
lassertion pour la loi @xmath89 sen dduit immdiatement grce . +
@xcite soit @xmath98 lespace vectoriel engendr par les arbres qui contiennent au moins une arte . on considre @xmath0 comme lalgbre symtrique
de @xmath98 i.e. @xmath99 . lunit de
@xmath0 est identifie larbre sans artes @xmath100 .
cette algbre est caractrise par son coproduit @xmath101 tel que @xmath102 et pour tout arbre @xmath46 diffrent de larbre vide on dfinit : @xmath103 o une sous - fort @xmath104 de @xmath46 est soit la fort triviale @xmath100 et dans ce cas @xmath105 , soit une collection @xmath106 de sous - arbres disjoints de @xmath46 et non triviaux ( i.e. ayant chacun au moins une arte ) .
larbre @xmath107 est larbre contract que lon obtient en crasant chaque composante de la sous - fort sur un point ( voir @xcite ) . @xmath108 cette algbre est une algbre de hopf gradue @xcite , la graduation tant suivant le nombre des artes .
soit @xmath6 le dual gradu de @xmath0 , soit @xmath109 la base duale du @xmath0 .
pour tout @xmath110 , @xmath111 est un caractre infinitsimal ( donc primitif dans @xmath6 ) .
si on note par @xmath112 le produit de convolution dans @xmath6 , on obtient que lensemble des caractres infinitsimaux de @xmath0 est une algbre de lie pour le crochet suivant : @xmath113=z_{t}\star z_{u}-z_{u}\star z_{t},\ ] ] o @xmath46 et @xmath114 sont des arbres dans @xmath98 . on a : + @xmath115 + o @xmath116 est le nombre de sous arbres de @xmath94 isomorphes
@xmath46 tel que le contract @xmath117 soit isomorphe @xmath114 .
on dfinit la loi @xmath2 comme linsertion dun arbre dans un autre ( voir @xcite ) , pour deux arbres @xmath46 et @xmath83 on dfinit : @xmath118 on obtient alors : @xmath119 si @xmath120 dsigne la projection de @xmath6 sur @xmath121 paralllement @xmath122 o @xmath123 est une fort non triviale ( produit dau moins deux arbres ) . on note @xmath124 , on aura : @xmath125 soit @xmath46 un arbre et @xmath123 une fort non vide .
alors : @xmath126 -si @xmath123 est un arbre on a directement le rsultat .
+ -si @xmath123 est une fort non triviale , soit @xmath114 un arbre .
alors : + @xmath127 ainsi : @xmath128 + [ pitilde ] soient @xmath129 deux arbres alors : @xmath130 les lments primitifs de @xmath6 forment une algbre de lie pour la convolution @xmath112 , ce qui prouve : + @xmath131 + la loi dinsertion est une loi pr - lie gauche .
soient @xmath129 et @xmath94 trois arbres , on a : + @xmath132 ce qui implique que : @xmath133 par suite @xmath2 est une loi pr - lie gauche .
+ la proposition [ greffe ] peut se montrer dune manire analogue .
lalgbre @xmath0 coagit gauche sur lalgbre de connes - kreimer @xmath5 ( voir @xcite ) par lunique morphisme dalgbre
@xmath134 dfini par : @xmath135 et @xmath136 si @xmath137 .
[ diagramme](voir le thorme @xmath138 de @xcite ) lapplication @xmath139 vrifie : @xmath140 i.e. le diagramme suivant commute o : @xmath141 [ neutre ] soient @xmath142 et @xmath143 alors on a : @xmath144 @xmath145 voir corollaire @xmath146 de @xcite .
en plus des deux lois pr - lie @xmath89 et @xmath90 dfinies au paragraphe [ gr ] , on dfinit pour @xmath147 : @xmath148 o @xmath149 dsigne le nombre de sous arbres de @xmath150 isomorphes @xmath114 tels que le contract @xmath151 soit isomorphe @xmath94 .
@xmath152 on dfinit maintenant une autre loi pr - lie lie la loi @xmath2 par @xmath153 tels que pour tout @xmath154 : @xmath155 o @xmath156 reprsente le nombre de manires dinserer @xmath114 dans @xmath94 pour obtenir @xmath150 .
@xmath157 on a de faon vidente une relation analogue : @xmath158 la question qui se pose ici est la suivante : y a - t - il une relation reliant les deux structures pr - lie @xmath3 et @xmath2 ( ou de faon quivalente , @xmath159 et @xmath160 ) ?
dans ce paragraphe nous montrons quil y a une relation reliant les deux structures pr - lie prcdentes .
en consquence si @xmath142 est un caractre infinitsimal , en particulier @xmath165 est un lment primitif donc : @xmath166 par suite pour tout @xmath167 on obtient : @xmath168 soient @xmath169 la projection de @xmath170 sur les lments primitifs de @xmath54 paralllement @xmath171 o @xmath123 est une fort non triviale .
- si @xmath123 est un arbre , on a directement le rsultat car @xmath173 .
+ - si @xmath123 est une fort non triviale , soit @xmath114 un arbre dans @xmath5 alors deux cas se prsentent : @xmath174 ou @xmath175 , o @xmath176 est la co - unit pour lalgbre de hopf @xmath5 . +
* si @xmath174 alors : @xmath177 or @xmath178 donc le rsultat est vrai pour larbre @xmath41 . +
* si @xmath179 on a : + @xmath180 car le contract @xmath181 est un arbre .
ainsi on obtient que : + @xmath182 par suite on conclut que : @xmath183 soit @xmath150 un arbre dans @xmath38 alors : + -si @xmath186 alors @xmath187 + -si @xmath188 alors : @xmath189 -si @xmath190 o @xmath191 deux arbres .
+ on a @xmath139 est un morphisme dalgbre donc : + @xmath192 soit @xmath193 et @xmath196 , daprs consquence du thorme [ premier theo ] on a : + @xmath197 + donc : + @xmath198 + par suite : + @xmath199 + ainsi + @xmath200 + do : + @xmath201 + ce qui prouve que : + @xmath202 on prend dans cet exemple : @xmath213 + - calcul de @xmath214 .
+ on a : + @xmath215 + donc : @xmath216 notons respectivement @xmath217 et @xmath218 les termes prcdents . on aura alors : + @xmath219 + et + @xmath220 .
+ par suite : @xmath221 -calcul de @xmath222 : + on a : + @xmath223 par suite : + @xmath224 notons @xmath225 et @xmath226 les termes de la somme prcdente on aura : + @xmath227 + ainsi : @xmath228 la diffrence ( [ differenceun])-([differencedeux ] ) , nous donne : + @xmath229 -calcul de @xmath230 : + on a : + @xmath231 + par suite : + @xmath232 ce qui est bien la diffrence @xmath233 le thorme [ joli formule ] est un cas particulier dun phnomne gnral de nature opradique .
rappelons quune oprade vectorielle unitaire @xmath234 est la donne pour tout @xmath235 dun espace vectoriel @xmath236 muni dune action du groupe symtrique @xmath237 , et dune loi de composition : @xmath238 satisfaisant des axiomes dassociativit et dquivariance par rapport aux actions des groupes symtriques en prsence .
lunitarit scrit @xmath239 ( @xmath9 est le corps de base ) o @xmath64 vrifie : @xmath240 @xmath241 pour tout @xmath242 et pour tout @xmath243 la composition partielle @xmath244 est dfinie par : @xmath245 la somme des compositions partielles : @xmath246 est une loi pr - lie droite @xcite .
+ lassociativit et lquivariance impliquent notamment : 1 .
pour tout @xmath247 et pour tout @xmath248 on a : @xmath249 o @xmath250 est la permutation de @xmath251 obtenue en permutant @xmath252 laide de @xmath253 , puis en permutant @xmath254 laide de @xmath255 .
+ 2 . pour tout @xmath256 , pour tout @xmath257 et pour tout @xmath258 on a lgalit suivante dans @xmath259 : @xmath260 3 . pour tout @xmath261 et
pour tout @xmath262 on a : @xmath263 o @xmath264 est la permutation de @xmath265 obtenue en permutant les lments @xmath266 du j - me bloc par @xmath267 , puis en permutant les blocs par @xmath255 .
lalgbre libre un gnrateur @xmath268 sur loprade @xmath234 @xcite est donne par @xmath269 .
les trois galits ci - dessus impliquent que les compositions partielles passent au quotient ainsi que la loi pr - lie @xmath270 , et que nous avons pour toute opration n - aire @xmath271 sur @xmath268 et @xmath272 @xmath273 en sommant sur tous les @xmath274 possibles de @xmath275 jusqu @xmath276 puis sur tout les @xmath277 de @xmath275 @xmath37 , on obtient donc : @xmath278 le thorme [ joli formule ] en dcoule en prenant pour @xmath234 loprade pr - lie @xcite et pour @xmath271 la loi de greffe droite : @xmath279 on obtient bien : @xmath280 le thorme [ joli formule ] sen dduit par le fait que : @xmath281 et @xmath282 . | dans larticle @xcite , damien calaque , kurusch ebrahimi - fard et dominique manchon ont introduit un nouveau coproduit sur une algbre commutative de forts darbres enracins @xmath0 . lalgbre de lie des lments primitifs du dual gradu @xmath1 est munie dune structure pr - lie gauche , note @xmath2 qui sexprime en termes dinsertion dun arbre dans un autre . dans ce travail nous montrons quil y a une relation `` de drivation '' reliant cette structure la structure pr - lie gauche de greffe dun arbre sur un autre @xcite , note @xmath3 , obtenue sur lalgbre de lie des lments primitifs du dual gradu @xmath4 de lalgbre de hopf de connes - kreimer @xmath5 @xcite . in @xcite , damien calaque , kurusch ebrahimi - fard and dominique manchon define a hopf algebra @xmath0 by introducing a new coproduct on a commutative algebra of rooted forests .
the space of primitive elements of the graded dual @xmath6 is endowed with a left pre - lie product denoted by @xmath7 defined in terms of insertion of a tree inside another . in this work
we prove a `` derivation '' relation between the pre - lie structure @xmath2 and the left pre - lie product @xmath3 @xcite on the space of primitive elements of the graded dual @xmath8 of the connes - kreimer hopf algebra @xmath5@xcite , defined by grafting . |
the recently discovered cluster of galaxies , 1e 0657@xmath056 ( @xmath1 ) , was observed by _ asca _ ( _ advanced satellite for cosmology and astrophysics _ ; see tanaka , inoue & holt , 1994 ) and found to have an unusually high temperature of @xmath7 kev ( tucker _ et al .
_ 1998 ; hereafter t98 ) .
a full account of this result and other properties of 1e 0657@xmath056 can be found in t98 .
cosmologies in which the density of the universe is critical ( @xmath3 ) predict such a small number of clusters with this high a temperature that an @xmath3 universe is strongly disfavored by this single measurement of the temperature of 1e 0657@xmath056 ( see t98 and references therein ) . since the cosmological implications are so strong , it is important to test the robustness of this observational result .
so far there are only two other clusters with temperatures greater than 10 kev ( ms 1054@xmath00321 , donahue _
_ 1998 ; a2163 , markevitch _
_ 1996 ; both have @xmath8 kev ) . even for the latter objects
the probability of observing @xmath9 kev for just one source in an @xmath3 universe is less than 0.01 ( as opposed to @xmath10 for 1e 0657@xmath056 ) . in this _
letter _ we re - analyse the _ asca _ data for 1e 0657@xmath056 using a variety of different data analysis methods and spectral analysis assumptions .
the different methods of background - subtraction and spectral extraction are described in detail in 2 and the spectral analysis of these data is described in 3 .
our conclusions are presented in 4 .
_ asca _ observed 1e 0657@xmath056 in 1996 may 1011 for a duration of @xmath11 ks .
since it is important that our results be reproducible , we describe the data reduction and analysis in some considerable detail .
the reader is referred to tanaka _
( 1994 ) for details of the instrumentation aboard _
asca_. the two solid state imaging spectrometers ( sis ) , hereafter sis0 and sis1 , with a bandpass of @xmath12 kev , were operated in 2-ccd faint and bright modes .
the two gas imaging scintillators ( gis ) , hereafter gis2 and gis3 , with a bandpass of @xmath13 kev , were operated in standard ph mode .
the sis faint and bright mode data were combined and the sis energy scale was fixed using a prescription based on measurements of cas a ( dotani _ et al .
this corrects for the continuing decline in the cti of the ccds with time and is based on interpolation or extrapolation of the cas a measurements .
the last cas a measurements incorporated into the analysis software were made about three months after the 1e 0657@xmath056 observation and since the cti changes very slowly with time , the interpolation of these measurements results in a systematic uncertainty in the gain which is much less than @xmath14 at 1 kev ( dotani _ et al .
version 1.1 of the sis response matrix generator was used in the analysis .
however , no corrections were made for any possible offsets , fluctuations or distortion of the ccd dark level distribution ( the latter also known as the ` rdd ' effect ) .
such corrections ( which themselves are subject to uncertainty ) can only be applied to faint mode data .
this can be done with the software tool ` correctrdd ' which , however , as of the time of writing , still does _ not _ correct for the loss in ccd efficiency .
we note that part of our objective is to reproduce the results of t98 who did not use correctrdd either because a working version of it was not available at the time they did their analysis . in 1-ccd mode ( the sis observation mode for most point sources )
the rdd effect in sis0 is negligible for data taken as late as early 1998 .
in addition to rdd there are two other effects which lead to a loss of low - energy efficiency which _ are _ important even in 1-ccd mode .
one of these is due to on - board software mis - grading events for some settings of the low - energy discriminator , and the origin of the other effect is not yet understood .
the latter can reduce the sis efficiency at 0.6 kev by as much as @xmath15 .
documentation of these effects can be found on the world - wide web pages of the _ asca _ guest observer facility ( hereafter _ asca _ gof ) .
this work is still very much on - going . for discussions of some calibration issues with respect to earlier data
( but still relevant here ) see grandi _
( 1997 ) and orr _ et al . _
( 1998 ) . considering the finite energy resolution and signal - to - noise of sis data ,
the loss of low energy efficiency ( which seems to be worse in sis1 ) can manifest itself as an apparently larger measured column density ( up to @xmath16 ) . in 2-ccd mode ,
the mode relevant for this observation of 1e 0657@xmath056 , the effect is worse in both sis0 and sis1 . typically , for data taken in late 1997 to early 1998 column densities measured for clusters by sis0 and sis1 can be a factor @xmath17 and @xmath18 larger than the corresponding column densities measured by the rosat _
pspc_. however , a systematic study and development of quantitative corrections is still underway ( see _ asca _ gof web pages ) .
data were screened so that accepted events satisfied the following criteria : ( i ) data were taken outside the saa ; ( ii ) the time since or before passage through the saa or a satellite day / night transition was @xmath19 s ; ( iii ) the elevation angle to earth was @xmath20 ; ( iv ) the magnetic cut - off rigidity ( cor ) was @xmath21 ; the deviation of the satellite from the nominal pointing position was @xmath22 ; ( v ) the sis parameters measuring active ccd pixels registered @xmath23 active pixels per second , and ( vi ) the radiation - belt monitor registered @xmath24
hot and flickering pixels in the sis were removed .
screening resulted in net ` good ' exposure times in the range @xmath2525.1 ks for the four instruments .
source events were extracted from the gis using circular regions with radii 6. for the sis , two different types of extraction region were used : ( i ) circular , with radii 4 ( which we call c ) and ( ii ) rectangular ( which we call r ) , using the exact dimensions used by t98 .
the reason for doing this is that no azimuthal dependence is included in the current xrt responses so it is useful to check whether there are any systematic errors introduced by using non - circular extraction regions .
background spectra were accumulated in three different ways .
the first , which we call ` b4 ' , used released blank sky fields which were taken in 4-ccd mode .
events from the entire field - of - view were used , excluding the outer ring in the gis .
the second method ( which we call ` ls ' , for local , same chip ) used local background events from the same chip as the source at radii greater than 5.5 from the source centroid .
for the gis , annulii centered on the source and having inner and outer radii of 8 and 12 respectively were used . the third method ( which we call ` lo ' for local , off - chip ) used the entire off - source sis chip for background ( slight differences in the spectral responses of the different chips are negligible compared to statistical and background - subtraction errors ) . for the gis , elliptical regions with minor and major axes of 6 and 12 respectively , offset as far as possible from the source centroid , were used .
all background data were subject to identical selection criteria as the on - source data .
using these different methods of background subtraction will then test the robustness of any spectral results to differences in the background in different ccd modes and sis chips , positions on the sky , epoch , and contamination from the source itself ( in the ls backgrounds ) .
the background - subtracted spectra extracted from circular regions resulted in count rates in the range 0.21 - 0.32 ct / s for the four instruments , whilst the smaller , rectangular sis regions yielded only 0.21 ct / s for sis0 and 0.14 ct / s for sis1 . for the x - ray telescope ( xrt ) effective areas we used version 2.72 of ascaarf with both empirical effective area factors activated ( these force residuals from fitting _ asca _ data for the crab nebula to be less than 3% ) .
the ` extended source ' option was used in generating the xrt effective areas .
we also tried some preliminary spectral fitting with responses made using the ` point source ' option and found negligible difference in the resulting source spectral parameters . from this
we conclude that systematics in the xrt responses due to the spatial extent of 1e 0657@xmath056 can be neglected in comparison to other sources of systematic error .
we fitted spectra from the four _ asca _ instruments simultaneously in the range 0.510 kev with a raymond - smith plasma model , plus a cold , neutral absorber at @xmath26 .
a total of three free , interesting parameters were involved ( the plasma temperature , @xmath27 , plasma abundances relative to solar , @xmath28 ( fe / h = @xmath29 ) , and the column density , @xmath30 .
the four independent instrument normalizations were also free but the deviation of any of the four normalizations from their mean was less than 20% .
such a large systematic difference between fluxes measured by the sis and gis is common for extended sources with photon - limited spectra .
hereafter we quote mean fluxes and luminosities from the two sis , which agree with each other to better than 1% , and are known to be more reliable than gis normalizations .
note that the galactic value of the column density , obtained from dickey and lockman ( 1990 ) is @xmath31 .
first we tested the effect of using the three different methods of background subtraction ( see 2 ) for circular source extraction regions .
the results are shown in table 1 ( fits 13 ) .
the ratios of the data to the best - fitting model for fit # 3 is shown in figure 1 .
for all three fits we obtain @xmath6 kev , @xmath32 , and @xmath28 @xmath33 .
for all three parameters , the systematic differences arising from the different methods of background subtraction are less than the statistical errors ( see table 1 ) . from fit
# 1 we obtain a 0.510 kev observed flux and intrinsic luminosity of @xmath34 and @xmath35 respectively ( @xmath36 and @xmath37 ) .
similar values were obtained for the remaining spectral fits . next we tested the effect of using rectangular extraction regions for the sis spectra ( as in t98 ) , using the exact dimensions as used by t98 ( fit # 4 ) and obtained very similar parameters to fits 13 , except for a slightly lower column density .
note that the background and region extractions for fit # 4 are the closest in similarity to those used by t98 .
thus neither the different background spectra nor different extraction regions can account for the significantly lower temperatures that we obtain .
note that our 90% confidence regions do not overlap with the t98 measurements of @xmath7 kev .
we then investigated the effect of fixing @xmath30 at the galactic value .
the results for fit # 5 ( see table 1 ) pertain to the use of the ` ls ' background ( 2 ) and circular source extraction .
the background ( ` b4 ' ) and extraction regions ( ` r ' ) for fit # 6 are similar to those used by t98 . it can be seen that the resulting values of @xmath27 are now in agreement with t98 . the ratios of data to best - fitting model for fit # 5 are shown in figure 2 in which the deficit at low energies is clearly seen in the data ( the fit is only marginally acceptable at the 90% confidence level ) .
this deficit is not seen in the data presented by t98 because they used smaller spectral extraction regions ( as in fit # 6 ) so their statistical errors were larger .
note that t98 do not explicitly state whether the absorption was fixed at the galactic value , nor do they give the actual value used , if it was .
we can easily test whether the high temperature of @xmath2 kev is an artifact of fixing the column density at the galactic value when the actual data clearly require a stronger low - energy turnover ( figure 2 ) .
we simulated a 40 ks exposure sis0 spectrum using an input model consisting of a raymond - smith plasma with @xmath38 kev , @xmath39 , and the same flux as the real 1e 0657@xmath056 data , and abundances of 0.2 relative to solar .
we then fitted the simulated data with the same model but with the column density fixed at the galactic value and all other parameters free .
we obtained a best - fitting temperature of @xmath40 kev .
this clearly demonstrates that if the column density in fitting the real 1e 0657@xmath056 data is fixed when the data really have a stronger low - energy cut - off ( whether or not this is due to extra intrinsic absorption or a calibration problem is irrelevant ) the result is an artificially high temperature .
further proof of this comes from excluding the low - energy part of the real data for spectral fitting .
we performed four - instrument fits in the restricted energy band 1.010 kev , with only galactic absorption included .
this significantly reduces the impact of any low - energy turnover .
we used background spectra ( ` b4 ' ) and extraction regions ( ` r ' ) most closely resembling those of t98 and the results are shown in table 1 , fit # 7 .
we obtain a lower temperature of @xmath41 kev . we found that excluding even more of the low - energy data reduced the measured temperature even further .
the next spectral fit in table 1 ( fit # 8) uses sis spectra made from only single - pixel ccd events ( grade 0 ) in order to investigate the effect on the measured temperature of the less uncertain calibration of two - pixel events ( under some circumstances - see the _ asca _ gof web pages ) . table 1 shows that the best - fitting temperature is still @xmath42 kev .
the basic _ asca _ event data used in t98 were produced from the raw data using ` revision 1 ' ( rev1 ) of the processing pipeline software .
the event data we used was produced later , using ` revision 2 ' ( rev2 ) of that software ( rev1 was phased out ) .
the final spectral fit in table 1 ( fit # 9 ) was performed to check whether the difference between rev1 and rev2 data could be responsible for the unusually high cluster temperature measured by t98 . in practice
the only differences between rev1 and rev2 are minor updates in the calibration files for both sis and gis that are used to compute temporal gain corrections . for the sis
, rev2 uses data from new calibration observations not available at the time rev1 was used ( as opposed to extrapolation using old data ) .
for the gis , rev2 incorporates temporal gain corrections not applied in rev1 .
another change in rev2 data is that pre - screened events files are made using different screening criteria than those used for rev1 .
however , this latter difference is not relevant to us because we dot use pre - screened event data and nor did t98 .
fit # 9 in table 1 is essentially a repeat of fit # 1 except that fit # 9 uses rev1 data and fit # 1 uses rev2 data .
it can be seen that the resulting spectral parameters for rev1 and rev2 are indistinguishable .
we have re - analysed the _ asca _ data for the cluster of galaxies 1e 0657@xmath056 using a variety of different data analysis and spectral fitting assumptions .
the actual methods of data extraction and background subtraction cover the entire range of methods used , plus more , in the majority of the hundreds of papers published using _
the one method of extraction and analysis used by tucker _
( 1998 ) yielded the unusually high temperature of @xmath43 kev .
all but one of our methods of analysis gave a much lower temperature , @xmath5 kev .
the only way we could reproduce the higher temperature was to fix the column density at the galactic value , even though this leads to a low - energy deficit in the data , which clearly require a stronger low - energy cut - off . we proved that fixing the column density in this way leads to an artificially high temperature by simulating _ asca _
data for a cluster with @xmath38 kev and a column density higher than galactic , and then re - fitting the data with the column fixed at the galactic value .
the additional low - energy cut - off in the real 1e 0657@xmath056 data may be astrophysical in origin ( intrinsic or line - of - sight absorption ) or it may be due to as yet uncalibrated loss of ccd efficiency in the sis .
the latter is a very difficult problem because there are a number of time - dependent effects which are causing a loss in the low - energy ccd efficiency and these are under on - going investigation by the _ asca _ gof and _ asca _ instrument teams . even aside from calibration problems
, it is recommended that the absorbing column density is not fixed at the galactic value when spectral fitting _
asca _ data .
this is because doing so assumes there is no other photo - electric absorption , either along the line - of - sight , or intrinsic . if this assumption is wrong then , as we have shown , one obtains artificially high cluster temperatures .
besides , such a spectral fitting assumption allows no room for systematic errors in the actual value for galactic absorption . even if there is no other absorption and no calibration problem , an error in the galactic column density will lead to an error in the measured temperature .
we note that our results will also be useful to anyone who wants to estimate the systematic effects of different background - subtraction methods for _ asca _ data for any source with a similar count - rate to that of 1e 0657@xmath056 during the observation reported here ( @xmath44 ct / s ) .
the temperature of @xmath5 kev for 1e 0657@xmath056 still puts it amongst the handful of clusters with @xmath9 kev , along with ms 1054@xmath00321 ( donahue _ et al .
_ 1998 ) and a2163 ( markevitch 1996 ) . as such
, their existence still poses a problem for a universe with @xmath45 in that the probability of observing only one of these clusters is @xmath46 .
the author thanks t. dotani , richard mushotzky and una hwang for valuable discussions , the anonymous referee for some extremely helpful comments , and the _ asca _ mission operations team at isas , japan , and all the instrument teams for their dedication and hard work in making these _ asca _ observations possible .
this research made use of the heasarc archives at the laboratory for high energy astrophysics , nasa / gsfc .
lcccccc 1 & ls & c & @xmath47 & @xmath48 & @xmath49 & 726.4/769 2 & lo & c & @xmath50 & @xmath51 & @xmath52 & 745.0/769 3 & b4 & c & @xmath53 & @xmath54 & @xmath55 & 748.4/769 4 & b4 & r & @xmath56 & @xmath57 & @xmath58 & 637.9/689 5 & ls & c & 6.55 fixed & @xmath59 & @xmath60 & 814.1/769 6 & b4 & r & 6.55 fixed & @xmath61 & @xmath62 & 670.2/690 7 & b4 ( 1 - 10 kev ) & r & 6.55 fixed & @xmath63 & @xmath64 & 577.0/649 8 & ls ( g0 ) & c & @xmath65 & @xmath66 & @xmath67 & 708.8/703 9 & ls ( rev1 ) & c & @xmath68 & @xmath48 & @xmath69 & 714.5/764 dickey , j.m . , & lockman , f.j . 1990 ,
28 , 215 donahue , m. , voit , m. , gioa , i. , luppino , g. , hughes , j. p. , & stocke , j. t. 1998 , apj , 502 , 550 dotani , t. , _ et al .
_ 1997 , _ ascanews _ , 5 , 14 grandi , p. , _ et al .
_ 1997 , , 325 , l17 markevitch , m. , mushotzky , r. f. , inoue , h. , yamashita , k. , furuzawa , a. , & tawara , y. 1996 , apj , 456 , 437 orr , a. , molendi , s. , fiore , f. , grandi , p. , parmar , a.n .
, & owens , a. 1997 , , 324 , l77 tanaka , y , inoue , h. , & holt , s.s .
1994 , , 46 , l37 tucker , w. , _ et al .
_ 1998 , , 496 , l5 ( t98 )
* figure 1 * + ratio of the _ asca _ data for 1e 0657@xmath056 ( @xmath1 ) to the best - fitting raymond - smith model ( corresponding to fit # 3 in table 1 ) .
the best - fitting temperature in this case is @xmath70 kev .
* figure 2 * + the result of fitting a raymond - smith model to the 1e 0657@xmath056 _ asca _ data with absorption fixed at the galactic value ( see text ) . shown
are the ratios of the _ asca _ data to the best - fitting raymond - smith model ( corresponding to fit # 5 in table 1 ) .
the best - fitting temperature in this case is @xmath71 kev . | a recent x - ray observation of the cluster 1e 0657@xmath056 ( @xmath1 ) with _ asca _ implied an unusually high temperature of @xmath2 kev .
such a high temperature would make it the hottest known cluster and severely constrain cosmological models since , in a universe with critical density ( @xmath3 ) the probability of observing such a cluster is only @xmath4 . here
we test the robustness of this observational result since it has such important implications .
we analysed the data using a variety of different data analysis methods and spectral analysis assumptions and find a temperature of @xmath5 kev in all cases , except for one class of spectral fits .
these are fits in which the absorbing column density is fixed at the galactic value . using simulated data for a 12 kev cluster ,
we show that a high temperature of @xmath2 kev is artificially obtained if the true spectrum has a stronger low - energy cut - off than that for galactic absorption only .
the apparent extra absorption may be astrophysical in origin , ( either intrinsic or line - of - sight ) , or it may be a problem with the low - energy ccd efficiency . although significantly lower than previous measurements , this temperature of @xmath6 kev is still relatively high since only a few clusters have been found to have temperatures higher than 10 kev and the data therefore still present some difficulty for an @xmath3 universe .
our results will also be useful to anyone who wants to estimate the systematic errors involved in different methods of background subtraction of _ asca _ data for sources with similar signal - to - noise to that of the 1e 0657@xmath056 data reported here . |
for the semilinear equation @xmath2 a traveling wave can connect the two steady states @xmath3 and @xmath4 in various situations .
this is the case when @xmath0 is unstable and @xmath1 is stable ( fisher / monostable ) .
this is also the case when @xmath0 and @xmath1 are both stable ( allen - cahn / bistable ) .
it is known that these waves are attractive and are obtained as the long time limit of the dynamics ( [ eq : semilnear ] ) associated with compactly supported initial data ( see @xcite ) .
when @xmath0 and @xmath1 are unstable , in between there is a stable steady state of @xmath5 which prevents any traveling wave to exists . for systems and non - local equations ,
the classification becomes more complicated because a steady state can be turing unstable , which means that @xmath1 is unstable with respect to some periodic perturbations in a bounded range of periods . in this note
we consider the nonlocal fisher equation ( [ eq : nonlocal ] ) , the simplest to produce turing instability .
@xmath6 with @xmath7 the steady state @xmath1 can be turing unstable when the fourier transform of @xmath8 changes sign and @xmath9 is large enough ( see @xcite ) .
this creates several differences with the monostable or bistable equations .
first of all , one can observe that the solution of ( [ eq : nonlocal ] ) associated with a compactly supported initial datum does not converge , for large times , toward the traveling wave but it converges towards a more complicated structure ( see @xcite , @xcite and @xcite ) .
the numerical simulation presented in b ) of figure [ fig : nonlocal1 ] shows that the solution of the evolution equation converges to a pulsating front , that is a function @xmath10 which is periodic in its second variable .
these types of fronts typically arise in the framework of reaction - diffusion equations with periodic coefficients ( @xcite , @xcite ) . in a ) of figure [ fig : nonlocal1 ] , the solution seems to converge to the superposition of a traveling wave and a pulsating front , with two different speeds .
these periodic patterns are a symptom of turing instability on the full line @xmath11 .
\a ) ) with kernel ( [ eq : phi2 ] ) .
the computational domain is @xmath12 $ ] .
left : the isovalues show that it is not a traveling wave but a more complicated structure ; the bottom subplots are zooms of the top subplots .
right : the function @xmath13 connects the ( dynamically ) unstable state @xmath0 with a periodic tail at @xmath14 .
two values of @xmath9 are used a ) @xmath15 ; b ) @xmath16.,title="fig:",scaledwidth=45.0% ] ) with kernel ( [ eq : phi2 ] ) .
the computational domain is @xmath12 $ ] .
left : the isovalues show that it is not a traveling wave but a more complicated structure ; the bottom subplots are zooms of the top subplots .
right : the function @xmath13 connects the ( dynamically ) unstable state @xmath0 with a periodic tail at @xmath14 .
two values of @xmath9 are used a ) @xmath15 ; b ) @xmath16.,title="fig:",scaledwidth=45.0% ] b ) ) with kernel ( [ eq : phi2 ] ) .
the computational domain is @xmath12 $ ] .
left : the isovalues show that it is not a traveling wave but a more complicated structure ; the bottom subplots are zooms of the top subplots .
right : the function @xmath13 connects the ( dynamically ) unstable state @xmath0 with a periodic tail at @xmath14 .
two values of @xmath9 are used a ) @xmath15 ; b ) @xmath16.,title="fig:",scaledwidth=45.0% ] ) with kernel ( [ eq : phi2 ] ) .
the computational domain is @xmath12 $ ] .
left : the isovalues show that it is not a traveling wave but a more complicated structure ; the bottom subplots are zooms of the top subplots .
right : the function @xmath13 connects the ( dynamically ) unstable state @xmath0 with a periodic tail at @xmath14 .
two values of @xmath9 are used a ) @xmath15 ; b ) @xmath16.,title="fig:",scaledwidth=45.0% ] secondly , it is proved in @xcite that a traveling wave @xmath17 always exists for all @xmath18 , with a generalized formulation @xmath19 the authors were only able to obtain , due to the nonlocal effect , the weak boundary condition at @xmath14 rather than the expected condition @xmath20 .
when @xmath9 is small enough or when the fourier transform of the kernel @xmath21 is positive everywhere , the traveling wave connects @xmath0 to @xmath1 . but this leaves open the question to know whether for @xmath9 large , the traveling wave solution of ( [ eq : phi2 ] ) connects the ( dynamically ) unstable state @xmath0 to a stable periodic state or to the turing unstable state @xmath1 .
also , when @xmath21 can take negative values , it is proved in @xcite that for @xmath9 large enough monotonic traveling waves can not exist . in the sequel
, we will firstly focus on the case @xmath22 and @xmath23}(x).\ ] ] then we also test two other cases [ eq : phinew ] @xmath24},\qquad \beta=\big(\frac{3}{4}\big)^{\frac{1}{3 } } \label{eq : phinew1}\ ] ] @xmath25}(x ) \label{eq : phinew2}\ ] ] in all these cases , @xmath26 takes negative values .
for instance , for ( [ eq : phi2 ] ) , @xmath27 . finally , to complete the tests , the traveling waves for the gaussian kernel @xmath28 which has positive fourier transform , are displayed . in section [ sec : algorithm ] , we develop a specific algorithm which allows us to build the traveling wave ( [ eq : nonlocaltw ] ) and not the pulsating front .
then , in section in [ sec : results ] , we perform numerical simulations to decide which alternative holds true .
numerically one can only solve the problem on a bounded domain of length @xmath29 ( this is also the analytical construction in @xcite ) @xmath30 the convolution is computed by extending @xmath13 by 1 on @xmath31 and @xmath0 on @xmath32 .
the parameter @xmath33 , small enough , is needed for technical reasons but intuitively its value has to be below the oscillations observed in figure [ fig : nonlocal2 ] . our algorithm for solving ( [ eq : nonlocaltwl ] ) is to divide the computational domain into two parts : @xmath34 $ ] and @xmath35 $ ] .
being given @xmath36 , in each interval an elliptic equation with dirichlet boundary conditions is solved @xmath37 the convolution term is computed by defining @xmath38 as @xmath39 on @xmath40 , @xmath41 and @xmath42 , @xmath1 on @xmath31 and @xmath0 on @xmath32 .
but the equation does not necessarily hold true at @xmath43 .
we define @xmath44 so as to impose that the jump of derivatives at zero vanishes @xmath45+\int_{x_l}^{x_r}\mu v ( 1-\phi\star v)dx.\ ] ] [ lemma : continuous ] when @xmath46 satisfies ( [ eq : uu1u2 ] ) and ( [ eq : sigmafv ] ) simultaneously , then @xmath13 is @xmath47 on @xmath48 and satisfies ( [ eq : nonlocaltwl ] ) .
we can write abstractly this problem as a fixed point for a system of two equations @xmath49 .
it is straightforward to make it discrete using finite differences . the most efficient way to solve
it is to use newton iterations .
the numerical results we present in this section are obtained with the hat function @xmath8 in ( [ eq : phi2 ] ) , @xmath50 and we study the effect of the bifurcation parameter @xmath9 .
the diffusion term is treated implicitly by centered three point finite difference while the reaction term is put explicit . in order to verify that the iterative scheme described in section [ sec : algorithm ] converges to the right solution , a crucial quantity to look at
is the truncation errors at zero @xmath51 here @xmath52 are the values of @xmath13 at the grid points @xmath53 .
the convergence results are displayed in table [ tab : fisher2mu64 ] .
one can see that @xmath54 converges to zero as @xmath55 , which shows that our numerical results is a good approximation of ( [ eq : nonlocaltwl ] ) on the whole computational domain .
note that better accuracy of @xmath54 can be obtained if we use higher order numerical integration methods for the convolution term .
.[tab : fisher2mu64 ] convergence of the truncation error at zero @xmath54 and of the traveling velocity @xmath44 for various values of @xmath56 and @xmath57 , with kernel ( [ eq : phi2 ] ) and @xmath58 .
for @xmath59 the speed is @xmath60 . [
cols="^,^,^,^,^,^,^,^,^,^,^,^",options="header " , ] the traveling wave shapes for @xmath61 , @xmath62 and @xmath63 for kernel ( [ eq : phi2 ] ) are depicted in figure [ fig : nonlocal2 ] .
when @xmath61 , we observe a monotone traveling wave which connects @xmath0 to @xmath1 , as for the local fisher equation . when @xmath9 grows , some oscillations appear .
numerically , when we increase @xmath57 , the amplitudes of the tail decrease and the bigger @xmath9 is , the slower the amplitudes decrease . the shapes of @xmath38 suggest that though @xmath1 is turing unstable with the kernel ( [ eq : phi2 ] ) when @xmath62 and @xmath63 , the traveling waves will still connect @xmath1 to @xmath0 .
we do not obtain the same type of structure than when we compute the solution of the evolution equation depicted in figure [ fig : nonlocal1 ] .
this means that there exist some traveling waves that connect @xmath0 to @xmath1 , but that these waves do not attract the solution of the cauchy problem associated with compactly supported initial data . ) with kernel ( [ eq : phi2 ] ) .
left : the numerical results for @xmath64 and @xmath62 .
right : the results for @xmath65 , the top subplot depicts @xmath38 while the bottom subplot is a zoom of the tail.,title="fig:",scaledwidth=45.0% ] ) with kernel ( [ eq : phi2 ] ) . left : the numerical results for @xmath64 and @xmath62 .
right : the results for @xmath65 , the top subplot depicts @xmath38 while the bottom subplot is a zoom of the tail.,title="fig:",scaledwidth=45.0% ] lastly , we consider the critical value of @xmath9 for which the monotonicity of the traveling waves is broken . since the monotone traveling waves always connect @xmath1 to @xmath0 , we perform the linearization , close to @xmath14 , by assuming @xmath66 with @xmath67 a real positive number .
are close to a linearization ] after inserting this form into ( [ eq : nonlocaltw ] ) , using @xmath22 and the smallness of @xmath68 , @xmath67 can be determined by @xmath9 through the equation @xmath69 a quadratic equation for @xmath70 , which gives @xmath71 .
thus the critical @xmath9 that makes @xmath67 no longer exist is @xmath72 .
we have checked that this threshold @xmath73 is correct on the numerical values .
the maximum of @xmath13 is @xmath1 when @xmath74 , but exceeds slightly @xmath1 when @xmath61 . in figure
[ fig : nonlocal2 ] with @xmath61 , the wave is nearly monotonic , but , checking the numerical values , the maximum of @xmath38 is @xmath75 .
this indicates that actually @xmath38 might be not monotone even for @xmath57 finite .
in @xcite the authors have proved that , when @xmath76 , the traveling wave is not monotone .
one open question is to know if the traveling wave is monotone when @xmath77 .
our simulation answers positively to this open question numerically .
the numerical results with kernels in ( [ eq : phinew ] ) are depicted in figure [ fig : newkernel ] . with large @xmath62
, we can see similar phenomena such that some oscillations appear and the waves become nonmonotone .
for the two kernels in ( [ eq : phinew ] ) , the state @xmath78 is turing unstable as for ( [ eq : phinew ] ) and figure [ fig : newkernel ] suggests again that the traveling waves will connect @xmath1 to @xmath0 . ) with @xmath62 and kernels in ( [ eq : phinew ] ) .
left : the numerical results with kernel ( [ eq : phinew1 ] ) .
right : the results for ( [ eq : phinew2 ] ) . in all these pictures ,
the top subplot depicts @xmath38 while the bottom subplot is a zoom of the tail.,title="fig:",scaledwidth=45.0% ] ) with @xmath62 and kernels in ( [ eq : phinew ] ) .
left : the numerical results with kernel ( [ eq : phinew1 ] ) .
right : the results for ( [ eq : phinew2 ] ) . in all these pictures ,
the top subplot depicts @xmath38 while the bottom subplot is a zoom of the tail.,title="fig:",scaledwidth=45.0% ] h. berestycki , g. nadin , b. perthame and l. ryzhik , the non - local fisher - kpp equation : traveling waves and steady states .
nonlinearity ( 22 ) , 28132844 , 2009 . m. a. fuentes , m. n. kuperman and v. m. kenkre , nonlocal interaction effects on pattern formation in population dynamics , phys .
91(15 ) , 15810414.115810414.4 , 2003 . a.n .
kolmogorov , i.g . petrovsky and n.s .
piskunov , etude de l quation de la diffusion avec croissance de la quantit de matire et son application un problme biologique , bulletin universit detat moscou ( bjul .
moskowskogo gos .
univ . ) , 126 , 1937 . | this note investigates the properties of the traveling waves solutions of the nonlocal fisher equation .
the existence of such solutions has been proved recently in @xcite but their asymptotic behavior was still unclear .
we use here a new numerical approximation of these traveling waves which shows that some traveling waves connect the two homogeneous steady states @xmath0 and @xmath1 , which is a striking fact since @xmath0 is dynamically unstable and @xmath1 is unstable in the sense of turing . *
key - words * nonlocal fisher equation , turing instability , traveling waves . + * ams subject classification * : 35b36 , 92b05 , 37m99 . |
there is strong evidence to suggest that the three gauge couplings of the strong , electromagnetic and weak interactions are unified at a high energy scale in a single gauge interaction based on a simple group , such as su(5 ) or so(10 ) , as long as the desert below the unification scale is described by the minimal supersymmetric extension of the standard model ( the mssm ) .
furthermore , the combination of supersymmetry ( susy ) and grand unification yields models with numerous attractive features : the embedding of the standard - model matter multiplets into a few irreducible representations of the gut group , the technically natural preservation of a hierarchy between the weak and gut scales , a longer proton lifetime to allow agreement with current experimental lower bounds , the correct prediction of the ratio of @xmath2 quark to @xmath3 lepton masses , simple anstze for the remaining fermion masses , and a picturesque scenario for radiative electroweak symmetry breaking .
we have chosen , therefore , to look beyond the gauge unification prediction of the weak mixing angle and examine the unification of third - family yukawa couplings @xcite , for the most part within its natural context of so(10 ) unification .
by yukawa couplings we mean the couplings of the top , bottom and tau to the higgs doublets which generate their masses when electroweak symmetry is broken .
the third family is singled out because of its relatively large yukawa couplings : it seems reasonable to suppose that they arise at tree - level from the simplest interactions , while the masses and mixings of the other generations require more complex , perhaps also higher - order and certainly very model - dependent structures .
our study also applies more generally to scenarios in which these yukawa couplings are unified but not in the context of an so(10 ) gut , or even models in which the yukawas are only comparable near the gut or planck scales .
we begin by considering the most immediate prediction of this yukawa unification , namely the top quark mass @xcite .
however , we are quickly led to consider in some depth the more general question of how the top - bottom mass hierarchy could be generated in the mssm , and how this hierarchy depends on the initial conditions of the renormalization group ( rg ) evolution at the gut scale @xcite
. we will conclude with a discussion of how natural ( or unnatural ! ) such a hierarchy seems in this context , what its other phenomenological predictions might be , and how one could hope to improve the theoretical picture or obtain experimental corroboration .
one consequence of yukawa unification is immediate , and independent of other assumptions except for the qualitative nature of the rg evolution equations of the mssm .
since the yukawa couplings of the top and bottom quarks ( and the @xmath3 lepton ) are always comparable , the large ratio of the top mass versus the bottom ( or tau ) mass must be due to a large ratio of the higgs vacuum expectation values ( vevs ) which give rise to their masses .
namely , since the up - type and down - type matter fermion masses arise from couplings to the up - type and down - type higgs multiplets ( @xmath4 ) , respectively , the large ratio @xmath5 is not a consequence of a large ratio of yukawas @xmath6 but rather of large @xmath7 .
thus yukawa unification generically implies @xmath8 .
further assumptions are necessary to make any precise predictions
. we will assume the following three throughout most of this work , although we will point out those conclusions which are more general : 1 .
the masses of the third generation , @xmath9 , @xmath10 and @xmath11 , originate from renormalizable yukawa couplings of the form @xmath12 in a supersymmetric gut with a gauge group containing ( the conventional ) so(10 ) ; @xmath13 denotes the 16-dimensional spinor representation of so(10 ) containing the third - generation standard - model fermions ( plus the right - handed neutrino which we assume to be superheavy ) and their superpartners .
2 . the evolution of the gauge and yukawa couplings in the effective theory beneath the so(10 ) breaking scale is described by the rg equations of the mssm .
the two higgs doublets lie predominantly in a single irreducible multiplet of so(10 ) .
the first and third assumptions serve to define what we mean by yukawa unification , while the second allows us to relate this gut - scale unification to weak - scale observables . from ( i ) and ( iii ) it follows that the third - generation yukawas must arise from either a @xmath14 or a @xmath15 interaction with an so(10 ) higgs multiplet .
the latter leads to the boundary conditions @xmath16 at the gut scale , but the resulting ratio of @xmath17 at low energies is far too low to be consistent with experiment ( at least within the perturbative regime , and unless very large threshold corrections to the @xmath2 mass arise at low energies @xcite ) .
thus we are restricted to using the @xmath18 , and hence the boundary condition @xmath19 with this boundary condition , and using the unification of gauge couplings to fix the unification scale and the gauge coupling at that scale , we can now evolve the yukawa couplings down to the weak scale for any given value of @xmath20 .
the idea is that the three observable masses @xmath9 , @xmath10 and @xmath11 are functions of the four parameters @xmath21 , @xmath22 , @xmath23 and @xmath24 , and @xmath25 are in turn determined by the unification in terms of @xmath20 and the gut scale @xmath26 .
since the latter is already known from gauge unification , we are left with three observable masses as functions of only two parameters , @xmath20 and @xmath24 .
thus we use two observables , @xmath10 and @xmath11 to fix @xmath20 and @xmath24 , and thereby predict the third observable @xmath9 . a detailed analysis of the rg evolution , and the consequent predictions ,
has already been presented @xcite .
the results , namely the values of @xmath25 and the ratio @xmath27 all at the weak scale , are plotted in fig . 1 as functions of @xmath20 .
( these curves actually use 2-loop rg evolution , but at this point the difference between 1- and 2-loop equations is not important .
for the final predictions of @xmath9 we use the full 2-loop evolution and 1-loop matching conditions . ) evidently , larger @xmath20 values correspond to a heavy top and to a smaller @xmath28 ratio .
the experimental value @xmath29 is found @xcite from the qcd sum rules value and is evolved to the weak scale using 2-loop qcd running .
we find , allowing for @xmath30 to vary between roughly 0.11 and 0.12 , the shaded range shown in fig . 1 .
thus , in the absence of any large corrections in the matching between the @xmath28 evolved down from the gut scale and the @xmath31 in the standard model , we find @xmath32 , which implies a heavy top .
140 160 700 500floridafig1.ps133
for a precise prediction , 2-loop rg equations must be used along with 1-loop matching functions .
these matching functions include logarithmic corrections from infinite counterterms as well as nonlogarithmic contributions from finite graphs .
the former are given elsewhere @xcite ; they are generally quite small , and invariably increase the top mass as the superpartner masses increase .
the latter are more interesting , since they can be very large @xcite .
the dominant corrections arise from the graphs of fig . 2 , which match the value of the @xmath2 mass as evolved down from the gut scale to the value in the low - energy theory . typically the gluino graph dominates , yielding a corrected value @xmath33 where @xmath34 , @xmath35 and @xmath36 is the gluino mass while @xmath37 is the mass of the heaviest superpartner in the loop ( more exact expressions may be found in our previous work ) .
the important observation here is that the bottom mass gets a large contribution from the up - type higgs at 1-loop order , whereas its tree - level mass was small due to the small vev of the down - type higgs . in the usual scenario with small @xmath24 ,
the bottom was light because its yukawa coupling was small , or in other words because it was protected by an approximate chiral symmetry .
thus any higher - order corrections would also be suppressed by this approximate symmetry . in the large @xmath24 scenario
these corrections are not suppressed : there is an enhancement of @xmath38 , which overcomes the usual @xmath39 loop factor to give a correction of order 1 to the @xmath2 mass at least if @xmath40 .
phenomenologically , the result is that @xmath10 can not be predicted with any certainty unless we know something about the superspectrum , and if @xmath10 is uncertain then so is the top mass prediction .
140 440 700 680floridafig2.ps2 42 in fig .
3 we show the top pole mass prediction , now to full 2-loop order , as a function of the @xmath41 running parameter @xmath42 in two cases .
the top curves and the higher horizontal axis correspond to a hierarchical spectrum , in which the squarks are heavy whereas the @xmath43 parameter and gaugino masses are light
. then the corrections @xmath44 are small and the top mass is predicted to be above 180 gev or so .
the bottom curves and the lower horizontal axis correspond to a roughly degenerate spectrum in which the corrections to the @xmath2 mass are @xmath45 and negative ( i.e. @xmath28 should be lowered by 25% in fig . 1 before matching to the experimental value ) .
now the top can be significantly lighter .
in fact , this last argument can be turned around : if the threshold corrections are too large in magnitude and negative then the top mass prediction will be below the experimental lower bound , while if they are large and positive then no value of @xmath20 will allow agreement with @xmath31 .
we thus find the following bounds on the superspectrum : if @xmath20 is allowed to vary , then @xmath46 whereas for fixed @xmath20 the appropriate limits can be read from fig .
1 . 150 180 650 600floridafig3.ps3 54
do we have any experimental information about these corrections ?
of course we have no direct evidence of any of the superpartners , but we can appeal to their indirect appearance in loop diagrams . consider again the diagrams of fig . 2 , but with one of the @xmath2 quarks replaced by a strange quark using a flavor - changing vertex , and with a photon attached in all possible ways .
we see that the same processes will lead to a contribution to the rare decay @xmath1 , and with a similar enhancement of @xmath47 over the usual mssm scenario @xcite .
the dependence of these diagrams on the superpartner masses is somewhat similar to that of @xmath48 , except that ( a ) now the operator is of higher dimension and so is suppressed by the mass of the heaviest superpartner , and ( b ) typically the higgsino - mediated diagram dominates .
if the parameters appearing in this higgsino diagram , namely @xmath43 , @xmath0 ( the trilinear soft susy - breaking parameter ) and the squark masses , are all comparable and of order the z mass , then this diagram gives a contribution to the amplitude for @xmath1 many times bigger than the standard model or the usual mssm amplitudes , and is clearly ruled out by the cleo limit @xcite . to restore agreement with experiment , either the overall superpartner mass scale must be raised far above the electroweak scale , or else @xmath43 or @xmath0 or both must be suppressed relative to the squark mass in the loop .
more quantitatively , we find that either the masses must all be raised to at least @xmath49 , or else if both @xmath43 and @xmath0 are near the z mass then the squarks must be above either @xmath50 gev or @xmath51 gev , depending on whether these diagrams interfere destructively or constructively with the 2-higgs standard model amplitudes . in any case
, however , these restrictions do not yet tell us anything about the combination @xmath52 which appears in @xmath44 ; the link between these two will be forged below , when we study the evolution of the entire set of mssm parameters .
we have assumed in the above that the top - bottom mass hierarchy would arise from a hierarchy of vevs in the higgs spectrum , namely @xmath53 .
we now consider how such a hierarchy could be generated in the mssm .
clearly this question is of interest for any model in which the yukawas themselves do not supply a sufficient hierarchy to explain the large ratio of third - generation quark masses , not just for the equal - yukawas case to which we have specialized . in studying this question
, however , we will pay attention to which spectra are favored by large @xmath24 scenarios , and whether with such spectra and our unification assumptions we can pin down the top mass prediction .
the scalar potential of the neutral higgs bosons which leads to electroweak symmetry breaking is given by @xmath54 where @xmath55 contain the soft - breaking masses and the @xmath43 parameter in the superpotential , @xmath56 is a soft - breaking mass parameter , and the quartic terms arise from d - terms and so are given by gauge couplings .
we will evolve these parameters from the gut to the weak scale using the 1-loop rg equations of the mssm , stopping the evolution at some typical scale ( of order the squark masses ) which minimizes the effects of higher - order corrections .
the subscript 0 indicates that we will restrict our attention to this ( rg - improved ) tree - level potential rather than calculate the full 1-loop effective potential or , better yet , explicitly integrate out massive particles and consider full 1-loop matching conditions .
we expect @xcite that our qualitative discussion of the radiative symmetry breaking will not be jeopardized by this simplification .
that is , a more complete calculation will change the numerical values of the gut parameters needed for correctly breaking the symmetry , but will not significantly alter the size of the domains in parameter space where such breaking is achieved .
the conditions for this breaking are well - known : @xmath57 ensures that the potential is bounded from below , and @xmath58 guarantees the existence of a minimum away from the origin and so breaks the symmetry . in practice ,
since @xmath59 will always be much less than or at most comparable to @xmath60 and @xmath61 , we can reduce these requirements to @xmath62 ( using the expression for the pseudoscalar higgs mass ) and @xmath63 ( noting that large @xmath24 means that the up - type higgs gets the large vev ) . in the usual and very attractive scenario of radiative breaking , the two mass parameters start out at the gut scale with a universal positive value : @xmath64 , where @xmath65 is the soft - breaking mass of the @xmath18 of higgs in so(10 ) , or of the @xmath66 and @xmath67 in su(5 ) . thus the symmetry is not broken at that scale . however , in the rg evolution to the electroweak scale , the large yukawa coupling of the top quark to @xmath68 , which gives the top its mass , also drives the mass - squared parameter @xmath69 of @xmath68 negative ( with the help of the qcd coupling ) , while the absence of a large yukawa in the down sector keeps the mass - squared of @xmath70 positive . in fact , conditions ( [ eq : bndbel ] ) and ( [ eq : ewsb ] ) are easily satisfied for a large range of initial conditions if @xmath71 , resulting in a very natural picture of radiative symmetry breaking .
this picture is essentially lost in the large @xmath24 scenario , for the following two reasons : 1 . since both yukawas are comparable [ and in fact initially equal in the so(10 )
case ] , the two higgs doublets tend to run in the same way , so either both stay positive at the electroweak scale and the symmetry does not break , or both become negative and the potential becomes unbounded from below ( a situation which breaks the symmetry but in a coleman - weinberg fashion , yielding an `` electroweak scale '' orders of magnitude higher than the susy - breaking scale ) . the effects which differentiate the evolution of the two higgs doublets , namely hypercharge and the absence of a right - handed neutrino , are small and a poor replacement for the usual @xmath72 splitting .
interestingly , an @xmath73 splitting between @xmath21 and @xmath74 is still of little use since it is quickly diminished by the fixed - point behavior of these couplings .
( some of these observation have been previously made by t. banks @xcite . ) 2 .
even when electroweak symmetry is broken , a large hierarchy of vevs must be generated between the two similarly - evolving higgs doublets .
minimizing the potential @xmath75 when @xmath76 yields + @xmath77 + and @xmath78 the first equation sets the scale , but from the second equation we see that a large hierarchy in vevs requires the large hierarchy @xmath79 .
this , as we show below , implies a degree of fine - tuning between some parameters in the lagrangian .
before analyzing the implications of these two criticisms , we present the 1-loop solutions @xcite of the rg equations for the mssm mass parameters , integrated between @xmath80 and a typical squark mass of 300 gev .
( none of our results is sensitive to the exact values of these starting and stopping scales . )
the solutions depend on the dimensionless initial values of the gauge and yukawa couplings @xmath81 and @xmath20 and on the dimensionful gut - scale parameters @xmath82 , @xmath65 , @xmath83 , @xmath84 , @xmath85 , @xmath86 and @xmath87 .
@xmath82 and @xmath65 are the soft - breaking masses of the @xmath18 and the @xmath13 respectively , and @xmath87 will be explained below ; we note for now that it vanishes for universal soft - breaking masses . whenever possible , capital letters denote values at the gut scale .
the rg equations themselves are well - known and will not be presented here .
these equations dictate that the low - energy values of the various mass parameters depend very simply on the dimensionful initial values , with coefficients that depend only on the dimensionless ones .
since @xmath81 is known from gauge unification , these coefficients depend only on @xmath20 . for the representative value @xmath88 , the solutions are : @xmath89 here @xmath90 and @xmath91 are the squark and slepton doublets , respectively , and @xmath92 , @xmath2 and @xmath3
are the su(2)-singlet squarks and sleptons . for clarity of presentation ,
we have dropped from the first 7 expressions above the terms proportional to @xmath93 and to @xmath94 , since their coefficients are small enough [ @xmath95 and exhibit sufficiently small custodial - su(2 ) breaking to be negligible for purposes of symmetry - breaking , at least if @xmath85 is not very much larger than the other mass parameters .
we have also used the low - energy value of @xmath43 in eqs .
( [ eq : mu]-[eq : ma ] ) .
these solutions are useful references for the discussions below .
these solutions can also be combined @xcite with the more analytic approach briefly described in eqs .
( [ eq : expsol ] ) to give a complete , semi - analytic solution to the rg equations when custodial su(2 ) is an approximate symmetry .
to understand the second criticism above , let us examine in more detail how eq .
( [ eq : tanbet ] ) may be satisfied .
we concentrate for the moment on six relevant electroweak - scale parameters : the up- and down - type higgs masses @xmath96 and @xmath97 , a typical squark mass @xmath98 , the @xmath56 parameter in the scalar potential , a gaugino mass ( specifically the wino mass ) @xmath99 , and the @xmath43 parameter .
this last one may be set to zero by imposing a peccei - quinn symmetry on the lagrangian , so the size of @xmath43 measures the breaking of this @xmath100 symmetry ; consequently , @xmath43 is multiplicatively renormalized .
the previous two , @xmath56 and @xmath99 , along with the @xmath0 parameter , transform in the same way under a continuous @xmath101 symmetry ( so they enter into each other s rg equations ) and may be made arbitrarily small by imposing this @xmath101 symmetry . with these two symmetries in mind @xcite , we consider three possible spectra having splittings which lead to a large @xmath24 according to eq .
( [ eq : tanbet ] ) : [ cols=">,^,^,^",options="header " , ] the @xmath102 and @xmath103 are examples of extra superheavy higgs representations which are typically added in order to break this @xmath104 ( in this case by acquiring vevs in the `` @xmath105 '' direction ) and reduce the rank of the group . as usual , the spontaneous breakdown of a u(1 ) leads to a vev for its d - term , which can induce masses for the various fields which appear in this d - term .
in particular , _
if we do not assume universal soft - breaking masses _ for all scalars , then the soft - breaking masses of the @xmath102 and @xmath103 need not be equal , and therefore their vevs are also split , in proportion to their mass splitting .
this splitting in turn generates a mass splitting in the low - energy mssm lagrangian through the cross - term : @xmath106 in fact , this mechanism splits any fields which have different charges under @xmath104 .
thus the boundary conditions for the scalar masses at the gut scale become @xmath107 where the capital letters on the left - hand side serve as reminders that these are the values at the gut scale , and @xmath108 is a new soft - breaking mass parameter in the low - energy theory . with this mass
we no longer need rely on large gaugino masses to split the higgs doublets : they can start out being different , and thus even with parallel evolution the symmetry - breaking conditions ( [ eq : bndbel][eq : ewsb ] ) can apparently be satisfied .
one problem with this mechanism is evident from the initial conditions in eq .
( [ eq : initcond ] ) : not just the higgs doublets but also the squarks and sleptons are split , so an excessively large @xmath109 could lower @xmath110 or @xmath111 sufficiently to make @xmath112 or @xmath113 negative at the electroweak scale , thereby spontaneously breaking the strong or electromagnetic gauge symmetries .
if rg effects were irrelevant , namely for small @xmath20 , then @xmath82 could always be raised enough to prevent this without affecting @xmath114 . however , for @xmath71 the squark masses strongly affect the evolution of the higgs doublets [ see eqs .
( [ eq : mu]-[eq : ma ] ) ] , and only for very constrained ranges of the initial parameters can the electroweak symmetry , and only that symmetry , be spontaneously broken at a reasonable scale .
in fact , as we show in brief below , there is a focusing effect that is inherent in the mssm rg equations when @xmath115 , and which inevitably requires an adjustment of the gut - scale parameters beyond the @xmath116 level derived above .
we first show this behavior of the rg equations for completely general initial conditions , and then return to discuss the specific case of eqs .
( [ eq : initcond ] ) .
consider the rg equations of the mssm in the limit of exact @xmath117 and @xmath101 symmetries , in which @xmath118 at all scales . for future reference
, we call this scenario _ the maximally symmetric case_. this limit is interesting for two reasons : first , no large corrections arise to the @xmath2 quark mass , and the @xmath119 prediction for all values of @xmath120 falls nicely within the range allowed by experiment ( see fig .
1 ) ; in other words , a heavy top quark near its fixed - point mass favors small @xmath44 .
second , as we saw above , having a large @xmath43 and @xmath99 calls for fine - tuning @xmath56 ( or some equivalent adjustment ) , so we d like to explore the opposite limit to see whether a more natural scenario can be achieved . of course , eventually we must relax this limit to agree with lep bounds , but the qualitative behavior we shall discover will persist . if we further approximate @xmath121 and neglect the sleptonic contributions ( thereby restoring custodial symmetry ) , the rg solutions simplify considerably .
there are now five relevant parameters . in terms of their initial conditions at the gut scale , @xmath122 , @xmath123 , @xmath124 , @xmath125 and @xmath110 , the solutions at the electroweak scale are : @xmath126 where @xmath127 and @xmath128 ( note again the use of capital letters to denote gut - scale initial parameters , and recall that @xmath129 , @xmath130 , @xmath90 , @xmath92 and @xmath2 refer to the up - type higgs , the down - type higgs , the su(2)-doublet third - generation squarks , the su(2)-singlet stop and the su(2)-singlet sbottom , respectively . ) evidently , two linear combinations of masses , labeled by @xmath131 and @xmath132 at the gut scale , renormalize multiplicatively and exponentially contract at low energies for @xmath71 .
the three other linear combinations , @xmath133 , @xmath134 and @xmath135 , are invariant .
the important observation here is that in the first contraction the sum rule @xmath136the coefficient of every term is positive , while we already know that each mass - squared itself must be positive for a proper electroweak - breaking scenario .
therefore each term by itself must be small , less than @xmath137 .
this can only happen if the various combinations of invariants ( and possibly also @xmath138 ) in the expressions ( [ eq : expsol ] ) for these terms are adjusted to be small relative to @xmath131 .
the exact constraints that follow from this requirement are given explicitly elsewhere @xcite .
they are of the form @xmath139 .
we learn that , for @xmath71 where this focusing effect is important , any given model for the gut - scale soft - breaking masses must either provide an explanation of why each invariant should be small relative to the sum @xmath140 , or else that invariant must be tuned by hand to be small .
we also learn that the conditions for successful symmetry - breaking are sensitive to any other small effects .
one such effect is custodial symmetry violation , which is parametrized above by @xmath141 and results from hypercharge and @xmath23 ( or the absence of @xmath105 ) . in the running of the yukawas , both of these cause @xmath21 to slightly exceed @xmath22 , and thus drive @xmath96 below @xmath97 , as in the conventional scenario of electroweak symmetry breaking . in the running of the masses ,
the contributions of the @xmath3 yukawa have an opposite and numerically more relevant impact .
therefore the custodial - breaking effects make it harder to break the symmetry correctly significantly harder in the specialized case discussed below . in any realistic scenario
there are also contributions from the gauginos and @xmath43 , so in the end the sum rule , and therefore the general limit which must be set on @xmath133 , @xmath134 and @xmath135 , takes the form @xmath142
m_s^2 \label{eq : sumrul}\ ] ] where @xmath143 is the largest mass parameter in the initial conditions at the gut scale .
if we now return to the more specialized boundary conditions of eq .
( [ eq : initcond ] ) , we find ( after setting @xmath144 ) : @xmath145 for this choice of boundary conditions , keeping the invariants small imposes only two requirements : @xmath146 and small @xmath87 .
the upper bound on the invariants involves both @xmath147 and @xmath141 , and as we hinted above they partially cancel in the combination @xmath148 which enters into the requirements : @xmath149 , is particularly sensitive to this cancellation@xcite . ) the first requirement , eq .
( [ eq : mhmsq ] ) , entails a definite tuning of parameters to a precision of @xmath148 , which does not apparently follow from any symmetry .
note that without this requirement , either color breaks when @xmath150 or a coleman - weinberg mechanism operates when @xmath151 .
the requirement of small @xmath87 may on the other hand be natural , since [ see eq .
( [ eq : mx ] ) ] the value of @xmath109 is smaller by an order of magnitude than the soft - breaking masses whose splitting generates the d - terms , and those masses may be expected to be comparable to @xmath82 and @xmath65 . in any case , we see that because of the focusing effect of the rg equations , the d - terms can not be _ allowed _ to induce splittings bigger than those we already had through custodial su(2)-breaking effects . hence they do not eliminate the criticism that the electroweak symmetry is hard to break when the yukawas are comparable .
but there is still a significant advantage in using these d - terms , since they can now substitute for large values of @xmath99 and @xmath43 , and with light gauginos and @xmath43 it is much easier to obtain a large @xmath24 , according to eqs .
( [ eq : tanbet ] ) and ( [ eq : btune ] ) .
before putting the various observations to use in examining specific scenarios and their merits , we point out another feature of the solutions to the rg equations which will further constrain the scenarios . as is evident from eqs .
( [ eq : at]-[eq : atau ] ) ( or directly from the rg equations ) , the initial value @xmath85 hardly affects the low - energy values of @xmath152 ; they are instead largely determined in magnitude and sign by the gaugino mass @xmath153 , which also fixes the low - energy gluino mass .
( it is difficult , though perhaps not impossible for sufficiently small @xmath20 , to construct models in which @xmath154 and yet the electroweak symmetry but neither color nor charge breaks spontaneously and correctly , so we shall disregard this possibility in these proceedings .
the implications of tuning @xmath85 to cancel the gaugino mass at low energies in the expression for @xmath155 will be considered elsewhere @xcite . )
this observation , which was also emphasized by carena et al .
@xcite , directly relates the @xmath44 corrections of eq .
( [ eq : delmb ] ) to the large @xmath1 graphs discussed above .
( more precisely , the gluino- and higgsino - exchange diagrams for each process are directly related . )
the sign of this correlation @xcite is such that when @xmath156 ( i.e. the predicted @xmath28 is lowered , and therefore so is the top mass ) then the large @xmath1 graphs interfere _
constructively _ with the usual 2-higgs standard model amplitude , and vice - versa . on one hand , we see from eq .
( [ eq : muglimits ] ) or from fig . 1 , that the bounds on @xmath44 are more severe when @xmath157 .
on the other hand , as noted above , when @xmath156 the interference is constructive and the bounds on the large @xmath1 graphs are stricter .
thus these two bounds are much stronger when taken together , and translate into the following statement : either ( a ) the gauginos or @xmath43 or both are significantly lighter than the squarks , or ( b ) the superpartners are much heavier than the z.
with these remarks in mind , we first examine the popular @xcite case of universal soft - breaking masses .
this scenario has also been recently studied in some detail by carena et al . @xcite .
if all soft - breaking scalar masses are equal then the d - term contributions vanish ( @xmath158 ) , and we are left with the three parameters @xmath43 , @xmath84 and @xmath159 , in addition to @xmath86 which is adjusted at the end to obtain the correct @xmath24 [ see eq .
( [ eq : tanbet ] ) ] .
we have already mentioned that in this case we need @xmath43 comparable to @xmath84 , and both at least as big as @xmath160 , to break electroweak symmetry correctly . we have also seen that these three parameters must be tuned in order to obtain a positive @xmath161 and @xmath162 : @xmath163 next , to achieve a hierarchy of higgs vevs , @xmath86 must be adjusted very precisely such that , at low energies , @xmath164 finally , since @xmath43 and the gauginos are not lighter than the squarks , @xmath44 is rather large , and so must be negative , as can be seen from eq .
( [ eq : muglimits ] ) or from fig . 1 .
hence the @xmath165 constraint is strong , necessitating large superpartner masses of at least @xmath166 and therefore a _ further _ tuning ( by roughly another order of magnitude ) of the three parameters to achieve correct electroweak breaking .
of course , such a highly - tuned scenario is also highly predictive : for example , the spectrum is highly constrained , and the top mass is predicted by the large @xmath44 corrections to be light . in search of a more natural scenario ,
we next relax the assumption of universal soft - breaking masses , which was perhaps arbitrary to begin with .
we begin with the maximally symmetric scenario studied previously ( see also scenario c above ) , in which @xmath167 so that the @xmath100 and @xmath101 symmetries are approximately obeyed .
since this hierarchy directly implies a small @xmath44 and therefore ( from fig .
1 ) a relatively large @xmath20 , the focusing effect of eq .
( [ eq : epsl ] ) takes its toll , and once again the initial parameters this time @xmath82 , @xmath65 , and @xmath87must be adjusted to at least @xmath168 $ ] .
then , to truly get a maximally symmetric scenario , we choose to obtain a large @xmath24 not by tuning to get a small @xmath56 but rather by ( equivalently ) tuning to get a small @xmath169 , which then allows small @xmath43 and @xmath99 relative to the typical low - energy soft - breaking masses and therefore establishes approximate @xmath100 and @xmath101 symmetries even at the electroweak scale : @xmath170 where @xmath171 . after this adjustment ,
large @xmath24 is automatic : @xmath172 and no further adjustment is necessary to suppress @xmath165 since @xmath43 and @xmath84 are small .
so this scenario requires much less adjusting than the universal case , although more than just the inevitable @xmath116 tuning .
( the actual tuning needed is in reality slightly more than indicated , due to the squark- and slepton - splittings induced by the d - terms and due to the effects of custodial symmetry violation , as mentioned above ; qualitatively , though , the picture we described remains . )
it is also predictive : the superspectrum is hierarchical with light charginos and neutralinos but heavy squarks , and since @xmath20 is large so is the top mass .
we can continuously retreat from this maximally symmetric case by increasing @xmath43 or @xmath84 ( or the related parameters ) , thereby losing @xmath100 or @xmath101 , first at low energies when @xmath43 or @xmath99 become comparable to @xmath173 , and then at all energies when @xmath43 or @xmath99 become comparable to @xmath143 itself .
both may be interesting for model - building ( for example , if @xmath43 is radiatively generated by @xmath0 terms at the gut scale , then @xmath100 rather than @xmath101 symmetry should be evident ) and for comparison with experiments once the superspectrum is measured .
in either case , the tuning is comparable to the maximally symmetric case , though less tuning is needed in eq .
( [ eq : sfta ] ) and more in eq .
( [ eq : sftb ] ) . we will however defer their discussion to our more complete study @xcite , and instead consider the case where _ both _ @xmath100 and @xmath101
symmetries are abandoned in favor of a smaller @xmath20 .
we will call this scenario , unimaginatively , the asymmetric case .
it alleviates the focusing effect of eq .
( [ eq : epsl ] ) since now @xmath174 , which in turn allows the initial conditions ( specifically the d - terms ) to split the higgs doublets and the other multiplets by a large amount , so now @xmath175
. small @xmath20 also entails larger ( negative ) @xmath44 corrections to correctly predict @xmath17 , and to this end we take @xmath176 .
we then see , however , that what we gain by eliminating the focusing effect we lose by restoring the @xmath165 problem : since @xmath43 and @xmath99 are no longer small , we are forced to raise the susy scale to @xmath177 , and therefore again to tune the initial parameters to make the z light : @xmath178 since the d - term splitting of the higgs is now large , @xmath179 , we find that @xmath86 requires only the typical tuning @xmath180 this scenario is comparable in its naturalness ( or lack thereof ) to the symmetric case . the superspectrum is uniformly heavy rather than hierarchical , and the top is light since @xmath20 is small .
these last two scenarios are qualitatively the best one can hope for from an so(10 ) model with yukawa unification , for two reasons .
first , obtaining a large @xmath24 and thereby the top - bottom hierarchy is never natural in the mssm , due to the lep bounds on @xmath43 and @xmath99 which force either @xmath162 to be much heavier than the z or @xmath56 to be much lighter than the gauginos .
second , the last two scenarios illustrate how , for large @xmath20 , the _ inherent _ focusing property of the rg equations in the symmetric limit necessitates a further tuning of the initial parameters , while for small @xmath20 a similar tuning is mandated by bounds on the rate of @xmath1 .
there are also intermediate scenarios , with only one of the symmetries , but they are apparently no more natural .
these various possibilities are considered in more detail elsewhere @xcite .
the universal case is generically much more tuned than most of these scenarios , as we have shown .
hence departures from universality , although possibly dangerous for the flavor - changing neutral current interactions they can induce in some models , are strongly favored in achieving a large @xmath24 . what is the status of the predictions ?
perhaps surprisingly , the top mass is not an independent prediction of yukawa unification , but rather depends strongly ( i.e. non - logarithmically ) on a certain ratio of superpartner masses appearing in @xmath44 . since the large- and small-@xmath20 cases are equally fine - tuned , naturalness arguments do not single out any particular top mass within this so(10 ) framework . instead , information about the top mass can be combined with fig . 1 and eq .
( [ eq : delmb ] ) to restrict the mass parameters which can be consistent with yukawa unification , and to determine a favored superspectrum .
it must be admitted that , with yukawa unification , the attractive conventional picture of _ radiative _ electroweak symmetry breaking due to a large @xmath21 but small @xmath22 is largely lost : the symmetry is either broken radiatively using small custodial isospin - violating effects and extraordinarily fine - tuned initial conditions , or else the higgs doublets are already split at the gut scale .
furthermore , we have seen that all the large @xmath24 scenarios are technically unnatural . on the other hand
( and to some extent because of the necessary fine - tuning ) , they are certainly predictive and so will be tested in future accelerators : for example , if the charginos are light but the squarks are heavy , the top should also be heavy ; if the @xmath181-singlet bottom squark or the doublet sleptons were lighter than the other squarks and sleptons , this would be a sign that the d - term splittings were large @xcite ; and finally , @xmath24 can itself eventually be measured to decisively confirm or dismiss the large @xmath24 hypothesis .
even before any further experimental input , there are a few theoretical avenues worth pursuing which would make yukawa unification much more attractive .
first , as discussed also by j. lykken in these proceedings @xcite , string models which lead to true grand - unified models as their low - energy effective lagrangians can exhibit higher symmetries in certain sectors of the gut than in other sectors .
in particular , a string theory with an so(10 ) gauge symmetry might break to an effective low - energy su(5 ) gut in a stringy way , leaving _ all three _
yukawa couplings unified at the planck scale la so(10 ) but splitting the @xmath66 from the @xmath67 soft masses , namely @xmath122 from @xmath123 .
note that unlike the d - term splittings we have considered before , @xmath122 and @xmath123 could now be split without necessarily decreasing some squark or slepton masses .
this approach can thus provide more freedom in the choice of boundary conditions although as we have shown , some features of the rg equations , in particular the focusing effect , are inherent in the equations themselves in certain limits and apply to any boundary conditions , so that freedom could either be the key to a more natural scenario ( see the following remark ) or could necessitate still more arbitrary fine - tunings at the gut scale .
second , we have shown that in the symmetric case with large @xmath20 , the ratio of the soft - breaking masses for the @xmath18 and the @xmath13 needs to have a certain value if the electroweak symmetry is to break correctly . while such a value is arbitrary in the context of an so(10 ) model and
therefore apparently requires a fine - tuning , perhaps this value could be explained as a ratio of integral conformal weights in the context of a string theory into which the gut is embedded .
( notice that this value is favored simply by the unification of yukawas at some large scale , and does not depend on an so(10 ) symmetry . ) if such an explanation could be found , then the symmetric large-@xmath20 case would now be strongly favored : it would only require the minimal @xmath116 tuning ( and would predict a heavy top ! ) .
in fact , if the squarks were to be _ experimentally determined _ to be heavy while the gauginos were light , then this case would be no more fine - tuned than the conventional small @xmath24 scenario , but would have the advantage of explaining the top - bottom mass hierarchy ( through the @xmath100 and @xmath101 symmetries)which , after all , was historically the motivation for studying yukawa unification .
we would like to thank m. carena and c. wagner for enjoyable discussions , in particular regarding the sign of the correlation between the diagrams in @xmath1 and those in @xmath48 .
we have also benefited from discussions with j. lykken and f. zwirner regarding string models .
u.s . would like to thank the organizers of the second ift workshop on yukawa couplings and the origins of mass for a stimulating conference .
this work was supported in part by nsf grants phy-89 - 17438 ( u.s . ) , phy-91 - 21039 ( r.r . ) and phy-90 - 21139 ( l.j.h . ) and in part by the united states department of energy under contract no .
de - ac03 - 76sf00098 ( l.j.h . ) .
fig . 1 : : : fig . 2 : : : fig . 3 : : : 99 g.f . giudice and g. ridolfi , z. phys .
c * 41 * , 447 ( 1988 ) ; m. olechowski and s. pokorski , phys .
b * 214 * , 393 ( 1988 ) ; w. majerotto and b. msslacher , z.phys .
c * 48 * , 273 ( 1990 ) ; p.h .
chankowski , phys .
d * 41 * , 2877 ( 1990 ) ; b. ananthanarayan , g. lazarides , and q. shafi , phys .
d * 44 * , 1613 ( 1991 ) ; q. shafi and b. ananthanarayan , _ proceedings of the 1991 trieste summer school _ , edited by e. gava _
( world scientific , singapore , 1992 ) , p. 233
[ bartol research institute report no .
ba-91 - 76 ( 1991 ) ] ; h. arason , d.j .
castao , b.e .
keszthelyi , s. mikaelian , e.j .
piard , p. ramond , and b.d .
wright , phys . rev
. lett . * 67 * , 2933 ( 1991 ) ; s. kelley , j.l .
lopez , and d.v .
nanopoulos , phys .
b * 274 * , 387 ( 1992 ) ; m. drees and m.m .
nojiri , nucl .
* b369 * , 54 ( 1992 ) ; v. barger , m.s .
berger , and p. ohmann , phys.rev .
d * 47 * , 1093 ( 1993 ) ; b. ananthanarayan , g. lazarides , and q. shafi , phys .
b * 300 * , 245 ( 1993 ) ; m. carena , s. pokorski and c.e.m .
wagner , nucl .
phys . * b406 * 59 ( 1993 ) ; b. ananthanarayan and q. shafi , bartol research institute report no .
ba-93 - 25 ( 1993 ) ; and carena et al .
l. j. hall , r. rattazzi and u. sarid , lawrence berkeley laboratory report no
. lbl-33997 ( june 1993 , revised march 1994 ) , submitted to phys .
d. r. rattazzi and u. sarid , stanford university report no . su - itp-94 - 16 , in preparation .
t. banks , nucl .
* b303 * , 172 ( 1988 ) .
r. garisto and j.n .
ng , phys .
* b315 * , 372 ( 1993 ) ; m.a .
daz , phys .
* b322 * , 207 ( 1994 ) ; f.m .
borzumati , desy report no . desy 93 - 090 ( october 1993 ) .
e. thorndike , cleo collab . , talk given at the 1993 meeting of the american physical society , washington , d.c .
, april 1993 .
nelson and l. randall , phys .
* b316 * , 516 ( 1993 ) . f. zwirner , in _ physics and experiments with linear colliders _ , edited by r. orava , p. eerola and m. nordberg ( world scientific , singapore , 1992 ) , vol .
i , p. 309
[ cern theoretical report no .
cern - th.6357/91 ] , and references therein .
m. carena , m. olechowski , s. pokorski and c.e.m .
wagner , cern theoretical report no .
cern - th.7163/94 ( 1994 ) .
y. kawamura , h. murayama and m. yamaguchi , phys .
b324 * , 52 ( 1994 ) . j. lykken , this volume . | we analyze some consequences of grand unification of the third - generation yukawa couplings , in the context of the minimal supersymmetric standard model .
we address two issues : the prediction of the top quark mass , and the generation of the top - bottom mass hierarchy through a hierarchy of higgs vacuum expectation values .
the top mass is strongly dependent on a certain ratio of superpartner masses .
and the vev hierarchy always entails some tuning of the gut - scale parameters .
we study the rg equations and their semi - analytic solutions , which exhibit several interesting features , such as a focusing effect for a large yukawa coupling in the limit of certain symmetries and a correlation between the @xmath0 terms ( which contribute to @xmath1 ) and the gaugino masses .
this study shows that non - universal soft - susy - breaking masses are favored ( in particular for splitting the higgs doublets via d - terms and for allowing more natural scenarios of symmetry breaking ) , and hints at features desired in yukawa - unified models .
several phenomenological implications are also revealed .
-.3 in [ 7 ] may 1994su - itp-94 - 15 + ru-94 - 37 + hep - ph/9405313 .2 in * yukawa unification : the good , the bad and the ugly * .3 in riccardo rattazzi + _ department of physics and astronomy + rutgers university + piscataway , nj 08855 _ uri
sarid ( presenter ) + _ physics department + stanford university + stanford , california 94305 _ lawrence j. hall + _ theoretical physics group , 50a/3115 + lawrence berkeley laboratory + 1 cyclotron road + berkeley , california 94720 _ .2 in # 1#2 # 1.3ex |
the last several years have been marked by the start of full scale data taking of large neutrino and muon telescopes located at lake baikal ( nt-200 ) and in deep polar ice ( amanda ) .
two underwater telescopes antares and nestor assuming installation at greater depths are being intensively constructed in the mediterranean .
the possibility of deployment of telescopes with huge detecting volumes up to 1 km@xmath0 is also under wide investigation .
so , the knowledge of expected angular distribution of integral flux of atmospheric muons deep underwater is of interest not only for cosmic ray physics but also for the estimation of the possible background for neutrino detection and at last for a test of the correctness of underwater telescope data interpretation by using the natural flux of atmospheric muons as calibration source .
the last item frequently implies the estimation with an appropriate accuracy ( e.g. , better than 5@xmath1 for a given sea level spectrum ) the underwater integral muon flux for various sets of depths , cut - off energies and angular bins especially for telescopes of big spacial dimensions . up to now the presentation of the results of calculations of muon propagation through thick layers of water
both for parent muon sea level spectra ( especially for angular dependence taking into account the sphericity of atmosphere ) and for underwater angular flux has not been quite convenient when applied to concrete underwater arrays .
in addition , a part of numerical results is available only in data tables ( often insufficient for accurate interpolation ) and figures .
the possibility of direct implementation of monte carlo methods depends on the availability of corresponding codes and usually assumes rather long computations and accurate choice of the grid for simulation parameters to avoid big systematic errors .
therefore , there remains the necessity of analytical expressions for underwater muon integral flux .
in addition , the possibility of reconstructing the parameters of a sea level spectrum by fitting measured underwater flux in the case of their direct relation looks rather attractive . in this paper
we present rather simple method allowing one to analytically calculate the angular distribution of integral muon flux deep under water for cut off energies ( 1@xmath2gev and slant depths of ( 116)@xmath3 km for conventional ( @xmath4 ) sea level atmospheric muon spectra fitted by means of five parameters .
the fluctuations of muon energy losses are taken into account .
according to the approach of work @xcite the analytical expression for calculations of underwater angular integral flux above cut - off energy @xmath5 for a slant depth @xmath6 seen at vertical depth @xmath7 at zenith angle @xmath8 and allowing for the fluctuations of energy loss is based on the relation @xmath9 where correction factor @xmath10 is expressed , by definition , by the ratio of theoretical integral flux calculated in the continuous loss approximation to that calculated by exact monte carlo .
the analytical parametrization for @xmath10 is presented in refs .
the dependencies of the correction factor on @xmath5 and @xmath11 , calculated for any reasonable sea level spectrum represent the set of rather smooth curves .
the angular flux @xmath12 based on effective linear continuous energy losses @xmath13 having 2 slopes , is calculated by the following rule : @xmath14 here @xmath15 is the energy in the point of slope change from @xmath16 to @xmath17 and @xmath18 is the muon path from the energy @xmath15 till @xmath5 which is given by @xmath19 the formula for integral muon angular flux in the assumption of linear continuous energy losses is as follows : @xmath20 where subscript @xmath21 stands over both pion ( @xmath22 ) and kaon ( @xmath23 ) terms and @xmath24 when using expression ( [ ad3 ] ) for slant depths @xmath25 one must substitute @xmath26 and @xmath27 and use the values ( @xmath28 ) for a loss description . for slant depths
@xmath29 the use of ( [ ad3 ] ) remains unchangeable and the loss values are expressed by ( @xmath30 ) .
this algorithm may be extended to computations with any number of slopes of the energy losses .
the 5 parameters ( @xmath31 ) are those of the differential sea level muon spectrum , for which we use the following parametrization : @xmath32 where @xmath33 is a spectral index and @xmath34 have approximate sense of critical energies of pions and kaons for given zenith angle and @xmath35 are those for vertical direction .
the corresponding angular distrubution should be introduced using an analytical description of effective cosine @xmath36 taking into account the sphericity of atmosphere .
it should be noted that the description of underwater angular flux with the 5 parameters of a sea level spectrum gives the possibility of their direct best fit by using the experimental underwater distribution .
flux value in ( [ ad3 ] ) is expressed in units of ( @xmath37s@xmath38sr@xmath38 ) and all energies are in ( gev ) , slant depth @xmath11 in units of ( g@xmath3@xmath37 ) , loss terms @xmath39 and @xmath40 in units of @xmath41gev@xmath42g@xmath43 and @xmath44 , correspondingly . : 1 , 2 , 3 , 5 , 10 , and 20 , from bottom to top .
curve labels correspond to values of sec@xmath45 .
solid curves : spectrum based on data tables from @xcite , dotted curves : spectrum defined by eq .
( [ bkfit ] ) , dashed curves : spectrum of ref .
@xcite shown only for two values of sec@xmath45 : 1 and 20 .
[ fig : seal],width=313 ] for the description of effective linear continuous energy losses we use the following values of parameters when substituting in ( [ ad2 ] ) : ( @xmath46=2.67 , @xmath47=3.40 ) and ( @[email protected] , @xmath50=3.66 ) with @xmath15=35.3 tev . to examine the angular behaviour of a flux given by the formula ( [ ad1 ] ) by means of the comparison with numerical calculations we used the following parameters of the sea level muon spectrum : @xmath51 these values
have been chosen according to splines computed in this work via the data tables kindly given us by authors of ref .
when checking the values of fit spectrum for @xmath52=(0.051.0 ) we realized that the standard description of effective cosine ( with geometry of spherical atmosphere and with definite value of effective height of muon generation ) is not enough and one should introduce an additional correction @xmath53 leading to ( 1020)@xmath1 increase of effective cosine value for @xmath54 0.1 .
the reason of an appearing of this correction is that the concept of an effective generation height is approximate one .
it fails at large zenith angles where the real geometrical size of the generation region becomes very large .
we should note that our expression ( [ sealds1 ] ) for the sea level muon spectrum does not contain a contrubution from atmospheric prompt muons . according to the most recent calculations based on perturbative qcd ,
this contribution becomes essential only at @xmath55 gev .
the predictions of nonperturbative models are slightly more optimistic .
we plan to generalize our approach and include this contribution in our following paper .
incidentally , inclusion of prompt muons should be done in parallel with taking into account the steepening of the sea level muon spectrum due to the knee in the primary cosmic ray spectrum .
the resulting fit of angular sea level spectrum in units of ( @xmath37s@xmath38sr@xmath38ge@xmath56 ) is given by @xmath57 with modified effective cosine expressed by @xmath58 where @xmath36 is derived from spherical atmosphere geometry and is given by the polynomial fit : @xmath59 with the coefficients of the decomposition assembled in table [ tab : ectab ] .
the accuracy of ( [ efcosfit ] ) is much better than 0.3@xmath1 except the region @xmath52=(0.30.38 ) where it may reach the value of 0.7@xmath1 .
note that for @xmath60 0.4 the influence of the curvature of real atmosphere is less than 4 @xmath61 but for @xmath54 0.1 it is greater than 40 @xmath61 .
@xmath53 is the correction which is given for @xmath62 by @xmath63 fig .
[ fig : seal ] illustrates the limits of applicability of angular spectrum given by eq .
( [ bkfit ] ) , for energy and zenith angle variables .
the energy region , inside which the deviation from parent spectrum is less than 5 @xmath61 , is shifted from ( 0.3200)@xmath3tev for @xmath52=1.0 to ( 1.5300)@xmath3tev for @xmath52=0.05 .
the sea level spectrum given by ( [ bkfit ] ) is valid only below the knee ( @xmath64300 tev ) of primary cosmic ray spectrum .
correspondingly , for critical energies in expression ( 3 ) one should use @xmath65 instead of @xmath66 .
the examination of ( [ ad3 ] ) showed rather quick convergence of series s@xmath67 with increase of @xmath11 and @xmath5 .
therefore , for the accuracy of @xmath68 computation better than 0.1 @xmath61 it is quite enough to take only four first terms of this series ( up to @xmath69 ) for all values @xmath70 1 km and @xmath5 in ( 110@xmath71 ) gev . even using the two terms
leads to the accuracy of 1.3 @xmath61 for ( @xmath11=1.15 km , @xmath5=1 gev ) and @xmath720.5@xmath1 for ( @xmath732.5 km , @xmath741 gev ) .
four curves correspond to vertical depths @xmath7 : 1.15 km , 1.61 km , 2.0 km , and 3.0 km , from top to bottom .
solid curves result from numerical computations by using the sea level spectrum based on data tables from @xcite and mum code of muon propagation .
the dots are result of analytical calculation with expression ( [ ad1 ] ) by using the sea level spectrum ( [ bkfit ] ) .
[ fig : form],width=313 ] at vertical depth of 1.15 km .
the curves are given for twelve zenith angles @xmath52 : 1.0 , 0.9 , 0.8 , 0.7 , 0.6 , 0.5 , 0.4 , 0.3 , 0.2 , 0.15 , 0.1 , and 0.05 , from top to bottom . solid curves result from numerical differentiation of integral flux computed by using the sea level spectrum based on data tables from @xcite and mum code of muon propagation .
dotted curves result from numerical differentiation of the analytical expression ( [ ad1 ] ) based on the sea level spectrum ( [ bkfit ] ) .
[ fig : diff_ef],width=313 ] fig .
[ fig : form ] shows the comparison of underwater angular integral fluxes allowing for loss fluctuations at different basic depths @xmath7 ( of location of existing and planned telescopes ) calculated both numerically using mum code @xcite for parent sea level spectrum and analytically ( [ ad1 ] ) for the spectrum given by ( [ bkfit ] ) .
we realized that the error given by formula ( [ ad1 ] ) for all mentioned sea level spectra is within the corridor of @xmath75(46)@xmath1 for all cut - off energies @xmath5=(110@xmath76gev and slant depths @xmath11=(116)@xmath3 km ( corresponding angle is expressed by @xmath77 for a given vertical depth @xmath7 ) .
this is proved for @xmath7 in a range ( 13)@xmath3 km . for bigger cut - offs of @xmath5=(110)@xmath3tev
the corridor of errors is @xmath75(57)@xmath1 for @xmath11=(113)@xmath3 km .
note that for the sea level spectrum ( [ bkfit ] ) , just used for @xmath10 parametrization , the errors are smaller on 2@xmath1 .
the accuracy of the parametrization , used for the correction factor as a function of @xmath5 and slant depth @xmath11 is rather high and is about @xmath78 for all angles and kinds of the sea level spectrum ( assuming that the spectral index @xmath33 is approximately within ( 2.652.78 ) ) .
it results in the possibility to use it for an estimating numerically from various sea level spectra the value of an angular integral flux allowing for fluctuations of losses without direct monte carlo simulations .
note that the expression ( [ ad1 ] ) may be directly used for an ice after substitution @xmath79 , with @xmath80 being the ice density , and , with an additional error of @xmath81 , for a sea water . in spite of seeming complexity of the formulae ( [ ad1 ] ) , ( [ ad2 ] ) and ( [ ad3 ] )
they may be easily programmed . the validity of proposed formula up to cut - off energies 10 tev allows a calculation of underwater angular differential spectrum @xmath82 by means of numerical differentiation of expression ( [ ad1 ] ) .
it leads to rather appropriate results up to slant depths ( 1112)@xmath3 km .
we illustrate this in fig .
[ fig : diff_ef ] by comparison the underwater angular spectra calculated by numerical differentiation of integral flux computed by using the sea level spectrum based on data tables from @xcite and mum code of muon propagation , and numerical differentiation of the analytical expression ( [ ad1 ] ) based on the sea level spectrum ( [ bkfit ] ) . for a vertical depth
@xmath7=1.15 km it leads to errors @xmath83 for @xmath5=(20@xmath84gev for the angles corresponding to slant depths @xmath85 of ( 13)@xmath3 km and @xmath75(68)@xmath1 for @xmath5=(30@xmath86gev for the slant depths ( 312)@xmath3 km . even for @xmath11=23.2
km the result is still valid within @xmath87 but for the very narrow energy region @xmath5=(90300)@xmath3gev .
the analytical expression presented in this work allows to estimate for fluctuating losses the integral flux of atmospheric muons in pure water expected for different zenith angles , @xmath52=(0.051.0 ) , at various vertical depths at least of @xmath7=(13)@xmath3 km for different parametrizations of the sea level muon spectra .
the errors of this expression are estimated to be smaller than @xmath75(46)@xmath1 for cut - off energies ranged in @xmath5=(110@xmath76gev and slant depths in @xmath88=(116)@xmath3 km .
the main advantage of the presented formula consists in the possibility of the direct best fit of 5 parameters of parent sea level spectrum using angular distribution of underwater integral flux measured experimentally at a given vertical depth . the validity of this analytical expression with an accuracy of @xmath75(57)@xmath1 for @xmath5=(10@xmath010@xmath89)@xmath3gev and slant depths of ( 112)@xmath3 km gives also the possibility of estimation the angular underwater differential spectrum ( by means of numerical differentiation ) with error smaller than @xmath75(68)@xmath1 for energies of ( 305@xmath90)@xmath3gev .
the accuracy of the parametrization , used for the correction factor as a function of @xmath5 and slant depth @xmath11 is rather high and is about @xmath78 for all angles and kinds of the sea level spectrum ( assuming that the spectral index @xmath33 is approximately within ( 2.652.78 ) ) .
it results in the possibility to use it for an estimating numerically from various sea level spectra the value of an angular integral flux allowing for fluctuations of losses without direct monte carlo simulations .
99 s. i. klimushin , e. v. bugaev , and i. a. sokalski , hep - ph/0012032 .
s. i. klimushin , e. v. bugaev , and i. a. sokalski , these proceedings .
misaki _ et al .
_ , in _ proceedings of the 26th icrc _ , salt lake city , utah , 1999 , edited by d. kieda , m. salamon , and b. dingus .
* 2 * , p. 139 .
i. a. sokalski , e. v. bugaev , and s. i. klimushin , hep - ph/0010322 .
l. v. volkova , g. t. zatsepin , and l. a. kuzmichev , yad .
* 29 * , 1252 ( 1979 ) [ sov .
* 29 * , 645 ( 1979 ) ] . | the analytical expression for integral energy spectra and zenith angle distributions of atmospheric muons at large depths is derived .
fluctuations of muon energy losses are described using the parametrized correction factor .
the fitting formula for the sea level muon spectrum at different zenith angles for spherical atmosphere is proposed .
the concrete calculations for pure water are presented . |
there is significant interest in modelling shallow water flows at the mesoscopic level , either utilising the mesoscopic information or directly simulating at the mesoscopic scale . for the first approach , the gas - kinetic scheme ( gks )
was proposed @xcite , in which a boltzmann type equation with the * * b**hatnagar-**g**ross-**k**rook ( bgk ) is used as an intermediate step to supplement the flux term of shallow water equations ( swes ) @xcite . in the literature
, the gks has been applied for a range of flows including strong shock and dam - breaking problems @xcite . for the second direction ,
the main approach is the lattice boltzmann method ( lbm ) .
the lbm can be considered as a special case of discrete velocity method ( dvm ) for which a minimal discrete velocity set is sought and tied to the discretisation in space @xcite . in this way , the scheme is sufficiently simple yet powerful enough to handle complex flow problems .
hence , the scheme has also be extended to modelling of the swes@xcite .
a critical challenge for lbm is to simulate supercritical shallow water flows characterised by a high froude ( @xmath0 ) number . for this purpose ,
an asymmetric model is proposed in @xcite for simulating one - dimensional flows with @xmath1 . in @xcite ,
a multi - speed scheme has been proposed recently by matching hydrodynamic moments and applied for both one - dimensional and two - dimensional supercritical flows .
since the discrete velocities are not integers in general , a finite difference scheme has to be used so that the scheme is best classified as a discrete boltzmann model . in this work ,
we have derived a hierarchy of * * d**iscrete * * b**oltzmann * * m**odels with * * p**olynomial * * e**quilibria ( dbmpe ) from the continuous boltzmann - bgk type equation using the hermite expansion approach @xcite . from this framework
, we can have both integer and non - integer discrete velocity sets , and then we are able to choose either particle - hoping like scheme or general finite difference scheme .
the order of expansion may be easily tuned according to the requirement of the flow to be simulated .
if the depth scale is much less than the horizontal length scale , the water flows are dominated by the nearly horizontal motio and are referred as shallow water flows . for these flows ,
the swes can be employed to simplify the modelling , which are written as @xcite @xmath2 and @xmath3 where @xmath4\thickapprox h\nu\triangle\bm{v},\ ] ] and @xmath5 ( i.e. , @xmath6 ) is the unit tensor .
the equations describe the evolution of depth , @xmath7 , and depth - averaged velocity , @xmath8 .
the flows are often driven by a body force , @xmath9 , which mathematically represents not only the effects of an actual body force , including geostrophic force and tide - raising forces , but also those of wind stress , surface slope and atmospheric pressure gradient .
the pressure @xmath10 originates from the hydrostatic assumption , where @xmath11 represents for the gravitational force . if necessary , the viscous term @xmath12 may also be considered where @xmath13 is the depth averaged kinetic viscosity .
as has been shown , the swes can be mathematically analogous to the 2d compressible flow equations .
in fact , it may be considered to be describing an ideal gas flow with a gas state equation eq .
( [ eq : state ] ) , and its ratio of specific heats is @xmath14 and the `` sound speed '' is @xmath15 ( cf . chapter 2.1 in @xcite ) .
a boltzmann - bgk type equation can then be constructed for swes @xcite , @xmath16 where @xmath17.\label{eq : swefeq}\ ] ] at the mesoscopic scale , the evolutionary variable becomes the distribution function @xmath18 which represents the number of particles in the volume @xmath19 centred at position @xmath20 with velocities within @xmath21 around velocity @xmath22 at time @xmath23 .
the macroscopic quantities can be obtained by integrating over the whole particle velocity space , i.e. , @xmath24 @xmath25 and @xmath26 moreover , the * * r**ight-**h**and * * s**ide ( rhs ) term of eq .
( [ eq : bkgswe ] ) obeys the conservation property of the collision integral as shown in eqs .
( [ eq : h ] ) and ( [ eq : v ] ) .
similar to a real ideal gas , the relation between the relaxation time and kinetic viscosity is @xmath27 which can be obtained by using the chapman - enskog expansion . in principle , this kinetic equation may be solved directly by using regular numerical discretisation for both physical and particle velocity , i.e. , the so - called discrete velocity method . compared with direct discretisation of swes , there are two more degrees of freedom for the particle velocity , which need to be treated carefully for both accuracy and efficiency . among various schemes ,
the gauss type quadrature can provide a very efficient yet simple to implement discretisation if properly truncating the equilibrium function eq .
( [ eq : swefeq ] ) .
we shall derive a hierarchy of discrete boltzmann models using a hermite expansion @xcite .
however , it is convenient to first introduce a non - dimensional system , @xmath28 @xmath29 where the hat symbol denotes the non - dimensional variables . the reference depth , @xmath30 , can be the characteristic depth of the system such as the depth at the inlet ( e.g. , @xmath31 in fig .
[ fig : ill ] ) , while the reference viscosity is represented by @xmath32 and the reference pressure @xmath33 is chosen as @xmath34 . by using these non - dimensional variables , eqs .
( [ eq : bkgswe ] ) and ( [ eq : swefeq ] ) become , @xmath35 and @xmath36,\label{eq : nonswefeq}\ ] ] while there is no need to change the form of eqs .
( [ eq : h ] ) -([eq : p ] ) .
we may also define a `` knudsen '' number @xmath37 and the local froude number becomes @xmath38 where the velocity magnitude is denoted by @xmath39 . using the `` knudsen '' number and considering the fact that the viscosity is often considered as constant , the rhs term of eq.(@xmath40 ) becomes , @xmath41 hence
, the actual relaxation time will change with time locally .
hereinafter , we shall use the non - dimensional version of quantities and equations by default , and the hat symbol will be omitted for clarity .
it is also convenient to use the symbol @xmath42 to substitute for @xmath43 in writing the equations .
first of all , the equilibrium distribution will be expanded on the basis of the hermite orthogonal polynomials@xmath44 in particle velocity space ( see ref .
@xcite for detail ) , i.e. , @xmath45 where the @xmath46 order terms are retained and the coefficient is given yb @xmath47 using gauss - hermite quadrature with weights @xmath48 and abscissae @xmath49 , @xmath50 , the integration in eq .
( [ eq : coeff ] ) has been converted into a summation .
the first few coefficients are given by @xmath51 @xmath52 @xmath53,\ ] ] @xmath54,\ ] ] and @xmath55,\label{eq : a4}\ ] ] where tensor products such as @xmath56 stands for @xmath57 .
it is easy to verify that , with an appropriate quadrature , the conservation property of the collision integral will be satisfied automatically using an expansion higher than the first order , which will simplify the algorithm .
the body force term @xmath58 can also be approximated as , @xmath59 where @xmath60 is the corresponding coefficients for the distribution function , @xmath61 .
the first two are same the @xmath62 and @xmath63 while the higher order terms can be related to stress and heat flux . through the expansion , the kinetic equation ( [ eq : nondimeq ] )
can be rewritten in its truncated form , i.e. , @xmath64 thus , we will be solving an approximation of the original kinetic equation .
the second step is to discretise eq .
( [ truncatedeq ] ) in particle velocity space .
the gauss - hermite quadrature is the natural choice . in one dimension ,
the discrete velocities @xmath65 are just the roots of the hermite polynomials , and the corresponding weights are determined by : @xmath66^{2}}.\label{weight}\ ] ] given one - dimensional velocity sets , those of a higher - dimension can be constructed using the production formulae @xcite .
once the discrete velocity set is chosen , the governing equation is discretised as @xmath67 where @xmath68 , @xmath69 and @xmath70 . according to eqs .
( [ eq : a0 ] ) - ( [ eq : a4 ] ) and ( [ approxfeq ] ) , the explicit form of the fourth order , @xmath71 , is @xmath72\nonumber \\ & + & \underline{\frac{c_{i}v_{i}}{6}[(c_{i}v_{i})^{2}-3v_{i}v_{i}+3(h-1)(c_{i}c_{i}-4)]}\nonumber \\ & + & \underline{\underline{\frac{1}{24}[(c_{i}v_{i})^{4}-6(v_{i}c_{i})^{2}v_{j}v_{j}+3(v_{j}v_{j})^{2}]}}\label{eq : discretefeq}\\ & + & \underline{\underline{\frac{h-1}{4}[(c_{i}c_{i}-4)((v_{i}c_{i})^{2}-v_{i}v_{i})-2(v_{i}c_{i})^{2}]}}\nonumber \\ & + & \underline{\underline{\frac{(h-1)^{2}}{8}\left[(c_{i}c_{i})^{2}-8c_{i}c_{i}+8)\right]}}\nonumber \\ \},\nonumber \end{aligned}\ ] ] where the underlined terms are of third order , those double underlined are of fourth order , and the others consist of the zero order , the first order and the second order terms .
we have already constructed a kind of dbmpe . in order to conduct a numerical simulation ,
the physical space @xmath73 and the time @xmath23 are yet to be discretised .
for this purpose , any available scheme may be utilised according to the property of flow , such as an upwind scheme for space discretisation . in particular , if using abscissae consisting of integers , the lattice boltzmann scheme @xmath74+\frac{\tau\mathcal{f}_{\alpha}dt}{\tau+0.5dt}\label{eq : scheme}\end{aligned}\ ] ] can be constructed by introducing @xmath75 which allows a kind of particle - hopping like feature .
due to this unique property , the lattice boltzmann method has attracted significant interests in broad areas including shallow water simulations . however ,
if a high - order expansion is used , the abscissae will not be integer in general so that more sophisticated schemes will be required to solve the dbmpe .
for instance , we will use an implicit - explicit ( imex ) scheme below for testing .
we are now ready to discuss a few aspects on the accuracy of dbmpe .
as previously shown , errors are introduced in truncating the equilibrium function and discretising the particle velocity , i.e. , from eq .
( [ eq : nondimeq ] ) to eq .
( [ truncatedeq ] ) and from eq .
( [ truncatedeq ] ) to eq .
( [ lbgk ] ) . in principle , if the order of the hermite expansion is sufficiently high , eq .
( [ truncatedeq ] ) is expected to accurately recovery eq .
( [ eq : nondimeq ] ) . in practice , however , only a few orders may be affordable in term of computational cost , and the approximation accuracy will be determined by the expansion order . in general , the accuracy will be related to the froude number , i.e. , higher froude numbers may need more expansion terms , cf .
( [ eq : nonswefeq ] ) and eq .
( [ eq : fr ] ) .
for instance , similar to gas dynamics , see ref .
@xcite , using a first - order expansion leads to a linear equation which is suitable when @xmath76 .
specifically , we may introduce a @xmath77 norm @xmath78 + to measure the displacement of @xmath79 from @xmath80 over the whole particle velocity space . in fig .
[ fig : et ] , the density plots of @xmath81 are shown over the macroscopic velocity range of @xmath82 and @xmath83 . for comparison ,
we consider both the second - order expansion and fourth - order expansion of the equilibrium distribution function . from the plots
, we clearly see that the error grows with increasing macroscopic velocities . in particular , with a smaller @xmath7 ,
the error may become relatively large ( note that @xmath81 represents an accumulation of errors from all discrete velocities ) with relatively small macroscopic velocities .
hence , simulations may be more prone to error and instability if there is a sharp horizontal gradient of depth which often induces fast water flow .
interestingly , it appears that a high - order expansion does not necessarily means smaller errors ( see the plots for @xmath84 ) .
this shows complex convergence behaviour of the hermite expansion .
however , since the conservation laws are retained naturally , even with limited expansion order , the dbmpe may be sufficient for a broad range of the hydrodynamic problems , particularly if the diffusion term plays a negligible role .
the second kind of error is from the chosen quadrature .
according to the chapman - enskog expansion , when the `` knudsen '' number @xmath85 is small , the first - order asymptotic solution of eq .
( [ truncatedeq ] ) may be written as @xmath86 where the zeroth - order solution , @xmath87 , is just the truncated equilibrium function , @xmath79 . from eq .
( [ eq:1stsol ] ) , @xmath88 will include a polynomial of @xmath89 of one order higher than that in @xmath87 .
therefore , to numerically evaluate the integration of eqs .
( [ eq : h ] ) to ( [ eq : p ] ) , we need a quadrature with the degree of precision @xmath90 to calculate the @xmath91 order of moment ( e.g. , @xmath92 for the zeroth - order moment , @xmath7 ) . specifically , if using a @xmath93 order hermite expansion and to get the stress right ( @xmath94 ) , we need at least a quadrature with the @xmath95 order of precision , which may need four discrete velocities .
in the following , we test the scheme by simulating the transient evolution of an initial discontinuity of water level , i.e. , the classical 1d dam - breaking problem . by utilising eq .
( [ approxfeq ] ) , the hemite abscissae and eq .
( [ weight ] ) , it is straightforward to construct various discrete velocity sets .
for instance , to capture the strong shock with high froude number , we may need to use a fourth order expansion for @xmath71 and a quadrature of at least @xmath95 order of precision .
in contrast , if the flow is subcritical , then it may be enough to employ commonly used abscissae of nine points and a second order expansion for @xmath71 .
we will test both scenarios .
when the fourth expansion terms are employed for @xmath71 , we use a set of 16 discrete velocities obtained from the roots of fourth order hermite polynomial , see eq .
( [ weight ] ) .
these 16 discrete velocities are not integer and require a finite difference scheme . for this purpose
, we employ an imex scheme @xcite , which is particularly suitable if an solution is pursued at the euler - limit when @xmath42 is typically small , i.e. , @xmath96 where @xmath97 and @xmath98 the force term is ignored here since it is not involved in the test case . to cope with the strong discontinuity
, the flux term is discretised as ( cf .
@xcite ) @xmath99(\sigma_{\alpha , i}-\sigma_{\alpha , i-1 } ) \label{spacediscrete}\ ] ] where @xmath100 and the van leer slope limiter @xmath101 for simplicity , only the discretisation in the @xmath102 direction is given here . the choice of the subscript @xmath103 in eq .
( [ spacediscrete ] ) also depends on the the sign @xmath104 . if @xmath105 ( @xmath106 ) the information is acquired from the grid point @xmath107 ( @xmath108 ) .
the mesh size is denoted by @xmath109 and @xmath110 . in simulations
, we set the whole domain length to be @xmath111 in the non - dimensional system ( i.e. , @xmath112 in the dimensional system since the initial depth @xmath31 at the inlet is chosen as the reference length ) .
an initial discontinuity will be put at the middle @xmath113 ( see fig . [
fig : ill ] ) . for supercritical flow ,
the initial water depth is @xmath114 at the left half domain ( i.e. , the reference length scale ) and @xmath115 at the right half , while for the subcritical case , the ratio is @xmath116 . ,
scaledwidth=40.0% ] for supercritical flow , the parameters for calculation are chosen as @xmath117 the results at @xmath118 and @xmath119 are shown in fig .
[ fig:1000vs1 ] .
excellent agreement is found for both water depth and velocity with the analytical solution ( see ref .
@xcite for details of theanalytical solution ) , although numerical error is found for the profile of local froude number at the position of discontinuity due to the phase lag between water depth and velocity caused by numerical diffusion .
the simulations confirms the capability of modelling supercritical flows with a fourth order expansion and corresponding quadrature .
by contrast , we also tested the combination of a second - order expansion and 16 discrete velocities , but found that the simulation is unstable using the same problem setup .
for the subcritical flow , we use the stream - collision scheme ( [ eq : scheme ] ) and the parameters are chosen as @xmath120 while the stream - collision scheme is simpler , it requires a smaller time step for stable simulations due to the stiffness of the right hand collision term when @xmath85 is small .
moreover , since the spatial mesh is tied to the time step , this requirement may cause quadratic increase of computational cost in 2d .
the above parameters present an economic combination which results in very efficient simulations .
the numerical results are shown in fig .
[ fig:4vs1 ] , where excellent agreement with the analytical solutions @xcite can be found .
moreover , it appears that there is almost no numerical diffusion .
however , possibly due to the low diffusion , the scheme limits the maximum local froude number . using the same expansion order and quadrature
, the scheme ( [ eq : fluxlimiter ] ) can simulate higher froude number in general , in particular for supercritical flow .
to summarise , we have derived a hierarchy of discrete boltzmann models with polynomial equilibria for modelling shallow water flows . utilising the hermite expansion and gauss - hermite quadrature
, the scheme automatically satisfies the conservation property of the collision integral , which greatly simplifies the algorithm .
we also discussed theoretically how to choose the expansion order and quadrature according the requirement of problems ( e.g. , @xmath0 ) , although the hermite expansion shows complex convergence behaviour .
in particular , if the quadrature consists of integer abscissae , we can obtain the very simple and efficient lattice boltzmann scheme .
the derived models are tested using the classical one - dimensional dam - breaking problem , excellent agreement with analytical solutions is found for both supercritical and subcritical flows . in the future , we will further investigate the application for more complicated problems , e.g. , the treatment of force terms .
authors from daresbury laboratory would like to thank the engineering and physical science research council ( epsrc ) for their support of collaborative computational project 5 ( ccp5 ) and uk consortium on mesoscale engineering sciences ( ukcomes , grant no .
ep / l00030x/1 ) .
authors from sichuan university would like to thank the support of the national natural science foundation of china under grant numbers : 51409183 , 51579166 and 5151101425 .
18ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ] + 12$12
& 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) .
@noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) . | a hierarchy of discrete boltzmann model is proposed for simulating shallow water flows . by using the hermite expansion and gauss - hermite quadrature ,
the conservation laws are automatically satisfied without extra effort .
moreover , the expansion order and quadrature can be chosen flexibly according to the problem for striking the balance of accuracy and efficiency .
the models are then tested using the classical one - dimensional dam - breaking problem , and successes are found for both supercritical and subcritical flows . |
gravitational wave bursts@xcite are short ( less than one second ) transients of gravitational radiation with poorly known waveforms .
they are one of the distinct types of signals ligo@xcite and other interferometric and resonant mass detectors@xcite are pursuing .
a search for such signals is generally a time - frequency search , since time and frequency information of the candidate events are essential for a multi - detector coincidence analysis as well as for relating events to astrophysical sources .
one of the well established tools for such an analysis is the short time fourier transform ( also known as the windowed fourier transform ) developed by denis gabor in 1946 .
this approach applies standard fourier transforms to short windowed segments of the original time series to create a uniform time - frequency map with a time resolution and frequency resolution that depends upon the window s duration .
however , it is impossible to pick a window duration appropriate for all frequencies .
low frequency signals require long duration windows to accumulate sufficient frequency information , while high frequency signals are better isolated in time using short duration windows .
time and frequency localization of transient signals is therefore not easily achieved with the short time fourier transform .
here we have also assumed that burst waveforms are limited in @xmath0 ; that is , that the number of oscillations fall within a particular range regardless of frequency . in this introductory methods paper
we explore multiresolution techniques for the detection of transients that rely on the dyadic wavelet transform@xcite and the constant @xmath0 transform@xcite .
both methods search for a statistically significant excess of signal power in the time - frequency plane .
however , contrary to existing excess power searches@xcite , we implement a bank of filters that produce a logarithmic tiling of the time - frequency plane .
this provides increasing time resolution at higher frequencies and thus naturally addresses the time - frequency localization problem .
this paper outlines these new methods and is organized as follows .
section [ sec : bursts ] defines a representation independent parameterization of gravitational - wave bursts and defines an optimal signal to noise ratio , which motivates our use of a multiresolution approach .
section [ sec : wavelet ] presents the basics of the continuous and discrete wavelet transforms , while section [ sec : q ] presents the continuous and discrete @xmath0 transform as a modification of the short time fourier transform .
section [ sec : lpef ] demonstrates how linear prediction can be used to whiten background noise prior to transform analysis .
section [ sec : events ] describes methods for the selection of significant events from the output of the proposed algorithms , and section [ sec : efficiency ] presents preliminary detection efficiencies of our search algorithms applied to gaussian and sine - gaussian bursts injected into simulated ligo detector noise .
tuning and application of the proposed algorithms to search for a broader set of signals in data from the first ligo science runs is currently in progress .
to further motivate the development of multiresolution techniques , we first consider the parameterization and detectability of gravitational - wave bursts that are well localized in both time and frequency .
the more general case of non - localized bursts can be treated as a superposition of localized bursts .
an arbitrary gravitational wave burst may be represented in both the time - domain and the frequency - domain by the fourier transform pair @xmath1 and @xmath2 . for bursts that are square - integrable
, we may also define a representation independent characteristic squared amplitude , @xmath3 for bursts that are well localized in both time and frequency , it is then meaningful to define a central time , central frequency , duration , and bandwidth : @xmath4 it can be shown that the duration and bandwidth , when defined in this way , obey an uncertainty relation of the form , @xmath5 which we will also take as an approximate criterion for bursts that are well localized in the time - frequency plane .
finally , we define the dimensionless quality factor of a burst , @xmath6 which for well localized bursts , is simply a measure of the burst s aspect ratio in the time - frequency plane .
in general , a search algorithm for gravitational - wave bursts projects the data under test onto a basis constructed to span the space of plausible bursts .
the optimal measurement occurs when a member of this basis exactly matches a gravitational - wave burst . in this case
we achieve a signal to noise ratio , @xmath7 , which for narrowband bursts or a flat detector noise spectrum , is simply the ratio of the total energy content of the signal , @xmath8 , to the in - band one - sided power spectral density , @xmath9 , of the detector noise : @xmath10 thus , @xmath11 , which has units of strain per square root hz , is a convenient quantity for evaluating the detectability of narrowband bursts since it is directly comparable to the amplitude spectral density of detector noise .
it is important to note , however , that the above signal to noise ratio is only achieved when a member of the measurement basis closely matches the signal in the time - frequency plane .
otherwise , the measurement encompasses either too little signal or too much noise , resulting in a loss in the measured signal to noise ratio .
since bursts are naturally characterized by their @xmath0 , we therefore choose to tile the time frequency plane by selecting a measurement basis which directly targets bursts within a finite range of @xmath0 .
this naturally leads to a logarithmic tiling of the time frequency plane in which individual measurement pixels are well localized signals , all with the same @xmath0 .
in the preceding section , we motivate a search for bursts which projects the data under test onto a basis of well localized bursts of constant @xmath0 .
the wavelet transform@xcite by construction satisfies these requirements , and is defined by the integral , @xmath12 where the wavelet , @xmath13 , is a time - localized function of zero average .
the coefficients @xmath14 are evaluated continuously over times , @xmath15 , and scales , @xmath16 . our ability to resolve in time and frequency
is then determined by the properties @xmath13 assumes at each scale . at large scale
, @xmath13 is highly dilated yielding improved frequency resolution at the expense of time resolution . at small scale
, we achieve good time resolution with large uncertainty in frequency .
however , since the number of oscillations is fixed , the resulting @xmath0 is constant over all scales
. for the case of discrete data , a computationally efficient algorithm exists for calculating wavelet coefficients over scales that vary as powers of two : @xmath17 .
this is the dyadic wavelet transform , which can be implemented for a limited family of wavelets using conjugate mirror filters .
the filters consist of a high pass filter , @xmath18 , and low pass filter , @xmath19 , which can be applied in a cascade to obtain the wavelet coefficients . beginning with the original time series , @xmath20 , of length @xmath21 ,
two sequences of length @xmath22 are obtained by application of the high pass and low pass filters followed by down - sampling . the sequence of detail coefficients , @xmath23 , and approximation coefficients , @xmath24 ,
are defined at each level , @xmath25 , of the decomposition by @xmath26 the detail coefficients for scale @xmath16 , where @xmath27 , calculated in this manner are the same as the wavelet coefficients obtained from equation [ eqn : cwt ] .
if @xmath21 is a power of two , so that @xmath28 , the final approximation sequence will be @xmath29 .
the simplest dyadic wavelet is the haar function : @xmath30 the corresponding high pass and low pass filters are @xmath31 / \sqrt{2 } \hfill \mbox{and } \hfill \hat{l}^{\mbox{\scriptsize haar } } = [ + 1 , + 1 ] / \sqrt{2 } , \hspace{0.5in}\ ] ] from which we see that the detail coefficients are related to the differences of each pair of points in the parent series , while the approximation coefficients are related to the averages of each pair . a technique called wavelet packets can be used to further decompose the detail coefficients into smaller frequency subbands of higher @xmath0 .
the _ waveburst_@xcite search method currently implements a full wavelet packet decomposition which results in a uniform time - frequency plane with time - frequency cell sizes equivalent to those of the highest scale in the dyadic decomposition .
it is possible , however , to choose a different subset of the wavelet packet coefficients which approximate a more constant @xmath0 time - frequency plane by further decomposing the detail coefficents only a constant number of times at each scale .
this would give much more flexibility to detect high @xmath0 signals which do not match well to the mother wavelet @xmath13 .
the @xmath0 transform is a modification of the standard short time fourier transform in which the analysis window duration varies inversely with frequency .
it is similar in design to the continuous wavelet transform and also permits efficient computation for the case of discrete data .
however , since reconstruction of the data sequence is not a concern , we permit violation of the zero mean requirement .
in addition , the discrete @xmath0 transform is not constrained to frequencies which are related by powers of two .
we begin by projecting the sequence under test , @xmath32 , onto windowed complex sinusoids of frequency @xmath33 : @xmath34 here @xmath35 is a time - domain window centered on time @xmath36 with a duration that is inversely proportional to the frequency under consideration@xcite .
the @xmath0 transform may also be expressed in an alternative form@xcite , which allows for efficient computation .
thus , the @xmath0 transform at a specific frequency is obtained by a standard fourier transform of the original time series , a shift in frequency , multiplication by the appropriate frequency domain window function , and an inverse fourier transform .
the benefit of equation [ eqn : qtransform : alt ] is that the fourier transform of the original time series need only be computed once .
we then perform the inverse fourier transform only for the logarithmically spaced frequencies that we are interested in .
we may also adapt the @xmath0 transform to the case of discrete data . in this case ,
equation [ eqn : qtransform ] takes the form : @xmath38 = \sum_{n = 0}^{n - 1 } x[n ] w[n - m , k ] e^{-
n k / n}.\ ] ] as was the case for the continuous @xmath0 transform , we may express the discrete @xmath0 transform in the alternative form : @xmath39 = \sum_{l = 0}^{n - 1 } \tilde{x}[l + k ] \tilde{w}^{*}[l , k ] e^{+ i 2 \pi m l / n}.\ ] ] this allows us to take advantage of the computational efficiency of the fast fourier transform for both the initial transform as well as the subsequent inverse transforms for each desired value of @xmath40 . due to the implementation of the discrete @xmath0 transform in equation [ eqn : dqt : alt ] ,
the window function is most conveniently defined in the frequency domain . in particular , we choose a frequency domain hanning window , which has finite support in the frequency domain while still providing good time - frequency localization .
in this section , we present linear prediction@xcite as a tool for whitening gravitational - wave data prior to analysis by both the dyadic wavelet transform and the discrete @xmath0 transform . for both transforms , the statistical distribution of transform coefficients
is well known for the case of input white noise .
whitened data therefore permits a single predefined test for significance , which greatly simplifies the subsequent identification of candidate bursts .
we also note that linear predictive whitening is a necessary component of current cross - correlation@xcite and time - domain@xcite searches for gravitational - wave bursts in ligo data .
linear prediction assumes that the @xmath41 sample of a sequence is well modeled by a linear combination of the previous @xmath42 samples .
we thus define the predicted sequence , @xmath43 = \sum_{m=1}^{m } c[m ] x[n - m],\ ] ] and the corresponding prediction error sequence , @xmath44 = x[n ] - \hat{x}[n].\ ] ] the coefficients @xmath45 $ ] are then chosen to minimize the mean squared prediction error , @xmath46 |^2,\ ] ] over a representative training sequence of length @xmath21 . assuming that @xmath47 $ ] is a stationary stochastic process , this results in the well known yule - walker@xcite equations , @xmath48 c[m ] = r[k ] \qquad \mbox{for } \qquad 1 \leq k \leq m,\ ] ] where @xmath49 $ ] is the autocorrelation of the training sequence @xmath47 $ ] evaluated at lag @xmath40 . in practice , we choose the biased autocorrelation estimate , @xmath50 = \frac{1}{n } \sum_{n = |k|}^{n } x[n ] x[n - |k|],\ ] ] which guarantees the existence of a solution to the yule - walker equations . when this process is applied to gravitational - wave data , the resulting prediction error sequence is composed primarily of sample to sample uncorrelated white noise .
however , it also contains any unpredictable transient signals which were present in the original data sequence .
thus , the prediction error sequence is the whitened data sequence which we desire to subject to further analysis by the algorithms proposed in this paper .
we therefore define the linear predictor error filter as the @xmath51 order finite impulse response filter , which when applied to a data sequence , @xmath47 $ ] , returns the corresponding prediction error sequence , @xmath52 $ ] : @xmath53 = \sum_{m=0}^{m } b[m ] x[n - m].\ ] ] the choice of the filter order , @xmath42 , is then determined by the measurement resolution of subsequent analysis . for a given filter order , the resulting prediction error sequence will be uncorrelated on time scales shorter than @xmath42 samples . in the frequency domain
, this corresponds to a spectrum which is white on frequency scales greater than a characteristic bandwidth , @xmath54 where @xmath55 is the sample frequency of the data .
thus , we can whiten to any desired frequency resolution by selecting the appropriate filter order , @xmath42 . in practice
, we simply choose the minimum filter order which is sufficient for the subsequent analysis .
for the training length , @xmath21 , we choose to train over times which are long compared to the typical gravitational - wave bursts we are searching for . [ 0.8 ] .
in addition , the right hand plot shows the effect of the same linear prediction error filter when applied to a simulated gravitational - wave burst injected into the data . in this case
, we inject a 235 hz sine - gaussian with a @xmath0 of 12.7 and signal to noise ratio of 20 as described in section [ sec : efficiency ] .
note the absence of a time delay between the whitened and unwhitened burst.,title="fig:",width=240 ] [ 0.8 ] .
in addition , the right hand plot shows the effect of the same linear prediction error filter when applied to a simulated gravitational - wave burst injected into the data . in this case
, we inject a 235 hz sine - gaussian with a @xmath0 of 12.7 and signal to noise ratio of 20 as described in section [ sec : efficiency ] .
note the absence of a time delay between the whitened and unwhitened burst.,title="fig:",width=240 ] in order to avoid timing errors due to the unspecified phase response of linear prediction error filters , we construct a new filter from the unnormalized autocorrelation of the original filter coefficients : @xmath56 = \sum_{m = |k|}^{m } b[m ] b[m - |k|].\ ] ] while this new filter has the desired property of zero phase response@xcite , its magnitude response has been squared . in practice ,
the construction of the zero phase filter is carried out in the frequency domain , where it is also possible to take the square root of the resulting frequency domain filter before transforming back to the time domain .
an example application of zero - phase linear prediction to sample gravitational - wave data and a simulated gravitational - wave burst is shown in figure [ fig : lpef ] .
we now seek to identify statistically significant events in the resulting dyadic wavelet transform and discrete @xmath0 transform coefficients .
we assume that linear predictive whitening ( section [ sec : lpef ] ) has been applied to the data prior to the dyadic wavelet transform . then , according to the central limit theorem , for a sufficiently large scale , @xmath25 , the wavelet coefficients , @xmath23 , within the scale will approach a zero - mean gaussian distribution with standard deviation @xmath57 .
we therefore define the sequence of squared normalized coefficients , or normalized pixel energies at scale @xmath58 , @xmath59 which is chi - squared distributed with one degree of freedom ( figure [ fig : histogram ] ) .
it is then simply a matter of thresholding on the energy of individual pixels , @xmath60 , to identify statistical outliers .
we may also choose to cluster nearby pixels to better detect bursts which deviate from the dyadic wavelet tiling of the time - frequency plane .
for a randomly selected cluster @xmath61 of @xmath21 pixels , we define the total cluster energy , @xmath62 which is also chi - square distributed , but with @xmath21 degrees of freedom .
it is therefore possible to select a threshold energy to achieve a desired white noise false event rate . alternatively
, sharp transients may also be detected by simply searching for vertical clusters of pixels within the top @xmath63 percent of pixel energies in each scale .
[ 0.8 ] transform ( right ) pixel energies in the scale and frequency bin encompassing 600 hz for the data of figure [ fig : lpef ] .
the distributions before ( dashed ) and after ( solid ) linear predictive whitening are compared to the theoretically expected distributions for ideal white noise ( shaded ) .
the whitened data show good agreement with theory except at high pixel energy , where the unpredictable non - stationary behavior ( glitches ) of the gravitational - wave data becomes apparent . for the discrete @xmath0 transform
, the inconsistency of the unwhitened data with theoretical white noise is due to the coherent 600 hz power line harmonic , which results in a rician rather than exponential distribution of pixel energies .
for the discrete haar wavelet transform , the presence of a coherent signal results in an apparent , but false , white noise distribution .
this is due to the use of a real valued wavelet which is sensitive to the relative phase between the signal and wavelet .
this effect actually masks true non - stationarities , which are only exposed after linear predictive whitening removes the coherent signal content.,title="fig:",width=240 ] [ 0.8 ] transform ( right ) pixel energies in the scale and frequency bin encompassing 600 hz for the data of figure [ fig : lpef ] .
the distributions before ( dashed ) and after ( solid ) linear predictive whitening are compared to the theoretically expected distributions for ideal white noise ( shaded ) .
the whitened data show good agreement with theory except at high pixel energy , where the unpredictable non - stationary behavior ( glitches ) of the gravitational - wave data becomes apparent . for the discrete @xmath0 transform ,
the inconsistency of the unwhitened data with theoretical white noise is due to the coherent 600 hz power line harmonic , which results in a rician rather than exponential distribution of pixel energies . for the discrete haar wavelet transform , the presence of a coherent signal results in an apparent , but false , white noise distribution
this is due to the use of a real valued wavelet which is sensitive to the relative phase between the signal and wavelet .
this effect actually masks true non - stationarities , which are only exposed after linear predictive whitening removes the coherent signal content.,title="fig:",width=240 ] for the discrete @xmath0 transform , we also make use of the central limit theorem by assuming that the data has first been whitened using linear prediction . for pixels of sufficiently long duration , the pixel energies , @xmath64 = | x[m , k ] |^2,\ ] ] at a particular frequency , @xmath40 ,
will then approach an exponential distribution ( figure [ fig : histogram ] ) .
thus for each pixel , we have a measure of significance : @xmath65 ) = \exp \left ( -\epsilon[m , k ] / \mu_k \right),\ ] ] where @xmath66 is the mean pixel energy at frequency @xmath40 . as with the discrete wavelet transform
, we may also consider clusters of @xmath21 pixels and select a threshold energy in order to achieve a desired white noise false event rate .
however in this case , the cluster energy is chi - square distributed with @xmath67 degrees of freedom . to efficiently detect bursts over a range of @xmath0 , we perform multiple @xmath0 transforms , each of which produces a time - frequency plane tiled for a particular @xmath0 within our range of interest . to best estimate the parameters of candidate bursts , we then select the most significant non - overlapping pixels among all @xmath0 planes using a simple exclusion algorithm .
pixels are considered in decreasing order of significance and any pixel which overlaps in time or frequency with a more significant pixel is discarded .
thus , for each localized candidate burst , only the single pixel which best represents the burst parameters is reported .
this does not exclude the possibility of clustering pixels from non - localized bursts , since the pixels which pass both the initial significance threshold as well as the exclusion algorithm will represent the strong localized features of such bursts . reducing a non - localized burst to its strong , localized features has the additional benefit of a much tighter time - frequency coincidence criterion when multiple detectors are used .
finally , we evaluate the detection efficiencies of the proposed algorithms over simulated ligo noise . the simulated data are composed of line sources and gaussian distributed white noise passed through shaping filters to produce a power spectrum which matches the typical h1 detector noise spectrum during the second ligo science run@xcite .
assuming optimal orientation , we evaluate the detection efficiency for each waveform at various @xmath11 by injecting 256 such bursts at random times into the simulated data ( see figure 3 ) .
the simulated data are first filtered by a 6th order 64 hz butterworth high pass filter followed by a linear prediction error filter with whitening resolution of 16 hz .
however , we do not model the non - stationary behavior of real ligo noise , which largely determines the false event rate of existing search algorithms . the algorithms considered in this paper
are equally applicable to the identification of statistically significant transient events in auxiliary interferometer or environmental data .
so , they may also prove useful for identifying potential vetoes . transform ( right ) .
the detection efficiency of the dyadic wavelet transform is shown for three durations of gaussian bursts with @xmath68 s of 0.35 , 0.71 , and 1.41 milliseconds as shown from left to right .
the detection efficiency of the @xmath0 transform is shown for a sine - gaussian burst with a @xmath0 of 12.7 and central frequency , @xmath69 , of 275 hz.,title="fig:",width=240 ] transform ( right ) .
the detection efficiency of the dyadic wavelet transform is shown for three durations of gaussian bursts with @xmath68 s of 0.35 , 0.71 , and 1.41 milliseconds as shown from left to right .
the detection efficiency of the @xmath0 transform is shown for a sine - gaussian burst with a @xmath0 of 12.7 and central frequency , @xmath69 , of 275 hz.,title="fig:",width=240 ] the low @xmath0 haar wavelet is best suited to broadband signals , and for a representative broadband waveform , we use gaussian injections of the form @xmath70 , for @xmath68 s of 0.35 , 0.71 , and 1.41 ms .
the filtered data are passed through the discrete wavelet transform and event selection algorithm as described in section [ sec : events : wavelet ] for a false event rate of 0.5 hz .
we allow clusters to be composed of a single pixel , or a combination of two pixels adjacent or overlapping in time and adjacent in scale for scales , @xmath25 , of 5 , 6 , and 7 . for @xmath68 s of 0.35 , 0.71 , and 1.41
ms , we obtain 50% efficiencies for @xmath11 equal to @xmath71 , @xmath72 , and @xmath73 hz@xmath74 , which correspond to signal to noise ratios , @xmath7 , of 3.6 , 3.7 , and 3.5 .
since the discrete @xmath0 transform targets well localized bursts , we demonstrate its efficiency for sine - gaussian bursts of the form , @xmath75 , with a @xmath0 of 12.7 and central frequency , @xmath69 , of 275 hz .
the discrete @xmath0 transform is applied to the whitened data with approximately 75 percent pixel overlap to search for single pixel bursts within the frequency band 644096 hz and with @xmath0 s in the vicinity of 10.6 , 14.1 , and 17.7 .
candidate events are then identified as described in section [ sec : events : dqt ] in order to obtain a 1 hz white noise false rate .
the resulting 50% detection efficiency occurs at an @xmath11 of @xmath76 hz@xmath74 , corresponding to a @xmath7 of 3.0 .
we have described two transient - finding algorithms using logarithmic tiling of the time - frequency plane and examined their applicability to the detection of gravitational wave bursts .
the methods work more efficiently when data are whitened and for this we have introduced a data conditioning algorithm based on linear prediction . as a first step in quantifying their performance , we have used simulated noise data according to the ligo noise spectra from its second science run .
moreover , _ ad hoc _ waveform morphologies were used to simulate the presence of a burst signal over the noise .
this allowed us to measure the algorithms efficiency at a fixed false alarm rate of o(1hz ) .
the results are encouraging and we are continuing the computation for mapping the efficiency of the algorithms as a function of their false alarm rate .
needless to say , even if the shaped ligo noise with some line features that we used for our simulation is a step forward , it does remain unrealistic .
real interferometric data have richer line structure , non - stationary and non - gaussian character that may be far from our assumptions ; at the same time though , it is extremely hard to simulate .
all of these make the perfomance estimates that we made using our simulation model optimistic . the use though of mathematically well described noise allows the straightforward reproduction and comparison of our results with other methods .
an end - to - end pipeline using the aforementioned methods for whitening and transient - finding on data collected by the ligo instruments is currently in development .
this will provide the ultimate test of the performance of the algorithms and will be reported separately within the context of the ligo scientific collaboration ( lsc ) data analysis ongoing research .
this work was supported by the us national science foundation under cooperative agreement no . | we present two search algorithms that implement logarithmic tiling of the time - frequency plane in order to efficiently detect astrophysically unmodeled bursts of gravitational radiation .
the first is a straightforward application of the dyadic wavelet transform .
the second is a modification of the windowed fourier transform which tiles the time - frequency plane for a specific @xmath0 .
in addition , we also demonstrate adaptive whitening by linear prediction , which greatly simplifies our statistical analysis .
this is a methodology paper that aims to describe the techniques for identifying significant events as well as the necessary pre - processing that is required in order to improve their performance .
for this reason we use simulated ligo noise in order to illustrate the methods and to present their preliminary performance . |
observations of hard x - ray ( hxr ) bursts with the ramaty high energy solar spectroscopic imager ( rhessi ) have produced a number of new properties of high energy electrons accelerated during solar flares .
one of its recent discoveries is a distinctive type of flares in which the bulk of the hxr emission is from the coronal part of the loop ( veronig & brown 2004 ; sui et al .
2004 ; krucker et al .
2008 ) . for such sources ,
the corona can be considered as not only the site of particle acceleration , but act as a thick target , stopping the accelerated electrons before they can penetrate to the chromosphere .
this is in contrast with the traditional idea of the collisional thick target model ( cttm ) that was believed to apply only to the dense chromosphere .
veronig & brown ( 2004 ) studied two events for which the coronal densities of the flaring loops were found sufficiently high to justify the cttm .
xu et al . (
2008 ) showed that for a certain group of events the observed coronal hxr source extent varies with photon energy as expected from a cttm with an extended acceleration region in the coronal loop .
the result not only serves as evidence for the coronal cttm , but also allows estimates of the plasma density and longitudinal extent of the acceleration region .
guo et al .
( 2012a ) have carried out a similar work but using the electron flux images instead of photon maps , as constructed by regularized spectral inversion of the visibility data in the count domain ( piana et al .
guo et al .
( 2012b ) further related those interpretations to the number of particles within the acceleration region , the filling factor , and the specific acceleration rate in units of electron per second and per ambient electron .
according to these studies those events with the coronal thick target hxr sources form an important ( though small ) class of solar flares for which acceleration parameters can directly be inferred .
if we plan to test the cttm for coronal hxr sources using other radiations , microwave radiation should be the first choice , for its close relationship with hxrs .
time profiles of microwave emissions tend to show a peak - to - peak correlation with hxr lightcurves , which led to the idea of the same population of electrons responsible for both radiations ( e.g. , dulk 1985 ) .
spatially both emissions should arise , at least , from the same magnetic loop or loop systems , but the location of each radiation maximum may differ along the loop as their radiation efficiencies differently respond to plasma parameters and magnetic field ( e.g. , sakao 1996 ) .
we expect that microwave spectrum will be particularly important in relation to the studies of the coronal cttm .
the main microwave radiation mechanism , gyrosynchrotron radiation , has primarily served as a diagnostic for magnetic field and electron distribution .
however , when thermal density is very high , two spectral features sensitive to density , the razin suppression ( razin 1960 ) and free - free emission , become prominent .
the razin suppression refers to less efficient radiation in the presence of ambient medium compared with the radiation in vacuum , and the razin frequency below which the effect is significant is given by @xmath7 hz where @xmath8 is electron density in units of @xmath1 and @xmath9 is the transverse magnetic field component in gauss .
the razin suppression will thus be effective for the condition of thick target looptop hxr sources .
the well - known free - free opacity is proportional to @xmath10 , with @xmath11 , the plasma temperature .
in addition , there is an _ absolute _
cut - off of microwave radiation at the plasma frequency which is a sole function of plasma density ( see , e.g. , dulk 1985 , melrose 1980 ) . with these features that become prominent at high density , a microwave spectrum should play a critical role in validating the cttm applied to coronal sources . in this letter
we compare the properties of a coronal hxr source with those of microwave source detected in an event which we consider a candidate for the coronal thick target source .
the event is the 2002 september 9 flare occurred in noaa ar 0105 and acquired goes soft x - ray class m2.3 .
this event was observed by the reuven ramaty high - energy solar spectroscopic imager ( rhessi ) and the owens valley solar array ( ovsa ) , respectively , and is one of the rare events for which spatial and spectral information at both wavelengths is available for this type of study .
in figure 1 we show the rhessi and ovsa spectra during the 2002 september 9 flare .
the top panel shows the spatially integrated rhessi spectrum at a single time corresponding to maximum emission measure obtained using ospex in the solarsoftware ( ssw ) .
ji et al . ( 2004 ) carried out the imaging spectroscopy for the same event . in their result and also in ours ( fig .
2 ) , the rhessi source appears to be a single source and the spectra obtained through ospex are indeed similar to those obtained from the imaging spectroscopy . as a characteristic feature , the spectrum shows steeply decreasing flux with increasing photon energy with spectral indices as high as 7@xmath128 .
such high index is typical for the events classified for the coronal thick target hxr sources ( guo et al .
2012ab , xu et al . 2007 ) .
the spectral index @xmath13 will later be used to provide the information on electron energy distribution to the modeling of the hxr source loop and that of the microwave spectra . in spite of such a soft nonthermal component , a thermal component could still be identified and the maximum temperature is found to be @xmath14k at around 17:50 ut , close to what we infer from goes soft x - ray data .
the bottom panel shows the total power microwave spectra for the 2002 september 9 flare at three selected times .
this is one of the calibrated microwave spectra in the period of 2001 - 2003 that can be downloaded from the ovsa web site . like the hxr spectra ,
these microwave spectra are also spatially integrated and include both r and l polarizations added together .
they appear in a typical shape , but with rapidly falling fluxes toward decreasing frequencies , which can be a sign for the razin suppression , and is suggestive of a high coronal density .
the high frequency spectrum is steeply decreasing with increasing frequency , which is qualitatively consistent with the high photon spectral index of the rhessi spectrum as found above .
we may also note that the peak frequency of microwave spectrum does not change much in the course of the flare .
previously , belkora ( 2007 ) claimed that such a feature may indicate the effect of the razin suppression .
it is thus qualitatively arguable that the spectral morphology of the ovsa spectra is supportive of the high coronal density required for the cttm of the coronal hxr source .
the ovsa imaging procedure is less known compared with that of the rhessi . since the ovsa uses many frequencies , all night observations of a quasar point source such as 3c84
are made several times a year , and are used to calibrate solar data obtained at nearby times .
daily calibration observations at selected frequencies are also made to provide corrections to the above calibration .
the calibrated visibilities are fed into the _ wimagr _ program available in the ssw , after which the imaging can be performed in the procedure very similar to that of the astronomical image processing system ( aips ) .
figure 2 shows hxr and microwave maps as contours on top of an h@xmath15 off - band blue - wing image obtained from the big bear solar observatory ( bbso ) .
based on the steep hxr spectrum and the noise level ( fig .
1 ) , the hxr imaging was performed below 50 kev . within the energy range only single loop - top source was found .
microwave sources generally vary with frequency more significantly , but the centroid of 3.0 ghz microwave source is co - spatial with that of the 2550 kev hxr source as shown in this figure . both the hxr and microwave sources appear as a single source lying between the two h@xmath15 kernels , and their morphologies change only a little with time ( see , for more details on the morphology of this event , ji et al .
2004 and lee et al .
we thus regard the hxr source as a loop - top coronal thick target source ( cf .
veronig & brown 2004 ) .
we followed the technique of xu et al .
( 2007 ) to determine the source sizes using the vis - forward fitting and made a model fit to the observed hxr source size as a function of photon energy . as for the model , we actually used equation ( 5 ) of guo et al .
( 2012b ) , which corresponds to xu et al.s ( 2007 ) model with an extended acceleration region of length , @xmath16 , uniformly filled with dense material of density @xmath17 .
this model also includes the electron energy distribution with a power law index for which we use @xmath18 under the thick target bremsstrahlung approximation ( brown 1971 ) .
figure 3 shows our fitting of the model to the observed hxr source size ( symbols ) at seven photon energies by varying @xmath19 and @xmath20 .
we obtained @xmath21 @xmath1 and @xmath22 from these data .
we calculated a set of theoretical microwave spectra for comparison with the ovsa spectrum using the numerical code developed by fleishman and kuznetsov ( 2010 ) , which includes full treatment of the razin effect . in figure 4
we show our search for the best set of model parameters with which we can reproduce the ovsa spectrum at the time of the maximum density ( @xmath2317:50 ut ) chosen as a representative spectrum . in the top panel
we first tried a modest density @xmath24 @xmath1 and the same electron energy distribution derived from the rhessi spectrum , i.e. , @xmath25 , except that the electron energy distribution is assumed to extend up to 1 mev .
this model ( shown as the dotted line in the top panel ) reproduces the high frequency spectrum to some extent , but fails to match the low frequency spectrum .
if a homogeneous model predicted a more rapidly increasing spectrum at low frequencies than observed , we may suspect that the disagreement is due to source inhomogeneity in reality ( e.g. , lee et al .
1994 ) . in this case
, however , the model predicts a more extended spectrum toward lower frequencies than observed .
we thus adjusted other factors maintaining the homogeneous source assumption . with a higher densities
@xmath26 @xmath1 the model ( solid line ) reproduces the low frequency slope under the enhanced razin suppression .
it closely fits to the high frequency spectrum as well , except somewhat excessive fluxes are predicted at the highest observed frequencies .
if we further raise the density as desired for supporting the thick target hxr source , we end up with a disagreement between the model ( dashed line ) and observation at both low and high frequencies , unfortunately . at this point
we should clarify one basic limitation encountered in the current modeling . even though we will further adjust other parameters to reproduce the microwave spectrum more closely , the density requirement for this microwave spectrum is rather strict .
that is , the minimum cut - off frequency for microwave radiation is the local plasma frequency , a sole function of density .
more specifically , the o - mode is cut - off at local plasma frequency and x - mode , at a higher frequency that depends on magnetic field as well ( melrose 1980 ) .
since this spectrum starts above 3 ghz , we derive @xmath27 @xmath1 .
admittedly this density is a few factor lower than the one needed for explaining the hxr source dimension varying with energy under the cttm , and this limit value of density can not be adjusted by change of other parameters .
we now proceed with the ` lowered ' target value of density .
since the excessive fluxes at high frequencies are due to free - free emission , we adopted higher temperatures as way of suppressing the free - free emission .
the three models shown in the middle panel inherit the same parameters from those in the top panel except the temperature increased to @xmath28 k. at this high temperature , the high frequency fluxes are significantly _ reduced _ , and all models agree to the observed spectrum at high frequencies .
we may further increase temperature without loosing this success , but too a high temperature is unwanted in a physical point of view , because the medium should be a cold - target for coulomb collisions .
emslie ( 2003 ) quantified the condition for cold target as @xmath29 where @xmath30 is the electron energy and @xmath11 is the target temperature .
kosugi et al .
( 1988 ) have shown that the electrons producing photons at @xmath31 have an energy in the range , @xmath32 . since we concern ourselves with the rhessi spectrum above 6 kev , the temperature @xmath28 k or 2 kev may qualify as cold target .
however , @xmath11 can not be much higher than this , especially considering that this photon spectrum is very soft and @xmath30 could be close to @xmath31 .
for instance , @xmath33 gives @xmath34 k for cold target . in
the above , change of the temperature helped to fit the high frequency spectrum , but now these models no longer fit the low frequency spectrum . to come up with this problem ,
we adjust the magnetic field , @xmath35 . in general , increasing ( decreasing ) @xmath35 results in a shift of the spectrum to higher ( lower ) frequencies . two models calculated with @xmath36 gauss and 350 gauss
are respectively shown as the dotted and dashed lines in the bottom panel .
each individual model spectrum appears to be narrower than the observed spectrum .
the low ( high ) field strength model could , however , reproduce the low ( high ) frequency spectrum .
we thus created an average model ( solid line ) of these two by linearly summing the flux from each model , which well agrees to the observed spectrum in overall .
we thus conclude that the source must have an inhomogeneous magnetic field distribution in the range of 250 gauss @xmath37 350 gauss .
the viewing angle of magnetic field , @xmath38 , is another variable and this should be high .
otherwise , the free - free emission tends to compete with the gyrosynchrotron radiation , which is unwanted in this case .
an extensive exercise shows that a high viewing angle is preferred , and we have fixed @xmath38 to 75@xmath39 throughout this modeling .
all these calculations were made with the isotropic pitch angle distribution of electrons .
if it were significantly anisotropic , the resulting spectral morphology may differ from the present result even for the same electron energy distribution . in that case , a few more assumptions regarding the electron energy distribution will be needed , which we do not regard essential for the current study .
we have studied the hxr and microwave data of the 2002 september 9 flare with a goal of testing the coronal thick target hypothesis for hxr sources based on the accompanying microwave spectra .
our study found that the microwave spectrum provides strong constraints on the thick target interpretation of hxr sources .
first of all , the lowest frequency of the microwave spectrum is strictly determined by the local plasma density , and the derived density @xmath40 @xmath1 is close to , but 5 times lower than the density found from the energy dependent variation of the coronal hxr source .
not only the density but temperature @xmath52 kev and magnetic field strength , 250350 gauss , are also required which will again be constrained by the condition for a cold - target for bremsstrahlung and the measurement of the photospheric magnetic field , respectively
. the next question will be whether these constraints could be alleviated by introducing a more realistic , inhomogeneous density structure instead of the simplified assumption for a uniform density source .
we commonly consider a source of dense and tenuous regions mixed , i.e. , filamentary structure , in which case the tenuous regions filled with nonthermal electrons work as microwave sources and the dense regions , the thick target hxr sources .
this scenario is not necessarily promising because microwaves should propagate through the dense plasma and can not avoid the highly enhanced free - free opacity .
the next conceivable structure is an inhomogeneous loop along its length such that a section in loop - top is tenuous , and the rest is filled with dense materials working as a thick target to coulomb collisions ( xu et al .
this model is actually favorable for explaining the microwave source , but has to be disputed because it is in apparent conflict with the hxr source lying in the looptop . in our current idea ,
the only plausible case would be a somewhat arbitrary coaxial loop structure which consists of the inner loop filled with dense colder plasma and the outer one with tenuous and hot plasma , so that the former is ideal for the thick target hxr source and the latter , more favorable for the microwave emission . in this case
the small disagreement between the densities derived from microwaves and hxrs , respectively , could be made acceptable .
this scenario would , however , need an explanation as to why the inner core has dense and colder materials compared with the outer part of the loop . in summary ,
microwave spectrum provides unique constraints on the coronal hxr sources through the cut - off frequency , the cold target temperature , the free - free opacity , and the razin effect .
these constraints are related to the fundamental physics and are thus robust .
microwave spectral observations should therefore be regarded as a new asset for validating those events of the coronal hxr sources and for determining additional source conditions for them that were not available from hxr observations alone .
future studies may be benefitted from the use of microwave polarizations and time profiles that will further enhance its diagnostic capabilities .
we thank _ rhessi _ team for excellent data set .
this work was supported by wcu program through nrf funded by mest of korea ( r31 - 10016 ) .
j.l . was supported by the international scholarship of kyung hee university , and also by nasa grants , nnx11ab49 g and nas5 - 01072 ( na03 ) 937836 to njit .
g.s.c . was supported by the korea research foundation grant funded by the korean government ( krf-2007 - 313-c00324 ) .
m. j. expresses his thanks to kyung hee university for granting him a sabbatical , during which his part of this research was performed . | recently a distinctive class of hard x ray ( hxr ) sources located in the corona was found , which implies that the collisionally thick target model ( cttm ) applies even to the corona .
we investigated whether this idea can independently be verified by microwave radiations that have been known as the best companion to hxrs .
the study is made for the goes m2.3 class flare occurred on 2002 september 9 that were observed by the reuven ramaty high energy solar spectroscopic imager ( rhessi ) and the owens valley solar array ( ovsa ) . interpreting the observed energy dependent variation of hxr source size under the cttm the coronal density should be as high as @xmath0 @xmath1 over the distance up to 12@xmath2 . to explain the cut - off feature of microwave spectrum at 3 ghz , we however , need density no higher than @xmath3 @xmath1 .
additional constraints need to be placed on temperature and magnetic field of the coronal source in order to reproduce the microwave spectrum as a whole .
firstly , a spectral feature called the razin suppression requires the magnetic field in a range of 250350 gauss along with high viewing angles around 75@xmath4 .
secondly , to avoid excess fluxes at high frequencies due to the free - free emission that were not observed , we need a high temperature @xmath52@xmath6 k. these two microwave spectral features , razzin suppression and free - free emissions , become more significant at regions of high thermal plasma density and are essential for validating and for determining additional parameters for the coronal hxr sources . |
a network of ground - based interferometric gravitational wave ( gw ) detectors ( ligo / virgo / geo/@xmath0 ) is currently taking data near its planned sensitivity . coalescing black hole binaries
are among the most promising , and most exciting , gw sources for these detectors . in order to successfully detect gws from coalescing black hole binaries , and to be able to reliably measure the physical parameters of the source ( masses , spins , @xmath0 ) ,
it is necessary to know in advance the shape of the gw signals emitted by inspiralling and merging black holes .
indeed , the detection and subsequent data analysis of gw signals is made by using a large bank of _ templates _ that accurately represent the gw waveforms emitted by the source .
here , we shall introduce the reader to one promising strategy toward having an accurate analytical numerical relativity simulations , numerically solving ode s is extremely fast , and can therefore be done ( possibly even in real time ) for a dense sampling of theoretical parameters , such as orbital ( @xmath1 ) or spin ( @xmath2 ) parameters . ]
description of the motion and radiation of binary black holes , which covers all its stages ( inspiral , plunge , merger and ring - down ) : the _ effective one body _
approach @xcite .
as early as 2000 @xcite this method made several quantitative and qualitative predictions concerning the dynamics of the coalescence , and the corresponding gw radiation , notably : ( i ) a blurred transition from inspiral to a ` plunge ' that is just a smooth continuation of the inspiral , ( ii ) a sharp transition , around the merger of the black holes , between a continued inspiral and a ring - down signal , and ( iii ) estimates of the radiated energy and of the spin of the final black hole .
in addition , the effects of the individual spins of the black holes were investigated within the eob @xcite and were shown to lead to a larger energy release for spins parallel to the orbital angular momentum , and to a dimensionless rotation parameter @xmath3 always smaller than unity at the end of the inspiral ( so that a kerr black hole can form right after the inspiral phase ) .
all those predictions have been broadly confirmed by the results of the recent numerical simulations performed by several independent groups @xcite ( for a review of numerical relativity results see @xcite ) .
note that , in spite of the high computer power used in these simulations , the calculation of one sufficiently long waveform ( corresponding to specific values of the many continuous parameters describing the two arbitrary masses , the initial spin vectors , and other initial data ) takes on the order of two weeks .
this is a very strong argument for developing analytical models of waveforms .
those recent breakthroughs in numerical relativity ( nr ) open the possibility of comparing in detail the eob description to nr results .
this eob / nr comparison has been recently initiated in several works @xcite .
the level of analytical / numerical agreement is unprecedented , compared to what has been previously achieved when comparing other types of analytical waveforms to numerical ones .
for instance , @xcite found that the maximal dephasing between a recent , very accurate nr simulation of 30 gw cycles during late inspiral @xcite and eob could be reduced below @xmath4 gw cycles .
the same comparison exhibited also an excellent agreement between the amplitudes of the nr and eob waveforms . if the reader wishes to put the eob results in contrast with other ( post - newtonian or hybrid ) approaches he can consult , _
e.g. _ , @xcite . before reviewing some of the technical aspects of the eob method ,
let us indicate some of the historical roots of this method .
first , we note that the eob approach comprises three , rather separate , ingredients : * a description of the conservative ( hamiltonian ) part of the dynamics of two black holes ; * an expression for the radiation - reaction part of the dynamics ; * a description of the gw waveform emitted by a coalescing binary system .
for each one of these ingredients , the essential inputs that are used in eob developments are high - order post - newtonian ( pn ) expanded results which have been obtained by many years of work , by many researchers ( see references below ) .
however , one of the key ideas in the eob philosophy is to avoid using pn results in their original ` taylor - expanded ' form ( _ i.e. _ @xmath5 , but to use them instead in some _ resummed _ form ( _ i.e. _ some non - polynomial function of @xmath6 , defined so as to incorporate some of the expected non - perturbative features of the exact result ) .
the basic ideas and techniques for resumming each ingredient of the eob are different and have different historical roots . concerning the ingredient ( 1 ) , _ i.e. _ the eob hamiltonian , it was inspired by an approach to electromagnetically interacting quantum two - body systems introduced by brzin , itzykson and zinn - justin @xcite .
the resummation of the ingredient ( 2 ) , _ i.e. _ the eob radiation - reaction force @xmath7 , was inspired by the pad resummation of the flux function introduced by damour , iyer and sathyaprakash @xcite .
as for the ingredient ( 3 ) , _ i.e. _ the eob description of the waveform emitted by a coalescing black hole binary , it was mainly inspired by the work of davis , ruffini and tiomno @xcite which discovered the transition between the plunge signal and a ringing tail when a particle falls into a black hole .
additional motivation for the eob treatment of the transition from plunge to ring - down came from work on the , so - called , ` close limit approximation ' @xcite .
let us finally note that the eob approach has been recently improved by following a methodology consisting of studying , element by element , the physics behind each feature of the waveform , and on systematically comparing various eob - based waveforms with ` exact ' waveforms obtained by nr approaches . among these ` exact ' nr waveforms , it has been useful to consider the small - mass - ratio limit @xmath8 , in which one can use the well controllable ` laboratory ' of numerical simulations of test particles ( with an added radiation - reaction force ) moving in black hole backgrounds @xcite .
before discussing the various resummation techniques used in the eob approach , let us briefly recall the ` taylor - expanded ' results that have been obtained by pushing to high accuracies the post - newtonian ( pn ) methods .
concerning the orbital dynamics of compact binaries , we recall that the 2.5pn - accurate - pn accuracy ' means that a result has been derived up to ( and including ) terms which are @xmath9 fractionally smaller than the leading contribution . ]
equations of motion have been derived in the 1980 s @xcite . pushing the accuracy of the equations of motion to the 3pn ( @xmath10 ) level proved to be a non - trivial task . at first , the representation of black holes by delta - function sources and the use of the ( non diffeomorphism invariant ) hadamard regularization method led to ambiguities in the computation of the badly divergent integrals that enter the 3pn equations of motion @xcite .
this problem was solved by using the ( diffeomorphism invariant ) _ dimensional regularization _ method ( _ i.e. _ analytic continuation in the dimension of space @xmath11 ) which allowed one to complete the determination of the 3pn - level equations of motion @xcite .
they have also been derived by an einstein - infeld - hoffmann - type surface - integral approach @xcite .
the 3.5pn terms in the equations of motion are also known @xcite . concerning the emission of gravitational radiation , two different _ gravitational - wave generation formalisms _
have been developed up to a high pn accuracy : ( i ) the blanchet - damour - iyer formalism @xcite combines a multipolar post - minkowskian ( mpm ) expansion in the exterior zone with a post - newtonian expansion in the near zone ; while ( ii ) the will - wiseman - pati formalism @xcite uses a direct integration of the relaxed einstein equations .
these formalisms were used to compute increasingly accurate estimates of the gravitational waveforms emitted by inspiralling binaries .
these estimates include both normal , near - zone generated post - newtonian effects ( at the 1pn @xcite , 2pn @xcite , and 3pn @xcite levels ) , and more subtle , wave - zone generated ( linear and non - linear ) ` tail effects ' @xcite . however , technical problems arose at the 3pn level .
the representation of black holes by ` delta - function ' sources causes the appearance of dangerously divergent integrals in the 3pn multipole moments . the use of hadamard ( partie finie ) regularization did not allow one to unambiguously compute the needed 3pn - accurate quadrupole moment .
only the use of the ( formally ) diffeomorphism - invariant _ dimensional regularization _ method allowed one to complete the 3pn - level gravitational - radiation formalism @xcite .
the works mentioned in this section ( see @xcite for a detailed account and more references ) finally lead to pn - expanded results for the motion and radiation of binary black holes .
for instance , the equations of motion are given in the form ( @xmath12 ; @xmath13 ) @xmath14 where @xmath15 denotes the ` conservative ' 3pn - accurate terms , while @xmath16 denotes the time - asymmetric contibutions , linked to ` radiation reaction ' .
on the other hand , if we consider for simplicity the inspiralling motion of a quasi - circular binary system , the essential quantity describing the emitted gravitational waveform is the _ phase _ @xmath17 of the quadrupolar gravitational wave amplitude @xmath18 .
pn theory allows one to derive several different functional expressions for the gravitational wave phase @xmath17 , as a function either of time or of the instantaneous frequency .
for instance , as a function of time , @xmath17 admits the following explicit expansion in powers of @xmath19 ( where @xmath20 denotes a formal ` time of coalescence ' , @xmath21 and @xmath22 ) @xmath23 with some numerical coefficients @xmath24 which depend only on the dimensionless ( symmetric ) mass ratio @xmath22 .
the derivation of the 3.5pn - accurate expansion ( [ eq6.4 ] ) uses both the 3pn - accurate conservative acceleration ( [ eq6.2 ] ) and a 3.5pn extension of the ( fractionally ) 1pn - accurate radiation reaction acceleration ( [ eq6.3 ] ) obtained by assuming a balance between the energy of the binary system and the gravitational - wave energy flux at infinity ( see , _
e.g. _ , @xcite ) .
the pn - expanded results briefly reviewed in the previous section are expected to yield accurate descriptions of the motion and radiation of binary black holes only during their _ early inspiralling _ stage , _ i.e. _ as long as the pn expansion parameter @xmath25 ( where @xmath26 is the distance between the two black holes ) stays significantly smaller than the value @xmath27 where the orbital motion is expected to become dynamically unstable ( ` last stable circular orbit ' and beginning of a ` plunge ' leading to the merger of the two black holes ) .
one needs a better description of the motion and radiation to describe the _ late inspiral _ ( say @xmath28 ) , as well as the subsequent _ plunge _ and _ merger_. one possible strategy for having a complete description of the motion and radiation of binary black holes , covering all the stages ( inspiral , plunge , merger , ring - down ) , would then be to try to ` stitch together ' pn - expanded analytical results describing the early inspiral phase with @xmath29 numerical results describing the end of the inspiral , the plunge , the merger and the ring - down of the final black hole , see , _
e.g. _ , refs .
. however , we wish to argue that the eob approach makes a better use of all the analytical information contained in the pn - expanded results ( [ eq6.1])-([eq6.3 ] ) .
the basic claim ( first made in @xcite ) is that the use of suitable _ resummation methods _ should allow one to describe , by analytical tools , a _ sufficiently accurate _
approximation of the _ entire waveform _ , from inspiral to ring - down , including the non - perturbative plunge and merger phases . to reach such a goal , one needs to make use of several tools : ( i ) resummation methods , ( ii ) exploitation of the flexibility of analytical approaches , ( iii ) extraction of the non - perturbative information contained in various numerical simulations , ( iv ) qualitative understanding of the basic physical features which determine the waveform .
let us start by discussing the first tool used in the eob approach : the systematic use of resummation methods .
two such methods have been employed ( and combined ) , and some evidence has been given that they do significantly improve the convergence properties of pn expansions .
the first method is the use of _ pad approximants_. it has been shown in ref . @xcite that near - diagonal pad approximants of the radiation reaction force the _ hamiltonian _ version of the radiation reaction term @xmath30 , eq .
( [ eq6.3 ] ) , in the ( pn - expanded ) equations of motion .
it can be heuristically computed up to ( absolute ) 5.5pn @xcite and even 6pn @xcite order by assuming that the energy radiated in gravitational waves at infinity is balanced by a loss of the dynamical energy of the binary system . ]
@xmath7 seemed to provide a good representation of @xmath7 down to the last stable orbit ( which is expected to occur when @xmath31 , _
i.e. _ when @xmath32 ) .
the second method is a novel approach to the dynamics of compact binaries , which constitutes the core of the effective one body ( eob ) method . for simplicity of exposition ,
let us first explain the eob method at the 2pn level .
the starting point of the method is the 2pn - accurate hamiltonian describing ( in arnowitt - deser - misner - type coordinates ) the conservative , or time symmetric , part of the equations of motion ( [ eq6.1 ] ) ( _ i.e. _ the truncation @xmath33 of eq .
( [ eq6.2 ] ) ) say @xmath34 . by going to the center of mass of the system @xmath35 , one obtains a pn - expanded hamiltonian describing the _ relative motion _ , @xmath36 , @xmath37 : @xmath38 where @xmath39 ( with @xmath21 and @xmath40 ) corresponds to the newtonian approximation to the relative motion , while @xmath41 describes 1pn corrections and @xmath42 2pn ones .
it is well known that , at the newtonian approximation , @xmath43 can be thought of as describing a ` test particle ' of mass @xmath44 orbiting around an ` external mass ' @xmath45 .
the eob approach is a _
general relativistic generalization _ of this fact .
it consists in looking for an ` external spacetime geometry ' @xmath46 such that the geodesic dynamics of a ` test particle ' of mass @xmath44 within @xmath47 is _ equivalent _ ( when expanded in powers of @xmath48 ) to the original , relative pn - expanded dynamics ( [ eq7.1 ] ) .
let us explain the idea , proposed in @xcite , for establishing a ` dictionary ' between the real relative - motion dynamics , ( [ eq7.1 ] ) , and the dynamics of an ` effective ' particle of mass @xmath44 moving in @xmath47 .
the idea consists in ` thinking quantum mechanically ' . instead of thinking in terms of a classical hamiltonian , @xmath49 ( such as @xmath50 , eq .
( [ eq7.1 ] ) ) , and of its classical bound orbits , we can think in terms of the quantized energy levels @xmath51 of the quantum bound states of the hamiltonian operator @xmath52 .
these energy levels will depend on two ( integer valued ) quantum numbers @xmath53 and @xmath54 .
here ( for a spherically symmetric interaction , as appropriate to @xmath55 ) , @xmath54 parametrizes the total orbital angular momentum ( @xmath56 ) , while @xmath53 represents the ` principal quantum number ' @xmath57 , where @xmath58 ( the ` radial quantum number ' ) denotes the number of nodes in the radial wave function .
the third ` magnetic quantum number ' @xmath59 ( with @xmath60 ) does not enter the energy levels because of the spherical symmetry of the two - body interaction ( in the center of of mass frame ) .
for instance , a non - relativistic coulomb ( or newton ! ) interaction @xmath61 gives rise to the well - known result @xmath62 which depends only on @xmath53 ( this is the famous coulomb degeneracy ) .
when considering the pn corrections to @xmath63 , as in eq .
( [ eq7.1 ] ) , one gets a more complicated expression of the form @xmath64 \ , , \end{aligned}\ ] ] where we have set @xmath65 , and where we consider , for simplicity , the ( quasi - classical ) limit where @xmath53 and @xmath54 are large numbers . the 2pn - accurate result ( [ eqn3 ] )
had been derived by damour and schfer @xcite as early as 1988 .
the dimensionless coefficients @xmath66 are functions of the symmetric mass ratio @xmath67 , for instance @xmath68 .
in classical mechanics ( _ i.e. _ for large @xmath53 and @xmath54 ) , it is called the ` delaunay hamiltonian ' , _ i.e. _ the hamiltonian expressed in terms of the _ action variables _ , _ i.e. _ , classically , @xmath69 . ]
@xmath70 , and @xmath71 , with @xmath72 . the energy levels ( [ eqn3 ] ) encode , in a gauge - invariant way , the 2pn - accurate relative dynamics of a ` real ' binary .
let us now consider an auxiliary problem : the ` effective ' dynamics of one body , of mass @xmath44 , following a geodesic in some ` external ' ( spherically symmetric ) metric differs from the two - body adm - coordinate relative distance @xmath73 .
the transformation between the two coordinate systems has been determined in refs .
@xmath74 here , the _ a priori unknown _ metric functions @xmath75 and @xmath76 will be constructed in the form of an expansion in @xmath77 : @xmath78 where the dimensionless coefficients @xmath79 depend on @xmath80 . from the newtonian limit ,
it is clear that we should set @xmath81 . by solving ( by separation of variables ) the ` effective ' hamilton - jacobi equation @xmath82 @xmath83 one can straightforwardly compute ( in the quasi - classical , large quantum numbers limit ) the delaunay hamiltonian @xmath84 , with @xmath85 , @xmath86 ( where @xmath87 , with @xmath88 , @xmath89 ) .
this yields a result of the form @xmath90 \ , , \nonumber \end{aligned}\ ] ] where the dimensionless coefficients @xmath91 are now functions of the unknown coefficients @xmath79 entering the looked for ` external ' metric coefficients ( [ eqn4 ] ) . at this stage
, one needs ( as in the famous ads / cft correspondence ) to define a ` dictionary ' between the real ( relative ) two - body dynamics , summarized in eq .
( [ eqn3 ] ) , and the effective one - body one , summarized in eq .
( [ eqn6 ] ) . as , on both sides , quantum mechanics tells us that the action variables are quantized in integers ( @xmath92 , @xmath93 , etc . )
it is most natural to identify @xmath94 and @xmath95 .
one then still needs a rule for relating the two different energies @xmath96 and @xmath97 .
@xcite proposed to look for a general map between the real energy levels and the effective ones ( which , as seen when comparing ( [ eqn3 ] ) and ( [ eqn6 ] ) , can not be directly identified because they do not include the same rest - mass contribution , while @xmath98 . ] ) , namely @xmath99 the ` correspondence ' between the real and effective energy levels is illustrated in fig . 1 .
denotes the ` principal quantum number ' ( @xmath100 , with @xmath101 denoting the number of nodes in the radial function ) , while @xmath54 denotes the ( relative ) orbital angular momentum @xmath102 .
though the eob method is purely classical , it is conceptually useful to think in terms of the underlying ( bohr - sommerfeld ) quantization conditions of the action variables @xmath103 and @xmath104 to motivate the identification between @xmath53 and @xmath54 in the two dynamics.,height=302 ] finally , identifying @xmath105 to @xmath106 yields six equations , relating the six coefficients @xmath107 to the six @xmath108 and to the two energy coefficients @xmath109 and @xmath110 .
it is natural to set @xmath111 ( so that the linearized effective metric coincides with the linearized schwarzschild metric with mass @xmath112 ) .
one then finds that there exists a _ unique _
solution for the remaining five unknown coefficients @xmath113 and @xmath110 .
this solution is very simple : @xmath114 note , in particular , that the map between the two energies is simply @xmath115 where @xmath116 is mandelstam s invariant .
note also that , at 2pn accuracy , the crucial ` @xmath117 ' metric coefficient @xmath75 ( which fully encodes the energetics of circular orbits ) is given by the remarkably simple pn expansion @xmath118 where @xmath119 and @xmath120 .
the dimensionless parameter @xmath67 varies between @xmath121 ( in the test mass limit @xmath122 ) and @xmath123 ( in the equal - mass case @xmath124 ) . when @xmath125 , eq .
( [ eq7.3 ] ) yields back , as expected , the well - known schwarzschild time - time metric coefficient @xmath126 .
one therefore sees in eq .
( [ eq7.3 ] ) the rle of @xmath80 as a deformation parameter connecting a well - known test - mass result to a non trivial and new 2pn result .
it is also to be noted that the 1pn eob result @xmath127 happens to be @xmath80-independent , and therefore identical to @xmath128 .
this is remarkable in view of the many non - trivial @xmath80-dependent terms in the 1pn relative dynamics .
the physically real 1pn @xmath80-dependence happens to be fully encoded in the function @xmath129 mapping the two energy spectra given in eq .
( [ eqn9 ] ) above .
let us emphasize the remarkable simplicity of the 2pn result ( [ eq7.3 ] ) .
the 2pn hamiltonian ( [ eq7.1 ] ) contains eleven rather complicated @xmath80-dependent terms .
after transformation to the eob format , the dynamical information contained in these eleven coefficients gets _ compactified _ into the very simple additional contribution @xmath130 in @xmath75 , together with an equally simple contribution in the radial metric coefficient : @xmath131 .
this compactification process is even more drastic when one goes to the next ( conservative ) post - newtonian order : the 3pn level , i.e. additional terms of order @xmath132 in the hamiltonian ( [ eq7.1 ] ) . as mentioned above
, the complete obtention of the 3pn dynamics has represented quite a theoretical challenge and the final , resulting hamiltonian is quite complicated . even after going to the center of mass frame , the 3pn additional contribution @xmath133 to eq .
( [ eq7.1 ] ) introduces eleven new complicated @xmath80-dependent coefficients .
after transformation to the eob format @xcite , these eleven new coefficients get ` compactified ' into only _ three _ additional terms : ( i ) an additional contribution to @xmath75 , ( ii ) an additional contribution to @xmath76 , and ( iii ) a @xmath134 modification of the ` external ' geodesic hamiltonian .
for instance , the crucial 3pn @xmath117 metric coefficient becomes @xmath135 where @xmath136 remarkably , it is found that the very simple 2pn energy map eq .
( [ eqn9 ] ) does not need to be modified at the 3pn level .
the fact that the 3pn coefficient @xmath137 in the crucial ` effective radial potential ' @xmath138 , eq .
( [ eq7.5 ] ) , is rather large and positive indicates that the @xmath80-dependent nonlinear gravitational effects lead , for comparable masses @xmath139 ) , to a last stable ( circular ) orbit ( lso ) which has a higher frequency and a larger binding energy than what a naive scaling from the test - particle limit @xmath140 would suggest .
actually , the pn - expanded form ( [ eq7.5 ] ) of @xmath138 does not seem to be a good representation of the ( unknown ) exact function @xmath141 when the ( schwarzschild - like ) relative coordinate @xmath26 becomes smaller than about @xmath142 ( which is the radius of the lso in the test - mass limit ) .
it was therefore suggested @xcite to further resum already represent a _
resummation _ of the pn dynamics in the sense that they have compactified the many terms of the original pn - expanded hamiltonian within a very concise format .
but one should not refrain to further resum the eob building blocks themselves , if this is physically motivated . ] @xmath138 by replacing it by a suitable pad @xmath143 approximant .
for instance , the replacement of @xmath138 by @xmath144 = \frac{1+n_1 u}{1+d_1 u + d_2 u^2 + d_3 u^3}\ ] ] ensures that the @xmath145 case is smoothly connected with the @xmath146 limit . the use of ( [ eq7.7 ] )
was suggested before one had any ( reliable ) non - perturbative information on the binding of close black hole binaries .
later , a comparison with some ` waveless ' numerical simulations of circular black hole binaries @xcite has given some evidence that ( [ eq7.7 ] ) is physically adequate . in refs .
@xcite it was also emphasized that , in principle , the comparison between numerical data and eob - based predictions should allow one to determine the effect of the unknown higher pn contributions to eq .
( [ eq7.5 ] ) .
for instance , one can add a 4pn - like term @xmath147 in eq .
( [ eq7.5 ] ) , and then pad the resulting radial function , say @xmath148 $ ] . comparing the predictions of @xmath149 $ ] to numerical data might then determine what is the physically preferred ` effective ' value of the unknown coefficient @xmath150 .
this is an example of the useful _ flexibility _
@xcite of analytical approaches : the fact that one can tap numerically - based , non - perturbative information to improve the eob approach .
in the previous section we have described how the eob method encodes the conservative part of the relative orbital dynamics into the dynamics of an effective particle .
let us now briefly discuss how to complete the eob dynamics by defining a _ resummed _ expression describing radiation reaction effects .
one is interested in circularized binaries , which have lost their initial eccentricity under the influence of radiation reaction .
for such systems , it is enough to include a radiation reaction force in the @xmath151 equation of motion .
more precisely , we are using phase space variables @xmath152 associated to polar coordinates ( in the equatorial plane @xmath69 ) .
actually it is convenient to replace the radial momentum @xmath153 by the momentum conjugate to the ` tortoise ' radial coordinate @xmath154 , _ i.e. _ @xmath155 .
the real eob hamiltonian is obtained by first solving eq .
( [ eqn9 ] ) to get @xmath156 in terms of @xmath97 , and then by solving the effective hamiltonian - jacobi equation terms that must be introduced at 3pn . ]
to get @xmath97 in terms of the effective phase space coordinates @xmath157 and @xmath158 .
the result is given by two nested square roots ( we henceforth set @xmath159 ) : @xmath160 where @xmath161 with @xmath162 .
here , we are using suitably rescaled dimensionless ( effective ) variables : @xmath163 , @xmath164 , @xmath165 , as well as a rescaled time @xmath166 .
this leads to equations of motion of the form @xmath167
@xmath168 @xmath169 @xmath170 as explained above the eob metric function @xmath171 is defined by pad resumming the taylor - expanded result ( [ eqn4 ] ) obtained from the matching between the real and effective energy levels ( one uses a similar pad resumming for @xmath172 , or rather for @xmath173 ) .
one similarly needs to resum the @xmath174 component of the radiation reaction which has been introduced on the r.h.s .
of the last equation ( [ eqn12 ] ) . during the quasi - circular inspiral @xmath175
is known ( from the pn work mentioned in section 2 above ) in the form of a taylor expansion of the form @xmath176 where @xmath177 , @xmath178 ( with @xmath179 defined as in eq .
( 22 ) of @xcite ) , and @xmath180 where we have added to the known 3.5pn - accurate comparable - mass result the small - mass - ratio 4pn contribution @xcite .
following @xcite , one resums @xmath181 by using the following pad resummation approach .
first , one chooses a certain number @xmath182 which is intended to represent the value of @xmath183 at which the exact angular momentum flux would become infinite if one were to formally analytically continue @xmath175 along _ unstable _ circular orbits below the last stable orbit ( lso ) : then , given @xmath182 , one defines the resummed @xmath184 as @xmath185 \
, , \ ] ] where @xmath186 denotes a @xmath187 pad approximant .
it has recently been shown @xcite that the _ flexibility _ in the choice of @xmath182 could be advantageously used to get a close agreement with nr data .
the quality of the agreement with nr data obtained by such a ` flexed ' pad resummation is illustrated ( in the small mass - ratio limit ) in fig .
2 . nicely agrees with the exact , numerical flux when using a suitably ` flexed ' value of @xmath182 .
the solid curve is the exact , numerical flux @xcite .
the empty circles denote the @xmath188 pad resummation using @xmath189 . the dashed curve , below the other two , denote the ( 3,3 ) pad resummation using the value @xmath190 , as originally suggested @xcite.,height=302 ]
let us note that the eob method is not restricted to planar motions .
in particular , the entire method has been extended to the case of circularized motions of _ spinning _ black holes @xcite . in this case
, one must work with more phase space variables , @xmath191 , @xmath192 , @xmath193 and @xmath194 , and include spin effects in the radiation reaction .
to end this brief review , let us sketch the definition of accurate waveforms within the eob approach .
the construction of eob waveforms is based on several separate inputs : * high - accuracy , pn - expanded inspiralling waveforms as basic inputs ( see references above ) ; * a specific resummation of the inspiral - plus - plunge waveform @xmath195 ( which includes the resummation of an infinite number of ` leading logarithms ' entering the multipolar - post - minkowskian expansion of tail effects , the resummation of the 5pn - accurate .
the 4pn and 5pn pieces are only known in the test - mass limit @xmath125 @xcite . ] ` non - tail ' @xmath196 , @xmath197 waveform , @xmath197 waveform .
] , and the parametrization of some non - quasi - circular effects ) @xcite : @xmath198 * a simplified representation of the transition between plunge and ring - down by smoothly _ matching _ , on a @xmath199-toothed `` comb '' @xmath200 centered around a matching time @xmath201 , the inspiral - plus - plunge waveform to a ring - down waveform , made of the superposition of several , _
i.e. _ a 5-teethed comb , and , correspondingly , 5 positive - frequency kerr quasi - normal modes . ]
quasi - normal - mode complex frequencies , @xmath202 with @xmath203 , and where the label @xmath204 refers to indices @xmath205 , with @xmath206 being the schwarzschild - background multipolarity of the considered ( metric ) waveform @xmath207 , with @xmath208 being the ` overtone number ' of the considered kerr - background quasi - normal - mode , and @xmath209 the degree of its associated spheroidal harmonics @xmath210 .
as discussed in @xcite and @xcite the physics of the transition between plunge and ring - down ( which was first understood in the classic work of davis , ruffini and tiomno @xcite ) suggests to choose as matching time @xmath201 , in the comparable - mass case , the eob time when the eob orbital frequency @xmath211 reaches its _ maximum _ value . here
, we have been describing the @xmath212-pn - accurate _ resummed _ eob waveform introduced in @xcite .
a less accurate , _ restricted _ ( @xmath121-pn ) eob waveform ( together with a simplified matching procedure ; @xmath213 ) has been used by buonanno and collaborators @xcite . , @xmath197 ) waveform ( [ eqn18 ] ) ( using @xmath216 , @xmath217 and a suitably flexed @xmath182 ) , with , about , 29 gw cycles of inspiral , @xmath218 gw cycle during plunge , and @xmath219 gw cycles of ring - down . ref .
@xcite has shown that the inspiral part of the eob waveform agreed with the caltech - cornell nr data @xcite within @xmath220 gw cycles , while ref .
@xcite has shown that the late - inspiral - plunge - ringdown part of the eob waveform agrees with an albert einstein institute nr waveform within @xmath221 gw cycles.,height=302 ] recent comparisons between eob waveforms and full , 3-dimensional nr waveforms have exhibited an excellent level of agreement .
after a preliminary comparison done in @xcite , buonanno et al .
@xcite compared _ restricted _ eob waveforms to nr waveforms computed by the nasa - goddard group . in the equal - mass case ( @xmath222 ) , they found that the dephasing between ( restricted ) eob and nr waveforms ( covering late inspiral , merger and ring - down ) stayed within @xmath223 gw cycles over 14 gw cycles . in the case of a mass ratio @xmath224 ( @xmath225 ) , the dephasing stayed within @xmath226 gw cycles over 9 gw cycles
the _ resummed _ eob waveform has been compared to two different ( equal - mass ) nr waveforms : ( i ) in the comparison with the very accurate inspiralling simulation of the caltech - cornell group @xcite the dephasing stayed smaller than @xmath220 gw cycles over 30 gw cycles ( and the amplitudes agreed at the @xmath227 level ) @xcite ; ( ii ) in the comparison with a late - inspiral - merger - ringdown nr waveform computed by the albert einstein institute group , the dephasing stayed smaller than @xmath221 gw cycles over 12 gw cycles .
see ref .
@xcite for another eob / nr comparison .
all these comparisons have used the natural _ flexibility _ of the eob formalism @xcite , _
i.e. _ the possibility to tune eob parameters representing yet uncalculated effects ( such as @xmath150 in @xmath171 or @xmath182 ) .
* acknowledgments . *
i thank alessandro nagar for many fruitful interactions , and the conveners of the second stueckelberg meeting for organizing a stimulating workshop .
i am also grateful to alessandro nagar and marie - claude vergne for help with the figures , and eric poisson for providing the data of fig . 2 .
t. damour , p. jaranowski and g. schfer : on the determination of the last stable orbit for circular general relativistic binaries at the third post - newtonian approximation , phys .
d * 62 * , 084011 ( 2000 ) [ arxiv : gr - qc/0005034 ] .
j. a. gonzalez , u. sperhake , b. bruegmann , m. hannam and s. husa : total recoil : the maximum kick from nonspinning black - hole binary inspiral , phys .
lett . * 98 * , 091101 ( 2007 ) [ arxiv : gr - qc/0610154 ] .
m. koppitz , d. pollney , c. reisswig , l. rezzolla , j. thornburg , p. diener and e. schnetter : getting a kick from equal - mass binary black hole mergers , phys .
99 * , 041102 ( 2007 ) [ arxiv : gr - qc/0701163 ] .
l. rezzolla , p. diener , e. n. dorband , d. pollney , c. reisswig , e. schnetter and j. seiler : the final spin from the coalescence of aligned - spin black - hole binaries , the astrophysical journal , volume 674 , issue 1 , pp .
l29-l32 ( 2008 ) , arxiv:0710.3345 [ gr - qc ] .
f. pretorius : binary black hole coalescence , the ` final version ' of this lecture note will appear in the book : ` relativistic objects in compact binaries : from birth to coalescence ' editor : colpi et al .
pulisher : springer verlag , canopus publishing limited , arxiv:0710.1338 [ gr - qc ] .
a. buonanno , y. pan , j. g. baker , j. centrella , b. j. kelly , s. t. mcwilliams and j. r. van meter : toward faithful templates for non - spinning binary black holes using the effective - one - body approach , phys .
d * 76 * , 104049 ( 2007 ) , [ arxiv:0706.3732 [ gr - qc ] ] .
t. damour , a. nagar , e. n. dorband , d. pollney and l. rezzolla : faithful effective - one - body waveforms of equal - mass coalescing black - hole binaries , submitted to phys .
d , arxiv:0712.3003v1 [ gr - qc ] .
a. gopakumar , m. hannam , s. husa and b. brugmann : comparison between numerical relativity and a new class of post - newtonian gravitational - wave phase evolutions : the non - spinning equal - mass case , arxiv:0712.3737 [ gr - qc ] .
p. jaranowski and g. sch " afer : 3rd post - newtonian higher order hamilton dynamics for two - body point - mass systems , phys .
d * 57 * ( 1998 ) 7274 - 7291 ; erratum - ibid .
d * 63 * ( 2001 ) 029902 , arxiv : gr - qc/9712075 .
l. blanchet , t. damour and g. esposito - farse : dimensional regularization of the third post - newtonian dynamics of point particles in harmonic coordinates , phys .
d * 69 * , 124007 ( 2004 ) , arxiv : gr - qc/0311052 .
m. e. pati and c. m. will : post - newtonian gravitational radiation and equations of motion via direct integration of the relaxed einstein equations .
ii : two - body equations of motion to second post - newtonian order , and radiation - reaction to 3.5 post - newton , phys . rev .
d * 65 * , 104008 ( 2002 ) , arxiv : gr - qc/0201001 . c. m. will and a. g. wiseman : gravitational radiation from compact binary systems : gravitational waveforms and energy loss to second post - newtonian order , phys.rev .
d * 54 * , 4813 ( 1996 ) , arxiv : gr - qc/9608012 .
m. e. pati and c. m. will : post - newtonian gravitational radiation and equations of motion via direct integration of the relaxed einstein equations . i. foundations , phys . rev .
d * 62 * ( 2000 ) 124015 , arxiv : gr - qc/0007087 .
l. blanchet , t. damour , b. r. iyer , c. m. will and a. g. wiseman : gravitational radiation damping of compact binary systems to second post - newtonian order , phys
* 74 * , 3515 ( 1995 ) , arxiv : gr - qc/9501027 .
l. blanchet , t. damour and b. r. iyer : gravitational waves from inspiralling compact binaries : energy loss and wave form to second post - newtonian order , phys .
d * 51 * , 5360 ( 1995 ) [ erratum - ibid .
d * 54 * , 1860 ( 1996 ) ] , arxiv : gr - qc/9501029 .
l. blanchet , b. r. iyer and b. joguet : gravitational waves from inspiralling compact binaries : energy flux to third post - newtonian order , phys .
d * 65 * ( 2002 ) 064005 ; erratum - ibid .
d71 ( 2005 ) 129903 , arxiv : gr - qc/0105098 .
l. blanchet , t. damour , g. esposito - farse and b. r. iyer : dimensional regularization of the third post - newtonian gravitational wave generation from two point masses , phys .
d * 71 * ( 2005 ) 124004 , arxiv : gr - qc/0503044 .
j. g. baker , s. t. mcwilliams , j. r. van meter , j. centrella , d .-
choi , b. j. kelly and m. koppitz : binary black hole late inspiral : simulations for gravitational wave observations , phys .
d * 75 * ( 2007 ) 124024 , [ gr - qc/0612117 ] .
l. blanchet , t. damour , g. esposito - farse and b. r. iyer : gravitational radiation from inspiralling compact binaries completed at the third post - newtonian order , phys .
* 93 * ( 2004 ) 091101 , arxiv : gr - qc/0406012 .
l. blanchet , g. faye , b. r. iyer and b. joguet : gravitational - wave inspiral of compact binary systems to 7/2 post - newtonian order , phys .
d * 65 * , 061501 ( 2002 ) [ erratum - ibid .
d * 71 * , 129902 ( 2005 ) ] , arxiv : gr - qc/0105099 .
t. damour , e. gourgoulhon and p. grandclement : circular orbits of corotating binary black holes : comparison between analytical and numerical results , phys .
d * 66 * , 024007 ( 2002 ) , arxiv : gr - qc/0204011 .
t. damour , b. r. iyer , p. jaranowski and b. s. sathyaprakash : gravitational waves from black hole binary inspiral and merger : the span of third post - newtonian effective - one - body templates , phys .
d * 67 * , 064028 ( 2003 ) [ arxiv : gr - qc/0211041 ] .
h. tagoshi and m. sasaki : postnewtonian expansion of gravitational waves from a particle in circular orbit around a schwarzschild black hole , prog .
phys . *
92 * , 745 ( 1994 ) [ arxiv : gr - qc/9405062 ] .
t. tanaka , h. tagoshi and m. sasaki : gravitational waves by a particle in circular orbits around a schwarzschild black hole
5.5 post - newtonian formula ,
* 96 * ( 1996 ) 1087 - 1101 , [ arxiv : gr - qc/9701050v1 ]
. l. e. kidder : using full information when computing modes of post - newtonian waveforms from inspiralling compact binaries in circular orbit , to be published in physical review d15 , arxiv:0710.0614v1 [ gr - qc ] .
e. poisson : gravitational radiation from a particle in circular orbit around a black hole . 6 .
accuracy of the postnewtonian expansion , phys .
d * 52 * , 5719 ( 1995 ) [ addendum - ibid .
d * 55 * , 7980 ( 1997 ) ] [ arxiv : gr - qc/9505030 ] . | the effective one body ( eob ) formalism is an analytical approach which aims at providing an accurate description of the motion and radiation of coalescing binary black holes .
we present a brief review of the basic elements of this approach . |
curved space - times are familiar from einstein s theory of gravitation,@xcite where the metric tensor @xmath0 , describing distances in a curved space - time with local lorentz invariance , is determined by the solution of the einstein equations .
a major problem for an experimental investigation of the ( kinematical as well as dynamical ) properties of curved space - times is that generating a significant curvature , equivalent to a ( relatively ) small curvature radius , is a close to impossible undertaking in manmade laboratories .
for example , the effect of the gravitation of the whole earth is to cause a deviation from flat space - time on this planet s surface of only the order of @xmath1 ( the ratio of schwarzschild and earth radii ) .
the fact that proper gravitational effects are intrinsically small is basically due to the smallness of newton s gravitational constant @xmath2m@xmath3kg@xmath4sec@xmath5 .
various fundamental classical and quantum effects in strong gravitational fields , a few of which we discuss below , are thus inaccessible for earth - based experiments .
the realm of strong gravitational fields ( or , equivalently , rapidly accelerating a reference frame to simulate gravity according to the equivalence principle ) , is therefore difficult to reach .
however , earth - based experiments are desirable , because they have the obvious advantage that they can be prepared and , in particular , repeated under possibly different conditions at will . the formalism to be described in what follows
is aimed at the realization of _ effective _ curved space - time geometries in perfect fluids , which can indeed be prepared on earth , and which mimic the effects of gravity inasmuch the kinematical properties of curved space - times are concerned . among such perfect fluids
are bose - einstein condensates , i.e. , the dilute matter - wave - coherent gases formed if cooled to ultralow temperatures , where the critical temperatures are of order @xmath6k ; for a short review of the ( relatively ) recent status of this rapidly developing field see ref.@xcite . in what follows , it will be of some importance that bose - einstein condensates belong to a special class of quantum perfect fluids , so - called superfluids.@xcite the curved space - times we have in mind are experienced by sound waves propagating on the background of a spatially and temporally inhomogeneous perfect fluid . of primary importance
is , first of all , to realize that the identification of sound waves propagating on an inhomogeneous background , which is itself determined by a solution of euler and continuity equations , and photons propagating in a curved space - time , which is determined by a solution of the einstein equations , is of a _ kinematical _ nature .
that is , the space - time metric is fixed externally by a background field obeying the laws of hydrodynamics ( which is prepared by the experimentalist ) , and not self - consistently by a solution of the einstein equations @xmath7 ( where @xmath8 and the speed of light are set to unity ) .
the latter equations relate space - time curvature represented by the einstein tensor @xmath9 , where @xmath10 is the ricci tensor and @xmath11 the ricci scalar with the energy - momentum content of all other fundamental quantum fields .
this energy - momentum content is represented by the classical quantity @xmath12 , which is the regularized expectation value of a quantum energy - momentum tensor.@xcite as a first introductory step to understand the nature of the proposed analogy , consider the wave equation for the velocity potential of the sound field @xmath13 , which in a homogeneous medium at rest reads @xmath14\phi = 0 , \label{waveeq}\ ] ] where @xmath15 is the sound speed .
it is a constant in space and time for such a medium .
this equation has lorentz invariance , that is , if we replace the speed of light by the speed of sound , it retains the form shown above in the new space - time co - ordinates obtained after lorentz - transforming to a frame moving at a constant speed less than the sound speed . just as the light field _ in vacuo _ is a proper relativistic field , sound is a `` relativistic '' field .
the lorentz invariance can be made more manifest by writing equation ( [ waveeq ] ) in the form @xmath16 , where @xmath17diag(@xmath18 is the ( contravariant ) flat space - time metric ( we choose throughout the signature of the metric as specified here ) , determining the fundamental light - cone structure of minkowski space;@xcite we employ the summation over equal greek indices @xmath19 assuming , then , the sound speed @xmath20 to be local in space and time , and employing the curved space - time version of the 3 + 1d laplacian @xmath21,@xcite one can write down the sound wave equation in an _ inhomogeneous medium _ in the generally covariant form@xcite @xmath22 here , @xmath23 $ ] is the determinant of the ( covariant ) metric tensor .
it is to be emphasized at this point that because the space and time derivatives @xmath24 are covariantly transforming objects in ( [ curvedspacewaveeq ] ) , the primary object in the condensed - matter identification of space - time metrics via the wave equation ( [ curvedspacewaveeq ] ) is the contravariant metric tensor @xmath25.@xcite in the condensed - matter understanding of analogue gravity , the quantities @xmath25 are _ material - dependent _ coefficients .
they occur in a dispersion relation of the form @xmath26 , where @xmath27 is the covariant wave vector , with @xmath28 the ordinary spatial momentum ( or quasi - momentum in a crystal ) .
the contravariant tensor components @xmath25 , for a perfect , irrotational liquid turn out to be@xcite @xmath29 where @xmath30 is the unit matrix and @xmath31 a space and time dependent function , to be determined from the proper equations of motion for the sound field ( see below ) . inverting this expression according to @xmath32 , to obtain the covariant metric @xmath0 , the fundamental tensor of distance reads @xmath33 where the line element is @xmath34 .
this form of the metric has been derived by unruh for an irrotational perfect fluid described by euler and continuity equations;@xcite its properties were later on explored in more detail in particular by visser.@xcite i also mention that an earlier derivation of unruh s form of the metric exists , from a somewhat different perspective ; it was performed by trautman.@xcite the conformal factor @xmath31 in ( [ gdown ] ) depends on the spatial dimension of the fluid .
it may be unambiguously determined by considering the action of the velocity potential fluctuations above an inhomogeneous background , identifying this action with the action of a minimally coupled scalar field in @xmath35-dimensional space - time:@xcite @xmath36 \nonumber\\ & \equiv & \frac12 \int d^{d+1}x % dt dz \sqrt{-{\sf g } } { \sf g}^{\mu\nu } \partial_\mu \phi \partial_\nu \phi \ , , \label{action}\end{aligned}\ ] ] where it is assumed that the compressibility @xmath37 of the ( barotropic ) fluid , @xmath38 , where @xmath39 is the pressure and @xmath40 the density of the fluid , is a constant . using the above canonical identification
, it may easily be shown that the conformal factor is given by @xmath41 .
it is mentioned here that the case of one spatial dimension ( @xmath42 ) is special , in that the so - called conformal invariance in two space - time dimensions implies that the classical equations of motion are invariant ( take the same form ) for any space and time dependent choice of the conformal factor @xmath31 .
the line element @xmath43 gives us the distances travelled by the phonons in an effective space - time world in which the scalar field @xmath13 `` lives '' .
in particular , quasiclassical ( large momentum ) phonons will follow _
light_-like , that is , here , _
sound_-like geodesics in that space - time , according to @xmath44 .
particularly noteworthy is the simple fact that the constant time slices obtained by setting @xmath45 in the line element are conformally flat , i.e. the quasiparticle world looks on constant time slices like the ordinary ( newtonian ) lab space , with a simple euclidean metric in the case of cartesian spatial co - ordinates we display .
all the intrinsic curvature of the effective space - time is therefore encoded in the metric tensor elements @xmath46 and @xmath47 .
together with matt visser , i described this curvature and its properties for an isotachic fluid ( i.e. , having a speed of sound independent of space and time).@xcite using that phonons move on geodesics , we discovered in@xcite the phenomenon that a vortex acts on quasiclassical phonons as an effective gravitational lens .
in ref.@xcite , using the fact that there is curvature in any spatially inhomogeneous flow ( that is , a flow which is not a simple superposition of translational motion and rigid body rotation ) , we have shown that there exists a sonic analogue of the `` warp - drive '' in general relativity permitting superluminal , i.e. , here `` superphononic '' motion.@xcite the point made by us is that in the acoustic curved space - times we consider , there is no violation of any condition on the positivity of energy necessary , which is in marked contrast to the original warp drives , where local energy densities by necessity must be negative to permit superluminal travel.@xcite this is due to the fact that the einstein equations , relating curvature and energy - momentum content of all fields other than gravitational fields , as already mentioned in the above do not need to be imposed in the analogy .
it is important to recognize that the form ( [ curvedspacewaveeq ] ) of the wave equation is valid generally ( with a possible additional scalar potential term ) .
that is , a generally covariant curved space - time wave equation can be formulated not just for the velocity perturbation ( sound ) potential in an irrotational euler fluid , for which we have introduced the effective metric concept . if the spectrum of excitation ( in the local rest frame ) is linear , @xmath48 , where @xmath49 is the propagation speed of _ some _ collective excitation , the statement that an effective space - time metric exists is true , provided we only consider wave perturbations of a single scalar field which constitutes the fixed classical background .
the argument to reach this conclusion is as follows . given that the action density @xmath50 is a functional of @xmath51 and its space - time derivatives @xmath52 , i.e. @xmath53 $ ] , we expand the action to quadratic order in the fluctuations around some stationary classical background solution @xmath54 of the euler - lagrange equations .
for any lagrangian of the specified form , the wave equation for perturbations @xmath55 above the background @xmath54
is@xcite @xmath56 the above equation in covariant notation reads @xmath57 where @xmath58 is a background - field dependent effective potential ( equal to the second term in round brackets in equation ( [ general ] ) divided by @xmath59 ) .
the potential @xmath60 may , for example , in the simplest case contain an effective mass of the scalar field , such that the wave equation becomes klein - gordon like , @xmath61 .
the effective metric coefficients are , up to an ( again dimension dependent ) conformal factor given by : @xmath62 the concept of an effective space - time metric therefore applies to every system having a single scalar wave equation of second order in both space and time derivatives , corresponding to perturbations propagating on a fixed classical background , where this background itself determines the metric coefficients .
we assumed in eq.([action ] ) that the compressibility @xmath37 is a constant .
this entails that the ( barotropic ) equation of state reads @xmath63 .
we then have , in terms of the interaction between the particles ( atoms ) constituting the fluid , a contact interaction ( pseudo-)potential , @xmath64 .
this is indeed the case for the _ dilute _ atomic gases forming a bose - einstein condensate .
well below the transition temperature , they are described to good accuracy by the gross - pitaevski mean field equation for the order parameter @xmath65 , breaks particle number conservation ( the global u(1 ) invariance ) ; for a review of the consequences see ref.@xcite . ] representing the expectation value of the quantum field operator @xmath66:@xmath67 \psi ( { \bm x},t ) .
\label{gpeq}\ ] ] the madelung transformation reads @xmath68 $ ] , where @xmath40 yields the condensate density and @xmath51 is the velocity potential ; it allows for an interpretation of quantum theory in terms of hydrodynamics.@xcite namely , identifying real and imaginary parts on left- and right - hand sides of ( [ gpeq ] ) , respectively , gives us the two equations @xmath69 the first of these equations is the josephson equation for the superfluid phase , which corresponds to the bernoulli equation of classical hydrodynamics , where the usual velocity potential of irrotational hydrodynamics equals the superfluid phase @xmath51 times @xmath70 , such that @xmath71 .
the latter equation implies that the flow is irrotational save on singular lines , around which the _ wave function phase _
@xmath51 is defined only modulo @xmath72 . therefore , circulation is quantized,@xcite and these singular lines are the center lines of quantized vortices
. the isothermal chemical potential @xmath73 ( which we have chosen to incorporate the kinetic energy term @xmath74 ) , is augmented by the `` quantum pressure term '' @xmath75 , which is the one genuine quantum term in ( [ jaeq ] ) , because @xmath76 ( observe that the first order in @xmath77 may be incorporated into the velocity potential ) . the second equation ( [ conteq ] )
is the continuity equation for conservation of particle number , i.e. , atom number in the superfluid gas .
the dynamics of the weakly interacting , dilute ensemble of atoms is thus that of a perfect euler fluid with quantized circulation of singular vortex lines , which are the only vortical excitations in a superfluid .
the fact that it is an euler fluid is true save for regions in which the density rapidly varies and the quantum pressure term @xmath78 becomes relevant , which happens on scales of order the coherence length @xmath79 , where @xmath80 is a constant asymptotic density far away from the density - depleted ( or possibly density - enhanced ) region .
the quantum pressure becomes relevant in the depleted - density cores of quantized vortices , or at the low - density boundaries of the system , and is negligible outside these domains of rapidly varying or low density .
the whole armoury of space - time metric description of excitations , explained in the last section , which is based on the euler and continuity equations , is then valid for phonon - like excitations of a bose - einstein condensate , with the space - time metric ( [ gdown ] ) , as long as we are outside the core of quantized vortices , ) arising therefrom , may be found in ref.@xcite . ] and far enough from the boundaries of the condensate .
quasiparticles are the fundamental entities used to describe an interacting condensed matter system in a particle picture , that is , in a suitable fock space . on the microscopic level ,
if the elementary constituents interact by two - body forces , we are given a second quantized hamiltonian operator of the form @xmath81 , where @xmath82 are the matrix elements for two - particle interaction in a plane wave basis , and @xmath83 are the bare single particle energies of the `` elementary '' bosons or fermions , which are created by the operators @xmath84 from the proper particle vacuum .
one then employs a unitary , i.e. operator - algebra - conserving bogoliubov transformation@xcite to another set of quasiparticle operators @xmath85 , @xmath86 , see eq.([bogotrafo1 ] ) below .
this gives the hamiltonian the reduced diagonal form @xmath87 , where @xmath88 represents additional terms of higher order than quadratic , which are supposed to be small compared to the leading diagonalized part of the hamiltonian , for the picture of noninteracting quasiparticles to make sense .
these quasiparticles possess a ( possibly spatially anisotropic ) dispersion relation @xmath89 which is linear for various important classes of quasiparticles , @xmath90 . among these classes of collective excitations
are , e.g. , phonons , antiferromagnetic magnons,@xcite or the excitations around the gap nodes of the @xmath39-wave superfluid @xmath91he - a.@xcite phonons are the small momentum quanta of the sound field in solids or fluids , with @xmath92 for a medium at rest .
their classical equation of motion in a perfect fluid is the generally covariant wave equation ( [ curvedspacewaveeq ] ) , with the effective space - time metric ( [ gdown ] ) . in the simplest case , for a homogeneous medium with space and time independent density , and a constant speed of sound @xmath15 ,
the quantized velocity potential of the phonons reads@xcite @xmath93 \left[{\hat{b}}_{\bm k}e^{-ic_s k t + i{\bm k}\cdot { \bm x } } + { \hat{b}^\dagger}_{\bm k } e^{ic_s k t - i{\bm k}\cdot { \bm x } } \right ] , \label{homo}\ ] ] where @xmath94 is the quantization volume of the system , and @xmath80 the constant background density .
that is , the quantum excitation field in a homogeneous medium may generally be decomposed into plane waves , with the appropriate frequencies as a function of momentum stemming from the dispersion relation , here @xmath95 .
note , in particular , that for this spatially and temporally homogeneous case , the statement that we observe positive frequency ( energy ) with respect to the laboratory time interval @xmath96 is unique , that is , it can be made independent of time and space . in an inhomogeneous fluid , this is ( generally ) no longer the case , and the notion of an excitation having positive energy may depend on where the detector is located in the fluid , if it is at rest relative to the fluid or moves , and what its natural time interval is
. the latter may be different from that of the laboratory , due to the particular way the detector couples to the fluid , see section [ gibbonshawking ] .
the number of particles assigned to the quantum field @xmath97 is unique with respect to a certain given state @xmath98 of the quantum field , @xmath99 , provided we decompose @xmath97 into modes associated to the operators @xmath100 and their hermitian conjugates .
the number of particles is , in particular , zero for the _ vacuum state with respect to the field operator @xmath101 _ , defined by @xmath102 .
however , it need not be zero with respect to another set of quasiparticle operators @xmath85 , which has a _ different _ vacuum @xmath103 . to demonstrate the basis dependence of quasiparticle number
, we use a general bogoliubov transformation for bosons , the class of ( quasi-)particles we consider , of the general form @xmath104 where @xmath105 represents a set of quantum numbers , not necessarily the plane waves used in eq.([homo ] ) .
the coefficients in the bogoliubov transformation must fulfill certain conditions for the transformation to be unitary , i.e. , to preserve the bosonic commutation relations for the new operators , @xmath106= \delta_{{\bm i}{\bm k}},\ , [ \hat b_{\bm i } , \hat b_{\bm k}]=0,\ , [ \hat b^\dagger_{\bm i } , \hat b^\dagger_{\bm k}]=0 $ ] . as a consequence of this defining unitary character , the following conditions on the transformation coefficients must hold : @xmath107 by using the transformation ( [ bogotrafo1 ] ) , it is straightforwardly shown that the number of @xmath101 particles in @xmath103 is given by @xmath108 : the old operator @xmath100 does not annihilate the new vacuum @xmath109 ( and vice versa ) , and what looks empty in one quasiparticle vacuum may be full of quasiparticles in another .
related to this ( formal ) operator - basis dependence is the fact that the actually _ detected _ number of particles is strongly _ observer _ dependent , as opposed to the formally defined quantity @xmath110 , which refers to one particular quasiparticle state @xmath111 .
the detected number of particles depends , in particular , on how the detector actually couples to the field @xmath97 whose quanta it measures .
various couplings of the detector , for example to different powers of the fluid density , will influence the quasiparticle basis in which the detector measures , and thus the quasiparticle number detected .
now , the salient point is that because the phonon field @xmath97 is a _ relativistic _ quantum field , we will be able to map the observer dependence just described , which is general and holds for any sort of quasiparticles , to the observer dependence experienced by proper relativistic quantum fields in curved or flat space - time.@xcite the observer dependence is of kinematical origin , i.e. , it originates in the fact that the relativistic quantum field propagates in a space - time of lorentzian signature , cf .
the discussion of the hawking radiation analogue phenomenon in ref.@xcite , and is therefore fully within the capabilities of our proposed analogy . to model the quasiparticle analogue of expanding universes
, we will make use of expanding bose - einstein condensates , which are produced by reducing the trapping potential strength , i.e. the harmonic trapping frequency with time , or by increasing the interaction coupling constant .
firstly , it is thus appropriate to describe the evolution of density and velocity distribution in the expanding gas by discussing the so - called scaling procedure established in refs.@xcite .
the scaling procedure introduces a set of generally three scaling variables , @xmath112 , which are a function of time only . using these scaling variables ,
one writes for the ( cartesian ) co - ordinate vector components @xmath113 ; for the scaled co - ordinate vector we use the shorthand notation @xmath114 .
it may then be shown that the evolution of the bose - einstein - condensed gas cloud is described , starting from the initial density and velocity potential distributions @xmath115 , @xmath116 , by the following density and velocity distributions ( from here on we take @xmath117):@xcite @xmath118 this is true provided the scaling parameters @xmath119 fulfill the equations@xcite @xmath120 where @xmath121 are the initial trapping frequencies and the dimensionless `` scaling volume '' reads @xmath122 .
the dots are time derivatives with respect to laboratory time @xmath123 here and in what follows . , but only over greek indices @xmath124 .
] we have taken into account in the above equation that the particle interaction can be varied in lab time , @xmath125 , by means of a suitable feshbach resonance;@xcite @xmath126 is the initial coupling constant .
we designate `` scaling basis '' quantities with a tilde .
for example , the ( stationary ) initial density distribution @xmath127 gives us @xmath128 if we replace @xmath129 .
the scaling evolution is exact in a thomas - fermi approximation which neglects the quantum pressure term @xmath78 and the kinetic energy @xmath74 .
since bose - einstein condensates were experimentally created , the scaling solution has routinely been employed to interpret the time - of - flight pictures with which they are visualized as well as analyzed.@xcite we define a `` scaling time '' variable by @xmath130 and the @xmath131 dependent _ scaling functions _ @xmath132 by @xmath133 in terms of these quantities , the effective second order action for the scaling basis fluctuations of the phase of the superfluid order parameter @xmath134 takes on the particularly simple diagonal form@xcite @xmath135 , \label{kappaaction}\ ] ] where the scaling speed of sound @xmath136 . because of the fact that this action does not mix spatial and temporal derivatives , the resulting line element _ in the scaling variables _ , according to the identification displayed in ( [ action ] ) is diagonal ( does not possess @xmath47 terms ) , and reads @xmath137.\label{blineelement}\ ] ] we now consider for simplicity the isotropic case , @xmath138 , which implies @xmath139 the generalization of the mode expansion in eq.([homo ] ) to inhomogeneous expanding bose - einstein condensates then takes , in this isotropic case , the form@xcite @xmath140.\label{3dsol}\end{aligned}\ ] ] the functions @xmath141 are the stationary solutions of the gross - pitaevski equations for excitations above the initial ground state , designated by the ( set of ) quantum numbers @xmath142 with initial energies @xmath143 .
the initial thomas - fermi quantization volume is @xmath144 ; in the hard - walled cubic box limit the modes are plane waves , @xmath145 $ ] .
the temporal mode functions @xmath146 satisfy the second order ordinary differential equation@xcite @xmath147 the case when @xmath148 is a constant ( unity ) is particular . in this case
, the quantum state of the quasiparticle excitations remains unchanged for increasing @xmath131 , and a given initial quasiparticle vacuum , _ in the scaling basis _ with the associated quasiparticle operators defined by ( [ 3dsol ] ) , remains empty forever .
that is , no quasiparticles are created in that basis , although the superfluid may be in a highly nonstationary motional state , obtained by changing the trapping @xmath149 rapidly with time .
however , a detector which measures in a quasiparticle basis different from the scaling basis , for example due to its particular coupling to the superfluid , may still detect that quasiparticles are `` created '' .
we will come back to this possibility in section [ gibbonshawking ] below , when we discuss the purely choice - of - observer related phenomenon of a thermal state in a quasiparticle basis belonging to one particular space - time , the de sitter space - time of cosmology .
consider now the general case that the scaling function @xmath150 is a function of scaling time @xmath131 .
the fact that @xmath148 depends on time implies that the statement `` the excitation is of positive frequency '' ( a particle ) or of `` negative frequency '' ( antiparticle ) for a given propagating wave can not be held up for all times @xmath131 .
this _ frequency mixing _ implies that quasiparticles are created from the quasiparticle vacuum , because an initially empty scaling basis vacuum state does not remain empty during the evolution of the system , i.e. initially @xmath151 , but at a later stage @xmath152 .
the fact that annihilation and creation operator parts of the initial vacuum are mixed , as a consequence of ( [ bogotrafo1 ] ) , is physically due to the fact that quasiparticles are scattered within the course of time at ( time dependent ) effective potentials .
physically , the @xmath131 dependence of @xmath148 furnishes such an effective potential , at which excitations are scattered from negative to positive frequency and vice versa : the equation ( [ chin ] ) is formally equivalent to scattering of a non - relativistic particle with energy @xmath153 by a potential in @xmath131 space , @xmath154 at large @xmath131 , the wkb scattering solution of ( [ chin ] ) therefore reads:@xcite @xmath155 + \beta_{n}\exp\left[i\epsilon_{n}\int_{-\infty}^{\tau_s } { d\tau_s'}\sqrt{f(\tau_s')}\right]\right ) , \label{chineqs}\ ] ] where the scattering amplitudes are related via the particle flux conservation condition @xmath156 .
the quantity @xmath157 can be interpreted as the number of _ scaling basis _
quasiparticles created from the initially empty scaling vacuum , due to the time dependent scattering of excitations moving in the nonstationary condensate . in the wkb approximation , the amplitudes @xmath158 and @xmath159 are connected in a simple way : @xmath160\alpha_{n } , \label{alphat0beta}\ ] ] where the inverse temperature @xmath161 is given by the integral @xmath162 , \label{deft0}\ ] ] and @xmath163 is the contour in the complex @xmath131-plane enclosing the closest to the real axis singular point of the function @xmath164.@xcite this gives us the number of particles created in the mode @xmath142 , using the flux conservation condition @xmath156 , @xmath165 - 1}.\ ] ] the distribution of the created quasiparticles follows a thermal bosonic distribution ( planck spectrum ) , at a temperature @xmath166 .
the adiabatic evolution of trapped gases hence leads to `` cosmological '' quasiparticle creation with thermal occupation numbers in the scaling basis .
the temperature @xmath167 occurring in the planck distribution above depends on the details of the scaling evolution , i.e. , on the specific superfluid dynamics imposed by the solution of eqs.([genbeq ] ) , that is , @xmath166 is a functional of the temporal evolution @xmath168.@xcite we use the term `` cosmological '' in context with the plain condensed - matter fact that quasiparticles are created in the scaling basis .
we now justify this by comparing our effective bose - einstein condensate metric ( [ blineelement ] ) to line elements which constitute cosmological solutions of the einstein equations .
for example , let @xmath169 by properly adjusting @xmath125 , thus choosing the coupling constant s time evolution to be given by @xmath170 , cf .
eq.([deftau ] ) .
we then obtain that ( [ blineelement ] ) equals ( up to the conformal factor @xmath171 ) ) an anisotropic version of the spatially flat friedmann - robertson - walker ( frw ) universe:@xcite @xmath172 . \label{blineelementiso}\ ] ] in the standard spatially isotropic form of the frw metric , all @xmath173 are equal . to obtain the exact equivalence to a spatially flat frw metric
, we have to assume in addition that @xmath174 is spatially independent , which is fulfilled close to the center of the gas cloud , where the parabolic density profile is approximately flat , and @xmath174 is essentially a constant . in the spatially isotropic case
, @xmath175 is the hubble `` constant '' , which obviously is a constant only if the gas expands exponentially in laboratory time , @xmath176 $ ] , just as we need exponentially rapid expansion for a constant hubble parameter in inflationary cosmological models.@xcite we therefore come to the remarkable conclusion that the co - ordinate scaling factor @xmath177 of the bose - einstein condensate quasiparticle universe , occurring in the equations of motion of a nonrelativistic condensed matter system , has exactly the same meaning , in the quasiparticle world , as the cosmological scale factor of the universe proper .
we now treat the case that the scaling factors @xmath132 in ( [ deff ] ) are constants , i.e. , do not depend on scaling time . therefore , following eqs .
( [ 3dsol ] ) and ( [ chin ] ) , no scaling basis quasiparticles are created through negative and positive frequency mixing .
the superfluid can still be in highly nonstationary motion , though : the time evolution is according to ( [ deff ] ) prescribed by @xmath178 cf .
( [ deff ] ) , where the @xmath179 are constants .
however , although the superfluid is in motion , no dissipation through intrinsic quasiparticle creation takes place , because there exists the fock space `` scaling '' basis , in which no quasiparticles are created from the scaling vacuum , and the energy of that particular superfluid vacuum is conserved .
a particular instance is the isotropic 2d case , @xmath180 , where for constant @xmath181 the condition on @xmath148 being constant is fulfilled . in de sitter space - time space
is empty and flat ( that is , the constant time slices are euclidean space ) , and all of the curvature of space - time is encoded in a nonvanishing cosmological constant @xmath182 .
for obvious reasons , the de sitter space - time is very popular in the quantum field theoretical treatment of cosmological theories,@xcite because it highlights the crucial cosmological role the energy density of all conceivable quantum fields taken together might play : the vacuum energy density may constitute the dominant effective source term for space - time curvature in the einstein equations . the gibbons - hawking effect for geodesic observers@xcite in such a de sitter space - time is the curved space - time analogue of the unruh - davies effect.@xcite the latter consists in the fact that a constantly accelerated detector moving in the flat ( purportedly `` empty '' ) space - time vacuum , responds as if it were placed in a thermal bath of ( quasi-)particles with temperature proportional to its acceleration .
observer - related phenomena are at the heart of quantum field theory on nontrivial , and generally curved , space - time backgrounds .
they tell us , in particular , that the particle content of a given quantum field depends on the ( motional ) state of the detection apparatus , which is verifying that there are particles by its `` clicks '' .
more technically speaking , the dependence of the particle content of quantum fields in curved space - time is rooted in the non - uniqueness of canonical field quantization in riemannian spaces.@xcite it is of fundamental importance to make these observer - dependent effects measurable , because such a measurement constitutes , _ inter alia _ , a consistency check for an all - important concept of standard quantum field theory , namely that the quantization of a given field is carried out on a _ fixed _ space - time background .
that the experimental verification of the observer dependence is exceedingly difficult with light becomes readily apparent if we calculate the unruh - davies temperature : the result is that it equals @xmath183 a = 4\ , { \rm k } \times a[10^{20 } g_{\oplus } ] \,,$ ] where @xmath184 is the acceleration of the detector in minkowski space ( @xmath185 is the gravity acceleration on the surface of the earth ) , and @xmath186 the speed of light .
the huge accelerations needed to obtain measurable values of @xmath187 make it obvious that an observation of the effect with light ( photons ) is decidedly a less than trivial undertaking . although proposals for a measurement with ultraintense short pulses of electromagnetic radiation have been put forward in , e.g. , refs.@xcite , it is less than obvious how the thermal spectra associated to the effect , which still furnish tiny contributions to the total energy , should be discernible from the background dominated by the ultraintense lasers used to create large accelerations of the elementary particles contained in the plasma .
we first discuss the simplest case of an isotropically expanding gas in a harmonic trap .
we will see that this case also serves a pedagogical purpose , because it forces us to distinguish between purely observer - related phenomena with thermal spectrum , which by definition all have @xmath188 , and `` cosmological '' particle creation with a thermal spectrum , as defined in section [ cosmological ] , for which the scaling functions @xmath189 in ( [ deff ] ) depend explicitly on scaling time .
one can create de sitter universes in an expanding gas by letting it expand exponentially , @xmath190 $ ] where @xmath191 is the hubble constant of cosmological expansion , equalling @xmath192 for the de sitter universe discussed , where @xmath193 is the spatial dimension .
the de sitter metric in its standard form , using as its time co - ordinate the `` cosmological '' time @xmath194 , reads@xcite @xmath195,\label{desittercomoving}\ ] ] where @xmath196 is the central scaling ( i.e. , initial ) speed of sound . in the present case of exponential expansion , the `` cosmological '' time interval occurring in the metric above is equal to both the laboratory and the scaling time interval , @xmath197 ; cf .
eq.([blineelementiso ] ) with @xmath198 $ ] .
however , using the isotropic harmonic expansion setup , one has to face the difficulty that in order to give the frw metric ( [ blineelementiso ] ) near the center of the trap ( i.e. , close to @xmath199 ) the de sitter form , one has to increase exponentially the interaction with laboratory time : because of ( [ defh ] ) , @xmath200 = \exp[c_0 \lambda t]$ ] .
though the central density also decreases exponentially ( like @xmath201 ) , the exponential increase of the coupling `` constant '' incurs strong three - body recombination losses,@xcite whose total rate ( in the dilute gas case and in three spatial dimensions ) is proportional to @xmath202 $ ] .
therefore , within a short time of order @xmath203 , the bose - einstein - condensed gas of interacting single atoms will simply no longer be in existence , because the atoms rapidly form bound states .
such an experiment will leave no time to measure phenomena which depend on the fact that an equilibrium is established ; in particular , the thermal equilibrium for the occupation numbers in the de sitter quasiparticle basis will not be established on such a short time scale . even more importantly , though one obtains indeed a thermal spectrum in the de sitter quasiparticle basis corresponding to the metric ( [ desittercomoving ] ) , @xmath204 $ ] depends on time , i.e. , it is not a constant in the quasiparticle basis corresponding to the mode functions @xmath205 $ ] .
the thermal spectrum obtained therefore is , by our physical definition , the thermal `` cosmological '' quasiparticle creation discussed in the previous section .
it is _ not _ what we want to observe in our effective de sitter space - time , namely the purely choice - of - observer related phenomenon gibbons - hawking effect , for which no actual quasiparticle `` creation '' in the scaling basis should take place . to circumvent the problem that the interaction coupling needs to be increased exponentially with time to obtain a de sitter universe in 2 + 1 or 3 + 1 isotropic space - time dimensions in an isotropic harmonic trap ,
i and petr fedichev have developed the model of a 1 + 1d de sitter universe .
this 1 + 1d toy model can be realized in a strongly anisotropic , cigar - shaped bose - einstein condensate,@xcite cf .
fig.[ellipse ] .
in particular , in the proposed experimental setup , no time variation of the coupling constant at all is necessary , which is thus a true `` constant '' also in time .
the analysis of the excitation modes in a strongly anisotropic , elongated bose - einstein condensate is based on the adiabatic separation ansatz@xcite @xmath206 where @xmath207 is the radial wavefunction characterized by the quantum number @xmath142 ( only zero angular momentum modes are considered here ) .
the above ansatz incorporates the fact that for strongly elongated traps , i.e. , traps for which @xmath208 , the dynamics of the condensate motion separates into a fast radial motion and a slow axial motion , which are essentially independent .
the @xmath209 are the mode functions for travelling wave solutions in the @xmath210 direction ( plane waves for a condensate at rest read @xmath211 $ ] ) .
the radial motion is assumed to be `` stiff '' such that the radial part is effectively time independent , because the radial time scale for adjustment of the density distribution after a perturbation is much less than the axial oscillation time scales of interest .
the ansatz ( [ sepansatz ] ) works independent from the ratio of healing length and radial size of the superfluid cigar . in the limit that the healing length is much less than the radial size
, thomas - fermi wave functions are used , in the opposite limit , a gaussian ansatz for the radial part of the wave function @xmath207 is appropriate .
the squared oscillation spectrum of the cigar - shaped condensate cloud reads @xmath212 , where @xmath213.@xcite a de sitter space - time is then obtained from the effective action for the phase fluctuations of the phonon ( @xmath214 ) modes near the center of the trap .
the corresponding 1 + 1d action is obtained after integrating out the transverse , strongly confined directions , and reads:@xcite @xmath215 \nonumber\\ & \equiv & \frac12 \int d^{d+1}x % dt dz \sqrt{-{\sf g } } { \sf g}^{\mu\nu } \partial_\mu \chi_0 \partial_\nu \chi_0 \ , .
\label{1daction}\end{aligned}\ ] ] the scaling parameters @xmath177 in the axial ( @xmath216 ) and @xmath217 in the perpendicular ( @xmath218 ) directions are functions of time , such that the action fulfills the identification with the action of a scalar field minimally coupled to gravity , as indicated in the second line above , where the metric coefficients are the @xmath0 in eq.([gdown ] ) . the identification of the above phase - fluctuations action with the action of a scalar minimally coupled to gravity works , that is , the two actions in the first and second lines of ( [ 1daction ] ) are consistent , if we impose the consistency condition that @xmath219 where @xmath220 is a renormalization factor according to @xmath221 , with @xmath222 the renormalized wave function , and @xmath223 the initial central density .
the factor @xmath220 does not influence the ( classical ) equation of motion @xmath224 ( it simply drops out ) , but does influence the response of a detector .
we will come back to this point in section [ desitterdetect ] below , when we discuss the explicit dependence on @xmath220 of the equilibration time scale for the stationary detector state considered , see eq.([tauequil ] ) . in refs.@xcite , we were using an alternative form of the 1 + 1d de sitter metric ( [ desittercomoving ] ) .
this alternative form is the one used in the original gibbons - hawking paper.@xcite it reads ( we leave out the conformal factor ) @xmath225 the time interval @xmath226 in the above metric is not the `` cosmological '' time interval @xmath227 in the version of the metric displayed in ( [ desittercomoving ] ) .
the two metrics may be transformed into each other using the following co - ordinate transformations : @xmath228 & = & \exp[-2\sqrt \lambda c_0 \tau_c ] -\lambda z_b^2 , % \quad \leftrightarrow \quad \tau = \tau ( \tau_c , z_b ) \nonumber\\ z & = & z_b \exp[\lambda c_0 \tau_c ] , \label{coordtrafo } % \quad \leftrightarrow \quad z = z ( t , z_b),\end{aligned}\ ] ] giving us @xmath229 , @xmath230 , which transforms ( [ desitterlineelement ] ) into ( [ desittercomoving ] ) .
the advantage of the form ( [ desitterlineelement ] ) is that it is plain in this form that the de sitter space - time has an event horizon , located in our 1 + 1d case at the constant values @xmath231 .
the quantity @xmath232 in the de sitter metric ( [ desitterlineelement ] ) must be independent of time .
this leads us to the requirement @xmath233 relating the dynamical parameters of our problem to each other ( @xmath234 is the instantaneous sound velocity at the center of the cloud ) .
these relations imply that @xmath235 const . , and thus that @xmath236 .
then , the cosmological constant @xmath182 becomes independent of time , as is necessary for an analogue de sitter space - time to be established .
the experimental procedure now is determined to be as follows : prepare a thomas - fermi ( i.e. , sufficiently large ) , strongly anisotropically trapped , cigar - shaped condensate .
let it expand in the axial direction linear in lab time by changing the trapping according to the scaling equations ( [ genbeq ] ) , such that @xmath237 , _ and _ simultaneously expand in the perpendicular direction with the square root of lab time , @xmath238 , such that @xmath239 in ( [ cond1 ] ) is a constant .
then , a detector `` tuned '' to the de sitter space - time ( [ desitterlineelement ] ) , i.e. , which works in the de sitter quasiparticle basis , will measure a thermal quasiparticle spectrum , with the de sitter temperature:@xcite @xmath240 the fact that a thermal spectrum is obtained can be directly derived from the equations of superfluid hydrodynamics , as expounded in refs.@xcite .
that is , it is not simply postulated due to the ( kinematical ) analogy with quantum field theory in de sitter space - time , but is a result of the quantized hydrodynamic equations determining the evolution of the quasiparticle content of the sound field in the de sitter basis .
the relation between de sitter time @xmath241 and the laboratory time is fixed by @xmath242 .
[ note that @xmath226 and the scaling time interval @xmath243 defined in eq.([deftau ] ) differ ; @xmath244 . ]
the transformation law between @xmath123 and the de sitter time @xmath241 ( on a constant @xmath210 detector , such that @xmath245 ) , is given by @xmath246 , \label{trafo}\ ] ] where the unit of lab time @xmath247 is set by the initial conditions for the scaling variables @xmath177 and @xmath217 .
it is important to recognize that an effective exponential `` acceleration '' of the oscillation frequencies , either because of the exponential dependence of laboratory time on scaling time , represented by ( [ trafo ] ) , or coming from the wkb approximation for the scattering amplitudes , and the exponentially small mixing of positive and negative frequency parts resulting therefrom , eq.([alphat0beta ] ) , is sufficient for the thermal spectrum to be obtained . in particular , though we have discussed a space - time which possesses a horizon , no pre - existing horizons in the parent space - time are necessary _ per se _ for thermal occupation number distributions to be established , as has been pointed out in ref.@xcite .
an explicit example is the unruh - davies effect , where this parent , global space - time is simply minkowski space . to detect the gibbons - hawking effect in de sitter space - time , one has to set up a detector which measures frequencies in units of the inverse de sitter time @xmath241 , rather than in units of the inverse laboratory time @xmath123 ; this corresponds to measuring in the proper de sitter quasiparticle vacuum , where , in particular , positive and negative frequency are defined with respect to @xmath241 .
only then does one detect quasiparticles which are defined with respect to the de sitter quasiparticle basis , and refer to a vacuum corresponding to exactly that space - time .
in refs.@xcite , i and petr fedichev have provided such a detector .
we have shown that a `` de sitter basis '' detector is realized by an `` atomic quantum dot '' ( aqd ) ( for a detailed exposition of aqd properties cf .
ref.@xcite ) .
the aqd can be implemented in a bose gas of atoms possessing two hyperfine ground states @xmath248 and @xmath249 ; the level scheme is represented in fig.[aqd ] .
the atoms in the state @xmath248 represent the expanding bose - einstein condensate , and are used to model the expanding de sitter universe .
the aqd itself is formed by trapping atoms in the state @xmath249 in a tightly confining optical potential @xmath250 created by a laser at the center of the cloud .
the interaction of atoms in the two internal levels is described by a set of coupling parameters @xmath251 ( @xmath252 ) , where @xmath253 are the @xmath254-wave scattering lengths characterizing short - range intra- and inter - species collisions ; @xmath255 , @xmath256 , and @xmath257 .
the on - site repulsion between the atoms @xmath249 in the dot is given by the energy level spacing @xmath258 between states with a occupation difference of one @xmath249 atom , where @xmath259 is the characteristic size of the ground state wavefunction of atoms @xmath249 localized in @xmath250 .
we consider the so - called collisional blockade limit of large @xmath260 , where only one atom of type @xmath249 can be trapped in the dot .
this limit assumes that @xmath261 is much larger than all other relevant frequency scales in the dynamics of both the aqd and the expanding superfluid , and corresponds to a large `` coulomb blockade gap '' in electronic quantum dots.@xcite as a result of these assumptions , the collective co - ordinate of the aqd is modeled by a pseudo - spin-@xmath262 degree of freedom @xmath263 , with spin - up / spin - down state corresponding to occupation of the aqd by a single atom or no atom in the hyperfine state @xmath249 .
a rabi laser of frequency @xmath264 , with a detuning @xmath265 from resonance between the two hyperfine levels @xmath248 and @xmath249 , couples atoms of the hyperfine species @xmath248 , constituting the expanding cigar - shaped superfluid , into the aqd .
the detector lagrangian reads ( ref.@xcite , cf . the hamiltonian formulation in ref.@xcite ) : @xmath266 \bar \eta^ * \bar \eta\ , . \label{lbareta}\end{aligned}\ ] ] the fact that the detector has the de sitter basis as its `` natural '' quasiparticle basis , and therefore measures in de sitter time , is due to the fact that the term linear in the detector co - ordinate @xmath267 in the lagrangian couples in a certain manner to the superfluid cigar in which it is embedded .
namely , the laser with frequency @xmath264 , causing transitions between the two hyperfine levels @xmath248 and @xmath249 , couples to the _ square root _ of the central mean - field particle density . this particular coupling ( represented by the second term in the first line of the above lagrangian ) is what we need , because @xmath268 .
the fact that the coupling coefficient is proportional to @xmath269 is required to establish that the detector can work as a de sitter detector , because it transforms the detector equations , which are obviously _ a priori _ in laboratory time , into equations in de sitter time ; see the equations ( 43 ) for the temporal evolution of the detector level occupations below . adjusting the detuning @xmath265 properly , such that @xmath270 , the first and the second term in the square brackets in the second line of ( [ lbareta ] ) cancel .
one then obtains a simple set of coupled equations for the occupation amplitudes of the state @xmath271 of the aqd : @xmath272 where @xmath226 is the _ de sitter time interval .
_ were it not for the density oscillations in the cigar , represented by the potential @xmath273 , the above equations ( [ desitterevol ] ) would represent a simple two - level system , with frequency splitting @xmath274 .
the density oscillations contained in the perturbation operator @xmath275 cause transitions between the two undisturbed _ detector eigenstates _ @xmath276 of the two - level system , which are separated by the energy @xmath277 .
the density perturbations in the expanding host superfluid lead to a damping of the rabi oscillations with frequency @xmath277 between these two states .
this constitutes the effect of the de sitter thermal bath to be observed , where the damping happens on the time scale displayed in ( [ tauequil ] ) below .
the response of the detector , that is , the transition rates between the detector states , can be calculated by evaluating a response function@xcite which makes use of the expectation value of the product of two @xmath278 operators . the probability per unit time for excitation ( @xmath279 , transition from @xmath280 to @xmath281 ) and de - excitation ( @xmath282 , transition from @xmath283 to @xmath284 ) of the detector takes the form:@xcite @xmath285 the second - quantized solution of the hydrodynamic equations for the density fluctuations above the superfluid ground state in the expanding cigar - shaped bose - einstein condensate reads@xcite @xmath286 \right ) \ !
+ { \rm h.c . } \label{deltarho } % { \hat a}^\dagger_{k } \exp\left[i\int \frac{dt \epsilon_{0k}}{bb^2}- ikz % \right\ } \nonumber\\\end{aligned}\ ] ] using this solution , and inserting into ( [ detectorresponse ] ) , we have shown@xcite that at late times @xmath241 the transition probabilities per unit de sitter , i.e. , per unit detector time satisfy detailed balance conditions .
they correspond to _ thermodynamic equilibrium _ at the temperature @xmath287 displayed in ( [ tds ] ) : @xmath288 where the bose - planck distribution function takes the form familiar from thermostatistics : @xmath289 - 1]}.\ ] ] the frequency @xmath290 can be varied by changing the undressed rabi frequency @xmath264 , varying the intensity of the rabi laser .
the detector thus has a changeable and therefore _ tunable _ frequency standard , which can be adjusted to scan the above distribution function for a given de sitter temperature @xmath287 . from the relation ( [ occup ] ) , we come to the remarkable conclusion that a properly designed detector can `` see '' a thermal equilibrium distribution in its quasiparticle basis , though it is embedded in a highly nonstationary system with respect to the laboratory frame .
the rapidly expanding bose - einstein condensate represents this highly nonstationary system , which hosts the de sitter quasiparticle detector aqd .
i stress here again that ( [ occup ] ) is an exact _ result _ obtained by quantizing hydrodynamic fluctuations in a nonstationary superfluid , and not just concluded from a mere comparison of the phonon dynamics in our expanding superfluid with the quantum field theory of photons in de sitter space - time .
the equilibration time scale of the detector , and thus the time scale on which the rabi oscillations between the detector states are damped out , is set by the detector frequency standard ( the level spacing ) @xmath277 , and by the renormalization factor @xmath220:@xcite @xmath291 the renormalization factor @xmath220 contained in ( [ cond1 ] ) determines the equilibration rapidity because it physically expresses the strength of detector - field coupling .
it is related to the initial diluteness parameter @xmath292 of the bose - einstein condensate and to the ratio @xmath293 , which determines inasmuch the system is effectively one - dimensional , @xmath294 . to obtain sufficiently fast equilibration
, the condensate thus has to be initially not too dilute as well as close to the quasi-1d rgime , for which the transverse harmonic oscillator energy scale is of order the energy per particle , @xmath295 .
these two conditions have another important implication .
the ratio of the instantaneous coherence length @xmath296 and the location of the horizons @xmath231 , which are _ stationary _ in the present setup , has to remain less than unity within the equilibration time scale .
if this is not the case , the coherence length , which plays the role of the analogue planck scale ( which is time dependent here ) , exceeds the length scale of the horizon at equilibration , and the concept of `` relativistic '' phonons propagating on a fixed curved space - time background with local lorentz symmetry becomes invalid .
the ratio @xmath297 at the lab equilibration time scale @xmath298 $ ] following from the de sitter equilibration time in eq.([tauequil ] ) , expressed in parameters relevant to the experiment , is given by @xmath299.\ ] ] we see that this ratio changes exponentially with both the initial diluteness parameter @xmath300 and the quasi-1d parameter @xmath293 . in
most currently realized bose - einstein condensates,@xcite the diluteness parameter @xmath301 . here , we initially need @xmath302 to have the condition @xmath303 fulfilled , assuming a reasonably large value of the de sitter temperature @xmath287 . though the condensate has to be _ initially
_ quite dense , it is to be stressed that the central density decays like @xmath304 during expansion .
therefore , the rate of three - body recombination losses quickly decreases during the expansion of the gas , and the initially relatively dense bose - einstein condensate , which would rapidly decay if left with a @xmath305 close to unity , can live sufficiently long , the total rate of three - body losses decreasing like @xmath306 .
the primary statement to be drawn from the present article is that phonons , i.e. , low - energy linear - dispersion quasiparticles , moving in a spatially and temporally inhomogeneous bose - einstein - condensed superfluid gas , are equivalent to photons , the quanta of the electromagnetic field , moving on geodesics in a given curved space - time .
we have explored the classical as well as the quantum aspects of this statement . on the classical side
, the analogy helps to provide us with a simple general means to study quasiparticle propagation in an inhomogeneous medium in motion .
an example of such an application is the gravitational lensing effect exerted by a superfluid vortex.@xcite on the quantum field theoretical side , we can access within the analogy phenomena which are extremely difficult if not impossible to access with light .
one of the fundamentals of quantum field theory , the fact that the particle content of a quantum field depends on the observer , can thus be experimentally verified for the first time .
the basic reason that the phenomena in question are ( comparatively ) easy to simulate in a condensed matter system is that the energy and temperature scales , under which they occur , relative to the typical energy scales of the system , can be changed at will by the experimentalist in a very controlled manner .
more particularly , the temperature of the thermal spectrum of phonons to be measured in the gibbons - hawking effect can be made relatively large compared to the axial phonon frequencies and the actual temperature of the gas itself , by expanding the condensate cloud rapidly enough . in the condensed matter analogue ,
the typical energy of the quanta produced can in principle be even made to approach the relevant `` planck '' scale , i.e. , the point in the quasiparticle energy spectrum where it begins to deviate from being linear .
finally , on a more adventurous side , one could conceive of carrying out experiments in `` experimental '' cosmology , as opposed to the currently existing purely `` observational '' cosmology .
in such an experimental approach to matters cosmological , one would try to reproduce under certain specified and , in particular , well - defined initial conditions large - scale features of the cosmos , in the laboratory setting of nonstationary , inhomogeneous superfluid gases .
misner , k.s .
thorne , and j.a .
wheeler , _ gravitation _ ( freeman , san francisco , 1973 ) .
j.r . anglin and w. ketterle ,
nature * 416 * , 211 ( 2002 ) .
khalatnikov , _ an introduction to the theory of superfluidity _
( addison wesley , reading , ma , 1965 ) .
n.d . birrell and p.c.w .
davies , _ quantum fields in curved space _
( cambridge university press , cambridge , england , 1984 ) .
h. minkowski , physik . zeitschr . * 10 * , 104 ( 1909 ) .
unruh , phys .
* 46 * , 1351 ( 1981 ) . m. visser , class .
quantum grav .
* 15 * , 1767 ( 1998 ) .
a. trautman , _ comparison of newtonian and relativistic theories of spacetime _ , in : `` perspectives in geometry and relativity '' ( indiana university press , bloomington , 1966 ) .
fedichev and u.r .
fischer , phys .
rev . lett . * 91 * , 240407 ( 2003 ) .
fedichev and u.r .
fischer , phys .
d * 69 * , 064021 ( 2004 ) .
m. visser , gr - qc/0309072 .
u.r . fischer and m. visser , ann
( n.y . ) * 304 * , 22 ( 2003 ) .
fischer and m. visser , phys .
* 88 * , 110201 ( 2002 ) .
fischer and m. visser , europhys .
62 * , 1 ( 2003 ) .
m. alcubierre , class .
quantum grav .
* 11 * , l73 ( 1994 ) .
olum , phys .
81 * , 3567 ( 1998 ) . c. barcel , s. liberati , and m. visser , class .
quantum grav . * 18 * , 3595 ( 2001 ) .
leggett , rev .
* 73 * , 307 ( 2001 ) ; _ ibid . _ * 75 * , 1083(e ) ( 2003 ) .
e. madelung , z. phys .
* 40 * , 322 ( 1927 ) .
l. onsager , nuovo cimento suppl .
* 6 * , 279 ( 1949 ) .
perez bergliaffa , k. hibberd , m. stone , and m. visser , physica d * 191 * , 121 ( 2004 ) .
bogoliubov , _ selected works , part ii : quantum and statistical mechanics _
( gordon and breach , new york , 1991 ) . c. kittel , _ quantum theory of solids _ ( wiley and sons , new york , 2nd revised printing 1987 ) .
volovik , _ the universe in a helium droplet _
( oxford university press , oxford , 2003 ) .
e.m . lifshitz and l.p .
pitaevski , _ statistical physics , part 2 _ ( pergamon press , new york , 1980 ) .
unruh , phys .
d * 14 * , 870 ( 1976 ) .
s.w . hawking and g.f.r .
ellis , _ the large scale structure of space - time _ ( cambridge university press , cambridge , england , 1973 ) .
gibbons and s.w .
hawking , phys .
d * 15 * , 2738 ( 1977 ) .
m. visser , phys .
* 80 * , 3436 ( 1998 ) .
fulling , phys .
d * 7 * , 2850 ( 1973 ) .
e. yablonovitch , phys . rev
. lett . * 62 * , 1742 ( 1989 ) .
p. chen and t. tajima , phys .
lett . * 83 * , 256 ( 1999 ) .
kagan , e.l .
surkov , and g.v .
shlyapnikov , phys .
a * 54 * , r1753 ( 1996 ) .
y. castin and r. dum , phys .
* 77 * , 5315 ( 1996 ) .
kagan , e.l .
surkov , and g.v .
shlyapnikov , phys .
lett . * 79 * , 2604 ( 1997 ) .
s. inouye _
et al . _ ,
nature * 392 * , 151 ( 1998 ) .
lett . * 89 * , 283202 ( 2002 ) . c. barcel , s. liberati , and m. visser , phys .
a * 68 * , 053613 ( 2003 ) .
fedichev and u.r .
fischer , phys .
a * 69 * , 033602 ( 2004 ) .
landau , e.m .
lifshitz , and l.p .
pitaevski , _ quantum mechanics : non - relativistic theory _
( 3rd edition , butterworth - heinemann , newton , ma , 1981 ) .
s. weinberg , _ gravitation and cosmology _ ( wiley and sons , new york , 1972 ) .
guth , phys .
d * 23 * , 347 ( 1981 ) .
lidsey , a.r .
liddle , e.w .
kolb , e.j .
copeland , t. barreiro , and m. abney , rev .
phys . * 69 * , 373 ( 1997 ) .
l.p . pitaevski and a. rosch , phys .
a * 55 * , r853 ( 1997 ) . f. chevy ,
v. bretin , p. rosenbusch , k.w .
madison , and j. dalibard , phys .
* 88 * , 250402 ( 2002 ) .
fedichev , m.w .
reynolds , and g.v .
shlyapnikov , phys .
lett . * 77 * , 2921 ( 1996 ) .
e. zaremba , phys . rev . a * 57 * , 518 ( 1998 ) .
b.l . hu and a. raval , mod .
phys . lett .
a * 11 * , 2625 ( 1996 ) .
a. recati , p.o .
fedichev , w. zwerger , j. von delft , and p. zoller , cond - mat/0404533 .
s.m . reimann and m. manninen , rev .
phys . * 74 * , 1283 ( 2002 ) . | recent developments in simulating fundamental quantum field theoretical effects in the kinematical context of analogue gravity are reviewed .
specifically , it is argued that a curved space - time generalization of the unruh - davies effect the gibbons - hawking effect in the de sitter space - time of inflationary cosmological models can be implemented and verified in an ultracold gas of bosonic atoms . |
studies on the decoherence dynamics of open quantum systems are of great importance to the field of quantum information science @xcite . any realistic analysis of quantum information protocols should take into account the decoherence effect of the environment . in many quantum communication and computation schemes , information is transmitted using photons . for instance , the first experimental verification of quantum teleportation @xcite used pairs of polarization entangled photons to transfer the polarization state of one photon onto another @xcite . within a year , unconditional quantum teleportation of optical fields was demonstrated experimentally using squeezed - state entanglement @xcite .
given their central role in these schemes and many others , much work has been carried out on the decoherence dynamics of optical fields . in particular ,
several authors have studied the continuous variable entanglement of optical fields ( see , for instance , refs .
@xcite ) .
conventional approaches not only treat the interactions between the quantum system @xmath0 of interest and its environment @xmath1 perturbatively , they also assume that the environmental correlation time @xmath2 is small compared to the time scale @xmath3 for significant change in @xmath0
. these yield approximate equations of motion , i.e. master equations , under the born - markov approximation @xcite . indeed , many studies on the entanglement dynamics of continuous variable system relied on this approximation @xcite .
however , it is evident from recent experiments ( see , for instance , refs .
@xcite ) , that there are many physically relevant situations where the markovian assumption does not hold , and a non - markovian treatment of the open system dynamics is necessary .
so , there has been an increasing interest in the understanding of the decoherence effect of open quantum system going beyond the born - markovian approximation in the last decades @xcite .
very recently , some phenomenological models on non - markovian entanglement dynamics of optical fields have been investigated @xcite .
it was found that in contrast to the monotonic decrease of entanglement over time in born - markovian entanglement dynamics @xcite , there are transient entanglement oscillations in non - markovian ones .
these oscillations are caused by the backactions of the environments on their respective local quantum systems @xcite .
the backaction , characteristic of non - markovian dynamics , means that the environments with their states changed due to interactions with the systems , in turn , exert their dynamical influences back on the systems . in this paper
we consider the exact decoherence dynamics of the continuous variable entangled squeezed state of two single - mode optical fields , @xmath4 and @xmath5 , that are spatially separated .
each optical field , @xmath6 , interacts with its own environment @xmath7 ( @xmath8 ) . @xmath9 and @xmath10 are independent and uncorrelated .
we study the exact entanglement dynamics of the two optical - field system for different @xmath2 s in comparsion with @xmath3 , and analyze when the system dynamics will exhibit novel non - markovian effects , and provide a detailed description of these . to this end
, we use the influence functional formalism @xcite , developed explicitly in refs .
our results show that besides the short - time oscillations , the non - markovian effect can affect the long - time behavior of the system dynamics and the steady state as well . in particular , when @xmath2 is comparable to @xmath3 , we find that the backaction effects counteract the dissipative effects of @xmath11 on @xmath12 respectively .
this leads to there being some nonzero residual entanglement in the steady state .
our paper is organized as follows . in sec .
ii , we introduce a model of two single - mode optical fields in two independent and uncorrelated environments ; and outline the exact dynamics that was derived in detail in ref . @xcite . in sec .
iii , using logarithmic negativity as an entanglement measure of continuous - variable states , we discuss the entanglement dynamics of the entangled squeezed state .
iv presents the numerical results of the entanglement dynamics , where we analyze explicitly the non - markovian effect of the environments on the system for different @xmath2 s in comparison with @xmath3 .
finally , we conclude in sec . v.
the total hamiltonian of the system @xmath12 plus environment @xmath11 is given by @xmath13 where @xmath14 are , respectively , the hamiltonian of the two optical fields , the two independent environments , and the interactions between them . the operators @xmath15 and @xmath16 ( @xmath8 ) are respectively the annihilation and creation operators of the @xmath17-th optical mode with frequency @xmath18 .
the two independent environments are modeled , as usual , by two sets of harmonic oscillators described by the annihilation and creation operators @xmath19 and @xmath20 .
the coupling constants between the @xmath17-th optical field and its environment are given by @xmath21 .
currently , most quantum optical experiments are performed at low temperatures and under vacuum condition . in this case ,
vacuum fluctuations are the main source of decoherence .
therefore , we take the environments to be at zero temperature throughout this paper . since we are only interested in the dynamics of @xmath12 , we like to eliminate the degrees of freedom of @xmath11 .
the influence - functional theory of feynman and vernon @xcite enables us to do that exactly . by expressing the forward and backward evolution operators of the density matrix of the system @xmath12 plus environment @xmath11 as a double path integral in the coherent - state representation @xcite , and performing the integration over the degrees of freedom of @xmath11
, we incorporate all the environmental effects on @xmath12 in a functional integral named influence functional @xcite .
the reduced density matrix , which fully describes the dynamics of @xmath12 is given by @xmath22where @xmath23 is the reduced density matrix expressed in coherent - state representation and @xmath24 is the propagating function . in the derivation of eq .
( [ rout ] ) , we have used the coherent - state representation @xmath25 which are the eigenstates of annihilation operators , i.e. @xmath26 and obey the resolution of identity , @xmath27 with the integration measures defined as @xmath28 .
@xmath29 denotes the complex conjugate of @xmath30 .
the time evolution of the reduced density matrix is determined by the propagating function @xmath31 .
the propagating function is expressed as the path integral governed by an effective action which consists of the free actions of the forward and backward propagators of the optical - field system and the influence functional obtained from the integration of environmental degrees of freedom .
after evaluation of the path integral , the final form of the propagating function is obtained as follows @xmath32\bar{\alpha}_{ki}^{\prime } \alpha _ { ki}% \big]\big\ } , \label{prord}\end{aligned}\ ] ] where @xmath33 satisfies @xmath34 with @xmath35 being used . combining eq .
( [ prord ] ) , we can get the exact time - dependent state from any initial state by the evaluation of the integration in eq .
( [ rout ] ) . to compare with the conventional master equation description of such system
, we now derive a master equation from the above results . after taking the time derivative to eq .
( [ rout ] ) and recalling the explicit form of eq .
( [ prord ] ) , we can derive an exact master equation @xmath36+\sum_{k=1 } ^2\gamma_k ( t)[2a_{k}\rho ( t)a_{k}^{\dag } \notag \\ & & -a_{k}^{\dag } a_{k}\rho ( t)-\rho ( t)a_{k}^{\dag } a_{k } ] , \label{mas}\end{aligned}\ ] ] where @xmath37 is the modified hamiltonian of the two optical modes and @xmath38 eq . ( [ mas ] ) is the exact master equation for the optical - field system .
@xmath39 plays the role of a time - dependent shifted frequency of the @xmath17-th optical field .
@xmath40 represents the corresponding time - dependent decay rate of the field .
we emphasize that the derivation of the master equation goes beyond the born - markovian approximation and contains all the backactions between the system and the environments self - consistently .
all the non - markovian character resides in the time - dependent coefficients of the exact master equation .
the time - dependent coefficients in the exact master equation , determined by eq .
( [ pa ] ) , essentially depend on the so - called spectral density , which characterizes the coupling strength of the environment to the system with respect to the frequencies of the environment .
it is defined as @xmath41 . in the continuum limit
the spectral density may have the form @xmath42 where @xmath43 is an exponential cutoff frequency , and @xmath44 is a dimensionless coupling constant between @xmath6 and @xmath7 . the environment is classified as ohmic if @xmath45 , sub - ohmic if @xmath46 , and super - ohmic for @xmath47 @xcite .
different spectral densities manifest different non - markovian decoherence dynamics .
we note that our exact master equation reduces to the conventional master equation under the relevant markov approximation .
the coefficients in the master equation ( [ mas ] ) become time - independent @xcite @xmath48 where @xmath49 denotes the cauchy principal value .
the coefficients in eqs .
( [ mp ] ) are precisely the corresponding ones in the markovian master equation of the optical system @xcite .
initially at time @xmath50 , @xmath12 is in an entangled squeezed state .
the entangled two - mode squeezed state is defined as the vacuum state acted on by the two - mode squeezing operator @xmath51 is the squeezing parameter . in the coherent - state representation , this initial state is given by @xmath52}{\cosh^2r}.\ ] ] the state approaches the ideal einstein - podolsky - rosen ( epr ) state @xcite in the limit of infinite squeezing ( @xmath53 ) .
the traditional way to generate the entangled two - mode squeezed state is via the nonlinear optical process of parametric down - conversion @xcite .
recently , a microwave cavity qed - based scheme to generate such states has also been proposed @xcite . after generating the entangled state given by eq .
( [ initial ] ) , the two cavity fields are then propagated , respectively , to the two locations separated between the sender and the receiver .
a quantum channel is thus established through the entangled two - mode squeezed state and is ready for teleporting unknown optical coherent states @xcite . at @xmath54 , due to interactions with @xmath11 , @xmath55 evolves to a mixed state .
a straightforward way to obtain the time - dependent mixed state is by integrating the propagating function over the initial state of eq .
( [ rout ] ) .
then the time - evolution solution of the reduced density matrix can be obtained exactly , @xmath56 , \label{final}\ ] ] where @xmath57 } , \\ b & = & \frac{-\tanh \left\vert r\right\vert u(t)^{2}}{1-\tanh ^{2}\left\vert r\right\vert ( 1-\left\vert u(t)\right\vert ^{2})^{2 } } , \\ c & = & \frac{\tanh ^{2}\left\vert r\right\vert ( 1-\left\vert u(t)\right\vert ^{2})\left\vert u(t)\right\vert ^{2}}{1-\tanh ^{2}\left\vert r\right\vert ( 1-\left\vert u(t)\right\vert ^{2})^{2}}.\end{aligned}\ ] ] to measure the entanglement in continuous variable system , one generally uses the logarithmic negativity @xcite .
the logarithmic negativity of a bipartite system was introduced originally as @xmath58 where @xmath59 is the negative eigenvalue of @xmath60 , and @xmath60 is a partial transpose of the bipartite state @xmath61 with respect to the degrees of freedom of the @xmath62-th party .
this measure is based on the peres - horodecki criterion @xcite that a bipartite quantum state is separable if and only if its partially transposed state is still positive .
for the continuous - variable ( gaussian - type ) bipartite state , its density matrix is characterized by the covariance matrix defined as the second moments of the quadrature vector @xmath63 , @xmath64 where @xmath65 , and @xmath66 , @xmath67 .
the canonical commutation relations take the form as @xmath68 = iu_{ij}$ ] , with @xmath69 and @xmath70 defining the symplectic structure of the system .
the property of the covariance matrix @xmath71 is fully determined by its symplectic spectrum @xmath72 , with @xmath73 ( @xmath74 ) the eigenvalues of the matrix : @xmath75 .
the uncertainty principle exerts a constraint on @xmath76 such that @xmath77 @xcite .
thus the peres - horodecki criterion for the continuous - variable state can be rephrased as the state being separable if and only if the uncertainty principle , @xmath78 , is still obeyed by the covariance matrix under the partial transposition with respect to the degrees of freedom of a specific subsystem @xcite . in terms of phase space ,
the action of partial transposition amounts to a mirror reflection with respect to one of the canonical variables of the related subsystem .
for instance , @xmath79 , and @xmath80 is the partial transposition with respect to the second subsystem .
if a gaussian - type bipartite state is nonseparable , the covariance matrix @xmath81 will violate the uncertainty principle and its symplectic spectrum @xmath82 will fail to satisfy the constraint @xmath83 .
the logarithmic negativity is then used to quantify this violation as @xcite @xmath84 where @xmath85 is the smaller one of the two symplectic eigenvalues @xmath86 .
it is evident from eq .
( [ measu ] ) that , if @xmath81 obeys the uncertainty principle , i.e. , @xmath83 , then @xmath87 , namely , the state is separable .
otherwise , it is entangled .
therefore , the symplectic eigenvalue @xmath85 encodes a qualitative feature of the entanglement for an arbitrary continuous - variable bipartite state .
with this entanglement measure at hand , we study now the entanglement dynamics of the squeezed - state quantum channel in our model . from the time - dependent state ,
the covariance matrix for the optical field can be calculated straightforwardly , @xmath88}{x } & \frac{a\text{im}[b]% } { x } \\ 0 & \frac{y(1+d)}{2(1-d)^{2 } } & \frac{a\text{im}[b]}{x } & \frac{-a\text{re}% [ b]}{x } \\ \frac{a\text{re}[b]}{x } & \frac{a\text{im}[b]}{x } & \frac{y(1+d)}{2(1-d)^{2 } } & 0 \\ \frac{a\text{im}[b]}{x } & \frac{-a\text{re}[b]}{x } & 0 & \frac{y(1+d)}{%
2(1-d)^{2}}% \end{array}% \right ) , \]]where @xmath89^{2}$ ] , @xmath90 , and @xmath91 . and
the logarithmic negativity @xmath92 can also be obtained exactly from eq .
( [ measu ] ) .
it is easy to verify that the initial entanglement is @xmath93 .
in the following , we analyze explicitly the exact decoherence dynamics of the entangled squeezed state of @xmath12 under the influence of @xmath11 . for simplicity , we assume from here on that the two optical fields are identical , i.e. , @xmath94 ; and they interact with the same strength , @xmath95 , with their individual environments . for definiteness , we consider both @xmath9 and @xmath10 to have ohmic spectral density .
the environmental correlation time @xmath2 in this case is roughly inversely proportional to the cutoff frequency @xmath96 in eq .
( [ spectral ] ) , i.e. , @xmath97 @xcite .
it is emphasized that the cutoff frequency @xmath98 , which is originally introduced to eliminate infinities in frequency integrations , therefore also determines if the dynamics of open system @xmath0 is markovian or non - markovian .
our non - perturbatively derived exact results allow us to explore all these possibilities . in fig .
[ f1 ] , we plot the numerical results of the decay rate @xmath99 and logarithmic negativity @xmath92 when @xmath100 .
the positivity of @xmath99 throughout the whole evolution process guarantees the monotonic decrease of @xmath92 .
accordingly , the entangled squeezed state eventually evolves to a product state , namely the ground state of the system : @xmath101 .
clearly , in this case , the backactions of @xmath11 have a negligible effect on the dynamics of @xmath12 , and we say the system dynamics is mainly governed by the dissipative effect of the environments .
there is thus no qualitative difference between the exact entanglement dynamics and the markovian results .
quantitatively , however , we note that for @xmath102 , the distinctive increase of @xmath99 results in @xmath92 decreasing rapidly .
this non - markovian effect only shows up in a very short time scale .
in fact , for @xmath103 , @xmath99 decreases and approaches gradually to a constant value as @xmath104 approaches @xmath3 ; and the rate of decrease of @xmath92 decreases .
[ f2 ] shows @xmath99 and @xmath92 when @xmath105 . in this case , the backactions of @xmath11 have a considerable impact on the dynamics of @xmath12 , and the markovian approximation is not applicable .
firstly , we note that @xmath99 can take negative values . physically , this corresponds to the systems reabsorbing photons from the environments , which leads to an increase in the photon number of the systems @xcite .
these negative decay rates provide evidences for backactions in non - markovian dynamics @xcite .
secondly , we observe that @xmath99 approaches zero asymptotically .
both results clearly differ from the markovian ones .
consequently , @xmath92 presnts distinctive behaviors that are absent in the markovian results .
firstly , due to the negative decay rates , @xmath92 shows oscillations . we must emphasize that these oscillations are fundamentally different from the transient entanglement oscillations previously obtained when two optical - fields interact with a common environment @xcite .
they are caused by the backactions of the environments on their respective local optical fields and are characteristic of non - markovian dynamics .
similar oscillations have been obtained in a system of two two - level atoms in two separated damping cavities @xcite . secondly , and more interestingly
, we find that there is some residual entanglement left in the steady state . from previous studies @xcite ,
one would have concluded that non - markovian effects only show up in short - time dynamics .
our results , however , clearly show on the contrary that non - markovian effects can also have an influence on the long - time behavior of the system dynamics and the final steady state of the system .
this counterintuitive behavior can be explained with the fact that the dissipative influence on the entanglement dynamics by the environments is strongly counteracted by the effect due to their backactions .
consequently , the decay of the entanglement ceases when the system evolves to some steady state , which is not the ground state @xmath106 .
the results when @xmath107 are shown in fig .
[ f3 ] ; a situation considered in ref .
@xcite . due to extremely long memory of the environments ,
the backactions on the systems are so strong that they govern the decoherence dynamics . as a result , @xmath99 and
hence @xmath92 oscillate over a very long duration .
these oscillations persist even as the state approaches the ground state .
the ` equilibrium ' position for the oscillation of @xmath99 is not at zero , but a small positive value .
this positivity means the systems dynamics will experience a weak dissipation , which is verified by the time evolution of @xmath92 in fig .
[ f3 ] . in summary
, we have studied the exact entanglement dynamics of the two optical - field system for different @xmath2 s in comparison with @xmath3 .
specifically , we have analyzed when the system dynamics will exhibit novel non - markovian effects , and provided a detailed description of these .
we have applied the influence - functional method of feynman and vernon to investigate the exact entanglement dynamics of two single - mode optical fields @xmath12 coupled to two independent and uncorrelated environments @xmath11 . from our analytical and numerical results
, it is seen that @xmath11 exert two competing influences on our system .
one effect , @xmath108 , is dissipative and is responsible for the decoherence of @xmath12 .
the other , @xmath109 , is due to the backactions of @xmath11 on @xmath12 .
the degree of manifestations of @xmath108 and @xmath109 in the dynamics of @xmath12 depends on @xmath2 in comparison with @xmath3 . for @xmath100 ,
@xmath108 dominates and @xmath109 only gives rise to a transient coherent oscillation of @xmath12 .
the state of @xmath12 evolves to the ground state @xmath106 , which is coincident with the markovian result . if @xmath105 , the near resonant interaction between @xmath12 and @xmath11 results in @xmath108 and @xmath109 being comparable and counteract each other .
these give rise to transient negative decay rates and asymptotical zero decay rate .
the state of @xmath12 thus evolves asymptotically to some steady state , which is not the ground state @xmath106 . finally , when @xmath107 , @xmath109 dominates and governs the dynamics of @xmath12 .
the decay rates of the system oscillate about some non - negative equilibrium position over a very long duration .
this slight positivity guarantees an overall weak dissipative effect on the system dynamics .
therefore , the state of @xmath12 eventually approaches the ground state with the entanglement oscillation persisting on for a very long time .
the theory we have established is a non - perturbative description of the exact decoherence dynamics of a system of two single - mode optical fields .
it can serve as a useful basic theoretical model in analyzing the non - markovian decoherence dynamics of optical fields employed in practical quantum information processing schemes .
it should be noted that although only the ohmic spectral density is considered here , it is straightforward to generalize our discussion to the non - ohmic cases .
the work is supported by nus research grant no .
r-144 - 000 - 189 - 305 .
j.h.a . also thanks the financial support of the nnsf of china under grant no .
10604025 , and the fundamental research fund for physics and mathematics of lanzhou university under grant no .
lzu05 - 02 .
99 m. a. nielsen and i. l. chuang , _ quantum computation and quantum information _
( cambridge university press , cambridge , u.k . , 2000 ) . c. h. bennett , g. brassard , c. crepeau , r. jozsa , a. peres , and w. k. wootters , phys .
lett . * 70 * , 1895 ( 1993 ) .
d. bouwmeester , j .- w .
pan , k. matter , m. eibl , h. weinfurter , and a. zeilinger , nature * 390 * , 575 ( 1997 ) . s. l. braunstein and h. j. kimble , phys .
lett . * 80 * , 869 ( 1998 ) .
a. furusawa , j. l. sorensen , s. l. braunstein , c. a. fuchs , h. j. kimble , and e. s. polzik , science * 282 * , 706 ( 1998 ) .
j. s. prauzner - bechcicki , j. phys .
a : math . gen . * 37 * , l173 ( 2004 ) .
an , s .- j .
wang , and h .-
luo , j. phys .
* 38 * , 3579 ( 2005 ) .
r. rossi jr . , a. r. bosco de magalhes , m. c. nemes , physica a * 365 * , 402 ( 2006 )
. m. ban , j. phy .
a : math . gen . * 39 * , 1927 ( 2006 ) .
liu and h .- s .
goan , phys .
a * 76 * , 022312 ( 2007 ) .
s. maniscalco , s. olivares , and m. g. a. paris , phys . rev .
a * 75 * , 062119 ( 2007 ) .
h . an and w. m. zhang , phys . rev .
a * 76 * , 042127 ( 2007 ) .
h. j. carmichael , _ an open systems approach to quantum optics _
, lecture notes in physics , vol .
m18 ( springer - verlag , berlin , 1993 ) .
breuer and f. petruccione , _ the theory of open quantum systems _
( oxford university press , oxford , 2002 ) .
f. dubin , d. rotter , m. mukherjee , c. russo , j. eschner , and r. blatt , phys .
lett . * 98 * , 183003 ( 2007 ) .
f. h. l. koppens , d. klauser , w. a. coish , k. c. nowack , l. p. kouwenhoven , d. loss , and l. m. k. vandersypen , phys .
99 * , 106803 ( 2007 ) .
d. mogilevtsev , a. p. nisovtsev , s. kilin , s. b. cavalcanti , h. s. brandi , and l. e. oliveira , phys .
lett . * 100 * , 017401 ( 2008 ) . | we study the exact decoherence dynamics of the entangled squeezed state of two single - mode optical fields interacting with two independent and uncorrelated environments .
we analyze in detail the non - markovian effects on the entanglement evolution of the initially entangled squeezed state for different environmental correlation time scales .
we find that the environments have dual actions on the system : backaction and dissipation .
in particular , when the environmental correlation time scale is comparable to the time scale for significant change in the system , the backaction would counteract the dissipative effect .
interestingly , this results in the survival of some residual entanglement in the final steady state . |
the observed direct cp asymmetries and branching ratios of the @xmath8 , @xmath9 decays @xcite , @xmath10 were regarded as puzzles , since they obviously contradict to the expected relations @xmath11 and @xmath12 .
these puzzles have been analyzed in the perturbative qcd ( pqcd ) approach @xcite up to next - to - leading - order ( nlo ) accuracy recently @xcite , where the contributions from the vertex corrections , the quark loops , and the magnetic penguin were taken into account .
it was found that the vertex corrections modify the color - suppressed tree contribution , such that the relative strong phase between the tree and penguin amplitudes involved in the @xmath13 decays decreases .
the predicted magnitude of the @xmath14 direct cp asymmetry then becomes smaller , and matches the data in eq .
( [ data ] ) . though the @xmath13 puzzle has been resolved , the @xmath1 puzzle remains , because the nlo color - suppressed tree amplitude does not increase the predicted @xmath2 branching ratio sufficiently .
a resolution to a puzzle usually demands an introduction of new mechanism .
it is thus essential to investigate whether the proposed new mechanism deteriorates the consistency of theoretical results with other data . to make sure the above nlo effects are reasonable
, we apply the same pqcd formalism to more two - body nonleptonic @xmath15 meson decays , concentrating on the @xmath0 branching ratios , which are also sensitive to the color - suppressed tree contribution .
it will be shown that the nlo pqcd predictions are in agreement with the data of the @xmath4 and @xmath5 branching ratios , and saturate the experimental upper bound of the @xmath3 branching ratio , @xmath16 @xcite .
therefore , our resolution to the @xmath13 puzzle makes sense , and the @xmath1 puzzle is confirmed .
the dramatic difference between the @xmath1 and @xmath17 data has been also noticed in @xcite , which stimulates the proposal of a new isospin amplitude with @xmath18 .
the possible new physics signals from the @xmath1 decays have been discussed in @xcite . it has been claimed that the @xmath1 puzzle is resolved in the qcd - improved factorization ( qcdf ) approach @xcite with an input from soft - collinear effective theory ( scet ) @xcite : the inclusion of the nlo jet function , one of the hard coefficients of scet@xmath19 , into the qcdf formula for the color - suppressed tree amplitude leads to enough enhancement of the @xmath2 branching ratio . following the argument made above
, we apply the same formalism @xcite to the @xmath13 , @xmath17 decays as a check .
it turns out that the effect of the nlo jet function deteriorates the qcdf results for the direct cp asymmetries in the @xmath6 and @xmath7 decays : the magnitude of the former increases , while that of the latter decreases , contrary to the tendency indicated by the data .
this nlo effect also overshoots the upper bound of the @xmath3 branching ratio very much .
this observation is expected : the @xmath3 and @xmath2 decays have the similar factorization formulas , so the branching ratio @xmath20 ought to be larger than @xmath21 due to the meson decay constants @xmath22 . therefore , the @xmath0 data have seriously constrained the possibility of resolving the @xmath1 puzzle in the available theoretical approaches .
there exists an alternative phenomenological application of scet @xcite , where the jet function , characterized by the scale of @xmath23 , @xmath24 being the @xmath25 quark mass and @xmath26 a hadronic scale , is regarded as being incalculable .
its contribution , together with other nonperturbative parameters , such as the charming penguin , were then determined by the @xmath1 data .
that is , the color - suppressed tree amplitude can not be explained , but the data are used to fit for the phenomenological parameters in the theory .
predictions for the @xmath13 , @xmath27 decays were then made based on the obtained parameters and partial su(3 ) flavor symmetry @xcite .
final - state interaction ( fsi ) is certainly a plausible resolution to the @xmath1 puzzle , but the estimate of its effect is quite model - dependent .
even opposite conclusions were drawn sometimes .
when including fsi either into naive factorization @xcite or into qcdf @xcite , the @xmath2 branching ratio was treated as an input in order to fix the involved free parameters .
hence , no resolution was really proposed .
it has been found that fsi , evaluated in the regge model , is insufficient to account for the observed @xmath2 branching ratio @xcite .
we conclude that there is no satisfactory resolution in the literature : the available proposals are either data fitting , or can not survive the constraints from the @xmath13 , @xmath17 data under the current theoretical development . in sec .
ii we compute the branching ratios , the direct cp asymmetries , and the polarization fractions of the @xmath0 decays using the nlo pqcd formalism .
the branching ratios and the direct cp asymmetries of the @xmath13 , @xmath9 , @xmath17 decays are calculated in sec .
iii by including the nlo jet function from scet@xmath19 into the qcdf formulas .
section iv is the discussion , where we comment on and compare the various analyses of the fsi effects in the @xmath13 , @xmath9 decays .
the nlo contributions from the vertex corrections , the quark loops , and the magnetic penguin to the @xmath13 and @xmath9 decays have been calculated in the naive dimensional regularization ( ndr ) scheme in the pqcd approach @xcite , and the results for the branching ratios and the direct cp asymmetries are quoted in tables [ br1 ] and [ cp1 ] , respectively .
we have taken this chance to correct a minor numerical mistake in the vertex corrections for the @xmath13 decays , whose branching ratios become smaller by @xmath28 .
note that a minus sign is missing for the @xmath29 term in the expression for the quark - loop contributions in eq .
( 27 ) of @xcite . nevertheless , this typo has nothing to do with the numerical outcomes .
our observations are summarized below .
the corrections from the quark loops and from the magnetic penguin come with opposite signs , and sum to about @xmath30 of the leading - order ( lo ) penguin amplitudes .
they mainly reduce the penguin - dominated @xmath13 branching ratios , but have a minor influence on the tree - dominated @xmath1 branching ratios , and on the direct cp asymmetries . on the contrary , the vertex corrections do not change the branching ratios , except the @xmath2 one .
they modify only the direct cp asymmetries of the @xmath31 , @xmath32 , and @xmath2 modes by increasing the color - suppressed tree amplitude @xmath33 few times .
the larger @xmath33 , leading to the nearly vanishing direct cp asymmetry @xmath34 , resolves the @xmath13 puzzle within the standard model .
.branching ratios from pqcd in the ndr scheme in units of @xmath35 .
the label lo@xmath36 means the lo results with the nlo wilson coefficients , and + vc , + ql , + mp , and + nlo mean the inclusions of the vertex corrections , of the quark loops , of the magnetic penguin , and of all the above nlo corrections , respectively .
the errors in the parentheses represent only the hadronic uncertainty @xcite . [
cols="^,^,^,^,^,^,^,^",options="header " , ] the predictions for the @xmath1 , @xmath37 , @xmath38 decays from qcdf with the input of the scet jet function are summarized in tables [ tab1 ] and [ tab2 ] .
the values in the parentheses are quoted from @xcite for the @xmath1 branching ratios , and from @xcite for the @xmath1 direct cp asymmetries and for the @xmath13 decays .
the small differences between our results and those from @xcite are attributed to the different choices of the ckm matrix elements , meson masses , etc .
all the calculations performed in this work , except those of the @xmath1 branching ratios , are new .
it is found that the scenario s4 plus the nlo jet function lead to the ratio @xmath39 , and accommodate at least the babar data of the @xmath2 branching ratio .
nevertheless , the same configuration overshoots the experimental upper bound of the @xmath3 branching ratio apparently , implying that the color - suppressed tree amplitude is enhanced overmuch by the nlo jet function . adopting the default scenario ,
qcdf satisfies the @xmath3 bound , but the predicted @xmath2 branching ratio becomes too small .
we have surveyed the other scenarios , and found the results from s1 and s3 ( s2 ) similar to those from the default ( s4 ) .
that is , it is also unlikely to accommodate the @xmath1 , @xmath17 data simultaneously in qcdf .
the @xmath13 branching ratios are not affected by the nlo jet function , because the color - suppressed tree amplitude is still subleading in the penguin - dominated modes .
the @xmath40 and @xmath41 branching ratios are not either , since they involve the larger color - allowed tree amplitude .
another indication against the resolution in @xcite is given by the direct cp asymmetries of the @xmath13 decays shown in table [ tab2 ] : both @xmath42 and @xmath43 deviate more from the data , a consequence expected from the discussion in @xcite . as explained in sec .
ii , the color - suppressed tree amplitude @xmath33 needs to be roughly orthogonal to @xmath44 in order to have a vanishing @xmath42 .
the nlo jet function , though increasing @xmath33 , does not introduce a large strong phase relative to @xmath44 .
that is , @xmath45 is large , but remains almost real as in the @xmath1 case @xcite .
this is exactly the same reason the @xmath13 puzzle can not be resolved in scet @xcite : the leading - power scet formalism demands a real ratio @xmath45 , such that a large @xmath45 just pushes the scet prediction for @xmath42 , about @xmath46 @xcite , further away from the data .
the direct cp asymmetry of the @xmath47 decays , whose tree contribution comes only from @xmath33 , is sensitive to the nlo jet function as indicated in table [ tab2 ] .
the direct cp asymmetries of the @xmath1 decays , which are tree - dominated , are relatively insensitive to the nlo jet function .
before concluding this work , we comment on and compare the various analyses of the fsi effects in the @xmath13 , @xmath9 decays .
the tiny @xmath2 branching ratio obtained in perturbative calculations naturally leads to the conjecture that fsi may play an essential role .
though the estimate of fsi effects is very model - dependent , the simultaneous applications to different decay modes can still impose a constraint .
the fsi effects from both the elastic and inelastic channels have been computed in the regge model for the @xmath1 decays @xcite and for the @xmath48 decays @xcite .
the conclusion is that fsi improves the agreement between the theoretical predictions and the experimental data , but does not suffice to resolve the @xmath1 puzzle : the @xmath2 branching ratio is increased by fsi only up to 0.10.65 @xcite .
moreover , the inelastic fsi through the long - distance charming penguin was found to be negligible in the @xmath1 decays , though it might be important in the @xmath13 ones .
the reason is that the contribution from the intermediate @xmath49 states is ckm suppressed in the former compared to the @xmath50 states in the latter .
this observation differs from that in @xcite , where a significant charming - penguin contribution was claimed .
we have pointed out in sec .
iii that the large charming penguin in @xcite is a consequence of fitting the scet parametrization to the data .
the inelastic fsi has been also evaluated as the absorptive part of charmed meson loops shown in figs .
[ fig1](a ) and [ fig1](b ) @xcite .
the two unknown cutoff parameters , appearing in the form factors associated with the three - meson vertices , were fixed by the measured @xmath13 branching ratios .
note that these parameters should be the same for @xmath51 and @xmath52 in the su(3 ) limit . applying the same formalism to the latter
, fsi can not resolve the @xmath53 puzzle , even allowing reasonable su(3 ) breaking effects for the cutoff parameters .
this result is understandable : the absorptive amplitudes from figs .
[ fig1](a ) and [ fig1](b ) are more or less orthogonal to the short - distance qcd penguin amplitudes in the @xmath53 decays , so that their effect is minor .
hence , the conclusion in @xcite is the same as in @xcite .
that is , the charming penguin is not enough to explain the observed @xmath1 branching ratios .
then additional dispersive amplitudes must be taken into account in @xcite .
those from figs . [ fig1](a ) and [ fig1](b ) , though calculable in the framework of @xcite , were not considered .
if considered , they , also contributing to the @xmath54 decays , would change the earlier predictions . therefore , a brand new mechanism , the dispersive amplitude from the meson annihilation @xmath55 shown in fig .
[ fig1](c ) , was introduced .
there is no corresponding diagram for the @xmath13 decays .
however , this amplitude is beyond the theoretical framework , i.e. , it can not be expressed in terms of the feynman rules derived in @xcite . the four free parameters , namely , the two cutoff parameters involved in figs . [ fig1](a ) and [ fig1](b ) , and the real and imaginary contributions from fig . [ fig1](c ) , were then determined by the four pieces of the @xmath1 data : the three branching ratios and the direct cp asymmetry @xmath56 .
that is , the @xmath57 branching ratio has been treated as an input .
the point of @xcite is to predict the direct cp asymmetries of the @xmath57 and @xmath58 decays , using the parameters fixed above .
the rescattering among the final states of the @xmath59 decays with @xmath60 , @xmath61 , and @xmath62 has been studied in @xcite .
these elastic fsi effects were parameterized in terms of two strong phases , which , together with the @xmath63 and @xmath64 form factors , and the chiral enhancement scale , were then determined by a global fit to the data , including the measured @xmath57 branching ratio .
nevertheless , the feature of the elastic fsi effects , i.e. , the correlated decrease and increase of the @xmath65 and @xmath57 branching ratios , respectively , was noticed @xcite . a fsi phase difference between the two @xmath1 isospin amplitudes with @xmath66
, @xmath67 has been introduced in @xcite , which was then varied to fit the @xmath1 data .
therefore , no explanation for the large @xmath57 branching ratio was provided from the viewpoint of fsi .
there exist other global fits based on different parametrizations for the charmless @xmath15 meson decays .
for example , the large ratio @xmath68 was extracted by fitting the quark - amplitude parametrization to the @xmath1 data @xcite .
no responsible mechanism was addressed , though the largeness of @xmath69 was translated into the largeness of the qcd penguin with an internal @xmath70 quark and/or of the exchange amplitudes in @xcite .
the qcdf formalism , in which the twist-3 spectator and annihilation amplitudes with the end - point singularities were parameterized as mentioned in sec .
iii , has been implemented into a global fit to the data @xcite . to reach a better fit
, the free parameters involved in qcdf must take different values for the @xmath59 , @xmath71 , @xmath72 modes .
these parameters have been tuned to account for the @xmath1 data in @xcite .
as emphasized before , the analysis must be also applied to other modes in order to obtain a consistent picture : the parameters preferred in @xcite lead to a large real @xmath68 , which is not favored by the data of the @xmath13 direct cp asymmetries as stated in sec .
after carefully investigating the proposals available in the literature , we have found that none of them can really resolve the @xmath1 puzzle .
the nlo pqcd analysis has confirmed that it is unlikely to accommodate the @xmath1 , @xmath17 data simultaneously ( the nlo pqcd predictions are consistent with the @xmath0 data ) .
the @xmath1 decays have been studied in the framework of light - cone sum rules ( lcsr ) @xcite , where a small @xmath2 branching ratio was also observed .
since there is only little difference between the sum rules for the @xmath1 and @xmath0 modes , we expect that the conclusion from lcsr will be the same as from pqcd .
the resolution with the input of the nlo scet jet function into qcdf @xcite does not survive the constraint from the @xmath0 data , and renders the @xmath73 and @xmath31 direct cp asymmetries deviate more away from the measured values .
we conclude that the @xmath0 data have seriously constrained the possibility of resolving the @xmath1 puzzle in the available theoretical approaches
. 1.0 cm we thank h.y .
cheng , c.k .
chua , d. pirjol , d. yang , and r. zwicky for useful discussions .
this work was supported by the national science council of r.o.c . under grant no .
nsc-94 - 2112-m-001 - 001 , by the taipei branch of the national center for theoretical sciences , and by the u.s .
department of energy under grant no
. de - fg02 - 90er40542 .
in this appendix we present the factorization formulas for the transverse helicity amplitudes : @xmath74 e_{e}(t ) h_{e}(a , b , b_1,b_2,x_2 ) \nonumber\\ & & \;\;\;\;\;\;\;\;\;\;\;\ ; + r_{2 } \left(\phi_{2}^v(\overline{x_2 } ) - \phi_{2}^a(\overline{x_2 } ) \right ) e_{e}(t ' ) h_{e}(a',b',b_2,b_1,x_1 ) \bigg\}\ ; , \\
f_{te4}(a ) & = & 32 \pi c_f m_b^2 \int_0 ^ 1 dx_1 dx_2 \int_0^{\infty } b_1db_1\ , b_2db_2\ , \phi_b(x_1,b_1 ) \nonumber \\ & & \times r_3 \ , \bigg\ { \left [ \phi_{2}^t(\overline{x_2 } ) - 2 r_{2 } \phi_{2}^a(\overline{x_2 } ) - r_{2 } x_2 \left(\phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \right ] e_{e}(t ) h_{e}(a , b , b_1,b_2,x_2 ) \nonumber\\ & & \;\;\;\;\;\;\;\;\;\;\;\ ; + r_{2 } \left(\phi_{2}^v(\overline{x_2 } ) - \phi_{2}^a(\overline{x_2 } ) \right ) e_{e}(t ' ) h_{e}(a',b',b_2,b_1,x_1 ) \bigg\}\ ; , \\
f_{ne6}(a ) & = & f_{te6}(a)\ = \ 0 \ ; , \\ f_{na4}(a ) & = & 16 \pi c_f m_b^2\ , r_{2}r_{3 } \int_0 ^ 1 dx_2 dx_3 \int_0^{\infty } b_2db_2\ , b_3db_3\ , \nonumber \\ & & \times \bigg\ { \left [ ( 1-x_3 ) \left ( \phi_{2}^v(\overline{x_2 } ) \phi_{3}^a(\overline{x_3 } ) + \phi_{2}^a(\overline{x_2 } ) \phi_{3}^v(\overline{x_3 } ) \right ) + ( 1+x_3 ) \left ( \phi_{2}^v(\overline{x_2 } ) \phi_{3}^v(\overline{x_3 } ) + \phi_{2}^a(\overline{x_2 } ) \phi_{3}^a(\overline{x_3 } ) \right ) \right ] \nonumber \\ & & \;\;\;\;\;\;\times e_{a}(t ) h_{e}(a , b , b_2,b_3,x_3 ) \nonumber\\ & & \;\;\;\;\;\ ; - \left [ ( 2-x_2 ) \left ( \phi_{2}^v(\overline{x_2 } ) \phi_{3}^v(\overline{x_3 } ) + \phi_{2}^a(\overline{x_2 } ) \phi_{3}^a(\overline{x_3 } ) \right ) - x_2 \left ( \phi_{2}^v(\overline{x_2 } ) \phi_{3}^a(\overline{x_3 } ) + \phi_{2}^a(\overline{x_2 } ) \phi_{3}^v(\overline{x_3 } ) \right ) \right ] \nonumber \\ & & \;\;\;\;\;\;\times e_{a}(t ' ) h_{e}(a',b',b_3,b_2,x_2 ) \bigg\}\ ; , \nonumber\\ & = & 16 \pi c_f m_b^2\ , r_{2}r_{3 } \int_0 ^ 1 dx_2 dx_3 \int_0^{\infty } b_2db_2\ , b_3db_3\ , \nonumber \\ & & \times \bigg\ { \left [ x_3 \left ( \phi_{2}^v(\overline{x_2 } ) - \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) + \left ( \phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) + \phi_{3}^a(\overline{x_3 } ) \right ) \right ] \nonumber \\ & & \;\;\;\;\;\;\times e_{a}(t ) h_{e}(a , b , b_2,b_3,x_3 ) \nonumber\\ & & \;\;\;\;\;\ ; - \left [ ( 1-x_2 ) \left ( \phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) + \phi_{3}^a(\overline{x_3 } ) \right ) + \left ( \phi_{2}^v(\overline{x_2 } ) - \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) \right ] \nonumber \\ & & \;\;\;\;\;\;\times e_{a}(t ' ) h_{e}(a',b',b_3,b_2,x_2 ) \bigg\}\ ; , \\
f_{ta4}(a ) & = & -\ , 32 \pi c_f m_b^2\ , r_{2}r_{3 } \int_0 ^ 1 dx_2 dx_3 \int_0^{\infty } b_2db_2\ , b_3db_3\ , \nonumber \\ & & \times \bigg\ { \left [ ( 1-x_3 ) \left ( \phi_{2}^v(\overline{x_2 } ) \phi_{3}^v(\overline{x_3 } ) + \phi_{2}^a(\overline{x_2 } ) \phi_{3}^a(\overline{x_3 } ) \right ) + ( 1+x_3 ) \left ( \phi_{2}^v(\overline{x_2 } ) \phi_{3}^a(\overline{x_3 } ) + \phi_{2}^a(\overline{x_2 } ) \phi_{3}^v(\overline{x_3 } ) \right ) \right ] \nonumber \\ & & \;\;\;\;\;\;\times e_{a}(t ) h_{e}(a , b , b_2,b_3,x_3 ) \nonumber\\ & & \;\;\;\;\;\ ; - \left [ ( 2-x_2 ) \left ( \phi_{2}^v(\overline{x_2 } ) \phi_{3}^a(\overline{x_3 } ) + \phi_{2}^a(\overline{x_2 } ) \phi_{3}^v(\overline{x_3 } ) \right ) - x_2 \left ( \phi_{2}^v(\overline{x_2 } ) \phi_{3}^v(\overline{x_3 } ) + \phi_{2}^a(\overline{x_2 } ) \phi_{3}^a(\overline{x_3 } ) \right ) \right ] \nonumber \\ & & \;\;\;\;\;\;\times e_{a}(t ' ) h_{e}(a',b',b_3,b_2,x_2 ) \bigg\}\ ; , \nonumber\\ & = & 32 \pi c_f m_b^2\ , r_{2}r_{3 } \int_0 ^ 1 dx_2 dx_3 \int_0^{\infty } b_2db_2\ , b_3db_3\ , \nonumber \\ & & \times \bigg\ { \left [ x_3 \left ( \phi_{2}^v(\overline{x_2 } ) - \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) - \left ( \phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) + \phi_{3}^a(\overline{x_3 } ) \right ) \right ] \nonumber \\ & & \;\;\;\;\;\;\times e_{a}(t ) h_{e}(a , b ,
b_2,b_3,x_3 ) \nonumber\\ & & \;\;\;\;\;\ ; + \left [ ( 1-x_2 ) \left ( \phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) + \phi_{3}^a(\overline{x_3 } ) \right ) - \left ( \phi_{2}^v(\overline{x_2 } ) - \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) \right ] \nonumber \\ & & \;\;\;\;\;\;\times e_{a}(t ' ) h_{e}(a',b',b_3,b_2,x_2 ) \bigg\}\ ; , \\
f_{na6}(a ) & = & 32 \pi c_f m_b^2 \int_0 ^ 1 dx_2 dx_3 \int_0^{\infty } b_2db_2\ , b_3db_3\ , \nonumber \\ & & \times \bigg\ { r_{2 } \left(\phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \phi_{3}^t(\overline{x_3 } ) e_{a}(t ) h_{e}(a , b , b_2,b_3,x_3 ) \nonumber\\ & & \;\;\;\;\;\ ; + r_{3 } \phi_{2}^t(\overline{x_2 } ) \left(\phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) e_{a}(t ' ) h_{e}(a',b',b_3,b_2,x_2 ) \bigg\}\ ; , \\
f_{ta6}(a ) & = & 2\,f_{na6}(a)\ ; , \\ { \cal m}_{ne4}(a ' ) & = & 32 \pi c_f \frac{\sqrt{2n_c}}{n_c}m_b^2\ , r_3 \int_0 ^ 1 dx_1dx_2dx_3 \int_0^{\infty } b_1 db_1\ , b_3 db_3\ , \phi_b(x_1,b_1 ) \nonumber \\ & & \times \bigg\ { ( 1-x_3 ) \phi_{2}^t(\overline{x_2 } ) \left(\phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) e'_{e}(t ) h_n(a , b , b_1,b_3 ) \nonumber \\ & & \;\;\;\;\;\ ; + \left [ x_3 \phi_{2}^t(\overline{x_2 } ) \left(\phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) - 2 r_2 ( x_2+x_3 ) \left ( \phi_{2}^v(\overline{x_2 } ) \phi_{3}^v(\overline{x_3 } ) + \phi_{2}^a(\overline{x_2 } ) \phi_{3}^a(\overline{x_3 } ) \right ) \right ] \nonumber\\ & & \;\;\;\;\;\;\times e'_{e}(t ' ) h_n(a',b',b_1,b_3 ) \bigg\}\ ; , \nonumber\\ & = & 32 \pi c_f \frac{\sqrt{2n_c}}{n_c}m_b^2\ , r_3 \int_0 ^ 1 dx_1dx_2dx_3 \int_0^{\infty } b_1 db_1\ , b_3 db_3\ , \phi_b(x_1,b_1 ) \nonumber \\ & & \times \bigg\ { ( 1-x_3 ) \phi_{2}^t(\overline{x_2 } ) \left(\phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) e'_{e}(t ) h_n(a , b , b_1,b_3 ) \nonumber \\ & & \;\;\;\;\;\ ; + \left [ x_3 \phi_{2}^t(\overline{x_2 } ) \left(\phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) - r_2 ( x_2+x_3 ) \left ( \phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) + \phi_{3}^a(\overline{x_3 } ) \right ) \right.\nonumber\\ & & \left.\;\;\;\;\;\;\;\;\;\;\ ; - r_2 ( x_2+x_3 ) \left ( \phi_{2}^v(\overline{x_2 } ) - \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) \right ] e'_{e}(t ' ) h_n(a',b',b_1,b_3 ) \bigg\}\ ; , \\ { \cal m}_{te4}(a ' ) & = & 64 \pi c_f \frac{\sqrt{2n_c}}{n_c}m_b^2\ , r_3 \int_0 ^ 1 dx_1dx_2dx_3 \int_0^{\infty } b_1 db_1\ , b_3 db_3\ , \phi_b(x_1,b_1 ) \nonumber \\ & & \times \bigg\ { ( 1-x_3 ) \phi_{2}^t(\overline{x_2 } ) \left(\phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) e'_{e}(t ) h_n(a , b , b_1,b_3 ) \nonumber \\ & & \;\;\;\;\;\ ; + \left [ x_3 \phi_{2}^t(\overline{x_2 } ) \left(\phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) + 2 r_2 ( x_2+x_3 ) \left ( \phi_{2}^v(\overline{x_2 } ) \phi_{3}^a(\overline{x_3 } ) + \phi_{2}^a(\overline{x_2 } ) \phi_{3}^v(\overline{x_3 } ) \right ) \right ] \nonumber \\ & & \;\;\;\;\;\;\times e'_{e}(t ' ) h_n(a',b',b_1,b_3 ) \bigg\}\ ; , \nonumber\\ & = & 64 \pi c_f \frac{\sqrt{2n_c}}{n_c}m_b^2\ , r_3 \int_0 ^ 1 dx_1dx_2dx_3 \int_0^{\infty } b_1 db_1\ , b_3 db_3\ , \phi_b(x_1,b_1 ) \nonumber \\ & & \times \bigg\ { ( 1-x_3 ) \phi_{2}^t(\overline{x_2 } ) \left(\phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) e'_{e}(t ) h_n(a , b , b_1,b_3 ) \nonumber \\ & & \;\;\;\;\;\ ; + \left [ x_3 \phi_{2}^t(\overline{x_2 } ) \left(\phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) + r_2 ( x_2+x_3 ) \left ( \phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) + \phi_{3}^a(\overline{x_3 } ) \right ) \right.\nonumber\\ & & \left.\;\;\;\;\;\;\;\;\;\;\ ; - r_2 ( x_2+x_3 ) \left ( \phi_{2}^v(\overline{x_2 } ) - \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) \right ] e'_{e}(t ' ) h_n(a',b',b_1,b_3 ) \bigg\}\ ; , \\ { \cal m}_{ne5}(a ' ) & = & - { \cal m}_{ne4}(a ' ) \ ; , \\ { \cal m}_{te5}(a ' ) & = & - { \cal m}_{te4}(a ' ) \ ; , \\ { \cal m}_{ne6}(a ' ) & = & -\ , 32 \pi c_f \frac{\sqrt{2n_c}}{n_c}m_b^2\ , r_{2 } \int_0 ^ 1 dx_1dx_2dx_3\int_0^{\infty } b_1 db_1\ , b_3 db_3\,\phi_b(x_1,b_1 ) \nonumber \\ & & \times x_2 \left(\phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \phi_{3}^t(\overline{x_3 } ) \left [ e'_{e}(t ) h_n(a , b , b_1,b_3 ) + e'_{e}(t ' ) h_n(a',b',b_1,b_3 ) \right]\ ; , \\ { \cal m}_{te6}(a ' ) & = & 2\ , { \cal m}_{ne6}(a')\ ; , \\ { \cal m}_{na4}(a ' ) & = & -\ , 64 \pi c_f \frac{\sqrt{2n_c}}{n_c}m_b^2\ , r_{2 } r_{3 } \int_0 ^ 1 dx_1dx_2dx_3 \int_0^{\infty } b_1 db_1\ , b_3 db_3\,\phi_b(x_1,b_1 ) \nonumber \\ & & \times \left ( \phi_{2}^v(\overline{x_2 } ) \phi_{3}^v(\overline{x_3 } ) + \phi_{2}^a(\overline{x_2 } ) \phi_{3}^a(\overline{x_3 } ) \right ) e'_{a}(t ' ) h_n(a',b',b_3,b_1)\ ; , \nonumber\\ & = & -\ , 32 \pi c_f \frac{\sqrt{2n_c}}{n_c}m_b^2\ , r_{2 } r_{3 } \int_0 ^ 1 dx_1dx_2dx_3 \int_0^{\infty } b_1 db_1\ , b_3 db_3\,\phi_b(x_1,b_1 ) \nonumber \\ & & \times \bigg\ { \left ( \phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) + \phi_{3}^a(\overline{x_3 } ) \right ) + \left ( \phi_{2}^v(\overline{x_2 } ) - \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) \bigg\ } \nonumber \\ & & \;\;\;\;\;\;\times e'_{a}(t ' ) h_n(a',b',b_3,b_1)\ ; , \\ { \cal m}_{ta4}(a ' ) & = & 128 \pi c_f \frac{\sqrt{2n_c}}{n_c}m_b^2\ , r_{2 } r_{3 } \int_0 ^ 1 dx_1dx_2dx_3 \int_0^{\infty } b_1 db_1\ , b_3 db_3\,\phi_b(x_1,b_1 ) \nonumber \\ & & \times \left ( \phi_{2}^v(\overline{x_2 } ) \phi_{3}^a(\overline{x_3 } ) + \phi_{2}^a(\overline{x_2 } ) \phi_{3}^v(\overline{x_3 } ) \right ) e'_{a}(t ' ) h_n(a',b',b_3,b_1)\ ; , \nonumber\\ & = & 64 \pi c_f \frac{\sqrt{2n_c}}{n_c}m_b^2\ , r_{2 } r_{3 } \int_0 ^ 1 dx_1dx_2dx_3 \int_0^{\infty } b_1 db_1\ , b_3 db_3\,\phi_b(x_1,b_1 ) \nonumber \\ & & \times \bigg\ { \left ( \phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) + \phi_{3}^a(\overline{x_3 } ) \right ) - \left ( \phi_{2}^v(\overline{x_2 } ) - \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) \bigg\ } \nonumber \\ & & \;\;\;\;\;\;\times e'_{a}(t ' ) h_n(a',b',b_3,b_1)\ ; , \\
{ \cal m}_{na5}(a ' ) & = & { \cal m}_{na4}(a')\ ; , \\ { \cal m}_{ta5}(a ' ) & = & { \cal m}_{ta4}(a')\ ; , \\ { \cal m}_{na6}(a ' ) & = & 32 \pi c_f \frac{\sqrt{2n_c}}{n_c}m_b^2 \int_0 ^ 1 dx_1dx_2dx_3 \int_0^{\infty } b_1 db_1\ , b_3 db_3\,\phi_b(x_1,b_1 ) \nonumber \\ & & \times \bigg\ { \left [ r_3 x_3 \phi_{2}^t(\overline{x_2 } ) \left(\phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) - r_2 ( 1-x_2 ) \left(\phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \phi_{3}^t(\overline{x_3 } ) \right ] \nonumber \\ & & \;\;\;\;\;\;\times e'_{a}(t ) h_n(a , b , b_3,b_1 ) \nonumber \\ & & \;\;\;\;\;\ ; + \left [ r_3 ( 2-x_3 ) \phi_{2}^t(\overline{x_2 } ) \left(\phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) - r_2 ( 1+x_2 ) \left(\phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \phi_{3}^t(\overline{x_3 } ) \right ] \nonumber \\ & & \;\;\;\;\;\;\times e'_{a}(t ' ) h_n(a',b',b_3,b_1 ) \bigg\}\ ; , \\ { \cal m}_{ta6}(a ' ) & = & 2\ , { \cal m}_{na6}(a')\;. \ ] ] the quark - loop corrections @xmath75 for @xmath76 , @xmath77 , and @xmath70 , and the magnetic - penguin corrections @xmath78 to the transverse components are written as @xmath79\nonumber \\ & & \;\;\;\;\;\ ; \times e^{(q)}(t_q , l^2 ) h_{e}(a , b , b_1,b_2,x_2 ) \nonumber\\ & & \;\;\;\;\;\ ; + \ , r_2 \phi_{2}^v(\overline{x_2 } ) \phi_{3}^v(\overline{x_3 } ) e^{(q)}(t_q^{\prime},l^{\prime 2 } ) h_{e}(a',b',b_2,b_1,x_1 ) \big\ } \ ; , \\ { \cal m}^{(q)}_{t } & = & 0 \ ; , \\ { \cal m}^{(g)}_{n } & = & 16 m_b^4 \frac{c_f^2}{\sqrt{2n_c } } \int_0 ^ 1 dx_1dx_2dx_3 \int_0^{\infty } b_1db_1\ , b_2db_2\ , b_3db_3\ , \phi_b(x_1,b_1 ) \nonumber \\ & & \times\ , \big\ { \left [ - r_2 ( 1-x_2 ^ 2 ) \left(\phi_{3}^v(\overline{x_3 } ) + \phi_{3}^a(\overline{x_3 } ) \right ) \phi_{3}^t(\overline{x_3 } ) - r_3 ( 1+x_2 ) x_3 \phi_{2}^t(\overline{x_2 } ) \left(\phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) \right.\nonumber\\ & & \left.\;\;\;\;\;\;\;\;\;\;\ ; - r_2 r_3 ( 1-x_2 ) \left ( \phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) + \phi_{3}^a(\overline{x_3 } ) \right ) \right.\nonumber\\ & & \left.\;\;\;\;\;\;\;\;\;\;\ ; - r_2 r_3 x_3(1 - 2x_2 ) \left ( \phi_{2}^v(\overline{x_2 } ) - \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) \right ] e_{g}(t_q ) h_g(a , b , c , b_1,b_2,b_3,x_2 ) \nonumber\\ & & \;\;\;\;\;\ ; - r_2 r_3 x_3 \left ( \phi_{2}^v(\overline{x_2 } ) - \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) e_{g}(t_q^{\prime } ) h_g(a',b',c',b_2,b_1,b_3,x_1 ) \big\ } \ ; , \\ { \cal m}^{(g)}_{t } & = & 32 m_b^4 \frac{c_f^2}{\sqrt{2n_c } } \int_0 ^ 1 dx_1dx_2dx_3 \int_0^{\infty } b_1db_1\ , b_2db_2\ , b_3db_3\ , \phi_b(x_1,b_1 ) \nonumber \\ & & \times\ , \big\ { \left [ - r_2 ( 1-x_2 ^ 2 ) \left(\phi_{3}^v(\overline{x_3 } ) + \phi_{3}^a(\overline{x_3 } ) \right ) \phi_{3}^t(\overline{x_3 } ) - r_3 ( 1+x_2 ) x_3 \phi_{2}^t(\overline{x_2 } ) \left(\phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) \right.\nonumber\\ & & \left.\;\;\;\;\;\;\;\;\;\;\ ; + r_2 r_3 ( 1-x_2 ) \left ( \phi_{2}^v(\overline{x_2 } ) + \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) + \phi_{3}^a(\overline{x_3 } ) \right ) \right.\nonumber\\ & & \left.\;\;\;\;\;\;\;\;\;\;\ ; - r_2 r_3 x_3(1 - 2x_2 ) \left ( \phi_{2}^v(\overline{x_2 } ) - \phi_{2}^a(\overline{x_2 } ) \right ) \left ( \phi_{3}^v(\overline{x_3 } ) - \phi_{3}^a(\overline{x_3 } ) \right ) \right ] e_{g}(t_q ) h_g(a , b , c , b_1,b_2,b_3,x_2 ) \nonumber\\ & & \;\;\;\;\;\ ; - r_2 \left [ ( 1-x_2 ) \phi_{2}^v(\overline{x_2})\phi_{3}^t(\overline{x_3 } ) - r_3 ( 1 - 2x_2 ) \phi_{2}^v(\overline{x_2})\phi_{3}^v(\overline{x_3 } ) - r_3 \phi_{2}^v(\overline{x_2})\phi_{3}^a(\overline{x_3 } ) \right.\nonumber\\ & & \left.\;\;\;\;\;\;\;\;\;\;\ ; - r_3 \phi_{2}^a(\overline{x_2})\phi_{3}^v(\overline{x_3 } ) + r_3 \phi_{2}^a(\overline{x_2})\phi_{3}^a(\overline{x_3 } ) \right ] e_{g}(t_q^{\prime } ) h_g(a',b',c',b_2,b_1,b_3,x_1 ) \big\ } \;.\end{aligned}\ ] ] the definitions of all the variables and the convolution factors in the above expressions are referred to @xcite .
99 heavy flavor averaging group , hep - ex/0505100 ; updated in http://www.slac.stanford.edu/xorg/hfag .
keum , h - n .
li , and a.i .
sanda , phys lett .
b * 504 * , 6 ( 2001 ) ; phys .
d * 63 * , 054008 ( 2001 ) .
l , k. ukai , and m.z .
yang , phys .
d * 63 * , 074009 ( 2001 ) .
li , s. mishima , and a.i .
sanda , phys .
d * 72 * , 114005 ( 2005 ) .
botella , d. london , and j.p .
silva , phys .
d * 73 * , 071501 ( 2006 ) .
yang , r.m .
wang , and g.r .
lu , phys .
d * 73 * , 015003 ( 2006 ) .
s. baek , f.j .
botella , d. london , and j.p .
silva , phys .
d * 72 * , 114007 ( 2005 ) .
cheng , y.n .
gao , c.s .
huang , and x.h .
wu , hep - ph/0512268 .
m. beneke , g. buchalla , m. neubert , and c.t .
sachrajda , phys .
lett . * 83 * , 1914 ( 1999 ) ; nucl .
b591 * , 313 ( 2000 ) ; nucl . phys . *
b606 * , 245 ( 2001 ) .
m. beneke and d. yang , nucl . phys . *
b736 * , 34 ( 2006 ) .
bauer , d. pirjol , i.z .
rothstein , and i.w .
stewart , phys .
d * 70 * , 054015 ( 2004 ) .
bauer , i.z .
rothstein , and i.w .
stewart , hep - ph/0510241 .
chua , w.s .
hou , and k.c .
yang , phys .
d * 65 * , 096007 ( 2002 ) ; mod .
a*18 * , 1763 ( 2003 ) .
cheng , c.k .
chua , and a. soni , phys .
d * 71 * , 014030 ( 2005 ) .
a. deandrea , m. ladisa , v. laporta , g. nardulli , and p. santorelli , hep - ph/0508083 .
y.y . charng and h - n .
li , phys .
d * 71 * , 014036 ( 2005 ) .
y. grossman , a. hocker , z. ligeti , and d. pirjol , phys .
d * 72 * , 094033 ( 2005 ) .
y. li and c.d .
l , phys .
d * 73 * , 014024 ( 2006 ) .
chen , hep - ph/0601019 .
t. kurimoto , h - n .
li , and a.i .
sanda , phys .
d * 65 * , 014007 ( 2002 ) .
p. ball , v.m .
braun , y. koike , and k. tanaka , nucl .
phys . * b529 * , 323 ( 1998 ) .
m. beneke and m. neubert , nucl . phys . *
b675 * , 333 ( 2003 ) .
p.k . das and k.c .
yang , phys .
d * 71 * , 094002 ( 2005 ) ; y.d .
yang , r.m .
wang , and g.r .
lu , phys .
d * 72 * , 015009 ( 2005 ) ; w.j .
zou and z.j .
xiao , phys .
d * 72 * , 094026 ( 2005 ) .
s. mishima and a.i .
sanda , prog .
. phys . * 110 * , 549 ( 2003 ) ; phys.rev .
d * 69 * , 054005 ( 2004 ) . p. ball and v.m .
braun , phys .
d * 58 * , 094016 ( 1998 ) . p. ball and r. zwicky , phys .
d * 71 * , 014029 ( 2005 ) .
a.l . kagan , phys .
b * 601 * , 151 ( 2004 ) ; hep - ph/0407076 .
chen , y.y .
keum , and h - n . li , phys .
d * 66 * , 054013 ( 2002 ) .
li and s. mishima , phys .
d * 71 * , 054025 ( 2005 ) .
p. colangelo , g. nardulli , n. paver , and riazuddin , z. phys .
c * 45 * , 575 ( 1990 ) ; m. ciuchini , e. franco , g. martinelli , and l. silvestrini , nucl . phys .
* b501 * , 271 ( 1997 ) .
a.r . williamson and j. zupan , hep - ph/0601214 .
cheng and k.c .
yang , phys . lett .
b * 511 * , 40 ( 2001 ) .
m. beneke and s. jager , hep - ph/0512101 .
m. beneke and s. jager , hep - ph/0512351 .
m. ladisa , v. laporta , g. nardulli , and p. santorelli , phys .
d * 70 * , 114025 ( 2004 ) .
e. kou and t.n .
pham , hep - ph/0601272 .
y.y . charng and h - n .
li , phys .
b. * 594 * , 185 ( 2004 ) .
t. yoshikawa , phys .
d * 68 * , 054023 ( 2003 ) ; s. mishima and t. yoshikawa , phys . rev .
d * 70 * , 094024 ( 2004 ) .
buras , r. fleischer , s. recksiegel , and f. schwab , phys .
lett . * 92 * , 101804 ( 2004 ) ; nucl.phys .
* b697 * , 133 ( 2004 ) .
x.g . he and b. mckellar , hep - ph/0410098 .
chiang , m. gronau , j.l .
rosner , and d.a .
suprun phys .
d * 70 * , 034020 ( 2004 ) .
z. ligeti , int . j. mod
a * 20 * , 5105 ( 2005 ) .
y.l . wu and y.f .
zhou , phys .
d * 71 * , 021701 ( 2005 ) .
s. baek , p. hamel , d. london , a. datta , and d.a .
suprun , phys .
d * 71 * , 057502 ( 2005 ) .
sun , d.s .
yang , and g.h .
zhu , phys .
d * 67 * , 014023 ( 2003 ) .
r. aleksan , p. f. giraud , v. morenas , o. pene , and a. s. safir , phys .
d * 67 * , 094019 ( 2003 ) .
cottingham , i.b .
whittingham , and f.f .
wilson , phys .
d * 71 * , 077301 ( 2005 ) .
a. khodjamirian , th .
mannel , m. melcher , and b. melic , phys .
d * 72 * , 094012 ( 2005 ) . | we point out that the @xmath0 data have seriously constrained the possibility of resolving the @xmath1 puzzle from the large observed @xmath2 branching ratio in the available theoretical approaches .
the next - to - leading - order ( nlo ) contributions from the vertex corrections , the quark loops , and the magnetic penguin evaluated in the perturbative qcd ( pqcd ) approach have saturated the experimental upper bound of the @xmath3 branching ratio , and do not help .
the nlo pqcd predictions for the @xmath4 and @xmath5 branching ratios are consistent with the data . the inclusion of the nlo jet function from the soft - collinear effective theory into the qcd - improved factorization approach , though enhancing the @xmath2 branching ratio sufficiently , overshoots the bound of the @xmath3 branching ratio , and deteriorates the predictions for the @xmath6 and @xmath7 direct cp asymmetries .
= 10000 |
when an exoplanet s orbital plane lies along our line of sight , the planet will transit its host star , periodically obscuring a small portion of the stellar disk and producing detectable dips in a photometric lightcurve .
the first transits ( of a planet previously discovered by doppler ) were observed in 1999 @xcite .
the first exoplanet to be detected with the transit technique was confirmed by doppler observations in 2002 @xcite . as of 18 october 2012 ,
288 confirmed transiting planets in 233 systems have been reported @xcite .
the _ kepler _ mission , operating since 2009 , has identified more than 2300 candidate transiting planets ( _ kepler _ objects of interest or kois ) ( * ? ? ? * hereafter b12 ) .
although only a small fraction of kois have been confirmed , the false positive rate is thought to be low @xcite , but see @xcite .
transit searches are more sensitive than doppler searches to the smallest planets because the transit signal scales with the square of the planet radius @xmath1 , while the doppler signal of a rocky planet scales approximately as @xmath2 @xcite .
_ kepler _ has already discovered 90 candidates possibly _ smaller _ than earth .
transiting planets are of special interest because their radii can be estimated from the transit signal .
if the transit is not grazing , the fractional decrease @xmath3 in the star s observed flux is @xmath4 where @xmath5 is the stellar radius .
a measurement of @xmath3 combined with knowledge of @xmath5 yields the planet radius . because the inclination of a transiting planet s orbit is near 90@xmath6
, the mass of the planet can also be unambiguously established from doppler observations .
combinations of mass and radius can be compared with predictions by models of interior structure @xcite .
spectroscopy or spectrophotometry during transits can detect or rule out constituents in an atmosphere @xcite , and secondary eclipses ( occultations of the planet ) can constrain temperature and albedo @xcite .
additional planets can be discovered by variation in transit times @xcite .
analyses of large samples of transiting planets , including the catalog of kois , have attempted to ascertain properties of transiting planet populations , e.g. , whether they are segregated into discrete groups @xcite , the distribution with planet radius @xcite the dependence of planet occurrence on the metallicity of the host star @xcite , the effect of stellar irradiance on giant planet radius @xcite , and the occurrence of planets compared to doppler surveys @xcite . in the case of _ kepler _ , lack of doppler confirmation for most candidate planets as well as detailed spectroscopic characterization of the stars make it important to properly account for any systematic effects or biases .
detection of a planet in a transit survey depends on the properties of the planet , most notably @xmath1 ( equation [ eqn.depth ] ) , but also on the orbital period because it determines the number of transits that are observed and the total transit signal .
@xcite , @xcite , and @xcite pointed out that transit - selected samples are biased toward large planets on short - period orbits .
these biases can be extreme in ground - based surveys which suffer correlated ( `` red '' ) noise from variations in atmospheric transmission and discontinuous observing windows .
equation [ eqn.depth ] also shows that planets of a given radius will be more readily detected around smaller stars .
this has , in part , motivated transit searches for planets around m dwarf stars @xcite . in this case
, a property of the host star , as opposed to the planet , influences the likelihood that a transiting planet will be detected , and that both star and planet will be included in a transit - selected sample .
thus a selection effect will act on stellar radius , or on any property that is related to stellar radius , such as metallicity .
this will produce systematic offsets or biases in the properties of stars hosting known transiting planets relative to the properties of the target sample .
the construction of a target catalog itself can also produce selection effects in a transit survey .
most notable among constraints on target stars is an apparent magnitude limit because of a signal - to - noise ratio ( snr ) requirement , or the need to confirm candidate transiting systems using doppler observations .
a magnitude limit will cause ( malmquist ) bias towards more luminous stars ; these can be included to larger distances and hence sample a larger volume of space . at a given
effective temperature , luminosity is uniquely related to stellar radius , and hence this is also a bias towards larger stars that , unmitigated , will affect the detection of planets and estimates of their radii .
some of these effects would disappear or could be corrected if stellar parameters , i.e. radius , were precisely established .
but , up to now , the large scale of transit surveys ( @xmath7 stars ) has precluded such determinations .
neither radius nor mass are directly observable for distant , single stars such as _ kepler _ targets .
the properties of most _ kepler _ stars have been inferred by comparing stellar models to the broad - band photometry of the _ kepler _ input catalog ( kic ) ( * ? ? ?
* hereafter br11 ) .
few spectra and almost no parallax ( distance ) measurements are available , and most stars have only upper limits on proper motion .
kic estimates of stellar radii have large uncertainties due to ( i ) errors in photometry ; ( ii ) degeneracies between stellar parameters and colors ; and ( iii ) errors in the models themselves . while kic estimates of stellar effective temperature are comparatively robust , those of surface gravity ( @xmath8 ) and metallicity ( fe / h ) are not as reliable ( br11 ) .
br11 concluded that no gravity or radius information could be inferred for stars hotter than 5400 k ( @xmath9 ) .
@xcite found that the kic and astroseismic radii of 500 solar - type stars have random discrepancies of order 50% and a systematic offset of about the same amount .
@xcite found a similar scatter but negligible systematic offset in @xmath8 ( and hence the radius ) of 93 solar - type _ kepler _ stars .
@xcite found that many m - type stars that were classified as dwarfs or were unclassified in the kic are actually evolved stars .
selection effects acting on uncertainties in stellar radius will bias the observed properties of planet - hosting stars with respect to their true distributions .
for example , while essentially all m - type hosts of kois are _ bona fida _ dwarf stars @xcite , the vast majority of the bright ( @xmath10 ) targets and some fainter stars are giants @xcite .
this disparity is a result of the strong selection effect on stellar radius described above ; planets are far more difficult to detect around giant stars due to their large size and higher variability @xcite .
because of the relation between planet radius and stellar radius ( equation [ eqn.depth ] ) , estimates of planet radius will likewise be affected . here
, we quantify five effects produced by selection bias and uncertainties in stellar parameters in the _ kepler _ survey . in section [ sec.scaling ]
we derive useful scaling relationships for selection effects on transiting planet detection and target star selection . in section [ sec.kepler ]
we apply these concepts to the _ kepler _ survey using the koi catalog , parameters from the kic , and models of stellar evolution and stellar populations .
we describe our methods and models in section [ sec.methods ] . in section [ sec.eddington ]
, we calculate the effect of eddington bias on the radius distribution of kois as a result of uncertainties in stellar radius . in section [ sec.malmquist ]
we describe the effect of malmquist bias on the magnitude - limited _ kepler _ target catalog and the preferential inclusion of more luminous , larger stars , thus biasing downwards the radius of some kois . in section [ sec.metallicity ]
we estimate the bias towards lower metallicity among koi - hosting stars as a consequence of the relationship between stellar metallicity and radius on the main sequence . in section
[ sec.inflated ] we describe how uncertainties in stellar radius produce correlated errors in planet radius and stellar luminosity , potentially affecting statistics describing the relationship between `` inflated '' giant planets and stellar irradiation .
in section [ sec.shrunken ] we consider the effect of stellar metallicity on transit depth and the interpretation of any correlation between metallicity and the radii of giant planets .
we summarize our results and describe current and future efforts to better determine the parameters of _ kepler _ stars in section [ sec.discussion ] .
in a transit survey , selection bias acts on a quantity @xmath11 ( a stellar or planetary parameter ) when the probability @xmath12 that a star is included in the survey , or that a planet is detected transiting a star , depend on that parameter .
this bias is superposed on any real correlations and will persist to the extent that the values of the parameter and its effect on inclusion or detection are imperfectly quantified .
the bias @xmath13 is the difference between the observed mean @xmath14 and the intrinsic mean @xmath15 , or @xmath16 where the brackets represent marginalizing over the population of stars , subject to any constraints . to derive useful scaling relations
, we chose apparent brightness ( magnitude ) and effective temperature @xmath17 as independently varying parameters .
the first fixes the noise level against which a transit must be detected .
morever , the _ kepler _ target catalog is magnitude - limited @xcite . among main sequence stars , @xmath17 is closely related to mass , an important parameter of planet populations @xcite . unlike other stellar parameters
, it can be robustly estimated from kic photometry ( br11 , * ? ? ?
effective temperature is thus a convenient plotting parameter which minimizes complications due to variation in the planet population along the main sequence .
nevertheless , values of @xmath17 do _ not _ map to unique values of stellar mass because stars have different ages and metallicities and plots with @xmath17 the dependence on mass should be considered `` blurred '' .
calculations using stellar models , as described below , explicitly take into account the effects of age and metallicity .
we first estimate the probability @xmath12 that a star is included in a catalog of transiting systems .
the probability of detecting a planet is calculated as a function of both stellar properties ( radius and mass @xmath5 and @xmath18 ) and planet properties ( radius @xmath1 and orbital period @xmath19 ) , and then marginalized over planet properties using an appropriate distribution function .
this yields @xmath12 as a function of @xmath5 and @xmath18 .
equation [ eqn.bias ] can then evaluated using a stellar model that describes the intrinsic distributions of these parameters and their relations to other observables . in
many instances we can use scaling relations rather than exact relations becaue equation [ eqn.bias ] is normalized .
we adopted a double power - law for the intrinsic distribution @xmath20 of planets with radius and orbital period @xcite : @xmath21 for @xmath19 larger than some minimum value @xmath22 where planets are found .
transit detection depends on the geometric probability that the planet is on a transiting orbit , as well as the the signal ( depth ) of the transit relative to noise . in the absence of coherent or `` red '' noise from the atmosphere ,
the signal - to - noise ratio of a single transit is @xmath23 , where @xmath24 is the total number of photons detected during the event . in an observation interval @xmath25 about @xmath26 transits will be observed , bringing the total number of photons to @xmath27
. therefore the signal - to - noise ratio of the co - added transits is @xmath28 at a given apparent brightness , @xmath29 , where @xmath30 is the transit duration .
when the transit impact parameter is low and the transit chord is close to the stellar diameter , @xmath31 . assuming a near - circular orbit , the transverse velocity @xmath32 can be expressed in terms of the orbital period and mass of the star and @xmath33 where @xmath34 is the gravitational constant .
we derive a scaling relation between snr and planet / star properties by substituting @xmath30 for @xmath24 in equation [ eqn.snr0 ] and ignoring constant factors : @xmath35 solving for @xmath1 gives a scaling relation for the radius of the smallest planet on a given orbital period that can be detected at a fixed snr threshold : @xmath36 likewise , there is a relation for the maximum orbital period at which a planet of a given radius @xmath1 can be detected at a fixed snr threshold : @xmath37 @xmath38 is a sensitive function of @xmath1 , underscoring why transit surveys are highly biased towards the largest planets @xcite . to obtain the observed distribution @xmath12 of planets with @xmath5 and @xmath18 , we multiply the intrinsic distribution ( equation [ eqn.distribution ] ) by the geometric probability that a planets is on a transiting orbit . for circular orbits
this is proportional to the ratio of the stellar radius to orbital semimajor axis @xmath39 which , based on newtonian orbital dynamics , scales as @xmath40 .
the observed planet distribution is : @xmath41 we marginalize equation [ eqn.observed ] over both @xmath19 and @xmath1 , first integrating from @xmath22 to @xmath38 .
the maximum period is also limited by the observing window and the requirement that more than one transit must be observed , e.g. @xmath42 .
integration of @xmath43 in equation [ eqn.observed ] yields a factor proportional to @xmath44 . if @xmath45 , then equation [ eqn.pmax ] , is used to re - express this as @xmath46 $ ] , where @xmath47 is the radius of the smallest planet that can be detected at @xmath48 , i.e. that can be detected at all ) .
integration of equation [ eqn.observed ] over @xmath49 from @xmath47 to @xmath50 produces : @xmath51 were @xmath52 .
because the @xmath22 factor and the integral depend only on @xmath53 and @xmath54 , which are planet population parameters and not stellar properties , they can be ignored when calculating biases in stellar properties .
substituting for @xmath47 and retaining only factors that depend on stellar properties , @xmath55 all else being equal , planets are more likely to be detected around stars with smaller radii ( because transit depths are larger ) and/or lower masses ( because transit durations are longer ) .
smaller stars are thus more likely to appear in a transit - selected sample . in the case of mass - radius relation @xmath56 for zero - age solar - type stars @xcite and a planet radius distibution power - law index @xmath57 @xcite , then @xmath58 .
this is simply a statement that smaller ( and more ) planets can be detected around lower mass stars .
older stars will have a steeper mass - radius relation , and as a result the dependence of @xmath12 on @xmath18 will be more pronounced . at a given apparent brightness ( observed flux ) ,
the quantity @xmath59 is fixed , where @xmath60 is the stellar surface brightness in the bandpass of interest and @xmath61 is the distance to the star .
substituting , @xmath62 into equation [ eqn.occur ] , and assuming that the planet population is distance - independent so that the distance factor can be moved outside the period and radius integrals , the scaling relation for observed occurrence becomes : @xmath63 if @xmath64 , ( * ? ? ?
* e.g. ) , closer and hotter host stars are more likely to be included in transit - selected samples .
stellar age and metallicity , which affect the relationships between stellar mass , radius , and surface brightness , are also biased as a result .
a correlation between stellar properties and distance can modulate the degree of this bias .
for example , if more distant stars tend to be more evolved along the main sequence and thus hotter , the bias will be less than if @xmath61 and @xmath60 are independent .
equation [ eqn.occur2 ] does not consider that star of a certain mass or radius may be _ over - represented _ in the parent population : this is discussed in the next section .
the target catalogs of transit surveys such as _ kepler _ are selected using a number of criteria , and chief among these is apparent magnitude .
a magnitude - limited sample of stars will be biased towards the most luminous objects , which will be included to greater distances @xcite .
these stars may be either more massive , more evolved , or both . at a given @xmath17 and thus fixed surface brightness @xmath60 ( ignoring the weak dependence of surface brightness on gravity and metallicity ) ,
the signal @xmath24 from a star during a transit will scale as @xmath65 . modifying equation [ eqn.snr0 ] appropriately , we find that the transit signal - to - noise ratio scales as @xmath66 the smallest planet that can be detected at a given snr will scale as @xmath67 multiplying a power - law distribution of planet radii ( equation [ eqn.distribution ] ) by the probability that a planet is on a transiting orbit ( @xmath68 ) and integrating over all planet radii down to @xmath69 gives the relation @xmath70 at a fixed color / temperature / surface brightness @xmath60 , a magnitude - limited survey will include stars of radius @xmath5 out to a distance @xmath71 .
assuming , for the moment , that transits can be detected to arbitrarily large distances , then integrating equation [ eqn.number ] over a homogeneous volume of radius @xmath72 yields @xmath73 for @xmath57 and at a given @xmath19 , @xmath12 scales as @xmath74 .
this relation illustrates how larger , more evolved stars can be preferentially included in a transit - selected sample despite the fact that transits of these stars are more difficult to detect .
although target stars in a magnitude - limited sample will be included to a distance @xmath71 , a planet of radius @xmath1 can only be detected to a distance @xmath75 where , according to equation [ eqn.snr2 ] , @xmath76 the detection limit decreases with @xmath5 while the inclusion limit @xmath72 _ increases _ with @xmath5 .
these limits coincide ( @xmath77 ) at a stellar radius @xmath78 : @xmath79 where @xmath80 is a constant factor , @xmath19 is in days , @xmath49 is in earth radii and @xmath18 is in solar masses .
( we calculate values of of @xmath80 for the _ kepler _ survey in section [ sec.malmquist ] . )
detections of planets of a given size around stars with @xmath81 is magnitude - limited and subject to a stellar radius bias that scales as @xmath82 , because the sample volume increases as @xmath83 and the transit probability increases as @xmath5 . for stars with @xmath84 , a survey is limited to a volume propoortional to @xmath85 ( see equation [ eqn.small ] ) , and the bias scales as @xmath86 , a weak dependence on @xmath5 in the opposite sense .
the critical stellar radius @xmath78 is most sensitive to planet radius and the dependence on period and stellar mass is weak .
to evaluate biases and selection effects in the _ kepler _ survey we modeled target stars with isochrones from the dartmouth stellar evolution database @xcite interpolated onto a 0.1-dex grid of metallicities using the on - line tool . for each star
, we compared adjusted kic parameters ( @xmath87 , @xmath88 , @xmath89 [ fe / h ] ) to model predictions using bayesian statistics .
specifically , we calculated a probability or weight @xmath90 for each model according to : @xmath91}p(m_*)p(t_*)p({\mathcal f})p(\zeta),\ ] ] where parameters with a `` hat '' are the dartmouth model values and @xmath92 , @xmath93 , @xmath94 , and @xmath95 are the priors for initial stellar mass ( initial mass function , imf ) , age , metallicity , and a modified distance modulus @xmath96 , where @xmath97 is the galactic latitude .
the modified distance modulus accounts for the finite dispersion of stars above the plane of the milky way , but neglects the vertical displacement of the sun .
we used an sdss @xmath98-band modulus @xmath99 , where @xmath100 is the observed apparent magnitude and @xmath101 is the absolute magnitude from the dartmouth models .
we ignored interstellar extinction , which will be @xmath102 magnitudes @xcite .
( while estimation of stellar parameters is sensitive to interstellar _ reddening _ , the amount of interstellar _ extinction _ is small compared to the uncertainties in the distance modulus . )
estimates of @xmath17 and [ fe / h ] from the kic were adjusted by -100 k and 0.17 dex , respectively and we used @xmath103 k , @xmath104 dex , and @xmath105 dex , based on a comparison of kic values with those spectroscopic values listed in b12 . for priors we adopted the @xcite imf , and a uniform age distribution over 1 - 13 gyr .
the latter corresponds to a constant rate of star formation since the advent of the galactic disk @xcite , but ignores the youngest stars , around which planets are more difficult to detect .
the metallicity distribution of _ kepler _ target stars is unknown and may be complex ; the field is not parallel to the galactic plane and may include members of a metal - poor `` thick disk '' population @xcite .
we used the metallicity distribution predicted by the the trilegal stellar population model @xcite as a prior .
stars in the direction of the center of the _ kepler _ field ( @xmath106 , @xmath107 ) were simulated to a cutoff magnitude @xmath108 . when compared to 2mass counts ,
trilegal counts agree with observations at least down to @xmath109 , but fail at @xmath110 , possibly due to incorrectly modeled bulge red giant branch stars and dust @xcite .
however , the kepler field cuts off at @xmath111 and and only 18 of the 84 ccd centers lie at @xmath112 .
the ( mostly default ) values for trilegal parameters are listed in table [ tab.params ] .
trilegal also reports a value of @xmath113 for each simulated star and we used these to construct a prior distribution of @xmath114 .
our priors are relaxed in the sense they only exclude very unlikely masses , ages , or metallicities .
it is also possible to impose priors on the stellar parameters @xmath17 and @xmath8 using the predictions of a stellar population model , but we consider such predictions too uncertain to justify this approach
. for each star , equation [ eqn.weight ] returns an array of values for @xmath90 corresponding to the grid of dartmouth models .
most values of @xmath90 are negligibly small and the corresponding models were ignored . from the remainder ,
the most probable ( highest @xmath90 ) model and accompanying parameters such as @xmath5 were identified .
statistics of the distribution of possible values were calculated , e.g : @xmath115 because the distributions can be very non - gaussian , we defined the fractional uncertainty in a stellar parameter to be one - half of the range encompassing 68% of the total probability ( normalized @xmath90 ) divided by the most probable value .
we found that uncertainties in the radii of late g- and k - type dwarf stars hosting kois is typically @xmath11615% , but are substantialy higher ( @xmath117% ) among some f- and g - type stars because of the coincidence of the dwarf and ( sub)giant branches ( figure [ fig.radunc ] ) .
evolved stars ( i.e. kic @xmath118 ) also have comparatively larger uncertainties .
the cluster of putative m `` dwarfs '' with radius uncertainties of @xmath11625% might be misclassified giant stars @xcite .
our estimated uncertainties are certainly lower bounds because ( 1 ) the errors in the stellar parameters @xmath17 , [ fe / h ] , and especially @xmath119 are themselves not gaussian - distributed , as presumed in equation [ eqn.weight ] ; and ( 2 ) we do not consider errors in the dartmouth models themselves .
eddington bias occurs when errors in measurement scatter more frequent values in a population to less frequent values at a higher rate than the reverse process .
this systematically inflates the observed frequency of rare members @xcite . because the distribution of planets with radius is a steep power law @xcite , errors in radius ( fractional standard deviation @xmath120 ) will bias the number of larger planets upwards
this will inflate the rate of planet occurrence @xmath12 above a given cutoff in radius @xmath121 .
planets with radius @xmath49 will appear to be larger than the cutoff if the error in stellar radius is larger than @xmath122 .
if errors in stellar radius are gaussian - distributed , the fraction of stars that satisfy that condition is @xmath123 .
the fractional upward bias in planet occurrence is the integral of this function with the normalized planet radius distribution , minus the intrinsic occurrence ( normalized to unity ) : @xmath124 where @xmath125 .
@xmath126 increases with @xmath120 and , if @xmath57 , reaches 18% when @xmath127% .
we estimated the amount of eddington bias in the apparent radius distribution of kois using the procedures described in section [ sec.methods ] .
for each koi we calculated the likelihood weight @xmath90 ( equation [ eqn.weight ] ) for all possible stellar models consistent using the parameters of the host star .
corresponding to each model we calculated a revised planet radius @xmath128 , where @xmath1 is the radius from b12 , @xmath129 is the model stellar radius and @xmath5 is the stellar radius of the maximum likelihood model ( highest @xmath90 ) .
the radius distribution , weighted by @xmath90 , is summed over all koi host stars and normalized .
this is compared to the observed distribution of @xmath1 ( figure [ fig.eddington ] ) .
the latter is not the _ intrinsic _ distribution , which must account for the probability that a planet transits and is detected @xcite .
as expected , eddington bias increases the apparent number of neptune - size and larger planets .
the bias is 17% above @xmath130 , demarcated by the vertical dashed line in figure [ fig.eddington ] , where the normalized distributions are equal .
the bias also suppresses the peak in the distribution at a jupiter radius .
corollaries of these results are that the actual occurrence rate of neptune - size planets is smaller than previously reported ( * ? ? ? *
i.e. ) , and that the intrinsic peak at @xmath131 is more pronounced than is apparent .
in addition , eddington bias decreases the apparent slope in the radius distribution ( figure [ fig.eddington ] ) .
this is a consequence of the observed turnover in the number of planets smaller than @xmath132 , and whether more large planets are scattered to smaller radii than vice versa .
_ kepler _ observations are incomplete for @xmath133 and while the intrinsic radius distribution of planets is presumed to turn over , the radius at which this actually occurs is not known and awaits a better understanding of the efficiency of _ kepler _ detection of small signals .
if the turnover below @xmath134 is real , then the intrinsic slope of the radius distribution is _ steeper _ than observed ( @xmath57 ) .
but if a scale - free power - law distribution continues to much smaller radii , then eddington bias affects the magnitude , but not the slope of the distribution . we simulated eddington bias on artificial samples of planets with radii drawn from a power - law distribution with variable index @xmath53 .
these radii replaced actual koi estimates in a repeat of the analysis described above .
the power - law index of the binned apparent distribution @xmath135 above some minimum radius @xmath136 is calculated by maximum likelihood : @xmath137 , where the summation is over all @xmath138 . as expected ,
while eddington bias significantly increases the fraction of planets with @xmath139 , the power - law index is relatively unchanged ( figure [ fig.eddington2 ] ) .
malmquist bias is the preferential inclusion of intrinsically luminous objects in a magnitude - limited survey due to the rapid increase in sampling volume @xmath140 with distance @xmath72 to which an object is included . among large , readily - detected objects ( planets ) in a magnitude - limited transit survey , the bias is even greater ( @xmath141 ) because the probability of a transiting geometry is proprtional to @xmath5 which , at a given effective temperature , scales with @xmath72 ( see section [ sec.targetbias ] ) . at a given apparent magnitude and planet radius ,
there is a maximum stellar radius @xmath142 to which a survey is essentially complete , i.e. not limited by the snr of a transit event .
we estimated @xmath78 as a function of @xmath1 by establishing the detection limit at different _ kepler _ magnitudes .
the _ kepler _ target catalog was constructed with different criteria for stars with @xmath10 and @xmath143 @xcite ; it is probably nearly complete for dwarf stars to @xmath144 but only includes selected dwarfs with @xmath143 @xcite .
we adopted a snr limit of 7.1 and an observation period of 487 d ( b12 ) . to estimate the noise of a typical dwarf star we performed a polynomial fit to a running median ( @xmath145 ) of 3 hr combined differential photometric precision ( cdpp ) values for _ kepler _ targets with @xmath146 , presumed mostly dwarfs .
this gave an estimate of the intrinsic 3 hr rms noise level as a function of @xmath147 ; @xmath148 the median noise at @xmath149 is 54 ppm .
we performed a similar analysis on stars with kic @xmath118 , presumably subgiants and giants , that constitute a locus of comparatively `` noisy '' targets , and found : @xmath150 for @xmath144 dwarfs , @xmath151 and at @xmath152 , @xmath153 . at @xmath144 ,
for a median orbital period @xmath154 d and @xmath155 , malmquist bias favors stars as large as @xmath156 . at @xmath108 , only stars with @xmath157 are favored because of higher noise at fainter magnitudes .
the situation is more extreme for giant planets ( @xmath158 ) , where malmquist bias will favor evolved stars as large as 10 - 20@xmath159 , presuming giant planets exist around such stars , as we discuss below .
bias towards larger stars , coupled with uncertainties in stellar radius , leads to underestimates of stellar - and hence planetary - radii .
we quantified this effect using the machinery described in section [ sec.methods ] , with the addition of a malmquist bias factor .
for each koi - hosting star , we evaluated the mean stellar radius by averaging over all stellar models weighted by @xmath90 ( from equation [ eqn.weight ] ) and multipled by either @xmath160 , where @xmath161 , or @xmath162 , if @xmath161 .
the ratio of the `` naive '' mean model radius to the bias - weighted mean radius is plotted in figure [ fig.malmquist ] vs. the nominal planet radius published in b12 .
deviation of this factor from unity can be considered the error in radius that results if malmquist bias is not taken into account .
about two - thirds of all koi - hosting stars , and the vast majority of those hosting planets smaller than neptune have predicted malmquist bias values @xmath010% . however , the majority of larger planets may have significantly underestimated radii , some by a factor of two . this dichotomy occurs because _ kepler _ detection of large planets is limited by the magnitude limit of the target catalog , not the snr of transit .
we emphasize that these calculations are _ statistical _ , i.e. we are calculating the expectation values of probability distributions with stellar radius , and that actual errors will vary .
nevertheless , the host stars of many giant planets may be more larger , more distant , and more luminous , and the radii of their planets may be significantly underestimated .
inclusion of larger , evolved stars means that some kois may be astrophysical false positives , e.g. m dwarf companions masquerading as planets @xcite , a possibility that we discuss in section [ sec.discussion ] .
the metallicity of host stars is an important parameter in studies of planet statistics .
a correlation between stellar metallicity and the presence of giant planets has been unambiguously established @xcite and is consistent with a prediction by the core - triggered instability theory of giant planet formation @xcite , i.e. that a solid core that initiates runaway accretion before the gas dissipates is more likely to form in a disk with a higher abundance of solids .
doppler surveys have failed to find any correlation between metallicity and the occurrence of neptune - size or smaller planets @xcite .
@xcite found that the average @xmath163-@xmath98 color of most _ kepler _ stars with small candidate planets was no different from the average of all stars at a given @xmath164-@xmath165 color , except for late k and early m - type stars ; those with planets have redder @xmath163-@xmath98 colors and @xcite argued that these are more metal - rich .
however , this difference may be an artifact of contamination of the sample by evolved stars , which have bluer @xmath163-@xmath98 colors than dwarfs and make the overall sample , but not the koi - hosting sample , bluer @xcite .
indeed , @xmath163-@xmath98 color might be insensitive to or depend only weakly on metallicity for these spectral types @xcite .
@xcite report metallicities of 78 late k and m dwarfs with kois based on infrared spectra .
the mean value , -0.09 , is consistent with the metallicity of m dwarfs in the solar neighborhood @xcite .
the average metallicity of _ kepler _ m dwarfs is not known but these intrinsically faint stars are within a few hundred pc of the sun @xcite .
the metallicities of stars of transiting planets need not be representative of the underlying population of planet - hosting stars .
metals are an important source of opacity in the atmospheres of cool stars , and , all else being equal , metal - poor dwarf stars should have smaller radii .
a transiting planet will be more detectable around a metal - poor subdwarf than a metal - rich dwarf star , and thus the host stars of kois will be biased towards metal - poor representatives of the overall population .
if sufficiently large , this bias could obfuscate any intrinsic relationship between stellar metallicity and the presence of planets .
we calculated the metallicity bias , i.e the expected metallicity of stars with detected transiting planets minus the expected metallicity , for all _ kepler _ quarter 6 target stars using eqns .
[ eqn.bias ] and [ eqn.occur ] , and the methods described in section [ sec.methods ] .
the difference between the `` naive '' mean metallicity of dartmouth models for each star , and the biased mean using the factor of equation [ eqn.occur ] , is plotted vs. adjusted kic effective temperature in figure [ fig.metalbias ] . as expected , the metallicity bias is negative except for a locus of positive values corresponding to evolved stars , for which radius _ decreases _ with increasing metallicity , e.g. @xcite .
the bias is small ( mean of -0.017 among dwarfs ) for the following reasons : ( i ) the geometric transit probability is proportional to stellar radius and thus increases with metallicity , countering the effect of metallicity on transit depth ; and ( ii ) the effect of metallicity on stellar radius is most pronounced among comparatively rare subdwarfs but has only a modest effect around solar metallicity , especially for the coolest stars @xcite . at the time
the first exoplanet around a main sequence star was confirmed , @xcite realized that highly - irradiated giant planets on close - in orbits may have anomalously large radii .
after sufficient numbers of transiting giant planets were discovered , it became apparent that some were `` inflated '' compared to theoretical predictions @xcite .
planets larger than @xmath166 can not be explained by conventional interior models of gas giants and require an additional source of internal energy to inflate them @xcite .
several non - exclusive explanations for the requisite energy source have been put forward @xcite .
one important clue is that planets experiencing higher irradiance or having higher emitting temperature are more likely to be inflated .
correlations between equilibrium temperature and radius have been reported among transiting giant planets discovered in ground - based surveys @xcite . among _ kepler _ giant planet candidates , inflation appears to occur only above an irradiance of about @xmath167 ergs s@xmath168 @xmath169 ( * ? ? ?
* hereafter d11 ) . where information about stellar parameters is limited
, spurious correlations can appear if two supposedly independent planetary parameters are related to the same , uncertain stellar parameter . in the absence of parallax or precise information on surface gravity ,
the radius of a star is constrained only by models of stellar atmospheres , stellar evolution , and galactic population .
uncertainty in stellar radius translates into corresponding uncertainties in both stellar luminosity and transiting planet radius .
because the radiation that a planet receives from a star is proportional to stellar luminosity , errors in irradiance and planet radius due to errors in stellar radius will be positively covariant . at least in principle , an apparent , positive trend between irradiation and planet radius could be created merely by errors in stellar radius .
we simulated the impact of this systematic with an analysis of kois similar to , but not identical to that of d11 .
we selected all kois with estimated radii of @xmath170 from b12 , excluding those listed as false positives or `` ambiguous '' in table 1 of d11 .
as in section [ sec.methods ] , we identified the best - fit dartmouth model for each host star based on a @xmath171 minimization of the difference with adjusted kic values of @xmath17 , @xmath8 , and [ fe / h ] , after applying corrections of -100 k to @xmath17 and 0.17 dex to [ fe / h ] ( br11 ) .
we assumed standard deviations of 200 k , 0.36 dex , and 0.3 dex , respectively based on br11 and 190 stars where both kic and spectroscopy - based parameters are available ( b12 ) .
if no kic value for [ fe / h ] was available we assumed solar metallicity . to estimate the maximum possible effect , no constraints other than the dartmouth evolutionary tracks were used , i.e. we equally weighted masses , ages between 1 - 13 gyr , and metallicities between -2.5 and + 0.5 dex .
orbit - averaged stellar irradition of the planet was calculated based on the model luminosity and mass , the orbital period , and assuming a circular orbit .
( non - circular orbits change the mean irradiance only slightly . )
planet radius was calculated from the transit depth and stellar model radius and we did not account for limb darkening .
the encircled points in figure [ fig.radirr ] indicate the best - fit planet radius vs. irradiance .
three kois ( 217.01 , 774.01 , and 1547.01 ) have estimated irradiances @xmath172 ergs s@xmath168 @xmath169 and @xmath173 , but only marginally so .
fifteen kois have re - estimated radii @xmath174 even though the values listed in b12 exceed the criterion @xmath175 .
twelve of these have kic impact parameters @xmath176 , suggesting problematic ( or extreme grazing ) transit solutions .
another ( koi 1419.01 ) has an implausible @xmath177 which is inconsistent with its transit duration of @xmath178 h and period @xmath179 d. koi 377.02 ( _ kepler _
9-b ) has an erroneous transit depth reported in the mast .
the best - fit dartmouth model assigns a somewhat smaller radius ( @xmath180 ) to the host star of koi 1193.01 and thus makes the planet smaller as well .
we excluded all planets with newly estimated radii @xmath181 from our analysis .
we assessed the trends produced by correlated errors in planet radius and irradiation by considering all dartmouth models that satisfy @xmath182 , where @xmath183 is the minimum ( best - fit ) value , and 8.02 is the @xmath184 corresponding to a 95.4% ( 2@xmath185 ) confidence interval for @xmath186 degrees of freedom ( stellar parameters ) .
because there are too many models to plot , we only show a random subsample of 200 such models for each koi as the small points in figure [ fig.radirr ] .
these clearly show that correlated errors will tend to scatter points between the high irradiation / inflated and the low irradiation / uninflated regions of the diagram .
the paucity of kois with inflated radii ( @xmath173 ) in the low irradiance region ( upper right hand domain of figure [ fig.radirr ] ) supports the contention that the inflation of giant planets is related to stellar irradiation or planet equilibrium temperature .
furthermore , kendall s and spearman s rank correlation tests of all kois with @xmath187 yield @xmath30 values of 0.246 and 0.364 , respectively , and corresponding @xmath188 ( significance ) values of @xmath189 and @xmath190 .
these low false - positive probabilities indicate a significant correlation between irradiation and plane radius . however , these statistics do not account for the systematic effect of correlated errors in radius and irradiation .
we simulated the effect of correlated errors by analyzing 10000 null realizations of the data where radii and orbital periods of kois were randomly shuffled among host stars and the transit depths were recomputed using equation [ eqn.depth ] , thus destroying any intrinsic correlation between radius and irradiation .
in computing each realization we include all kois with @xmath191 to account for small planets that may appear larger , but in each monte carlo realization , as with the real sample , we limited the statistical analysis to 8 - 22@xmath192 .
new ( `` observed '' ) estimates of kic stellar parameters were constructed from the `` true '' values by adding random , gaussian - distributed offsets with standard deviations of 200 k for @xmath17 , 0.36 dex for @xmath8 , and 0.3 dex for [ fe / h ] .
best - fit dartmouth models were found for each parameter set , the planet radii and irradiation values were determined , and the correlation statistics were calculated .
new @xmath188 values for the fraction of kois in the low - irradiance / inflated - radius zone , and kendall s @xmath30 , and spearman s @xmath30 were computed as the fraction of mc realizations that are smaller ( more significant ) than the observed values .
the distributions for the first two metrics are shown in figs .
[ fig.ninflate ] and [ fig.kendall ] and the @xmath188 values are @xmath193 and @xmath194 , respectively .
the result for the spearman s rank coefficient is similar , with @xmath195 .
* hereafter dr12 ) reported a weakly significant ( @xmath196 or @xmath197 ) trend of decreasing radius of _ kepler _ ( candidate ) giant planet with increasing metallicity of the host star .
she examined the ratio @xmath198 of 218 kois from @xcite with estimated radii of 5 - 20 @xmath192 and the correlation with estimated values of [ fe / h ] from the kic .
she interpreted the decline as evidence that giant planets around metal - rich stars have larger solid cores and , for the same total planet mass , smaller radii @xcite .
figure [ fig.depth ] is an updated version of figure 1 in dr12 based on the more recent release of kois with revised radii @xcite .
it includes 225 kois with @xmath199 and host stars with kic - determined metallicities . as in figure 1 from dr12 , a running median ( @xmath200 )
is plotted .
the kendall @xmath30 correlation coefficient is -0.032 , indicating no signficant correlation ( @xmath201 ) .
we were unable to reproduce the result of dr12 by simple cuts on this sample to approximate the earlier koi sample , perhaps because many stellar radii ( and hence planet radii ) have been revised @xcite .
we also emphasize that the values of [ fe / h ] in the kic are no more accurate than @xmath2020.3 dex ( br11 ) .
irrespective of any physical phenomenon , one would expect to observe a decrease in @xmath198 with increasing metallicity simply because metal - rich dwarfs tend to be larger than metal - poor dwarfs , and hence transit depths will be smaller ( equation [ eqn.depth ] ) .
we modeled this effect with 10000 monte carlo realizations of the koi catalog .
there are two effects from increasing the radii of the host stars of a given planet population :
one is that transit depths will become smaller and the planets will appear to be smaller .
the other is that some planets may fall below the lower radius cutoff ( 5@xmath192 ) and be excluded from the analysis .
the reverse is true for lower metallicity ; planets appear larger and a few planets may exceed the maximum cutoff ( 20@xmath192 ) .
we therefore considered kois over a broader ( 3 - 25@xmath192 ) range of radii , adopted this sample as representing the intrinsic ( `` true '' ) distribution of radii , estimated their apparent radii from the radius of the star and transit depth , and then applied the same radius criteria as dr12 .
we randomly shuffled the planet population among the host stars , thus destroying any intrinsic radius - metallicity correlation , computed the radii of the stars using the dartmouth stellar evolution models , and re - calculated the transit depths .
each monte carlo host star was assigned the corrected @xmath17 of the actual star it replaced .
we assigned an _ observed _ metallicity based on the kic value , a systematic correction @xmath203 of 0.17 dex ( br11 ) , a random normally - distributed error @xmath185 of 0.3 dex , and a prior distribution of _ intrinsic _ metallicities that is a guassian with mean @xmath204 and standard deviation @xmath205 .
this is equivalent to drawing metallicities from a single normal distribution with mean @xmath206 and standard deviation @xmath207 .
the radius of each monte carlo star was taken to be the median of all model radii with @xmath146 ( presuming they are dwarf stars ) , [ fe / h ] within 0.15 dex of the monte carlo model , and @xmath17 within 100 k. we did not apply any age criterion other than 1 - 13 gyr .
we then calculated @xmath198 using the shuffled planet radius and the median model radius stellar radius . for each monte carlo sample , we calculated kendall s @xmath30 and false positive probability for a correlation between the _ observed _ metallicities and the artificial transit depths .
median - filtered ( @xmath208 ) curves from these monte carlo realizations typically show a decline of @xmath198 with increasing metallicity .
figure [ fig.tau ] shows the distribution of @xmath30 from 10000 null realizations .
the value of @xmath30 from the actual koi sample is plotted as the dashed line .
61.6% of these null realizations produce a significant ( @xmath209 ) correlation and 71.6% of values are below ( and thus more significant than ) the actual value of -0.032 . for comparison
the dr12 value is -0.17 .
thus , negative correlations between metallicity and @xmath198 are to be expected soley as a consequence of the metallicity - radius relation of stars , although these monte carlo simulations indicate that there is a @xmath11640% chance that random errors in kic [ fe / h ] values would prevent such a correlation from being detected .
we have shown that selection effects for both transiting planets , and the target stars of transit surveys , combined with uncertainties in stellar radii , can bias the properties of host stars and their planets .
these effects are in addition to those previously identified by @xcite , @xcite , and @xcite , which concern effects arising from the sensitivity of detection efficiency to planet radius and period .
we have analyzed the effects of these systematics on the _ kepler _ survey and its catalogs of target stars and candidate planets , using current models of stellar evolution and galactic stellar populations to infer the properties of _ kepler _ stars .
we did not apply constraints from the relation between stellar density , transit duration , and orbital period because the relation also depends on unknown orbital eccentricity and argument of periastron , and is not applicable to non - koi stars .
we found that eddington bias from the steep distribution of kois with radius results in an overestimation of the overall frequency of planets with @xmath210 by about 15 - 20% of the actual value .
we also find that eddington bias acts to soften the `` bump '' in the distribution at jupiter - size planets .
this leads us to predict that the intrinsic peak at that radius is more pronounced .
the effect on the distribution of smaller planets depends on whether the turnover in the radius distribution below 2@xmath192 is real , or the result of incompleteness .
if the former , eddington bias acts to flatten the apparent slope of the radius distribution , and in this case we predict that the actual slope is steeper than the @xmath57 power - law . otherwise , the effect of eddington bias on the power - law index is about 0.1 .
we made statistical estimates of malmquist bias as a consequence of the magnitude limit of the target catalog .
the estimated bias for two - thirds of koi systems , including most kois smaller than neptune , is @xmath211% .
however , we found that bias is more prevalent and pronounced ( up to a factor of two in radius ) among larger candidate planets and their host stars , resulting from detection of these systems being governed by the apparent magnitude limit of the target catalog , rather than the snr of transit detection .
a malmquist bias towards more luminous stars raises the possibility of inclusion of unidentified evolved stars within the _ kepler _ target catalog ( in addition to a number of deliberately selected and clearly identified giant stars ) .
nominally , stars with large radii were removed by a vetting process that used a criterion of _ kepler _ detection of a @xmath134 planet @xcite .
however , kic - derived stellar radii are based on estimates of log @xmath163 and many of these are problematic .
kic photometry provides no information for the gravity of stars with @xmath212k ( @xmath213 ) , and subgiants would be assigned erroneously high log @xmath163 ( br11 ) .
there are bona fida subgiants hosting kois , e.g. the f5 subgiant hd 179070 @xcite .
spectroscopy of stars hosting candidate giant planets has revealed other instances in which subgiants were misclassified as cooler , main sequence dwarfs in the kic .
@xcite report a hot - jupiter - hosting f - type subgiant ( @xmath214 , @xmath215 ) .
based on spectra , they estimate @xmath216 , which is in contrast to its kic value of 4.55 .
likewise , the host of koi-423b , assigned @xmath217 in the kic , is an f7iv subgiant with @xmath218 @xcite .
three of five undiluted eclipsing binaries identified by @xcite as false positives among _ kepler _ giant planet candidates have masses larger than 1 m@xmath219 , and one of these is definitely an evolved star .
the mean difference between 190 pairs of kic and spectroscopic values of @xmath8 reported in @xcite is only 0.02 dex ( standard deviation of 0.36 dex ) .
nevertheless , astroseismically - derived @xmath8 values average 0.05 - 0.17 dex lower than kic values and astroseismically - determined radii are up to 50% larger @xcite . among koi - hosting stars
whose radius has been underestimated , small planets may actually be larger , even jupiter - size planets . in turn ,
giant `` planets '' may turn out to be diluted or undiluted stellar companions , a significant source of astrophysical false positives in transit surveys @xcite . based on a preliminary doppler survey
, @xcite estimated that about 40% of candidate giant planets are false positives and about one quarter of those are undiluted eclipsing binaries .
this also means that estimates of the occurrence of jupiters on close - in orbits @xcite must be revised downwards .
@xcite report that the occurrence of `` hot jupiters '' ( @xmath220 d ) in the _ kepler _ catalog is only half that seen in doppler surveys , and adjustment for a high false - positive rate would worsen this discrepancy .
one explanation for the discrepancy between the _ kepler _ and doppler surveys might be the presence of misidentified subgiant stars in the _ kepler _ target catalog
. the intrinsic distribution of planets may be different around evolved stars compared to main sequence stars .
planets have been discovered around subgiant stars @xcite , but giant planets appear to be rare with 0.6 au ( @xmath221 d ) of clump gk giants @xcite - corot-21b may be an exception @xcite .
the timescale of the decay of a planet s orbit due to dissipation of tides in a star s convective envelope scales as @xmath222 , where @xmath223 is the mass of the envelope .
hot jupiters are likely to be destroyed by infall and disruption inside the roche lobe as a star evolves off the main sequence , expands , and its convective envelopes thicken @xcite .
thus , one explanation for the comparative paucity of hot jupiters in the koi catalog is that , because of malmquist bias , many _ kepler _ targets are older stars or subgiants for which hot jupiters can not be detected , have been miscategorized as neptunes , or have been destroyed by orbital decay .
a comparison between the distributions of @xmath8 predicted by trilegal and that of the kic suggest no large ( @xmath22410% ) population of unidentified subgiants , however spectroscopy of candidate subgiants is needed to actually test this conjecture .
we have shown that , because metal - poor stars tend to have smaller radii than their metal - rich counterparts , stars with transiting planets will be biased towards metal - poor members , independent of any correlation between planets and metallicity . however , we estimate that this metallicity bias is only about -0.02 dex and can be neglected .
thus a comparison between the mean metallicity of stars with transiting planets and that of the overall target population is appropriate .
the mean metallicity of m dwarfs with kois , -0.09 @xcite , and solar - type stars with small planets , -0.01 @xcite , appears similar to that of the solar neighborhood : @xcite report a mean metallicity of @xmath225 for a volume - limited local sample of m dwarfs using a photometric calibration , and @xcite report a median metallicity of -0.06 for all stars in the solar neighborhood . whether the overall
_ kepler _ target population has a similar metallicity distribution is not yet known and additional observations are required . from our calculations
we conclude that such a comparison would not suffer from significant metallicity bias , but must take into account a dilution factor because stars without transiting planets are not necessarily stars without planets .
this dilution factor is large for a high planet occurrence @xcite .
we have shown how uncertainties in stellar radius or distance produce correlated errors in a planet s radius and the radiation received from the host star .
this effect can produce an artificial correlation in populations of planets where none exists .
recently , such a correlation has been found in both ground - based transit surveys and the _ kepler _ catalog , and highlighted as a test of mechanisms to explain the `` inflation '' of giant planets on close - in orbits .
we quantified the systematic effect of correlated errors in stellar radius in the case of the _ kepler _ kois and show that , despite this systematic , the result of d11 , i.e. that inflated planets are absent at low irradiance , is still significant . to maximize any systematic effect
, we used a very broad range of metallicities ( -2.5 to + 0.5 ) and no constraint on stellar distance ( e.g. , from a model of galactic structure ) , thus further strengthening our conclusion .
finally , we have shown how searches for trends of transiting planet radius with stellar properties may engender systematic errors unless the effect of those properties on apparent stellar radius - and hence planet radius - is taken into account .
we examined the tentative ( 2.3@xmath185 ) claim of dr12 that giant planets around metal - rich stars tend to have smaller transit depths , because they are smaller and perhaps have larger rocky cores . performing a similar analysis on the most recent koi catalog
, we were unable to reproduce that trend .
moreover , we performed simulations that show that the trend observed by dr12 could be easily explained by the dependence of stellar radius on metallicity
. two limitations of our analysis are that ( i ) we have asssumed gaussian - distributed errors in the corrected kic parameters @xmath17 , @xmath8 , and [ fe / h ] , and ( ii ) that the construction of bayesian priors on mass , age , and metallicity treat them as independent variables .
neither of these is absolutely correct ; the first assumption probably produces an underestimate of the uncertainty in stellar radius while the second assumption produces an overestimate of the uncertainty .
of course , any inadequacies in the dartmouth stellar evolution models themselves are not accounted for .
there are other systematics effects which may be present in transit surveys .
two - thirds of solar - type ( f6-k3 ) stars are found in multiple systems @xcite . at the typical distance of _ kepler _ kois with solar - type hosts ( 950 pc ) , one 4 arc - second pix subtends about 3800 au , sufficient to include nearly all companions to primaries @xcite .
the presence of an unresolved companion , or any background star , will dilute the transit signal .
transits otherwise just above the detection threshold might be rendered invisible . as a consequence
, members of multiple systems may be underrepresented among stars with transiting planets .
for equal - mass binaries ( twins ) where the transit signal is lower by a factor of 2 , the fractional noise will decrease by @xmath226 ( due to the doubling of the signal compared to a comparable single star ) and thus the radius of the smallest detectable planet will increase by a factor of @xmath227 , or about 1.2 . for a power - law size distribution ( equation [ eqn.distribution ] ) , the number of detectable planets _ per star _ will decrease by a factor of @xmath228 , or 0.64 for @xmath229 .
however , nearly - equal mass binaries represent only 12% of all binaries
@xcite and systems with mass ratios @xmath230 and luminosity ratios @xmath231 , where the dilution will be much smaller , are the norm .
star counts reach @xmath1161000 mag@xmath168 deg@xmath232 at @xmath233 , and so there is only a few % chance of significant dilution by an unrelated star . to the extent that stellar variability inhibits transit detection , younger , and more active stars will be also underrepresented among kois . the best defense against the systematic errors we have described is better characterization of the target stars of transit surveys , especially those hosting planets .
this will reduce , but not entirely eliminate , these biases .
spectroscopic characterization and refinement of the properties of a fully representative sample of _ kepler _ target stars , not just the koi hosts , is vital to robust statistical analyses of the properties of transiting planets and their parent stars stars , and such programs are underway @xcite .
spectra of modest resolution ( @xmath234 ) @xcite or snr ( @xmath235 ) @xcite ( but not both ) can provide substantial improvements over photometry alone . the _ gaia _
( originally global astrometric interferometer for astrophysics ) mission , scheduled for launch in august 2013 , will obtain parallaxes of stars as faint as 16th magnitude with a standard error of @xmath23640 @xmath113as @xcite .
this will allow the luminosity of a solar - type star to be determined with an error about 15% and its radius with an error of about 8% .
the distance to brighter stars will be measured with even greater precision .
_ gaia _ will also obtain moderate - resolution spectra in a narrow region centered on the ca ii triplet region which can be used to classify stars @xcite and measure their radial velocities to a precision of a few km sec@xmath168 .
radial velocites , combined with parallaxes , yield space motions and membership in distinct stellar populations ( e.g. thin disk , halo ) .
_ gaia _ data will also benefit future transit surveys that will cover all of or a large part of the sky @xcite .
this research was supported by nsf grants ast-09 - 08406 and nasa grants nnx10ai90 g and nnx11ac33 g to eg .
the _ kepler _ mission is funded by the nasa science mission directorate , and data were obtained from the mukulski archive at the space telescope science institute , funded by nasa grant nnx09af08 g .
ll parameter & value + dust : & + extinction at @xmath50 & 0.0378 + scale height & 110 pc + scale length & 100 kpc + position of sun : & + galactocentric radius & 8700 pc + height above disk & 24.2 pc + thin disk : & + zero - age scale height & 95 pc + radial length scale & 2.8 kpc + local surface density & 59 m@xmath219 pc@xmath232 + star formation rate & 2-step + thick disk : & + scale height & 800 pc + radial length scale & 2.8 kpc + local density & @xmath237 m@xmath219 pc@xmath232 + star formation rate & 11 - 12 gyr constant + halo : & + shape & @xmath238 spheroid + scale length & 2.8 kpc + oblateness & 0.65 + local density & @xmath239 m@xmath219 pc@xmath232 + star formation rate & 12 - 13 gyr + | statistical analyses of large surveys for transiting planets such as the _ kepler _ mission must account for systematic errors and biases .
transit detection depends not only on the planet s radius and orbital period , but also on host star properties .
thus , a sample of stars with transiting planets may not accurately represent the target population .
moreover , targets are selected using criteria such as a limiting apparent magnitude .
these selection effects , combined with uncertainties in stellar radius , lead to biases in the properties of transiting planets and their host stars .
we quantify possible biases in the _ kepler _ survey .
first , eddington bias produced by a steep planet radius distribution and uncertainties in stellar radius results in a 15 - 20% overestimate of planet occurrence .
second , the magnitude limit of the _ kepler _ target catalog induces malmquist bias towards large , more luminous stars and underestimation of the radii of about one third of candidate planets , especially those larger than neptune .
third , because metal - poor stars are smaller , stars with detected planets will be very slightly ( @xmath00.02 dex ) more metal - poor than the target average .
fourth , uncertainties in stellar radii produce correlated errors in planet radius and stellar irradiation . a previous finding
, that highly - irradiated giant are more likely to have `` inflated '' radii , remains significant , even accounting for this effect .
in contrast , transit depth is negatively correlated with stellar metallicity even in the absence of any intrinsic correlation , and a previous claim of a negative correlation between giant planet transit depth and stellar metallicity is probably an artifact . |
in the last two decades , there has been a revival on the study of classical wormhole solutions either in general relativity ( gr ) or in alternative theories of gravitation . a cornerstone in this revival
has been the paper of morris and thorne @xcite in which , they derived the properties that a space - time must have in order to hold up such a geometry .
let us recall that the most salient feature of these space - times is that an embedding of one of their space - like sections in euclidean space displays two asymptotically flat regions joined by a throat .
there are several reasons that support the interest in these solutions .
one of them is the possibility of constructing time machines @xcite which violate hawking s chronology protection conjecture @xcite .
another reason is related to the nature of the matter that generates this solution : it must be `` exotic '' , _
i.e. _ its energy density takes negative values in some reference systems .
this entails violations of the weak energy condition ( wec ) which , although not based in any strong physical evidence @xcite , is deep enough in the body of relativists belief so as to abandon it without serious reasons .
these violations are in fact encoded in the evolution of the expansion scalar , governed by raychaudhuri s equation @xcite . as wec is violated for any static spherically symmetric and has recently been shown to be the case also for any static in gr @xcite , the attemps to get around wec violations have led two different branches of research .
firstly , there have been an increasing number of works in non - standard gravity theories .
brans - dicke gravity supports static wormholes both in vacuum @xcite and with matter content that do not violate the wec by itself @xcite . also , there are analysis in several others alternative gravitational theories , such as @xmath0 @xcite , einstein - gauss - bonnet @xcite and einstein - cartan models @xcite . in all cases ,
the violation of wec is a necessary condition for static to exist , although these changes in the gravitational action allows one to have normal matter while relegating the exoticity to non - standard fields .
this lead to explore the issue of wec violation in non - static situations and see which is the case for dynamic , time - dependent wormholes .
the first of this kind of analysis , due to roman @xcite , was devoted to the case of inflating lorentzian wormholes .
these are wormholes of the morris - thorne type embedded in an inflating background , with all non - temporal components of the metric tensor multiplied by a factor of the form @xmath1 , being @xmath2 related to the cosmological constant .
the aim of roman was to show that a microscopical could be enlarged by inflationary processes .
however , this construction also violate the wec .
the other works , due to kar and kar and sahdev @xcite , pointed at testing whether , within classical gr , a class of non - static , non - wec - violating wormholes could exists .
they used a metric with a conformal time - dependent factor , whose spacelike sections are @xmath3 with a metric , like the ones analyzed in @xcite . in this work , we shall extend this last analysis by studying more general cases of evolving wormholes , given by a generic morris - thorne metric , in the presence of matter described by a given stress - energy tensor . to proceed
further we shall make two anstze .
firstly , we give a specification of the redshift function @xmath4 of morris - thorne consistent , for any hypersurface of constant @xmath5 , with asymptotic flatness and a metric free of horizons , as was the one made for the static case in @xcite .
then we shall show how to choose a particular functional form for the trace of the stress - energy tensor matter in order to get different solutions for the conformal factor . at this point
, we shall derive an analytical solution which we use later to analyze the wec violation scenario and some human tranversability criteria .
we adopt the following diagonal metric for the space - time : @xmath6 with @xmath7 a conformal factor , finite and positive defined throughout the domain of @xmath5 , and @xmath8 , the @xmath9 line element as usual . when @xmath10 is constant , this general metric stands for the ones firstly analyzed in the work of morris and thorne @xcite .
when @xmath11 , it reduces to the particular case studied by kar @xcite . in the spirit of @xcite
, we make the ansatz @xmath12 , where @xmath13 is a positive constant to be determined .
this choice guarantees that the redshift function is finite everywhere , and consequently there is no event horizon . for the stress - energy tensor
we take that of an imperfect fluid represents time - like unit vectors and @xmath14 space - like unit vectors in the radial direction respectively.]@xcite , @xmath15 + 2 \,q \,\zeta_{(\mu } \xi_{\nu)}\ ] ] where @xmath16 stands for the energy density , @xmath17 for the isotropic fluid pressure , and @xmath18 for the energy flux . the anisotropy pressure
@xmath19 is the difference between the local radial ( @xmath20 ) and lateral ( @xmath21 ) stresses . with the foregoing assumptions , the nonvanishing einstein s equations are @xmath22 @xmath23 @xmath24 @xmath25 in the expressions given above ,
dashes denote derivatives with respect to @xmath26 while dots , derivatives with respect to @xmath5 .
it seems convenient to assume that the trace of the stress energy tensor will be a separated function of @xmath5 and @xmath26 , t_^ = + \ { - 2 e^-2 } + & : = & ( t ) [ traza ] after separating variables , equation ( [ traza ] ) can be rewritten as , @xmath27 @xmath28 = { \cal c } \label{radial}\ ] ] where @xmath29 is an arbitrary constant which for simplicity , we shall take equal to 0 . performing the change of variables
@xmath30 , equation ( [ radial ] ) transforms into a bernoulli equation , @xmath31 the additional change @xmath32}$ ] , allows for the general solution to be written as @xmath33 with @xmath34 , and @xmath35 an integration constant to be determined .
note that , far from the throat , _
r e^-2 = 1 , or equivalently , _
r = 0 , where @xmath36 is the embedding function @xcite .
moreover , for any hypersurface of constant @xmath5 the mouths of the wormhole need to connect two asymptotically flat spacetimes , thus the geometry at the wormhole s throat is severely constrained .
indeed , the definition of the throat ( minimum wormhole radius ) , entails for a vertical slope of the embedding surface , _
rr^+ = _ rr^+ = ; [ martikahermosa ] besides , the solution must satisfy the _ flaring out condition _
@xcite , which stated mathematically reads , - > 0 .
[ martikaestamashermosaaun ] equations ( [ martikahermosa ] ) and ( [ martikaestamashermosaaun ] ) will be satisfied if and only if @xmath37 thus , in order to fix the constant @xmath35 , we we must select a value for the dimensionless radius of the throat ( @xmath38 ) such that eq .
( [ fo ] ) be satisfied .
nevertheless , the absolute size of the throat also depends on @xmath13 . as an example let us impose @xmath39 which , using ( [ solucion ] ) , sets @xmath40 and @xmath41 .
the aforementioned properties of @xmath42 , together with the definition of @xmath4 and @xmath43 , entail that the metric tensor describes , for @xmath44 constant , two asymptotically flat spacetimes joined by a throat .
thus , equation ( [ solucion ] ) represents an analytical solution for an evolving geometry , which , due to the choice we made for the trace , is independent of @xmath7 . to obtain the complete behavior of the metric
, we have to define @xmath45 in order to solve for @xmath10 in equation ( [ 5a ] ) .
we take instead eq .
( [ 5a ] ) as the definition of @xmath46 in what follows .
having at our disposal an explicit solution , given in the form of @xmath47 , we can now turn to the question of whether it entails wec violations .
of course , we have to note that , in the general case , it will ultimately depend on the specific choice of @xmath7 , but we use here the fact that , due to the form in which the solution for @xmath47 was computed , it is independent of the explicit form of the conformal factor . thus , we are able to study different functional forms of @xmath7 for the same @xmath47 .
the wec asserts that for any time - like vector @xmath48 : t _ v^v^0 .
the physical significance of this condition is that the observed energy density , as measured by any time - like observer , has to be positive .
the connection of wec violations with the existence of wormholes geometries has been widely studied before @xcite . in the case of evolving wormholes ,
kar @xcite has demonstrated that there exist solutions which do not violate the wec for a specific choice of the redshift function .
his analysis is indeed general , in the sense that even without having worked out any explicit solution , he could conclude that there exist some lorentzian evolving geometries with the required matter not violating the wec .
we must recall however that kar dealt with the particular case @xmath49 .
we pursue here to refine his analysis , by means of the solution for the metric component @xmath50 given in the previous section . with the stress - energy tensor of equation ( [ tensor ] ) ,
wec entails 0 + p_r+2q 0 + p_0 . starting from the einstein equations , the three previous inequalities can be written as follows : f_1(t)+g_1(r ) 0 ( t , r ) f_2(t)+g_2(r ) + 2q(r , t)0 ( t , r ) f_3(t)+g_3(r ) 0 ( t , r ) where the functions @xmath51 and @xmath52 are as in table [ tabla1 ] .
it is a trivial exercise to write all radial functions in terms of the @xmath53-variable by means of the definition @xmath54 .
after this is done , it is seen that these functions display an explicit dependence on the @xmath13 parameter and on @xmath55 , through @xmath35 .
the radius of the throat will be fixed when @xmath13 and @xmath55 are simultaneously known .
these radial functions are plotted in figs .
1,2 and 3 for different values of @xmath13 and for @xmath56 .
note that , at or near the throat , some of them are negative .
this just reflects the fact that no static wormholes can be built , within gr , without using exotic matter .
far enough from the throat , all functions tend to zero , being strong the dependence on @xmath13 of the shape of the curves . also from the figures , we may note that if we demand that the time - dependent functions @xmath57 and @xmath58 be greater than the modulus of the minimum of each of the corresponding radial functions , all wec inequalities will be fulfilled . in the case of the second inequality ,
the term coming from the energy flux must be taken into account to determine the non - exotic region . in table
[ tabla2 ] , we provide several examples of possible solutions for @xmath7 .
the procedure to obtain these particular functions as solutions for the metric is , as stated , to define a suitable functional form for the temporal part of the trace of the stress - energy tensor for matter , i.e. @xmath45 .
this values are also given together with the values of the temporal functions @xmath51 for each choice .
this entries would allow comparison with wec inequalities in order to search for constraints on the values of the constants involved in each choice of @xmath7 , to be fulfilled in order to obtain non - exotic matter at the throat of the evolving wormhole . finally , the domain of @xmath5 for which @xmath43 is finite and without zeros is also given in table [ tabla2 ] .
the comments made by kar @xcite when he considered different @xmath10 functions in his scheme are totally applicable here .
for instance , we also have exponentially expanding or contracting phases -for the first two solutions- and a -like universe ( with a bang and a crunch ) for @xmath59 .
in our scheme , the particular choice of @xmath7 will determine the energy flux , through the field equation ( [ flux ] ) .
the sign of the energy flux ( which depends on @xmath60 ) will decide in turn if the is attractive or repulsive , in the sense explained in @xcite . for large enough and positive @xmath18
, the second wec inequality will be fulfilled easily .
morover , this would imply a quick process of expansion which in turn favors a possible trip by diminishing the tidal forces , as we shall show in the next section .
we shall deal now with the possibility that an explorer might enter into the kind of evolving studied here , pass through the tunnel , and exit into an external space again without being crushed during the journey .
we are not going to discuss here the stability of the wormhole , being this beyond the scope of this paper .
instead , we content ourselves with an analysis of the tidal gravitational forces that an infalling radial observer must bear during the trip .
the extension to a non - radial motion follows from the recipe given in @xcite ( chap .
we recall the desired traversability criteria concerning the tidal forces that an observer should feel during the journey : they must not exceed the ones due to earth gravity @xcite .
from now on , the algebra will be simplified by switching to an orthonormal reference frame in which @xmath61 . in terms of this basis , the proper reference frame is given by , @xmath62 where @xmath63 and @xmath64 are the usual parameters of the lorentz transformation formula .
thus , stated mathematically the traversability criteria read , @xmath65 @xmath66 @xmath67 + \beta\,\frac { \alpha\ , \dot{\omega } e^{\alpha / r}}{r^2\,\omega\,e^{\lambda}}\right\}\right| \leq \frac{g_{_{\oplus}}}{2m}. \label{t2}\ ] ] note that , contrary to the static case , as the geometry is changing with time , the tidal forces that the hypothetical observer feels also depend on time .
the first of these inequalities could be interpreted as a constraint on the gradient of the redshift , while the second acts as a constraint on the speed of the spacecraft .
it is important to stress that we have here a term associated with the energy flux , the last one of the left hand side in equation ( [ t2 ] ) .
let us take as an example the case of an expanding wormhole , with @xmath68 and domain for @xmath5 in the interval @xmath69 .
now , equations ( [ t1 ] ) and ( [ t2 ] ) can be rewritten in the following compact form , @xmath70 @xmath71 + 2 \,\beta\,\frac{\alpha \,\omega\ , e^{\alpha / r}}{r\,e^{\lambda}}\right\}\right| \leq \frac{e^{2\omega t}}{(10^{10 } { \rm cm})^2}.\ ] ] it is seen from these equations that in this expanding `` wormhole universe '' the tidal forces felt by an observer at fixed @xmath26 diminish with time .
the traversability criteria will be then satisfied for any value of @xmath72 if @xmath5 is big enough .
on the other hand , when the wormhole is contracting ( _ i.e. _ @xmath73 ) , the traversability criteria constrain the possible values of @xmath13 , @xmath64 , and @xmath74 more and more as the geometry evolves , eventually rendering the trip impossible .
the behaviour of the tidal forces is then intimately related to the conformal factor @xmath10 , and one would generically expect that if there is an expansion ( contraction ) , the tidal forces will diminish ( grow ) with @xmath5 .
we presented the complete analytical solution for a restricted class of evolving geometries by imposing a suitable form of the trace of the stress - energy tensor for matter .
this , in turn , depends on the particular election of the conformal factor . with this solution
we have analyzed different scenarios of wec violation , extending in this way previous works by kar and kar and sahdev @xcite , in the sense that the redshift function in our case is not zero .
it is worth noting that we have introduced a nonzero energy flux in the stress - energy tensor , which also renders our non - static solution more general than the ones presented before .
we should also remark that one of the keys for the existence of these solutions is their non - flat asymptotic behaviour . in an asymptotically flat space - time , the existence of traversable solutions in gr is forbidden by the theorem of topological censorship @xcite .
we have also studied some traversability criteria concerning tidal forces that a traveler might feel .
we found out that , since the geometry itself is evolving in time , there could be in some cases an expansion of the throat of the which may favor the diminishing of destructive forces upon the trip .
this work has been partially supported by conicet and unlp .
held partial support from fomec .
one of us ( spb ) would like to thank the international center for theoretical physics in trieste for hospitality . | we present here analytical solutions of general relativity that describe evolving wormholes with a non - constant redshift function .
we show that the matter that threads these wormholes is not necessarily exotic .
finally , we investigate some issues concerning wec violation and human traversability in these time - dependent geometries . _
pacs number(s ) : _ 04.20.-q , 04.20.jb |
the field of quantum information theory has grown considerably in the last two decades , achieving important results on the characterization of quantum correlations and on their applicability to quantum information processing @xcite .
quantum protocols consist in two or more agents sharing quantum correlated states , which can help them to perform computational and technological tasks .
the vast majority of the work in quantum information theory has been developed in the setting of distinguishable particles where each agent owns a given number of particles .
however , quantum mechanics predicts that particles behave differently when they are identical @xcite .
the key point is that identical particles can not be individually addressed : an agent can not distinguish which particle he is manipulating .
experimental realizations of quantum information processes consist of identical particles , e.g. photons , atoms in optical lattices , or electrons in solid state systems .
for instance , ultracold atomic gases @xcite can be controlled with very high precision , and therefore are a promising arena for the study of many - body physics and applications in quantum information .
identical particles can be distinguished if appropriate degrees of freedom , e.g. spatial confinements , unambiguously characterize each particle .
the properties in terms of these degrees of freedom , e.g. their individual positions , are employed to label each particle @xcite , without being further manipulated .
based on this remark , many proposals suggest to encode states of distinguishable particles into systems of identical particles @xcite , in order to implement quantum information processing . nevertheless , if one aims to build an integrated architecture , it is difficult to effectively distinguish identical particles . in particular , it can be required to exploit all the accessible degrees of freedom , in order to encode the desired protocol , with no way to label or distinguish particles . for these reasons , the peculiarities of identical particles should be properly considered for a good definition of quantum correlations and for practical purposes of quantum information processing . in the present paper ,
we focus on the analysis of a specific quantum protocol , the generalization of teleportation @xcite with massive , identical particles , and on the use of entangled states to improve achievable teleportation performances .
entanglement , which is considered a key resource for teleportation , was debated for identical particles under several assumptions @xcite .
we apply the approach developed in @xcite , where entanglement is defined via non - classical correlations between subsets of observables .
this is a very general and powerful approach that recovers the standard definition of entanglement for distinguishable particles , while for systems of identical particles it accounts for quantum correlations between occupations of orthogonal modes in the fock space .
applying this framework to quantum information processing , each agent owns and locally manipulates modes , such as wells in optical lattices or hyperfine levels of molecular systems , which can be experimentally addressed @xcite .
a key feature of massive particles , like atoms and constituents of condensed matter systems , is the conservation of the number of particles , mathematically described by a superselection rule @xcite .
this feature is responsible for a very different behaviour of entanglement among identical particles as compared to the case of photons and distinguishable particles , manifest , for instance , in a simpler detectability of entanglement @xcite , a high robustness against noise @xcite , and a different geometry of entangled states @xcite .
the properties of entanglement in its presently used definition have been studied in fermionic superconducting systems @xcite , electrons in low - dimentional semiconductors @xcite , and bosonic ultracold gases @xcite , and exploited in several applications , such as quantum data hiding @xcite , teleportation @xcite , bell s inequalities @xcite , dense coding @xcite , and quantum metrology @xcite .
we shall discuss teleportation in detail , which plays a fundamental role in quantum computation @xcite .
the present analysis unifies and largely generalizes all the previous proposals of teleportation of identical particles , in the light of properties studied in @xcite . in the original teleportation protocol
@xcite one agent , alice , wants to teleport an arbitrary , perhaps unknown , state to another agent , bob .
alice owns the state to be teleported and a share of a resource state , and bob owns the remaining part of the resource state .
the algorithm of the standard teleportation is the following : _
i ) _ alice performs a projective measurement onto the basis of maximally entangled states of her states ; _ ii ) _ alice sends bob the result of the measurement ; _ iii ) _ bob performs a suitable operation on his state , conditioned on the message he got from alice . in the setting of distinguishable particles , if the shared state is a pure , maximally entangled state , bob ends up with a state identical to the initial state to be teleported , while the state on alice s side has been transformed by the measurement .
otherwise , alice and bob can distil maximally entangled states from a larger number of not maximally entangled states , in order to perform an optimal teleportation @xcite .
the teleportation can also be applied to a part of an entangled state . in this case ,
alice initially owns a subsystem entangled with another subsystem in possession of a third party , and uncorrelated with the resource state and bob s subsystem .
after the teleportation , alice s subsystem is entangled neither with bob s nor with the third party subsystem , while the subsystems respectively owned by bob and the third party become entangled .
this application is called entanglement swapping , and can be useful for sharing entanglement at long distances , required for quantum networks @xcite .
this paper is organized as follows : in section [ entanglement ] , we give the basic definition of entanglement and review some of its properties . in section [ twomode ]
, we introduce the teleportation protocol , where one mode of a two - mode state is teleported using a two - mode resource state .
the conservation of the total number of particles for the independent operations performed during the teleportation protocol reduces the performance .
this results in the impossibility of a perfect teleportation for any finite number of particles in the resource state . in section [ perf ] ,
we analyse the teleportation efficiency with some exemplary resource states , and show that the efficiency grows with the number of particles in the resource state .
these resource states include the maximally entangled states , but also states that can be more easily prepared , such as su(2 ) coherent states @xcite , and the ground state of a double well potential with intrawell interactions @xcite .
additional details are shown in appendix [ perf.app ] .
we comment on the more cumbersome case of many - mode states to be teleported , as compared to two - mode states , in section [ multimode ] . in section
[ reference ] , we prove the impossibility of perfect teleportation with a finite number of particles , for any teleportation protocol . as a byproduct
, we prove that it is not possible to recover perfect teleportation even considering the usual way to overcome the particle number superselection rule , i.e. to couple the system to a reservoir @xcite .
finally , we summarize our conclusions in section [ discussions ] .
we use a mathematical style ( * definition * , * proposition * ) , in order to emphasize a few key statements .
we now introduce basic definitions and notation which we will rely on in the rest of this paper .
specialist readers already fluent in the algebraic characterization of the entanglement of indistinguishable particles may skip this section and directly proceed to section [ twomode ] .
we start with an algebraic framework which generalizes and recovers both entanglement of distinguishable particles and entanglement of identical particles @xcite .
let us consider a many - body system described by the hilbert space @xmath0 .
the algebra of all bounded operators , including all the observables , is denoted by @xmath1 is generated by differentiation of the so - called weyl operators .
any function of the creation and annihilation operators is obtained by a proper differentiation of the weyl operators @xcite . ] . unlike the usual definition @xcite
, we move the attention from the states and the tensor product induced partitioning of @xmath0 to the observables and local structures of @xmath1 . an _ algebraic bipartition _ of @xmath1
is any pair @xmath2 of commuting subalgebras , @xmath3 .
any element of @xmath4 commutes with any element of @xmath5 , @xmath6=0 $ ] .
the notion of locality lies in the commutativity of the subalgebras , ensuring the compatibility of observables of @xmath4 with observables of @xmath5 .
an operator is said to be _ local _ with respect to the bipartition @xmath2 , if it is the product @xmath7 of an operator @xmath8 in @xmath4 and some @xmath9 in @xmath5 . with the above definitions
, we can now state the definition of separable and entangled states .
[ ent.state ] a state @xmath10 is said to be _
separable _ with respect to the bipartition @xmath2 if the expectation of any local operator @xmath7 can be decomposed into a convex combination of products of local expectations : @xmath11 with @xmath12 and @xmath13 admissible states of the system .
otherwise , the state is _
entangled_. let us now focus on many - body systems whose constituents are @xmath14 bosons which fill @xmath15 different modes .
the formalism of second quantization is more convenient for such systems .
let us introduce creation and annihilation operators @xmath16 , @xmath17 , for each mode , with the bosonic commutation relations , @xmath18=\delta_{jl}$ ] .
the total hilbert space @xmath19 of the system is spanned by the fock states @xmath20 is the occupation number of the @xmath21-th mode such that @xmath22 .
we use tensor product notation to write fock states : for instance , @xmath23 .
we have to keep in mind that this tensor product structure is constrained by the conservation of the number of particles .
the norm - closure of the set of polynomials in all creation and annihilation operators , @xmath24 , @xmath25 , is the algebra @xmath26 .
we define bipartitions of this algebra by splitting the set of creation and annihilation operators into two disjoint sets @xmath27 and @xmath28 with arbitrary @xmath29 .
the norm - closure of all polynomials in the creation and annihilation operators of the first ( second ) set is the subalgebra @xmath4 ( @xmath5 ) .
according to the previous definition [ ent.state ] , a pure state is @xmath30-separable if and only if @xmath31 where @xmath32 and @xmath33 are arbitrary functions .
mixed @xmath30-separable states are convex combinations of pure @xmath30-separable states .
see @xcite for a detailed analysis .
since the total number of particles is conserved , all the observables and the density matrices commute with the total number operator . in other words
, there is a superselection rule @xcite which forbids any coherent superpositions of states with different total numbers of particles .
this is a key feature of such systems , and at the very heart of some recent results @xcite .
any orthonormal basis of @xmath30-separable states can be relabelled as @xmath34 the integer @xmath35 in eq .
counts the number of particles in the first @xmath29 modes , while @xmath36 labels the different ways in which @xmath35 particles can fill those modes .
similarly , @xmath37 labels the ways in which the remaining @xmath38 particles can occupy the other @xmath39 modes .
any pure @xmath30-entangled state is a coherent superposition of at least two @xmath30-separable states .
there are two qualitatively different ways to superimpose separable states .
the first is the superposition of states labeled by different @xmath40 , keeping @xmath35 fixed .
physically , the local numbers of particles are fixed in such states , thus the mode - bipartition corresponds to a particle - bipartition . a state which exhibits only these coherences
has a block - diagonal structure in the label @xmath35 : @xmath41 this class of block - diagonal states includes all the states which are separable and positive under partial transposition ( ppt ) @xcite .
the set of states in each diagonal block , i.e. for each fixed value of @xmath35 , is defined on the span of @xmath42 .
this is a hilbert subspace of dimension @xmath43 , isomorphic to the unconstrained tensor product space @xmath44 .
therefore , the mathematical features , and thus their physical consequences , are qualitatively analogous to those of distinguishable particles @xcite .
the sum of the single block dimensions @xcite is the dimension of the total hilbert space @xmath45 because of the above isomorphism , such states are natural candidates for realizing quantum protocols developed for distinguishable particles with identical bosons @xcite .
the second kind of superposition is the superposition of states labeled by different @xmath35 , which is not compatible with a particle - bipartition .
these states are not block - diagonal and , thus , not ppt @xcite .
the entanglement of such states is more powerful and robust than entanglement of distinguishable particles , because separable states in our framework are condensed to a zero measure subset .
in fact , this kind of entanglement can not be washed out by any mixtures with separable states , and mixtures with entangled states can erase such entanglement only under very special conditions @xcite . on the other hand
, local noisy dynamics can not completely dissipate entanglement at finite times @xcite , contrary to what happens to distinguishable particles @xcite .
all these phenomena reflect the fact that the set of separable states has zero measure . besides its definition ,
entanglement of identical particles can also be quantified with so - called entanglement measures .
the theory of entanglement measures is based on the same principles of that for distinguishable particles @xcite , provided that any physical state commutes with the total number of particles in our framework .
some entanglement measures for identical particles have been discussed in @xcite : negativity , robustness of entanglement and entanglement entropy . in accordance with these entanglement measures
, we call a pure bipartite state of @xmath14 particles maximally entangled if its reduced density matrices are maximally mixed within the subspace with no more than @xmath14 particles . in the rest of this paper
, we use negativity as entanglement measure . in the following ,
we employ entangled states of identical particles , as defined above , as a resource for quantum teleportation .
we start with discussing a teleportation protocol that uses two - mode states .
the single block dimensions of two - mode states are @xmath46 .
thus , the only contribution to entanglement is due to coherent superpositions among different local particle numbers @xmath35 .
we disuss the teleportation of one mode of a two - mode state @xmath47 with the help of a two - mode , shared resource state @xmath48 . the labels @xmath49 number the modes .
the initial global state is @xmath50 , where latexmath:[\[\label{initial.state } and @xmath48 is a general two - mode state with @xmath52 particles @xmath53 the second and the third modes are owned by the sender , alice .
she aims to teleport the state of the second mode to the target mode , the forth one , owned by the receiver , bob . it does not matter whether the first mode is in possession of alice or not . first
, if @xmath54 alice performs a projective measurement of her modes , that involves projectors @xmath55 , onto the following orthogonal states @xmath56 ( where the phases @xmath57 ensure the orthogonality ) .
we assess the entanglement and completeness of the basis for a comparison with teleportation protocols for distinguishable particles , because there a complete projective measurement onto maximally entangled states is required for perfect teleportation @xcite .
the states have the same amount of entanglement as the maximally entangled states of @xmath14 two - mode particles .
moreover , they span a @xmath58-dimensional hilbert space .
nevertheless , since the second ( third ) mode can be filled at most by @xmath14 ( @xmath52 ) particles , all the possible pure states of these modes span an @xmath59-dimensional hilbert space . in other words ,
the conservation of the total number of particles applies to each state that is prepared independently , i.e. @xmath47 and @xmath48 , and not to the reduced state of the second and the third mode which have interacted with the other modes . on the other hand
, alice performs the measurement independently from the first and the fourth mode , thus the projections on the states ( [ states.meas ] ) must commute with the total number of particles .
therefore , the states ( [ states.meas ] ) do not define a complete measurement on the second and the third mode .
additional projectors onto the missing subspace should be considered , in order to make the measurement without post - selection trace preserving .
however , the additional projectors can not project onto states with the same amount of entanglement .
we will consider the complete set of orthogonal projectors @xmath55 onto the following basis @xmath60 where we have defined the function @xmath61 which equals the cardinality of the sum in equation .
the projection onto the states take into account also the cases when @xmath62 .
the records of alice s measurement are @xmath63 which label the outcome corresponding to the projection onto each basis state .
if alice records @xmath63 , the state changes into @xmath64 where @xmath65 is the identity operator on the @xmath21-th mode , and @xmath66 are the coefficients of the initial state . at this point the state of bob s mode depends on the outcome of alice s measurement . in order to reconstruct the state
to be teleported , alice sends bob the outcome @xmath63 of her measurement , via a classical channel , and bob subsequently applies the operation @xmath67 to the fourth mode , where @xmath68 since the conservation of the total number of particles is central in our analysis , we now show that the operations @xmath69 are fully consistent with this conservation law .
indeed , @xmath69 can be implemented by a unitary operation which preserves the total particle number on the forth mode and on an additional fifth mode : @xmath70 where @xmath71 is the trace over the fifth mode , and @xmath72 @xmath73 is the identity matrix on the subspace orthogonal to the support of each @xmath74 .
thus , @xmath75 is a unitary transformation which commutes with the total number of particles .
now , we come to the analysis of the final state and how similar to the original state to be teleported it is .
since bob knows the outcome @xmath63 on alice s side , and accordingly transforms his state , he ends up with the mode @xmath76 of one of the states @xmath77 with probability @xmath78 , as given by the partial trace over the second and the third mode @xmath79= \nonumber \\ & = & \sum_{k , j=\max\{0,-l\}}^{\min\{n,\nu - l\}}\left(\rho_{34}\right)_{k+l , j+l}\frac{c_k\bar c_j}{{\cal c}_l}|k\rangle_1{\,}_1\langle j|\otimes|n - k\rangle_4{\,}_4\langle n - j|,\end{aligned}\ ] ] with @xmath80 .
the average of the teleported state over all possible outcomes on alice s side , which we will use to discuss the performances of the above protocol , is given by the following local operations with classical communication ( locc ) : @xmath81=\sum_{l =- n}^\nu\sum_{\lambda=0}^{{\cal c}_l-1}p_{(l,\lambda)}\rho_{14}^{(l,\lambda)}=\sum_{l =- n}^{\nu}\sum_{k , j=\max\{0,-l\}}^{\min\{n,\nu - l\ } } c_k\bar c_j\left(\rho_{34}\right)_{k+l , j+l}|k\rangle_1{\,}_1\langle j|\otimes|n - k\rangle_4{\,}_4\langle n - j|.\ ] ] we stress that the state ( [ teleported ] ) is not bob s final state , but the average over all possible final states at bob s end .
following @xcite , the faithfulness of the teleportation is quantified by the fidelity @xmath82|\psi\rangle,\ ] ] where @xmath83 is the uniform distribution over all pure states , and the states @xmath84 and the average teleported state @xmath85 $ ] stem from the same hilbert space .
the teleportation fidelity is the average overlap between the final state of the first and fourth mode and the initial state of the first and second mode , thus measures how similar these states are . in order to define the uniform distribution ,
consider an arbitrary state @xmath86 with @xmath87 , in a hilbert space spanned by the orthonormal basis @xmath88 .
re - writing the coefficients @xmath89 , with @xmath90 and @xmath91 , the uniform distribution over the pure states is induced by the haar measure of the group of the unitary transformations : can be considered .
for instance , a flat distribution in the variables @xmath92 and @xmath93 has been studied in @xcite .
the only differences are some numerical factors , while the scaling with @xmath14 does not change .
we focus on the measure ( [ measure ] ) for comparison with the results in @xcite . ]
@xmath94 see e.g. @xcite .
for instance , the average values of @xmath95 and @xmath96 with @xmath97 are @xmath98 where @xmath99 is the euler s gamma function @xcite and @xmath100 .
inserting equation for the average teleported state into the definition of the fidelity , and using the averages and , we derive the following expression @xmath101 where the function @xmath102 comes from counting the number of times each term @xmath103 appears within the triple sum inherited by equation . , with @xmath104 and @xmath105 , appears in the triple sum of equation @xmath106 times if @xmath107 , @xmath14 times if @xmath108 , @xmath109 times if @xmath110 , ... only once if @xmath111 , and never if @xmath112 . in order to see this ,
look at the sum over @xmath113 as a squared @xmath114 shadow shifting along the diagonal blocks of @xmath48 .
the matrix elements of @xmath48 covered at each sweep of the shadow are @xmath115 for @xmath35 and @xmath21 contributing to the sum of . ] if the agents are not interested in teleporting the state itself , but in sharing as much entanglement as possible between the first and the fourth mode , then a more relevant figure of merit is the measure of entanglement of the final state between the first and the fourth mode . in general , the final state is mixed , and the negativity @xcite is the most easily computable measure of entanglement for mixed states : @xmath116 where @xmath117 denotes partial transposition @xcite . while in general it vanishes for some entangled states , the negativity of a two - mode state with a fixed number of particles is a faithful measure of entanglement @xcite .
given a two - mode state @xmath10 , its negativity is @xmath118 we consider the double average of the negativity over the outcomes @xmath63 and the uniform distribution of the initial state as a quantifier of the final entanglement : @xmath119 where the function @xmath102 has the same origin as in equation .
equation ranges between zero and @xmath120 .
the upper bound @xmath121 is proven by noting that the teleportation protocol does not act on the first mode and can not increase the entanglement between the first mode and the rest .
thus , the average final entanglement @xmath122 is not larger than the average entanglement over all pure initial states , namely @xmath123 .
the entanglement of each final state @xmath77 is independent of whether the local operation @xmath69 has been performed .
we notice that @xmath124 follows from the triangle inequality for the absolute value applied to , and from the positivity of the fidelity @xmath125 .
we will use this inequality to translate properties of the fidelity to properties of the average teleported entanglement .
let us briefly comment on the case of a non - entangled state @xmath47 instead of the state .
this case is not of practical interest for two reasons .
first , it does not allow for the protocol of entanglement swapping .
second , the perfect teleportation of the second mode of such a state is possible without any entangled resource state .
indeed , the non - entangled two - mode states are only the fock states in the choosen basis of modes @xcite .
in particular , the states of the second mode are the fock states @xmath126 , with @xmath127 . in order to broadcast the information of this state ,
it is enough that alice measures the number operator @xmath128 , which provides the full information on @xmath129 , and communicates the result to bob .
then , bob can prepare locally the state @xmath130 .
it is however instructive to look at the fidelity of the teleportation protocol in this case .
equation ( [ fidelity ] ) with @xmath127 tells us that the state @xmath126 can not be teleported if @xmath131 or @xmath132 .
thus , the teleportation protocol - fails to teleport any state of the fock basis of the second mode , and the linearity of the protocol implies the following property : [ prop1 ] the above teleportation protocol - can not perfectly teleport an arbitrarily entangled state @xmath47 .
the teleportation fidelity is @xmath133 .
one can wonder whether there is a better protocol which achieves @xmath134 .
the answer is negative , as we prove below , even considering the most general class of teleportation protocols . before discussing generalizations of the above teleportation
, we analyse the performances of the protocol presented in this section with some exemplary resource states .
in this section , we discuss the teleportation performances quantified by the fidelity and the average final entanglement of several interesting resource states , i.e. maximally entangled states , n00n states , su(2 ) coherent states , and the ground state of a double well potential , as regards their relevance in other physical processes .
additional details are exposed in appendix [ perf.app ] .
we identify the useful states for teleportation , namely those states that outperform the separable states when employed for the teleportation protocol .
the additional contribution to the fidelity , with respect to that of separable states , can be negative , unlike the contribution to the average final entanglement
. moreover , not all of the off - diagonal entries are relevant for teleportation .
indeed , the entries @xmath135 with @xmath112 do not enter in the formula of the fidelity ( [ fidelity ] ) and of the average final entanglement ( [ av.ent ] ) .
as mentioned above , separable resource states , namely @xmath136 provide threshold performances that define useful resource states .
the teleportation fidelity and the average entanglement are the same for all resource states of this type : @xmath137 it is worth mentioning that @xmath138 equals the highest teleportation fidelity achievable with resource states of distinguishable @xmath139-level systems , which are separable or positive under partial transposition @xcite .
this latter observation is consistent since resource states with a fixed number of particles are special cases of unconstrained states , e.g. photonic modes , which in turn are mathematically equivalent to states of distinguishable @xmath139-level systems .
let us now consider the pure , maximally entangled resource state @xmath140 of @xmath52 two - mode particles @xcite , where @xmath141 , the state of the second mode is perfectly teleported to the fourth mode .
these outcomes occur with probability @xmath142 , see appendix [ perf.app ] .
the direct computation of the teleportation fidelity ( [ fidelity ] ) with elementary summations leads to @xmath143 which is always larger than the fidelity @xmath138 of separable resources , and is arbitrary close to one if @xmath144 . under this condition ,
the probability of the outcome @xmath63 with @xmath145 or @xmath146 is negligible , and the initial state is almost perfectly teleported .
the average entanglement of the teleported state is straightforwardly computed : @xmath147 which is smaller than the average entanglement @xmath120 over all pure states , and converges to this value for @xmath144 .
exemplary entangled resource states which are _ not _ useful for teleportation are the so - called n00n sates with @xmath52 particles : @xmath148 for the case @xmath149 , which slightly outperforms separable resource states without qualitative improvement .
these features of n00n states are remarkable , because they are the most useful states in the high accuracy estimation of the relative phase between the two arms of an atomic interferometer @xcite .
the apparent imbalance of their performance for different purposes can be explained because coherence among all fock states , thus strong mode - entanglement , is needed to implement accurate teleportation , whereas only coherence between two fock states with highly unbalanced population in two modes , thus weak mode - entanglement , is required for precise phase estimation .
indeed , accurate phase estimation is achieved with states that sensitively vary under phase shifts , and corresponds to a heisenberg - like relation where the larger the variance of relative occupation number , the smaller the accuracy of the phase .
moreover , the counterpart of n00n states in first quantization , i.e. the so - called ghz states , are used to perfectly teleport states of distinguishable particles in low dimensional hilbert spaces @xcite .
states studied in the context of mean field approximation and mesoscopic quantum coherent phenomena @xcite are the so - called su(2 ) coherent states @xcite , also known as atomic coherent states @xcite .
these are states where all particles occupy the same combination of two modes defined by the population probabilities of both modes , @xmath150 and @xmath151 , and the relative phase @xmath152 : @xmath153 we numerically compute the fidelity @xmath154 and the average final entanglement @xmath155 of the teleportation protocol when the resource state is a symmetric coherent state @xmath156 , i.e. with balanced population probability @xmath157 , that is routinely prepared in the laboratory @xcite . we plot teleportation performances and compare them with the performances of the maximally entangled resource state , with those of separable resource states , and with their maximum values in figures [ fid.sym.coh ] and [ ent.sym.coh ] .
teleportation performances are always better than performances of separable states , and can be very close to their maximum values for large @xmath52 .
in such limit , the binomial distribution in the definition of the coherent state becomes more and more flat , and approaches the uniform distribution that characterizes the superposition of the maximally entangled state .
[ htbp ] [ htbp ] we also plot the teleportation fidelity @xmath158 and the average final entanglement @xmath159 which can be generated with the help of a general with a general su(2 ) coherent state , with fixed @xmath160 and @xmath161 in figure [ perf.coh ] .
maximum fidelity is produced with the symmetric for the symmetric coherent state @xmath162 , and maximum average final entanglement is achieved for the symmetric coherent state @xmath163 for any @xmath152 .
the average final entanglement does not depend on the phase @xmath152 , since only the modulus of the off - diagonal elements of the density matrix enters in its definition .
recall that perfect teleportation with identical particles , meaning fidelity @xmath134 , is impossible , while for distinguishable particles it is attained with maximally entangled resource states . in this context , it is remarkable that some coherent states can attain almost perfect teleportation , and that they can on average even outperform maximally entangled resource states , in some parameter regimes .
indeed , coherent states are easy to prepare in experiments , see e.g. @xcite , and @xcite for a theoretical proposal of dissipative preparation . in order to stress the application of the notion of entanglement discussed in this paper ,
notice that su(2 ) coherent states are entangled and they are useful resources for teleportation .
in this context , su(2 ) coherent states never outperform classical metrology for phase estimation @xcite .
thus , teleportation with identical particles shows the effects and the usefulness of mode - entanglement , even for states where phase estimation does not . in this section
, we discuss the ground state of the two - mode bose - hubbard hamiltonian @xcite as a resource for the teleportation protocol .
this application is potentially relevant because this state can be prepared with present days technology , such as magnetic traps and evaporative cooling @xcite .
the bose - hubbard hamilonian @xmath164 is fixed by the tunneling amplitude @xmath165 between the two well sites and the on - site interaction strength @xmath166 .
let us consider this ground state as the resource state of the teleportation protocol @xmath167 , where @xmath168 1 . if @xmath169 or @xmath170 , the tunneling term can be treated as a perturbation , and the ground state emerges as a superposition of few fock states . with a similar analysis as compared to that in section [ n00n.states ]
, we can argue that the resulting state has poor entantanglement and is not very useful for teleportation .
2 . if @xmath171 , then the ground state is a gaussian superposition + @xmath172 + with the normalization + @xmath173 + for large @xmath52 .
this range of @xmath174 shall be called _ single gaussian regime_. 3 . in a different regime ,
characterized by @xmath175 , the ground state is a superposition of two gaussians @xcite : + @xmath176 + with the normalization + @xmath177 + for large @xmath52 .
we shall refer to this instance as _
double gaussian regime_. 4 .
the transition range between the single and the double gaussian regimes is the _ critical regime _ @xmath178 . in the critical regime ,
the ground state is a superposition peaked around @xmath179 , but less confined than a gaussian since it starts to split into two imbalanced occupations of the sites .
[ htbp ] considering the ground state of the bose - hubbard model as a resource for the teleportation protocol , we numerically compute the fidelity @xmath180 , equation , and the average final entanglement @xmath181 , equation .
we plot teleportation performances and compare them with the performances of the maximally entangled resource state , with those of separable resource states , and with their maximum values in figures [ fid.gauss ] and [ ent.gauss ] .
we choose @xmath182 because we do not observe any qualitative difference within the single and the double gaussian regimes .
teleportation performances are always better than those of separable resource states , and can be very close to their maximum values for large @xmath52 because , as for the symmetric coherent state , the wave function of the ground state spreads and approaches the uniform distribution realized by the superposition in the maximally entangled state . in figures [ fid.bh ] and [ ent.bh ] , we plot teleportation performances as functions of @xmath174 , and comparisons with those of the maximally entangled resource state .
we notice that there is a region in both regimes where the fidelity @xmath180 and the average final entanglement @xmath181 are larger than the corresponding quantities of the maximally entangled state .
when @xmath14 grows , this region shrinks around the boundary with the critical regime @xmath183 .
thus , the single gaussian and the double gaussian approximations may be no longer reliable and these numerical results may not necessarily match the exact behaviour of the ground state of ( [ bh ] ) .
neverthless , ( [ gauss ] ) and ( [ 2gauss ] ) are _ bona fide _ states which outperform the average performances of the maximally entangled states , even if they do not coincide with the ground state of the double well potential .
[ htbp ] [ htbp ] this concludes our analysis of the teleportation protocol described in section [ twomode ] . in the next two sections we discuss generalizations of this protocol and the corresponding performances
in this section we generalize the teleportation protocol by the use of many - mode states . as an introductory scenario ,
let us replace the first mode with a set of @xmath29 modes .
the initial state is @xmath184 , and @xmath36 is an additional index that distinguishes different orthogonal occupations of the first @xmath29 modes with @xmath35 particles .
since the goal is to teleport the single mode labelled by 2 , we apply a two - mode resource state with @xmath52 particles ( [ res ] ) , and the teleportation protocol discussed in section [ twomode ] , which does not act on the first @xmath29 modes .
the average teleported state and the average fidelity are , respectively , @xmath185=\sum_{l =- n}^\nu\sum_{k , j=\max\{0,-l\}}^{\min\{n,\nu - l\}}\sum_{\sigma=1}^{d_k^{(m)}}\sum_{\tau=1}^{d_j^{(m)}}(\rho_{34})_{k+l , j+l } \ ,
c_{k\sigma}\bar c_{j\tau}|k,\sigma\rangle_1{\,}_1\langle j,\tau|\otimes|n - k\rangle_4{\,}_4\langle n - j| , \\ \label{fidelity.modes } & & f_m=\frac{1}{d(d+1)}\sum_{l =- n}^\nu\left(\sum_{k , j=\max\{0,-l\}}^{\min\{n,\nu - l\}}d_k^{(m)}d_j^{(m)}(\rho_{34})_{k+l , j+l}+\sum_{k=\max\{0,-l\}}^{\min\{n,\nu - l\}}d_k^{(m)}(\rho_{34})_{k+l , k+l}\right).\end{aligned}\ ] ] the fidelity @xmath186 is straightforwardly computed insering the initial state and the average teleported state into the general equation , and using averages of the probability distribution with @xmath106 replaced by @xmath187 , being the hilbert space dimension in the present case . in figure [ fid.modes ] , we plot the fidelity @xmath188 for the maximally entangled resource state ( [ max.ent.state ] ) , and its difference with respect to the fidelity @xmath189 of the same protocol with @xmath190 discussed above .
we consider the special case of @xmath160 particles in the state to be teleported , and notice that the fidelity increases with @xmath29 .
this is an effect of the dimensionality of the subsystem @xmath191 .
thus , as soon as @xmath192 , the fidelity is always larger than the fidelity with @xmath190 . furthermore , alice measures the pairs @xmath63 which allow a perfect teleportation with probability @xmath142 . the fidelity for the maximally entangled resource states
is bounded from below by such probability .
figure [ fid.modes ] shows that this bound is not saturated in general , since the fidelity @xmath188 can be larger than @xmath189 that is in turn larger than the probability of a perfect teleportation .
moreover , the fidelity increases with the number @xmath52 of particles in the resource state .
this can be explained with the inceasing amount of coherence and entanglement , which reduces more and more the probability of imperfect teleportation , as happens for two - mode initial states .
let us now move to a more general setting .
alice wants to teleport a set @xmath193 of modes , which are entangled with another set @xmath194 of modes in the initial state @xmath195 .
the shared resource state is @xmath196 , where @xmath197 labels the set of modes owned by alice , and @xmath198 labels the set of modes owned by bob .
alice can teleport each mode of @xmath193 one at a time , applying the protocol described in section [ twomode ] .
high fidelity is achieved if the two agents share a maximally entangled two - mode state with @xmath52 particles for each mode in the set @xmath193 .
therefore , if there are @xmath29 modes in each of the sets @xmath193 , @xmath197 and @xmath198 , the resource state is @xmath199 , where @xmath200 and @xmath201 label the modes in the set @xmath197 and @xmath198 , respectively .
the fidelity of this protocol is bounded from below by @xmath202 , which is the probability that each mode is perfectly teleported .
figure [ fid.modes ] shows that this is not a sharp bound , since the fidelity of the teleportation of each single mode can be larger than this lower bound , as discussed above .
the lower bound , and thus the fidelity , are arbitrarily close to one if @xmath144 and @xmath29 is finite . [ htbp ] the state ( [ many.two.mode ] ) is not a maximally entangled state between the sets of modes @xmath197 and @xmath198 , because the reduced states are not completely mixed .
it is however an extremely useful resource for the just mentioned procol .
a natural question is whether we can achieve a higher fidelity by teleporting all the modes in @xmath193 in one single run
. such a protocol would be the straightforward generalization of the protocol with two - mode resource states : _
i ) _ alice performs a projective measurement including projectors onto highly entangled states between @xmath193 and @xmath197 , then _ ii ) _ she communicates the outcome to bob , and finally _ iii ) _ bob transforms his modes according to the received outcome .
let us first consider the simpler case of an initial state @xmath195 lying on one diagonal block ( [ block.diag ] ) , i.e. with a fixed local number @xmath35 of particles in the modes @xmath194 and @xmath38 in the modes @xmath193 .
the subspace of these states is isomorphic to the unconstrained tensor product space @xmath44 , and thus to the space of two distinguishable systems of dimensions @xmath203 and @xmath204 , respectivelty .
hence , the set of modes @xmath193 can be perfectly teleported by the usual teleportation protocol @xcite for a @xmath204-level system , translated to the formalism of fock space .
to this aim , we need a maximally entangled resource state between two @xmath204-level systems .
this is realized by a pure state @xmath205 , where both @xmath197 and @xmath198 are sets of @xmath39 modes with a fixed number @xmath38 of particles in each of them .
this protocol turns out to be exactly the same as discussed in @xcite , once the local number of particles and the number of modes are set , respectively , to @xmath206 and @xmath207 .
this latter example has been derived within the formalism of first quantization @xcite , and seems a bit intricate because of the permutation invariance of states . on the other hand , in second quantization
we only need to straightforwardly apply the usual teleportation protocol , because the symmetrization is implicitly included in the formalism .
now , we can wonder whether we can teleport all the modes in @xmath193 in one single run even for non - block - diagonal initial states . generalizing the protocol in section [ twomode ]
, alice performs a projective measurement on the modes in @xmath193 and @xmath197 .
the measurement projects onto states that generalise to the case of many - mode subsystems .
some of these states are @xmath14-particle maximally entangled with respect to the bipartition @xmath208 , in analogy to the two - mode states ( [ states.meas ] ) with @xmath209 , and other states have more than @xmath14 particles and the same amount of entanglement than the previous ones , generalising the two - mode states ( [ states.meas ] ) with @xmath210 .
two main difficulties arise in this protocol , which decrease the fidelity and forbid a perfect teleportation .
the first comes from the mismatch between the dimension of the reduced states of @xmath193 and @xmath197 and the number of orthogonal projectors onto states with the same amount of entanglement , as happens for the two - mode states where we completed the measurement with additional projectors ( [ states.meas.2 ] ) .
the second difficulty concerns the non - existence of a complete orthonormal basis of maximally entangled pure states with a fixed total number of identical particles .
such a basis always exists for distinguishable particles and unconstrained hilbert spaces @xcite .
it exists for the fock space of @xmath14 two - mode particles , for the fock space of one @xmath15-mode particle symmetrically halved in two mode - partitions ( @xmath211 ) , and in the special case of @xmath212 particles that fill @xmath213 modes divided into two equal parties ( @xmath214 ) @xcite . on the other hand
, a straightforward computation shows that a complete basis of maximally entangled states does not exist for one particle in @xmath15 modes which are split into two unequal sets ( @xmath215 ) , or for @xmath212 @xmath216-mode particles .
it is not clear whether the existence of such bases is related to the symmetry of the algebraic bipartition .
the impossibility of such a complete orthonormal basis reduces the number of projectors in alice s measurement , that generalise ( [ states.meas ] ) and allow for a perfect teleportation .
we showed that the performance of the above teleportation protocol with two - mode states is limited by the conservation of the total number of particles .
the limitation comes from the impossibility of a complete projective measurement such that each outcome provides perfect teleportation .
this in turn stems from the fact that the second and the third mode are entangled with the other modes , thus their reduced state does not have a fixed number of particles , while alice s projectors do .
one can wonder whether there is a different protocol that provides perfect teleportation .
possible generalizations allow many - mode states , different perhaps non - projective measurements on alice s side , and the use of a general resource state .
a more interesting generalization is to relax the constraint of total particle number conservation .
nevertheless , in this section , we will prove that the teleportation fidelity is never one , for any general teleportation protocol using a finite number of particles . before discussing this statement
, we recall the basic idea of relaxing conservation laws in quantum information protocols . the presence of a conservation law is formulated in terms of a superselection rule , namely the requirement that all the physically addressable states and observables commute with the conserved quantities .
this requirement was connected to the lack of a reference frame , as reviewed in @xcite .
for instance , any operator @xmath217 compatible with the conservation of the total particle number is invariant under the twirling operation @xmath218 , where @xmath219 is the total number operator .
this means that the phase @xmath220 is not observable if the particle number is conserved , and that the physical operators are uniformly averaged over the phase shift @xmath221 . on the contrary , if the different values of @xmath220 can be distinguished with respect to a phase reference , the physical operators are no longer invariant under the twirling operation .
the standard method to bypass the constraints imposed by the superselection rule @xcite involves embedding the original system into a larger one , allowing interactions with a reservoir .
the conservation law applies only to the total system , while the original system is much less constrained .
we then apply the invariance of physical operators under twirling operation with respect to the number of particles @xmath222 of the systems plus that of the reservoir , @xmath223 : @xmath224 .
equivalently , the phase @xmath220 conjugated to the total number of particles of the system plus the reservoir is not observable .
nevertheless , since the relative particle number @xmath225 is not a conserved quantity , the local phase conjugated to @xmath222 can be distinguished by measurements of the reservoir that , thus , provides a quantum phase reference . to give a specific example , the two - mode state @xmath226 and the operator @xmath227 do not satisfy the particle number superselection rule , while the three - mode state @xmath228 and the operator @xmath229 do , and reproduce the same statistics of the previous two - mode case .
a natural question is whether this approach allows to overcome all the restrictions imposed by the superselection rule , such as the impossibility of perfect teleportation shown in proposition [ prop1 ] and in section [ perf ] .
for instance , any unconstrained operations on the original systems can be mimicked by suitable operations on the total system @xcite .
this consideration was used to prove that any unconstrained quantum protocol consisting of local unitary operations can be mimicked , controlling the interactions with local reservoirs @xcite .
this result is not directly applicable to general processes which consist of local operations and classical communication ( locc ) . in the present paper
, we are interested in the possibility of mimicking the unconstrained teleportation , which occurs with fidelity one , by means of interactions with a reservoir .
we explicitely show that the teleportation fidelity is strictly smaller than one , for any teleportation protocol consisting of finitely many particles and of locc on the system and a resource state possibly correlated with a reservoir .
[ prop.imp ] deterministic perfect teleportation , namely with fidelity one , is never possible for a fixed and finite number of identical particles . for the rest of this section
we prove the above proposition .
we start with a special generalized teleportation protocol that captures the salient features of the impossibility of perfect teleportation , and will discuss all possible extensions later .
consider sets @xmath230 with a number @xmath231 of modes and @xmath232 .
using the orthonormal basis , the generalization of the initial state ( [ initial.state ] ) is latexmath:[\[\label{initial.state.reserv } where @xmath194 ( @xmath193 ) is a set of modes that generalizes the first ( second ) mode of the protocol in section [ twomode ] , and we aim to teleport the set of modes @xmath193 .
consider pure resource states of @xmath52 four - mode particles latexmath:[\[\label{res.reserv } @xmath197 is a set of modes owned by alice , while @xmath198 and @xmath235 are the sets of modes owned by bob .
we consider resource states that can be coupled to a shared particle reservoir which is necessary for mimicking a general operation unconstrained by the particle number superselection rule .
the part of the reservoir possessed by alice is included in the set @xmath197 .
considering arbitrarily large sets of modes , there is neither physical reason nor notational convenience to identify the reservoir from the modes in @xmath197 .
it is however convenient to divide the modes possessed by bob into two sets @xmath198 and @xmath235 , such that @xmath198 has the same number of modes as @xmath193 .
thus , @xmath198 is the target of the teleportation , and @xmath235 plays the role only of the reservoir .
alice performs a projective measurement in this extended setting , then communicates the result to bob , and bob performs a suitable operation on his modes . in general , alice projects onto an orthonormal basis of pure states
@xmath236 with @xmath237 particles , and @xmath238 labels the elements of the orthonormal basis : @xmath239 afterwards , alice communicates the result @xmath238 of her measurement to bob who performs an operation @xmath240 on his modes .
the operation is aimed to recover a state which maximizes the fidelity , such that the final state of the modes in @xmath194 and @xmath198 is as similar as possible to the initial state .
if a protocol with fidelity one is possibile , namely perfect teleportation from the modes in @xmath193 to the modes in @xmath198 , then the final state has to be pure and bob s modes have to be factorized with the remaining modes , because these are features of the initial state .
therefore , @xmath240 has to preserve the norm of the state ( [ measured ] ) . comparing the state after the measurement and the initial state , in order to maximize the fidelity , the state of the set @xmath198 that multiplies @xmath241 should be transformed into @xmath242 .
since these properties must be satisfied for all the initial states , i.e. for all the coefficients @xmath243 , the operator @xmath240 must transform the states @xmath244 into @xmath245 for all @xmath35 and @xmath165 , where @xmath246 is an arbitrary phase , and @xmath247 is a state with @xmath248 particles in the modes @xmath235 which is independent on the other indices . this transformation should rely on correlations among the indices @xmath165 , @xmath249 and @xmath250 induced by the coefficients @xmath251 and @xmath252 .
the result is @xmath253 after tracing out all modes but those in @xmath194 and @xmath198 , we get the unnormalized state @xmath254 where @xmath255 is the same as the initial state with the only difference that the set of modes @xmath193 is replaced by @xmath198 . equation is the average of @xmath256 with probability @xmath257 . assuming that the fidelity is one ,
the overlap between any initial state and the normalized state resulting from the @xmath238-th outcome should be one , namely @xmath258 with @xmath259 for all the initial states
@xmath84 . recalling the notation , and considering , among all possible initial states , @xmath260 ,
the condition ( [ fid.part ] ) can be re - written , after some simple algebra , as @xmath261 the latter equation holds for all @xmath35 , @xmath262 , @xmath165 and @xmath263 that can be arbitrary chosen in the exemplary initial state , and thus implies @xmath264 are indendent on @xmath35 and @xmath165 . plugging this result into the definition of the fidelity
, we get @xmath265 .
the fidelity is maximized if the cardinality of the sum over @xmath238 is maximal , that corresponds to a projective measurement onto the complete orthonormal basis @xmath266 .
the completeness of the measurement implies the identity @xmath267 hence , a perfect teleportation implies @xmath268 which is a contradiction .
the only assumption we made is that the teleportation is perfect , namely that the fidelity is one .
therefore , a perfect teleportation can not be implemented exploiting particle reservoirs .
if each set @xmath269 is made of one mode and the modes in @xmath235 are factorized from the others in the resource state ( [ res.reserv ] ) and in the projectors ( [ proj.reserv ] ) , the proof recovers a generalization of the teleportation in section [ twomode ] without reservoirs . in the following ,
the proof is generalized along several directions in order to recover all possible teleportation protocols described by locc .
* we only considered pure resource states because the maximum fidelity is attained by a pure resource state .
indeed , the linearity of teleportation implies that the fidelity provided by a mixture of pure resource states is the convex combination of the fidelities provided by each pure resource state .
thus , if the fidelity is strictly smaller than one for pure resource states , it is as well for mixed resource states .
* the above proof can be generalized to a scenario where alice projects onto pure states with different total numbers of particles and when the resource state is a mixture of states with different total numbers of particles .
indeed , the operation @xmath240 can be optimized only for one choice of the couple @xmath270 , introducing addional errors for the other values .
* we can further generalize the argument when alice projects onto degenerate subspaces . in this case
, the modes in @xmath193 and @xmath197 will be entangled with the other modes , providing a decrease of the fidelity when they are traced out .
* there is no loss of generality in considering projective measurements on alice s side .
in fact , it is known @xcite that any quantum operation , e.g a generalized measurement or completely positive map , is equivalent to a unitary operation @xmath166 followed by a projective measurement on an enlarged system .
thus , unitary operations @xmath166 on alice s side do not change bob s final state , and the extension of the system can be gathered in the set @xmath197 .
hence , this more general scheme is absorbed by the previous case .
* if no projective measurements are involved , there is no broadcast of information to bob s end . when a projective measurement is performed , information is teleported to bob s modes .
the cost of the projection is the complete erasure of information on alice s side .
for this reason , no further iterations or backward communication from bob to alice can improve the teleportation fidelity .
this proves the statement for a general locc protcol .
there is still the possibility to consider an infinite number of particles .
we have already shown that such an asymptotic , perfect teleportation is possible with the teleportation protocol described in section [ twomode ] , without any further generalization .
we have discussed how the teleportation protocol can be applied to the case of identical massive particles . due to the indistinguishability of particles , we applied a general notion of entanglement between subalgebras of observables . as a consequence , we identified local parties with orthogonal modes , such as in optical lattices where we can split the wells into groups .
we considered the following situation : a sender , alice , wants to teleport the state of one of her modes to a mode owned by a receiver , bob . to this aim
, they use an entangled shared two - mode state . in general , the mode whose state alice wants to teleport can be entangled with another mode .
we computed the general formula of the teleportation fidelity , that is the average overlap between the initial pure state and the teleported state , and the average entanglement of the teleported state , when one mode of a two - mode state is teleported .
we proved that the conservation of the total number of particles forbids the fidelity to be one for a finite number of particles , even if coherent interactions with a reservoir are allowed to overcome the superselection rule .
we computed the teleportation performances for several states .
in particular , the maximally entangled state ( [ max.ent.state ] ) , the symmetric coherent state and the ground state of the two - mode bose - hubbard hamiltonian ( [ bh ] ) provide perfect teleportation in the limit of large numbers of particles .
each of these states is interesting for different reasons .
the maximally entangled state ( [ max.ent.state ] ) provides perfect teleportation with high probability for a finite number of particles .
the symmetric coherent state can be prepared with current technologies in systems of ultracold atoms @xcite .
moreover , it is closly related to the mean field approximation and mesoscopic quantum coherent phenomena @xcite , and is considered a classical state for metrological purposes @xcite , while it is very useful for teleportation . in comparison ,
n00n states are extremely useful in quantum metrology @xcite while they are not for teleportation .
these differences are a consequence of the fact that teleportation and metrological performaces require different state properties : large entanglement , thus coherence among all the fock state , for teleportation , and coherence between two fock states with highly unbalanced population in two modes for phase estimation .
the teleportation performances of the ground state of the double well potential are appealing because this state can be generated with available techniques , i.e. magnetic traps and evaporative cooling @xcite .
finally , we briefly discussed possible generalizations to the teleportation of many modes and related difficulties in achieving high performances . * acknowledgement * u.m . acknowledges stefano zippilli for useful discussions , and funding by the grant j1 - 5439 of slovenian research agency .
u.m . and a.b .
acknoledge funding by deutsche forschungsgemeinschaft and by evaluierter fonds der albert ludwigs - universitaet freiburg .
in this appendix , we describe some additional details on teleportation performaces with the resource states discussed in section [ perf ] . which is diagonal in the fock basis , because , after alice s measurement , the state is always separable .
the teleportation fidelity and the average teleported entanglement are given respectively by equation and @xmath272 .
[ [ maximally - entangled - resources - and - probabilistic - perfect - teleportation ] ] maximally entangled resources and probabilistic , perfect teleportation ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@xmath278=\frac{\nu - n+1}{\nu+1}|\psi_{14}\rangle\langle\psi_{14}|+\frac{1}{\nu+1}\left(\sum_{l =- n}^{-1}q_l|\psi_{14}\rangle\langle\psi_{14}|q_l+\sum_{l=\nu - n+1}^{\nu}q_l|\psi_{14}\rangle\langle\psi_{14}|q_l\right).\ ] ] if alice s measurement results in a projection onto with a specific @xmath63 , the component @xmath279 of the initial state is teleported .
this projection exactly recovers the initial state , if @xmath280 , resulting in a perfect teleportation , namely the mapping from the second mode to the fourth mode , which occurs with probability @xmath142 .
all the other outcomes result in the partial teleportation of the component @xmath281 , with probability @xmath282 .
this lead to teleportation fidelity and to average maximally entanglement .
a possible strategy to optimize the performance of the teleportation protocol is to increase the probability of the perfect teleportation . on the one hand ,
such probability @xmath283 increases with the ratio @xmath284 and goes to one when @xmath285 .
this requires the ability to prepare states for a very large number of particles @xmath52 . on the other hand , considering a fixed and finite number of particles @xmath52 , we can improve the probability of perfect teleportation with repeated teleportations .
if alice s projection onto corresponds to a value @xmath145 or @xmath146 , which does not allow a perfect teleportation , she teleports another copy of the original state until she gets the outcome @xmath280 , namely a perfect teleportation .
after @xmath286 runs of the teleportation protocol , the probability of perfect teleportation is therefore all the maximally entangled resource states provide the same performance , up to local unitary operations .
these local unitary operations can be reabsorbed in the protocol , for instance redefining @xmath69 .
the choice in maximizes the fidelity of the maximally entangled state , whereas it is not optimal for other maximally entangled states which can lead to a fidelity smaller than that of separable states .
the redefinition of @xmath69 , by absorption of the unitaries , does not affect the average final entanglement @xmath292 . that are directly computed from the general equations and .
the second term of is the additional contribution to the fidelity with respect to separable resources .
it is negligible for large @xmath297 , increases with decreasing @xmath298 , and contributes to the teleportation fidelity at the same order as @xmath138 .
thus , the resulting improvement gives at most a larger prefactor , without changing the scaling with the numbers of particles .
moreover , the average final entanglement is of order one , while the maximum value of the negativity and its average over all pure states scale linearly with @xmath14 .
we stress that the operations @xmath299 are fully consistent with the conservation of the total particle number , which plays a crucial role in our study .
indeed , @xmath300 can be implemented with a unitary operation on the third mode and on an additional fifth mode , to preserve the total particle number : @xmath301 , where @xmath71 is the trace over the fifth mode , and @xmath303 is the identity matrix on the subspace orthogonal to the support of each of the other terms .
thus , @xmath304 is a unitary transformation which commutes with the total number of particles .
analogously , @xmath305 can be implemented with a total number preserving unitary operation on the fourth mode and an additional sixth mode .
in addition to the discussion in section [ su2.coh.states ] and from figures [ fid.sym.coh ] and [ ent.sym.coh ] , we note that the su(2 ) symmetric coherent state @xmath162 outperforms the maximally entangled states for @xmath306 up to large particle numbers @xmath52 .
the intuitive reason is the following : if a maximally entangled state is employed as a resource , the teleportation is perfect whenever alice measures @xmath63 with @xmath307 , which happens with a probability @xmath142 .
the teleported state resulting form any other of alice s outcomes is a projection ( [ proj.q ] ) of the original state . for a symmetric coherent resource state ,
the teleportation corresponding to alice s outcomes @xmath63 , with @xmath307 , is slightly distorted with respect to the perfect teleportation , but these outcomes do occur with higher probability than in the case of the maximally entangled resource state .
as an example of entangled resource state that does not outperform separable states , we numerically checked that the teleportation fidelity of the coherent state @xmath308 is smaller than the fidelity of separable states @xmath138 , for @xmath309 $ ] and @xmath310 $ ] .
indeed , from figure [ perf.coh ] the maximum fidelity with su(2 ) coherent resource states is achieved for the symmetric state @xmath162 .
any coherent state is equivalent to any other coherent state with the same value of @xmath150 and different phases @xmath152 , up to local unitary operations on the modes , i.e. of the form @xmath311 with a given function @xmath312 .
this means that coherent states with the same @xmath150 but different @xmath152 have the same entanglement @xcite .
nevertheless , these local operations can be reabsorbed in the operations @xmath69 , changing the teleportation protocol and thus its performance .
consistently , we find a dependence of the fidelity on the phase @xmath152 , even if neither the entanglement of su(2 ) coherent resource states nor the average final entanglement they produce change with @xmath152 . in this section ,
we stress some features of teleportation performances with the resource state being the ground state of the bose - hubbard model .
we numerically observed that the behaviour of the fidelity and of the average final entanglement , when the ground state is chosen in the double gaussian regime , are qualitatively the same as in the single gaussian regime .
the reason is that for small ( large ) @xmath52 the gaussians are highly peaked ( strongly spread out ) , and the small ( large ) coherence among fock states , thus the entanglement , justify low ( high ) teleportation performances . increasing @xmath52 ,
the spread of the gaussians increases monotonically , as well as the teleportation performances .
unfortunately , there is no explicit formula for the ground state in the critical regime .
the fidelity @xmath180 and the average final entanglement @xmath181 are larger than the fidelity @xmath189 and the average final entanglement @xmath292 of the maximally entangled state , respectively , for small values of @xmath14 up to large @xmath52 .
the intuitive reason is the same as the one given above for the symmetric coherent state . the number of such values of @xmath14 increases when @xmath174 approaches @xmath313 , and decreases to zero when @xmath174 moves away from @xmath313 . in the absence of interactions ,
i.e. @xmath314 , the exact ground state of the hamiltonian ( [ bh ] ) is the symmetric coherent state with @xmath157 and @xmath315 .
indeed , the symmetric coherent state is approximated by the gaussian state ( [ gauss ] ) with @xmath314 , as can be seen by application of stirling s approximation @xcite for a large number of particles @xmath52 .
consistently , the fidelities and the average final entanglement of these resource states have very similar quantitative behaviours , like in figures [ fid.sym.coh ] and [ ent.sym.coh ] . 102ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ] + 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop _
( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop `` , '' ( ) , @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) @noop ( ) , @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) and , eds . ,
@noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop `` , '' ( ) @noop * * , ( ) @noop _ _ ( , ) @noop * * , ( ) | we study teleportation with identical massive particles .
indistinguishability imposes that the relevant degrees of freedom to be teleported are not particles , but rather addressable orthogonal modes .
we discuss the performances of teleportation under the constraint of conservation of the total number of particles .
the latter inevitably decreases the teleportation fidelity . moreover ,
even though a phase reference , given by the coupling to a reservoir , circumverts the constraint , it does not restore perfect deterministic teleportation .
the latter is only achievable with some special resource entangled states and when the number of particles tends to infinity .
interestingly , some of such states are the many - particle atomic coherent states and the ground state of cold atoms loaded into a double well potential , which are routinely prepared in experiments . |
a permutation is a bijection from @xmath5=\{1,2,\ldots , n\}$ ] to @xmath5 $ ] .
let @xmath0 be the set of permutations of length @xmath6 .
let @xmath7 and @xmath8 be two permutations .
we say that @xmath9 _ contains _
@xmath2 if there exists a subsequence @xmath10 such that @xmath11 is order - isomorphic to @xmath2 ; in such a context @xmath2 is usually called a _
pattern_. we say that @xmath9 _ avoids _
@xmath2 , or is @xmath2-_avoiding _ , if such a subsequence does not exist .
the set of all @xmath2-avoiding permutations in @xmath0 is denoted @xmath12 . for an arbitrary finite collection of patterns @xmath13
, we say that @xmath9 avoids @xmath13 if @xmath9 avoids any @xmath14 ; the corresponding subset of @xmath0 is denoted @xmath15 .
while the case of permutations avoiding a single pattern has attracted much attention , the case of multiple pattern avoidance remains less investigated .
in particular , it is natural , as the next step , to consider permutations avoiding pairs of patterns @xmath16 , @xmath17 .
this problem was solved completely for @xmath18 ( see @xcite ) , for @xmath19 and @xmath20 ( see @xcite ) , and for @xmath21 ( see @xcite and references therein ) .
several recent papers @xcite deal with the case @xmath19 , @xmath22 for various pairs @xmath23 .
another natural question is to study permutations avoiding @xmath16 and containing @xmath17 exactly @xmath24 times .
such a problem for certain @xmath18 and @xmath25 was investigated in @xcite , and for certain @xmath19 , @xmath22 in @xcite .
the tools involved in these papers include continued fractions , chebyshev polynomials , and dyck words . _ chebyshev polynomials of the second kind _
( in what follows just chebyshev polynomials ) are defined by @xmath26 for @xmath27 .
evidently , @xmath28 is a polynomial of degree @xmath29 in @xmath30 with integer coefficients .
chebyshev polynomials were invented for the needs of approximation theory , but are also widely used in various other branches of mathematics , including algebra , combinatorics , and number theory ( see @xcite ) .
_ dyck words _ are words @xmath31 of @xmath32 verifying that @xmath33 and that for all @xmath34 , @xmath35 .
dyck words of length @xmath36 are enumerated by the @xmath6th catalan number @xmath37 whose generating function is @xmath38 .
we also consider words of @xmath39 of length @xmath6 enumerated by the @xmath6th fibonacci number @xmath40 with @xmath41 and @xmath42 whose generating function is @xmath43 .
apparently , for the first time the relation between restricted permutations and chebyshev polynomials was discovered by chow and west in @xcite , and later by mansour and vainshtein @xcite , krattenthaler @xcite .
these results related to a rational function @xmath44 for all @xmath45 .
+ an involution is a permutation such that its cycles are of length @xmath46 or @xmath47 that is @xmath7 is an involution if and only if @xmath48 for all @xmath49 $ ] .
some authors considered involutions with forbidden patterns . + regev in @xcite provided asymptotic formula for @xmath50-avoiding involutions of length @xmath6 and he also established that @xmath51-avoiding involutions of length @xmath6 are enumerated by motzkin numbers .
gessel @xcite exhibited the enumeration of such @xmath52-avoiding involutions of length @xmath6 .
moreover , gouyou
beauchamps @xcite obtained by an entirely bijective proof very nice exact formulas for the number of @xmath53-avoiding and @xmath54-avoiding involutions of length @xmath6 .
+ gire @xcite studied some permutations with forbidden subsequences and established a one - to - one correspondence between 1 - 2 trees having @xmath6 edges and permutations of length @xmath6 avoiding patterns @xmath55 and @xmath56 , the latter being allowed in the case where it is itself a subsequence of the pattern @xmath57 .
guibert @xcite also established bijections between 1 - 2 trees having @xmath6 edges and another set of permutations with forbidden subsequences and @xmath58-avoiding involutions of length @xmath6 and @xmath59-avoiding involutions of length @xmath6 ( and so with @xmath51-avoiding involutions of length @xmath6 by transposing the corresponding young tableaux obtained by applying the robinson - schensted algorithm ) .
he also shown @xcite a bijection between vexillary involutions ( that is @xmath60-avoiding involutions ) and @xmath61-avoiding involutions .
more recently , guibert , pergola and pinzani @xcite established a one - to - one correspondence between 1 - 2 trees having @xmath6 edges and vexillary involutions of length @xmath6 .
so all these sets are enumerated by the @xmath6th motzkin number @xmath62 .
it remains a connected open problem : in @xcite conjectures that @xmath63-avoiding involutions of length @xmath6 are also enumerated by the @xmath6th motzkin number .
in this paper we present a general approach to the study of involutions in @xmath0 avoiding @xmath1 ( or containing @xmath1 exactly once ) , and avoiding ( or containing exactly once ) an arbitrary pattern @xmath64 . as a consequence ,
we derive all the previously known results for this kind of problems , as well as many new results .
some results are also established by bijections as for an example a bijection between @xmath1-avoiding involutions and primitive dyck words .
the paper is organized as follows .
the case of involutions avoiding both @xmath1 and @xmath2 is treated in section @xmath47 .
we present an explicit expression in terms of chebyshev polynomials for several interesting cases .
the case of involutions avoiding @xmath1 and containing @xmath2 exactly once is treated in section @xmath65 . here
again we present an explicit expression in terms of chebyshev polynomials for several interesting cases .
finally , the cases of involutions containing @xmath1 exactly once and either avoiding or containing exactly once an arbitrary pattern @xmath2 is treated in sections @xmath66 and @xmath67 ; respectively .
let @xmath68 denote the number of involutions in @xmath69 avoiding @xmath13 , and let @xmath70 be the corresponding generating function .
the following proposition is the base of all the other results in this section , which holds immediately from definitions .
[ prom ] for any involution @xmath71 such that @xmath72 holds either , 1 . for @xmath73 $ ] , @xmath74 , where * @xmath75 is a @xmath1-avoiding permutation of the numbers @xmath76 , *
@xmath77 is a @xmath1-avoiding permutation of the numbers @xmath78 such that @xmath79 is the identity permutation of @xmath80 , * @xmath81 is a @xmath1-avoiding involution of the numbers @xmath82 ; 2 . for @xmath83 , @xmath84 where @xmath75 is an involution in @xmath85 .
as a corollary of proposition [ prom ] we get the generating function for the number of involutions in @xmath69 as follows .
( see ( * ? ? ?
* prop . 5 ) ) let @xmath86 be the generating function for the catalan numbers ; then @xmath87 proposition [ prom ] yields for all @xmath88 , @xmath89 } c_{j-1}i_\varnothing(n-2j)+i_\varnothing(n-1),\ ] ] where @xmath90 is the @xmath91th catalan number .
besides @xmath92 , therefore in terms of generating function we get that @xmath93 we can also prove this result by a bijective point of view .
let @xmath94 be the language of primitive dyck words .
+ the number of such words of @xmath95 of length @xmath6 is the central binomial coefficient @xmath96 with @xmath97 .
indeed , any primitive dyck word @xmath31 of @xmath95 can be uniquely written as @xmath98 where @xmath99 is a dyck word ( that is @xmath100 and latexmath:[$|w_i|_x = @xmath31 can also be uniquely written as @xmath102 .
so primitive dyck words @xmath31 of @xmath95 of length @xmath6 are in bijection with bilateral words of latexmath:[$\ { w \in \{x,\overline{x}\}^ * : |w|_x = length @xmath6 trivially enumerated by @xmath96 .
[ phi ] there is a bijection @xmath104 between involutions in @xmath69 and primitive dyck words of @xmath95 of length @xmath6 .
moreover , the number of fixed points of the involution corresponds to the difference between the number of letters @xmath30 and @xmath105 into the primitive dyck word .
let @xmath106 be a @xmath1-avoiding involution on @xmath5 $ ] having @xmath107 fixed points .
according to proposition [ prom ] we have @xmath108 with @xmath109 ( @xmath110 has no fixed points and constitutes cycles with @xmath111 or @xmath112 ) , @xmath111 does not contain fixed point and @xmath113 ( @xmath30 is the first fixed point ) .
we obtain two @xmath1-avoiding involutions on @xmath114 $ ] from @xmath106 : the first one is given by inserting a fixed point between @xmath110 and @xmath111 , and the second one ( iff @xmath106 has at least one fixed point ) is given by modifying the first fixed point @xmath30 by a cycle starting between @xmath110 and @xmath111 .
all @xmath1-avoiding involutions can be obtained ( and only once ) by applying this rule , starting from the empty involution of length @xmath115 .
+ let @xmath31 be a primitive dyck word of @xmath95 of length @xmath6 such that @xmath116 .
so we have @xmath117 where @xmath99 are dyck words for all @xmath118 .
we obtain two primitive dyck words of length @xmath119 from @xmath31 : @xmath120 and @xmath121 ( iff @xmath122 ) .
all primitive dyck words can be obtained ( and only once ) by applying this rule , starting from the empty word of length @xmath115 . clearly ,
these two generating trees for the @xmath1-avoiding involutions and the primitive dyck words can be characterized by the following succession system : + @xmath123 figure [ fig - ag ] shows the bijection @xmath104 between @xmath1-avoiding involutions and primitive dyck words ( and the labels of the succession system which characterizes them ) for the first values .
= 590.0pt [ nb132ptfix ] the number of @xmath1-avoiding involutions of length @xmath6 is @xmath124 .
moreover , the number of @xmath1-avoiding involutions of length @xmath6 having @xmath107 fixed points with @xmath125 ( and @xmath107 is odd iff @xmath6 is odd ) is the ballot number @xmath126 .
indeed , the number of primitive dyck words @xmath31 of @xmath95 according to the length and @xmath127 is given by the ballot number ( or delannoy number @xcite or distribution @xmath9 of the catalan number @xcite ) .
in particular , the number of @xmath1-avoiding involutions of length @xmath36 without fixed points is @xmath128 the @xmath6th catalan number .
the following theorem is the base of all the other results in this section .
[ thg ] let @xmath13 set of patterns , @xmath129 , and let @xmath130 be the generating function for the number of @xmath13-avoiding permutations in @xmath69 .
then @xmath131 proposition [ prom ] with definitions of @xmath132 yields for @xmath133 , @xmath134 } s_t(j-1)i_{t'}(n-2j),\ ] ] where @xmath135 is the number of permutations in @xmath136 .
hence , in terms of generating functions we have @xmath137 so the theorem holds .
[ ex1 ] ( see @xcite ) let us find @xmath138 ; let @xmath139 and @xmath140 , so theorem [ thg ] gives @xmath141 where by definitions @xmath142 , hence @xmath143 which means the number of involutions @xmath144 is given by @xmath145}$ ] for all @xmath146 .
similarly , @xmath147 , so the number of involutions @xmath148 is given by @xmath40 the @xmath6th fibonacci number .
the case of varying @xmath3 is more interesting . as an extension of example [ ex1 ]
let us consider the case @xmath149 .
[ th12k ] for all @xmath45 , @xmath150 immediately , the theorem holds for @xmath151 .
let @xmath152 ; theorem [ thg ] gives @xmath153 on the other hand , the generating function for the sequence @xmath154 is given by @xmath155 ( see ( * ? ? ?
* th . 1 ) ) with @xmath156 ( see @xcite ) we get that @xmath157 besides @xmath158 , hence by use induction on @xmath3 and definitions of @xmath159 the theorem holds .
we consider now a combinatorial point of view for this result .
let @xmath106 be a @xmath1-avoiding involution . clearly , if @xmath106 avoids @xmath4 then @xmath106 has less than @xmath3 fixed points . moreover , if @xmath106 of length @xmath6 having less than @xmath3 fixed points is obtained from an @xmath1-avoiding involution @xmath160 of length less than @xmath6 having @xmath3 fixed points ( and take @xmath160 as big as possible ) by applying the rules described for bijection @xmath104 given by theorem [ phi ] , then @xmath106 contains a subsequence @xmath161 because the first fixed points of @xmath160 become cycles into @xmath106 such that the beginning of these cycles and the last remaining fixed points of @xmath160 into @xmath106 constitute a subsequence of type @xmath162 .
so the succession system @xmath163 + @xmath164 + characterizes the generating tree of the involutions avoiding both @xmath1 and @xmath162 .
it is easy to see that for @xmath3 odd , the number of involutions of length @xmath165 avoiding both @xmath1 and @xmath4 is the twice of the number of involutions of length @xmath166 avoiding both @xmath1 and @xmath4 .
moreover , the reader can note that the set of labels of this succession system is finite and so the corresponding generating function is rational .
more precisely , we immediately deduce from the previous succession system that the number of involutions of length @xmath6 avoiding both @xmath1 and @xmath4 and having @xmath107 fixed points is given by the @xmath167th component of the vector given by @xmath168 where @xmath169 is a vector of @xmath3 elements and @xmath170 is a @xmath171 matrix .
+ in another way , we can see that as an automaton where the states are @xmath172 and the transitions are arrows from @xmath173 to @xmath174 for @xmath175 and from @xmath173 to @xmath176 for @xmath177 . the bijection @xmath104 establishes a one - to - one correspondence between involutions of length @xmath6 avoiding both @xmath1 and @xmath178 and having @xmath107 fixed points , and primitive dyck words @xmath179 of @xmath95 of length @xmath6 such that @xmath99 is a dyck word of height less than @xmath180 ( that is @xmath100 and @xmath181 and for all @xmath182 , @xmath183 ) for all @xmath184 . + in particular involutions of length @xmath36 avoiding both @xmath1 and @xmath4 without fixed points are in bijection by @xmath104 with dyck words of length @xmath36 of height less than @xmath3 .
these dyck words of bounded height was considered by kreweras @xcite and viennot @xcite .
in particular , dyck words of length @xmath36 of height less than @xmath46 , @xmath47 , @xmath65 , @xmath66 , @xmath67 are respectively enumerated by @xmath115 , @xmath46 , @xmath185 , @xmath186 , @xmath187 for all @xmath188 . + we provide some simple bijections for special cases @xmath189 ( related to example [ ex1 ] ) by generating some well known words in the same way as involutions avoiding both @xmath1 and @xmath162 .
fist of all , we consider the case @xmath190 and the words of @xmath191 or @xmath192 enumerated by the powers of @xmath47 we can generate from the empty word labeled @xmath193 by the rules : + @xmath194 + such that the words labeled @xmath193 start by @xmath195 whereas the words labeled @xmath196 or @xmath197 start by @xmath198 .
+ so , words of @xmath199 ( respectively @xmath200 ) are in bijection with involutions avoiding both @xmath1 and @xmath201 of length @xmath36 ( respectively @xmath202 ) enumerated by @xmath203 ( respectively @xmath203 ) .
next we consider the case @xmath204 and the words of @xmath205 enumerated by the fibonacci numbers we can generate from the empty word labeled @xmath193 by the rules : + @xmath206 + such that the words labeled @xmath193 or @xmath207 start by @xmath208 whereas the words labeled @xmath196 or @xmath197 start by @xmath198 .
+ so , words of @xmath205 of length @xmath6 are in bijection with involutions in @xmath209 enumerated by @xmath40 the @xmath6th fibonacci number .
now we consider the case @xmath210 and the words of @xmath211 or @xmath212 enumerated by the powers of @xmath65 we can generate from the empty word labeled @xmath193 by the rules : + @xmath213 + such that the words labeled @xmath193 or @xmath196 start by @xmath214 , the words labeled @xmath207 or @xmath215 start by @xmath216 , and the words labeled @xmath197 start by @xmath195 ( and have one letter more than words labeled @xmath193 or @xmath215 at the same level ) .
+ so , words of @xmath217 ( respectively @xmath218 ) are in bijection with involutions avoiding both @xmath1 and @xmath53 of length @xmath202 ( respectively @xmath219 ) enumerated by @xmath220 ( respectively @xmath221 ) . figure [ fig - forbid ] ( an output of the software ` forbid ` @xcite ) shows the first values for the number of involutions avoiding both @xmath1 and @xmath4 for @xmath222 according to the number of fixed points .
involutions @xmath223 according to @xmath224 for @xmath225 .... = 1 = 2 = 2 = 4 = 4 = 8 = 8 = 16 = 16 = 32 = 32 = 64 = 64 = 128 = 128 2 : - 1 0 2 0 4 0 8 0 16 0 32 0 64 0 1 : 1 0 2 0 4 0 8 0 16 0 32 0 64 0 128 0 : 0 1 0 2 0 4 0 8 0 16 0 32 0 64 0 1 : 2 : 3 : 4 : 5 : 6 : 7 : 8 : 9 : 10 : 11 : 12 : 13 : 14 : 15 : [ n ] .... involutions @xmath226 according to @xmath224 for @xmath225 .... = 1 = 2 =3 = 5 = 8 = 13 = 21 = 34 = 55 = 89 = 144 = 233 = 377 = 610 = 987 3 : - - 1 0 3 0 8 0 21 0 55 0 144 0 377 2 : - 1 0 3 0 8 0 21 0 55 0 144 0 377 0 1 : 1 0 2 0 5 0 13 0 34 0 89 0 233 0 610 0 : 0 1 0 2 0 5 0 13 0 34 0 89 0 233 0 1 : 2 : 3 : 4 : 5 : 6 : 7 : 8 : 9 : 10 : 11 : 12 : 13 : 14 : 15 : [ n ] ....
involutions @xmath227 according to @xmath224 for @xmath225 .... = 1 = 2 =3 = 6 = 9 = 18 = 27 = 54 = 81 = 162 = 243 = 486 = 729 = 1458 = 2187 4 : - - - 1 0 4 0 13 0 40 0 121 0 364 0 3 : - - 1 0 4 0 13 0 40 0 121 0 364 0 1093 2 : - 1 0 3 0 9 0 27 0 81 0 243 0 729 0 1 : 1 0 2 0 5 0 14 0 41 0 122 0 365 0 1094 0 : 0 1 0 2 0 5 0 14 0 41 0 122 0 365 0 1 : 2 : 3 : 4 : 5 : 6 : 7 : 8 : 9 : 10 : 11 : 12 : 13 : 14 : 15 : [ n ] .... [ ex2 ] let us find @xmath229 ; let @xmath230 and @xmath231 , so theorem [ thg ] gives @xmath232 where by definitions @xmath233 , hence @xmath234 which means the number of involutions @xmath235 is given by @xmath145}$ ] for all @xmath146 . similarly , @xmath236 , so the number of involutions @xmath237 is given by @xmath40 the @xmath6th fibonacci number .
we can easily prove by a combinatorial way that @xmath235 is given by @xmath145}$ ] .
+ an involution @xmath106 avoiding both @xmath1 and @xmath238 of length @xmath6 can be either @xmath239 with @xmath240 or @xmath241 such that @xmath110 is also an involution avoiding both @xmath1 and @xmath238 ( of length @xmath242 , and if we subtract @xmath173 to each element ) .
we code this recursive decomposition of an involution @xmath106 avoiding both @xmath1 and @xmath238 by a word of nonnegative integers formed by the successive positive numbers @xmath173 and whose last nonnegative integer is ( the smallest integer of ) the number of the fixed points in @xmath106 divided by @xmath47 .
this coding is clearly bijective .
+ for example , the involutions @xmath243 , @xmath46 , @xmath244 , @xmath245 , @xmath201 , @xmath55 , @xmath51 , @xmath246 , @xmath58 and @xmath59 are respectively coded by @xmath115 , @xmath115 , @xmath46 , @xmath247 , @xmath46 , @xmath247 , @xmath47 , @xmath248 , @xmath249 and @xmath250 . moreover , the involution @xmath251
\ 19 \linebreak[0 ] \ 20 \linebreak[0 ] \ 16 \linebreak[0 ] \ 17 \linebreak[0 ] \ 18 \linebreak[0 ] \ 15 \linebreak[0 ]
\ 14 \linebreak[0 ] \ 9 \linebreak[0 ] \
10 \linebreak[0 ] \ 11 \linebreak[0 ] \ 12 \linebreak[0 ] \ 13 \linebreak[0 ] \ 8 \linebreak[0 ] \ 7 \linebreak[0 ] \ 4 \linebreak[0 ]
\ 5 \linebreak[0 ] \ 6 \linebreak[0 ] \ 2 \linebreak[0 ] \ 3 \linebreak[0 ] \ 1 $ ] in @xmath252 is coded by @xmath253 .
+ thus , involutions avoiding both @xmath1 and @xmath238 of length @xmath6 are coded by words @xmath254 with @xmath255 , @xmath256 for all @xmath257 , @xmath258 and @xmath259 .
trivially , words @xmath31 are in bijection with words @xmath260 which are compositions of @xmath261 into @xmath262 positive parts enumerated by @xmath263 . the case of varying @xmath3 is more interesting . as an extension of example [ ex2 ] let us consider the case @xmath264 .
similarly as theorem [ th12k ] we have [ th21k ] for all @xmath45 , @xmath265 therefore , theorem [ th12k ] and theorem [ th21k ] yields @xmath266 .
we establish a bijection for this result .
[ 12k=2134k ] there is a bijection between involutions avoiding both @xmath1 and @xmath267 of length @xmath6 and involutions avoiding both @xmath1 and @xmath268 of length @xmath6 , for any @xmath269 .
+ moreover , two involutions in bijection have the same number of fixed points @xmath107 for all @xmath270 whereas the involutions avoiding both @xmath1 and @xmath267 having @xmath271 or @xmath272 fixed points correspond to the involutions avoiding both @xmath1 and @xmath268 having @xmath271 or more fixed points . in order to establish this result
we consider a generating tree for the involutions avoiding both @xmath1 and @xmath268 which is characterized by the same succession system @xmath163 given in subsection [ subsec21 ] characterizing a generating tree for the involutions avoiding both @xmath1 and @xmath267 .
+ so let @xmath106 be an involution avoiding both @xmath1 and @xmath268 of length @xmath6 and let @xmath273 be the number of fixed points of @xmath106 .
the label @xmath274 of @xmath106 is defined by @xmath275 if @xmath276 or by @xmath277 if @xmath278 and @xmath279 or by @xmath280 if @xmath278 and @xmath281 .
of course , the empty involution of length @xmath115 has label @xmath193 .
we obtain @xmath160 an involution avoiding both @xmath1 and @xmath268 of length @xmath119 by applying the following rules : * if @xmath282 $ ] , we have @xmath283 with @xmath109 , and then @xmath160 obtained by inserting a fixed point between @xmath110 and @xmath111 has label @xmath167 . *
if @xmath277 , we have @xmath284 with @xmath285 , and then @xmath160 obtained by inserting a fixed point between @xmath110 and @xmath30 has label @xmath286 .
* if @xmath287 $ ] , we have @xmath108 with @xmath109 , @xmath113 and @xmath288 for all @xmath289 . then @xmath160 obtained by modifying the first fixed point @xmath30 by a cycle starting between @xmath110 and @xmath111 ( and ending in @xmath30 ) has label @xmath290 .
* if @xmath280 , we have @xmath291 with @xmath292 , and then @xmath160 obtained by inserting a fixed point between @xmath110 and @xmath30 has label @xmath293 . *
if @xmath277 , we have @xmath294 with @xmath295 , @xmath285 , @xmath296 and @xmath297 for all @xmath298 . then @xmath160 obtained by modifying the @xmath299 fixed points between @xmath110 and @xmath111 by @xmath300 consecutive cycles each of difference ( between the index and the value ) @xmath300 that is @xmath301 has label @xmath302 .
there is a bijection between permutations avoiding both @xmath1 and @xmath267 of length @xmath6 and permutations avoiding both @xmath1 and @xmath268 of length @xmath6 , for any @xmath269 . by proposition [ prom ]
, we deduce that permutations @xmath106 avoiding both @xmath1 and @xmath267 ( respectively @xmath268 ) of length @xmath6 are in bijection with involutions without fixed points @xmath303 avoiding both @xmath1 and @xmath267 ( respectively @xmath268 ) of length @xmath36 . moreover , a particular case of theorem [ 12k=2134k ] establishes a one - to - one correspondence between involutions avoiding both @xmath1 and @xmath267 without fixed points and involutions avoiding both @xmath1 and @xmath268 without fixed points .
[ ex22 ] by proposition [ prom ] it is easy to obtain for @xmath305 , @xmath306 + 1.\ ] ] we consider a combinatorial approach to show example [ ex22 ] .
clearly , we have that involutions avoiding both @xmath1 and @xmath56 of length @xmath6 are @xmath307 for all @xmath308 and that involutions avoiding both @xmath1 and @xmath55 of length @xmath6 are @xmath309 for all @xmath310 . as an extension of example [ ex22 ]
let us consider the case @xmath311\}$ ] , where @xmath312=(d+1,d+2,\dots , k,1,2,\dots , d)$ ] .
[ thkd ] for any @xmath152 , @xmath313 , @xmath314}=\frac{1}{x(u_d(t)-u_{d-1}(t))}\left [ u_{d-1}(t)+\frac{u_{k-2d-1}(t)}{u_{k - d}(t)u_{k - d-1}(t)}\sum_{j=0}^{k - d-1 } u_j(t ) \right],\quad t=\frac{1}{2x}.\ ] ] proposition [ prom ] yields , in the second case the generating function for the number of involutions @xmath312$]-avoiding permutations is @xmath315}(x)$ ] . in the first case
, we assume that @xmath81 either @xmath196 avoiding @xmath316 , or @xmath197 containing @xmath316 . in @xmath196 , @xmath75 and @xmath77 avoiding @xmath317 ,
so the generating function for these number of involutions is @xmath318 ( similarly theorem [ th12k ] ) . in @xmath197 , @xmath75 and @xmath77 avoiding @xmath319 , so the generating function for these number of involutions is @xmath320}(x)-i_{12\dots(k - d)}(x))$ ] ( the generating function for the number of involutions in @xmath321)$ ] such containing @xmath322 is given @xmath323}(x)-i_{12\dots(k - d)}(x)$ ] ) .
therefore @xmath324}(x)&=1+xi_{[k , d]}(x)+x^2r_{k - d-1}(x^2)i_{12\dots(k - d)}(x)+\\ & + x^2r_{d-1}(x^2)(i_{[k , d]}(x)-i_{12\dots(k - d)}(x ) ) , \end{array}\ ] ] which means that @xmath314}(x)=\frac{1}{1-x - x^2r_{k - d-1}(x^2)}\cdot(1+x^2i_{12\dots ( k - d)}(x)(r_{k - d-1}(x^2)-r_{d-1}(x^2))).\ ] ] hence , by use the identities @xmath156 and @xmath325 , the theorem holds .
[ ex3412 ] theorem [ thkd ] yields for @xmath204 and @xmath326 , the number of involutions @xmath327 is given by @xmath40 the @xmath6th fibonacci number .
we consider a combinatorial approach to show example [ ex3412 ] .
an involution @xmath106 avoiding both @xmath1 and @xmath58 of length @xmath6 can be written @xmath328 with @xmath329 such that @xmath110 is also an involution avoiding both @xmath1 and @xmath58 ( of length @xmath330 , and if we subtract @xmath46 to each element ) .
we code @xmath106 by a word of @xmath331 of length @xmath6 in that way : @xmath198 if @xmath332 , @xmath208 if @xmath333 and nothing if @xmath334 for all @xmath329 .
this coding is clearly bijective .
+ following @xcite we say that @xmath8 is a _
pattern if it can be represented as @xmath335 so that each of @xmath336 is nonempty , @xmath337 is a layered permutation of @xmath338 for some @xmath339 , and @xmath340 .
for example , @xmath341 and @xmath342 are wedge patterns . for a further generalization of theorem [ th12k ] , theorem [ th21k ] and ( * ? ? ?
2.6 ) , consider the following definition .
we say that @xmath343 is a _
double - wedge _ pattern if there exist a wedge pattern @xmath344 such that @xmath345 for example , the double - wedge patterns of length @xmath247 are @xmath346 5 $ ] , @xmath347 213\linebreak[0 ] 4\linebreak[0 ] 5\linebreak[0]$ ] , @xmath348 31245\linebreak[0]$ ] , @xmath349 41235\linebreak[0]$ ] , @xmath350 23145\linebreak[0]$ ] , @xmath351 42135\linebreak[0]$ ] , @xmath352 34125\linebreak[0]$ ] , @xmath353 23415\linebreak[0]$ ] and @xmath354 32415\linebreak[0]$ ] .
[ dwedge ] for any double - wedge pattern @xmath355 @xmath356 first of all , let us find the generating function @xmath357 where @xmath358 . by use proposition [ prom ]
we obtain in the first case @xmath359 , and in the second case @xmath360 where @xmath361 is the generating function for the number of permutations in @xmath362 , therefore ( @xmath46 for the empty permutation ) @xmath363 on the other hand , mansour and vainshtein proved @xmath364 for any wedge pattern @xmath160 , so @xmath365 by use the identity @xmath366 we have @xmath367 now , let us find @xmath368 in terms of @xmath369 .
so , by use the identity @xmath370 and use the symmetric inverse operation , the first part of the theorem holds .
[ extwedge ] for any wedge pattern @xmath344 the generating function for the number of permutations in @xmath371 ( or @xmath372 , or @xmath373 ) is given by @xmath374 , for all @xmath375 .
let @xmath376 and let @xmath377 be the generating function for the number of permutations in @xmath378 . by ( * ? ? ?
1 ) we have @xmath379 on the other hand , by ( * ? ? ?
2.6 ) and @xmath160 a wedge pattern in @xmath380 we have @xmath381 .
therefore , by use the identity @xmath366 we get @xmath382 by use the definitions of chebyshev polynomials of the second kind it is easy to see @xmath383 hence by use again theorem ( * ? ? ?
1 ) we have @xmath384 .
similarly we obtain the other cases . as a corollary of theorem [ dwedge ] we have [ idwedge ] for any double wedge pattern @xmath385 @xmath386 since , if @xmath387 and @xmath388 , then theorem [ thg ] yields @xmath389 , and by use ( * ? ? ?
1 ) we have @xmath390 , where @xmath391 and @xmath392 two patterns in @xmath393 .
hence , the theorem holds by use theorem [ dwedge ] , theorem [ extwedge ] , and induction on @xmath107 . in view of theorem [ dwedge ] and theorem [ idwedge ]
it is a challenge to find a bijective proof .
now , let us restrict more than two patterns ( @xmath1 and two other patterns ) .
[ ex3 ] let us find @xmath394 ; let @xmath395 and @xmath396 , so theorem [ thg ] gives @xmath397 where by definitions @xmath398 , hence @xmath399 which means the number of involutions @xmath400 is given by @xmath401 , and @xmath402 is given by @xmath403 for all @xmath146 , where @xmath404 is the @xmath405th fibonacci number .
we consider a combinatorial approach to show example [ ex3 ] by distinguishing the cases of odd and even length .
+ an involution @xmath106 avoiding @xmath1 , @xmath201 and @xmath238 of length @xmath202 can be written either @xmath406 or @xmath407 or @xmath46 ( if @xmath408 ) such that @xmath110 and @xmath111 are also involutions avoiding @xmath1 , @xmath201 and @xmath238 ( of length @xmath409 for @xmath110 if we subtract @xmath46 to each element , of length @xmath410 for @xmath111 if we subtract @xmath47 to each element ) .
we code @xmath106 by a word of @xmath39 of length @xmath6 in that way : @xmath198 if @xmath411 , @xmath208 if @xmath412 and nothing if @xmath413 for all @xmath329 .
this coding is clearly bijective .
+ an involution @xmath106 avoiding @xmath1 , @xmath201 and @xmath238 of length @xmath36 can be written either @xmath414 ( that includes @xmath415 if @xmath416 ) or @xmath417 or @xmath418 ( if @xmath416 ) or the empty involution ( if @xmath408 ) such that @xmath110 and @xmath111 are also involutions avoiding @xmath1 , @xmath201 and @xmath238 ( of length @xmath419 for @xmath110 if we subtract @xmath46 to each element , of length @xmath420 for @xmath111 if we subtract @xmath47 to each element )
. we code @xmath106 by a word of @xmath39 of length @xmath119 in that way : @xmath198 if @xmath412 for all @xmath421 , @xmath208 if @xmath422 , @xmath208 if @xmath423 for all @xmath424 , @xmath425 if @xmath426 and @xmath427 if @xmath428 .
moreover , the empty involution is coded by @xmath198 .
this coding is clearly bijective .
using definitions and theorem [ thg ] it is easy to see the following .
[ th12k213 ] for all @xmath45 , @xmath429 [ ex312decr ] using proposition [ prom ] it is easy to see for @xmath88 , @xmath430 + 1.\ ] ] we consider a combinatorial approach to show example [ ex312decr ] . clearly , we have that involutions avoiding @xmath1 , @xmath238 and @xmath55 of length @xmath6 are @xmath431 and also @xmath432 if @xmath433 with @xmath434
. we also have that involutions avoiding @xmath1 , @xmath238 and @xmath59 of length @xmath6 are @xmath435 for all @xmath310 .
let @xmath436 denote the number of involutions in @xmath69 containing @xmath2 exactly @xmath29 times , and let @xmath437 be the corresponding generating function . let us start by the following example .
[ exbb0 ] by proposition [ prom ] it is easy to see @xmath438 which means @xmath439 . as extension of example [ exbb0 ]
let us consider the case @xmath440 .
[ th1_12 ] for all @xmath45 ; @xmath441 by proposition [ prom ] we have for @xmath442 , @xmath443 } s_{12\dots ( k-1)}(j-1)i_{12\dots k}^{1}(n-2j),\ ] ] where @xmath444 is the number of @xmath4-avoiding permutations in @xmath445 .
besides @xmath446 for all @xmath447 , and @xmath448 .
similarly as proof of theorem [ th12k ] we have @xmath449 hence , by induction on @xmath3 with initial condition @xmath450 , the theorem holds .
similarly as theorem [ th1_12 ] we have an explicit formula when @xmath451 or @xmath452 .
for all @xmath152 ; @xmath453 more generally , by proposition [ prom ] and the argument proof of theorem [ th12k ] we get [ th12kr ] for any @xmath454 @xmath455 where @xmath456 is the generating function for the number of permutations in @xmath0 containing @xmath457 exactly @xmath198 times .
in @xcite found an explicit formula for @xmath458 , so theorem [ th12kr ] yields a recurrence for @xmath459 .
for example , following @xcite ( ( * ? ? ?
3.1 ) ) we have a recurrence for @xmath460 where @xmath461 .
let @xmath45 ; for @xmath461 @xmath462 the above theorem yields for @xmath463 an explicit formula for @xmath464 . for all @xmath45 , @xmath465
we first relate involutions containing @xmath1 once to @xmath1-avoiding involutions .
[ bij132once-132 ] there is a bijection @xmath466 between involutions containing @xmath1 exactly once of length @xmath6 having @xmath107 fixed points with @xmath467 and @xmath1-avoiding involutions of length @xmath468 having also @xmath107 fixed points .
let @xmath469 with @xmath113 , @xmath470 and @xmath471 be an involution containing @xmath1 once ( that is subsequence @xmath472 ) of length @xmath6 having @xmath107 fixed points .
we replace the subsequence @xmath472 by a fixed point between @xmath111 and @xmath112 in order to obtain an @xmath1-avoiding involution of length @xmath468 having @xmath107 fixed points .
note that the only possibility to have exactly once @xmath1 subsequence is a cycle with a fixed point just to its left .
moreover , @xmath473 in order to forbid another @xmath1 subsequence and cycles are only allowed from @xmath110 to @xmath111 and from @xmath110 to @xmath112 ( and not from @xmath111 , @xmath110 , @xmath111 , @xmath112 respectively to @xmath112 , @xmath110 , @xmath111 , @xmath112 ) whereas fixed points can uniquely be into @xmath112 .
clearly the involution we obtain avoids @xmath1 and in particular , the fixed point @xmath474 can not be a part of an @xmath1-subsequence because it can not be the @xmath65 or @xmath47 ( all the elements on its left are greater than it ) and it can not be the @xmath46 ( there is no cycle starting on its right ) .
+ let @xmath475 with @xmath476 and @xmath477 for all @xmath478 ( that is @xmath24 is the first fixed point ) , @xmath479 for all @xmath480 ( all the elements of @xmath481 are cycles ending into @xmath482 ) , @xmath483 $ ] for all @xmath484 and @xmath485 $ ] for all @xmath486 ( @xmath487 is entirely constituted by cycles from @xmath488 to @xmath489 ) be an @xmath1-avoiding involution of length @xmath468 having @xmath107 fixed points .
we modify the fixed point @xmath24 by a cycle starting between @xmath488 and @xmath489 ( and ending between @xmath489 and @xmath482 ) and by adding a fixed point just to the right of @xmath488 in order to obtain an involution containing @xmath1 once of length @xmath6 having @xmath107 fixed points .
proposition [ prom ] leads immediately to the decomposition of @xmath160 .
the involution we obtain contains @xmath1 exactly once that is the subsequence we modify and insert .
there is no other @xmath1-subsequence and in particular , the fixed point inserted and the start of the new cycle can not be the @xmath65 or @xmath47 of another @xmath1-subsequence ( all the elements on their left are greater than them ) , the fixed point inserted and the start of the new cycle and the end of the new cycle can not be the @xmath46 of another @xmath1-subsequence ( there is no cycle starting on their right ) , the end of the new cycle can not be the @xmath65 of another @xmath1-subsequence ( because in that case the @xmath47 must be connected to @xmath481 and the @xmath46 must be the fixed point inserted or an element of @xmath489 that forms an @xmath56-subsequence ) , and the end of the new cycle can not be the @xmath47 of another @xmath1-subsequence ( because in that case the @xmath46 must be an element of @xmath490 or the start of the new cycle and the @xmath65 must be the fixed point inserted or an element of @xmath489 that forms an @xmath491-subsequence excepted for the fixed point inserted and the new cycle ) .
+ so we have established a bijection between @xmath106 an involution containing @xmath1 once and @xmath160 an @xmath1-avoiding involution where @xmath492 , @xmath110 corresponds to @xmath490 , @xmath493 and @xmath112 corresponds to @xmath482 .
for example , the involution @xmath494 \ 19 \linebreak[0 ]
\ 17 \linebreak[0 ] \ 18 \linebreak[0 ] \ 16 \linebreak[0 ] \ 12 \linebreak[0 ] \
11 \linebreak[0 ] \ 13 \linebreak[0 ] \ { \bf 9 } \linebreak[0 ] \ { \bf 14 } \linebreak[0 ] \ 7 \linebreak[0 ] \ 6 \linebreak[0 ] \ 8 \linebreak[0 ] \ { \bf 10 } \linebreak[0 ] \ 15 \linebreak[0 ] \ 5 \linebreak[0 ] \ 3 \linebreak[0 ] \ 4 \linebreak[0 ] \ 2 \linebreak[0 ] \ 20 \linebreak[0 ] \ 21 \linebreak[0 ]
\ 1 \linebreak[0 ] \ 23 $ ] containing @xmath1 once ( the subsequence
@xmath495 @xmath496 @xmath247 ) corresponds to the @xmath1-avoiding involution @xmath497 \
17 \linebreak[0 ] \ 15 \linebreak[0 ]
\ 16 \linebreak[0 ] \ 14 \linebreak[0 ] \ 10 \linebreak[0 ] \ 9 \linebreak[0 ] \ 11 \linebreak[0 ] \ 7 \linebreak[0 ] \ 6 \linebreak[0 ] \ 8 \linebreak[0 ] \ { \bf 12 } \linebreak[0 ] \ 13 \linebreak[0 ] \ 5 \linebreak[0 ] \ 3 \linebreak[0 ] \ 4 \linebreak[0 ]
\ 2 \linebreak[0 ] \ 18 \linebreak[0 ] \ 19 \linebreak[0 ] \ 1 \linebreak[0 ] \ 21 $ ] .
the number of involutions containing @xmath1 exactly once of length @xmath6 having @xmath107 fixed points with @xmath498 is the ballot number @xmath499 .
moreover , the number of involutions containing @xmath1 exactly once of length @xmath6 is @xmath500 .
we immediately deduce this result from bijection @xmath466 of theorem [ bij132once-132 ] and corollary [ nb132ptfix ] .
in fact , the number of involutions containing @xmath1 once of length @xmath6 is either the number of @xmath1-avoiding involutions of length @xmath468 if @xmath6 is odd or the number of @xmath1-avoiding involutions of length @xmath468 having more than one fixed point if @xmath6 is even . of course
, some of the following results can immediately be obtained from bijection @xmath466 of theorem [ bij132once-132 ] and results of section [ sec2 ] .
let @xmath501 denote the number of involutions in @xmath12 such containing @xmath1 exactly once , and let @xmath502 be the corresponding generating function .
the following proposition is the base of all the other results in this section , which holds immediately from definitions .
[ prom2 ] let @xmath106 an involution in @xmath0 such that contains @xmath1 exactly once , and let @xmath72 .
then holds either 1 . or
@xmath428 ; 2 . or @xmath503 where @xmath504 , @xmath505 and @xmath110 avoids @xmath1 .
3 . or @xmath506 where @xmath507 , @xmath508 and @xmath509 . another approach to find the generating function of involutions in @xmath0 containing @xmath1 exactly once is by use proposition [ prom2 ] .
[ thcc1 ] let @xmath86 be the generating function for the catalan numbers ; then @xmath510 according to proposition [ prom2 ] with terms of generating functions we get the following : the first part of the proposition yields @xmath511 , the second part of the proposition yields @xmath512 , and the third part of the proposition gives @xmath513 .
hence @xmath514 [ excc1 ] from proposition [ prom2 ] it is easy to see that , the number of the involutions in @xmath515 and containing @xmath1 exactly once is @xmath516 for @xmath6 odd , otherwise is @xmath115 . also , @xmath517 the @xmath518th fibonacci number , @xmath519}$ ] .
again , the case of varying @xmath3 is more interesting . as an extension of example [ excc1 ]
let us consider the case @xmath440 .
[ thcc2 ] for all @xmath45 , @xmath520 proposition [ prom2 ] with use the generating function of permutations in @xmath521 given by @xmath159 , yields @xmath522 by use the relation @xmath523 we get that @xmath524 so induction on @xmath3 with example [ excc1 ] gives the theorem .
similarly , we obtain another case @xmath451 .
[ thcc22 ] for all @xmath525 , @xmath526.\ ] ] similarly as proof of theorem [ thcc2 ] with use the generating function for the number of @xmath228-avoiding permutations in @xmath69 is given by @xmath159 ( see @xcite ) , we obtain that @xmath527 and by induction with @xmath528 ( it is easy to see ) the theorem holds .
theorem [ thcc22 ] yields @xmath529 the @xmath530th fibonacci number for all @xmath97 .
[ excc2 ] proposition [ prom2 ] yields , @xmath531 for all @xmath88 , and @xmath532}-1 $ ] for all @xmath88 . once again , the case of varying @xmath3 is more interesting . as an extension of example [ excc2 ]
let us consider the case @xmath452 .
[ thcc3 ] for all @xmath525 , @xmath533.\ ] ] similarly as proof theorem [ thcc2 ] we have that @xmath534 so by using theorem [ thcc2 ] the theorem holds . more generally , we present an explicit expression when @xmath535 $ ] as follows .
[ thcc4 ] for @xmath536 , @xmath537 , @xmath538}(x)=\frac{r_d(x^2)}{1-xr_d(x^2)}\left [ x^2r_{k - d-1}(x^2)+\frac{x^2(r_{k - d-1}(x^2)-r_{d-1}(x^2))}{u_{k - d}\left(\frac{1}{2x}\right)}\sum_{j=1}^{k - d-2 } u_j\left(\frac{1}{2x}\right ) \right].\ ] ] according to proposition [ prom2 ] in terms of generating functions we get the following . in first case @xmath539}(x)$ ] . in the third case ,
if @xmath110 contains @xmath540 then @xmath106 contains @xmath312 $ ] which is a contradiction , we get that @xmath110 avoids @xmath540 , hence @xmath541 . finally ,
in the second case , let us observe two subcases @xmath111 contains @xmath316 or avoids @xmath316 ; so by use the same argument of the third case we get @xmath542}(x)-j_{12\dots(k - d)}(x)).\ ] ] therefore , if we add all these cases we get @xmath543}(x)$ ] .
hence , by theorem [ thcc2 ] this theorem holds .
let @xmath544 denote the number of involutions in @xmath0 such containing @xmath1 exactly once and containing @xmath2 exactly @xmath29 times .
let @xmath545 be the corresponding generating function .
let us start be the following result . | we study generating functions for the number of involutions in @xmath0 avoiding ( or containing once ) @xmath1 , and avoiding ( or containing once ) an arbitrary permutation @xmath2 on @xmath3 letters . in several interesting cases
the generating function depends only on @xmath3 and is expressed via chebyshev polynomials of the second kind . in particular , we establish that involutions avoiding both @xmath1 and @xmath4 have the same enumerative formula according to the length than involutions avoiding both @xmath1 and any _ double - wedge pattern _ possibly followed by fixed points of total length @xmath3 .
many results are also shown with a combinatorial point of view . |
in many protocols of quantum information processing , it is crucial to transmit the quantum state of qubits with high fidelity @xcite .
while various schemes of quantum state transfer ( qst ) were proposed and demonstrated experimentally @xcite for the optical system even with atom ensemble @xcite , people have tried the best to implement this task based on the solid - state systems , which are believed as the best candidates for the scalable quantum computing @xcite .
it was recognized that the quantum spin chain @xcite and the bloch electron system @xcite with the artificial nearest neighbor ( nn ) couplings can be used as quantum data buses to transfer quantum information perfectly . in ref .
@xcite , we discovered that the spectrum - parity matching is responsible for the perfectness of most protocols of qst .
however , these solid - state based schemes for implementing quantum data bus are too artificial with very specially designed nn couplings and only single - particle cases were considered .
though it is believed that such an engineered quantum system will be realized in the future experiments , at least , this difficulty should be overcome in theoretical aspect . to this end , we have tried to employ the higher - dimensional system or the complicated quantum network , such as the spin ladder @xcite . though the spin ladder is not an ideal medium serving as a perfect data bus , it shows that the long - distance qst is possible via the multi - exciton system due to the existence of the spin gap , which is much larger than the coupling strength connecting two separated qubits to the two ends of the ladder respectively .
a medium can be a robust quantum data bus if the fidelity of transferring a quantum state between two sites attached to it approaches unity at a finite temperature .
these consideration motivates us to use more natural energy gapped materials to act as a quantum data bus . in this paper
, we study the function of quantum data bus for the conducting polymers ( polyacetylene ) , modeled as a peierls distorted tight - binding chain ( pdtbc ) or the su - schrieffer - heeger ( ssh ) model @xcite .
it is well known that , taking the coupling between electron and photon into account , such a system exhibits an instability against the lattice distortion , which induces an energy gap for the spectrum in half - filled case as a result of dimerization .
therefore , the ssh chain is a good candidate of quantum data bus for fermion , which is just similar to the spin ladder for spin @xcite .
furthermore , such a practical system allows us to investigate the case with larger particle number by performing analytical analysis for the two distant separated probe points @xmath2 since the ssh model can be solved exactly .
being ( b ) odd and ( c ) even . , width=6 ] as a protocol for the quantum communication with always - on interaction , two points @xmath0 and @xmath1 are connected to two sites of the ssh chain ( see the fig .
1(a ) ) . for the half - filled case ,
when the connections of @xmath0 and @xmath3 with ssh chain are switched off , the ground states are two - fold degenerate due to the vanishing on - site chemical potential . when the connections switch on , the two degenerate levels split while the eigenstates remain nearly half - filled for the subsystem @xmath2 due to the energy gap of the ssh chain .
it leads to the effective interaction between @xmath0 and @xmath1 .
thus the level spacing of two lowest eigenstates directly corresponds to the effective hopping integral between @xmath0 and @xmath1 and determines the validity of quantum state transfer via such a medium . generally ,
if the energy gap is large enough , an effective hamiltonian @xmath4 of these two sites is induced by this quantum data bus to perform perfect qst at low temperature .
thus the validity of @xmath4 and the behavior of the effective hopping integral ( level spacing of two lowest eigenstates ) as the distance between @xmath5 and @xmath1 increases are crucial for the quality of a quantum data bus . to demonstrate the properties of our scheme , we compare the eigenstates of @xmath6 with those reduced density matrices from the ground and the first excited states of the total system .
* * * * we find that the peierls energy gap of the ssh chain plays an important role in protecting the robustness of the qst .
the main conclusions are achieved both with the analytical study and numerical calculation .
this paper is organized as follows . in sec .
ii , the model setup of our protocol and the spectrum of ssh chain are introduced . as a reasonable approximation up to second order , the effective hamiltonian @xmath4 with respect to the two separated points @xmath0 and @xmath1
is deduced by using the frhlich transformation @xcite to remove the degree of freedom of ssh model in sec .
iii . in sec .
iv , the density matrices of qubits @xmath0 and @xmath1 for the ground and the first excited states of the whole system are calculated to demonstrate the quantum entanglement of them over a long distance .
the qst scheme is also investigated numerically in the region of the crossover between two types of dimerization .
it shows that , at this transition point , the qst becomes fastest , but with the similar lost of fidelity to what has been discussed in the ref . @xcite . in sec.v , we discuss the quantum decoherence problem due to the quasi - excitations of the ssh chain at finite temperature .
the summary and remarks are presented finally in sec .
in a polyacetylene , @xmath7-electrons , which are localized between the two bonded nuclei , connect one @xmath8 atom with the neighboring two @xmath8 and one @xmath9 atoms to form an elementary linear configuration ; @xmath10-electrons , which are less localized , behave as bloch electrons .
such a system can be treated by using tight - binding approximation and @xmath10-electrons can be the candidate as a carrier of quantum information @xcite . here , we also consider the coupling between electrons and dispersionless phonons represented by local oscillators .
let @xmath11 be the displacement of the @xmath12th @xmath13 ( the unit of a polyacetylene ) from the equilibrium position .
then , the tight - binding hamiltonian reads as @xmath14where @xmath15\]]describes the lattice motion . here , @xmath16 is the hopping integral and @xmath17 is the electron - lattice coupling constant .
@xmath18 denotes the canonical momentum operator conjugate to @xmath11 and @xmath19 is the effective mass of the unit of a polyacetylene ( ch ) .
@xmath20 is the spring constant due to @xmath7-electrons , and @xmath21 ( @xmath22 ) is annihilation ( creation ) operator of @xmath10-electrons on site @xmath23 with spin @xmath24 .
this hamiltonian is proposed by su , schrieffer and heeger @xcite to describe the dimerization phenomenon of the polyacetylene in association with its conductivity . in our scheme
, we consider the qst between sites @xmath0 and @xmath1 connected to the two sites of the ssh chain .
our purpose is to transfer a qubit state , which is in a superposition state of electron spin up and spin down , through the hopping of an polarized electron from the site @xmath0 to @xmath1 ( see the fig .
since the hamiltonian ( [ hssh ] ) does not contain spin dependent interaction , the spin polarization of every electron is conservative . in this sense
, the spatial motion of electrons along the chain will carry a spin state from one location to another .
thus , we can study the fidelity of spin state transfer by considering the fidelity of charge state transfer .
the connection of sites @xmath0 and @xmath1 to the ssh chain is described by thehamiltonian @xmath25where @xmath26 and @xmath27 ( @xmath28 and @xmath29 ) are annihilation ( creation ) operators of electrons on sites @xmath0 and @xmath1 with spin @xmath7 , while @xmath30 and @xmath31 denote the connecting sites in the ssh chain . and @xmath1.,width=6 ] in fig .
2 , we intuitively demonstrate how the above hamiltonian can lead to the effective hopping of electron between the additional sites @xmath0 and @xmath1 .
we will also show how the effective hopping integral depends on the distance between them .
the original energy degeneracy at @xmath32 of the subsystem formed by @xmath0 and @xmath1 will be removed by the switching on @xmath33 and then split into two sub - levels with the level spacing @xmath34 .
here , @xmath35 is the effective hopping integral , which can be obtained analytically as follows . to consider how a ssh chain can be used as a quantum data bus to transfer a quantum state , we first summarize the known results about diagonalization of the ssh hamiltonian .
the most important discovery is the prediction about the peierls transition happens when the polyacetylene is dimerized as @xmath36 .
it minimizes the energy of the one - dimensional electronic gas when the lattice vibrates slowly so that the born - oppenheirmer approximation can determine an instantaneous eigen energy of electron for a fixed lattice configuration .
after dimerization , the hamiltonian then reads @xmath37 \left ( c_{n+1,\sigma } ^{\dagger } c_{n,\sigma } + h.c.\right ) + e_{0}\text{,}\]]where @xmath38 denotes the distortion of the hopping integral and @xmath39 is a constant .
taking periodic boundary condition , it can be diagonalized as @xmath40with the dispersion relation @xmath41for the excitations described by the fermion operators @xmath42and @xmath43where @xmath44 is the number of @xmath13 in polyacetylene chain ( or the length of the chain ) , @xmath45 , @xmath46 , and @xmath47 \text{.}\]]the single - particle spectrum ( [ dispersion ] ) is illustrated in the fig .
2(a ) and the energy gap between the two bands is@xmath48 in next section , we will show that such an energy gap is necessarily required for a desirable robust quantum data bus for qst .
usually , as illustrated in fig .
2(b ) , the large energy gap above the lowest two levels can provide a kind of quantum protectfor the qst via the virtual excitations of the quantum data bus , which induce a long range hopping of electron between sites @xmath0 and @xmath1 in the low temperature .
the dense continuity of spectrum above the gap ensures that the strength of the effective hopping integral across @xmath0 and @xmath1 is so strong that the fast entanglement can be generated . in the following
, we will demonstrate that such a dimerized gap plays a same role for qst as the spin gap in spin ladder @xcite .
to deduce the effective hamiltonian about the indirect coupling between two attached sites @xmath0 and @xmath1 , we utilize the frhlich transformation , whose original approach was used successfully for the bcs theory of superconductivity .
the effective hamiltonian @xmath49 can be achieved by a unitary operator @xmath50 , where @xmath9 is the hamiltonian of the total system with the perturbation decomposition @xmath51 in the second order perturbation theory , we require the anti - hermitian operator @xmath52 obeys @xmath53 = 0\text{. } \label{s1}\]]thus the effective hamiltonian can be approximated as @xmath54 \text{.}\ ] ] there are two ways to connect the two sites @xmath0 and @xmath1 to the ssh chain , with ( without ) the mirror inversion symmetry as showed in fig .
1(b ) ( fig .
1(c ) ) . here , only the case with both @xmath30 and @xmath55 being odd is discussed .
the result can be used to all cases , because the system is invariant under exchanging @xmath0 and @xmath1 . as the solution of eq .
( [ s1 ] ) , @xmath52 can be expressed explicitly as @xmath56 , \notag\end{aligned}\]]which leads to the effective hamiltonian @xmath57with @xmath58=0.\]]here @xmath59denotes the effective hopping of electron between @xmath0 and @xmath1 with the strength @xmath60while @xmath61\end{aligned}\]]describes the dynamics of the data bus , where @xmath62 \text { , } \notag \\
b\left ( k , k^{\prime } \right ) & = & e^{-i\left ( k - k^{\prime } \right ) \frac{% l_{0}+1}{2}}\left [ 1-e^{i\left ( \theta _ { k}-\theta _ { k^{\prime } } \right ) } % \right ] \text{.}\end{aligned}\ ] ] straightforward calculation shows that , in the thermodynamic limit , @xmath63 , the hopping constant becomes @xmath64here , we have used the one - dimensional integral to replace the discrete sum as @xmath65 it needs to be pointed out that the effective hamiltonian @xmath4 works well only in the half - filled case .
when the lower quasi - band is filled fully , the next excitation should jump over the gap and then virtual excitations can effectively induce the effective coupling . to demonstrate the above result about the effective long range hopping induced by the ssh chain , we schematically sketch the virtual process in fig .
3 , where the quanta of data bus between the two sites are exchanged . if the energy gap of the pdtbc is much larger than the connection couplings , @xmath66 , the virtual transition is essentially second order process .
there are two kinds of virtual process for an initially half - filled state of the pdtbc .
the first one happens via the exchange of the electron in the upper quasi - band , while the second one is based on the exchange of the hole in the lower quasi - band .
we would like to remind that the two quasi - bands will become continuous in the thermodynamic limit .
nevertheless , there still exists indirect interaction between @xmath0 and @xmath1 as the effective results of the second order process . and the upper quasi - band and ( b ) holes between lower quasi - band and @xmath67 .
these processes lead to effective hopping of electron between @xmath0 and @xmath1 directly .
, width=8 ] now we consider the scheme using the ssh chain to transfer the quantum state via the above effective virtual excitation process .
let alice hold an electron with spin state @xmath68at the site @xmath0 , where @xmath69 ( @xmath70 ) denotes the spin up ( down ) state .
thus , the initial state @xmath71 of the total system is @xmath72here , @xmath73 @xmath74 denotes the empty state , i.e. , there is no electron at site @xmath0 ( both sites @xmath75 and @xmath1 ) . at the instant @xmath76the total system
evolves into a new factorized state @xmath77 or @xmath78to realize a perfect quantum swapping .
then bob at the site @xmath1 can receive an electron with spin state @xmath79 .
here,@xmath80 determines the characteristic time of quantum state transfer between two locations @xmath0 and @xmath1 , so the behavior of @xmath81 vs. @xmath55 is crucial for quantum information transfer .
it is desirable that the above scheme for qst based on the ssh virtual excitation can work well over a longer distance , namely , the effective hopping integral @xmath35 should not decay too fast as the distance of two points @xmath2 increases . according to eq .
( [ teff - ii ] ) , @xmath82 decays exponentially as @xmath55 increases .
however , @xmath83 closes to @xmath84 for small @xmath85 .
so @xmath86 does not decay rapidly for any finite @xmath55 . to demonstrate this ,
numerical simulation is performed .
the value of @xmath86 is computed from the energy spacing between the ground and first excited states of the total system by exact diagonalization method for the finite system with @xmath87 , @xmath88 , @xmath89 $ ] , @xmath90 , @xmath91 , and @xmath92 . the exact numerical result and the analytical result obtained from eq .
( [ teff - ii ] ) are plotted in fig .
it shows that they are in agreement with each other well and @xmath86 exhibits power law decay for finite @xmath55 .
it indicates that the characteristic time @xmath93 is proportional to the distance @xmath55 , which is crucial for the scalable quantum information processing . for the finite systems obtained by numerical exact diagonalization ( solid dot ) and the approximate analytical results from eq .
[ teff - ii ] ( line ) . here
only odd @xmath55 is plotted .
it shows that @xmath35 decays not so fast in this range of @xmath55 and the exact numerical and approximate analytical results are in agreement with each other very well.,width=7 ]
in the above section , we studied the behavior of @xmath82 changing as the distance @xmath55 increases .
it determines the speed of the qst via this ssh chain .
however , all the conclusions obtained should be based on the fact that the effective hamiltonian @xmath4 given by ( [ hab ] ) is valid . in this section ,
we investigate the validity of the effective hamiltonian @xmath4 by comparing the eigenstates of @xmath94 with the density matrix reduced from the ground and first excited states of the total system ( [ h ] ) .
notice that the hamiltonian ( [ h ] ) does not contain the spin - spin interaction term .
then we can regard the system as a spinless fermion system and then apply the result obtained feasibly for the spinless case to the original system .
therefore , in the following discussion we will ignore the spin degree of freedom for simplicity .
we define the states @xmath95 by @xmath96to denote the eigenstates of @xmath4 in the half - filled subspace , where @xmath97 , @xmath98 are spinless fermion operator at sites @xmath0 and @xmath1 . here
, @xmath99 @xmath100 is the number of particles in the anti - bonding ( bonding ) state .
then we calculate the fidelity of eigenstate defined by @xmath101 or @xmath102obviously , it can be described as the expectation value of the effective density matrix @xmath103 in the exact eigenstates @xmath104 of the total system with the reduced density matrix @xmath105 . here
, @xmath106 denotes the eigenstates of ( [ h ] ) with @xmath107 , while @xmath108 denotes the groundstate ( @xmath109 ) and excited eigenstates ( @xmath110 ) of the half - filled ssh chain .
state @xmath111 is the ground or first excited states of the whole system ( the hamiltonian ( [ h ] ) with @xmath112 ) , which are spanned by the states possessing the same parity as the states @xmath113 . on the other hand , it is known that the eigenstates of @xmath4 with unity concurrence are maximally entangled states @xcite .
thus by calculating the mode concurrences of two sites @xmath0 and @xmath1 for the ground and first excited states of the total system ( [ h ] ) , one can verify whether the exact solutions can be well approximated by that from the effective hamiltonian .
if fidelities or concurrences is near unity , it can be concluded that the effective hamiltonian @xmath4 is indeed valid for the perfect qst . in the following discussion , we try to explicitly express the exact eigenstates @xmath111 of the total system with the help of the gellmann - low theorem @xcite . in quantum field theory
, this theorem is used successfully to obtain the ground state of the interaction field from the ground state of the free field .
the exact eigenstates can be expressed with the input field @xmath114 as @xmath115for @xmath116 . here , @xmath117is the time - evolution operator in the interaction picture , where @xmath118 is the time - ordering symbol and @xmath119 .
similarly , we have the output field@xmath120with these formal expressions for the exact eigenstates of the total system , we can calculate reduced density matrix with respect to the @xmath2-subsystem .
then the expectation value of the projection operators of @xmath121 in @xmath122 can be calculated as @xmath123where@xmath124 \right\ } \text{,}\]]and @xmath125 notice that the above equation can not be calculated directly , because there is not an explicit expression for @xmath111 with respect to the well known basis vectors @xmath126 .
to get rid of this difficulty , we use the normalization condition @xmath127 or @xmath128then the eq .
( [ pp ] ) is rewritten as @xmath129 if the connection between sites @xmath2 and ssh chain is switched off , i.e. , the hopping integral in @xmath4 vanishes , states @xmath130 and @xmath131 are the eigenstates corresponding to vanishing eigen energies .
thus we can expand @xmath132 , which equals to unity when @xmath4 is exact , as the series of @xmath33 .
the dyson expansion of@xmath133 also gives the numerator and the denominator in the form of the series of @xmath33 .
comparing the terms of each order with respect to small coupling @xmath33 , the zero order of @xmath134 ( @xmath135 ) is unity , while the first order of @xmath134 ( @xmath136)is zero as a result of the particle number conservation , and second order of @xmath134 ( @xmath137 ) is non - zero and can be expressed as @xmath138 \text{. } \label{p2}\]]it approximately describes the difference between the eigenstates of @xmath4 and those reduced states from the exact eigenstates of the total system . in order to clearly characterize our model by the correction @xmath139
, we have the upper bound of @xmath139 directly from ( [ p2 ] ) as @xmath140 in the thermodynamic limit , this upper bound becomes @xmath141this inequality provides us a necessary condition for the validity of the effective hamiltonian @xmath4 .
actually , the necessary condition is @xmath142 , which requires the validity condition @xmath143on the other hand , the sufficient condition is @xmath144 for this scheme .
then , if we take @xmath145 , it should satisfy the sufficient and necessary conditions under which the effective hamiltonian can work well . in this paper ,
we take @xmath146 for the numerical calculation to demonstrate the validity of this scheme . furthermore
, the behavior of @xmath132 around the point @xmath147 , in which region the above condition ( [ valid cond ] ) is violated , is also investigated by numerical method .
it shows that @xmath134 deviate from unity within this region .
we will discuss this problem in detail in the following .
moreover , the mode concurrence between @xmath0 and @xmath1 for the state @xmath122 can be gained from @xmath134 .
it could also be used to characterize the validity of the effective hamiltonian @xmath6 .
the mode concurrence for the state @xmath122 is @xmath148 where @xmath149 , @xmath150 , and @xmath151 denotes the average value for an arbitrary operator @xmath152 ( @xcite ) . since states @xmath104 and @xmath153 , possess the same symmetry with respect to the points @xmath0 and @xmath1 , we have @xmath154similarly , we have @xmath155 \text{% { } } \]]under the validity condition ( [ valid cond ] ) , @xmath134 is near unity and @xmath156 should be larger than zero .
so we have mode entanglement @xmath157obviously , when the effective hamiltonian @xmath4 is valid , the sites @xmath0 and @xmath1 are maximally entangled for the ground and first excited states of the whole system . .@xmath158 and @xmath159 for the finite system with @xmath87 , @xmath88 , @xmath160 $ ] , @xmath161 , @xmath91 , and @xmath92 obtained by numerical exact diagonalization and from the approximately analytical expression eq .
( [ p2 ] ) .
they are in agreement with each other very well , which shows that the effective hamiltonian @xmath4 can well describe the states of points @xmath0 and @xmath1 in the ground and first excited states of the total system . [ cols="^,^,^,^,^ " , ] from the eq .
( [ teff - ii ] ) about effective hopping integral @xmath35 , one can see that when @xmath162 @xmath163 @xmath164 @xmath165 @xmath166 , @xmath167 is the ground state of the total system , while @xmath168 is the ground state for @xmath169 , due to the sign of @xmath35 being positive or negative .
we denote the expectation value @xmath134 of the projection operator for the ground and first excited states as @xmath170 and @xmath171 , respectively .
to demonstrate the validity of the effective hamiltonian of the sites @xmath2 , numerical simulation is performed for the finite system .
we calculate @xmath170 and @xmath171 for the finite system with @xmath87 , @xmath88 , @xmath172 $ ] , @xmath90 , @xmath91 , and @xmath92 by numerical exact diagonalization and from the approximately analytical expression eq .
the results are listed in table 1 .
it shows that they are in agreement well and the single - particle ( half - filled ) eigenstates of the effective hamiltonian @xmath4 can well describe the states of points @xmath0 and @xmath1 in the ground and first excited state of the total system .
so @xmath94 * * * * is valid when temperature * * * * @xmath173**. * * in the above studies , it is found that the magnitude of distortion @xmath174 is crucial for the quantum data bus to perform qst and the speed of qst is sensitive to the dimerization .
there are two types of dimerization corresponding to @xmath175 .
it is interesting to investigate what happens near the transition point @xmath147 . in this region
the analytical results obtained above is no longer valid due to the vanishment of the gap .
numerical simulation should be employed to calculate @xmath176 , @xmath170 and @xmath171 , which characterize the property of qst .
the numerical results obtained by exact diagonalization for the systems with @xmath177 , @xmath178 , @xmath179 , @xmath180 , @xmath181 is plotted in fig .
it shows that when the systems approach to the critical point @xmath147 , the gap @xmath182 between the ground and first excited states has a sharp peak , while @xmath170 tends to @xmath183 , @xmath171 remains unitary approximately with only a very slight jump around the transition point .
it indicates that although qst becomes fast around the critical point , the fidelity decreases rapidly .
this phenomenon is very similar to that discovered by ref .
this fact also implies that the features of two probing sites @xmath0 and @xmath1 , such as the fidelity ( mode concurrence ) and the recurrent time @xmath184 , reveal the intrinsic nature of the quantum data bus , a quantum phase transition - like behavior of the ssh chain . between the ground and first excited states , @xmath185 , and @xmath186 of the system with ( a ) @xmath187 ; ( b ) @xmath188 , and @xmath189 obtained by numerical exact diagonalization as the function of @xmath190 .
when @xmath191 approaches to zero , the gap @xmath182 has a sharp peak , while @xmath170 tends to @xmath183 , @xmath171 remains unitary approximately .
it shows that although qst becomes fast , the fidelity decreases rapidly around @xmath192 .
this observation indicates that the features of two sites @xmath0 and @xmath1 reveals the transition of the data bus.,title="fig:",width=4 ] between the ground and first excited states , @xmath185 , and @xmath186 of the system with ( a ) @xmath187 ; ( b ) @xmath188 , and @xmath189 obtained by numerical exact diagonalization as the function of @xmath190 .
when @xmath191 approaches to zero , the gap @xmath182 has a sharp peak , while @xmath170 tends to @xmath183 , @xmath171 remains unitary approximately .
it shows that although qst becomes fast , the fidelity decreases rapidly around @xmath192 .
this observation indicates that the features of two sites @xmath0 and @xmath1 reveals the transition of the data bus.,title="fig:",width=4 ]
finally we consider the decoherence problem due to the couplings with auxiliary modes concerning phonon excitations . to this end , we need to calculate the effect of electron - vibration coupling in the conducting polymers , which is off - resonate to the electron motion .
the electron - vibration coupling gives rise to the disadvantage of the qst scheme based on ssh chain , though it does not dissipate the energy of the electronic subsystem .
now we revisit the role of the electron - vibration coupling hamiltonian ( hssh ) in the process of quantum information transfer based on the virtual excitation of the ssh model .
we notice that there is no direct interaction between the electron motion and the quasi - excitation of ssh model when the energy gap is much larger than the connection couplings since @xmath193=0 $ ] .
thus the main source of decoherence is due to electron - phonon couplings .
this will realize a typical quantum decoherence model for a two level system coupled to a bath of harmonic oscillators sun , decohernce - song . in order to analyze this decoherence problem more quantitatively
, we describe the single mode phonon by the perturbation @xmath194induced by the quantized vibration of the ssh chain , where @xmath195 is a constant and @xmath196 ( @xmath197 ) is the bosonic creation ( annihilation ) operator of the phonon excitation . in small distortion case , according to ( [ teff - ii ] ) the effective hamiltonian depends on the phonon excitation through @xmath198\]]where@xmath199@xmath200 then we can rewrite the total hamiltonian @xmath201 describing the quantum decoherence problem with @xmath202where the hamiltonian @xmath203 corresponds to @xmath4 and has been rewritten in terms of the quasi - spin operators @xmath204here , we have ignored the indices of the spin degree of freedom for avoiding confusion , because all results we obtain in the following are unrelated to spin .
obviously , the effective electron - phonon coupling @xmath205 can only lead to a phase shift in the spin qubit initially prepared in the quasi - spin states@xmath206it is also pointed out that the ground state ( an eigenstate of the quasi - spin @xmath207 ) of the effective hamiltonian is a mode entanglement state .
we consider the vibration mode initially prepared in a thermal equilibrium state @xmath208at the temperature @xmath209 , where the partition function is@xmath210and @xmath211 is the boltzman constant .
let the electron be initially in a pure state @xmath212 .
after a straightforward calculation we get the density matrix at time @xmath213 @xmath214and the corresponding reduced density matrix @xmath215 $ ] of @xmath2 can be obtained by tracing over the phonon modes .
the off - diagonal elements of @xmath216 can be given explicitly as @xmath217where @xmath218 \text{,}\]]is the so - called decoherence factor@xcite .
the time independent factor @xmath219depends on the temperature @xmath209 and the distance @xmath55 between @xmath0 and @xmath1
. the time dependent behavior of the decoherence factor is mainly determined by the oscillating factors @xmath220 .
it indicates that , by increasing @xmath55 and @xmath209 , @xmath221 becomes larger .
it leads to @xmath222 except for some special instants at @xmath223here , @xmath224 decays to a minimum @xmath225 @xmath226 as temperature increases .
this result shows that thermal excitation of the phonons will block a perfect qst to some extent and this effect becomes prominent as the distance between @xmath0 and @xmath1increases*. * to consider the extent of this influence about the qst scheme based on the ssh chain , we can calculate the fidelity or the purity @xmath227:$ ] @xmath228 \label{qqq}\]]with the ideal state@xmath229the eq .
( [ qqq ] ) shows the intrinsic relation between the purity and the decoherence factor .
actually the off - diagonal elements are responsible for the intrinsic physics of qst , because the time evolution driven by @xmath4 for @xmath0 and @xmath1 will stop if off - diagonal elements vanish .
we notice that the norms of off - diagonal elements is proportional to the decoherence factor @xmath230 . in order to show the influence of the electron - vibration coupling on the qst , 3d diagram of @xmath224
is plotted in fig .
6(a ) for the case with @xmath231 , @xmath232 , @xmath233 , @xmath234 , @xmath235 $ ] , and @xmath236 $ ] . in fig . 6(b ) and ( c ) , the cross sections of the 3d diagram for @xmath224 are plotted to demonstrate the behavior of the decoherence factor .
actually , there exist other modes of phonon excitations in the ssh chain .
but in the case of dimerization , the mode of phonon we described above dominates the quantum decoherence and thus it is reasonable that we only focus on the dimerization mode . as the function of the temperature and time for the case with @xmath237 , @xmath238 , @xmath239 , @xmath234 , @xmath240 $ ] , and @xmath241 $ ] .
( b ) , ( c ) the cross section of @xmath224 for different @xmath209 and @xmath55 .
it shows that thermal excitation of the phonons will block a perfect qst as @xmath209 or @xmath55 increases.,title="fig:",width=7 ] as the function of the temperature and time for the case with @xmath237 , @xmath238 , @xmath239 , @xmath234 , @xmath240 $ ] , and @xmath241 $ ] .
( b ) , ( c ) the cross section of @xmath224 for different @xmath209 and @xmath55 .
it shows that thermal excitation of the phonons will block a perfect qst as @xmath209 or @xmath55 increases.,title="fig:",width=4 ] as the function of the temperature and time for the case with @xmath237 , @xmath238 , @xmath239 , @xmath234 , @xmath240 $ ] , and @xmath241 $ ] .
( b ) , ( c ) the cross section of @xmath224 for different @xmath209 and @xmath55 .
it shows that thermal excitation of the phonons will block a perfect qst as @xmath209 or @xmath55 increases.,title="fig:",width=4 ]
in this paper we proposed a qst scheme based on a desirable quantum data bus , which can be implemented by a more practical system , the conducting polymers ( polyacetylene ) modeled as a ssh chain .
this system serves as a quantum data bus with the following advantages .
firstly , the strength of the induced effective hopping of a qubit between two distant sites does not decay rapidly over the distance between them , and thus it can realize a fast quantum entanglement to transfer quantum information with the always - on coupling and minimal control . secondly , there exists an energy gap to ensure the virtual excitation of the electrons in the ssh chain which induces the effective hopping of a qubit between two distant sites , but does not dissipate their energy .
this shows the similar results to that of the spin ladder @xcite .
thus a higher fidelity can be achieved for our qst scheme based on the virtual excitation of the ssh chain in lower temperature .
technologically , the more natural and less artificial designs with practical modulation about coupling constants for our protocol push the studies of qst based on real physical solid state system . furthermore , based on the analytical and numerical results
, it is found that the validity of @xmath4 strongly depends on the distortion @xmath174 of the ssh chain .
thus , as a measurable issue , it essentially reflects the intrinsic property of the medium itself . in this sense , such a scheme can also be employed to explore the intrinsic property of the quantum system .
these observations have some universality and may motivate us to investigate the functions of quantum data bus based on other real physical systems .
we acknowledge the support of the nsfc ( grant no . 90203018 , 10474104 , 10447133 ) , the knowledge innovation program ( kip ) of chinese academy of sciences , the national fundamental research program of china ( no .
2001cb309310 ) .
s. m. barnett , o hirota , and e andersson ( eds . ) , quantum communication , measurement and computing : the seventh international conference on quantum communication , measurement and computing international conference on quantum comm , springer verlag ( 2004 ) c.p .
sun , phys . rev . * a * 48 , 898 ( 1993 ) , in quantum - classical correspondence , ed . by d.h .
feng and b.l.hu 99 - 106 ( international press,1994 ) ; in quantum co - herence and decoherence , ed . by k. fujikawa and y.a.ono ,
331 - 334 , ( amsterdam : elsevier science press,1996 ) ; c.p .
sun , x.x .
yi and x.j .
liu , fortschr .
phys.*43 * , 585 ( 1995 ) . | we systematically study the transfer of quantum state of electron spin as the flying qubit along a half - filled peierls distorted tight - binding chain described by the su - schrieffer - heeger ( ssh ) model , which behaves as a quantum data bus .
this enables a novel physical mechanism for quantum communication with always - on interaction : the effective hopping of the spin carrier between sites @xmath0 and @xmath1 connected to two sites in this ssh chain can be induced by the quasi - excitations of the ssh model .
as we prove , it is the peierls energy gap of the ssh quasi - excitations that plays a crucial role to protect the robustness of the quantum state transfer process .
moreover , our observation also indicates that such a scheme can also be employed to explore the intrinsic property of the quantum system . |
the main reason why supernova remnants ( snrs ) are considered as a cosmic ray ( cr ) source is a simple argument about the energy required to sustain the galactic cosmic ray ( gcr ) population against loss by escape , nuclear interactions and ionization energy loss .
the mechanical energy input to the galaxy from each supernova ( sn ) is about @xmath1 erg so that with a rate of about one every 30 years the total mechanical power input from supernovae is of the order @xmath2 erg / s ( e.g. * ? ? ?
thus supernovae have enough power to drive the gcr acceleration if there exists a mechanism for channeling about 10% of the mechanical energy into relativistic particles .
the high velocity ejecta produced in the supernova explosion interacts with the ambient medium to produce a system of strong shocks .
the shocks in turn can pick up a few particles from the plasma flowing into the shock fronts and accelerate them to high energies . the only theory of particle acceleration
which at present is sufficiently well developed and specific to allow quantitative model calculations is diffusive acceleration applied to the strong shocks associated with snrs ( e.g. see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* for review ) .
considerable efforts have been made during the last years to empirically confirm the theoretical expectation that the main part of gcrs indeed originates in snrs.theoretically progress in the solution of this problem has been due to the development of the kinetic nonlinear theory of diffusive shock acceleration @xcite .
the theory includes all the most relevant physical factors , essential for snr evolution and cr acceleration , and it is able to make quantitative predictions of the expected properties of crs produced in snrs and their nonthermal radiation .
detail information about high - energy cr population in snrs can be obtained from the observations of the nonthermal emission produced by accelerated crs in snrs .
the electron cr component is evident in a wide wave length range by the radiation that they produce in snrs , from radio to @xmath3-ray emission , whereas in the case of the nuclear cr component a @xmath3-ray detection is the only possibility to find it .
if this nuclear component is strongly enhanced inside snrs then through inelastic nuclear collisions , leading to pion production and subsequent decay , @xmath3-rays will be produced at the detectable level .
high - energy @xmath3-rays can also be produced by cr electrons due to the inverse compton ( ic ) scattering of the background photons .
therefore it is not so obvious which cr component ( nuclear or electron ) produces tev emission , detected from cassiopeia a ( cas a ) @xcite , rxj1713.7 - 3946 @xcite and vela jr @xcite . for this determination
the snr magnetic field value plays the crucial role .
the relative role of electrons in @xmath3-ray production is low if snr magnetic field is strongly amplified .
in addition , if the magnetic fields in snrs are considerably larger than the typical interstellar value , then snrs are able to generate a power - law spectrum of accelerated crs at least up to the knee energy @xcite .
in fact , there is strong evidence that the actual magnetic field in snrs is indeed intrinsically enhanced : the analysis of the nonthermal emission of young snrs show that all of them have amplified fields @xcite . required strength of the magnetic field , that is significantly higher than the interstellar medium ( ism ) field , has to be attributed to nonlinear field amplification at the sn shock by cr acceleration itself , in fact by the dominant nuclear cr component @xcite .
the application of kinetic theory to individual snrs ( see * ? ? ? * for a review ) has demonstrated its power in explaining the observed snr properties and in predicting new effects like the extent of magnetic field amplification , leading to the concentration of the highest - energy electrons in a very thin shell just behind the shock .
the theory also predicted the enhancement of the high - energy galactic @xmath3-ray background radiation due to the contribution of crs confined in parent snrs , that was partly confirmed by the measured background tev - emission . in order to perform the detail comparison of theoretical expectation with the experiment one needs sufficient number of individual snrs with known values of relevant physical parameters such as the age , distance , ism density .
unfortunately the number of such snrs are very limited , especially thous which are seen in all wavelength from radio to @xmath3-ray .
it is why every new experimentally established property of nonthermal emission of snr represents considerable interest for theoretical analysis .
therefore we start this review with a brief consideration of three snrs for which essentially new information about the efficiency of cr production was recently obtained
. then we shall consider the correspondence of the interior magnetic field values extracted from the analysis of the experimental data with theoretical expectation .
finally , we discuss the maximum energy of crs which can be achieved during their acceleration in snrs , taking into account realistic values of magnetic field in snrs .
experimental and theoretical study of young galactic snrs , tycho s snr , rx j1713.7 - 3946 and sn 1987a , performed recently provides very useful information about cr acceleration by sn shocks . therefore below we
briefly analyse these results .
the kinetic theory has been applied in detail to tycho s snr , the result of a type ia sn explosion in a ( roughly ) uniform interstellar medium ( ism ) , in order to compare the results with existing data , using a stellar ejecta mass @xmath4 , distance @xmath5 kpc , and ism number density @xmath6 h - atoms @xmath7 @xcite . for these parameters
a total hydrodynamic explosion energy @xmath8 erg was derived to fit the observed size @xmath9 and expansion speed @xmath10 .
a rather high downstream magnetic field strength @xmath11 @xmath12 g is needed to reproduce the observed steep and concave radio spectrum ( see also * ? ? ? * ) and to ensure a smooth cutoff of the synchrotron emission in the x - ray region .
overall , very good consistency of the predictions of the nonlinear theory with the existing observational data was achieved .
using chandra x - ray observations @xcite have recently estimated the ratio between the radius @xmath13 of the contact discontinuity ( cd ) , separating the swept - up ism and the ejecta material , and the radius @xmath14 of the forward shock .
the large mean value @xmath15 of this ratio was interpreted as evidence for efficient cr acceleration , which makes the medium between those two discontinuities more compressible .
this result was analysed within already performed kinetic nonlinear approach @xcite .
note , that the increase of the ratio @xmath16 due to the sn shock modification by efficiently accelerated crs was predicted for the case of kepler s snr by @xcite .
fig.[f1 ] and fig.[f2 ] show the calculations of shock and cd related quantities which were part of the earlier considerations @xcite , together with the azimuthally averaged experimental data available at the time .
the calculated shock as well as cd radii and speeds are shown as a function of time for the two different cases of interior magnetic field strengths @xmath17 @xmath12 g and @xmath18 @xmath12 g required by the value of the synchrotron frequency , where the spectrum hardens relative to the low - energy radio spectrum . according to fig.[f1]a tycho s snr
is nearing the adiabatic phase .
the softness of the observed low - energy radio spectrum
relative to a test particle spectrum required a proton injection rate @xmath19 .
this implies a significant nonlinear modification of the shock at the current age of @xmath20 yrs .
a larger magnetic field lowers the alfvnic mach number and therefore leads to a decrease of the shock compression ratio , as seen in fig.[f1]b .
the result is a total compression ratio @xmath21 and a subshock compression ratio @xmath22 for @xmath17 @xmath12 g . in turn
@xmath23 , @xmath24 , for @xmath18 @xmath12 g .
therefore , including cr acceleration at the outer blast wave , the calculated value of the ratio @xmath16 for @xmath18 @xmath12 g is slightly lower than for @xmath17 @xmath12 g . at the current epoch
we have @xmath25 which is somewhat lower than the value @xmath15 inferred from the chandra observations . projecting a highly structured shell onto the plane of
the sky tends to favor protruding parts of the shell .
therefore the average radius measured in projection is an overestimate of the true average radius . analyzing the amount of bias from the projection for the shock and cd radii @xcite
found a corrected `` true '' value @xmath15 which is lower than their measured `` projected average '' value @xmath26 .
one needs to make such a purely geometrical correction independently of the nature of the factors which produce the shell structures , if one measures the size of the quasispherical structured shell in projection instead of dealing with a 2-dimensional shell crossection . in the latter case
there would be no need for such a correction .
irrespective of this need to correct the observations of the cd position , if one starts from a spherically symmetric calculation of the cd radius as we do , one has to take into account that the actual cd is subject to the rayleigh - taylor ( r - t ) instability ( e.g. * ? ? ?
* ; * ? ? ?
* ) , and thus a correction is needed to compare such a 1-d calculation with observations of the cd . in the nonlinear regime
the instability leads to effective mixing of the ejecta and swept - up ism material with `` fingers '' of the ejecta on top of this mixing region .
the latter has a radius larger than @xmath27 .
the minimum correction to @xmath27 is the radial extent @xmath28 of the mixing region above @xmath27 .
we estimate @xmath29 , where @xmath30 is the longest finger size according to the numerical modeling of @xcite ( albeit without particle acceleration ) .
this leads to a rough estimate of the corrected cd radius @xmath31 .
as it is seen in fig.2 , the comparison of the corrected values @xmath32 , according to the earlier calculations @xcite , with this experimentally estimated value @xmath15 shows quite good agreement .
the agreement between theoretical solutions for effective particle acceleration with the measurements of the shock and discontinuity radii has to be considered as new evidence for strong nonthermal effects in tycho s snr .
new northern hemisphere tev detectors is expected to detect this source at tev - energies in , predominantly , hadronic @xmath3-rays at an energy flux level above @xmath33 . as a corollary
the detection of a tev signal is not only important by itself , but it is also crucial for the correct determination of all other key snr parameters .
rx j1713 - 3946 is a shell - type supernova remnant ( snr ) , located in the galactic plane , that was discovered in x - rays with rosat @xcite
. further study of this snr with the asca satellite by @xcite and later by @xcite have shown that the observable x - ray emission is entirely non - thermal , and this property was confirmed in later xmm - newton observations @xcite .
the radio emission of this snr is weak : only part of the shell could be detected in radio synchrotron emission up to now , with a poorly known spectral form @xcite .
rx j1713 - 3946 was also detected in very high energy @xmath3-rays with the cangaroo @xcite and h.e.s.s .
@xcite telescopes .
especially the latter observations show a clear shell structure at tev energies which correlates well with the asca contours .
the difficulty for the theoretical description is the fact that several key parameters of this source are either not known or poorly constrained .
this already concerns the distance and age of the object .
it was demonstrated that consistent description of this object is achieved @xcite following present consensus which puts the distance at 1 kpc , the age to about 1600 years and that the primary explosion must have been a type ii / ib sn event with a massive progenitor star whose mass loss in the main sequence phase created a hot wind bubble in a high - density environment .
the solution for the overall remnant dynamics then yields the value for the expansion velocity of the outer shock , given the total mechanical energy @xmath34 erg released in the explosion . to obtain a consistent solution for the broadband nonthermal emission
the injection rate @xmath19 and interior magnetic field @xmath35 g are needed .
the calculated overall broadband spectral energy distribution is displayed in fig.[f3 ] , together with the experimental data from atca at radio wavelengths , as estimated for the full remnant by @xcite , the x - ray data from asca , and the most recently measured tev gamma - ray spectrum from h.e.s.s .
the overall fit is impressive , noting that the choice of a few key parameters like @xmath36 , @xmath37 , @xmath38 in the theory allows a spectrum determination over more than 19 decades .
note , that calculated gamma - ray stectrum ( berezhko and vlk , 2006 ) agrees with the new h.e.s.s .
data ( aharonian et al . , 2007 ) even better than with earlier h.e.s.s .
measurements ( aharonian et al . , 2006a ) .
the properties of small scale structures of snr rx j1713.7 - 3946 seen in x - rays by @xcite , and in particular by @xcite , provide even stronger evidence that the magnetic field inside the snr is indeed considerably amplified . in order to find out
if the filamentary structure found by @xcite is indeed consistent with the efficient cr production projected radial profile @xmath39 calculated for the x - ray energy @xmath40 kev is shown in fig.[f4 ] .
the theory predicts the peak of the emission just behind the shock front with projected thickness @xmath41 that corresponds to an angular width @xmath42 .
this width is significantly thinner than the observed width of @xmath43 .
however , one should take into account the smoothing procedure with which the data were obtained / presented .
if we apply a point spread function of width @xmath44 , corresponding to the smoothed xmm data , to the calculated profile , we obtain the broadened profile , which is shown in fig.[f4 ] by the dashed curve . it has a width @xmath45 , or @xmath46 , and is consistent with the observational profile of @xcite .
in addition the projected radial profile ( dash - dotted line ) which corresponds to the test particle limit is presented in fig.[f4 ] . in this case
the proton injection rate was taken artificially so small @xmath47 , that accelerated crs do not produce any significant nonlinear effect , like shock modification and/or magnetic field amplification .
therefore interior magnetic field value @xmath48 @xmath12 g is much lower in this case . due to this fact synchrotron losses are not relevant on account of the low magnetic field , the peak of the emission is broader by a factor of five , inconsistent with the xmm observations .
the projected radial @xmath3-ray emission profile , calculated for the energy @xmath49 tev , is presented in fig.[f5 ] . as a result of the large radial gradient of the gas density and the cr distribution inside the snr , the theoretically predicted radial profile of the tev - emission
is concentrated within a thin shell of width @xmath50 . since the h.e.s.s .
instrument has a finite angular resolution we present in fig.[f4 ] also the modified radial profile smoothed with the gaussian point spread function with @xmath51 , that corresponds to the angular resolution @xmath52 .
as shown by fig.[f5 ] the smoothed radial profile of the tev - emission is much broader and what is most interesting it is characterised by a maximum to minimum intensity ratio @xmath53 .
such a ratio is consistent with the h.e.s.s .
measurement @xcite which obviously gives only a lower limit to the sharpness of the @xmath3-ray profile .
note that the broad radial profile of the tev - emission with @xmath54 measured by the h.e.s.s .
instrument is indirect evidence that the actual radial profile is significantly sharper , with a higher ratio @xmath55 .
we conclude that the present observational knowledge of snr rx j1713.7 - 3946 can be interpreted by a source which ultimately converts about 10% of the mechanical explosion energy into nuclear crs and that the observed high energy @xmath3-ray emission of snr rx j1713.7 - 3946 is of hadronic origin .
note , that @xcite have obtained the same type of conclusion suggesting a rather different @xmath3-ray spectrum formation scenario , compared with @xcite .
supernova 1987a occurred in the large magellanic cloud .
it has been extensively studied in all wavelengths from radio to @xmath3-ray .
the initial short outburst of radio emission @xcite is attributed to the synchrotron emission of electrons accelerated by the sn shock propagated in the free wind of presupernova star , which was the blue supergiant ( bsg ) @xcite .
after about 3 years radio emission was detected again @xcite as well as the monotonically increased x - ray emission @xcite .
this second increase of emission is attributed to the entrance of the outer sn shock into the thermalized bsg wind and then in the h ii region occupied by much more dense matter , consists of the swept up wind of red supergiant ( rsg ) progenitor star @xcite .
the advantage of sn 1987a for theoretical consideration is that the key parameter values are very well determined :
stellar ejecta mass @xmath56 , distance @xmath57 kpc , hydrodynamic explosion energy @xmath58 erg ( e.g. * ? ? ?
during an initial period the shell material has a broad distribution in velocity @xmath59 .
the fastest part of this ejecta distribution is described by a power law @xmath60 with @xmath61 .
the kinetic nonlinear model for cr acceleration in snrs has been in detail applied to sn 1987a , in order to compare its results with observational properties @xcite . a rather high downstream magnetic field strength @xmath62 mg is needed to reproduce the observed steep radio spectrum @xcite .
the required strength of the magnetic field have to be attributed to nonlinear field amplification at the sn shock by cr acceleration itself .
since for all the thoroughly studied young snrs , the ratio of magnetic field energy density @xmath63 in the upstream region of the shock precursor to the cr pressure @xmath64 is about @xmath65 @xcite and cr pressure in young snrs has a typical value @xmath66 , the upstream magnetic field @xmath67 was taken in the form @xmath68 calculated shock radius @xmath9 and speed @xmath10 shown in fig.[f6]a as a function of time are in satisfactory agreement with the values obtained on the basis of radio and x - ray measurements .
note that radio data compared with x - ray data gives larger shock size at any given time @xmath69 and calculated radius @xmath70 goes between these two sets of data . to fit the spectral shape of the observed radio emission a proton injection rate @xmath71 is needed .
this leads to a significant nonlinear modification of the shock : as it is seen in fig.[f6]b total shock compression ratio @xmath72 is essentially larger and a subshock compression ratio @xmath73 is lower than classical value 4 . since during the last 15 yrs the shock speed and the gas density are almost constant cr energy content grows roughly linearly with time ( see fig.[f6]c )
it gives the natural explanation of the linear increase of the radio emission detected during this evolutionary period .
strongly modified sn shock generates cr spectrum @xmath74 , which is very steep at momenta @xmath75 , with index @xmath76 .
cr electrons with such a spectrum produces synchrotron radioemission spectrum @xmath77 with spectral index @xmath78 , that very well corresponds to the experiment , as it is seen in fig.[f7 ] , where synchrotron energy spectra @xmath79 , calculated for five subsequent epoch together with the experimental data are presented . note that cr spectrum has a concave shape : it becomes flatter at higher momenta @xmath80 . as a consequence synchrotron spectrum @xmath81 is also concave as it is clearly seen in fig.[f7 ] at @xmath82 hz .
radio data reveal this property in good consistency with theoretical prediction .
strong downstream magnetic field @xmath83 mg , that corresponds to the upstream field @xmath84 mg ( see fig.[f7 ] ) , provides synchrotron cooling of electrons with momenta @xmath85 that in turn makes synchrotron spectrum at high frequencies @xmath86 hz very steep ( see fig.[f7 ] ) . concave shape of electrons continuously produced at the shock front together with their synchrotron cooling lead to a formation of two peaks in synchrotron energy spectrum @xmath79 .
the first one at @xmath87 hz corresponds to cr electron momentum @xmath88 above which synchrotron energy looses are relevant , whereas the second peak at @xmath89 hz corresponds to the maximum momentum @xmath90 of accelerated electrons . under this condition calculated synchrotron flux at frequency @xmath91 hz ,
which corresponds to the photon energy @xmath92 kev , is below the measured flux at the epochs @xmath93 d. since the contribution of the nonthermal radiation in the observed x - ray emission of sn 1987a is not very well known ( e.g. * ? ? ?
* ) , the observed x - ray flux has to be considered as the upper limit for the expected nonthermal emission . at early epoch @xmath94
d however the calculated flux exceeds the measured one ( see the curve , corresponding @xmath95 d in fig.[f7 ] ) .
this can be considered as indication , that the actual magnetic field @xmath96 is few times larger then given by the eq.(1 ) . at lower magnetic field @xmath97
mg synchrotron losses of high energy cr electrons are considerably smaller compared with the previous case .
due to this fact synchrotron spectra considerably exceeds at any given epoch the measured chandra flux @xcite .
therefore one can conclude , that the actual interior magnetic field strength is not lower than @xmath98 mg .
we see that quite a reasonable consistency of kinetic nonlinear theory with most of the observational data of sn 1987a is achieved , that gives the evidence of the efficient cr production leading to a strong shock modification and strong magnetic field amplification .
calculated @xmath3-ray integral energy flux at all energies is dominated by the @xmath99-decay component @xcite . at the current epoch the expected @xmath3-ray energy flux at tev - energies is about @xmath100 erg/(@xmath101s ) and during the next four years it is expected to grow by a factor of two .
the detection of high energy @xmath3-rays from sn 1987a , which could be done by instruments like h.e.s.s . , would provide very clear evidence for the efficient acceleration of nuclear crs .
detailed consideration performed within a frame of nonlinear kinetic theory demonstrates that the cr production efficiency in young snrs is consistent with requirements for the galactic cr sources .
recent experimental and theoretical studies of tycho s snr , rx j1713.7 - 3946 and sn 1987a provides new very strong evidence for that .
consideration performed by @xcite led him to the opposite conclusion : that tev @xmath3-ray emission from the mostly fairly young snrs ( sn 1006 , cas a and tycho s ) is unexpectedly low .
note , however , that this consideration is based on the simplified estimates , which does not take into account such an important aspect as the necessity of renormalization of cr production efficiency , predicted in spherical approach .
we have argued before that ion injection is quite efficient at the quasiparallel portions of the shock surface and it is expected to be strongly suppressed at the quasiperpendicular part of the shock ( see * ? ? ? * for details ) .
therefore one should renormalize the results for the nucleonic spectrum , calculated within the spherically symmetric model .
the lack of symmetry in the actual snr can be approximately taken into account by a renormalization factor @xmath102 , roughly @xmath103 to @xmath104 , which diminishes the nucleonic cr production efficiency , calculated in the spherical model , and all effects associated with it . for sn 1006 which appears to approximate this ideal case ,
efficient cr production is expected to arise within two polar regions , where the sn shock is quasi - parallel .
the calculated size of the efficient cr production regions which amounts to about 20% of the shock surface corresponds very well to the observed sizes of the bright x - ray synchrotron emission regions @xcite .
this implies a renormalization factor of 0.2 .
in addition , the distance to the object @xmath105 kpc used by @xcite for the case of sn 1006 is considerably lower than the most reliable determination of the distance @xmath106 kpc to sn 1006 @xcite . due to the above two factors the estimate of tev @xmath3-ray flux from sn 1006 , made by @xcite ,
should be reduced by a factor of 20 , so that it will be in agreement with h.e.s.s .
upper limit .
note , that the study of sn 1006 on the basis of kinetic nonlinear theory demonstrated that the lack of a tev signal that follows from the non - detection by the h.e.s.s .
instrument does not invalidate the theoretical picture which gives a consistent description of the nonthermal emission characteristics ; it rather implies a constraint on the ambient gas density @xmath107 @xmath7 @xcite . using xmm - newton x - ray observations @xcite
recently confirmed that the ism density is indeed low , @xmath107 @xmath7 .
the example of sn 1006 illustrates that real difficulty in reliable prediction of expected @xmath3-ray emission from snrs is not a theoretical one , but rather due to poorly known values of relevant astronomical parameters , like distance to the object and ambient ism density .
> from theoretical consideration @xcite it becomes clear that in the case of sn 1006 and in other similar cases @xmath3-ray measurements will at the same time give a reliable estimate of the ambient ism density .
similar situation is for the two other cases of cas a and tycho s snr , considered by @xcite : renormalization of his estimates , made in spherical approach , and the adoption the most reliable astronomical parameters provide the value of the expected @xmath3-ray flux consistent with the existing measurements ( or upper limits ) ( see * ? ? ?
all young snrs ( including three above cases ) to which nonlinear kinetic acceleration theory has been successfully applied up to now exhibit the strongly amplified magnetic field as result of very efficient acceleration of nuclear crs @xcite .
magnetic field extracted from the spectral properties of the overall synchrotron emission and from the observed filamentary structure of the nonthermal x - ray emission agree very well and their value can be expressed in the form @xmath108 where @xmath64 is cr pressure at the shock front , @xmath109 is the upstream field , which is already amplified and therefore considerably exceeds magnetic field value @xmath110 @xmath12 g in the ambient ism .
let us consider the consistency of such a field strength with the plasma physics .
several attempts have been undertaken to model highly nonlinear turbulent magnetic field , created by accelerated crs upstream of the strong shock @xcite .
two aspects of this process are of significant importance : the amount of magnetic field amplification and the extent of of gas heating due to the turbulence dissipation . unfortunately up to now it is not clear in which proportions cr working on the gas in the precursor region is divided between gas heating and magnetic field amplification .
therefore we restrict ourselves here by a simple estimates in order to demonstrate that considerable magnetic field amplification @xmath109 , consistent with eq.(2 ) is indeed possible upstream of the strong shock .
there are two different mechanisms operating upstream of the fast super - alfvnic shock and leading to the amplification of the chaotic magnetic field .
the first one is related with the resonant interaction of crs with the alfvn waves , which due to the cr anisotropy triggers an instability , followed by the intense alfvn wave excitation @xcite .
the other one , recently proposed by @xcite , is non - resonant amplification of mhd perturbations due to lorentz force , created by cr current .
integral effect of resonant alfvn wave excitation can be written in the form @xcite @xmath111 if we suggest that the cr working @xmath112 goes on the excitation of the alfvn waves . here
@xmath113 is magnetic component of the alfvn wave energy density , @xmath114 is alfvnic speed , @xmath115 is the plasma flow speed directed along the @xmath116 axis to the shock situated at the point @xmath117 . for the purpose of estimates we suggest quasistationary condition and neglect the gas speed variation @xmath118 in the upstream region @xmath119 due to the shock modification by cr backreaction , taking @xmath120 . assuming that the far upstream field @xmath121 coincides with the ism field @xmath122 and performing the integration of the above equation we have magnetic field energy density at the shock front @xmath123 where @xmath124 is alfvnic shock mach number , @xmath125 is the energy density of the far upstream field .
since the strong shock provides efficient cr acceleration with @xmath126 the above expression shows that alfvn wave amplitude grows to a large value @xmath127 in the case of strong shock with mach number @xmath128 .
resonant magnetic field amplification should be supplemented by two effects : non - resonant instability and the alfvn speed variation @xmath129 across the precursor due to the increase of the mean effective field @xmath130 . in this case
( see * ? ? ?
* for details ) instead of ( 4 ) we have @xmath131 where @xmath132 is the mean field amplified non - resonantly . here
@xmath133 , @xmath134 is the maximum cr momentum .
substituting @xmath135 in the expression ( 5 ) we have @xmath136 since for all known young galactic snrs the shock speed is about @xmath137 , this expression shows , that existing mechanisms is able to amplify magnetic field within the precursor to the level , which considerably exceeds the required value ( 2 ) .
therefore we can suggest that the main amount of the energy given by the expression ( 5 ) in fact goes on the gas heating due to dissipation of mhd perturbations and only small fraction of it , given by ( 2 ) , indeed goes on the magnetic field amplification .
numerical simulation of considered process performed by @xcite confirmed , that the gas thermal energy an order of magnitude larger than the energy content of amplified magnetic field .
note that the process of magnetic field amplification is not included in kinetic nonlinear theory ( e.g. * ? ? ?
* ; * ? ? ?
it simply postulates the existence of already amplified far upstream field @xmath96 , which value is consistent with experimental requirements , given by eq.(2 ) .
the consistency of such a picture is provided by the fact , that the spectrum of crs produced by strong modified shock is very hard so that crs with highest energies have a largest contribution in their energy content .
these the most energetic crs produce field amplification on their spacial scale that is the precursor size .
therefore crs with lower energies already see the amplified field @xmath138 .
note also , that kinetic nonlinear model contains in the explicit form the suggestion , that the whole cr working goes on the gas heating presumably due to some kind of dissipation mechanism . since according to the above estimates ,
the magnetic field energy which can be potentially achieved in the precursor region is considerably larger than required energy , given by eq.(2 ) , it can be considered as indirect evidence for the existence of such dissipation processes , which efficiently converts the energy of magnetic field perturbations into the gas thermal energy .
the existence of nonadiabatic gas heating within the precursor plays an important role for the shock structure and dynamics .
it restricts the shock compression ratio by the value @xcite @xmath139 if the whole cr working @xmath140 goes on the gas heating . for the appropriate upstream magnetic field values ,
consistent with eq.(2 ) , it provides moderate shock modification with @xmath141 . in the opposite case ,
when within the precursor there are no any additional gas heating except adiabatic one , the expected shock compression ratio is @xcite @xmath142 where @xmath143 is the sonic mach number , @xmath144 is the far upstream sound speed . for the typical ism parameters this expression leads to much larger shock compression @xmath145 . in the case
when cr working in the upstream region goes on the gas heating , the expected gas thermal energy just ahead of the subshock is @xmath146 appropriate magnetic field value @xmath96 given by eq.(2 ) leads to a typical shock compression ratio @xmath147 , that according to eq.(8 ) corresponds to @xmath148 , that in turn according to eq.(10 ) provides gas thermal energy @xmath149 .
this value @xmath150 very well corresponds to what is expected from the expression ( 7 ) , if indeed main amount of magnetic field perturbation energy is dissipated .
we therefore conclude that kinetic nonlinear theory provides the consistent description at least at shock speeds @xmath151 km / s , that correspond to the most active phase of the snr evolution that is the beginning of the sedov phase .
the most important consequence of magnetic field amplification in snrs is the substantial increase of the maximum energy of crs @xmath152 , accelerated by sn shocks .
let us briefly analyse what value of @xmath153 can be expected taking into account the amplification of snrs magnetic field . according to theoretical consideration @xcite and the analysis of the observational data @xcite cr diffusion in the amplified magnetic field is close to the bohm limit .
therefore cutoff momentum of crs accelerated at any given snr evolutionary stage can be expressed in the form @xcite @xmath154,\ ] ] where @xmath155 is cr diffusion coefficient , corresponding to bohm limit ; @xmath156 is the charge number of cr nuclei .
the value of parameter @xmath157 depends on the shock expansion law @xmath70 and on the downstream gas flow radial profile @xmath158 .
note that the main factors which determines the value of cr cutoff momentum @xmath159 are : i ) finite shock size ; ii ) shock deceleration and iii ) cr adiabatic cooling in the expended downstream region @xmath160 .
these factors are more relevant compared with the time factor , studied by @xcite , that in turn prevents the complete shock modification due to cr backreaction @xcite . first consider the most simple case of uniform ism , that is typical for type ia sne . in such a case the main part of cr spectrum formed during the snr evolution is produced at the sedov phase . since the cr cutoff momentum at the sedov phase @xmath161 decreases with time @xmath69 the most energetic crs in the power law part of this spectrum
are produced at the very end of the free expansion phase .
therefore the maximum cr momentum @xmath162 depends on physical parameters as @xmath163 , where @xmath164 , @xmath165 and @xmath166 are mean ejecta speed , sweep up radius and sweep up time respectively .
taking into account the value of upstream magnetic field ( 2 ) and assuming @xmath167 , we have @xcite @xmath168 it has a strong dependence on the sn parameters @xmath169 and @xmath170 , but is only weakly dependent upon the ism density @xmath171 .
maximum momentum of protons , given by eq.(12 ) for @xmath172 erg , @xmath4 and @xmath173 @xmath7 is @xmath174 .
this required condition to reproduce the spectrum of gcrs up to the knee energy is fulfilled due to the magnetic field amplification . for cr iron nuclei @xmath175 that provides the possibility for the formation of gcr spectrum up to the energy @xmath0 ev inside snrs . since during the early free expansion snr evolutionary phase
the sn shock speed @xmath176 km / s is considerably higher than on the subsequent sedov phase ( @xmath177 km / s ) , one could expect the production of crs with energies @xmath178 essentially higher than produced on sedov phase @xcite .
let us briefly consider this possibility , which was studied by @xcite and by @xcite . during the free expansion phase the sn shock evolution
is determined by the velocity distribution of the fastest part of the ejecta which is described by a power law @xmath179 with @xmath180 to 12 @xcite .
the ejecta kinetic energy contained in the fraction with speed @xmath59 within the interval @xmath181 is @xmath182 .
the cr spectrum created by the strongly modified shock is quite hard : @xmath183 , @xmath184 . for the sake of a rough estimate
we assume that all the above ejecta energy goes into crs with maximum momenta @xmath185 , where the shock speed is roughly the ejecta speed @xmath186 .
this gives @xmath187 in the ultrarelativistic case .
taking into account that @xmath188 , we obtain @xmath189 since in the case of uniform ism @xmath190 we have @xmath191 if we suggest for the simplicity @xmath192 . even in the case of type ia sne with a typical value @xmath180 eq.(15 ) gives @xmath193 @xcite .
it means that the spectrum of crs with momenta @xmath178 accelerated during the free expansion phase is so steep , that it hardly plays any role for the formation of the gcr spectrum . in the case of type ii
/ ib sne , which are more numerous in our galaxy @xcite , during the initial snr evolutionary phase sn shock propagates through the progenitor star wind .
therefore in this case ambient gas density at the shock front varies as @xmath194 that gives @xmath195 . even at the lowest value @xmath196
the resulting cr spectrum is very steep , @xmath197 .
we therefore conclude , that crs accelerated during the free expansion phase do not play a role for gcr spectrum , except the case if the actual value of the parameter @xmath198 is as small as @xmath199 for type ia sne and @xmath200 for type ii / ib sne or the csm has more complicated structure than considered above . in fig.8
we present gcr energy spectra , expected in the galaxy , assuming that snrs are the main gcr source . compared with our previous calculations
@xcite much higher magnetic field in snr is used , which varies during the snr evolution according to eq.(2 ) . to obtain these spectra @xmath201 the corresponding source spectra @xmath202 , produced in snr during its evolution ,
were calculated for a typical set of relevant parameters ( @xmath172 erg , @xmath4 ,
@xmath173 @xmath7 ) , taking into account their steepening due to the gcr escape from the galaxy with a time scale @xmath203 , where @xmath204 is the rigidity , @xmath205 is the kinetic energy of cr particle @xcite . for a comparison with the experiment
we represent in fig.8 the data of only three experiments , atic-2 @xcite , jacee @xcite and kascade @xcite , which on our view are representative in the corresponding energy ranges .
one can see , that the theory in a satisfactory way fits the existing data up to the energy @xmath206 ev .
the essential exception is the measured in the recent atic-2 balloon experiment @xcite helium spectrum , which opposite to the theoretical expectation is noticeably flatter than the protons spectrum . note that the spectral shape of protons and helium measured in the previous atic-1 experiment at energies @xmath207 gev @xcite are very similar .
similar form of the spectra for different gcr nuclei at energies @xmath208 gev if observed could be considered as the indirect confirmation of gcr origin in snrs .
therefore the confirmation of atic results for protons and helium by other experiments with comparable or even better precision is very much needed .
another difficulty for the theory is the fact that to get a consistency with the observed gcr spectra the value @xmath209 is needed , which is a bit beyond the experimentally determined interval @xmath210 @xcite .
in addition , strong energy dependence of the gcr escape time @xmath211 makes it difficult to interpret the observed cr anisotropy , which is very low , less than one percent at energies @xmath212 ev @xcite .
here @xmath213 is total energy of cr particle .
total flux of gcrs escaping from the galaxy @xmath214 has energy dependence like cr source spectrum @xmath215 with @xmath216 .
therefore gcr anisotropy outside the region where cr sources are situated ( galactic disk volume ) increases with energy @xmath217 , like inverse residence time @xmath218 . if @xmath12 is as big as 0.6 the expected anisotropy at @xmath219 ev exceeds the level of 100% @xcite . since our observing position is not outside the source layer
, we should see a smaller gcr anisotropy @xcite .
in addition one can also suggest that in the region , occupied by cr sources , gcr diffusion coefficient is higher than in outer part of gcr confinement volume , due to enhanced level of turbulence produced by gcrs streaming away from the galaxy . in such a case
the gcr anisotropy in the source layer could be sufficiently small and consistent with the observations .
note that the theory predicts concave spectra at energies @xmath220 gev for individual cr species and there are some experimental evidences that the actual gcr spectra are indeed concave , eventhough one needs more precise measurements at @xmath221 gev in order to make more strict conclusion . at the same time calculated all particle spectrum has almost pure power law form up to the knee energy ( see fig.8 ) . according to fig.8 the knee in the observed all particles gcr spectrum has to be attributed to the maximum energy of protons , produced in snrs .
the steepening of the all particle gcr spectrum above the knee energy @xmath222 ev is a result of progressive depression of contribution of light cr nuclei with the energy increase .
such a scenario is confirmed by kascade experiment which shows relatively sharp cutoff spectra of gcr species at energies @xmath223 ev @xcite . at energy @xmath224 ev gcr spectrum
is dominated by iron nuclei contribution .
according to the analysis , performed by @xcite the contribution of galactic cr sources into the observed gcr spectrum dominates up to the energy @xmath225 ev above which it progressively goes down so that at energy @xmath226 ev the contribution of extragalactic sources becomes dominant .
at so - called second knee energy @xmath227 ev these two sources provide roughly equal contributions into gcr spectrum . as it is seen from fig.8
, calculations provide a good fit of the observed gcr up to the required energy @xmath228 ev , that is a strong argument for snrs to be considered as the main galactic cr source
. essentially different scenario of the formation of gcr spectrum and the first knee in particular was proposed and discussed by erlykin and wolfandale in a series of papers ( see * ? ? ?
* and references there ) .
based on the analysis of measurements performed by ground based installations they claim the existence of complex feature of gcr spectrum in the knee energy region .
this leads them to a conclusion that observed gcr spectrum consists of smooth background spectrum ( without the knee ) , produced by many cr sources , and superimposed single source ( presumably nearby snr ) contribution , that forms a bump on the smooth background spectrum usually interpreted as a ( first ) knee .
the contribution of the single nearby snr , leading to some peculiarity of gcr spectrum at the detectable level in the knee region , is indeed quite possible @xcite .
however , the existing measurements of cr composition demonstrate , that the knee is indeed peculiar point in gcr spectrum , where not only the shape of gcr spectrum but also their chemical composition undergoes a noticeable change , that is consistent with cr production in typical snrs , as described above .
detailed consideration performed within a frame of nonlinear kinetic theory demonstrates that the cr production efficiency in young snrs is consistent with requirements for the galactic cr sources .
recent experimental and theoretical studies of tycho s snr , rx j1713.7 - 3946 and sn 1987a provides new very strong evidence for that .
synchrotron radiation from young snrs provides evidence that efficient cr acceleration followed by the significant shock modification and magnetic field amplification takes place there .
magnetic field amplification leads to the considerable increase of maximum energy of crs , accelerated in snrs .
calculations performed within kinetic nonlinear theory demonstrate that the expected gcr spectrum , produced in snrs , in satisfactory way fits the existing gcr data up to the energy @xmath225 ev .
the first knee in the observed all particles gcr spectrum is attributed to the maximum energy of protons , produced in snrs .
the steepening of the all particle gcr spectrum above the knee energy @xmath222 ev is a result of progressive depression of contribution of light cr nuclei with the energy increase .
such a scenario is confirmed by kascade experiment which shows relatively sharp cutoff spectra of gcr species at energies @xmath223 ev @xcite . the nature of the whole gcr spectrum , which does not require any significant contribution of galactic cr sources other then snrs , can be the following .
gcr spectrum up to the energy @xmath229 ev is dominated by the contribution of galactic sources , whereas at @xmath230 ev gcrs are predominantly extragalactic .
such a picture is consistent with the analysis , presented here , and with the analysis of ultra high energy cr properties @xcite as well .
detail experimental study of gcr spectrum and composition could find whether indeed the transition from galactic to extragalactic component takes place in energy range @xmath231 ev .
due to high interior magnetic field in young snrs the @xmath99-decay @xmath3-rays generated by the nuclear cr component as a rule dominates over @xmath3-rays , generated by electron cr component , and calculated @xmath3-ray flux fits existing data .
new generation of ground - based @xmath3-ray detectors cangaroo iii , h.e.s.s . , magic , veritas will presumably measure significantly more accurate @xmath3-ray spectra from already detected snrs , like cas a , rxj1713.7 and tycho s snr , as well as from a number of new sources . in this regard considerable interest as a potential @xmath3-ray sources represent tycho s snr , sn 1987a and kepler s snr @xcite eventhough the expected @xmath3-ray fluxes from these objects only slightly above the sensitivity of modern tev @xmath3-ray instruments .
together with data , which soon will be obtained by space @xmath3-ray observatory glast , they will reproduce the full @xmath3-ray spectra of some number of individual snrs , as well as the galactic background spectrum .
although the existing data together with their theoretical interpretation presents a consistent picture of gcr origin , these new measurements will allow for detailed study of snrs and the cr production inside them .
this work has been supported by the russian foundation for rasic research ( grants 07 - 02 - 00221 , 06 - 02 - 96008 , 05 - 02 - 16412 ) .
the author thanks the organizers of the cospar meeting e1.4 `` new high - energy results on supernova remnants and pulsar wind nebula '' for the invitation to present this paper and for the financial support , prof . m.i.panasyuk and dr.a.d.panov for their kind presentation of atic-2 data and prof .
h.j.vlk for many useful discussions .
acero , a. , ballet , b. inner density of sn 1006 .
36th cospar sci .
assembly , beijing , abstracts p.1598 , 2006 .
aharonian , f. a. , akhperjanian , a. , barrio , j. , et al .
evidence for tev gamma ray emission from cassiopeia a. astron .
370 , 112 - 120 , 2001a .
aharonian , f.a . ,
akhperjanian , a.g.,bazer - bachi , a.r .
high - energy particle acceleration in the shell supernova remnant .
nature 432 , 75 - 77 , 2004 .
aharonian , f.a . ,
akhperjanian , a.g.,bazer - bachi , a.r .
et al . a detailed spectral and morphological study of the gamma - ray supernova remnant rx j1713.7 - 3946 with hess , astron .
449 , 223 - 242 , 2006a .
aharonian , f.a . ,
akhperjanian , a.g.,bazer - bachi , a.r .
first detection of a vhe gamma - ray spectral maximum from a cosmic source : hess discovery of the vela x nebula , astron .
448 , l43-l47 , 2006b .
aharonian , f.a . ,
akhperjanian , a.g.,bazer - bachi , a.r .
primary particle acceleration above 100 tev in the shell - type supernova remnant rxj1713.7 - 3946 , astron .
464 , 235 - 243 , 2007 .
ahn , h. s. , seo , e. s. , adams , j. h. et al .
the energy spectra of protons and helium measured with the atic experiment .
37 , 1950 - 1954 , 2006 .
aloisio , r. , berezinsky , v. , blasi .
a dip in the uhecr spectrum and the transition from galactic to extragalactic cosmic rays .
arxiv : astro - ph/0608219 , 2006 .
amato , e. , blasi , p. non - linear particle acceleration at non - relativistic shock waves in the presence of self - generated turbulence .
mon . not .
r. astron .
371 , 1251 - 1258 , 2006 .
antoni , t. , apel , w.d . ,
badea , a.f .
kascade measurements of energy spectra for elemental groups of cosmic rays : results and open problems .
24 , 1 - 25 , 2005 .
bell , a.r .
the acceleration of cosmic rays in shock fronts.i . mon . not .
182 , 147 - 156 , 1978 . bell , a. r. , lucek , s.g . cosmic ray acceleration to very high energy through the non - linear amplification by cosmic rays of the seed magnetic field .
mon . not .
321 , 433 - 438 , 2001 .
bell , a.r .
turbulent amplification of magnetic field and diffusive shock acceleration of cosmic rays .
mon . not .
353 , 550 - 558 , 2004 .
berezhko , e.g. , krymsky , g.f . acceleration of cosmic rays by shock waves .
soviet phys .- uspekhi 31 , 27 - 51 , 1988 .
berezhko , e.g. , yelshin , v.k . , ksenofontov , l.t .
cosmic ray acceleration in supernova remnants .
jetph 82 , 1 - 21 , 1996 .
berezhko , e.g. maximum energy of cosmic rays accelerated by supernova shocks .
astropart .
, 5 , 367 - 378 , 1996 .
berezhko , e.g. , vlk , h.j . kinetic theory of cosmic rays and gamma rays in supernova remnants .
i. uniform interstellar medium . astropart .
phys . 7 , 183 - 202 , 1997 .
berezhko , e.g. , ksenofontov , l.t .
composition of cosmic rays accelerated in supernova remnants .
jetph 89 , 391 - 403 , 1999 .
berezhko , e.g. , ellison , d.c . a simple model of nonlinear diffusive shock acceleration . astrophys
. j. 526 , 385 - 399 , 1999 .
berezhko , e. g. , ksenofontov , l. t. cosmic rays , radio and gamma - ray emission from the remnant of supernova 1987a .
astronomy letters 26 , 639 - 656 , 2000 .
berezhko , e.g. , vlk , h.j .
kinetic theory of cosmic ray and gamma - ray production in supernova remnants expanding into wind bubbles .
357 , 283 - 300 , 2000a .
berezhko , e.g. , vlk , h.j .
theory of synchrotron emission from supernova remnants .
427 , 525 - 536 , 2004 .
berezhko , e.g. gamma - ray astronomy and cosmic ray origin problem . adv .
space res . , 35 , 1031 - 1040 , 2005 berezhko , e.g. , vlk , h.j .
theory of cosmic ray production in the supernova remnant rx j1713.7 - 3946 .
451 , 981 - 990 , 2006 .
berezhko , e.g. , ksenofontov , l. t. , vlk , h.j .
gamma - ray emission expected from kepler s supernova remnant .
452 , 217 - 221 , 2006 .
berezhko , e. g. , ksenofontov , l. t. magnetic field in supernova remnant sn 1987a .
j. 650 , l59-l62 , 2006 .
berezhko , e. g. , vlk , h. j. spectrum of cosmic rays , produced in supernova remnants .
j. 661 , l175 - 178 , 2007 .
berezinsky , v.s . ,
bulanov , s.v . ,
ginzburg , v.l . , et al .
1990 , astrophysics of cosmic rays , north - holland publ .
blandford , r. , eichler , d. particle acceleration in astrophysical shocks : a theory of cosmic ray origin .
154 , 1 - 75 , 1987 .
cassam - chena , g. , decourchelle , a. , ballet , j. , et al .
xmm - newton observations of the supernova remnant rx j1713.7 - 3946 and its central source observations of snr rx j1713.7 - 3946 .
427 , 199 - 216 , 2004 .
chevalier , r.a .
self - similar solutions for the interaction of stellar ejecta with an external medium .
258 , 790 - 797 , 1982 .
chevalier , r. a. , fransson , c. circumstellar matter and the nature of the sn1987a progenitor star .
nature , 328 , 44 - 45 , 1987 .
chevalier , r.a . , blondin , j.m .
, emmering , r.t .
hydrodynamic instabilities in supernova remnants - self - similar driven waves .
j. 392 , 118 - 130 , 1992 .
chevalier , r. a. , dwarkadas , v. v. the presupernova h ii region around sn 1987a .
j. 452 , l45-l48 , 1995 .
decourchelle , a. , ellison , d.c . , ballet , j. thermal x - ray emission and cosmic ray production in young supernova remnants .
j. 543 , l57-l60 , 2000 .
drury , l. oc . ,
vlk , h. j. , berezhko , e. g.existence and interpretation of smooth cosmic - ray dominated shock structures in supernova remnants .
299 , 222 - 224 , 1995 .
drury , lo.c .
an introduction to the theory of diffusive shock acceleration of energetic particles in tenuous plasmas .
46 , 973 - 1027 , 1983 .
enomoto , r. , tanimori , t. , naito , t. , et al .
the acceleration of cosmic - ray protons in the supernova remnant rx j1713.7 - 3946 nature 416 , 823 - 826 , 2002 .
erlykin , a.d . ,
wolfendale , a.w . the origin of pev cosmic rays .
356 , l63-l65 , 2000 .
gaensler , b. m. , manchester , r. n. , staveley - smith , l. et al . the asymmetric radio remnant of sn 1987a .
j. 479 , 845 - 860 , 1997 gorenstein , p. , hughes , j. p. , tucker , w. h. astrophys .
j. 420 , l25-l28 , 1994 .
hasinger , g. , ashenbach , b. , trmper , j. the x - ray lightcurve of sn 1987a .
312 , l9-l12 , 1996 .
hillas , a. m. can diffusive shock acceleration in supernova remnant account for high - energy galactic cosmic rays ? .
g : nucl . part .
31 , r95-r131 , 2005 .
hiraga , j.s . ,
uchiyama , y. , takahshi , t. , aharonian , f.a .
spectral properties of nonthermal x - ray emission from the shell - type snr rx j1713.7 3946 as revealed by xmm - newton astron .
431 , 953 - 961 , 2005 .
jones , e.m . ,
smith , b.w .
& straka , w.c .
formation of supernova remnants - the pre - blast - wave phase .
j. 249 , 185 - 194 , 1981 .
jones , f.c . , ellison , d.c . the plasma physics of shock acceleration .
space sci . rev .
58 , 259 - 346 , 1991 .
koyama , k. , kinagasa k. , matsuzaki , k. , et al .
discovery of non - thermal x - rays from the northwest shell of the new snr rx j1713.7 - 3946 : the second sn 1006 ?
pasj 49 , l7-l11 , 1997 .
ksenofontov , l. t. , berezhko , e. g. and vlk , h. j. dependence of the gamma - ray emission from sn 1006 on the astronomical parameters .
443 , 973 - 980 , 2005 .
lagage , p. o. , cesarsky , c. j. the maximum energy of cosmic rays accelerated by supernova shocks .
125 , 249 - 257 , 1983 .
lazendic , j. s. , slane , p.o . ,
gaensler , b.m .
, et al . a high - resolution study of nonthermal radio and x - ray emission from supernova remnant g347.3 - 0.5 .
j. 602 , 271 - 285 , 2004 .
lucek , s.g . ,
bell , a.r .
non - linear amplification of a magnetic field driven by cosmic ray streaming .
mon . not .
314 , 65 - 74 , 2000 .
malkov , m.a . ,
drury , l.,oc .
nonlinear theory of diffusive acceleration of particles by shock waves .
64 , 429 - 481 , 2001 .
malkov , m. a. , diamond , p.h . , sagdeev , r.z . on the gamma - ray spectra radiated by protons accelerated in supernova remnant shocks near molecular clouds
: the case of supernova remnant rx j1713.7 - 3946 .
j. 624 , l37-l40 , 2005 .
manchester , r. n. , gaensler , b. m. , wheaton , v. c. , et al .
evolution of the radio remnant of sn 1987a : 1990 - 2001 .
pasa 19 , 207 - 221 , 2002 .
manchester , r. n. , gaensler , b. m. supernova 1987a in the large magellanic cloud .
iau circ . 7757 , 2001 .
mccray , r. supernova 1987a revisited . annual rev
31 , 175 - 216 , 1993 .
mckenzie , j.f . ,
vlk , j.h .
non - linear theory of cosmic ray shocks including selfgenerated alfvn waves .
116 , 191 - 200 , 1982 .
michael , e. , zhekov , s. , mccray , r. , et al .
the x - ray spectrum of supernova remnant 1987a astrophys .
j. 574 , 166 - 178 , 2002 .
muraishi , h. , tanimori , t. , yanagita , s. et al .
evidence for tev gamma - ray emission from the shell type snr rx j1713.7 - 3946 .
354 , l57-l61 , 2000 .
panov , a. d. , adams , j. h. , ahn , h. s. et al .
the results of atic-2 experiment for elemental spectra of cosmic rays , @xmath232arxiv : astro - ph/0612377@xmath2332006 .
parizot , e. , marcowith , a. , ballet , j. , gallant , y. a. observational constraints on energetic particle diffusion in young supernovae remnants : amplified magnetic field and maximum energy .
453 , 387 - 395 , 2006 .
park , s. zhekov , s. a. , burrows , d. n. et al . a chandra view of the morphological and spectral evolution of supernova remnant 1987a astrophys .
j. 610 , 275 - 284 , 2004 .
pelletier , g. , lemoine , m. , marcowith , a. turbulence and particle acceleration in collisionless supernovae remnant shocks .
i - anisotropic spectra solutions .
453 , 181 - 191 , 2006 .
pfeffermann , e. , aschenbach , b. 1996 , in roentgenstrahlung from the universe , ed .
zimmermann , j. trmper , & h. yorke ( mpe rep .
263 ; garching : mpe ) , 267 .
ptuskin , v.s .
, zirakashvili , v.n . on the spectrum of high - energy cosmic rays produced by supernova remnants in the presence of strong cosmic - ray streaming instability and wave dissipation .
429 , 755 - 765 , 2005 .
reynolds , s.p . , ellison , d.c .
electron acceleration in tycho s and kepler s supernova remnants - spectral evidence of fermi shock acceleration .
j. 399 , l75-l78 , 1992 .
reynolds , p.t . , akerlof , c.w . , cawley , m.f .
survey of candidate gamma - ray sources at tev energies using a high - resolution cerenkov imaging system - 1988 - 1991 .
j. 404 , 206 - 218 , 1993 .
slane , p. , gaensler , b.m . , dame , t.m .
nonthermal x - ray emission from the shell - type supernova remnant g347.3 - 0.5 .
j. 525 , 357 - 367 , 1999 .
staveley - smith , l. manchester , r. n. , kesteven , m. j. et al .
birth of a radio supernova remnant in supernova 1987a .
nature 355 , 147 - 149 , 1992 .
tammann , g.a .
, lfler , w. , schrder , a. the galactic supernova rate .
92 , 487 - 493 , 1994 .
tan , s.m . ,
gull , s.f .
the expansion of tycho s supernova remnant as determined by a new algorithm for comparing data .
mon . not .
216 , 949 - 970 , 1985 .
takahashi , y. , asakimori , k. , burnett , t.h .
et al . elemental abundance of high energy cosmic rays nuclear phys .
b ( proc . suppl . )
60 , 83 - 92 , 1998 .
turtle , a. j. , campbell - wilson , d. , bunton , j. d. et al . a prompt radio burst from supernova 1987a in the large magellanic cloud .
nature 327 , 38 - 40 , 1987 .
vladimirov , a. , ellison , d. c. , bykov , a. nonlinear diffusive shock acceleration with magnetic field amplification .
j. 652 , 1246 - 1258 , 2006 .
vlk , h.j . ,
berezhko , e.g. , ksenofontov , l.t . ,
rowell , g.p .
the high energy gamma - ray emission expected from tycho s supernova remnant .
396 , 649 - 656 , 2002 .
vlk , h.j . ,
berezhko , e.g. , ksenofontov , l.t .
variation of cosmic ray injection across supernova shocks .
409 , 563 - 571 , 2003 .
vlk , h.j .
, berezhko , e.g. , ksenofontov , l.t . magnetic field amplification in tycho and other shell type supernova remnants .
433 , 229 - 240 , 2005 .
vlk , h.j . ,
berezhko , e.g. , ksenofontov , l.t .
new evidence for strong nonthermal effects in tycho s snr . accepted in astrophys .
space sci .
uchiyama , y. , aharonian , f.a . & takahshi , t. fine - structure in the nonthermal x - ray emission of snr rx j1713.7 - 3946 revealed by chandra .
400 , 567 - 574 , 2003 .
wang , c .- y . ,
chevalier , r.a .
instabilities and clumping in type ia supernova remnants .
j. 549 , 1119 - 1134 , 2001 .
warren , j.s . ,
hughes , j.p .
badenes , c. et al .
cosmic - ray acceleration at the forward shock in tycho s supernova remnant : evidence from chandra x - ray observations .
j. 634,376 - 389 , 2005 .
winkler , p. f. , gupta , g. and long , k. s. the sn 1006 remnant : optical proper motions , deep imaging , distance , and brightness at maximum .
j. 585 , 324 - 335 , 2003 . , contact discontinuity radius @xmath13 , shock speed @xmath234 , and contact discontinuity speed @xmath235 , for tychos snr as functions of time , including particle acceleration ; ( b ) total shock ( @xmath236 ) and subshock ( @xmath237 ) compression ratios @xcite .
the _ dotted vertical line _ marks the current epoch .
the _ solid and dashed lines _ correspond to the internal magnetic field strength @xmath17 @xmath12 g and ( @xmath18 @xmath12 g ) , respectively .
the observed mean size and speed of the shock , as determined by radio measurements @xcite , are shown as well.,width=283 ] of the radii of the contact discontinuity and the forward shock of tycho s snr as a function of time @xcite .
solid and dashed lines correspond to the same two cases as in fig.[f1 ] .
the experimental point is taken from @xcite.,width=283 ] ev corresponds to @xmath99-decay @xmath3-ray emission , whereas the dashed and dash - dotted curves indicate the inverse compton ( ic ) and nonthermal bremsstrahlung ( nb ) emissions , respectively .
, width=510 ] kev from rx j1713.7 - 3946 @xcite .
the solid line corresponds to the high - injection model ; the dashed line again represents the above profile , but smoothed to the resolution of the xmm - newton data used in @xcite ; the dash - dotted line corresponds to the test - particle limit .
, width=283 ] -ray emission for @xmath3-ray energies
@xmath49 tev @xcite .
the calculated radial profile is given by the solid line ; the dashed line represents the calculated profile smoothed with a gaussian point spread function of the width @xmath238 . for purposes of presentation both profiles are normalised to their peak values in this figure . ,
width=283 ] , shock speed @xmath234 , gas density @xmath239 and upstream magnetic field @xmath96 at the current shock position ; ( b ) total shock ( @xmath236 ) and subshock ( @xmath237 ) compression ratios ; ( c ) kinetic energy of ejecta @xmath240 and accelerated cr energy content @xmath241 as functions of time for sn 1987a @xcite .
the observed radius of the sn shock , as determined by radio @xcite and x - ray measurements @xcite , are shown by circles and stars respectively .
the scaling values are @xmath242 cm and @xmath243 km / s .
the _ dotted vertical line _ marks the current epoch .
[ f6],width=245 ] | we analyse the results of recent measurements of nonthermal emission from individual supernova remnants ( snrs ) and their correspondence to the nonlinear kinetic theory of cosmic ray ( cr ) acceleration in snrs .
it is shown that the theory fits these data in a satisfactory way and provides the strong evidences for the efficient cr production in snrs accompanied by significant magnetic field amplification .
magnetic field amplification leads to considerable increase of cr maximum energy so that the spectrum of crs accelerated in snrs is consistent with the requirements for the formation of galactic cr spectrum up to the energy @xmath0 ev .
galactic cosmic rays ; supernova remnants ; shock acceleration ; nonthermal emission : radioemission , x - rays , gamma - rays |
the modelling and control of quantum linear systems is an important emerging application area which is motivated by the fact that quantum mechanical features emerge as the systems being controlled approach sub - nanometer scales and as the required levels of accuracy in control and estimation approach quantum noise limits . in recent years
, there has been considerable interest in the feedback control and modeling of linear quantum systems ; e.g. , see @xcite .
such linear quantum systems commonly arise in the area of quantum optics ; e.g. , see @xcite . the feedback control of quantum optical systems has applications in areas such as quantum communications , quantum teleportation , and gravity wave detection . in particular , the papers @xcite have been concerned with a class of linear quantum systems in which the system can be defined in terms of a set of linear quantum stochastic differential equations ( qsdes ) expressed purely in terms of annihilation operators .
such linear quantum systems correspond to optical systems made up of passive optical components such as optical cavities , beam - splitters , and phase shifters .
the main results of this paper apply to this class of linear quantum systems for the square case in which the number of outputs is equal to the number of inputs .
this paper is concerned with the use of singular perturbation approximations in order to obtain reduced dimension models for the class of linear quantum systems under consideration .
singular perturbation approximations are is widely used for obtaining reduced dimension models for classical systems ; e.g. , see @xcite . in the case of quantum systems ,
a reduced dimension model may be desired for a quantum plant to be controlled in order to simplify the controller design process which can be very complicated using existing quantum controller design methods such as the quantum lqg method of @xcite .
another application of model reduction for linear quantum systems arises in the case of controller reduction where a reduced dimension controller is obtained from a high order synthesized controller . in the case of coherent quantum control such as considered in @xcite ,
the controller is required to be a quantum system itself and thus the reduced dimension system must be physically realizable . in the physics literature , a commonly used technique in the modeling of quantum systems is the method of adiabatic elimination , which is closely connected to the singular perturbation method in linear systems theory ; e.g , see @xcite .
the papers @xcite also consider the issue of convergence of these singular perturbation approximations . in this paper , we consider the properties of the singular perturbation approximation to a linear quantum system from a linear systems point of view ; e.g. , see @xcite for a detailed description of singular perturbation methods in linear systems theory including error characterization in both the time and frequency domains . in particular , we are concerned with the physical realizability properties of the singular perturbation approximation to a linear quantum system .
the issue of physical realizability for linear quantum systems was considered in the papers @xcite .
this notion relates to whether a given qsde model represents a physical quantum system which obeys the laws of quantum mechanics . in particular , the results of the papers @xcite show that the notion of physical realizability enables a direct connection between results in quantum linear systems theory and linear systems theory . in applying singular perturbation methods to obtain approximate models of quantum systems , it is important that model obtained is a physically realizable quantum system so that it retains the essential features of a quantum system . also , if the approximate model of a quantum plant is to be used for controller synthesis , the controller synthesis procedure may need to exploit the physical realizability of the plant model . in addition , if model order reduction is applied to a coherent feedback controller which is to be implemented as a quantum system , then this reduced order controller model must be physically realizable . in the paper @xcite , the notion of physical realizability is shown to be equivalent to the lossless bounded real property for the class of square linear quantum systems under consideration .
this property requires that the system matrix is hurwitz and that the system transfer function is unitary for all frequencies .
the main result of this paper shows that if a singularly perturbed linear quantum system is physically realizable for all values of the singular perturbation parameter , then the corresponding reduced dimension approximate system has the property that all of its poles are in the closed left half of the complex plane and its transfer function is unitary for all frequencies .
these properties indicate that in all but pathological cases , the singular perturbation approximation method will yield a physically realizable reduced dimension system .
in addition , an example is given showing one such pathological system in which the singular perturbation approximation is not strictly hurwitz .
the paper also presents a result for a special case of the singularly perturbed linear quantum systems considered in this paper .
this special case corresponds to singular perturbations which arise physically from a perturbation in the system hamiltonian . in this case
, the result shows that the corresponding reduced dimension approximate system is always physically realizable .
this result can in fact be derived from the nonlinear quantum system results presented in the papers @xcite .
however , we have included this result , along with a straightforward proof , for the sake of completeness .
we have also included an example of a singularly perturbed linear quantum optical system which fits into the subclass of singularly perturbed quantum systems for which this result applies .
this example illustrates how such singularly perturbed quantum systems can arise naturally in physical quantum optical systems .
the remainder of this paper proceeds as follows . in section [ sec : systems ] , we define the class of linear quantum systems under consideration and recall some preliminary results on the physical realizability of such systems . in section [ sec : sing_pert ] , we consider the singular perturbation approximation to a linear quantum system .
we first present a result for the class of singularly perturbed linear quantum systems under consideration which relates to the lossless bounded real property .
we then consider a special class of singular perturbations which is related to corresponding perturbations of the quantum system coupling operator and hamiltonian operator .
we present a result which relates to this class of singular perturbations and shows that the corresponding approximate reduced dimension system is guaranteed to be physically realizable . in section [ sec : example ] , we present a simple example from the field of quantum optics to illustrate the proposed theory . in section [ sec : conclusions ] , we present some conclusions .
we consider a class of linear quantum systems described in terms of the annihilation operator by the following quantum stochastic differential equations ( qsdes ) : @xmath0 where @xmath1 , @xmath2 , @xmath3 and @xmath4 ; e.g. , see @xcite . here
@xmath5^t$ ] is a vector of ( linear combinations of ) annihilation operators .
the vector @xmath6 represents the input signals and is assumed to admit the decomposition : @xmath7 where @xmath8 is the noise part of @xmath6 and @xmath9 is an adapted process ( see @xcite and @xcite ) .
the noise @xmath8 is a vector of quantum noises .
the noise processes can be represented as operators on an appropriate fock space ( for more details see @xcite ) .
the process @xmath9 represents variables of other systems which may be passed to the system ( [ sys ] ) via an interaction .
more details concerning this class of quantum systems can be found in the references @xcite , @xcite .
( see @xcite . )
[ phys_real ] a linear quantum system of the form ( [ sys ] ) is said to be _ physically realizable _ if there exists a commutation matrix @xmath10 , a coupling matrix @xmath11 , a hamiltonian matrix @xmath12 , and a scattering matrix @xmath13 such that @xmath14 and @xmath15 . here
, the notation @xmath16 represents conjugate transpose . in this definition
, if the system ( [ sys ] ) is physically realizable , then the matrices @xmath13 , @xmath17 and @xmath11 define a open harmonic oscillator with scattering matrix @xmath13 , coupling operator @xmath18 and a hamiltonian operator @xmath19 ; e.g. , see @xcite , @xcite , @xcite and @xcite .
this definition is an extension of the definition given in @xcite to allow for a general scattering matrix @xmath13 ; e.g. , see @xcite . the following theorem is a straightforward extension of theorem 5.1 of @xcite to allow for a general scattering matrix @xmath13 .
( see @xcite . )
[ t1 ] a linear quantum system of the form ( [ sys ] ) is physically realizable if and only if there exists a matrix @xmath20 such that @xmath21 in this case , the corresponding hamiltonian matrix _
m _ is given by @xmath22 the corresponding coupling matrix @xmath11 is given by @xmath23 and the corresponding scattering matrix is given by @xmath24 .
note that _ m _ is a hermitian matrix .
[ d2 ] the linear quantum system ( [ sys ] ) is said to be _
lossless bounded real _ if the following conditions hold : 1 .
f is a hurwitz matrix ; i.e. , all of its eigenvalues have strictly negative real parts ; 2 .
the transfer function matrix @xmath25 satisfies @xmath26 for all @xmath27 the following definition extends the standard linear systems notion of minimal realization to linear quantum systems of the form ( [ sys ] ) ; see also @xcite .
[ d3 ] a linear quantum system of the form ( [ sys ] ) is said to be _ minimal _ if the following conditions hold : 1 .
_ controllability .
_ @xmath28 for some @xmath29 and @xmath30 implies @xmath31 ; 2 . _
observability_. @xmath32 for some @xmath29 and @xmath33 implies @xmath31 .
the following theorem is an straightforward extension of theorem 6.6 of @xcite to allow for a general scattering matrix @xmath13 .
[ t2 ] a minimal linear quantum system of the form ( [ sys ] ) is physically realizable if and only if the system is lossless bounded real .
we now consider a class of quantum systems of the form ( [ sys ] ) dependent on a parameter @xmath34 which are referred to as singularly perturbed quantum systems : @xmath35 this system can be re - written in the more standard singularly perturbed form ( e.g. , see @xcite ) : @xmath36 if the matrix @xmath37 is non - singular , we can define the corresponding reduced dimension slow subsystem ( e.g. , see @xcite ) by formally setting @xmath38 in ( [ sys2 ] ) to obtain @xmath39 where @xmath40 this is the singular perturbation approximation to the system ( [ sys1 ] ) .
we are interested in whether the reduced dimension quantum system ( [ sys3 ] ) , ( [ slow_matrices ] ) is physically realizable if the singularly perturbed quantum system ( [ sys2 ] ) is physically realizable for all @xmath34 .
one approach to addressing this question might be to apply theorem [ t2 ] and indeed , we can obtain the following theorem which is the main result of the paper : [ t3 ] if the singularly perturbed linear quantum system ( [ sys2 ] ) is physically realizable for all @xmath34 and the matrix @xmath37 is non - singular , then the corresponding reduced dimension quantum system ( [ sys3 ] ) , ( [ slow_matrices ] ) is such that the matrix @xmath41 has all of its eigenvalues in the closed left half of the complex plane and the transfer function matrix @xmath42 satisfies @xmath43 for all @xmath44 . the proof of this theorem is given in the appendix . note that this result is not sufficient to prove the physical realizability of the reduced dimension quantum system ( [ sys3 ] ) , ( [ slow_matrices ] ) since the application of theorem [ t2 ] requires that the system realization be minimal and hence the conditions of theorem [ t2 ] will only be satisfied if the matrix @xmath41 is hurwitz .
however , the properties established in this theorem indicate that in all but pathological cases , the singular perturbation approximation will yield a physically realizable reduced dimension system .
these pathological cases can be detected by testing the eigenvalues and minimality of the reduced order system .
in addition , the following example shows one such pathological system in which the singular perturbation approximation is not strictly hurwitz and not minimal .
* example * we consider a singularly perturbed quantum linear system of the form ( [ sys2 ] ) where @xmath45 ; f_{12 } = i , f_{21 } = \frac{1}{2}i , f_{22 } = -i , \nonumber \\
g_1 & = & -i , g_2 = i , h_1 = i , h_2 = -2 i , k = i. \end{aligned}\ ] ] for each @xmath34 , we calculate the characteristic polynomial of the matrix @xmath46 $ ] to be @xmath47 from this , it follows using the routh - hurwitz criterion that the matrix @xmath48 is hurwitz for all @xmath34 .
furthermore , it is straightforward to verify that the matrix @xmath49 > 0 $ ] satisfies the conditions @xmath50 where @xmath51 $ ] and @xmath52 $ ] .
hence , it follows from theorem [ t1 ] that this singularly perturbed quantum system is physically realizable for all @xmath34 .
furthermore , it follows from ( [ phys_real_ex ] ) that this system is in fact minimal for all @xmath34 .
however , when we consider the reduced order approximate system , we calculate @xmath53 $ ] which is not hurwitz . also , @xmath54 , @xmath55 , @xmath56 and thus , the reduced order system is not minimal .
this example shows that a stronger result than theorem [ t3 ] , which guarantees minimality and hurwitzness of the approximate system , can not be obtained in the general case . in the next subsection
, we consider a special class of singular perturbations for which the physical realizability of the reduced dimension system can be guaranteed .
we now consider a special class of singularly perturbed physically realizable quantum systems of the form ( [ sys1 ] ) defined in terms of the matrices @xmath13 , @xmath11 and @xmath17 in definition [ phys_real ] . indeed
, we consider the case in which @xmath57 , @xmath58;~ m = \left[\begin{array}{cc } m_{11 } & \frac{1}{\sqrt{\epsilon}}m_{12}\\ \frac{1}{\sqrt{\epsilon}}m_{12}^\dagger & \frac{1}{\epsilon}m_{22 } \end{array}\right]\ ] ] for all @xmath34 where @xmath15 , and @xmath59 and @xmath60 are hermitian matrices .
then , substituting these values into ( [ harmonic ] ) , we obtain the following linear quantum system of the form ( [ sys ] ) : @xmath61 if we make the change of variables @xmath62 , this leads to the following singularly perturbed quantum system of the form ( [ sys1 ] ) : @xmath63
note that even though we formally let @xmath64 in the singular perturbation approximation , this state space transformation can be applied for each fixed @xmath34 . then , for the singularly perturbed linear quantum system ( [ sys5 ] ) , we can obtain the corresponding reduced dimension approximate system according to equations ( [ sys3 ] ) , ( [ slow_matrices ] ) .
the following result is obtained for singularly perturbed linear quantum systems of the form ( [ sys5 ] ) .
this result can also be derived from the general nonlinear results presented in the papers @xcite . however , this result for the linear case is included here for the sake of completeness .
[ t4 ] consider a singularly perturbed linear quantum system ( [ sys5 ] ) which is physically realizable for all @xmath34 and suppose that the matrix @xmath65 is nonsingular
. then the corresponding reduced dimension approximate system defined by equations ( [ sys3 ] ) , ( [ slow_matrices ] ) is physically realizable .
the proof of this theorem is given in the appendix .
we consider an example from quantum optics involving the interconnection of two optical cavities as shown in figure [ f1 ] .
each optical cavity consists of two partially reflective mirrors which are spaced at a specified distance to give a cavity resonant frequency which corresponds to the frequency of the driving laser ; e.g. , see @xcite . in practice
, optical isolators would also need to be included in the optical connections between the cavities to ensure that the light traveled only in one direction . here
@xmath66 and @xmath67 are the coupling parameters of the first cavity and @xmath68 is the coupling parameter of the second cavity .
these parameters are determined by the physical characteristics of each cavity including the mirror reflectivities .
the qsde of the form ( [ sys ] ) describing this quantum system is as follows : @xmath69 & = & \left[\begin{array}{cc}-\frac{k_1+k_2}{2 } -\sqrt{k_1 k_2 } & -\sqrt{k_1 \tilde \gamma } \\
-\sqrt{k_2 \tilde \gamma } & -\frac{\tilde \gamma}{2}\end{array}\right ] \left[\begin{array}{c}a \\
\tilde a\end{array}\right]dt -\left[\begin{array}{c}\sqrt{k_1}+\sqrt{k_2 } \\ \sqrt{\tilde \gamma } \end{array}\right ] du_2 ; \nonumber \\ dy_1 & = & \left[\begin{array}{cc}\sqrt{k_1}+\sqrt{k_2 } & \sqrt { \tilde \gamma } \end{array}\right]\left[\begin{array}{c}a \\ \tilde
a\end{array}\right]dt + du_2.\end{aligned}\ ] ] we wish to consider the reduced dimension approximation to this system which is obtained by letting @xmath70 .
this corresponds to the case in which the mirrors in cavity 2 are perfectly reflecting and so there is a direct optical feedback from the output @xmath71 of cavity 1 into the input @xmath72 of cavity 1 .
if we let @xmath73 , it is straightforward to verify that this system is a system of the form ( [ sys4 ] ) with @xmath57 , @xmath74 , @xmath75 , @xmath76 , @xmath77 , @xmath78 , @xmath79 . with the change of variables @xmath80 ,
the system becomes @xmath81 & = & \left[\begin{array}{cc}-\frac{k_1+k_2}{2 } -\sqrt{k_1 k_2 } & -\sqrt{k_1 } \\
-\sqrt{k_2 } & -\frac{1}{2}\end{array}\right ] \left[\begin{array}{c}a \\ \tilde a\end{array}\right]dt-\left[\begin{array}{c}\sqrt{k_1}+\sqrt{k_2 } \\ 1 \end{array}\right ] du_2 ; \nonumber \\ dy_1 & = & \left[\begin{array}{cc}\sqrt{k_1}+\sqrt{k_2 } & 1 \end{array}\right]\left[\begin{array}{c}a \\
\tilde a\end{array}\right]dt + du_2\end{aligned}\ ] ] which is a singularly perturbed quantum system of the form ( [ sys2 ] ) .
hence , the corresponding reduced dimension slow subsystem ( [ sys3 ] ) , ( [ slow_matrices ] ) is given by @xmath82 since the system ( [ sys8 ] ) satisfies the conditions of theorem [ t4 ] , it follows from this theorem that the system ( [ sys9 ] ) will be physically realizable .
this can also be verified directly by noting that the system ( [ sys9 ] ) satisfies the conditions of theorem [ t1 ] with @xmath83 .
note that for this example , if @xmath84 , then the reduced dimension quantum system is uncontrollable , unobservable and has a pole at the origin .
in this paper , we have considered the physical realizability properties of the singular perturbation approximation to a class of singularly perturbed linear quantum systems
. these results may be useful in the modeling of linear quantum systems such as gravity wave detectors where a simplified model is required without sacrificing physical realizability .
the author is wishes to acknowledge useful research discussions with matthew james , elanor huntington , michele heurs and hendra nurdin .
if the singularly perturbed quantum system ( [ sys2 ] ) is physically realizable for all @xmath34 , then it follows from theorem [ t1 ] that for all @xmath34 , there exists a matrix @xmath85 such that the matrices @xmath86;~ g_\epsilon = \left[\begin{array}{c } g_{1}\\ \frac{1}{\epsilon}g_{2}\end{array}\right];\nonumber \\ h & = & \left[\begin{array}{cc } h_{1 } & h_{2 } \end{array}\right];~k\end{aligned}\ ] ] satisfy the conditions ( [ physrea ] ) .
hence , it follows from the first of these equalities and fact 12.21.3 of @xcite that matrix @xmath48 has all of its eigenvalues in the closed left half of the complex plane for all @xmath34 .
then , using a standard result on singularly perturbed linear systems ( e.g. , see theorem 3.1 on page 57 of @xcite ) it follows that the matrix @xmath41 has all of its eigenvalues in the closed left half of the complex plane . with the matrices @xmath48 , @xmath87 , @xmath88 and @xmath89 defined as above ,
it follows by a straightforward but tedious calculation that we can write the transfer function @xmath90 in the form : @xmath91^{-1}\nonumber \\ & & \times \left(g_0+f_{12}\left(i-\frac{f_{22}}{\epsilon s}\right)^{-1}g_2\right)\nonumber \\ & & + k_0+h_2f_{22}^{-1}\left(i-\frac{f_{22}}{\epsilon s}\right)^{-1}g_2\end{aligned}\ ] ] where the matrices @xmath41 , @xmath92 , @xmath93 , @xmath94 are defined as in ( [ slow_matrices ] ) .
now for small values of @xmath34 , we can approximate the term @xmath95 in the above expression as follows : @xmath96 from this , it follows that we can write @xmath97^{-1 } \nonumber \\ & & \times \left(g_0-\epsilon s f_{12}f_{22}^{-1}g_2\right)\nonumber \\ & & + k_0- \epsilon sh_2f_{22}^{-2}g_2 + o(\epsilon^2).\end{aligned}\ ] ] from this and some further straightforward manipulations and simplifications , we can obtain @xmath98 now using the fact that the matrices @xmath48 , @xmath87 , @xmath88 and @xmath89 satisfy the conditions ( [ physrea ] ) , we will show that transfer function matrix @xmath99is unitary at all frequencies . indeed using ( [ physrea ] )
, we have for all @xmath44 @xmath100\left(-i\omega i - f_\epsilon^\dagger\right)^{-1}h^\dagger\nonumber \\ & & = h\theta\left(-i\omega i - f_\epsilon^\dagger\right)^{-1}h^\dagger + h\left(i\omega i - f_\epsilon\right)^{-1}\theta h^\dagger \nonumber \\ & & = -kg_\epsilon^\dagger\left(-i\omega i - f_\epsilon^\dagger\right)^{-1}h^\dagger - h\left(i\omega i - f_\epsilon\right)^{-1}g_\epsilon k^\dagger . \end{aligned}\ ] ] now using the third equation of ( [ physrea ] ) and the fact that @xmath89 is square , we have for all @xmath44 @xmath101 therefore , since @xmath102 is square we have @xmath103 for all @xmath34 .
hence , it follows from ( [ tf_pert ] ) and the fact that ( [ unitary_e ] ) holds for all @xmath34 that we must have @xmath104 for all @xmath44 .
this completes the proof of the theorem .
@xmath105 for the singularly perturbed linear quantum system ( [ sys5 ] ) , it is straightforward but tedious to verify that the corresponding reduced dimension slow subsystem ( [ sys3 ] ) , ( [ slow_matrices ] ) is given by @xmath106 where @xmath107 furthermore , it is straightforward to verify that @xmath108 and the matrix @xmath109 is hermitian . hence , it follows from definition [ phys_real ] that the system ( [ sys6 ] ) is physically realizable with the matrices @xmath110 , @xmath111 , @xmath109 defined as above and with @xmath57 .
this completes the proof of the theorem .
@xmath105 m. yanagisawa and h. kimura , `` transfer function approach to quantum control - part i : dynamics of quantum feedback systems , '' _ ieee transactions on automatic control _ ,
48 , no . 12 , pp . 21072120 , 2003 .
v. p. belavkin and s. c. edwards , `` quantum filtering and optimal control , '' in _ quantum stochastics and information : statistics , filtering and control ( university of nottingham , uk , 15 - 22 july 2006 ) _ , v. p. belavkin and m. guta , eds.1em plus 0.5em minus 0.4emsingapore : world scientific , 2008 . j. gough and r. van handel , `` singular perturbation of quantum stochastic differential equations with coupling through an oscillator mode , '' _ journal of statistical physics _ , vol .
127 , no . 3 , pp .
575607 , 2007 .
l. bouten , r. van handel , and a. silberfalb , `` approximation and limit theorems for quantum stochastic models with unbounded coefficients , '' _ journal of functional analysis _ , vol .
254 , pp . 31233147 , 2008 .
j. gough , h. nurdin , and s. wildfeuer , `` commutativity of the adiabatic elimination limit of fast oscillatory components and the instantaneous feedback limit in quantum feedback networks , '' _ journal of mathematical physics _
51 , no . 12 , pp .
123518.1123518.25 , 2010 . | this paper considers the use of singular perturbation approximations for a class of linear quantum systems arising in the area of linear quantum optics .
the paper presents results on the physical realizability properties of the approximate system arising from singular perturbation model reduction . |
the investigation of hypernuclei is a rapidly progressing field of nuclear physics , since these nuclei provide complementary methods to improve traditional nuclear studies and open new horizons for studying nuclear physics aspects related to particle physics and nuclear astrophysics ( see , e.g. , @xcite and references therein ) .
indeed , baryons with strangeness embedded in a nuclear environment are the only available tool to approach the many - body aspects of the three - flavor strong interaction .
hyperon - nucleon and hyperon - hyperon interactions are also an essential ingredient for the nuclear equation of state ( eos ) at high density and low temperature .
another novel aspect of contemporary hypernuclear studies is the exploration of the limits of stability in isospin and strangeness space .
presently , hypernuclear physics is still focused on spectroscopic information and is dominated by a quite limited set of reactions .
these are reactions induced by high - energy hadrons and leptons leading to the production of only few particles , including kaons which are often used to tag the production of hypernuclei in their ground and low excited states . in such reactions hyper - systems with baryon density around the nuclear saturation density , @xmath0 @xmath1
are formed .
therefore , most previous theoretical studies concentrated on the calculation of the structure of nearly cold hypernuclei . however , many experimental collaborations ( e.g. , panda @xcite , fopi / cbm , and super - frs / nustar at fair @xcite ; star at rhic @xcite ; alice at lhc @xcite ; bm@n and mpd at nica @xcite ) have started or plan to investigate hypernuclei and their properties in hadron and heavy ion induced reactions .
this represents an essential extension of nuclear / hypernuclear studies : the isospin space , particle unstable states , multiple strange nuclei , the production of hypermatter , and precision lifetime measurements are unique topics of these fragmentation reactions .
it is relevant in this respect to note that the very first experimental observation of a hypernucleus was obtained in the 1950-s in nuclear multifragmentation reactions induced by cosmic rays @xcite . in recent years
a remarkable progress was made in the investigation of the fragmentation and multifragmentation reactions associated with relativistic heavy - ion collisions ( see , e.g. , @xcite and references therein ) .
this gives us an opportunity to apply well known theoretical methods , which were adopted for the description of the conventional reactions , also for the formation of hypernuclei @xcite .
the task of this work is to develop new realistic models of hypernuclear production which are able to provide detailed predictions in order to optimise the experimental conditions when searching for both @xmath2-hypernuclei and normal exotic nuclei .
the formation processes of hypernuclei are quite different in central and peripheral ion collisions .
there are indications that in high energetic central collisions the coalescence mechanism , which assemble light hyper - fragments from the produced hyperons and nucleons ( including anti - baryons ) is dominating @xcite .
because of the very high temperature of the fireball ( t@xmath3160 mev ) only lightest clusters , with mass numbers a@xmath44 , can be produced in this way with a reasonable yield @xcite .
on the other hand , it was noticed sometime ago that the capture of hyperons in spectator regions after peripheral nuclear collisions is a promising way to produce hypernuclei @xcite .
nuclear matter created in peripheral collisions shows distinctly different properties compared to nuclear matter at mid - rapidity .
it is well established that moderately excited spectator residues ( t@xmath45 - 6 mev ) are produced in such reactions @xcite .
a hyperon bound in these residues should not change the picture since the hyperon - nucleon forces are of the same order as the nucleon - nucleon ones .
general features of the decay of such hyper - residues into hyper - fragments could be investigated with statistical models ( e.g.,generalized statistical multifragmentation model smm @xcite ) , which successfully describe the production of normal fragments @xcite .
the models predict the formation of exotic hypernuclei and hypernuclei beyond the drip - lines , which are difficult to create in other reactions @xcite .
there is an alternative treatment of the process that considers first statistical smm decay of excited residues , and , afterwards , a coalescence model for final production of hyper - fragments @xcite .
both theoretical mechanisms are under discussion and waiting for a test by experiments .
spectator heavy fissioning hypernuclei were identified with a relatively high probability in reactions induced by protons with energy around the threshold @xcite , and in annihilation of antiprotons @xcite .
very encouraging results on hypernuclei come from experiments with light projectiles : in addition to well - known hypernuclei @xcite , evidences for unexpected exotic hypernuclear states , like a @xmath2 hyperon bound to two neutrons , were reported @xcite , which were never observed in other reactions .
as was discussed , the production of such new exotic states could be naturally explained within the break - up of excited hypernuclear systems @xcite . in previous publications we have considered the formation of hypernuclei within the dubna cascade model ( dcm ) @xcite and the ultra - relativistic quantum molecular dynamics model ( urqmd ) @xcite .
these calculations include the capture of the produced hyperons in the potential of the spectator residues @xcite , and the coalescence into lightest clusters together with their thermal production in central collisions @xcite .
involving new transport models is very important since we obtain knowledge about uncertainties in such calculations . in this work , besides urqmd , we employ the hadron - string dynamics ( hsd ) model @xcite , which were used successfully for description of strangeness production @xcite .
we develop a generalization of the coalescence model @xcite , the coalescence of baryons ( cb ) , which is applied after urqmd and hsd stage . in this way it is possible to form fragments of all sizes , from the lightest nuclei to the heavy residues , including hypernuclei within the same mechanism .
the advantage of this procedure is the possibility to predict the correlations of yields of hypernuclei , including their sizes , with the rapidity on the event - by - event basis , that is very essential for the planning of future experiments .
a detailed picture of peripheral relativistic heavy - ion collisions has been established in many experimental and theoretical studies .
nucleons from the overlapping parts of the projectile and target ( participant zone ) interact intensively between each other and with other hadrons produced in primary and secondary collisions .
nucleons from the non - overlapping parts interact rarely , and they form the residual nuclear systems , which we call spectators .
we apply two dynamical models to describe the processes leading to the production of strange particles in nucleus - nucleus collisions before their accumulation in nuclear matter .
using different models allows us to estimate the theoretical uncertainties associated with the different treatment of the dynamical stage .
the first model is the ultra - relativistic quantum molecular dynamics model ( urqmd ) @xcite .
the model is based on an effective microscopic solution of the relativistic boltzmann equation .
products of binary interactions of particles include 39 different hadronic species ( and their anti - particles ) which scatter according to their geometrical cross section .
the allowed processes include elastic scattering and @xmath5 processes via resonance creation ( and decays ) as well as string excitations for large center - of - mass energies ( @xmath6 gev ) . the current version 3.4 of urqmd also includes important strangeness exchange reactions , e.g. , @xmath7 ( where @xmath8 is a strange baryon ) @xcite .
another model is the off - shell hadron - string - dynamics ( hsd ) transport model @xcite .
it is based on the solution of the generalized transport equation @xcite including covariant self energies for the baryons .
we recall that in the hsd approach nucleons , @xmath9 s , n@xmath10(1440 ) , n@xmath10(1535 ) , @xmath2 , @xmath11 and @xmath12 hyperons , @xmath13 s , @xmath14 s and @xmath15 s as well as their antiparticles are included on the baryonic side , whereas the @xmath16 and @xmath17 octet states are incorporated in the mesonic sector .
inelastic baryon baryon ( and meson - baryon ) collisions with energies above @xmath18 gev ( and @xmath19 gev , respectively ) are described by the fritiof string model @xcite , whereas low energy hadron
hadron collisions are modelled using experimental cross sections . in the both hsd and urqmd models
the initial state of colliding nuclei is generated similarly : the nucleon s coordinates are initialized according to a woods - saxon profile in coordinate space and their momenta are assigned randomly according to the fermi distribution .
a composite particle can be formed from two or more nucleons if they are close to each other in phase space .
this simple prescription is known as coalescence model and it is based on the properties of the nucleon
nucleon interaction .
one can use the coalescence in both momentum ( velocity ) space and the coordinate space .
the coalescence in the momentum space model has proven successful in reproducing experimental data on the production of light clusters ( see e.g. @xcite ) .
recently , we developed an alternative formulation of the coalescence model , the coalescence of baryons ( cb ) , which is suitable for computer event by event simulations @xcite .
baryons ( nucleons and hyperons ) can produce a cluster with mass number @xmath20 if their velocities relative to the center - of - mass velocity of the cluster is less than @xmath21 .
accordingly we require @xmath22 for all @xmath23 , where @xmath24 ( @xmath25 are momenta and @xmath26 is the sum energy of the baryons in the cluster ) .
this is performed by sequential comparison of the velocities of all baryons .
if we consider only the production of lightest clusters ( a@xmath44 ) the coalescence velocity parameter @xmath27 gives a good description of the data , as was shown in previous analyses @xcite . however , the coalescence mechanism may also be applied to construct heavy nuclei @xcite . in this case
the parameter @xmath28 should be larger , in order to incorporate higher velocities of the hyperons which can be captured in the deeper potentials of big nuclei .
this potential well saturates at around @xmath293040 mev . it was demonstrated in ref .
@xcite ( fig .
10 ) that according to this potential criterion the momentum distribution of the captured @xmath2 hyperons can be approximated by a step - like function , and that hyperons with relative momenta less than 200 - 250 mev / c can be bound .
therefore , relative velocities up to @xmath30 should be taken into account as coalescence parameter , which is naturally close to the fermi velocity .
we would like to note a problem which is sometimes disregarded in coalescence simulations .
some nucleons may have velocities such that they can belong to different ( or even more than two ) coalescent clusters according to the coalescence criterion . in these cases
the final yield will depend on the sequence of nucleons within the algorithm . to avoid this uncertainty we developed an iterative coalescence procedure .
@xmath31 steps are calculated in the coalescence routine with the radius @xmath32 which is increased at each step @xmath33 : @xmath34 ( with @xmath35 ) .
clusters produced at earlier steps participate as a whole in the following steps . in this case
the final clusters not only meet the coalescence criterion but also their nucleons have the minimum distance in the velocity space .
this procedure gives a mathematically correct result in the limit @xmath36 , however , we found that in practical calculations it is sufficient to confine the steps to @xmath31=5 . for more reliable identification of the clusters ,
we apply in addition a coordinate proximity criterion . a single baryon and a cluster with mass number a is confirmed to compose the new cluster if the relative distances of all baryons from the cluster s center of mass @xmath37 is less than @xmath38(a+1)@xmath39 .
here is @xmath40=2 fm , which can be justified by multifragmentation studies ( see , e.g. , @xcite ) : it was established that an excited nuclear system ( with mass a ) before its disintegration may reach a big freeze - out volume @xmath41 . in this volume
the system can be considered in thermal equilibrium and live for a short time ( @xmath29100 fm / c ) .
we assume that the coalescence should be a rather fast process which happens when particles leave the interaction zone and the rate of the secondary interactions decreased considerably . from the urqmd model calculation for 20 a gev
we evaluate this time as @xmath450 fm / c for big targets and projectiles @xcite .
that is usually smaller than the decay time of nuclear systems in both multifragmentation and evaporation / fission processes : it is therefore naturally that such a coalescence prescription may introduce an excitation energy in the clusters , which can decay afterwards . as our analysis shows , the criterion in the coordinate space correlates with the velocity criterion .
namely , when the projectile and target are well separated at later times of the reaction , the proximity in the velocity space means the proximity in the coordinate space too .
this correlation appears naturally within the potential capture criterion @xcite .
another important development of our coalescence procedure is that we assign the primary nucleons of initial nuclei to the residual nuclei if they did not interact with any particle during the collision .
in some approaches these residues are formed by default @xcite . however , in the transport models used here ( urqmd , hsd ) these nucleons preserve initialized momenta .
the stability of the initial nuclei is not pursued in this case , since low - energy interactions inside nucleus are not precisely determined .
we think it is a very good approximation to combine such nucleons into a residual cluster , especially for peripheral collisions , because of their initial proximity in momentum and coordinate space .
we have checked that it is not effective to resolve the residues problem by simply increasing the coalescence parameters within our procedure , since we would enforce artificially the formation of the lightest clusters too .
in addition we verify the assignment of these nucleons to the residues by controlling their rapidities @xmath42 : most nucleons deviate by less than @xmath43 from the rapidities of the target and projectile .
( this range @xmath44 is associated with the fermi momenta of nucleons in the initial nucleus @xcite . ) as we know from the production of normal fragments @xcite , the realistic description of residues is important : according to the general picture of these reactions @xcite , many hyperons can be captured by a sufficiently large piece of excited spectator matter , leading to the formation of hot hypermatter , which in the following decay via evaporation , fermi - break - up , fission , or multifragmentation .
for our analysis we have selected both small and heavy colliding nuclei , however , symmetric systems . for carbon projectile and targets
we have performed calculations for different energies which are relevant for gsi / fair facility @xcite .
the reason is that the light hypernuclei can be quite easily identified in future experiments , and that such experiments are already planned @xcite .
the reactions with gold nuclei are added to generalize our conclusions for the production of heavy fragments at high beam energy .
such energies are easily reached with available accelerators ( in particular , rhic ) and our predictions can help to prepare measurements of both light and heavy hypernuclei at all possible rapidities .
we generated from @xmath45 to @xmath46 inelastic events for each energy while integrating over all impact parameters ( minimal bias calculations ) .
for collisions of light ( carbon ) nuclei we stop our transport calculations at the time moment of 20 fm / c after the maximum overlap between the target and projectile has been reached .
we have checked that at later time cuts the number of produced particles and their momenta change very little , since they are far separated from each other .
the result of the coalescence process remains nearly the same if we increase this time scale .
for the gold nuclei the corresponding time was taken as 40 fm / c , since they are larger . in fig . 1 we show the urqmd results of the total rapidity distributions of all baryons and @xmath2 hyperons produced in the collisions of a carbon projectile and target with laboratory beam energies of 2 and 10 gev per nucleon , and for a gold on gold system at 15 gev per nucleon .
we have specially separated the remaining spectator nucleons which participate mostly in producing fragments and hyper - fragments in the target and projectile region .
the rapidity distributions obtained with hsd model have practically the same form .
one can see wide baryon and hyperon distributions which cover the whole rapidity range of the reaction .
the peaks at projectile and target rapidities do mainly consist of the spectator nucleons which did not interact and which form the residues .
all these particles are both the output of the transport models and the input for the coalescence approach leading to formation of nuclear and hyper - nuclear matter . large excited pieces of hyper - matter can be produced by the capture of @xmath2 hyperons within the nuclear residues , as demonstrated also in ref .
similar to nuclear reactions without the involvement of strange particles , we expect that these hyper - residues will be excited and decay afterwards @xcite producing final nuclei and hypernuclei .
however , the formation of light hyper - clusters can take place at all rapidities .
this is an advantage of the coalescence procedure as it can account for all these phenomena on the same footing , and , therefore , systematic comparisons can be performed .
the total mass yields of the normal fragments and hyper - fragments ( with one bound @xmath2 ) are shown in fig . 2 .
the coalescence of baryons ( the cb model ) was applied after the urqmd , for the reactions demonstrated in fig . 1
. the yields are normalized per one inelastic event .
however , one should take into account that only events with production of hyperons have been analysed in this case .
for this reason there is no characteristic increase of the yield of normal fragments with masses around the projectile / target mass , which are caused by very peripheral collisions .
the explanation of this behaviour was already suggested in ref .
@xcite : the production of hyperons needs usually many particle collisions leading to a considerable emission of fast nucleons from the residues .
the hsd+cb calculations show similar distributions .
one can see that the production of fragments of all sizes is possible . as expected
the yield of conventional fragments is by few orders of magnitude higher than the yield of hyper - fragments .
nevertheless , the production of hyper - fragments is sufficient to be experimentally measured ( see also @xcite ) .
it is a natural result of the coalescence that the yield of the lightest hyper - fragments is dominating .
however , the capture of hyperons by residues saturates the yield for large masses and leads to abundant production of heavy hyper - fragments . within this approach
one can see clearly that nearly all normal fragments and hyper - fragments with @xmath47 in the carbon collisions , and with @xmath48 in the gold collisions originate from the capture of @xmath2 hyperons by spectator residues ( dotted lines ) .
as was mentioned we believe that these hyper - fragments represent excited pieces of hyper - matter whose evolution can be calculated with statistical models @xcite .
the excitation energy of such primary fragments can also be evaluated from the analysis of experimental data @xcite . for this calculation
we have used the coalescence parameter @xmath49 in order to take into account the higher velocities possible in big clusters formed by the residues .
it is also consistent with the values obtained in our previous analysis in ref .
decreasing @xmath28 leads to a smaller yield of hyper - fragments , without changing the form of their distribution . in this case the yield of light normal fragments is reduced while residues are hardly affected . to complement
the analysis of the fragment masses we provide information about the velocities of all produced hyper - fragments . here and in the following figures we consider the hyper - fragments and hyper - residues with mass numbers @xmath50 .
their total rapidity distributions are demonstrated in figs . 3 and 4 , for urqmd+cb and hsd+cb calculations respectively .
the hyper - fragments can be produced at any rapidity available for hyperons in the reaction ( solid lines ) . however , as seen from the figures the big fragments , which can come only from the residues , are concentrated around the target and projectile rapidity ( dashed lines ) .
the small fragments formed after the coalescence of fast baryons can populate the midrapidity region also . as can be seen from a comparison of fig . 3 and fig . 4 the shape of the obtained distributions do not change with the employed transport models , urqmd and hsd .
however , the yields of light hyper - fragments are slightly larger in the hsd case .
it is instructive that in the carbon collisions the hyper - residues are responsible for producing nearly all hyper - fragments in their kinematic regions . in the gold case , many new particles are produced in this region , therefore , besides big hyper - residues additional light hypernuclei can be formed too .
the light hypernuclei @xmath51h and @xmath52h are specially interesting : they can be easily identified by their decay into @xmath53 and @xmath54he , and into @xmath53 and @xmath55he , respectively .
these correlations have been observed already in many heavy - ion experiments at high energies @xcite .
such hypernuclei can serve as indicators for the production of hyper - matter . in figs . 5 and 6 we show the rapidity distributions of these light hypernuclei produced in the same reactions .
the simulations are performed within the urqmd and cb models .
as before , the coalescence parameter @xmath49 has been used for the calculations shown in fig . 5 .
for comparison , the results obtained with a smaller parameter @xmath56 are presented in fig .
the latter may be more adequate for these small nuclei , since previously the yields of normal small clusters have been well described with a such low coalescence parameter @xcite . in this case
the fragments can be treated already as nuclei in final state without secondary de - excitation , since the later one is mainly relevant for big residues .
one can see an interesting behaviour : the @xmath51h nuclei are essentially formed over all rapidities .
it is obvious , that the production of the clusters is smaller at low coalescent parameters ( compare figs . 5 and 6 ) .
at low @xmath28 , however , the fragments are more grouped at the target and projectile rapidities .
this concentration is more evident for lager nuclei @xmath52h .
this is the consequence of the applied coalescent mechanism : if the velocity space is reduced the large clusters are more efficient in the capture of hyperons because of the larger coordinate space . with increasing energy the fraction of nuclei around residues increases , since more particles are produced in this region as a result of secondary interactions . whereas particles originating from midrapidity have higher energy and they are more separated in the phase space . therefore , despite of the general increase the number of such particles , the total number of clusters may not increase .
the dependence of the results on the coalescence parameter was specially investigated in these reactions , since it is related to the main uncertainty of the predictions .
7 and 8 show the total yields of hyper - fragments and hyper - residues in collisions of carbon on carbon , and gold on gold , respectively . as usual , the yields are normalized per inelastic event and integrated over all impact parameters .
one can see that the urqmd as well as the hsd model gives similar results . as expected all yields increase with @xmath28 .
however , at small @xmath28 the capture of hyperons take place on residual nuclei predominantly .
the big residues cover a large coordinate space region that becomes important for this mechanism in the case of reducing the momentum space .
it is also essential that the secondary interactions which contribute considerably to the formation of hyperons with relatively low momenta happen mostly in the residue region . on the other hand ,
primary interactions leading to the production of high - energy hyperons take place in the central ( midrapidity ) region . in this case particles
are far from each other in momentum space , therefore , in order to construct a cluster a larger @xmath28 is required . in the gold collisions the difference between the calculations with urqmd and hsd for light hyper - fragments is more prominent , by about a factor 2 ( see fig .
8) . within the models
this is related to the treatment of the secondary interactions and hsd leads to an enlarged production of these hyper - fragments .
however , the values predicted by both models look quite reasonable , and they can be checked by analysing experimental data . in turn , the possible variation of the predictions is important for planing future measurements .
since increasing the beam energy results in a larger number of produced hyperons , the yields of hyper - fragments may increase too .
this is clearly seen in fig . 7 .
more details are shown in fig . 9 for carbon collisions for a wide range of beam energies .
there is a saturation of the yields of hyper - fragments , both light and heavy ones , at energies higher than 510 gev per nucleon .
this effect is found for both models and for all coalescent parameters .
depending on @xmath28 this saturation happens at a different level . by comparing these results with the previous ones which were obtained with the dcm and the capture of hyperons by the nuclear potential
( see fig . 2 in ref .
@xcite ) we note that they will be similar if we take the values of @xmath28 in - between the ones shown now in fig .
9 ( i.e. , in - between 0.1 and 0.22 ) .
the uncertainty obtained with this parameter should be clarified by a comparison with experimental data and with more sophisticated theory calculations . because of the saturation of the yield at high energies the experimental hypernuclear studies can be pursued at the accelerators of moderate relativistic energies , above the threshold ( @xmath571.6 a gev ) .
generally , this combination of the transport and coalescence models can be used for analysis of yields of non - strange fragments too .
this can give an additional insight into the reaction mechanism .
besides light fragments which were already tested @xcite , the intermediate and large fragments are also of considerable interest , e.g. , see the aladin data @xcite . as mentioned
, a detailed comparison may require a connection with the secondary de - excitation processes .
we conclude that relativistic heavy - ion reactions are a very promising source of hyper - matter and hyper - fragments .
a large amount of hypernuclei of all masses can be produced .
their properties can also be investigated taking into account the advantages of relativistic velocities , e.g. , for the life - time and correlation measurements .
the well established urqmd and hsd transport models have been used in order to describe the strangeness and hyperon formation .
they give a quite reliable picture of the reactions and they also consistent with other dynamical approaches ( e.g. , dcm ) used by us previously .
the interaction of hyperons with nucleons leads to their capture and to the formation of hyper - matter .
we describe this process within a generalized coalescence model .
the coalescence of baryons is consistent with the hyperon capture in a potential well of large nuclear residues , and the coalescence parameters are expected to be of the same order as for normal fragments .
this procedure gives a possibility to consider the formation of light hypernuclei on the same footing .
we demonstrate that big hyper - fragments are mostly produced from the spectator residues , while the light ones can be formed at all rapidities .
we expect , however , that some large species of hypermatter will be excited , and decay afterwards with production final hypernuclei and normal nuclei , as in usual fragmentation and multifragmentation reactions .
such a mechanism should allow the investigation of possible phase transitions in hypermatter with statistical models describing the secondary disintegration . by summing up the results obtained with various models we note that the production of hyper - fragments in relativistic heavy - ion collisions is universal and
well established theoretically .
it is very instructive to investigate the whole reaction mechanism by measuring simultaneously big hypernuclei originating from the residues and light hypernuclei which can be formed in the hot midrapidity region .
the saturation of all yields takes place at the beam energies higher than 510 gev per nucleon .
this opens the possibility to study hypernuclei at gsi / fair ( darmstadt ) , nuclotron / nica ( dubna ) , rhic ( brookhaven ) , hiaf ( lanzhou ) and other heavy - ion accelerators of moderate relativistic energies . in the following we plan to analyze theoretically the formation of multi - hyperon nuclei , which can be abundantly produced in these reactions .
another promising opportunity would be to study unstable ( resonance ) states of hypernuclei via particle correlations .
in addition , exotic hypernuclei may be formed and investigated in the secondary evaporation , fission , and multifragmentation like processes .
comparing these theoretical predictions with future experiments may provide new information on the @xmath58 and @xmath59 interaction at low energies , as well as about properties of hyper - matter .
this work was supported by the the gsi helmholtzzentrum fr schwerionenforschung and hessian initiative for excellence ( loewe ) through the helmholtz international center for fair ( hic for fair ) we also acknowledge the support by the research infrastructure integrating activity study of strongly interacting matter hadronphysics3 under the 7th framework programme of eu ( sphere network ) .
the computational resources were provided by the loewe frankfurt center for scientific computing ( loewe - csc ) .
c.rappold , t.r .
saito , and c. scheidenberger .
simulation study of the production of exotic hypernuclei at the super - frs ( at gsi scientific report 2012 ) , gsi report 2013 - 1 , 176 p. ( 2013 ) . | within a combined approach we investigate the main features of the production of hyper - fragments in relativistic heavy - ion collisions .
the formation of hyperons is modelled within the urqmd and hsd transport codes . to describe the hyperon capture by nucleons and nuclear residues a coalescence of baryons ( cb ) model was developed .
we demonstrate that the origin of hypernuclei of various masses can be explained by typical baryon interactions , and that it is similar to processes leading to the production of conventional nuclei . at high beam energies
we predict a saturation of the yields of all hyper - fragments , therefore , this kind of reactions can be studied with high yields even at the accelerators of moderate relativistic energies .
hypernuclei , relativistic heavy - ion reactions , nuclear fragments production |
the hydrogen @xmath0 emission is one of the strongest lines in the uv band .
it provides a unique and powerful tool to search for distant galaxies , as first suggested by @xcite , and evidenced by the successful detection of hundreds of @xmath0 emitting galaxies , or @xmath0 emitters ( laes ) , at redshifts @xmath4 over the past decade ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
more recently , @xcite have discovered the most distant galaxy at @xmath5 using the @xmath0 line . to obtain a full picture of the physical properties of the high - redshift laes
, efforts have been made recently to observe these objects in broad - band continuum ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
one remarkable result from these multi - wavelength surveys is that , laes appear to evolve with cosmic time : the laes at high redshifts @xmath6 tend to be smaller and younger galaxies , and have less dust extinction and higher equivalent widths ( ews ) than their counterparts at lower redshifts @xmath7 . despite the rapid progress in the @xmath0 detections ,
the origin and nature of these distant laes , however , remain largely unknown , because the radiative transfer ( rt ) of @xmath0 line is a complicated process .
it depends on the geometry , kinematics , ionization state , and dust content of the surrounding medium , as well as the stellar population and other ionizing sources of the galaxy . to date
, there have been a number of theoretical studies on the resonant scattering of @xmath0 photons , analytically in simple geometry ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , and numerically ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . among the numerical codes ,
monte carlo method is widely employed to solve the @xmath0 rt , owing to the irregular geometry and inhomogeneous distribution of the interstellar medium ( ism ) , and the slow convergence of the numerical techniques at high optical depths . by applying @xmath0 rt codes to cosmological simulations
, several aspects of the laes have been investigated by various groups , including observed properties of laes at @xmath8 @xcite , escape fraction of @xmath0 photons from high - z galaxies @xcite , effects of ionization on the emergent @xmath0 spectra @xcite , and mechanism of high ews in dusty multi - phase ism @xcite .
however , most of these previous works have focused only on the @xmath0 line , and were therefore limited to some aspect of the @xmath0 properties , such as flux and line profile . on the other hand ,
there has been an array of rt codes in continuum over the last several decades , using either ray - tracing methods ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , or monte carlo methods ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in particular , @xcite developed a multi - wavelength rt code that employs radiative equilibrium and an adaptive refinement grid , which reproduced the observed dust properties of the most distant quasars detected at @xmath9 when applied to the hydrodynamic quasar simulations of @xcite .
however , most of these developments focused mainly on radiation transport in dusty environments , in particular the absorption and re - emission by dust , of local star - forming regions . in order to explain the diverse properties of laes and their evolution with redshift , it is critical to couple @xmath0 with continuum , in the context of galaxy formation and evolution .
several aspects of such an approach have been attempted in previous studies . for examples , @xcite incorporated pre - computed grid of @xmath0 rt into ionization calculations , @xcite considered a combination of @xmath0 and ionization rt , @xcite studied the effect of ionization and dust on transfer of @xmath0 photons , while @xcite included dust and non - ionizing uv continuum in the @xmath0 rt calculations .
an ultimate goal would be to tackle the ionization of hydrogen , non - ionizing continuum , interstellar dust , and @xmath0 propagation and scattering simultaneously .
furthermore , the implementation of @xmath0 emission should include two major generation mechanisms , the recombination of ionizing photons , and collisional excitation of hydrogen gas .
in addition , relevant ionizing sources such as stars , active galactic nucleus ( agn ) , and uv background radiation should be included .
last , but not the least , the rt code should incorporate an adaptive refinement grid in order to cover a large dynamical range and resolve dense gas regions where star formation takes place and prominent @xmath0 emission comes from @xcite .
in this work , we present a new , 3-d monte carlo rt code , all - wavelength radiative transfer with adaptive refinement tree ( @xmath1 ) , which couples @xmath0 line and multi - wavelength continuum , for the study of panchromatic properties of galaxies and ism .
( note that the `` coupling '' in this paper means the rt of @xmath0 with other continuum photons , not the coupling between rt and hydrodynamics . )
our code improves over the original continuum - only version of @xcite , and features three essential modules : continuum emission from x - ray to radio , @xmath0 emission from both recombination and collisional excitation , and ionization of neutral hydrogen .
the coupling of these three modules enables a self - consistent and accurate calculation of the @xmath0 and multi - wavelength properties of galaxies , as the equivalent width of the @xmath0 line depends on the uv continuum , and the escape fraction of @xmath0 photons strongly depends on the ionization structure and the dust content of the object .
moreover , the adaptive refinement grid handles arbitrary geometry and efficiently traces inhomogeneous density distribution in galaxies and ism .
furthermore , this code takes into account radiation from both stars and accreting black holes , so it can be used to study both galaxies and agns .
@xmath1 can produce a number of observables , including spectral energy distribution ( sed ) from x - ray to radio , multi - band fluxes and images , @xmath0 emission line and its ew , which can be directly compared to real observations .
it has a wide range of applications , and can be easily applied to simulations using either grid - based or smoothed particle hydrodynamic ( sph ) codes by converting the simulation snapshot to the grid structure of @xmath1 .
the paper is organized as follows . in 2 , we describe the @xmath1 code , which includes implementation of three modules of continuum , @xmath0 line , and ionization of hydrogen , the adaptive refinement grid , and the dust model . in 3 , we present the application of @xmath1 to a sph cosmological simulation , and study in details the @xmath0 emission and multi - band properties from individual galaxies at redshifts z=8.5 , 6.2 , and 3.1 from the simulation .
we discuss in 4 contribution to @xmath0 emission from agns , stars , and excitation cooling from gas accretion in our model , and effects of numerical resolutions on the rt calculations , and summarize in 5 .
art@xmath10 is a 3-d , monte carlo rt code based on the original version of @xcite , which included the continuum emission from x - ray to radio , and an adaptive refinement grid .
we have added two more modules , @xmath0 emission from both recombination and collisional excitation , and ionization of neutral hydrogen , to the current version and couple them self - consistently .
the continuum part was described in details in @xcite . here
we briefly outline the continuum procedure , and focus on the new implementations of @xmath0 line and ionization .
the continuum module in @xmath1 was developed in @xcite , which adopted the radiative equilibrium algorithm of @xcite . in dusty environments , absorbed
radiation energy by dust is re - emitted as thermal emission .
the re - emitted spectrum depends on the temperature of the dust , which is assumed to be in thermal equilibrium with the radiation field .
the radiative equilibrium is ensured by performing the monte carlo transfer iteratively until the dust temperature distribution converges , which can be computationally expensive . to accelerate the calculation
, @xmath1 uses the
immediate reemission " scheme @xcite , in which the dust temperature is immediately updated on absorption of a photon packet , and the frequency of re - emitted photons are sampled from a spectrum that takes into account the modified temperature .
the temperature is determined by the balance between the stacked energy absorbed by dust in the cell and the thermal emission from them .
the emitted energy in a cell in the time interval @xmath11 is @xmath12 where @xmath13 is the planck mean opacity , @xmath14 is the frequency integrated planck function , and @xmath15 is the dust mass in the cell , the subscript @xmath16 indicates the @xmath16th cell .
solving the balance between the absorbed and emitted energy , we obtain the dust temperature as follows after absorbing @xmath17 packets , @xmath18 where @xmath19 is the total number of photon packets in the simulation , and @xmath20 is the total source luminosity .
note that because the dust opacity is temperature - independent , the product @xmath21 increases monotonically with temperature .
consequently , @xmath22 always increases when the cell absorbs an additional packet .
the added energy to be radiated owing to the temperature increase @xmath23 is determined by a temperature - corrected emissivity @xmath24 in the following approximation when the temperature increase , @xmath23 , is small : @xmath25 the re - emitted packets , which comprise the diffuse radiation field , then continue to be scattered , absorbed , and re - emitted until they finally escape from the system .
this method conserves the total energy exactly , and does not require any iteration as the emergent sed , @xmath26 , corresponds to the equilibrium temperature distribution .
hydrogen @xmath0 photon corresponds to the transition between the @xmath27 and @xmath28 levels of a hydrogen atom .
it is the strongest hi transition .
the rt of @xmath0 photons is determined by @xmath0 resonant scattering , dust absorption and scattering , and ionization state of the medium .
the process is highly complicated in galaxies owing to the complex geometry and gas distribution . in order to accelerate the numerical convergence of the rt process ,
monte carlo method has been commonly used in a number of @xmath0 codes ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
? * ; * ? ? ?
our implementation of @xmath0 line transfer adopts the monte carlo method , and the major improvements over many of these codes are that , it is coupled with ionization and multi - wavelength continuum which enables a self - consistent and accurate calculation of the @xmath0 properties , and is incorporated with a 3-d adaptive - mesh refinement grid which efficiently handles arbitrary geometry and inhomogeneous density distribution .
moreover , we treat both recombination and collisional excitation for @xmath0 emission . the optical depth @xmath29 of a @xmath0 photon with frequency @xmath30 traveling a path of length @xmath31 is determined by @xmath32 where @xmath33 is the number density of neutral hydrogen gas with parallel velocity component @xmath34 , @xmath35 is the scattering cross section as a function of frequency . in the rest frame of the hydrogen atom
, @xmath35 takes the form @xmath36 where @xmath37 is the @xmath0 oscillator strength , @xmath38 is the line - center frequency , @xmath39 is the natural line width , and the other symbols have their usual meaning . assuming a maxwellian distribution for the thermal velocity of the encountering atoms , the resulting average cross section is @xmath40 where @xmath41 is the voigt function , @xmath42^{1/2}\nu_{0}$ ] is the doppler width , @xmath43 is the relative frequency of the incident photon in the laboratory frame , and @xmath44 is the relative line width .
some previous work approximated the voigt function with a gaussian fitting in the core and a power law fitting in the wing .
however , since this approximation causes a relative error of a few tens per cent at the transit region , we use the analytical fit of @xcite , @xmath45 where @xmath46 with @xmath47 and @xmath48 .
this approximation fits the voigt function well for all frequencies , and the relative error is always less than 1 percent above a temperature of 2k .
the optical depth of dust absorption and scattering is estimated as , @xmath49 where @xmath50 is the dust mass and @xmath51 is the mass opacity coefficient of absorption and scattering .
when the stacked optical depth through passing cells achieves an optical depth determined by equation ( [ eq : tau ] ) , the encountering medium is chosen using a random number by comparing with the fraction , @xmath52 .
when @xmath0 photons are scattered by neutral hydrogen atoms , the frequency in laboratory frame is changed depending on the velocity components of the atoms and the direction of the incidence and the scattering .
then the velocity components of the directions perpendicular to the incident direction will follow a gaussian distribution , and can be generated by a simple box - muller method @xcite . however , the parallel component of the velocity depends on relative frequency @xmath53 of the incident photon owing to the resonance nature of the scattering .
the probability distribution of the parallel component is drawn from @xmath54 where @xmath55 is velocity of the parallel component normalized by thermal velocity @xmath56 ( hereafter @xmath57 means normalized velocity by @xmath56 ) . to follow this distribution , we use the method of @xcite .
when a photon has a frequency @xmath58 in the optically - thick cell , where @xmath59 is the boundary between the core and the wing of the voigt profile , i.e. , where @xmath60 , it can not travel long distance and is confined in the cell with numerous scattering .
only when it has a large x from scattering by high velocity atoms can it escape from the cell and travel long distance .
the photon usually experiences many scatterings ( @xmath61 ) before it moves into the wing , it is therefore extremely computation costly to trace this process for all photon packets . in order to speed up the calculation , in particular to avoid the huge number of scatterings in the core
, we use a core skipping method developed by @xcite and follow the procedure of @xcite and @xcite .
it artificially push the photon in the wing by scattering with atoms which have high velocity in the direction perpendicular to the incident direction .
the velocity of the perpendicular component @xmath62 is generated from a truncated gaussian , @xmath63 where @xmath64 and @xmath65 are two univariates , and we use the critical frequency @xmath66 introduced in @xcite , i.e. , @xmath67 where @xmath68 for @xmath69 , or @xmath70 for @xmath71 .
this acceleration scheme can reduce the calculation time by several orders of magnitude , and can produce a line profile which agrees well with analytical solutions in a static slab . the frequency after scattering depends on the direction in which the photon is scattered .
the direction is given by the phase function , @xmath72 , where @xmath73 is the angle between the incident direction @xmath74 and the outgoing direction @xmath75 , and @xmath76 is the degree of polarization .
the ratio @xmath76 becomes 3/7 for the @xmath77 state in @xmath58 @xcite , and 1 for the scattering in the wing @xcite .
the transition from @xmath78 in the core , together with the core skipping scheme , results in isotropic scattering .
the difference in the phase function does not affect the @xmath0 properties such as the escape fraction @xmath79 , the emergent line profile and luminosity , even if a single phase function is used for all scatterings ( e.g. , * ? ? ?
therefore , we simply assume an isotropic scattering for all the scattering process in this work .
we should point out , however , that the phase function becomes important in the polarization effect of @xmath0 line ( e.g. , * ? ? ?
* ) , and so care must be taken in dealing with that . in the scattering process , the final frequency in the laboratory frame is then @xmath80 where the factor @xmath81 takes into account the recoil effect @xcite with @xmath82 being the planck constant .
we trace the @xmath0 photon packet until it is absorbed by dust or escapes from a galaxy ( e.g. , when the photons move out of the calculation box typically 10 times larger than the virial radius of a galaxy ) .
@xmath0 emission is generated by two major mechanisms : recombination of ionizing photons and collisional excitation of hydrogen gas .
* * recombination * : ionizing radiation from stars , agns and uvb can ionize the hydrogen gas in galaxies .
the collision by high - temperature gas can also ionize the hydrogen .
the ionized hydrogen atoms then recombine and create @xmath0 photons via the state transition @xmath83 .
the @xmath0 emissivity by the recombination is @xmath84 where @xmath85 is the case b recombination coefficient , and @xmath86 is the average number of @xmath0 photons produced per case b recombination .
here we use @xmath85 derived in @xcite .
since the temperature dependence of @xmath86 is not strong , @xmath87 is assumed everywhere @xcite .
the product @xmath88 is the energy of a @xmath0 photon , 10.2 ev . *
* collisional excitation * : high temperature electrons can excite the quantum state of hydrogen gas by the collision .
due to the large einstein a coefficient , the hydrogen gas can occur de - excitation with the @xmath0 emission .
the @xmath0 emissivity by the collisional excitation is estimated by @xmath89 where @xmath90 is the collisional excitation coefficient , @xmath91 @xcite . once the ionization structure have been determined ( see section [ sec : ion ] ) , we estimate the intrinsic @xmath0 emissivity in each cell by the sum of above @xmath0 emissivity , @xmath92 .
the excitation cooling dominates at @xmath93 , but becomes smaller than the recombination cooling at @xmath94 ( e.g. , * ? ? ?
* ) . there is a large ambiguity in the estimation of the excitation cooling from multi - component ism in sph simulations @xcite .
however , in the present work , we find that the power from stars and agns is always larger than or comparable with the excitation cooling . hence , even the largest possible cooling rate is still sub - dominant to the nebular @xmath0 emission ( i.e. , coming from hii regions around young hot stars ) . in this paper
, we estimate the ionization rate and @xmath0 emissivity for each cell by using the mixed physical quantities of multi - component ism at first , then reduce it by weighting with the mass fraction of cold gas @xmath95 , i.e. , @xmath96 , where @xmath95 can reach @xmath97 depending on the situation . to test our implementation of the @xmath0 rt , we perform some standard tests against analytical solutions , as well as other numerical results in the literature .
1 . neufield test as the first test , we carry out the rt calculation in a dust - free slab of uniform gas .
the uniform slab was analytically studied by @xcite in the optically thick limit . in this case , the emergent @xmath0 line profile is given by @xmath98},\ ] ] where @xmath99 is the optical depth at the line center from mid - plane to the boundary of the slab and @xmath100 is the injection frequency @xmath53 .
profiles of monochromatic line radiation emitted from a dust - free uniform slab .
the black lines are our simulation results , and the color lines represent analytical solutions by @xcite for different optical depths : red @xmath101 , blue
@xmath102 , and green @xmath103 .
the temperature of gas is 10 k. ] figure [ fig : slab ] shows our test result of emergent spectra from a dust - free gas slab with a temperature of @xmath104 and a central plane source , in comparison with analytical solutions for optical depths @xmath105 and @xmath106 , respectively .
the width of the peaks becomes larger with larger @xmath99 , in agreement with the dependence of the optical depth on frequency .
our simulation results agree very well with the analytical solutions , and the agreement becomes better with higher @xmath99 . the dust absorption is crucial in the study of escape fraction , flux , image and profile of @xmath0 from galaxies . @xcite has derived an approximate expression for the escape fraction of @xmath0 photons in a static dusty slab , @xmath107},\ ] ] where @xmath108 is the absorption optical depth of dust , and @xmath109 , with @xmath110 being a fitting parameter .
this solution is valid for extremely optically thick which @xmath111 , and in the limit of @xmath112 .
it suggests that the escape fraction decreases rapidly with increasing @xmath113 .
photons in a dusty slab as a function of @xmath114 .
the solid line is an approximate analytical solution of @xcite , and the open circles are our simulation results . ] in figure [ fig : dustslab ] , we compare our simulation results of @xmath0 escape fraction in a dusty slab with the analytical solutions of @xcite . as is shown , our results agree with the analytical curve very well . 1 .
loeb & rybicki test @xcite analytically derived the intensity field in a spherically symmetric , uniform , radially expanding neutral hydrogen cloud surrounding a central point source of @xmath0 photons .
no thermal motion is included ( @xmath115 k ) . in the diffusion limit ,
the mean intensity @xmath116 as a function of distance from the source @xmath117 and frequency shift @xmath118 is given by @xmath119 with @xmath120 , where @xmath121 is the comoving frequency shift of @xmath0 at which the optical depth to infinity is unity , and @xmath122 is the scaled distance , where @xmath123 is the physical distance at which the doppler shift from the source due to the hubble - like expansion equals @xmath121 .
source in a uniformly expanding medium at various frequencies .
the dotted lines are analytical solutions by @xcite in the diffusion limit , and the color lines are our simulation results . ]
we use a simulation setup similar to that of @xcite for the test .
the results are presented in figure [ fig : lr ] in comparison with the analytical solutions of @xcite .
our simulation results are very close to those from previous tests ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , and are in good agreement with the analytical solutions in regimes where the diffusion limit is valid . however , the simulations diverge from the analytic solutions at larger @xmath117 , where the assumption of optically thick is no longer valid . 1 .
bulk motion test the bulk motion of gas affects the escape of @xmath0 photons as it decreases the effective optical depth @xcite . in practice
, the bulk speed of the gas to the center can be @xmath124 for inflow caused by gravitation , or outflow by supernovae .
the relative velocity can decrease the optical depth significantly .
unfortunately , there is no analytical solution of emergent spectrum for moving medium of @xmath125 k. however , @xcite calculated the emergent spectra of an isothermal and homogeneous sphere undergoing isotropic expansion or contraction . here , we follow the procedure of @xcite to set up the test .
the velocity @xmath126 of a fluid element at a distance @xmath127 from the center is set to be @xmath128 where the hubble - like parameter @xmath129 is fixed such that the velocity increases linearly from 0 in the center to a maximal absolute velocity @xmath130 at the edge of the sphere ( @xmath131 ) : @xmath132 with @xmath130 positive for an expanding sphere , or negative for a collapsing one .
spectrum from an isothermal and homogeneous sphere undergoing isotropic expansion with different maximal velocities at the edge of the sphere .
the initial conditions are the same as those used for figure 8 in @xcite , with a temperature of the sphere @xmath133 k , and the hydrogen column density @xmath134 . ]
figure [ fig : vout ] shows the our test results of the emergent spectrum from an isothermal and homogeneous sphere undergoing isotropic expansion with different maximal velocities @xmath135 at the edge of the sphere . as @xmath135 increases , the blue wing is suppressed while the red wing is broadened . this is because in an outflowing sphere , the photons in the blue wing have a higher probability of encountering hydrogen atoms and being scattered than those in the red wind , which can escape more easily .
this plot suggests that the peak of the profile is pushed further away from the line center for increasing @xmath135 .
however , above a certain threshold value , the peak moves back toward the center again owing to the decrease of optical depth caused by the fast bulk motion .
our results show a good agreement with the previous simulations by @xcite .
ionizing photons from young stars play an important role in galaxy evolution and the cosmic reionization of the intergalactic medium ( igm ) in the early universe .
stellar radiation can ionize interstellar gas , and a fraction of them can escape from galaxies , and reionize the neutral hydrogen ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the ionization structure in igm is highly relevant to the detectability of high-@xmath136 laes ( e.g. , * ? ? ?
* ; * ? ? ?
in particular , ionization affects significantly the properties of distant laes , as the emissivity , scattering and escaping of @xmath0 photons depend sensitively on the ionization state of the ism .
the existing algorithms of ionization rt can be broadly divided into three categories : moments , monte carlo , and ray - tracing methods ( see @xcite for a recent review ) .
recently , @xcite and @xcite conducted a comparison of various cosmological ionization codes .
it was shown that the ray - tracing method provides the most accurate solution but is computationally expensive ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , the moments method is usually diffusive ( e.g. , * ? ? ?
* ) , and the monte carlo method is efficient and flexible but requires a large number of photon packets ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
our implementation of the ionization rt in @xmath1 code is based on the monte carlo method , which is similar to crash @xcite .
the rt of ionizing photons in our code also includes absorption and scattering process by hydrogen gas and interstellar dust .
the optical depth is estimated similar to equation ( [ eq : tau ] ) , @xmath137 , where @xmath138 is the ionizing cross section of hydrogen with lyman limit frequency @xmath139 @xcite .
the time evolution of ionization degree of hydrogen gas is estimated by @xmath140 where @xmath141 , @xmath142 are the photo - ionization efficiency by star and uvb , respectively , @xmath143 is the collisional - ionization efficiency , and @xmath144 is the recombination coefficient . both @xmath143 and @xmath144 are taken from @xcite .
the @xmath145 by uvb is estimated by self - shielding model , in which uvb is optically thin at @xmath146 , zero intensity at @xmath147 .
we set the critical density to be @xmath148 based on the observation data of column density distribution of neutral hydrogen gas @xcite .
the photo - ionization rate by stars and agns is estimated by absorbed - photon number in a cell @xmath149 , i.e. , @xmath150 .
we treat the time integration with an adaptive time step introduced in @xcite , by using three time steps , @xmath151 , @xmath152 and @xmath153 , where @xmath151 is an evolution time step to renew the ionization state .
it is determined by splitting the recombination time scale of volume - mean density by @xmath154 , which is close to that of low - density region . in each @xmath151 , the photon - budget of @xmath155 are traced and the ionization rate is estimated .
the evolution of ionization degree is renewed with @xmath151 , which is divided by @xmath153 , a time step estimated from recombination time - scale in each cell such that the ionization rate is constant . however , when the ratio of the absorbed photon number to that of neutral hydrogen in a cell exceeds a pre - set limit ( @xmath156 in this work ) , the ionization fraction is renewed with @xmath152 .
in addition , the frequency of the photon is sampled from the source spectrum , similar to that in the continuum rt calculation .
we use the stellar population synthesis model of galaxev @xcite to produce intrinsic seds of stars for a grid of metallicity and age , and we use a simple , broken power law for the agn @xcite . a @xcite initial mass function is used in our calculations . to test our code ,
we simulate an hii bubble by a central point source in a hydrogen sphere with uniform density , i.e. , the strmgren sphere @xcite .
the radius of the hii bubble , @xmath157 , is estimated by the balance between the ionization rate of the source and the recombination rate of the gas @xmath158 where @xmath159 is the number of ionizing photons from the source per second , and @xmath85 is the recombination coefficient to all excitation states . and the position of the ionizing front of the hii bubble ,
@xmath160 is analytically derived by @xcite , @xmath161 where @xmath162 is the recombination timescale .
when @xmath163 , @xmath160 approaches to @xmath157 . in the presence of dust , the size of hii bubble decreases due to dust extinction
@xcite has also derived analytically the size of the hii bubble @xmath164 in a dusty medium : @xmath165 where @xmath166 , @xmath167 , and @xmath168 is the optical depth of dust from the source to the distance of strmgren sphere .
( red line ) , and the dotted lines are the analytical solutions by @xcite . ] . ] in the tests , we calculate the ionization structure of a uniform density field .
the source is placed at the corner of the simulation box with @xmath169 .
the number density is @xmath170 , and the recombination coefficient @xmath171 ( @xmath172 ) .
figure [ fig : h2img ] shows the neutral fraction of hydrogen gas at mid - plane at @xmath173 , in a dust - free ( left panel ) and a dusty medium ( right panel ) .
the hii bubble becomes significantly smaller than that in dust - free medium .
figure [ fig : x1dist ] shows the corresponding neutral fraction of hydrogen gas as a function of the distance from the source for both dust - free and dusty tests , and figure [ fig : iprop ] shows the position of the ionization front of the dust - free medium .
all our results are in good agreements with the analytical solutions of @xcite .
our rt calculation is done on a grid . in order to handle arbitrary geometry , and cover a large dynamical range
while resolving small - scale , high - density gas regions in galaxies and ism , we use an adaptive - mesh refinement grid in 3-d cartesian coordinates .
the detailed description of the adaptive grid is given in @xcite .
here we briefly summarize the basic method and procedures .
we typically start with a base grid of @xmath174 box covering the entire simulation volume .
each cell is then adaptively refined by dividing it into @xmath175 sub - cells .
the refinement is stopped if a predefined maximum refinement level , rl , is reached , or if the total number of particles in the cell becomes less than a certain threshold ( typically set to be 32 for sph simulations ) , whichever criterion is satisfied first .
the resolution of the finest level is therefore @xmath176 , where @xmath177 is the box length , and rl is again the maximum refinement level .
the @xmath1 code can be easily applied to either grid- or particle - based simulations , once the snapshot is converted to the format of our grid , which is an effective octo - tree . for particle - based simulations ,
the physical quantities of particles are interpolated onto a grid .
for example , the gas properties at the center of each grid cell , such as density , temperature , and metallicity , are calculated using the sph smoothing kernel of the original simulation .
all physical quantities are assumed to be uniform across a single cell . from the cosmological simulation on top of the gas density distribution of the system .
the box size is 1 mpc in comoving scale .
the maximum refinement level is 11 , and the size of the finest cell is 59.5 pc in physical scale . ]
figure [ fig : plot_grid ] shows an example of the adaptive grid applied to a snapshot redshift @xmath178 from the cosmological simulation used in this work ( see section [ sec_cos ] for details ) . for the parameters used in this application , @xmath179 mpc , rl = 11 , we use a total grid number of only @xmath180 and reach a minimum cell size of @xmath181 pc in physical scale , which corresponds to the spatial resolution of 244 pc in comoving scale of the hydrodynamic simulation .
this grid resolution is equivalent to a @xmath182 uniform grid , which is prohibitive with current computational capabilities .
@xmath1 implements a multi - phase model of the ism @xcite , as adopted in the hydrodynamic simulations @xcite .
it consists of a `` hot - phase '' ( diffuse ) and a `` cold - phase '' ( dense molecular and @xmath183 core ) components , which co - exist under pressure equilibrium but have different mass fractions and volume filling factors ( i.e. , the hot - phase gas is @xmath184 in mass but @xmath185 in volume ) .
the cold gas is assumed to follow two empirical relations observed in giant molecular clouds : a power - law mass distribution ( e.g. , @xcite ) , and a power - law mass - radius relation ( e.g , @xcite ) . within each grid cell in the rt calculation , the hot gas is uniformly distributed , while the cold , dense cores are randomly embedded .
the dust is assumed to be associated with both the cold and hot phase gases , and it follow the distribution of the gas with certain dust - to - gas ratio . in the rt calculation , when a photon enters a cell ,
we first determine the distance it travels in the hot phase gas before hitting a cold cloud , then the distance it travels in the cold cloud before it is absorbed .
the combination of the two gives the total optical depth in the multi - phase ism , which is then compared with a random number to determine whether the photon should be stopped for scattering or absorption ( see @xcite for a detailed description and implementation ) .
the @xmath0 rt in a multi - phase ism and dust model is one of the new features in our rt code compared with previous works . in @xmath1 , depending on the hydrodynamic simulation and the nature of the calculation , we can choose either a multi - phase or a single - phase model , in which the gas and the dust are uniformly distributed with the mean density in a cell , for the ism .
galaxies and quasars at high redshifts show extinction by interstellar dust @xcite . in the early universe , most of the dust
is thought to be created by type ii supernovae ( sne ) . theoretically ,
several groups have calculated dust formation in the ejecta of type ii sne @xcite . in particular , @xcite have developed a dust model based on standard nucleation theory and tested it on the well - studied case of sn1987a .
they find that sne with masses in the range of @xmath186 produce about 0.01 of the mass in dust per supernova for primordial metallicity , and @xmath187 for solar metallicity . in @xmath1
, we adopt the dust size distribution of @xcite for solar metallicity and a @xmath188 sn model , as in figure @xmath189 in their paper . the size distribution
is then combined with the dust absorption and scattering cross section of @xcite to calculate dust absorption opacity curves .
the sn models include silicate fractions @xmath190 , which shows the silicate feature @xmath191 ( see the resultant opacity curve @xcite ) .
the dust - to - gas mass ratio in each cell is linearly proportional to the metallicity in the cell .
the ratio is normalized to galactic value when the metallicity is the solar abundance . to date
, there is little work on the coupling of continuum and @xmath0 line transfer . in the seminar work of @xcite
, they incorporated pre - computed tables of the properties of the @xmath0 photons in the calculation of ionizing uv continuum radiation .
our approach differs from theirs in that we follow the actual propagation and scattering of the @xmath0 photons in the medium , and take into account the effects of ionization and dust absorption . in @xmath1 ,
the photons are treated in three regimes : ionizing continuum at @xmath192 , @xmath0 line with center @xmath193 , and non - ionizing continuum at @xmath194 .
we follow the following steps to couple the radiative transfer of these three components : 1 .
the gas and dust content of each grid cell is determined before the rt process .
the neutral hydrogen fraction @xmath195 in each cell is determined by following the ionizing continuum photons .
the resulting ionization structure is then input to calculate the rt of @xmath0 photons .
the absorption of the ionizing photons by dust will be taken into account in the calculation of the dust re - emission in longer wavelengths of the non - ionizing continuum .
the monte carlo rt method follows the propagation , scattering , absorption , and reemission of photon packets " ( groups of equal - energy , monochromatic photons that travel in the same direction ) , by randomly sampling the various probability distribution functions that determine the optical depth , scattering angle , and absorption rates . in detail
, the monte carlo ray - tracing procedure involves the following steps : 1 .
a photon packet of continuum is emitted from either a stellar source or an accreting black hole with random frequencies consistent with the source spectra . in the case of @xmath0 photon
, it is emitted from ionized hydrogen gas .
the photon is emitted with a uniformly distributed random direction .
the probability of a photon being emitted by any given source is determined by its luminosity relative to the total .
2 . a random optical depth over which the photon must travel before an interaction with gas or dust occurs , @xmath196 , is drawn to determine the interaction location .
the interaction includes scattering and absorption . in our method ,
the photon energies are not weighted , only one event is allowed .
that is , at any given interaction site , the photon is either scattered or absorbed , but not both .
starting from the location of the photon emission , the cumulative optical depth of the photon , @xmath197 , is calculated stochastically for both hot and cold ism in a mulit - phase model , or single - phase medium with the mean density along the ray . if the photon is stopped for interaction within a single cell , then @xmath197 is the sum of contributions from possibly multiple segments of both hot and cold for the multi - phase model , or a single - phase gas / dust within this cell . if the photon passes through multiple cells before an interaction occurs , then @xmath197 is the sum of all contributions from relevant segments in these cells .
4 . at each boundary between the hot and cold phase gas clouds , or at the boundary of the grid cell , the next interaction point
is determined by the comparison between @xmath198 and @xmath197 . if @xmath199 , then the photon is either scattered or absorbed , with a probability given by the scattering albedo . the exact interaction location is then determined inside either hot or cold phase gas , such that @xmath197 becomes exactly @xmath198 .
if the photon is scattered , its direction is altered using a phase function , and the ray tracing of the new photon is repeated from step 2 .
the phase function is the henyey - greenstein for dust , and simply isotropic for gas .
if the photon is absorbed by dust , the dust temperature is raised , and a new photon is re - emitted according to the scheme described below .
the ray tracing of the newly emitted photon again restarts from step 2 . 5 .
if the photon escapes from the system without reaching the optical depth @xmath198 , it is then collected in the output spectrum and image .
the next photon will be picked up from the source , and the whole monte carlo procedure from step 1 will be restarted . in @xmath1
, we incorporate a `` peeling off '' procedure using weighted photons to obtain high - resolution images from a particular viewing direction . from a particular viewing direction , the contribution from direct ( radiation source or thermal dust emission ) and scattered
photons are tracked separately with different weight . at each location of emission ,
the contribution from direct photons to the image is computed as @xmath200 where @xmath201 is the energy of each photon packet , @xmath53 and @xmath202 are the position of the emission site projected onto the image plane , and @xmath203 and @xmath204 are the total optical depth and distance from the emission site to the observer , respectively . at each scattering site ,
the contribution from scattered photons to the image is @xmath205 where @xmath206 is the optical depth from the scattering location to the observer , and @xmath207 is the scattering phase function .
the total intensity of the image is the sum of @xmath208 and @xmath209 , and is accumulated as each photon packet is emitted and scattered during the course of the monte carlo simulation . to obtain an image for a given telescope at a given waveband , a specific filter function
is then applied to convolve the image .
the cosmological simulation presented here follows the formation and evolution of a milky way - size galaxy and its substructures .
the simulation includes dark matter , gasdynamics , star formation , black hole growth , and feedback processes .
the initial condition is originally from the aquarius project @xcite , which produced the largest ever particle simulation of a milky way - sized dark matter halo .
the hydrodynamical initial condition is reconstructed from the original collionsionless one by splitting each original particle into a dark matter and gas particle pair , displaced slightly with respect to each other ( at fixed center of mass ) for a regular distribution of the mean particle separation , and with a mass ratio corresponding to a baryon fraction of 0.16 @xcite .
the whole simulation falls in a periodic box of 100 @xmath210 on each side with a zoom - in region of a size @xmath211 .
the spatial resolution is @xmath212 pc in the zoom - in region .
the mass resolution of this zoom - in region is @xmath213 for dark matter particles , @xmath214 for gas and star particles .
the cosmological parameters used in the simulation are @xmath215 , @xmath216 , @xmath217 and @xmath218 , consistent with the five - year results of the wmap @xcite .
the simulation evolves from @xmath219 to @xmath220 .
the simulation was performed using the parallel , n - body / smoothed particle hydrodynamics ( sph ) code gadget-3 , which is an improved version of that described in @xcite and @xcite . for the computation of gravitational forces
, the code uses the `` treepm '' method @xcite that combines a `` tree '' algorithm @xcite for short - range forces and a fourier transform particle - mesh method @xcite for long - range forces .
gadget implements the entropy - conserving formulation of sph @xcite with adaptive particle smoothing , as in @xcite .
radiative cooling and heating processes are calculated assuming collisional ionization equilibrium @xcite .
star formation is modeled in a multi - phase ism , with a rate that follows the schmidt - kennicutt law ( @xcite ) .
feedback from supernovae is captured through a multi - phase model of the ism by an effective equation of state for star - forming gas @xcite .
the uv background model of @xcite is used . black hole growth and feedback
are also included in our simulation based on the model of @xcite , where black holes are represented by collisionless `` sink '' particles that interact gravitationally with other components and accrete gas from their surroundings .
the accretion rate is estimated from the local gas density and sound speed using a spherical bondi - hoyle @xcite model that is limited by the eddington rate .
feedback from black hole accretion is modeled as thermal energy , @xmath221 of the radiation , injected into surrounding gas isotropically , as described in @xcite and @xcite .
this feedback scheme self - regulates the growth of the black hole and has been demonstrated to successfully reproduce many observed properties of local elliptical galaxies ( e.g , , @xcite ) and the most distant quasars at @xmath9 @xcite .
we follow the black hole seeding scheme of @xcite and @xcite in the simulation : a seed black hole of mass @xmath222 was planted in the gravitational potential minimum of each new halo identified by the friends - of - friends ( fof ) group finding algorithms with a total mass greater than @xmath223 .
each snapshot is processed by an on - flying fof group finding algorithm with a dark matter linking length less than 20% of their mean spacing .
other types of particles are then linked to the nearest dark matter particle .
a substructure detection algorithm subfind @xcite is then applied to each group to search satellite halos . in the simulation ,
each fof group with its substructure is identified as a galaxy ( a detailed description of the simulation is presented in zhu et al . , in preparation ) . in this work
, we apply the @xmath1 code to the ten most massive galaxies in each snapshot at six redshift bins from @xmath224 and focus on the @xmath0 and multi - band properties . in our post - processing procedure ,
we first calculate the rt of ionizing photons ( @xmath225 ) and estimate the ionization fraction of the ism .
the resulting ionization structure is then used to run the @xmath0 rt to derive the emissivity , followed by the calculation of non - ionizing continuum photons ( @xmath226 ) in each cell .
our fiducial run is done with @xmath227 photon packets for each ionizing , @xmath0 , and non - ionizing components . because the spatial resolution of the cosmological simulation is not adequate to resolve the multiple phase of the ism
, we assume a single - phase medium in each density grid .
the highest refinement of the grid is @xmath228 .
the galaxies in our sample are the 10 most massive ones from the cosmological simulation at six snapshots from @xmath2 .
note that because our simulation focuses on a milky way - size galaxy , our sample represents the progenitors of the milky way at different cosmic time , they are not the typical `` massive '' galaxies one expect from a general cosmological simulation with a mean overdensity .
the total and stellar masses of these galaxies become larger with time ( decreasing redshift ) due to gas accretion and merging process .
for example , the total mass of the most massive galaxy is @xmath229 , @xmath230 , @xmath231 , and @xmath232 at @xmath233 and @xmath234 respectively , and the corresponding stellar mass is @xmath235 , @xmath236 , @xmath237 , and @xmath238 .
figure [ fig : img_cosmo ] ( left panels ) show the projected density of gas and stars of each snapshot at the mentioned redshift .
filamentary structure is clearly seen , and the most massive galaxy resides in the intersection of the filaments .
the resulting @xmath0 and near infrared ( nir ) fluxes of the most massive galaxy from our rt calculations are shown in the middle and right panels of figure [ fig : img_cosmo ] , respectively .
the nir band corresponds to irac-1 ( @xmath239 ) of the _ spitzer _ telescope .
the flux image is generated from a `` peeling off '' technique in the rt calculation as described in section [ sec_peeloff ] .
the @xmath0 emission appears to follow the distribution of gas and stars of the galaxy .
its surface brightness and size increase with time ( decreasing redshift ) , while those of the nir show different pattern .
the @xmath0 emission depend on the star formation rate ( sfr ) of the galaxy and the photon escape fraction .
although sfr decreases at lower redshift , the escape fraction increases as the galaxy grows larger , resulting in an increase of @xmath0 luminosity .
in addition , the @xmath0 emission is bright over extended region , owing to many scattering processes .
on the other hand , the observed irac-1 in these redshifts corresponds to wavelengths @xmath240 in the rest frame , hence it depends on both sfr and stellar mass of the galaxy .
the cross section of dust decreases steeply with increasing wavelength , it becomes very small at these wavelengths .
therefore , a significant fraction of photons emitted by the stars in this wavelength range can escape without being scattered or absorbed by dust .
thus , unlike @xmath0 , the irac-1 image is more compact , and it concentrate on the high density peak of stars .
we note that the infrared fluxes are higher than those of the lae sample at @xmath178 by @xcite , but comparable with those of the irac - detected laes in @xcite
. this may be due to high sfr in these galaxies .
the nir wavelength shown here is similar to the f356w filter of the next generation telescope , the _ james webb space telescope _ ( jwst ) .
the detection limit of the filter in ab magnitude can achieve @xmath241 with 1 hour exposure time ( e.g. , * ? ? ?
our calculations show that the model galaxies have @xmath242 , 28.5 ( z = 8.5 ) , and 29.8 ( z = 10.2 ) at this wavelength .
hence , the jwst will be able to detect the infrared emission from galaxies such as the progenitor of the milky way at @xmath9 with a @xmath243 minute exposure time . even at higher redshifts
, these galaxies may be detected with an integration of @xmath244 hours for @xmath245 , and @xmath246 hours for @xmath247 .
the corresponding multi - wavelength seds of the most massive galaxy at selected redshift are shown in figure [ fig : sed_cosmo ] .
the shape of the sed changes significantly from @xmath247 to @xmath248 , as a result of changes in radiation from stars , absorption of ionizing photons by gas and dust , and re - emission by dust in the infrared .
the @xmath0 line appears to be strong in all cases .
the deep decline of uv continuum at @xmath249 is caused by strong absorption of ionizing photons by the dense gas .
galaxy at lower redshift has a higher floor of continuum emission from stars and accreting bhs , a higher ionization fraction of the gas , and a higher infrared bump owing to increasing amount of dust and absorption .
moreover , due to the effect of negative k - correction , the flux at @xmath250 stays close in different redshifts .
our calculations show that the model galaxies have a flux of @xmath251 , @xmath252 , and @xmath253 at @xmath254 in observed frame .
the new radio telescope , _ atacama large millimeter / submillimeter array _ ( alma ) may be able to detect such galaxies at @xmath9 with @xmath244 hours integration , and at @xmath255 for @xmath256 hours with 16 antennas . even with alma , it may be very difficult to detect such a galaxy at @xmath247 ( more than 70 days integration ) .
( green ) , 8.5 ( orange ) , 7.2 ( blue ) , 6.2 ( purple ) , and 3.1 ( black ) , respectively . ] from the seds , it is interesting to compare the emergent @xmath0 luminosity and that of the infrared integrated from 8 to 1000 @xmath257 in the rest frame .
a simple relation between infrared and @xmath0 luminosity has been suggested by @xcite through star formation , namely @xmath258 and @xmath259 with the assumption of @xmath260 ( case b ) .
however , our calculation suggests a more complex relation between @xmath0 and infrared luminosity , as inferred from figure [ fig : sed_cosmo ] .
this is because @xmath261 depends on @xmath262 , while the @xmath263 depends on @xmath264 , and the two escape fractions differ from each other . in our simulations ,
the @xmath264 mostly decreases with redshift , and is below 0.5 at @xmath265 .
hence , the uv continuum may serve as a good tracer of sfr at @xmath266 , while @xmath265 the infrared flux may be one .
in addition , since the modeled galaxies at @xmath265 show a smaller value of @xmath262 than 0.5 ( see next subsection ) , hence @xmath0 flux also may not be a good indicator of sfr .
properties of modeled galaxies at different redshifts , in clock wise direction : star formation rate , emergent @xmath0 luminosity , equivalent width of @xmath0 line in rest frame , and photon escape fraction of @xmath0 ( open circles ) and uv continuum ( 1300 - 1600 @xmath267 , crosses ) .
red and blue filled circles are the values of the most massive galaxies and the median in our sample at each snapshot , respectively . filled and open red circles in the panel of escape fraction shows the @xmath0 and uv continuum escape fraction of the most massive galaxies , filled and open blue circles are the median values of @xmath0 and uv continuum , respectively . ]
the derived properties of @xmath0 emission from these modeled galaxies are shown in figure [ fig : flux_cosmo ] , in comparison with their corresponding sfr .
the most massive galaxies at very high redshift ( @xmath268 ) maintain a high sfr of @xmath269 fueled by accretion of cold gas and merging process .
the sfr decreases to @xmath270 at @xmath7 owing to depletion of star - forming gas and suppression by radiation feedback from both stars and accreting bhs .
the resulting emergent @xmath0 luminosity from recombination of ionizing photons and from excitation cooling increases from @xmath271 at @xmath272 to @xmath273 at @xmath178 .
the observed lae luminosity functions show a characteristic luminosity @xmath274 at @xmath275 @xcite , and @xmath276 at @xmath178 @xcite , which are comparable to the luminosity range of our modeled galaxies .
the median values of sfr moderately decrease with increasing redshift , and show @xmath277 .
however , the @xmath261 does not decreases largely by the redshift evolution of @xmath79 .
in addition , a few galaxies in each snap shot are @xmath0 bright , and have @xmath0 luminosity of @xmath278 , which is detectable with recent observational studies ( e.g. , * ? ? ? * ) .
the lower - left panel of figure [ fig : flux_cosmo ] shows the photon escape fraction of @xmath0 , @xmath279 and uv continuum @xmath280 , where @xmath280 is calculated at @xmath281 . both escape fractions appear to increase with time ( decreasing redshift ) , with @xmath282 at @xmath272 to @xmath283 at @xmath178 , while @xmath284 at @xmath272 to @xmath285 at @xmath178 .
in the early evolution phase , the most massive galaxies are compact and gas - rich .
they grow rapidly from accretion of cold gas which fueled intense star formation .
the highly concentrated gas and dust efficiently absorb the @xmath0 and uv photons , and suppress their escape . at a later time , the gas and stars distribute to a more extended region , and a large fraction of gas is converted to stars .
hence , the photon may easily escape from the galaxy , contributing to a higher photon escape fraction .
on the other hand , the median value of @xmath79 of @xmath0 and uv increases monotonically with redshift , e.g. , @xmath286 at @xmath272 to @xmath287 at @xmath178 , while @xmath288 at @xmath272 to @xmath289 at @xmath178 .
such a trend is consistent with observations by @xcite . unlike the most massive galaxies , the low - mass ones have lower sfr and lower dust content , the ism is less dense and less clumpy , and the metallicity is lower . as a result , the escape fractions are higher in small galaxies than the massive ones .
the ew of @xmath0 line is defined by the ratio between the @xmath0 flux and the uv flux density @xmath290 in rest frame , where the mean flux density of @xmath291 in rest frame is used . in practice ,
the ew is estimated by @xmath292 , where @xmath293 is the intrinsic equivalent width .
the @xmath293 depends on the stellar sed and the excitation @xmath0 cooling : the younger sed ( or top - heavy imf ) and the more efficient @xmath0 cooling , the higher @xmath293 .
the resulting @xmath0 ews of the most massive galaxies are shown in the lower right panel of figure [ fig : flux_cosmo ] .
since the photon escape fraction of @xmath0 is close to that of the uv continuum , the ew shown is basically @xmath293 .
all three modeled galaxies have @xmath294 , they are therefore classified as laes ( e.g. , * ? ? ?
the ew evolves gradually with redshift . at @xmath268 ,
the ews are high , @xmath295 , owing to strong @xmath0 emission from excitation cooling and agn activity , while at @xmath178 , it drops to @xmath296 as the @xmath0 emission is mainly produced by stellar radiation .
the median value increases more steeply with redshift , from @xmath297 at @xmath178 to @xmath298 at @xmath245 .
these ews are within the observed range , and the trend is in broad agreement with observations that galaxies at higher redshifts appear to have higher ew than their counterparts at lower redshifts ( e.g. , * ? ? ?
* ; * ? ? ?
some galaxies at higher redshift @xmath299 show very high ew because the excitation @xmath0 cooling becomes dominant @xcite .
however , it will be suppressed significantly by igm at such a high redshift .
the recent discovery of the most distant lae , udfy-38135539 , at @xmath5 by @xcite marks a milestone in observational cosmology that galaxies form less than 600 million years after the big bang .
the detected @xmath0 line flux of this object is @xmath300 , which implies a @xmath0 luminosity of @xmath301 .
this is close to that of our galaxy at @xmath272 .
our model suggests that this lae may have a total mass of @xmath302 , a stellar mass of @xmath303 , a sfr of the order of @xmath304 , and @xmath282 .
we note that @xcite suggested a lower sfr of @xmath305 for this galaxy by assuming a @xmath0 escape fraction of 100 per cent .
however , our detailed rt calculations show that the escape fraction is much smaller than unity , which means it would require a higher sfr in order to produce the observed @xmath0 flux
. profile .
dot and solid lines are the profiles of intrinsically emitted and escaped photons respectively . ]
figure [ fig : profile_cosmo ] shows the profiles of @xmath0 line of the modeled galaxies .
the frequency of the intrinsic @xmath0 photon is sampled from a maxwellian distribution with the gas temperature at the emission location in the rest flame of the gas .
we collect all escaped @xmath0 photons and estimate the line profile . in practice ,
the inhomogeneous gas structure can cause the change of flux depending on the viewing angle .
although the @xmath0 flux can change by some factors @xcite , we focus on only the mean value in this paper . the @xmath0 lines of our galaxies show mostly a single peak , a common profile of laes at high redshifts ( e.g. , * ? ? ? * ) . in a static and optically thick medium
, the profile can be double peaked , but when the effective optical depth is small due to high relative gas speed or ionization state , there might be only a single peak @xcite . in our case ,
the flow speed of gas is up to @xmath306 , and the gas is highly ionized by stellar and agn radiation , which result in a single peak .
in addition , the profile at higher redshift shifts to shorter wavelength , and it shows the characteristic shape of gas inflow @xcite .
although our simulation includes feedback of stellar wind similar to that of @xcite , the @xmath0 line profile indicates gas inflow or symetric in these galaxies .
our result suggests that high - redshift star - forming galaxies may be fueled by efficient inflow of cold gas from the filaments .
we will address this issue in detail in a forthcoming paper ( yajima et al . in preparation ) .
we note that there is a number of studies on inflow and outflow in @xmath0 lines , both observationally @xcite and theoretically ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
it has been suggested that the igm correction of @xmath0 emission may be enhanced due to inflow ( e.g. , * ? ? ? * ) or decreased via outflows ( e.g. , * ? ? ?
however , there is a large uncertainty in the prescription of wind feedback in current simulations . to investigate the velocity field of ism in high - redshift galaxies
, one needs simulations with a realistic wind model and a high spatial resolution comparable to that of line observations , which are currently unavailable and are beyond the scope of this paper .
moreover , we should point out that the line profile can change due to the difference of the sub - grid ism models .
in addition , the line profile may be highly suppressed and changed by scattering in the intergalactic medium ( igm ) ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the igm effectively absorbs the @xmath0 photons at the line center and at shorter wavelengths by the hubble flow ( e.g. , * ? ? ?
. therefore , the inflow feature in our profiles may be disappeared and the shape may change to an asymmetric single peak with only photons at red wing .
we will investigate the igm correction in future work that includes propagation and scattering of @xmath0 and ionizing photons in the igm . .
red and blue filled circles are the values of the most massive galaxies and the median in our sample at each snapshot , respectively . ]
figure [ fig : fesc_ion ] shows the escape fraction of ionizing photons , @xmath307 , and the resulting emissivity from both stars and agns of the modeled galaxies .
the escape fraction of the most massive galaxies has a range of @xmath308 , consistent with the results of @xcite .
in addition , @xmath307 decreases with increasing redshift .
unlike the @xmath0 and non - ionizing uv photons , the ionizing photons can be absorbed by hydrogen gas and dust . at high redshift ,
the star formation efficiency is small , so a large fraction of hydrogen gas remains neutral and prevent the escape of ionizing photons .
the median values do not decrease with redshift due to lower dust mass .
the emissivity of the ionizing photons roughly increase with decreasing redshift , reaching @xmath309 at @xmath178 . in this work ,
we do not calculate the rt in the igm , but the size of the hii bubble around the galaxy may be estimated by a modified strmgrem sphere . assuming a clumpiness factor of @xmath310 @xcite , the radius of the hii bubble is estimated to be @xmath311 and @xmath312 kpc at @xmath313 and @xmath314 , respectively .
the @xmath0 photons may pass through the hii bubbles , resulting in a correlation between high - redshift laes and hii regions , as suggested by @xcite .
in addition , the median value of ionizing - photons emissivity is smaller due to lower sfr , and hence the hii bubble becomes smaller than that of the most massive galaxy by a factor of @xmath315 .
in observations of @xmath0 emission from high - redshift galaxies , there are a number of uncertainties regarding the contributions from agns ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , stars , and excitation cooling from cold accretion . with our model , we can separate the radiation components and disentangle the contribution from different sources .
also , in numerical simulations , resolution plays an important role in the robustness of the results .
so we discuss these physical and numerical effects on @xmath0 emission in this section .
photons ( top panel ) and the emergent luminosity ( bottom panel ) .
the ionization and @xmath0 luminosity are estimated with radiation from the following sources : uvb only ( triangles ) , uvb + star ( squares ) , and uvb + agn + star ( circles ) . ] the effects of agn and stars on the @xmath0 emission is shown in figure [ fig : fesc_comp_star ] .
it appears that the presence of agn does not change significantly the @xmath0 escape fraction .
it is because agns are usually located in galactic center where the gas density peaks , most of the ionizing photons near lyman - limit ( @xmath316 ) are trapped there ( e.g. , * ? ? ?
* ; * ? ? ?
* ) and can not escape . on the other hand
, a fraction of high energy photons can escape from the galaxy without ionizing interstellar gas due to small cross section of gas and dust .
therefore , agn radiation has little effect on the @xmath79 of @xmath0 photons .
there is no change in @xmath79 in the case where there is no stellar and agn radiation .
in addition , the agn does not affect the @xmath0 luminosity , as shown in the bottom panel of figure [ fig : fesc_comp_star ] .
this may be because the bhs in these galaxies are still quite small and the accretion rates are quite low . on the other hand , the presence of stars appears to change the escape fraction and emergent luminosity of @xmath0 , because the stellar radiation is much stronger than that of agn in the model galaxies , and the effect becomes stronger at lower redshift . because @xmath0 emissivity is produced by the recombination of ionizing photons and collisional excitation from hydrogen gas , the interplay between these two mechanisms determines the outcome . at redshifts @xmath268 ,
excitation cooling from cold accretion of dense gas becomes comparable to recombination cooling , while at @xmath7 , the number of @xmath0 photons by recombination in hii regions by stars and agns outnumbers that from excitation cooling .
therefore , the contribution from stars to the @xmath0 emission changes with cosmic time and environment .
the @xmath0 emissivity from excitation cooling can be reduced due to the stellar ionization ( e.g. , * ? ? ?
* ) . moreover , the distribution of excitation @xmath0 cooling can be more extended than star - forming regions , resulting in higher escape fraction .
thus , the escape fraction and emergent luminosity of @xmath0 somewhat increase in the case without stellar radiation .
figure [ fig : fesc_comp_n ] demonstrates effects of photon number on the escape fraction of @xmath0 photons , with different number of photon packets , @xmath317 and @xmath318 .
it appears that the results converge once the photon number is above @xmath319 , which is an order of magnitude lower than the fiducial @xmath227 in our general rt calculations .
the relative difference between @xmath320 and @xmath318 is 1 - 6 per cent .
of course , the photon number should increase with galaxy mass , but for our modeled galaxies , @xmath227 should be sufficient . photons .
open triangles , squares and circles indicate photon packet number @xmath321 , and @xmath318 , respectively .
the fiducial number used in our general rt calculations is @xmath227 . ]
different color indicates different maximum refinement level ( rl ) : purple rl = 4 , blue
rl = 6 , green rl = 8 , orange rl = 10 , and red
the fiducial rl used in our general rt calculations is 11 .
the cell size of the highest refinement level is @xmath322 mpc in comoving scale . ]
the effects of grid resolution on the escape fraction of @xmath0 photons is presented in figure [ fig : fesc_comp_l ] , where results from different maximum refinement level ( rl ) are compared .
it appears that @xmath79 changes dramatically with rl , but it converges at @xmath323 .
grids with poor resolution can produce artificially high @xmath79 by a factor of more than 20 . because most of the @xmath0 photons are created in high - density gas clumps , where they are also effectively absorbed by numerous scatterings , sufficiently resolving
the gas density field is important in the calculation of @xmath0 rt .
if we follow the resolution of @xmath324 with uniform grids , it would require a number of cells of @xmath325 , too expensive for rt simulations with the current computational facility .
in fact , the current limit is @xmath326 , as demonstrated by recent work on three - dimensional rt calculations using uniform grid ( e.g. , * ? ? ?
* ; * ? ? ?
. therefore , adaptive refinement grid is critical in the rt calculation of @xmath0 emission .
we have presented an improved , three - dimensional , monte carlo radiative transfer code @xmath1 to study multi - wavelength properties of galaxies .
it has the following essential features : * it couples @xmath0 line , ionization of neutral hydrogen , and multi - wavelength continuum radiative transfer .
the @xmath0 module includes emission from both recombination and collisional excitation , and the continuum includes emission from x - ray to radio . *
it employs an adaptive grid scheme in 3-d cartesian coordinates , which handles an arbitrary geometry and covers a large dynamic range over several orders of magnitude .
such a grid is critical in resolving the clumpy , highly inhomogeneous density field in the rt calculations . *
it adopts a two - phase ism model in which the cold , dense clouds are embedded in a hot , diffuse medium in pressure equilibrium . in the case where hydrodynamic simulations have sufficient resolution to resolve the multi - phase ism
, this model ensures an appropriate prescription of the ism physics , which is important for studying dust properties in galaxies . *
it includes a supernova - origin dust model in which the dust is produced by type - ii supernovae , which is especially relevant for dust in high - redshift , young objects , because there are insufficient low - mass , old agb stars in these systems to produce the observed amount of dust from the classical dust models .
* it follows the radiation from stars and agns , and calculates the scattering , absorption , and re - emission of the surrounding medium .
it efficiently and self - consistently produces seds and images in a full spectrum , including @xmath0 emission line and continuum from x - ray to millimeter , as well as ionization structures of the medium , for direct comparison with multi - band observations . *
it can be easily applied to either grid- or particle - based hydrodynamical simulations and has broad applications in cosmology .
we have tested the new @xmath1 extensively , and found that the coupling of @xmath0 line and continuum enables a self - consistent and accurate calculation of the @xmath0 properties of galaxies , as the equivalent width of the @xmath0 line depends on the uv continuum , and the escape fraction of @xmath0 photons strongly depends on the ionization structure and the dust content of the object .
we applied the code to a cosmological sph simulation of a milky way - like galaxy , and studied in details a sample of massive galaxies at redshift z=3.1 - 10.2 .
we find that most of these galaxies are @xmath0 emitters .
the escape fraction of @xmath0 photons of the most massive galaxy in each snap shot increases with decreasing redshift , from about 0.18 at redshift @xmath255 to 0.6 at @xmath7 .
the emergent @xmath0 luminosity shows similar trend , increasing from @xmath327 at @xmath255 to @xmath328 at @xmath7 .
the equivalent widths of these galaxies change with redshift as well , from @xmath329 at @xmath255 to @xmath330 at @xmath7 .
the profiles of the resulting @xmath0 lines commonly show single peak , and the peak of the profile at @xmath266 is shifted to shorter wavelength , implying gas inflow from surrounding filaments . on the other hand ,
the median values in our galaxy sample shows somewhat different trend from that of the most massive ones .
the escape fraction monotonically increases with redshift , while the @xmath0 luminosity shows a weak evolution , as a result of the interplay between sfr and escape fraction .
our results suggest that the laes and their properties evolve with cosmic time , and that the first laes in the early universe , such as currently the most distant one detected at @xmath3 @xcite , may be dwarf galaxies with a total mass of @xmath302 fueled by cold gas accretion .
they may have a star formation rate of order of @xmath331 , and a @xmath0 escape fraction of @xmath256 per cent .
we thank mark dijkstra , claude - andr faucher - gigure , caryl gronwall , robin ciardullo , lars hernquist , and avi loeb for stimulating discussions and helpful comments , as well as the referee for an insightful and constructive report which has helped improve the manuscript significantly . support from nsf grants
ast-0965694 , ast-1009867 ( to yl ) , and ast-0807075 ( to ta ) is gratefully acknowledged .
yl thanks the institute for theory and computation ( itc ) at harvard university where the project was started for warm hospitality .
we acknowledge the research computing and cyberinfrastructure unit of information technology services at the pennsylvania state university for providing computational resources and services that have contributed to the research results reported in this paper ( url : http://rcc.its.psu.edu ) .
the institute for gravitation and the cosmos is supported by the eberly college of science and the office of the senior vice president for research at the pennsylvania state university .
a. , parry i. , bland - hawthorn j. , cianci s. , king d. , mcmahon r. , medlen s. , 2004 , in society of photo - optical instrumentation engineers ( spie ) conference series , vol . 5492 , society of photo - optical instrumentation engineers ( spie ) conference series , a. f. m. moorwood & m. iye , ed .
, pp . 10221032 w. h. , teukolsky s. a. , vetterling w. t. , flannery b. p. , 1992
, numerical recipes in fortran . the art of scientific computing , press , w. h. , teukolsky , s. a. , vetterling , w. t. , & flannery , b. p. , ed | narrow - band @xmath0 line and broad - band continuum have played important roles in the discovery of high - redshift galaxies in recent years .
hence , it is crucial to study the radiative transfer of both @xmath0 and continuum photons in the context of galaxy formation and evolution in order to understand the nature of distant galaxies . here
, we present a three - dimensional monte carlo radiative transfer code , all - wavelength radiative transfer with adaptive refinement tree ( @xmath1 ) , which couples @xmath0 line and multi - wavelength continuum , for the study of panchromatic properties of galaxies and interstellar medium . this code is based on the original version of li et al . , and features three essential modules : continuum emission from x - ray to radio , @xmath0 emission from both recombination and collisional excitation , and ionization of neutral hydrogen .
the coupling of these three modules , together with an adaptive refinement grid , enables a self - consistent and accurate calculation of the @xmath0 properties , which depend strongly on the uv continuum , ionization structure , and dust content of the object .
moreover , it efficiently produces multi - wavelength properties , such as the spectral energy distribution and images , for direct comparison with multi - band observations . as an example
, we apply @xmath1 to a cosmological simulation that includes both star formation and black hole growth , and study in detail a sample of massive galaxies at redshifts @xmath2 .
we find that these galaxies are @xmath0 emitters ( laes ) , whose @xmath0 emission traces the dense gas region , and that their @xmath0 lines show a shape characteristic of gas inflow .
furthermore , the @xmath0 properties , including photon escape fraction , emergent luminosity , and equivalent width , change with time and environment .
our results suggest that laes evolve with redshift , and that early laes such as the most distant one detected at @xmath3 may be dwarf galaxies with a high star formation rate fueled by infall of cold gas , and a low @xmath0 escape fraction .
[ firstpage ] radiative transfer line : profiles ism : dust , extinction galaxies : evolution galaxies : formation galaxies : high - redshift |
the rise of deep large area surveys such as 2mass ( the two micron all sky survey ) , sdss ( the sloan digital sky survey ) and ukidss ( the uk infrared deep sky survey ) has increased the number of known brown dwarfs to several hundred , since the discovery of gliese 229b @xcite and tiede 1 @xcite just over a decade ago .
these populations have helped shape our understanding of ultracool dwarfs ( ucds ; with spectral type @xmath15m7 e.g. @xcite ) and revolutionized the classification system for sub - stellar objects including the creation of two new spectral types l and t. the latest m dwarfs ( @xmath16m7 - 9 ) have effective temperature ( @xmath3 ) reaching down to @xmath162300k . at lower @xmath3 ( @xmath162300 - 1400k )
are the l dwarfs , which have very dusty upper atmospheres and generally very red colours .
t dwarfs are even cooler having @xmath3 in the range @xmath161400 - 650k , where the low @xmath3 limit is currently defined by the recently discovered t8@xmath17 dwarfs , ulas j0034 - 0052 @xcite and cfbds j005910.90 - 011401.3 @xcite .
t dwarf spectra are dominated by strong water vapour and methane bands , and generally appear very blue in the near infrared ( @xcite ; @xcite ) .
the physics of ultracool atmospheres is complex , and very difficult to accurately model .
atmospheric dust formation is particularly challenging for theory ( @xcite ; @xcite ) , and there are a variety of other important issues that are not well understood , including the completeness of ch@xmath18/h@xmath19o molecular opacities , their dependence on @xmath3 , gravity and metallicity ( e.g. @xcite ; @xcite ; @xcite ) , as well as the possible presence of vertical mixing in such atmospheres @xcite .
the emergent spectra from ultracool atmospheres are strongly affected by factors such as gravity and metallicity ( e.g. @xcite ; @xcite ; @xcite ) , and we must improve our understanding of such effects if we are to be able to spectroscopically constrain atmospheric and physical properties ( such as mass , age and composition ) of low - mass / sub - stellar field populations .
discovering ucds whose properties can be inferred indirectly ( without the need for atmospheric models ) is an excellent way to provide a test - bed for theory , and observationally pin down how physical properties affect spectra .
we refer to such ucds as _ benchmark _ objects ( e.g. @xcite )
. a population of benchmark ucds with a broad range of atmospheric properties will be invaluable in the task of determining the full extent of spectral sensitivity to variations in ucd physical properties .
however , such benchmarks are not common , and the constraints on their properties are not always particularly strong .
one variety of benchmark ucd that could yield accurate ages and surface gravities over a broad age range are in binary systems with a white dwarf ( wd ) companion .
in particular , if the wd is relatively high mass , then the main sequence progenitor star will have a short lifetime , and the age of the binary system ( and the ucd ) will essentially be the same as the cooling age of the wd , which can be well determined from theory .
there have been several searches to find ucd companions to wds . despite this , only a small number of detached ucd
wd binaries have been identified ; gd 165b(l4 ) @xcite , gd 1400(l6/7 ) ( @xcite ; @xcite ) , wd@xmath20(l8 ) ( @xcite ; @xcite ) and pg1234 + 482 ( l0 ) ( @xcite ; @xcite ) .
the two components in gd165 are separated by 120au , the separation of the components in gd1400 and pg@xmath21 are currently unknown and wd@xmath20 is a close binary ( semi - major axis @xmath22r@xmath23 ) . @xcite and @xcite also identified three late m companions to white dwarfs ; wd@xmath24 ( m8 at 23au ) , wd@xmath25 ( m8 at 2054au ) and wd@xmath26 ( m9 at 284au ) . the widest system previously known was an m8.5 dwarf in a triple system a wide companion to the m4/wd binary lhs 4039 and lhs 4040 @xcite , with a separation of 2200au .
there are several other known ucd
wd binaries , however these are cataclysmic variables ( e.g. sdss 1035 ; @xcite , sdss1212 ; @xcite , @xcite , ef eri ; @xcite ) but these are unlikely to provide the type of information that will be useful as benchmarks , as they have either evolved to low masses via mass transfer or their ages can not be determined because of previous evolutionary phases ( e.g. common envelope ) .
wd0137 - 349 is also not likely to be useful as a benchmark due to the large uncertainty in age caused by the common envelope stage and the heating from the white dwarf .
recent analysis from @xcite shows that the fraction of brown dwarf companions ( l and t ) at separations within a few hundred au of white dwarfs is @xmath270.6 per cent .
however , despite this ucds in wide binary systems are not uncommon ( revealed through common proper motion ) around main sequence stars at separations of @xmath28au ( @xcite ; @xcite ) . however ,
when a star sheds its envelope post - main sequence , we may expect a ucd companion to migrate outwards to even wider separation ( @xcite ; @xcite ) , and ucd wd binaries could thus have separations up to a few tens of thousands of au . although some of the widest binaries may be dynamically broken apart quite rapidly by gravitational interactions with neighbouring stars , some systems may survive , offering a significant repository of benchmark ucds .
we present here the first results from our search of 2mass and the digitized schmidt plate archive supercosmos , for wide ucd
wd binaries in the southern sky , and present our first discovery of a wide ucd wd binary confirmed through common proper motion and spectroscopy . in section 2
we describe the techniques we have developed to select candidate ucd
wd binary systems using database photometry and proper motion .
section 3 describes our second epoch imaging of candidates , and followup infrared and optical spectroscopy along with proper motion analysis .
we present spectroscopic analysis in section 4 .
section 5 discusses the newly discovered system , and section 6 presents a discussion of future work .
candidate wds were selected from the supercosmos science archive ( ssa ) , following a similar technique to that of @xcite .
we selected candidates based on their position on a reduced proper motion ( rpm ) diagram , a technique that uses proper motion as a distance proxy , taking advantage of the fact that nearby objects in general have higher proper motion .
[ rpmd ] shows our supercosmos rpm diagram .
our initial ssa sample ( shown as dots ) was a magnitude selected sample ( @xmath2920 ) to avoid sources near the plate limit .
we also selected against non - moving sources ( by requiring pm@xmath1510mas yr@xmath30 , pm/@xmath315 ) , and against non stellar sources ( requiring a database star - like class=2 ) .
in addition we avoid the galactic plane ( requiring @xmath3225@xmath33 ) , require ssa @xmath34- band coverage ( @xmath35 + 3@xmath33 ; which offers possible additional epochs for nearby ucd candidates ; see section 3.3 ) , and impose a quality constraint on the database photometry ( qual@xmath361040 in each of the @xmath37- , @xmath38- and @xmath34- bands , indicating the the object is unlikely to be a bright star artifact . ) . the main sequence and wd sequence can be clearly seen in fig .
[ rpmd ] , and display good separation for @xmath391.3 . to further illustrate the location of the wd sequence , we over plot the positions of wds from the spectroscopically confirmed mccook & sion wd catalogue ( @xcite ; here after ms99 ) , along with pre - wds ( i.e. sdo stars ) and halo objects from @xcite . we also overplot subdwarf b stars , hot , cool and extreme subdwarfs from @xcite , @xcite , @xcite and @xcite respectively .
these , like halo objects can have high velocity dispersions and can masquerade as white dwarfs in rpm diagrams .
the overplotted populations help confirm the location of the wd sequence for @xmath391.3 , and allow us to assess the location of wds out to redder colour .
our final rpm selection criteria were chosen to strike the best balance between wd selection and contamination minimisation , and are shown in fig .
[ rpmd ] as dashed lines . at the red end in particular ,
we have attempted to include as many wd candidates as possible , while avoiding contamination from cool subdwarfs .
the rpm selection criteria are @xmath40 for @xmath41 and + @xmath42 for @xmath43 ; our selection criteria resulted in a sample of 1532 wd candidates .
ucd candidates were photometrically selected by their near infrared ( nir ) colours from the 2mass all sky point source catalogue , using the nasa / ipac gator search engine .
we used colour cut criteria devised by @xcite , which are based on the colours of known l - dwarfs with reliable @xmath44- , @xmath45- and @xmath46- band 2mass photometry ( snr@xmath1520 ) from the caltech cool dwarf archive ( dwarfarchives.org ) , requiring the following photometric criteria to be met : @xmath47 + @xmath48 + @xmath49 + @xmath50 + @xmath51 + and @xmath52 .
+ we also imposed an optical - nir colour restriction using the usno - a2.0 cross - match facility within the gator , ruling out objects with @xmath535.5 ( @xmath54 combined with [email protected] , or nopt_mchs=0 ) .
in addition certain sky coverage constraints were imposed , and we made use of the 2mass data quality criteria and some additional database flags , to remove contamination from artefacts and low quality photometric data ( requiring cc_flag=000 , ph_qual@xmath36ccc , jhk_snr@xmath145 , gal_contam=0 , prox@xmath146 ) and known contaminants ( mp_flg=0 ) .
as with the wd candidate selection , we searched areas away from the galactic plane ( @xmath5625@xmath33 ) to avoid confusion from over crowding and contamination from reddened stars and giants .
we also avoided @xmath5786@xmath33 , for which 2mass suffers from incompleteness issues with its optical cross - matching .
additional areas surrounding the small and large magellanic clouds were also ignored .
finally two additional uncatalogued , reddened regions were avoided following the approach of @xcite .
these excluded regions are listed in table [ cloud_table ] , and the resultant sky coverage of our survey is 13,216 sq .
degs , or 32 per cent of the sky .
visual inspection of all these candidates ( 13,338 in number ) was neither practical or necessary at this stage , and we thus postponed our visual inspection until after our binary pair selection procedure ( see next section ) .
.galactic co - ordinates of contaminated and overcrowded regions removed from our selection .
[ cols="<,^,^,^,^",options="header " , ] [ properties ]
a parallax measured distance would allow the luminosity of 2massj@xmath9 to be directly measured .
our age constraint would then facilitate a tight model constraint on the radius , and thus an accurate determination of @xmath3 and log(@xmath5 ) .
this ucd could then have a very useful role as part of a testbed for our understanding of ultracool atmospheres .
for example , @xcite studied a set of m dwarfs with young and intermediate ages , and showed that a variety of spectral lines including the @xmath44- band ki lines , are gravity sensitive .
it is also clear that the effects of dust in very late m dwarf atmospheres are not well understood ( e.g. @xcite ) . removing degeneracies in the physical properties of such objects
will allow for a more focused testing of available models in this region of @xmath58log(@xmath5 ) space .
a more expansive search for benchmark ucd - wd binary systems can be made using the combination of ukidss and sdss ( e.g. @xcite ) .
these surveys probe a significantly larger volume for ucds and wds than 2mass and supercosmos , and will thus yield a larger number of wide benchmark binaries , with a broader range of properties .
it is likely that significant numbers of binary systems could be identified containing an l / t dwarf and a wd ( see @xcite ) .
also , if the frequency of wide ucd companions to higher mass stars is of a similar level to those found around solar type stars , there could be numerous detectable systems with high mass wd components .
tight constraints on binary age will be particularly important when determining the properties of cooler , lower mass ucd companions , since sub - stellar brown dwarfs cool continuously , and their mass and surface gravity depend strongly on age .
the broad band spectral morphology of l and t dwarf populations show a significant scatter independent of spectral type , presumably due to variations in surface gravity and metallicity .
thus , the physical constraints offered by benchmark systems of this type could hold great potential to reveal trends linking such spectral changes to variations in ucd physical properties .
we acknowledge science and technology facilities council ( stfc ) for their support to acd , bb , slf , djw and jrc , we would also like to acknowledge the helpful advice given from detlev koester in this work . this publication has made use of the nasa / ipac infrared science archive , which is operated by the jet propulsion laboratory , california institute of technology , under contract with the national aeronautics and space administration .
we have also made use of the data obtained from the supercosmos science archive , prepared and hosted by the wide field astronomy unit , institute for astronomy , university of edinburgh , which is funded by the stfc .
we gratefully acknowledge the uk and australian government support of the anglo - australian telescope through their stfc and detya funding as well as nasa grant nag5 - 8299 & nsf grants ast95 - 20443 and ast-9988087 and sun microsystems .
we would also like to acknowledge the denis project , that has been partly funded by the science and the hcm plans of the european commission under grants ct920791 and ct940627 .
it is supported by insu , men and cnrs in france , by the state of baden - wrttemberg in germany , by dgicyt in spain , by cnr in italy , by ffwfbwf in austria , by fapesp in brazil , by otka grants f-4239 and f-013990 in hungary , and by the eso c&ee grant a-04 - 046 .
jean claude renault from the iap was the project manager .
observations were carried out thanks to the contribution of numerous students and young scientists from all involved institutions , under the supervision of p. fouqu , survey astronomer resident in chile .
allard f. , hauschildt p.h . ,
alexander d.r .
, tamanai a. , schweitzer a. , 2001 , apj , 556,357 baraffe i. , chabrier g. , allard f. , hauschildt p.h . , 2002 ,
a&a , 382 , 563 benvenuto o.g . ,
althaus l.g . , 1999 ,
mnras 303 , 30 bergeron p. , gianninas a. , boudreault s. , 2007 , in asp conf .
372,15th european workshop on white dwarfs , eds .
r.napiwotzki & m.r.burleigh , ( san francisco : asp ) , 29 bessell m.s . , 1986 ,
pasp , 98 , 1303 burgasser a.j . ,
geballe t.r . ,
leggett s.k . ,
kirkpatrick j.d . ,
golimowski d.a . , 2006 ,
apj , 637 , 1067 burgasser a.j .
, burrows a. , kirkpatrick j.d . , 2006 , apj , 639,1095 burleigh m.r . , clarke f.j . ,
hodgkin s.t . , 2002 , mnras , 331 , 41 burleigh m.r . ,
hogan e. , dobbie p.d .
, napiwotzki r. , maxted p.f.l . , 2006 , mnras , 373 , l55 burleigh m.r . ,
marsh t.r . ,
gnsicke b.t .
, goad m.r .
, dhillon v.s . ,
littlefair s.p . , wells m. , bannister n.p . , hurkett c.p . ,
martindale a. , dobbie p.d . , casewell s.l . ,
baker d.e.a . , duke j. , farihi j. , irwin m.j . , hewett p.c . , roche p. , lewis f. , 2006 , mnras , 373 , 1416 burrows a. , sudarsky d. , hubeny i , 2006 , apj , 640,1063 cataln s. , isern j. , garca - berro e. , ribas i. , allende prieto c. , bonanos a.z .
, 2008 , a&a , 477 , 213 cavanagh b. , hirst p. , jenness t. , economou f. , currie m.j .
, todd s. , ryder s.d . , 2003,aspc , 295,237 chabrier g. , baraffe i. , allard f. , hauschildt p. , 2000
, apj , 542 , 464 chabrier g. , bassard p. , fontaine g. , saumon d. , 2000 , apj , 543 , 216 cruz k.l . , reid i.n . , liebert j. , kirkpatrick j.d . ,
lowrance p.j .
, 2003 , aj , 126 , 2421 cruz k.l . , et al . , 2007 , apj , 133 , 439 cushing m.c .
, rayner j.t . , vacca w.d . , 2005 ,
apj , 623 , 1115 dahn c.c . , et al . , 2002 , aj , 124 , 1170 day - jones a.c . , pinfield d.j .
, napiwotzki r. , burningham b. , jones h.r.a . , jenkins j.s . , 2007 , aas , 221 , 16204 delorme p. , delfosse x. , albert , l. , artigau e. , forveille t. , reyl c. , allard f. , homeier d. , robin a. , willott c.j . , liu m. , dupuy t. , 2008 , arxiv0802.4387d dobbie p.d . , pinfield d.j
, napiwotzki r. , hambly n.c .
, burleigh m.r . ,
barstow m.a . ,
jameson r.f . , hubeny i. , 2004 , mnras , 355,39 dobbie p.d . , burleigh m.r . ,
levan a.j .
, barstow m.a .
, napiwotzki r. , holberg j.b .
, hubeny i. , howell s.b . , 2005 , mnras , 357 , 1049 dobbie p.d .
, napiwotzki r. , burleigh m.r .
, barstow m.a . ,
boyce d.d . ,
casewell s.l . ,
jameson r.f . , hubeny i. , fontaine g. , 2006 , mnras , 369 , 383 engelbrecht a. , koester d.2007 , in asp conf .
372 , 15th european workshop on white dwarfs , eds .
r.napiwotzki & m.r.burleigh , ( san francisco : asp ) , 289 farihi j. , christopher m. , 2004 , aj , 128 , 1868 farihi j. , becklin e.e . , zuckerman b. , 2005 , aj , 161 , 394 farihi j. , hoard d.w . , wachter s. , 2006 , apj , 646 , 480 farihi j. , burleigh m.r . , hoard d.w . , 2008 ,
apj , 674 , 421 farihi j. , becklin e.e . , zuckerman b. , 2008 , arxive:0804.0237 ferrario l. , wickramasinghe d. , liebert j. , williams k.a . , 2005 , mnras , 361 , 1131 finley d.s . , koester d. , basri g.1997 , apj 488 , 375 folkes s.l . , pinfield d.j . ,
kendall t.r .
, jones h.r.a
. , 2007 , mnras , 378,901 geballe t.r . , knapp g.r . ,
leggett s.k . ,
fan x. , golimowski d.a . , 2003 ,
csss , 12,670 gizis j.e . ,
kirkpatrick j.d .
, burgasser a. , reid i.n . ,
monet d.g .
, liebert j. , wilson j.c . , 2001 , apj , 551 , 163 golimowski d.a .
, et al . , 2004 , aj , 127 , 3516 holberg j.b . , oswalt t.d . , sion e.m , 2002 , apj , 571 , 512 holberg j.b . , bergeron p. , 2006
, aj , 132 , 1221 howell s.b . ,
ciardi d.r , 2001 , apj , 550 , 57 jeans j.h . , 1924 , mnras , 85 , 2 jones h.r.a . , steele i.a . , 2001
, bolin heidelberg springer jones h.r.a . , pavlenko y. , viti s. , barber r.j .
, yakovina l.a . , pinfield d. , tennyson j. , 2005 , mnras , 358 , 105 kalirai j.s . , hansen b.m.s .
, kelson d.d . ,
reitzel d.b . ,
rich r.m . , richer h.b . , 2008 , apj , 676 , 594 kilkenny d. , heber u. , drilling j.s .
, 1988 , aaoc , 12 , 1 kleinman s.j . , et al . , 2004 ,
apj , 607 , 426 knapp g.r . , et al . , 2004 ,
aj , 127 , 3553 knox r.a . ,
hawkins m.r.s . ,
hambley n.c . , 1999 ,
mnras , 306 , 736 leggett s.k . , 1992 , apj , 82 , 351 littlefair s.p . ,
dhillon v.s . ,
marsh t.r . , gnsicke b.t . , southworth j. , watson c.a . , 2006 , sci , 314 , 1578 liu m.c . , leggett s.k . , chiu k. , 2007 , 660 , 1507 maxted p.f.l . , napiwotzki r. , dobbie p.d . , burleigh m.r .
, 2006 , nat , 442 , 543 mccook g.p .
, sion e.m . , 1999 ,
apjs , 121 mcgovern m.r .
, mclean i.s . ,
kirkpatrick j.d .
, burgasser a.j . , prato l. , 2004 , aas , 20513510 mclean i.s . , et al .
, 2000 , apj , 533 , 45 mclean i.s .
, mcgovern m.r . , burgasser a.j .
, kirkpatrick j.d .
, prato l. , kim s.s .
, 2003 , apj , 596 , 561 metchev s.a . , hillenbrand l.a . , 2006 ,
apj , 651 , 1166 monet d.g . ,
dahn c.c . ,
vrba f.j . ,
harris h.c . ,
pier j.r . ,
luginbuhl c.b .
, ables h.d .
, 1992 , aj , 103,638 monteiro h. , jao w.c . , henry t. , subasavage j. , beaulieu t. , 2006 , apj , 638 , 446 mullally f. , kilic m. , reach w.t . , kuchner m.j . , von hippel t. , burrows a. , winget d.e . ,
apjs , 171 , 206 nakajima t. , oppenheimer b.r . , kulkarni s.r .
, golimowski d.a . , matthews k. , durrance s.t . , 1995 , nat , 378 , 463 napiwotzki r. , green p.j . , saffer r.a .
, 1999 , apj 517 , 399 napiwotzki r. , et al . , 2001 , an , 322 , 411 natali f. , natali g. , pompei e. , pedichini f. , 1994 , a&a , 289 , 756 padoan p. , nordlund 2002 , apj , 576,870 panei j.a .
, althaus l.g . ,
benvenuto o.g . , 2000 ,
a&a , 353 , 970 pinfield d.j .
, jones h.r.a . ,
lucas p.w . ,
kendall t.r .
, folkes s.l .
, day - jones a.c . ,
chappelle r.j . ,
steele i.a . , 2006 ,
mnras , 368,1281 rebolo r. , zapatero - osorio m.r .
, martin e.l .
, 1995 , nat,377,129 reid i.n .
, burgasser a. , j. , cruz k.l . ,
kirkpatrick j.d .
, gizis j.l . , 2001 ,
aj , 121 , 1710 reipurth b. , clarke c. , 2001 , aj , 122 , 432 rice w.k.m . , armitage p.j . , bonnell i.a . ,
bate m.r . , jeffers s.v . ,
vine s.g . , 2003 ,
mnras , 346,36 saumon d. , et al . , 2007 ,
apj , 656,1136 scholz r.d .
, lodieu n. , ibata r. , bienaym o. , irwin m. , mccaughrean m.j .
, schwope a. , 2004 , mnras , 347 , 685 schrder k.p .
, pauli e.m . , napiwotzki r. , 2004 , mnras , 354 , 727 silvestri n.m . , hawley s.l . ,
oswalt t.d . , 2005 ,
aj , 129 , 2428 slesnick c.l . ,
hillenbrand l.a . ,
carpenter j.m . , 2004 , apj , 610 , 1045 smith j.a . ,
tucker d.l .
, allam s.s . , jorgensen a.m. , 2002 , aas , 20110408 stark m.a . ,
wade r.a . , 2003 ,
aj , 126,1455 steele p.r . , burleigh m.r .
, dobbie p.d . ,
barstow m.a . , 2007 , mnras , tmp , 1021 warren s.j . , et al . , 2007 , mnras , 381 , 1400 weidemann v. , 2000 , a&a , 363 , 647 wesemael f. , van - horn h.m .
, savedoff m.p . ,
auer l.h . , 1980 ,
apj , 43 , 159 whitworth a.p . , goodwin s.p . , 2005 , an , 326 , 899 williams k.a . , bolte m. , koester d. , 2004 , apj , 615 , 49 yong d. , lambert d.l . , 2003 , pasp , 115 , 22 zuckerman b. , becklin e.e . , 1992 , apj , 386 , 260 | we present the discovery of the widest known ultracool dwarf white dwarf binary .
this binary is the first spectroscopically confirmed widely separated system from our target sample .
we have used the 2mass and supercosmos archives in the southern hemisphere , searching for very widely separated ultracool dwarf white dwarf dwarf binaries , and find one common proper motion system , with a separation of @xmath0au at an estimated distance of @xmath1pc , making it the widest known system of this type .
spectroscopy reveals 2mass j@xmath2 is a da white dwarf with @xmath3=@xmath4k , log(@xmath5)@xmath6 and m@xmath7m@xmath8 .
we spectroscopically type the ultracool dwarf companion ( 2mass j@xmath9 ) as m9@xmath101 and estimate a mass of @xmath11m@xmath8 , @xmath12k and log(@xmath5)=@xmath13 , placing it near the mass limit for brown dwarfs .
we estimate the age of the system to be @xmath141.94gyrs ( from the white dwarf cooling age and the likely length of the main sequence lifetime of the progenitor ) and suggest that this system and other such wide binaries can be used as benchmark ultracool dwarfs .
[ firstpage ] stars : low mass , brown dwarfs - stars : white dwarfs - binaries : general |
gauge theories in a slice of anti - de sitter ( ads ) space have attracted a great deal of attention in recent years primarily because they provide geometrical explanations of particle physics mysteries , including the gauge hierarchy problem @xcite , and a window into strongly coupled gauge theories via the ads / cft correspondence @xcite .
massless gauge fields propagating in the bulk of ads@xmath0 have been thoroughly examined in the literature @xcite and have many applications ( see for example @xcite ) .
it is well - known that the zero mode of this field has a flat profile ( constant wavefunction ) in the extra - dimension , and that the dual theory is described by a ( mostly ) elementary source field interacting with a strongly coupled conformal field theory ( cft ) via a conserved cft current @xcite .
not so familiar is the fact that gauge fields can be localized in the extra dimension , at least for the abelian case @xcite . in order to localize the zero mode , the bulk equation of motion and boundary conditions
must be modified , which is accomplished through the introduction of mass terms in the bulk and on the boundaries . by tuning these masses , a zero mode solution with an exponential profile
is allowed .
the phenomenology and holographic description of localized u(1 ) gauge fields is described in @xcite .
extending this mechanism to the non - abelian case is not so straightforward . in order to preserve four - dimensional ( 4d ) gauge invariance ,
the zero mode cubic and quartic gauge interactions must have the same coupling .
these couplings are proportional to wavefunction overlap integrals , and for exponential zero mode wavefunctions they do not necessarily match . thus it seems the localization mechanism violates 4d gauge invariance . to have a viable low energy gauge theory , the equality of these couplings must be restored . in this paper we show that this can be achieved by considering interactions on the boundary , which provide the correct contributions to zero mode cubic and quartic terms to force the gauge couplings to be equal .
these interactions can be added to the theory in a gauge invariant way via boundary kinetic terms .
localizing both abelian and non - abelian gauge fields is important for both 5d model building and phenomenology as well as holographic physics . because the gauge field can be localized anywhere in the bulk , an entire new range of parameter space can now be explored .
for example , if the standard model gauge bosons all propagate in the bulk , each field could in principle be localized at different places in the extra dimension .
this may help to alleviate experimental bounds , as in the u(1 ) case , and could perhaps lead to novel model building scenarios .
perhaps more interesting is the notion that if zero modes are localized near the infrared boundary , there may exist a dual theory in which the gauge fields are composite cft states . in this case , the dual gauge theory would be a valid effective field theory below the cutoff corresponding to the scale of conformal symmetry breaking .
our approach is not without a few caveats that need to be addressed .
first , a particular boundary kinetic term must be added to restore the gauge invariance at the nonlinear level .
this represents an additional tuning beyond the tuning already required between bulk and boundary masses ( as well as the tuning between bulk and boundary cosmological constants in the original randall - sundrum model ! ) .
more importantly , mass terms in the 5d theory must be generated through spontaneous symmetry breaking if we hope to preserve a 4d gauge symmetry .
therefore , we must also account for the additional scalar fields present in the theory .
these scalar fields complicate the analysis and may also introduce more severe problems , such as ghosts or strong coupling . although we are not able to offer a general analysis of these scalar issues , we present some simple examples illustrating the problems and speculate on possible resolutions .
even if these issues can only be resolved in the uv completion of the model the nice features of the yang - mills localization will still be preserved at low energies and therefore makes the present study relevant .
we also compare our approach to other methods of localizing gauge fields on the brane .
as first shown in @xcite , gauge fields can be given a localized wavefunction through a dilatonic coupling .
a similar mechanism is also motivated by deconstruction , in which the 4d gauge theories have nonuniversal gauge couplings at each site @xcite .
these approaches share many of the features of our method , but there are also differences , which we point out . the scalar component @xmath1 is also utilized in models of electroweak symmetry breaking as an alternative to the higgs boson @xcite . the zero mode profile of this field is restricted by gauge invariance .
however , we show that like the vector @xmath2 , it is possible to change the profile , and thus localize the @xmath1 zero mode anywhere in the bulk . this could affect the phenomenology of gauge - higgs unification models in novel ways .
the organization of this paper is as follows . in section 2
we begin by reviewing how the addition of bulk and boundary mass terms allow for a localized zero mode gauge field .
we then move on to the full non - abelian analysis , showing how boundary kinetic terms , if appropriately tuned , can restore 4d gauge invariance . in section 3
we discuss the additional scalar fields in the theory .
we present a flat space analysis of the nambu - goldstone bosons associated with the massive vectors in the 4d theory .
we also discuss the possibility of strong coupling and ghosts by considering a simple example containing only boundary masses ( no bulk masses ) . in section 4
we review other methods of gauge field localization , including localization with a dilaton and via deconstruction , and compare these methods with our approach .
we discuss the localization of the scalar @xmath1 in section 5 .
finally we conclude in section 6 , and suggest possible directions for future investigation .
we will work with the following metric describing 5d anti - de sitter space : @xmath3 where @xmath4 is the extra coordinate and @xmath5 is the curvature scale related to the bulk cosmological constant .
the fifth dimension @xmath4 is compactified on a @xmath6 orbifold and two three - branes are located at the orbifold fixed points @xmath7 and @xmath8 , referred to as the ultraviolet ( uv ) and infrared ( ir ) brane , respectively .
the indices of 5d coordinates are labeled with latin letters ( @xmath9,@xmath10 , @xmath11 ) while 4d indices are labeled by greek letters ( @xmath12 , @xmath13 , @xmath11 ) .
all 4d indices are raised and lowered with the minkowski metric @xmath14 .
it is well - known that massless gauge fields propagating on the background ( [ adsmetric ] ) lead to a massless zero mode that is not localized in the extra dimension , but rather has a flat profile @xcite . instead to localize a zero mode , bulk and boundary mass terms must be added which modify both the bulk equation of motion and the boundary conditions for the gauge field @xcite .
the massless mode then has an exponential profile and can be localized anywhere in the bulk .
we will show how this is accomplished in section 2.1 . of course
such mass terms must be generated in a gauge invariant way , so that a 4d gauge symmetry is preserved for the zero mode .
the most obvious way to generate such terms is by the spontaneous breaking of gauge symmetry .
this introduces extra scalar fields into the theory which may cause problems .
for instance , if massless scalar modes persist in the 4d theory they may be undesirable for model building .
other , more severe problems , such as the existence of ghosts or strongly coupled light scalar fields , may also occur .
we will discuss these issues in section 3 .
however , the gauge sector can be isolated from the scalar sector , and none of these more severe problems exist for the gauge field .
a separate issue arises when we consider the localization of non - abelian gauge fields . as we will see in this section ,
the zero mode cubic and quartic couplings do not match , and thus the conventional 4d gauge symmetry does not exist in the simplest formulation .
the effective couplings are proportional to the overlap integral of the wavefunctions , and are only equal when the zero mode is flat . to remedy this
, we will add cubic and quartic interactions on the branes which will restore the equality of the zero mode gauge couplings .
the only way this can be done while preserving gauge invariance is by adding a kinetic term to the brane , which we show in section 2.2 .
we study only pure yang - mills theory in this paper .
however , if fermions propagate in the bulk , we can generalize the mechanism and restore 4d gauge invariance by adding gauge covariant boundary kinetic terms for fermions .
we begin by briefly reviewing how to localize the zero mode gauge field in the extra dimension . for simplicity we will consider a u(1 ) gauge field , since the generalization to the non - abelian case is straightforward at the quadratic level in the action .
consider the following 5d equation of motion for a spin-1 field @xmath15 : @xmath16 the mass term arises from spontaneous symmetry breaking @xcite and contains both bulk and boundary terms : @xmath17 where we have defined @xmath18 .
the bulk and boundary masses are written in terms of the ads curvature @xmath5 with dimensionless parameters @xmath19 and @xmath20 .
as shown in @xcite , we must tune the parameters , @xmath21 to allow a zero mode solution .
only real values for the mass parameters , @xmath22 , are considered and therefore @xmath20 can take any real value .
let us choose the gauge @xmath23 for simplicity .
the equation of motion ( [ eom1 ] ) becomes @xmath24a_\mu(x , y)=0~. \label{eom2}\ ] ] decomposing the vector @xmath2 as @xmath25 the solution of ( [ eom2 ] ) for the massless mode is @xmath26 since @xmath20 can be any real number ( [ tune ] ) , the zero mode can be localized anywhere in the fifth dimension . to see the necessity of tuning the parameters @xmath19 and @xmath20 , note
the boundary condition derived by integrating the equation of motion ( [ eom2 ] ) across the boundary is @xmath27 the zero mode satisfies this equation only if the relation ( [ tune ] ) holds . otherwise , the boundary conditions project out the massless mode .
it is possible that this tuning is enforced in a supersymmetric extension , as happens for bulk scalar fields @xcite . to summarize , the bulk mass term changes the profile of the zero mode from flat to exponential .
clearly , an exponential wavefunction is incompatible with ordinary dirichlet or neumann conditions .
the brane - localized mass modifies the boundary condition to be compatible with this exponential wavefunction .
furthermore , this leads to interesting phenomenological possibilities and remarkably the localized u(1 ) gauge bosons can be given a holographic interpretation @xcite .
let us now consider the following action for a non - abelian gauge field @xmath28 in a slice of ads@xmath0 : @xmath29~ , \label{a1}\end{aligned}\ ] ] where the field strength tensor is defined as @xmath30 with group structure constants @xmath31 and 5d coupling @xmath32 .
again , the mass term is assumed to be generated by spontaneous symmetry breaking .
the technique described in the previous section can also be applied to localizing non - abelian gauge fields . at the quadratic level ,
the equation of motion is the same as in the abelian case , and therefore the procedure is identical .
however , when we examine the nonlinear terms of the yang - mills theory to compute the zero mode cubic and quartic interactions , we find that the couplings do not match .
thus it seems that there is no longer any 4d gauge invariance .
it is possible to restore the zero mode cubic and quartic couplings consistent with 4d gauge invariance if we consider brane - localized interactions .
these terms play a similar role to the boundary mass terms , which cancels the bulk mass contribution to the zero mode action .
we can not add these terms by hand to the boundary action without violating gauge invariance . to preserve the symmetry
, we must add a boundary kinetic term , so that the action becomes @xmath33 . \label{a2}\ ] ] note the boundary term is suppressed by @xmath34 relative to the bulk term .
it will be necessary to tune the dimensionless coefficient @xmath35 to match the zero mode couplings .
for simplicity , we have chosen the same coupling on each brane , but it is possible to have them different , which we will discuss at the end of this section . also notice that one of the kinetic terms has a negative coefficient and therefore could introduce a ghost into the theory . however , it is the sum of the bulk and boundary kinetic energies which is important .
when we perform the kaluza - klein reduction , we will require the wavefunction overlap integral , including the bulk and boundary terms , to be positive , which then implies the particle has positive energy .
we will see that this requirement is satisfied for all values of @xmath35 that we will consider .
in fact negative boundary kinetic terms have been previously considered in the literature . in ref .
@xcite the effect of brane kinetic terms in the randall - sundrum model was studied and it was found that a ghost does indeed appear in the spectrum , but only if one of the boundary terms is made so strong and negative that it overwhelms the positive energy contributions from the bulk and other boundary .
similar conclusions were drawn in ref .
@xcite where negative kinetic terms in flat space were examined .
thus , it seems from the 4d effective field theory vantage point , negative boundary kinetic terms in principle pose no problem to model building .
however , note that this cancellation mechanism is nonlocal since locally there is always a negative kinetic term . with this modification to the action ,
the equation of motion at quadratic level is @xmath36a_\mu(x , y)=0~. \label{eom4}\ ] ] decomposing the field ( [ kk ] ) and demanding @xmath37 the equation of motion for the eigenfunction @xmath38 becomes @xmath39f^n(y)=-e^{2ky}m_n^2\left(1+\frac{\beta}{k^2}\sigma''(y)\right)~. \label{eom5}\ ] ] the bulk equation of motion is unaffected by the presence of the boundary term , but the boundary condition is now altered : @xmath40 the massive modes are influenced by the brane kinetic term and thus the spectrum will be modified . however , the boundary condition for the massless mode @xmath41 is the same as in the abelian case ( [ bc1 ] ) , and therefore the zero mode can be localized provided we demand the tuning condition ( [ tune ] ) .
the normalized zero mode solution is given by @xmath42 note that for a flat zero mode ( @xmath43 ) , the normalization is unaffected by the presence of the brane - localized kinetic term since the wavefunction has the same value at each boundary .
the normalization @xmath44 must be real , or equivalently the overlap integral ( [ normal ] ) must be positive so that the field is not ghostlike .
thus , we must require that @xmath45 .
we will see shortly that this limit is always satisfied if we require that a 4d gauge invariance be preserved for the zero mode .
the interaction terms following from the action ( [ a2 ] ) are @xmath46~. \label{int1}\ ] ] localizing a massless mode comes at a price : the loss of _ 4d gauge invariance _ of the zero mode . without boundary interaction terms
, the cubic and quartic vertices will have different strengths .
the 5d symmetry may still be realized in the effective theory , but it will not take the standard form familiar in 4d theories .
the gauge invariance can be restored if we add a boundary kinetic term with a precise coefficient ( [ a2 ] ) .
let us investigate the effective 4d interactions and see the problem and its resolution explicitly .
inserting the zero mode wavefunction ( [ wf ] ) into ( [ int1 ] ) , the couplings can be calculated by performing the overlap integrals .
the cubic coupling is given by : @xmath47 while the quartic coupling is @xmath48 clearly for @xmath49 , @xmath50 and the 4d gauge symmetry is lost for general values of @xmath20 .
thus we see the necessity of including a boundary kinetic term for nonzero @xmath20 .
only when the mode is flat ( @xmath43 ) are the couplings identical . moreover , when the mode is flat arbitrary boundary kinetic terms ( with different coefficients ) may be added to the action without affecting the zero mode gauge invariance .
we will demand that a 4d gauge symmetry be present in the theory for general values of @xmath20 , or in other words , we will require that @xmath51 . to find the value of @xmath35 that satisfies this requirement
, we square ( [ cubic ] ) and equate it to ( [ quartic ] ) . solving for @xmath35 , we find @xmath52 , \label{beta}\ ] ] where we have defined @xmath53)^2}~.\ ] ] the exponential modulating function @xmath54 ranges from as low as @xmath55 when @xmath43 and as high as @xmath56 , when @xmath57 is very large .
for any particular value of the localization parameter @xmath20 , there are two possible boundary kinetic terms that will preserve 4d gauge invariance , with strength @xmath58 or @xmath59 .
the @xmath58 and @xmath59 branches are continuously connected at @xmath43 and this behavior is plotted in fig .
[ fig1 ] .
( in units of @xmath60 ) as a function of @xmath61 .
the @xmath58 branch is indicated by the solid line while the @xmath59 branch is indicated by the dashed line .
the @xmath62 and @xmath63 branches are continuously connected at @xmath43 corresponding to a constant zero mode wavefunction . at @xmath43 , @xmath35 can take any value . ]
it is instructive to study two limiting cases . for small values of @xmath20 , we find that the coefficient @xmath35 becomes @xmath64 this equation seems to imply that for the flat mode ( @xmath43 ) , we must also include boundary kinetic terms to maintain 4d gauge invariance .
of course this is not true .
the presence of boundary kinetic terms has no effect on the 4d gauge invariance when the mode is flat , which can easily be seen , for example , from the wavefunction overlap integrals that determine the 4d couplings , eqs .
( [ cubic ] ) and ( [ quartic ] ) .
the boundary terms simply cancel when the wavefunction is constant . in the more generic case of different coefficients for each brane - localized kinetic term ,
the zero mode couplings are still equal .
we can also study the limit where @xmath65 , in which case the coefficient @xmath35 reduces to @xmath66 we see that either a very small or very large boundary kinetic term must be added to restore gauge invariance in the low energy theory .
now we turn to the effective 4d gauge couplings , which are computed by inserting ( [ beta ] ) back into ( [ quartic ] ) .
the couplings can also take two possible values depending on which boundary kinetic term ( @xmath58 or @xmath59 ) is chosen .
this is shown clearly in fig .
[ fig2 ] with @xmath67 plotted as a function of @xmath20 in units of @xmath68 .
plotted ( in units of @xmath68 ) as a function of @xmath20 for @xmath69 . for any given value of @xmath20
there are two characteristic behaviors , the @xmath59 branch ( dashed line ) , and the @xmath58 branch ( solid line ) .
the @xmath62 and @xmath63 branches are continuously connected at @xmath43 . ]
the two branches are again connected at @xmath43 , in which case the 4d gauge coupling becomes the familiar @xmath70 . for large @xmath57
the gauge coupling on the @xmath59 branch is approximately @xmath71 notice that a large value of @xmath72 is required to obtain a 4d gauge coupling of order one and therefore perturbativity of the 5d gauge theory does not allow for arbitrary localization on the @xmath59 branch . on the other hand for the @xmath58 branch ( shown as the solid line in fig .
[ fig2 ] ) , the gauge coupling for highly localized modes ( @xmath73 ) becomes : @xmath74 in this case a 4d gauge coupling of order one for large @xmath57 can always be obtained for perturbative values of the 5d gauge coupling .
thus , for extremely localized zero modes , only the @xmath58 branch is consistent with perturbativity for all large values of @xmath57 . for weak localization where @xmath75 , a 4d coupling of order one can be defined for both branches . as @xmath57 becomes of order unity and the mode
is localized , only the @xmath58 branch would be consistent with perturbative values of @xmath68 .
finally , let us verify that the zero mode is not a ghost for all values of @xmath35 . at the end of section 2.2 we found that @xmath76 , otherwise @xmath77 has negative kinetic energy and is ghostlike . from eq .
( [ beta - largealpha ] ) we see that the maximum value of @xmath78 is @xmath79 which occurs for very large values of @xmath57 on the @xmath58 branch .
thus we see that the limit is always satisfied and the zero mode has positive energy .
notice that in our approach , there are two relations that must be tuned : 1 ) the bulk and boundary masses must satisfy relation ( [ tune ] ) to permit a zero mode solution consistent with boundary conditions , and 2 ) a precise boundary kinetic term ( one on each brane ) with a coefficient given by ( [ beta ] ) , which is related to the boundary mass @xmath20 must be added to the action . of course
, all of these parameters will receive quantum corrections , and therefore these relations are inherently radiatively unstable .
this question will not be addressed in this paper and we simply assume that the tuning is enforced in the underlying theory to cancel any quantum effects that leads to deviations in the tunings .
one might hope that there is some symmetry which could enforce the tuning relations .
one possibility is supersymmetry .
similar bulk and boundary mass relations for scalar fields are demanded by supersymmetry @xcite , and it is not unrealistic to think that the same thing could happen for gauge fields .
however there are no known examples where boundary kinetic terms are demanded by supersymmetry , so it remains an unanswered question whether the second tuning will be enforced by any symmetry .
let us briefly discuss the possibility of having different kinetic terms on the uv and ir brane , with strengths @xmath80 and @xmath81 . modifying the action ( [ a2 ] )
@xmath82 , \label{adiff}\ ] ] we can follow the procedure outlined in section 2.2 and compute the zero mode wavefunction and cubic and quartic couplings .
demanding that the couplings are equal gives a constraint on the allowed @xmath35 values .
for a given localization @xmath20 , we can choose , say , @xmath81 and then compute @xmath80 in terms of @xmath20 and @xmath81 .
we must also demand that the normalization of the zero mode wavefunction is real ; otherwise the zero mode will be a ghost .
this is illustrated in fig .
[ fig3 ] for the case @xmath83 . and @xmath81 for the case @xmath83 , indicated by the solid line .
the shaded region indicates the @xmath35 values which produce a zero mode ghost , and are thus not allowed .
the line @xmath84 is also shown ( dotted line ) , and the @xmath58 and @xmath59 branches are also indicated . ] notice that it is not possible to have two positive kinetic terms ; at least one must be negative .
this implies that for high energy processes , locally the kinetic term will appear to be negative , and the theory will have a ghost . at large distances ,
the nonlocal effects of both brane terms combine to form a perfectly sensible effective field theory , so long as the boundary terms are not so strong and negative that they overwhelm the positive bulk contribution .
the shaded region indicates where this happens ; the negative ( and large ) brane kinetic terms produce a zero mode ghostlike gauge field .
note also the case @xmath84 studied in the previous section lies in the ghost - free region . the structure of this plot is similar for different values of the localization parameter @xmath20 . as we localize the zero mode towards the ir brane ( increasing @xmath20 ) the asymptote moves towards , but never reaches , the @xmath4-axis .
instead localizing the massless mode towards the uv brane ( decreasing @xmath20 ) , causes the asymptote to move further away towards large negative values .
the qualitative behavior of the allowed values for @xmath80 and @xmath81 is the same in these cases . in particular , it is never possible to have both brane kinetic terms positive , and there is always a region where the brane kinetic terms produce a zero mode ghost .
to localize a massless gauge field in the bulk , we found that it was necessary to add mass terms in the bulk and on the boundary .
if we simply add a hard mass term which breaks the 5d gauge symmetry , there is no hope of recovering a 4d gauge symmetry in the effective theory .
thus , it is necessary to generate the mass term through spontaneous symmetry breaking , i.e. a higgs mechanism . thus localizing a gauge field in this approach
comes at the price of introducing additional scalar fields in the theory .
these additional scalars complicate the analysis and obscure the 4d gauge symmetry compared to the case of a massless 5d bulk gauge field , in which the 4d symmetry is realized in a rather simple way . in that case , the scalar mode @xmath85 is the nambu - goldstone boson which is eaten by the massive vector @xmath86 @xcite . with additional scalar fields ,
the nambu - goldstone boson will be a particular linear combination of scalar fields , while another combination will be a physical scalar degree of freedom .
although the analysis is complicated in warped space , we illustrate this feature with a flat space example in section 3.1 .
more pressing is the possibility that these scalar modes may be massless , strongly coupled , or even ghostlike .
we discuss these issues in section 3.2 using a simple example in warped space with only boundary masses .
we find that natural boundary conditions allow a zero mode physical scalar field to exist . to preserve the gauge symmetry a boundary mass term for the vector @xmath2
necessarily contains a scalar kinetic term .
when the bulk and boundary mass parameters are tuned to cancel the zero mode mass , we might then expect that the zero mode scalar kinetic term is canceled or even becomes negative , corresponding to a strong coupling problem or a ghostlike state , respectively .
although we find no ghost in this example , the physical scalar zero modes are generically strongly coupled .
it seems likely that a different choice of boundary conditions ( e.g. dirichlet ) will eliminate the scalar zero mode , but the general analysis is then complicated , as we will discuss . in this section our discussion is limited to the abelian case for simplicity , but of course similar issues will be faced in non - abelian extensions .
let us revisit the abelian case , and consider the 5d action for the gauge field @xmath15 with bulk mass @xmath87 and the 5d nambu - goldstone boson @xmath88 which appears after spontaneous symmetry breaking : @xmath89~ , \label{a3}\ ] ] where we have only included quadratic terms in the lagrangian involving @xmath90 and @xmath91 . from a 5d point of view
, @xmath91 is the nambu - goldstone boson which gives a mass to the vector @xmath90 .
however when we compactify the extra dimension , the modes @xmath85 will play some role in giving the 4d vectors @xmath92 a mass @xmath93 .
we might then expect that it is some combination of modes @xmath94 and @xmath95 which form the true nambu - goldstone boson for @xmath92 , while the orthogonal combination is a physical scalar particle .
previously we did not take into account the nambu - goldstone boson which couples to @xmath2 at the quadratic level in the action .
here we show that it is indeed possible to isolate @xmath2 by defining @xmath96 gauges . by separating the action ( [ a3 ] ) into components
@xmath97 we obtain @xmath98~. \label{a4}\end{aligned}\ ] ] we can see that there is nontrivial mixing between @xmath99 , and @xmath91 , and it is not clear which combination of fields is the nambu - goldstone boson .
we can remove the mixing in the @xmath2 sector by adding a gauge fixing term @xmath100 ^ 2~. \label{lgf}\ ] ] the bulk equation of motion ( [ eom2 ] ) then follows if we choose the unitary gauge given by @xmath101 . with boundary mass terms
there will also be terms involving @xmath91 in the boundary action that mix with @xmath2 .
this mixing can be removed in a similar manner by gauge fixing , except that in this case there will be terms proportional to delta - functions . still the vector @xmath2 can be isolated and its equation of motion can be analyzed straightforwardly as in section 2.1 .
ideally , we would like to decompose the action ( [ a4 ] ) in order to see how the 5d symmetry manifests itself in the kaluza - klein action ( i.e. identify the nambu - goldstone boson ) , as is done with massless gauge fields @xcite . in warped space ,
however , it is unclear how to decompose the fields since they obey different equations of motion .
the analysis can be performed in flat space , and we do find that the 4d nambu - goldstone boson is a linear combination of @xmath85 and @xmath102 . consider the flat space limit @xmath103 of the action ( [ a4 ] ) : @xmath104~. \label{flataction}\end{aligned}\ ] ] the structure of the action gives us a hint as to how to decompose the fields .
consider the following expansion : @xmath105 where @xmath106 and the normalization is determined by @xmath107 the resulting 4d action is diagonal in kaluza - klein modes , but there is still mixing between @xmath108 , and @xmath102 : @xmath109~.\end{aligned}\ ] ] examining the action , it is natural to guess that the nambu - goldstone boson is proportional to the linear combination @xmath110 .
however , we must also show that there is a scalar massless eigenstate .
the mass mixing matrix @xmath111 has eigenvalues ( @xmath112 , @xmath113 ) and the eigenstates can be written as @xmath114 rewriting the action in terms of scalar mass eigenstates , we find @xmath115 , \label{bulk}\end{aligned}\ ] ] it is clear that @xmath116 is the nambu - goldstone boson .
the coefficient of the mixing term between @xmath92 and @xmath116 is exactly the one needed to make the propagator for @xmath92 transverse .
the unitary gauge is simply @xmath117 .
notice that @xmath118 is a physical scalar degree of freedom , as we would expect from counting the degrees of freedom contained in a 5d massive vector field .
if no additional boundary terms are considered , then the low energy theory contains no massless modes .
it is possible to obtain massless modes by augmenting the bulk action with boundary potentials which modify the boundary conditions . in flat space
the kaluza - klein decomposition is simple since the eigenfunctions are sinusoidal . in warped space ,
it is not clear that the eigenfunctions of the vector @xmath2 and the physical scalar and nambu - goldstone bosons will be related in a simple manner . in the case with no bulk and boundary masses @xcite , the decomposition proceeds as in ( [ aex ] ) and ( [ a5ex ] ) and the 4d action is simple .
besides complicating the general analysis , scalar fields can in principle introduce more severe problems into the theory .
to add boundary mass terms for the vector , we must also add kinetic terms for the scalar fields on each brane .
the tuning condition ( [ tune ] ) implies that one of these kinetic terms has the wrong sign , associated with negative energy . at low energies
this does not necessarily mean that the scalar has negative kinetic energy , since there are contributions from the bulk and the brane to the kinetic term of the scalar field .
one must explicitly examine the low energy theory to see if the particle is a ghost .
indeed , even though negative kinetic energy terms were added for the gauge field in section 1.2 to maintain 4d gauge invariance , the zero mode did have overall positive energy ( [ a2 ] ) .
however this cancellation mechanism was nonlocal so that at high energies there is a ghost locally which eventually needs to be resolved in the uv completion of the theory .
another potential problem is that the zero mode scalars might have a small or even vanishing kinetic term due to cancellations from bulk and brane contributions .
the kinetic term must then be canonically normalized , leading to strong coupling in any interactions in which the scalar participates .
these issues are difficult to examine in the general warped case with bulk and boundary masses because the action ( [ a4 ] ) does not allow for a straightforward kaluza - klein decomposition of the scalar fields .
however , there is one warped special case , the `` specular '' solution in which the bulk mass is zero but boundary mass terms localize the gauge field , where we can analyze these scalar issues in detail .
we find in this case that although there is no ghostlike particle , there is a strong coupling problem ( i.e. the kinetic term vanishes ) .
we give suggestions as to how this problem might be eliminated .
a massless gauge field in the bulk of ads@xmath0 is described by the action @xmath119~,\nonumber \\ & = & \int d^5x\left[-\frac{1}{4}f_{\mu\nu}f^{\mu\nu}-\frac{1}{2}e^{-2ky}(\partial_5 a_\mu -\partial_\mu a_5)^2 \right]~.\end{aligned}\ ] ] following the standard procedure @xcite , we expand the 5d fields in the following way : @xmath120 where the eigenfunctions @xmath38 satisfy the differential equation @xmath121 and the orthogonality condition @xmath122 we will focus on the massless mode , @xmath123 . normally the solution is taken to be @xmath124 , but since the differential equation is second order , there are in fact two linearly independent solutions .
the general solution is given by @xmath125 we must impose boundary conditions that are consistent with the variational principle .
the boundary terms that arise from variation of the action are @xmath126 \bigg\vert^{\pi r}_0~ , \label{b1}\ ] ] which forces us to impose dirichlet or neumann boundary conditions on @xmath127 .
these choices force @xmath128 , and the exponential solution is removed .
however , it is possible to keep the exponential solution if we add the following boundary mass term to the action : @xmath129~. \label{bmass}\ ] ] we will see that the boundary action is invariant under a 4d gauge symmetry .
varying the boundary action yields @xmath130 \bigg\vert^{\pi r}_0~. \label{b2}\end{aligned}\ ] ] combining the terms in ( [ b1 ] ) and ( [ b2 ] ) and using the bulk equation of motion for @xmath1 : @xmath131 we find the natural boundary condition for @xmath38 to be @xmath132 this corresponds to @xmath133 in ( [ bc1 ] ) , which is consistent with ( [ tune ] ) .
notice that there are other possibilities for boundary conditions consistent with the variational principle .
for instance we can choose dirichlet conditions for @xmath1 .
in this case the analysis is not as straightforward and the 4d gauge invariance is obscured , but there will be no @xmath1 zero mode .
we will revisit this discussion shortly when we discuss strong coupling .
for the zero mode , the exponential solution satisfies the boundary condition ( [ bc3 ] ) and the constant @xmath134 .
the zero mode solution is given by @xmath135 note that the zero mode field @xmath136 is not eliminated in this case , but has the same bulk profile as @xmath137 in the fundamental domain @xmath138 .
this is consistent because the wavefunction for @xmath139 is @xmath140 and is therefore odd with respect to orbifold symmetry . as mentioned above
, the boundary action ( [ bmass ] ) respects a 4d gauge invariance : @xmath141 where , noting ( [ bc3 ] ) , the second relation is equivalent to @xmath142 on the boundary .
it is instructive to examine how the _ abelian _ gauge symmetry , which involves no nambu - goldstone fields , arises for the zero mode . to this end
, we will detune the condition ( [ bc3 ] ) to be @xmath143 where @xmath144 is now an arbitrary constant .
the lowest lying mode becomes massive .
the eigenfunctions are given by @xmath145~,\ ] ] where @xmath146 and @xmath147 are constants .
the coefficient @xmath147 is found by applying the boundary conditions : @xmath148 and the masses are found by equating @xmath149 .
we will now show that there is a smooth limit @xmath150 starting from the detuned case . performing the kaluza - klein reduction ,
the action becomes @xmath151~ , \nonumber \\
& = & \int d^4x\sum_n\left[-\frac{1}{4}(f^n_{\mu\nu})^2-\frac{1}{2}(\partial_\mu a^n_5)^2-\frac{1}{2}m_n^2(a^n_\mu)^2+m_n a^n_\mu \partial_\mu a^n_5 \right]~. \label{a8}\end{aligned}\ ] ] it is clear from the action that @xmath85 is the nambu - goldstone boson , which provides a correct contribution to the vacuum polarization amplitude of @xmath92 . expanding the gauge parameter @xmath152 , we see that the 4d gauge invariance is given by @xmath153 in the unitary gauge , @xmath23 , the nambu - goldstone @xmath85 is eaten by @xmath92 .
now if we take the limit @xmath154 , the zero mode becomes massless @xmath155 .
the zero mode action reduces to @xmath156~,\ ] ] and there is an abelian gauge invariance , as is evident from ( [ gs ] ) by taking @xmath123 .
it is also clear from ( [ gs ] ) that when @xmath157 that we do not have freedom to set the zero mode @xmath158 .
the expansion ( [ exp ] ) indicates that in the limit @xmath150 , the wavefunction for @xmath139 becomes infinite .
we might argue that the expansion is not valid in this limit and instead use the following expansion : @xmath159 in this case we would find that although the wavefunction is finite , the kinetic term for @xmath139 vanishes .
actually , this implies that as we tune the boundary mass parameter @xmath154 in ( [ bc4 ] ) to get a massless mode , we should properly canonically normalize the kinetic term .
the smallness of the kinetic term then reappears as the large wavefunction in ( [ exp ] ) .
the large wavefunction in ( [ exp ] ) means that any interactions will become strongly coupled , and therefore the theory is not well - defined if such interactions exist .
indeed , any bulk interactions with @xmath90 would have this problem because of 5d gauge invariance .
the problem is even more severe because the mode is massless , meaning that the strong coupling problem is present in the low energy theory .
however there are no ghosts in the theory at low energies , as can be seen clearly by taking the limit @xmath150 in the action ( [ a8 ] ) .
the kinetic term for @xmath1 is always properly normalized with the correct sign corresponding to positive energy .
how can we avoid this strong coupling problem ?
the most obvious possibility would be to impose a different boundary condition for @xmath1 consistent with the variational principle .
a dirichlet condition , @xmath160 is our other option consistent with the variational principle , as can be seen from examining ( [ b2 ] ) .
this implies @xmath161 on the boundary and will make the second term in ( [ b2 ] ) vanish .
more importantly , this boundary condition will kill the zero mode @xmath162 . at the massless level
, we would have an ordinary gauge theory with only the zero mode @xmath77 .
however , for the massive modes , the action would no longer be diagonal with respect to kaluza - klein level @xmath163 because the wavefunctions associated with @xmath164 will be altered .
in particular the zero mode would mix with these massive modes quadratically and the 4d gauge symmetry would be obscured at high energies . in principle
there is no problem with this philosophy , and in fact this is akin to what happens in non - abelian theories when the zero mode interacts with the kaluza - klein modes via cubic and quartic terms , and the gauge invariance is realized in a non - trivial way ( each kaluza - klein mode transforms into other kaluza - klein modes ) .
another way the strong coupling problem might be addressed is via quantum corrections . in the @xmath157 limit
, the mass of the @xmath162 is not protected by a gauge symmetry in the 4d theory ( at least in the abelian case ) , and thus we would expect the mass to receive large corrections , quadratically dependent on the uv cutoff of the theory . in this way @xmath162 could decouple at low energies , relegating the problem to the uv . on the other hand ,
the masslessness of @xmath137 is protected by 4d gauge symmetry .
we have seen that bulk and boundary masses alter the equations of motion and boundary conditions to permit a localized zero mode non - abelian gauge field .
however , this is not the only way to achieve localization . another possibility is to couple the gauge field to a dilaton with a nontrivial background solution , as first illustrated in @xcite .
interestingly , it has been suggested that this approach is related to localization via bulk and boundary mass terms , essentially through a field redefinition @xcite .
this connection has also been alluded to via deconstruction , in which the wavefunction of the zero mode results from a position dependent gauge coupling varying from site to site @xcite . in this section ,
we wish to comment on this relation and point out the similarities and differences in each approach . in the end
, these other approaches to gauge field localization may provide simpler models , in particular with regard to the scalar sector , which we have seen in our approach is quite difficult to analyze in a satisfactory way .
let us begin by reviewing the basic idea considered in @xcite .
the theory consists of 5d gravity coupled to a scalar field @xmath165 .
the scalar @xmath165 acquires a bounce - like configuration , which can be idealized to a `` brane '' by taking a particular limit in the scalar potential .
rs2-like solutions appear for the metric in this brane limit .
if we introduce another scalar , the dilaton @xmath166 , it also acquires a nontrivial background , which in the brane limit turns out to be linear in @xmath4 , @xmath167 , where @xmath168 is a dimensionless constant .
in fact if we simply begin with a brane , that is , without reference to the field @xmath165 , we still find a linear dilaton solution @xcite . in these cases
, however , the dilaton @xmath166 back - reacts on the geometry , which is modified from pure ads@xmath0 . in our discussion below , we will assume the geometry is pure ads@xmath0 to compare with our model , even though we know of no 5d model with a linear dilaton solution and pure ads@xmath0 geometry .
following @xcite , we can easily see how the gauge field is localized .
consider the following action of a massless gauge field coupled to a dilaton @xmath166 , expanded about its background configuration : @xmath169~,\nonumber \\ & = & \int d^5x~ e^{-2 b k y}\left[-\frac{1}{4}f_{\mu\nu}f^{\mu\nu}-\frac{1}{2}e^{-2 k y } ( \partial_5 a_\mu-\partial_\mu a_5)^2 \right]~. \label{a5}\end{aligned}\ ] ] performing a kaluza - klein decomposition as in ( [ kk ] ) , the solution for the massless mode is a constant , @xmath170 .
the zero mode action can then be written as @xmath171~.\ ] ] the profile of the gauge boson with respect to a flat metric is then seen to be @xmath172 . the coupling to the dilaton simply becomes the gauge boson wavefunction when the dilaton acquires a vacuum expectation value ( vev ) .
this model can be related to localization with bulk and boundary masses by performing a field redefinition , @xmath173 .
the action ( [ a5 ] ) transforms to @xmath174~. \label{a6}\ ] ] it appears that we have broken the 5d gauge symmetry by performing this @xmath4-dependent redefinition , but of course the symmetry is manifested in a different way .
gauge transformations on the vector @xmath2 are the same while transformations on @xmath1 become rather unusual . under field
redefinition , the gauge parameter transforms to @xmath175 , and we find @xmath176 we can still choose the gauge @xmath23 ) .
] , in which case the action ( [ a6 ] ) can be rewritten as @xmath177\nonumber \\ & & + \int d^4x~\frac{1}{2}b k\ , e^{-2 k y } ( a_\mu)^2 \bigg\vert_0~. \label{a7}\end{aligned}\ ] ] we see that the field redefinition has generated bulk and boundary mass terms .
the relation of the coefficients of these mass terms is precisely the one necessary to force a localized massless mode , which can easily be seen by identifying @xmath178 and @xmath179 , in which case the tuning relation ( [ tune ] ) holds . again
the wavefunction is given by @xmath180 .
notice that in this model @xcite , there is only one brane , so that the wavefunction is normalizable only if @xmath168 is positive .
why ca nt we simply start from the action ( [ a6 ] ) , and demand the strange gauge symmetry ( [ gauge ] ) without ever referring to the dilaton @xmath166 ? this would allow us to have a theory with a localized gauge field and _ no _ extra scalar fields .
we can do this , but there are a couple of objections to this approach .
the action ( [ a6 ] ) explicitly violates 5d lorentz invariance in the bulk .
of course the dilatonic action ( [ a5 ] ) appears to violate the symmetry as well , but this is only because the vev of the dilaton spontaneously breaks lorentz invariance , and the symmetry is therefore realized non - linearly through fluctuations about the background . if we do nt have this extra scalar , then the theory breaks 5d lorentz invariance . on the other hand ,
4d lorentz invariance is a symmetry of the action ( [ a6 ] ) , which is the only symmetry we need for phenomenological purposes .
another objection , of course , would be that the unorthodox gauge symmetry in ( [ gauge ] ) would seem a bit unnatural if we did nt know about its origin from the dilaton coupling .
it is a local symmetry nonetheless , indicating a redundancy in the description of the physical degrees of freedom in the system . the pure action ( [ a6 ] ) without a dilaton finds a stronger motivation through deconstruction as shown in @xcite .
in this section we will review this approach and mention some of the advantages of the deconstruction viewpoint .
consider @xmath44 sites each with an su@xmath181 gauge theory .
the sites are connected by bifundamental scalar fields @xmath182 .
the 4d lagrangian is then @xmath183 where the covariant derivative is defined as @xmath184 .
the gauge symmetry is broken to a diagonal su@xmath185 by assuming that the scalar fields acquire vevs proportional to the unit matrix , @xmath186 .
the mass lagrangian is then @xmath187 to reproduce the familiar case of a massless gauge field in ads@xmath0 in the continuum limit , the prescription is to take universal gauge couplings @xmath188 and @xmath4-dependent vevs @xmath189 , where the site `` coordinate '' is @xmath190 .
this gives a flat zero mode gauge field .
however , if we make different choices for the couplings and vevs , in particular assuming nonuniversal gauge couplings of the form @xmath191 and vevs of the form @xmath192 the continuum limit mass lagrangian becomes @xmath193~.\ ] ] this action can be rewritten identically to ( [ a7 ] ) , containing the precise bulk and boundary mass terms which will localize a massless mode . from the deconstructed point of view , this makes complete sense because the su@xmath185 gauge symmetry is still intact , implying the existence of a massless eigenstate which is a linear combination of the @xmath194 .
what is appealing about the deconstruction approach to localizing the gauge field is that there is no reference to a dilaton .
the fluctuations about the vacuum @xmath195 are related in the continuum to @xmath1 , and no other scalar field is present in the theory . indeed , after taking into account these fluctuations , the full action is given by ( [ a6 ] ) .
the zero mode localization is simply a consequence of different gauge theories @xmath196 having different couplings @xmath197 , which is certainly allowed .
the absence of the dilaton is much like the absence of the radion in generic deconstructed theories compared to their full five dimensional counterparts .
the approaches to localizing gauge fields described in the previous section seem very similar to our approach .
the localized wavefunction can be ascribed to a dilaton coupling in a nontrivial background , or to a position dependent gauge coupling . upon a field redefinition , bulk and boundary masses
appear with precisely the right coefficients necessary to localize a zero mode , identical to ( [ mass ] ) .
these observations were first made in @xcite .
although the technical details of each approach are similar , the philosophy behind each is somewhat different .
we are beginning with a pure gauge theory in a slice of ads@xmath0 , and attempting to generate the necessary mass terms through the spontaneous breaking of the symmetry .
the 5d bulk lorentz symmetry is never broken . in the other approaches ,
the localized wavefunction is generated through an auxiliary mechanism , either the dilaton profile or a position dependent coupling , and is crucially dependent on breaking 5d lorentz invariance in the bulk .
moreover , it remains to be shown whether there exists a solution to einstein s equations that produces a linear dilaton background with pure ads@xmath0 geometry .
the number of additional physical scalar particles in each theory is different as well . in our approach ,
the number of scalar fields depends on the representation of the symmetry group in which scalar fields reside , and this number could be quite large .
the analysis of these scalars is also complicated since there is mixing at the quadratic level in the action , and can in principle lead to strong coupling problems in the theory , as elaborated on in section 3 . on the other hand , in order to localize the field with a dilaton , only one scalar is needed per gauge field @xmath90 .
fluctuations about the dilaton background in ( [ a5 ] ) need to be studied , but the analysis should be simpler because these perturbations will not mix with the gauge field at the quadratic level , but only appear as interactions .
on the other hand , the deconstructed approach is even simpler because there are no additional scalar fields , although the continuum theory explicitly breaks 5d lorentz symmetry . at the non - abelian level , these approaches really diverge .
we found it necessary to add brane - localized kinetic terms to restore the equality of the zero mode gauge couplings .
the other approaches accomplish this by having position dependent couplings . to see this , consider the non - abelian generalization of ( [ a5 ] ) . to compare with our approach ,
we redefine the field @xmath198 to generate bulk and boundary mass terms . upon this redefinition ,
the cubic and quartic interactions become @xmath199~. \label{int2}\ ] ] we can view the 5d gauge couplings as position dependent : @xmath200 . inserting the zero mode wavefunction @xmath201 and computing the overlap integrals
, we find that the 4d cubic and quartic couplings are equal , and thus the low energy theory is gauge invariant . in our approach , the 5d gauge couplings are constant in the bulk
. the phenomenology of each localization mechanism would also be different . in our approach , brane kinetic terms
must be included , which can drastically alter the mass spectrum , wavefunctions , and couplings to matter , as illustrated by a number of studies @xcite .
brane kinetic terms are not necessary in the other approaches , but of course may be included for study . finally , the scalar content of each theory is much different , which would affect the phenomenology .
although we have focused on the gauge field @xmath2 in this paper , many models utilize the @xmath1 field in one way or another .
one well known example is the hosotani mechanism @xcite , or radiative symmetry breaking , where quantum corrections induce a mass for the zero - mode @xmath139 , which is then identified as the higgs boson .
a recent twist on this idea , motivated by holography , are composite higgs models @xcite , where the dual of the @xmath1 is a composite pseudo - nambu - goldstone boson of a strongly coupled theory .
the ability to localize the @xmath1 field in these contexts could have interesting effects on the phenomenology of these models . with a massless 5d gauge boson in the bulk
, the profile of @xmath139 is fixed by gauge invariance , and in fact is localized near the ir brane . as is the case with the vector @xmath2 , it is possible to localize the scalar @xmath1 zero mode at different points in the bulk . using bulk and boundary masses generated by spontaneous symmetry breaking to localize
this field goes against the philosophy of the gauge - higgs models where @xmath1 is to be used as an _ alternative _ to spontaneous symmetry breaking .
hence , we will show how the @xmath1 can be localized either through a dilaton coupling , or through deconstruction , similar to the case for @xmath2 in section 4 .
we reiterate that at this point it is unknown whether a 5d system with a linear dilaton solution in a pure ads@xmath0 background exists .
starting from the action ( [ a5 ] ) , we add the following gauge fixing term to diagonalize the action : @xmath202 ^ 2~. \label{lgf2}\ ] ] the @xmath1 action becomes @xmath203~. \label{a9}\ ] ] decomposing the field , @xmath204 , we find that the kaluza - klein modes have a @xmath205-dependent mass , indicating that they are not physical particles .
however for the zero mode this is not the case .
the eigenfunction @xmath206 obeys the equation of motion @xmath207 and with the boundary condition @xmath208 has solution @xmath209 inserting the solution into the action ( [ a9 ] ) , we have @xmath210~,\ ] ] indicating that with respect to a flat metric , the wavefunction is @xmath211 .
the physical zero mode is localized on the uv brane when @xmath212 while for @xmath213 it is localized on the ir brane .
note that this agrees with the zero mode localization of a pure bulk scalar with bulk mass @xmath214 and boundary mass @xmath215 @xcite .
zero mode gauge fields in a compact extra dimension can be localized through the addition of bulk and brane mass terms .
the inclusion of these terms modifies the bulk equation of motion and the boundary conditions such that a zero mode with an exponential wavefunction exists in the theory . in the non - abelian case , however , the cubic and quartic couplings of this zero mode are not equal , meaning that the gauge symmetry of the 4d theory is lost .
we have shown how to recover the 4d gauge symmetry by adding gauge interactions on the boundaries . in this way
the sum of the bulk plus boundary contributions to the interactions of the zero mode restores the equality of the gauge couplings , and then the 4d gauge invariance appears in its usual form .
we can add these brane interactions in a gauge invariant way via brane - localized kinetic terms .
one caveat in this approach is that there are two relations in this theory that must be tuned , one between the bulk and boundary masses to produce a localized massless mode wavefunction , and another between kinetic terms and the boundary mass parameter .
furthermore even though the zero mode had overall positive kinetic energy , via a nonlocal cancellation mechanism , negative kinetic energy terms were required on at least one brane .
these terms would need to be resolved in any sensible uv completion of the model .
there are also caveats related to the additional scalar fields that must be present in the theory . to add mass terms in the bulk and on the boundary in a gauge invariant way
, the terms must be generated by breaking the 5d gauge symmetry , which necessitates introducing additional scalar fields in the theory . in the warped case ,
a complete analysis of these extra scalars is limited because of the nontrivial mixing .
we have instead presented this analysis in flat space , where the eigenfunctions of the different fields are similar .
if one could extend this analysis to the warped case , many of the scalar issues , such as strong coupling , could be addressed more generally .
we have shown , however , that in the warped case with only boundary masses , a strong coupling problem may exist in the scalar sector .
this problem may be resolved by a different choice of boundary conditions or by radiative corrections to the scalar zero mode mass . even if the problems of the scalar sector can not be addressed without a more fundamental uv description , the yang - mills localization will still be preserved at low energies .
the phenomenology of localized gauge fields in the randall - sundrum model would therefore be interesting to explore . if we consider the standard model in the bulk , then
each gauge boson can in principle be localized at different points in the extra dimension .
gauge fields in the bulk were originally considered in @xcite , and it was found that the mass of the first kaluza - klein excitation must be pushed to a rather high scale to suppress four - fermion operators . however , localizing gauge fields can help lower the kaluza - klein mass scale , as illustrated in the abelian case @xcite .
it is conceivable that localizing gauge fields could have other phenomenological benefits .
also , the phenomenology would be different depending on the localization mechanism , either bulk and boundary masses , dilaton coupling , or deconstruction .
perhaps the most interesting aspect of localized gauge fields is the holographic interpretation @xcite .
if the zero mode is localized on the uv brane , we find that the massless particle in the dual theory is primarily an elementary field external to the cft . on the other hand , a zero mode localized on the ir brane corresponds in the dual theory to a composite massless gauge field built out of cft states .
for example , if we consider the standard model gauge fields propagating in the bulk with zero modes localized towards the ir brane , then in the dual theory the forces that we experience - electromagnetic , weak , and strong - are really emergent phenomena , a result of strongly coupled ( unknown ) cft dynamics at low energy .
this is similar to the emergent gravity scenario considered in ref .
@xcite , and adding boundary kinetic terms for gravity as done for non - abelian gauge fields in this paper , suggests a similar nonlinear extension for gravity . finally , we discussed the localization of the @xmath1 field , the scalar component of the 5d gauge field .
we have shown that it is possible to localize this field anywhere in the bulk with a dilaton coupling or via deconstruction .
this opens up new phenomenological possibilities in models where @xmath1 is the pseudo - nambu - goldstone boson responsible for electroweak symmetry breaking .
in particular , motivated by the dual interpretation , the dual higgs field can become even more composite as the zero mode is localized on the ir brane .
we thank emilian dudas and erich poppitz for helpful discussions . t.g .
acknowledges the cern theory division and the galileo galilei institute for theoretical physics in florence for their warm hospitality and the infn for partial support where portions of this work were done . this work was supported in part by a department of energy
grant de - fg02 - 94er40823 at the university of minnesota , and an award from research corporation .
l. randall and r. sundrum , phys .
lett . * 83 * , 3370 ( 1999 ) [ arxiv : hep - ph/9905221 ] .
j. m. maldacena , adv .
* 2 * ( 1998 ) 231 [ int .
j. theor .
* 38 * ( 1999 ) 1113 ] [ arxiv : hep - th/9711200 ] .
s. s. gubser , i. r. klebanov and a. m. polyakov , phys .
b * 428 * ( 1998 ) 105 [ arxiv : hep - th/9802109 ] .
e. witten , adv .
* 2 * ( 1998 ) 253 [ arxiv : hep - th/9802150 ] .
n. arkani - hamed , m. porrati and l. randall , jhep * 0108 * ( 2001 ) 017 [ arxiv : hep - th/0012148 ] .
r. rattazzi and a. zaffaroni , jhep * 0104 * ( 2001 ) 021 [ arxiv : hep - th/0012248 ]
. m. perez - victoria , jhep * 0105 * ( 2001 ) 064 [ arxiv : hep - th/0105048 ] .
k. agashe and a. delgado , phys .
d * 67 * , 046003 ( 2003 ) [ arxiv : hep - th/0209212 ] .
h. davoudiasl , j. l. hewett and t. g. rizzo , phys .
b * 473 * , 43 ( 2000 ) [ arxiv : hep - ph/9911262 ] .
a. pomarol , phys .
b * 486 * , 153 ( 2000 ) [ arxiv : hep - ph/9911294 ] .
t. gherghetta , arxiv : hep - ph/0601213 . k. ghoroku and a. nakamura , phys .
d * 65 * , 084017 ( 2002 ) [ arxiv : hep - th/0106145 ] .
i. kogan , s. mouslopoulos , a. papazoglou , and g. ross , nucl .
b * 615 * , ( 191 2001 ) [ arxiv : hep - ph/0107307 ] .
b. batell and t. gherghetta , phys .
d * 73 * , 045016 ( 2006 ) [ arxiv : hep - ph/0512356 ] .
a. kehagias and k. tamvakis , phys .
b * 504 * , 38 ( 2001 ) [ arxiv : hep - th/0010112 ] .
h. abe , t. kobayashi , n. maru , and k. yoshioka phys .
d * 67 * , 045019 ( 2003 ) [ arxiv : hep - ph/0205344 ] .
y. hosotani , phys .
b * 126 * , 309 ( 1983 ) ; phys .
b * 129 * , 193 ( 1983 ) .
s , j. huber and q. shafi , phys .
d * 63 * , 045010 ( 2001 ) [ arxiv : hep - ph/0005286 ] .
t. gherghetta and a. pomarol , nucl .
b * 586 * , 141 ( 2000 ) [ arxiv : hep - ph/0003129 ] .
h. davoudiasl , j.l .
hewett , and t.g .
rizzo phys .
d * 68 * , 045002 ( 2003 ) [ arxiv : hep - ph/0212279 ] .
f. del aguila , m. perez - victoria , and j. santiago , jhep * 0303 * ( 2003 ) 051 [ arxiv : hep - th/0302023 ] .
l. randall and m. schwartz , jhep * 0111 * , 003 ( 2001 ) [ arxiv : hep - th/0108114 ] .
m. tachibana , [ arxiv : hep - th/0108164 ] .
m. carena , e. ponton , t. tait , and c.e.m .
wagner phys .
d * 67 * , 096006 ( 2003 ) [ arxiv : hep - ph/0212307 ] .
r. contino , y. nomura and a. pomarol , nucl .
b * 671 * , 148 ( 2003 ) [ arxiv : hep - ph/0306259 ] .
k. agashe , r. contino , and a. pomarol , nucl .
b * 719 * , 165 ( 2005 ) [ arxiv : hep - ph/0412089 ] .
t. gherghetta , m. peloso and e. poppitz , phys . rev .
d * 72 * , 104003 ( 2005 ) [ arxiv : hep - th/0507245 ] . | we present a mechanism to localize zero mode non - abelian gauge fields in a slice of ads@xmath0 . as in the u(1 ) case , bulk and boundary mass terms allow for a massless mode with an exponential profile that can be localized anywhere in the bulk . however in the non - abelian extension , the cubic and quartic zero mode gauge couplings do not match , implying a loss of 4d gauge invariance .
we show that the symmetry can be restored at the nonlinear level by considering brane - localized interactions , which are added in a gauge invariant way using boundary kinetic terms .
possible issues related to the scalar sector of the theory , such as strong coupling and ghosts , are also discussed .
our approach is then compared with other localization mechanisms motivated by dilaton gravity and deconstruction .
finally , we show how to localize the scalar component @xmath1 zero mode anywhere in the bulk which could be relevant in gauge - higgs unification models .
umn - th-2527/06 + november 2006 * yang - mills localization in warped space * brian batell and tony gherghetta + + |
the instability of hydrodynamic accretion from a moving medium and its relevance to the temporal behavior of accreting x - ray sources has been debated since the discovery of a flip - flop instability in two - dimensional hydrodynamic simulations of an accreting compact object ( * ? ? ? * hereafter ft ) .
this non - axisymmetric instability has been subsequently seen in numerous hydrodynamic simulations of two - dimensional planar accretion @xcite , but has not been unambiguously identified in three - dimensional simulations @xcite . given that accretion - powered x - ray sources in be - type binary systems are expected to accrete from a planar disk expelled by the companion be star , attention has focussed on the relationship of the flip - flop instability to these be - type x - ray pulsars .
@xcite interpreted the recurrent flares in exo 2030 + 375 in terms of the episodic accretion resulting from the flip - flop instability .
@xcite argued that the variable accretion torques associated with the flip - flop instability may explain the erratic spin behavior in these systems . the first quantitative description of hydrodynamic accretion was given by @xcite , who were attempting to explain climate changes on earth in terms of an increase in solar output due to the addition of accretion energy as the sun passed through an interstellar cloud .
shortly thereafter the same authors proposed hydrodynamic accretion from interstellar gas as a solution to the paradox of the extremely short lifetimes of massive stars compared to geologic timescales , which at the time was considered an exceedingly embarrassing consequence of the present theory of stellar evolution " @xcite .
the hoyle - lyttleton accretion model is parameterized by an object of mass @xmath7 moving at a speed @xmath8 through a uniform medium of density @xmath9 .
using ballistic orbits ( neglecting pressure effects ) , @xcite defined an accretion radius given by @xmath10 such that gas approaching the star with an impact parameter less than @xmath0 would collide on an accretion line behind the star and would be left with insufficient kinetic energy to escape the gravitational potential of the star .
the mass accretion rate is expected to be comparable to the mass flux through a circle of radius @xmath0 far upstream : @xmath11 a thorough review of hoyle - lyttleton accretion theory and subsequent numerical studies is given by @xcite .
we provide only a brief synopsis here .
this basic model of hydrodynamic accretion was first investigated numerically by @xcite , who used 2d axisymmetric steady - state solutions to quantify the mass accretion rate . for an upstream mach number of 2.4 he found an accretion rate of @xmath12 .
subsequent time - dependent 2d axisymmetric simulations have , for the most part , shown that axisymmetric hydrodynamic accretion is stable @xcite . @xcite found steady axisymmetric accretion flows provided that the inner boundary condition was totally absorbing .
@xcite found steady solutions for low mach number flows , but they found the accretion flow to be unsteady if the upstream mach number was greater than 2 .
these latter results , however , are likely plagued by poor resolution near the accreting object .
the authors attributed the unsteady flow to gas flowing towards the accretor within the accretion column , but slightly missing the accretor .
one might expect the gas in the accretion column to be completely absorbed if a higher grid resolution near the accretor fully resolved the flow .
this picture of steady , axisymmetric accretion is altered in the context of an x - ray binary system in which a compact object is accreting from the stellar wind of a companion star . in this case
gradients in both density and velocity of the flow past the compact star will break the axisymmetry , and in particular may lead to the accretion of angular momentum as well as mass @xcite .
@xcite used 2d hydrodynamic simulations to study the dynamics of such wind accretion in a close binary system .
they presented some models for which the flow near the accreting star sporadically changed its direction of rotation - a phenomenon the authors referred to as a ` flip - flop flow . ' shortly thereafter , numerical studies of the more idealized problem of hoyle - lyttleton accretion in 2d planar flow found a similar behavior , with the accretion wake flipping from side to side . the result of this flip - flopping
is the formation of temporary accretion disks spinning in alternating directions , with a burst of mass accretion when the direction of rotation is flipped .
this phenomenon has been seen in many subsequent simulations , and has come to be known as the ` flip - flop ' instability . while hoyle - lyttleton accretion is a very idealized problem , it is important to note that the flip - flop instability was first seen in more realistic simulations of gas flow within a binary star system @xcite and in subsequent simulations that incorporated several key physical processes including x - ray heating and radiative cooling @xcite .
although in the latter case it was found that strong compton heating due to a high x - ray luminosity of the accreting star suppressed the flip - flop instability .
however , both of theses studies of accretion in binary systems were restricted to 2d .
the increase in density and pressure as the accretion flow approaches the compact star will be different in three dimensions compared to two , raising the possibility that the stability of 3d accretion may differ from these 2d results .
the only published simulations of 3d hoyle - lyttleton accretion with a reasonably small accretor were a series of studies using a nested cartesian grid @xcite .
the accretion in these 3d studies was found to be unstable , but the evolution did not resemble the flip - flop behavior found in 2d planar flow .
in particular , the amplitude of variations remained considerably smaller than that found in 2d , and no evidence of steady quasi - keplerian disk flow was ever seen .
nonetheless , some of the 3d models reported in @xcite did exhibit a quasi - periodic swinging of the accretion shock leading to brief epochs of strong rotational flow near the accretor .
this behavior was particularly evident for the models with the smallest accretors .
although these 3d runs used nested grids to improve spatial resolution near the accreting star , each nested grid was only @xmath13 zones .
this resolution is relatively low compared to that used in many of the 2d simulations exhibiting the flip - flop , and may have contributed to the lack of detection of the flip - flop instability in 3d .
independent of the question of whether adiabatic , 2d planar flow is an appropriate model for astrophysical systems , the flip - flop instability itself has drawn significant scrutiny and criticism over the past two decades .
many examples of 2d planar accretion simulations have been presented in the literature , with mixed results as tabulated by @xcite .
several authors have pointed out various shortcomings of existing 2d planar hoyle - lyttleton simulations .
these include , but are not necessarily limited to , poor numerical resolution @xcite , inaccurate algorithms @xcite , inappropriate boundary conditions upstream and at the surface of the accretor @xcite , lack of conservation of angular momentum @xcite , and an unrealistically large accretor .
these criticisms continue to cast doubt and confusion over the existence of the flip - flop instability .
in this paper we present an in - depth study intended to provide a better understanding of the properties of the flip - flop instability .
we take care to address the many issues raised about previous numerical simulations and we use the increased availability of computing power to more fully explore parameter space .
we describe our numerical method in section 2 , building from the original flip - flop simulations of .
we present our results in section 3 , where we show the flip - flop instability is a true overstability of the accretion wake , and provide a summary of our conclusions in section 4 .
we use the time - dependent hydrodynamics code vh-1 to solve the euler equations for an ideal gas characterized by an adiabatic index @xmath1 in 2d polar coordinates with a point source of gravity at the coordinate origin .
this code is similar to the code used by .
both codes are based on the piece - wise parabolic method @xcite , but the code used here uses the lagrange - remap formulation while the code used by used the direct - eulerian formulation .
none - the - less , we expect very similar results from these two codes .
slightly greater than @xmath0
. the parameters of the model shown here are @xmath3 and @xmath14,width=336 ] to this end , we begin by presenting results for a simulation set up to mimic a simulation in , using the same grid , boundary conditions and simulation parameters .
specifically , we use a numerical grid with 100 radial zones and 200 azimuthal zones .
the inner radius corresponding to the surface of the accreting star is @xmath15 , the first zone has a width of @xmath16 , and the subsequent radial zones increase in width by a factor of 1.03 .
this puts the outer radius of the computational grid at @xmath17 .
the flow at the upstream boundary is given a uniform velocity with a mach number of 4 .
the gradients for all flow variables at the outer boundary were set to zero for the downstream half of the grid , with the further condition that the radial velocity remained positive .
we imposed an absorbing boundary condition at the surface of the accreting star by setting the density and pressure to very small values inside of this boundary @xcite .
we note that the flip - flop instability has been shown to be relatively insensitive to the specific boundary conditions imposed on the accreting surface @xcite .
we imposed a transverse density gradient at the upstream boundary of the form @xmath18 with @xmath19 .
the results of this simulation , shown in figure [ fig : bruce ] as a trace of the specific angular momentum of the accreted gas , is qualitatively similar to the corresponding simulation presented in .
( note that we normalize the specific angular momentum to that of a circular orbit at the radius of the accretor , @xmath20 , so that the saturation value when a temporary accretion disk has formed is independent of the accretor radius . ) in particular , the accretion is marked by phases of disk accretion with relatively constant specific angular momentum corresponding roughly to that of a circular orbit at the surface of the accretor .
these disk phases occur for both directions of rotation , with sometimes very abrupt transitions between the two .
given the dramatic increase in available computing power in the two decades since the discovery of the flip - flop instability , we are at liberty to run a large number of simulations to refine our computational model , to run at higher spatial resolution , and to explore the parameter space more completely . in the following subsections
we delineate the changes made to this initial numerical model in order to arrive at what we consider to be high - fidelity hydrodynamic simulations demonstrating the exponential growth of the flip - flop instability .
a serious shortcoming of our initial model was the presence of density inhomogeneities in the upstream flow due to numerical noise .
the source of these numerical fluctuations was traced to the conservation of total energy at the expense of ensuring constant entropy in the upstream supersonic flow traveling at various angles with respect to the computational grid . to reduce errors in internal energy created by subtracting kinetic energy from the conserved total energy
, we modified the algorithm to remap internal energy rather than total energy when the mach number of the flow exceeds 3.0 .
this modification still provides for accurate shock jump conditions because the original conservation of total energy is used in zones containing a shock , but it now allows one to model uniform supersonic ( mach = 4 ) planar flow across a cylindrical polar grid with minimal numerical artifacts . assuming a uniform planar flow at the upstream boundary neglects any gravitational effects of the flow from infinity to that boundary . as a result ,
the flow near the accretor becomes dependent on the chosen radius of the outer boundary @xcite .
a more realistic boundary condition is to assume a ballistic trajectory from infinity to the outer boundary @xcite .
the resulting flow velocity at a given point ( @xmath21 ) in space is then @xcite @xmath22 @xmath23 where @xmath24 is the impact parameter at infinity of the stream line passing through that point , @xmath25 .
\label{eqn : zeta}\ ] ] one can then find the local gas density using mass conservation , equating the mass flux through a region bounded by @xmath24 and @xmath26 at infinity to the mass flux through a region bounded by @xmath27 and @xmath28 for a point along the same stream line .
for the 2d planar geometry used in this paper the result is @xmath29 once the density is known , the pressure can be found by assuming the flow is isentropic .
the flow variables upstream of the accretion shock in our simulations remain within 0.5 percent of this analytic solution for an upstream mach number of 4 .
although often not explicitly stated , previous simulations were typically started from an initial state of perfectly uniform planar flow .
the early evolution is then dominated by the formation of a tail shock along the symmetry axis behind the accretor and the subsequent expansion of the region of shocked gas .
this highly variable initial flow masks the origin of any possible instability , and indeed makes it impossible to assess whether the observed behavior is a true instability in the sense that an initial equilibrium state will tend to evolve away from equilibrium .
we address this problem by creating a steady - state solution for initial conditions .
we evolve our model on only half of the grid , enforcing reflection symmetry about the line of centers .
these ` half ' simulations are initialized with the ballistic flow model and allowed to evolve for several flow crossing times , during which the accretion shock settles into a steady state .
these relaxed solutions , an example of which is shown in figure [ fig : schematic ] , are then used to initialize the full 2d simulations .
the downstream boundary condition in previous numerical simulations has traditionally been a condition of zero gradients , allowing flow to smoothly move off the grid .
this is a reasonable approach if the flow is moving supersonically off the grid , but can lead to spurious reflections if the flow across the boundary is subsonic . to investigate the downstream conditions in the accretion wake we examined the mach number of the flow both in our steady - state initial conditions and in the time - dependent unstable flow . in the steady - state flow the radial velocity on the symmetry axis does not become positive until @xmath30\,r_a$ ] , so we focus our attention to flow slightly off this axis . in figure
[ fig : downstream ] we plot the mach number of the radial flow as a function of distance from the accretor for angles of 7 and 14 degrees from the symmetry axis . for reference ,
the shock is at an angle of about 21 degrees far downstream . at 7 degrees from the axis ,
the flow exceeds mach 1 only after traveling @xmath31 , and even then it barely rises above unity .
we also show the minimum mach number inside of the accretion wake for a simulation subject to the flip - flop instability .
the unsteady flow falls in an envelope defined by the 7 and 14 degree traces . setting the outer radius to greater than 10 @xmath0 , as done by @xcite ,
ensures that the gas will propagate supersonically off of the grid under most circumstances , thereby minimizing wave reflection from the boundary . ) .
the two dashed lines are from a steady - state solution , tracing the mach number along angles of 7 and 14 degrees from the axis .
the solid lines show the minimum mach number inside the wake at different times in a simulation of unstable accretion flow.,width=336 ] @xcite demonstrate that the dynamics of the temporary accretion disks formed during the flip - flop instability are less dependent on the underlying computational mesh if the hydrodynamic algorithm strictly conserves angular momentum rather than linear momentum .
this result is not unexpected given that during the disk epochs mass accretion is driven by spiral shocks within a roughly keplerian flow , and any mass accretion is closely tied to the transport of angular momentum . to this end
we made a minor change to the ppmlr algorithm used by vh-1 in order to explicitly conserve angular momentum .
this code modification has been used to quantify disk accretion driven by spiral shocks @xcite , measuring effective values of the disk @xmath32 parameter as small as @xmath33 . the computational grid used by resulted in zones with extreme aspect ratios ( 8:1 ) near the surface of the accretor , while other researchers used a single zone to represent the accretor @xcite .
more recent work has attempted to improve this shortcoming by running identical simulations with increasing numerical resolution @xcite or by applying adaptive mesh refinement to more accurately resolve the flow near the accretor @xcite .
we chose to use a grid similar to that of , namely a uniform grid in @xmath34 and a logarithmically increasing grid in @xmath27 .
however , we followed the approach of @xcite and chose parameters that kept the aspect ratio of all zones roughly square .
this means both that @xmath35 is constant throughout the grid , and also that @xmath36 is much smaller than the radius of the accretor given that we use a minimum of 200 zones in the angular direction and we require that @xmath36 be comparable to @xmath37 .
for our standard model ( a in table 1 ) we chose parameters comparable to that of , namely an upstream mach number of four , @xmath38 , and @xmath3 .
we did not impose any asymmetry in the upstream flow . this simulation does exhibit the flip - flop instability ( figure [ fig : aso08_seq ] ) , however the oscillation of the accretion wake is less abrupt than found in the original simulations of .
much of the evolution is characterized by the accretion wake swinging back and forth , but remaining attached to the accretor . only for brief epochs
does the shock wrap around the leading edge of the accretor and ` detach ' , resulting in a temporary accretion disk as seen in the last frame of figure [ fig : aso08_seq ] .
moreover , the flow never exhibits an abrupt transition from a disk of one rotation directly to a counter - rotating disk .
instead , the disk disappears and the accretion wake swings back and forth one or more times before forming another disk . the mass accretion rate and the specific angular momentum of the accreted gas are shown in figure [ fig : aso01chart ] . the mass accretion rate is normalized to the 2d planar equivalent of equation ( [ eqn : mdot ] ) : @xmath39 during the initial evolution , before the oscillations of the accretion shock grow to the point of driving the shock around the leading edge of the accretor , the mass accretion rate remains at a roughly constant value of @xmath40 .
this is in good agreement with the original calculation of @xcite for axisymmetric hoyle - lyttleton accretion as well as subsequent simulations of planar accretion flow @xcite . during the brief epochs of disk accretion ,
the mass accretion rate is both substantially lower and highly variable .
the termination of each disk phase is marked by a sudden spike in the mass accretion rate as the bulk of the mass that had been orbiting in a quasi - keplerian accretion disk is rapidly accreted on the relatively short free - fall timescale . to quantify the growth of the instability we fit the early evolution ( before the oscillations reach a maximum amplitude ) with an exponentially growing sinusoid : @xmath41
the fitted function shown in figure [ fig : aso01chart ] corresponds to an oscillation frequency of @xmath42 and an exponential growth rate of @xmath43 .
this corresponds to the amplitude approximately doubling each period of oscillation .
note that the oscillation period ( @xmath44 ) is comparable to the keplerian period at the accretion radius ( @xmath45 ) .
eventually some of the peaks in the trace of @xmath46 become choppy , indicating those points when an accretion disk has formed ( figure [ fig : aso01chart ] ) .
, of the accreted gas .
the dashed curve in the top panel is the analytic fit described in equation ( 9 ) .
the mass accretion rate , shown in the bottom panel , is relatively constant except during brief periods of time when a disk forms and the mass accretion rate drops substantially .
this low accretion phase is always followed by a large jump , when the mass in such a disk is flushed onto the accretor resulting in a momentary spike in mass accretion.,width=336 ] these basic results were found to be relatively insensitive to the mach number of the flow or the presence of a small asymmetry in the upstream flow .
the original discovery of the flip - flop instability came from simulations in which a density gradient was imposed on the ambient medium . to see
if the instability is affected by small gradients , we ran a simulation ( model c ) with a transverse density gradient as in equation ( [ eqn : epsilon ] ) , but with @xmath47 replaced by @xmath24 as given in equation ( [ eqn : zeta ] ) .
we used an asymmetry value of @xmath48 .
we also ran a simulation ( model b ) with no imposed asymmetry but with a higher mach number of 10 .
the results for these runs are shown in figure [ fig : compare ] in the form of the evolution of the accreted specific angular momentum . in each case
the growth of the flip - flop instability was similar to the canonical run with @xmath49 and a mach number of 4 .
the growth rate of the instability in the higher mach number flow was roughly 25% higher , but otherwise the three simulations were very similar .
is shown for models a , b ( mach = 10 ) , and c ( @xmath50).,width=336 ] is shown for an accretor of radius @xmath51 in the top plot , along with @xmath52 for the canonical model with @xmath38 ( dashed line ) .
the bottom plot shows @xmath52 for an accretor of radius @xmath53 , again compared against the canonical model . , width=336 ] for applications to accreting neutron stars and black holes , we are interested in values of @xmath54 that are far smaller than our canonical value of @xmath55 .
most simulations in the literature are limited to relatively large accretors because decreasing @xmath54 becomes computationally very expensive .
the timestep in these simulations is set by the courant condition near the surface of the accretor , and hence scales as @xmath56 .
some of our extended runs with small accretors ran for more than @xmath57 time steps .
in addition to our canonical model we present results for models with an accretor radius of @xmath58 and @xmath59 in figure [ fig : radiij ] . a smaller accretor is more unstable .
the growth rate is larger and the flow forms an accretion disk quicker . in the intermediate case
the growth rate was roughly 50% higher and the first disk was formed half an oscillation earlier . in the smallest accretor case the growth rate was more than twice as high as the canonical run and the first disk formed in half the time - after only one oscillation of the accretion shock . the accretion disk also persists longer for smaller accretors . in the canonical model
a temporary disk is formed for only brief epochs ( typically one flow time across the accretion diameter ) with little impact on the quasi - periodic behavior of @xmath46 . in the case of the smallest accretor
a given disk phase can last for several flow times , with very abrupt transitions between counter - rotating disks .
the structure of the temporary accretion disks as a function of accretor size is illustrated in figure [ fig : disks ] .
one can see a general trend of decoupling between the spiral shocks of the accretion disk and the leading bowshock as the size of the accretor is reduced . in model
a with the largest accretor the leading bowshock is a direct extension of a spiral shock reaching down to the surface of the accretor . in the intermediate model
the spiral shock wraps at least one full revolution around the accretor before it becomes a part of the external bowshock . in the case of the smallest accretor ( model e )
, the spiral shocks can be traced out to the bowshock , but it involves several windings . the result in this last case is a disk - like accretion flow that is only weakly coupled to the bowshock .
the oscillation period of the flip - flop instability in the small amplitude regime ( before the formation of any accretion disk ) is relatively insensitive to the accretor radius , with values comparable to the keplerian period at the accretion radius .
the dependence of @xmath60 on @xmath54 is hard to quantify because @xmath60 decreases somewhat as the amplitude of oscillation increases . as a result ,
the value of @xmath60 derived from the simulation data depends on the time window within which one fits the analytic function ( eq . [ [ eqn : jdot ] ] ) .
this window is forced to be early in the evolution for the small accretor case because of the rapid growth , but a better fit is found if a later window is used in the runs with a larger accretor .
one can say , however , that the differences in @xmath60 for different values of @xmath54 are smaller than the variations in @xmath60 with oscillation amplitude . .
the density scale is different in each of the three images to allow similar detail to be seen in each . ]
clrrrr a & 0.037 & 4 & 1.333 & 0 & 0.070 + b & 0.037 & 10 & 1.333 & 0 & 0.088 + c & 0.037 & 4 & 1.333 & 0.06 & 0.070 + d & 0.0125 & 4 & 1.333 & 0 & 0.104 + e & 0.0037 & 4 & 1.333 & 0 & 0.170 + f & 0.037 & 4 & 1.667 & 0 & 0.000 + g & 0.037 & 4 & 1.400 & 0 & 0.068 + h & 0.037 & 4 & 1.500 & 0 &
0.040 + i & 0.037 & 4 & 1.550 & 0 & 0.031 + j & 0.0125 & 4 & 1.600 & 0 & 0.011 + k & 0.00185 & 4 & 1.667 & 0 & 0.008 + the mass accretion during the disk phase is driven by spiral shocks @xcite .
this is illustrated in figure [ fig : orbit ] , where we show a streamline computed from the gas velocity at a fixed moment in time .
note that this streamline is not the same as the path taken by a fluid element over time , but it provides a better illustration of the effects of spiral shocks on the flow . when orbiting gas encounters a spiral shock , its orbital velocity is decreased , resulting in the kinks seen in the streamline .
this process robs the gas of angular momentum , which is transferred to gas at larger radii through the pressure gradient behind the shock .
the loss of angular momentum in turn leads to an inward radial drift , and eventually accretion onto the central object .
the specific angular momentum of the gas accreted during a disk phase is @xmath61 , independent of the size of the accretor .
this is slightly larger than the keplerian value of unity at the surface of the accretor , consistent with the eccentricity of the orbiting gas near the accretor .
illustrates the elliptical nature of the orbits and the influence of spiral shocks on the accretion of the orbiting gas .
the shading represents gas pressure .
the spiral shocks are marked by discontinuities in pressure as well as slight kinks in the computed streamline.,width=326 ] finally , we found that the adiabatic index has a dramatic influence on this instability .
we ran several simulations varying mach number and density asymmetry with @xmath5 and @xmath38 , and in all cases the oscillation of the accretion shock remained bounded .
typically the shock cone begins oscillation very early on but the oscillations never grow beyond a relatively small amplitude and no accretion disk is ever formed .
this dependence on @xmath1 is consistent with the results of @xcite .
they ran a simulation with a homogeneous upstream flow and @xmath5 and reported steady flow for an absorbing inner boundary .
they also ran a simulation similar to with @xmath3 and @xmath19 and found the flip - flop instability , regardless of the inner boundary condition .
the authors attributed this difference to @xmath62 , but given the results reported here this difference is more likely due to @xmath1 .
, for @xmath15 .
the simulation with @xmath63 exhibited small oscillations at a constant amplitude , but no evidence for growth.,width=326 ] to further investigate this dependence on @xmath1 we ran a series of simulations in which @xmath1 was the only parameter changed . to quantify the instability we fit @xmath46 as in figure [ fig : aso01chart ] .
the results , shown in figure [ fig : omega ] , show a monotonic decrease in the growth rate with increasing @xmath1 . by @xmath64 the accretion wake
is no longer unstable .
the wake still swings back and forth , but the amplitude of those oscillations is only a few degrees and it remains remarkably constant . and adiabatic index @xmath65.,width=326 ] this critical value of @xmath1 , above which the flow is ( at least marginally ) stable , depends on the size of the accretor . as shown above , a smaller accretor produces a faster growth rate of the instability .
a smaller accretor also means a larger critical value of @xmath1 .
we ran simulations with @xmath64 for smaller accretors ( see table 1 ) and found the flow to be unstable , with increasing growth rate for decreasing @xmath54 as in the case with @xmath66 .
we also ran simulations with @xmath65 , searching to see if such flow is unstable if the size of the accretor is sufficiently small . with @xmath67 , the amplitude of the the oscillating accretion shock grew very slowly with @xmath68 , but then seemed to level off before the amplitude became big enough for the shock to flip over the leading edge and produce a disk .
the growth rate was about the same for a slightly smaller accretor , @xmath69 , but as shown in figure [ fig : flipped ] the amplitude was big enough to produce a flip after only six oscillations .
we have presented high - fidelity simulations of 2d planar accretion onto a compact object in order to better understand the flip - flop instability originally discovered by . from these simulations
we can draw the following conclusions : \1 ) first and foremost , we have shown that the flip - flop is a true overstability , where a small displacement of an equilibrium accretion shock leads to oscillations with exponentially growing amplitude .
\5 ) for a given accretor size , the stability of the flow is dependent on the adiabatic index of the gas , with the growth rate decreasing as @xmath1 increases from 4/3 .
planar accretion is stable for @xmath70 in the case of a large accretor ( @xmath38 ) , while @xmath65 is only unstable for the smallest accretor used in our simulations , @xmath71 .
this work was supported in part by nsf grant oci-749248 .
t. c. p. was supported in part by an undergraduate research award from north carolina state university .
the authors acknowledge the texas advanced computing center ( tacc ) at the university of texas at austin for providing computing resources that have contributed to the research results reported within this paper .
we thank the anonymous referee for pointing out that the critical @xmath1 for unstable flow might depend on the radius of the accretor . | we revisit the flip - flop instability of two - dimensional planar accretion using high - fidelity numerical simulations . by starting from an initially steady - state axisymmetric solution , we are able to follow the growth of this overstability from small amplitudes . in the small - amplitude limit , before any transient accretion disk is formed , the oscillation period of the accretion shock is comparable to the keplerian period at the hoyle - lyttleton accretion radius ( @xmath0 ) , independent of the size of the accreting object .
the growth rate of the overstability increases dramatically with decreasing size of the accretor , but is relatively insensitive to the upstream mach number of the flow .
we confirm that the flip - flop does not require any gradient in the upstream flow . indeed
, a small density gradient as used in the discovery simulations has virtually no influence on the growth rate of the overstability .
the ratio of specific heats does influence the overstability , with smaller @xmath1 leading to faster growth of the instability . for a relatively large accretor ( a radius of @xmath2 )
planar accretion is unstable for @xmath3 , but stable for @xmath4 .
planar accretion is unstable even for @xmath5 provided the accretor has a radius of @xmath6 .
we also confirm that when the accretor is sufficiently small , the secular evolution is described by sudden jumps between states with counter - rotating quasi - keplerian accretion disks . |
one of the most impressive experiments using trapped bose gases is the interference experiment of ketterle and co - workers.@xcite two condensates are are separately prepared and allowed to overlap .
an interference pattern then arises showing the remarkable quantum coherence of the condensates .
there have been other interesting condensate interference experiments as well.@xcite@xmath0@xcite if one assumes that the separated clouds initially have a definite phase relation , then the experiments are well - described by straight - forward theory.@xcite however , questions immediately arise .
do separately prepared condensates have a phase relation?@xcite the preparation of the sample certainly did not involve the establishment of a state with known phase .
more likely the particle number in each cloud would have , or could have , been initially known ahead of time .
nevertheless an interference pattern emerges with some well - established phase .
so how is this possible ?
the question was answered in several theoretical papers showing how a phase appears even when the two clouds are prepared in fock states , that is , states with sharply known particle numbers.@xcite@xmath0@xcite in this paper we revisit the question and present a derivation of this result .
this result is quite satisfying , since it in a sense justifies the usual simple assumption of the interferring coherent systems having a well - defined , but completely unknown , initial phase relationship . in order to discuss the properties of condensates and superfluids
one usually introduces theoretically a so - called `` order parameter '' or `` condensate wave function '' defined as @xmath1 where @xmath2 is an operator destroying a particle at position @xmath3 the resulting quantity has a magnitude equal to the square root of the condensate density and also a phase . obviously the state in which @xmath1 is non - zero
can not have a fixed particle number .
with one of these wave functions for each condensate , it is straightforward to discuss interference of the two , since each one has its own phase , and an interference pattern arises with a relative phase equal to the difference between the individual phases .
in essence one has described each condensate by a single - particle wave function so that interference is no more than the overlap and interference of two classical waves .
however , one can ask how this single - particle wave function might have arisen .
indeed its existence involves the question of `` spontaneously broken gauge symmetry''@xcite@xmath4@xcite , the necessity of which has been brought into question in recent years.@xcite@xmath0@xcite suppose we consider a condensate described by a wave function @xmath5 one might describe the direction specified by the angle @xmath6 by a `` spin '' in a two - dimensional plane .
how do we prepare such a state ?
what is it that picks out the direction of this pseudo - spin from the all the degenerate possible directions ?
there is an analogy with ferromagnetism , where there is symmetry in the possible degenerate three - dimensional directions of the magnetization .
somehow , say , because of a small external field , a particuliar direction is space is picked out for the magnetization . in a similar way the phase angle is picked out .
one can suppose in the ferromagnetism case that , in actual practice , there is always some small field to pick out a preferred direction , so that the symmetry is broken .
however , the local field that is used theoretically to pick a phase direction does not exist in nature .
indeed one would have to prepare a state that does not have a fixed particle number .
the treatment of phase ( actually relative phase ) in sec .
iv , and the spontaneous appearance of a relative phase under the effect of measurement of particle position in fock states , avoids violating particle conservation and does not require use of any symmetry - breaking field and so helps in this regard . a closely related feature of the discussion is the idea that the phase emerging from successive measurements of particle position starting with a fock state is somewhat like the emergence of a `` hidden '' or `` additional '' variable in quantum mechanics.@xcite@xmath7@xcite@xmath7@xcite was the phase there before the experiment started , or did the experiment itself cause it to take on its final value ?
hidden variables can be invoked to specify non - commuting variables . in standard
quantum mechanics , particle number and relative phase can be considered conjugate variables ; the knowledge of one excludes that of the other . as we measure particle position we will see that the knowledge of the particle number becomes less certain while the uncertainty of relative phase decreases . in the next sections we first discuss the kind of state that has a known phase .
obviously with this state the interference pattern emerges with just the prepared phase . among these states
are are the coherent states of glauber.@xcite these can either have particle number completely unknown or have the total number of particles in two clouds known ( in which case they are called `` phase states '' ) , although the number in each cloud is still unknown .
we then derive the interference pattern starting with fock states and see the emergence of a relative phase even though no phase was present at the beginning of the experiment ( or was at least `` hidden '' ) .
we even find a way to prepare a state that has a known relative phase .
the controversal theoretical constructs are seen to be unnecessary .
a gas ( or liquid ) undergoing bose - einstein condensation ( bec ) is often described by a classical field known as an `` order parameter '' or `` condensate wave function . ''
such a quantity can arise in several ways .
suppose that @xmath2 represents a second - quantized operator destroying a boson at position @xmath8
. then the one - particle density matrix is defined as @xmath9 penrose and onsager@xcite showed that a criterion for a bose condensate or `` off - diagonal long - range order '' is that the density matrix have the form @xmath10 where @xmath11 vanishes when @xmath12 .
the function @xmath13 is the condensate wave function .
one often assumes the system in a state such that the destruction operator itself has a non - zero average : @xmath14 where @xmath15 is the condensate density and @xmath16 its phase .
such a state is said to have `` spontaneously broken gauge symmetry '' because a particular phase ( out of many possible degenerate phase states ) has been chosen.@xcite@xmath0 @xcite the gross - pitaevskii equation is a non - linear schrodinger equation for @xmath17 which has been used extensively , with remarkable success , to describe interacting trapped bec gases within mean - field approximation.@xcite in describing the interference pattern of the experiment of ref .
one has to consider the overlap of bose clouds released from harmonic oscillator traps.@xcite this leads to some interesting features such as fringes whose separation changes with time . in our analysis here
we will consider only plane waves and ignore any time evolution .
thus suppose we have an order parameter that involves two condensate clouds , having condensate densities @xmath18 and @xmath19 in momentum states @xmath20 and @xmath21 this dual order parameter has the form @xmath22 the density of the combined system is then @xmath23 where @xmath24 @xmath25 , @xmath26 and @xmath27 we have an interference pattern with _ relative _ phase @xmath28 the phase shift @xmath16 is measureable although the individual phases @xmath29 and @xmath30 are not .
this analysis is simple , but it requires the preparation of the system in a state with known individual phases .
how do we do that ?
what is the nature of such a state ?
clearly the expectation value of eq .
( [ orderpar ] ) can not be in a state of definite particle number or the expectation value would vanish .
next we investigate this question more deeply .
as noted by johnston@xcite in this journal , the coherent states introduced by glauber@xcite for photons are appropriate for superfluids .
these well - know states have been reviewed in this journal on occasion,@xcite and appear in texts@xcite as well .
they are also called `` classical states '' and are the minimum uncertainty states of the harmonic oscillator.@xcite here we do not use them in full generality , but rather use a subset of them known as `` phase states . ''
( those wishing to see the full treatment of coherent states in a treatment of a condensate wave function are referred to appendix a. ) phase states describe two condensates ( in states @xmath20 and @xmath31 ) with variable particle numbers , @xmath32 @xmath33 ( both macroscopic ) , but fixed total number @xmath34 no other momentum states are occupied .
if particle creation operators @xmath35 and @xmath36 ( obeying bose commutation relations ) for the two states act on the vacuum to put particles into these two states , then we define the ( properly normalized ) state as @xmath37 where we define the quantities @xmath38 as complex and separate them into magnitudes @xmath39 and phases @xmath40 according to the notation @xmath41 also @xmath42 we can easily compute the average number of particles @xmath43 in this state@xmath44 use the first form of eq .
( [ expform ] ) to give @xmath45 where we have taken @xmath46 thus @xmath47 similarly we get @xmath48 so that @xmath49 and @xmath50 the fact that @xmath51 is known in our phase state does not affect the results for interference patterns , which depend just on relative phase . indeed such states have been used on many occasions to discuss the interference of two condensates.@xcite@xmath7@xcite@xmath7@xcite@xmath7@xcite our state can be used to discuss the how relative phase can be conjugate to particle number . write it in a form that makes the phases explicit : @xmath52 now change the phases to relative phase @xmath53 and total phase @xmath54 and then take the derivative with respect to @xmath55@xmath56 the phase derivative operator gives the same result as the number difference operator so that @xmath16 and @xmath57 are conjugate variables.@xcite@xmath7@xcite note that the total phase appears only as an external factor @xmath58 $ ] and so has no physical significance . thus the individual phases have no physical significance ; only the relative phase @xmath16 is a meaningful quantity . in order to emphasize this last point and put the phase state in a more compact form to treat interference we rename it and rewrite it as @xmath59 where now @xmath60 and @xmath16 is the relative phase . also now @xmath61 we have simply dropped a meaningless factor of unit magnitude . since we have just two occupied states , the terms in eq .
( [ field exp ] ) not referring to states @xmath62 and @xmath63 never contribute and we can more simply write @xmath64 we can also make this more compact by writing @xmath65 with @xmath66 the quantity extracted out @xmath67 can again be dropped as a meaningless unit - magnitude factor .
we want to consider how @xmath68 acts on the phase state .
we have @xmath69 by changing variables in the first state to @xmath70 we can put both terms in the same form so @xmath71 where @xmath72 we can now easily compute the condensate wave function as @xmath73 if for the moment we put back the previously neglected leading phase factors we have @xmath74 which has just the same form as eq .
( [ simplecase ] ) , with average densities in place of precise densities and well - defined individual phases .
however the individual phases are not measurable only the relative phase is .
indeed the average density follows immediately as @xmath75 just like eq .
( [ simpint ] ) where now @xmath76 and @xmath77 have obvious definitions in terms of averages .
thus a phase state provides a rigorous background for discussion of condensate wave functions and for the simplified form of treating interference between the two condensates .
how we might actually prepare one _ before _ the experiment is a separate difficult question , which we treat below
. we will find it useful and necessary in the next section to consider more general cases in which we make measurements of many particle positions essentially simultaneously .
this allows interference fringes to emerge where they would otherwise not occur . for our phase state
such calculations are straightforward and add no additional information since the the multiparticle densities all factor in the phase states .
for example , consider the expectation value of @xmath78 of course , @xmath79 and @xmath80 do nt commute , but in the approximation shown we are dropping a term of order @xmath51 compared to one of order @xmath81 the last form is more convenient to use . by the eigenfunction behavior of the phase state we easily get @xmath82
this result generalizes to @xmath83 considering a state in the form @xmath84 is useful for interpreting an experiment . here
we can consider our experiment as detecting particle 1 and then 2 shortly thereafter , and so on .
after @xmath85 detections the wave function evolves to a state missing several particles .
what is the nature of the state to which it has evolved ? in the case of a phase state it is @xmath86 however , it is more interesting to consider the case of @xmath87 a fock state as we do in the next section .
it is not evident that experimentalists can prepare a phase state as described above .
it would seem more likely that they are working with fock states , that is , states in which the particles numbers , @xmath88 and @xmath89 in the two condensates are known rather well .
it seems likely , in any case , that one has more chance of initially preparing such a state .
however , as several workers@xcite@xmath0@xcite have realized in recent years , and as we will show , that an interference pattern with some phase can still arise in a fock state . denoting the state sharp in particle number as @xmath90 and using the bose annihilation relation
@xmath91 the one - body density in a fock state is @xmath92 and of course there is no interference . phase and particle number are conjugates , and the particle number is initially known . however ,
if we consider measuring the position of two particles simultaneously , some correlation should arise .
we have @xmath93 \label{nlesscert}\end{aligned}\ ] ] so that the particle number is now slightly less certain . the two - body fock correlation function is @xmath94 \nonumber \\ & = & n^{2}\left [ 1+x\cos \mathbf{k\cdot } ( \mathbf{r}_{1}-\mathbf{r}% _ { 2})\right ] \label{twopart}\end{aligned}\ ] ] for two particles there is indeed a position correlation .
as we make more and more measurements the state gets more and more mixed among states with various numbers of particles .
we can rewrite eq .
( [ twopart ] ) in a somewhat different and useful way .
a simple integration shows that @xmath95 .\ ] ] compare this with eq .
( [ 2bodyphase ] ) .
we have a similar looking result except that we now integrate over all relative phases .
remarkably this result can be extended to higher order correlation functions .
indeed ref .
shows that @xmath96 \label{mfock}\ ] ] where it is assumed that @xmath97 . in order to derive this result ,
we invert eq .
( [ phasestate ] ) .
multiply both sides by @xmath98 and integrate over @xmath16 to give @xmath99 thus , by eq .
( [ eigen ] ) , we find @xmath100 we will return to analyze this interesting state below .
first consider the @xmath101-body fock correlation function : @xmath102 phase states are not actually orthogonal , but for large @xmath51 they are essentially so as we show in appendix a , so , if @xmath103 we can write @xmath104 and @xmath105 just as we claimed in eq .
( [ mfock ] ) .
we will show how to get this result by direct calculation , as done in ref .
, in appendix b. this result eq .
( [ mfock ] ) has the form of eq .
( [ mphase ] ) but with the unknown phase integrated over . that makes sense in that our initial fock state did not have a phase defined and can ,
indeed , itself be expressed as a sum over phase states as in eq .
( [ phasesum ] ) .
suppose we have started with a fock state and have made @xmath106 particle measurements and have found particles at positions @xmath107 and are about to measure the @xmath85th . then the probability of finding the @xmath85th particle at position @xmath108 is @xmath109 \label{mthprob}\ ] ] where @xmath110 \label{gphi}\ ] ] as we show by direct simulation , @xmath111 develops a sharp peak at some _ a priori _ unpredictable phase .
if one makes measurements from the first to the @xmath85th particle by this prescription , the peak becomes narrower as one proceeds .
of course , as more measurements are made , the particle number in each condensate becomes less certain ( as , for example , in eq .
( [ nlesscert ] ) ) , so phase can be more sharply defined .
now look back at eq .
( [ ftops ] ) .
after a fair number @xmath85 of measurements , the real part of @xmath112 peaks up sharply at some @xmath16 value , call in @xmath113 .
that means that the measurements have converted the fock wave function into a narrow sum of phase states around @xmath114 the more measurements that are made , the better is the definition of the phase state .
measurements in a fock state provide a way to prepare a phase state .
one can understand the mit experiments@xcite in this way .
the starting state was prepared as two separate condensates , whose particle numbers could have been known ; many subsequent particle measurements sharpened the phase to some random value and the final overall observation showed that phase .
we use eqs . ( [ mthprob ] ) and ( [ gphi ] ) .
one choses the initial position @xmath115 randomly , and then the next particle is chosen from the probability distribution @xmath116 and so on .
we will find that , if @xmath101 is large enough , @xmath117 peaks up at some _ a priori _ unpredictable phase angle , @xmath113 which may fluctuate somewhat as @xmath85 changes but gradually settles down .
starting a new experiment from the fock state will lead to a randomly different phase angle .
we will consider only the case in which the initial fock state has @xmath118 i.e. , @xmath119 it is convenient to fourier expand @xmath120 that is , we write @xmath121\ ] ] in the integration of eq .
( [ mthprob ] ) only @xmath122 @xmath123 and @xmath124 will contribute and doing the integrals gives @xmath125 if we define @xmath126 , @xmath127 and @xmath128 we can write @xmath129=k\left ( 1+a_{m}\cos ( \mathbf{k\cdot r}_{m}+\phi _ { m})\right ) .
\label{prob2}\ ] ] where @xmath130 is a normalization factor , and @xmath131 gives the the value of an angle in the @xmath85th experiment . since @xmath132 is a probability , we must have @xmath133 so that @xmath132 is always positive .
since @xmath134 has that property we can write @xmath135 where @xmath136 behaves like a polar angle
. then the emerging phase actually has a space angle designation ( @xmath137 we will find numerically that @xmath138 rapidly as we make measurements . in that case
the probability of eq .
( [ prob2 ] ) looks just like the density prediction of eq .
( [ simpint ] ) and moreover since @xmath139 is a narrow function peaking at @xmath140 the phase defined by the fourier coefficients is the same as that defined by the peak of @xmath141 as seen using eq .
( [ mthprob ] ) .
we work in just one dimension . to choose from the probability of eq .
( [ prob ] ) , we need to find the cumulative probability @xmath142 and then solve the equation @xmath143 for @xmath144 where @xmath145 is a random number uniformly distributed on @xmath146 to @xmath147 we take the box size @xmath148 and choose a @xmath149 value such that @xmath150 is an integer times 2@xmath151 to provide periodic boundary conditions .
the normalization of the probability in eq.([prob2 ] ) is just the factor @xmath152 figs .
1 and 2 are plots of @xmath153 and @xmath134 versus iteration number in a particular run of 200 interations ; each of these is found from the amplitudes @xmath122 @xmath154 @xmath124 at each iteration .
there is no reason why @xmath134 should be unity from the outset .
however , @xmath134 does always proceed to unity after a small number of iterations .
the result is that @xmath155 approaches some sharply - defined random phase angle as predicted .
of course , for small @xmath85 the fluctuations are relatively large and settle down only after many measurements .
this corresponds to an initially wide distribution , @xmath156 , which , however , progressively narrows as more information is gathered .
3 plots the final angular distribution @xmath157 ; it is indeed sharply peaked at the same value found from the iteration limit .
4 plots the final probability distribution , eq .
( [ prob2 ] ) , versus postion @xmath158 and also shows a histogram of the positions found in the 200 iterations on the same scale .
one sees that these positions really do fall in the given distribution with the expected oscillations and with the same phase as gotten in the two other ways .
we have shown how one might rigorously treat the interference of two bose condensates .
the usual assumption of two condensates with individually known phases gets involved with the thorny questions about whether one can usefully define the phase of a single condensate . however , using such states does lead directly to the usual fomulas for interference patterns based on very simple assumptions .
this procedure remains unsatisfying since it may not be very obvious to the experimenter how to prepare such phase states before looking at the interference .
experimentally it seems to make no difference , since without special preparation experiments , even with fock states , we have seen how an interference pattern arises using bose condensates . the discussion of sec .
iv shows explicitly why such preparation was not necessary .
even if one starts with a state where the particle number in each cloud is precisely known and many particles are involved in the measurement , one finds a perfect interference pattern emerging , with a well - defined relative phase . starting from a state having a definite number of particles , the experiment will end up with a state having a quite definite value of the relative phase .
thus this procedure actually provides a method of preparing the phase state discussed in sec .
starting from a fock state , make , say , two hundred position measurements to get to a narrow @xmath159 the phase of the wave function of the remaining state of @xmath160 total particles will be well - defined .
the final result is likely a phase state with a known _ total _ number of particles such as that discussed at eq .
( [ phasestate ] ) , but an unknown number in each condensate . in theoretical superfluid calculations
it is simpler to treat the problem with definite phases than to use a fock state .
however , the actual existence of such a broken symmetry state is subject to question.@xcite@xmath0@xcite in a ferromagnet one actually has broken the symmetry of the various directions of magnetization from the presence of some small external field .
the existence of a field that would make @xmath161 non - zero is not so clear , since such states do not conserve particle number .
the existence of a well - defined relative phase established by measuring particle positions can be established without worrying about broken symmetry .
if a person feels uncomfortable with the idea of a phase _ emerging _ from the series of measurements done on particle position , then he or she might assume , with no change in theoretical prediction , that that phase pre - existed within the clouds of particles .
that is , the individual condensates had some relative phase , before they met , as a so - called `` hidden variable '' and the experiments simply bring out this previously hidden phase . of course
, in the next realization of the experiment , starting again from a fock state , the phase will surely emerge with a randomly different value , in accordance with conventional quantum mechanics , which expresses the fock state as a sum over all phase states as given in eq .
( [ phasesum ] ) . consider the following wave function to describe a single momentum state @xmath162 which is made up of a mixture of states of known particle number @xmath163 @xmath164 this function is properly normalized .
the parameter @xmath165 is complex and we separate it into magnitude @xmath166 and phase @xmath167 according to the notation @xmath168 we can easily compute the average number of particles @xmath169 in this state . let @xmath170 be the destruction operator for particles in state @xmath171 then @xmath172 and @xmath173 .
the state @xmath174 has the nice property that it is an eigenstate of the lowering operator @xmath175@xmath176 thus @xmath170 has a non - zero expectation value in this state : @xmath177 clearly the states @xmath174 provide a definite phase .
of course they are then not eigenstates of the number operator .
next construct a multi - level many - body state with many @xmath149 values possible .
this takes the form @xmath178@xmath179
\left| n_{k_{1}},n_{k_{2}},n_{k_{3}}\cdots \right\rangle\ ] ] where @xmath180 means sum over all possible numbers of particles @xmath181 in all the @xmath149-states . where @xmath184 is the volume of the system .
we have @xmath185 if one of the @xmath149-states is macroscopically occupied , say , the momentum state @xmath186 then we can write @xmath187 where @xmath188 and @xmath189 is the total contribution of the non - condensed states . the leading term @xmath190 represents a condensate wave function having a definite phase @xmath29 but non - sharp number of particles .
consider the case of the interference of a double condensate in momentum states @xmath62 and @xmath191 with coherent states having only these two momentum states occupied we can write @xmath192 where the averages @xmath193 and @xmath194 are both macroscopic quantities .
we manipulate the sums slightly in terms of particle creation operators @xmath35 and @xmath36 for the two states .
if @xmath195 then we have @xmath196 we calculate the inner product of two phase states to show that they are nearly orthogonal for large particle number . from eq .
( [ phasestate ] ) we find @xmath198 ^{n - n}=\frac{1}{g^{n}}\left [ 1+\gamma ^{2}e^{i\left ( \phi -\phi ^{\prime } \right ) } \right ] ^{n}\ ] ] this is a very sharply peaked function of @xmath199 to see this , taylor expand the logarithm of this in powers of @xmath200 and then exponentiate the result keeping only terms to @xmath201 .
the result is @xmath202 \exp \left [ -i\frac{n}{(1+\gamma ^{2})}\left ( \phi ^{\prime } -\phi \right ) \right]\ ] ] in the limit of very large @xmath203 this is proportional to a delta function of the @xmath204 as we assumed in the discussion of sec .
iv . we want to derive the general expression of eq .
( [ mfock ] ) for the correlation function @xmath205 consider this quantity in its original form for a fock state : @xmath206 because this is a diagonal matrix element in fock space , each time an @xmath207 occurs , there must be a matching @xmath208 similarly for the @xmath209 operators .
we are assuming @xmath210 or @xmath33 so that we can always write @xmath211 @xmath212 etc .
thus each @xmath213 gives @xmath214 and each @xmath215 an @xmath33 .
consider a particular combination product : @xmath216 with the restriction that every time an @xmath217-type term occurs there _ must _ be a corresponding @xmath218-type term somewhere in the overall product to give the proper balance of creation and destruction operators .
thus one gets a series of terms of the form @xmath219 where @xmath220 and the sum of all the @xmath221 vanishes .
that is , we have @xmath222 where @xmath223 means sum on all @xmath224 with the restriction that @xmath225 the restriction on the @xmath226 values can be lifted if we insert the integral @xmath227 which allows us to write @xmath228 \nonumber \\ & = & n^{m}\int_{0}^{2\pi } \frac{d\phi } { 2\pi } \prod_{i=1}^{m}\left [ 1+x\cos \left ( \mathbf{k}\cdot \mathbf{r}_{i}+\phi \right ) \right]\end{aligned}\ ] ] as we wished to prove .
markus greiner , immanuel bloch , olaf mandel , theodor w. hnsch , and tilman esslinger , `` exploring phase coherence in a 2d lattice of bose - einstein condensates , '' phys .
lett . * 87 * , 160405 - 14 ( 2001 ) .
mark h.wheeler , kevin m. mertes , jessie d. erwin , and david s. hall , `` spontaneous macroscopic spin polarization in independent spinor bose - einstein condensates , '' phys .
* 93 * , 170402 - 14 ( 2004 ) .
m. naraschewski , h. wallis , and a. schenzle , j. i. cirac , p. zoller , `` interference of bose condensates , '' phys . rev . * a 54 * , 21852196 ( 1996 )
; h. wallis , a. rhrl , m. naraschewski , and a. schenzle , `` phase - space dynamics of bose condensates : interference versus interaction , '' phys . rev . * a 55 * , 2109 - 2119 ( 1997 ) .
for example , stephen howard and sanat k. roy , `` coherent states of a harmonic oscillator , '' am .
55 , 1109 - 1117 ( 1987 ) ; ivan h. deutsch , `` a basis - independent approach to quantum optics , '' am .
59 , 834 - 839 ( 1991 ) . | we consider the interference of two overlapping ideal bose - einstein condensates .
the usual description of this phenomenon involves the introduction of a so - called condensate wave functions having a definite phase .
we investigate the origin of this phase and the theoretical basis of treating interference .
it is possible to construct a phase state , for which the particle number is uncertain , but phase is known .
however , how one would prepare such a state before an experiment is not obvious .
we show that a phase can also arise from experiments using condensates in fock states , that is , having known particle numbers .
analysis of measurements in such states also gives us a prescription for preparing phase states .
the connection of this procedure to questions of `` spontaneously broken gauge symmetry '' and to `` hidden variables '' is mentioned . |
thin film photo - voltaic devices made of inorganic semiconductors such as tin sulfide ( sns ) and cadmium telluride ( cdte ) etc . have gained attention in recent years due to their lower cost compared to the made from silicon @xcite .
sns is consider a better option considering they are not toxic like cadmium compounds @xcite .
sns is a semiconducting material that has a band - gap ranging between 1.1 - 2.2 ev @xcite and refractive index in the range of 1.6 - 2.2 ev , depending on the fabrication techniques @xcite .
however , the major limitation of inorganic semiconductors are their light trapping ability thereby which their efficiency as solar cells are lowered ( compared to silicon ) @xcite . in order , to increase the light trapping ability of these semiconductors
, researchers are adopting a new methods using noble metal nano - particles dispersed in the semiconducting film @xcite .
noble metal nano - particles like silver ( ag ) , gold ( au ) , aluminum ( al ) and copper ( cu ) embedded in a dielectric / semiconducting medium exhibit interesting optical properties and have gained a lot of attention due to their potential applications in plasmonic solar cells @xcite , photonic devices @xcite , surface plasmon enhanced sensors @xcite .
localized surface plasmon resonance ( lspr ) occurs when the electromagnetic wave travels along the metal and dielectric/ semiconductor interface @xcite .
lspr peaks and their positions are influenced by the size , the shape of the metal nano - particle @xcite , its dielectric properties and those of the local surrounding medium * ? ? ? * .
for nano - particles of ag , au and copper the lspr peak occurs in the uv - visible region of the spectrum depending on the density of the electrons in the metal nano - particle@xcite .
literature consists of many articles on metal nano - particles embedded in different dielectric / semiconducting mediums like @xmath2 @xcite , si @xcite and zno @xcite etc . out of all the noble metals , silver nano - particles showed the best spr response @xcite , making it very useful for applications .
we hence selected ag to study the surface plasmon resonance in tin sulfide ( sns ) .
very few research groups have carried out work on ag - sns films @xcite . in this manuscript ,
we report the preparation of sns : ag nano - composite films of varying thicknesses and investigate the size - dependent optical properties .
composite thin films of tin sulfide ( sns ) and silver ( ag ) were grown by thermal evaporation with vacuum better than @xmath3 torr .
the pellets used as starting material were made by mixing sns powder and ag nano - powder ( average grain size 20 nm ) .
the mass ratio of sns and ag powder used was 2:1 .
the 99.99% purity sns powder was provided by himedia ( mumbai ) and the ag nano - powder was obtained from nanoshel ( usa ) .
the thickness of the films were measured using veeco s dektak surface profiler ( 150 ) .
the structural properties of the films were studied using a bruker d8 diffractometer at an operating voltage of 40 kv in the @xmath4 mode with cu target giving x - ray at @xmath5 and technai g2 t30u twin transmission electron microscope ( tem ) at an accelerating voltage of 200 kv .
the chemical composition analysis of the films were examined using kratos axis ultra dld s x - ray photo - electron spectroscopy ( xps ) with al @xmath6 target .
the beam energy used for the analysis was 1490 ev and binding energy was referenced with respect to carbon .
the surface morphology was analyzed by a field emission scanning electron microscopy ( fe - sem fei - quanta 200f ) .
the optical absorption and transmittance spectra were recorded using a systronics uv - vis double beam spectrophotometer ( 2202 ) in the 300 - 900 nm wavelength range . [ uv ]
figure 1 shows the optical absorption spectra of the sns / ag composite thin films . unlike the absorption spectra of sns films @xcite , these films show two peaks at @xmath7 and @xmath1 in all the films without exception .
considering the starting silver powder had particle size of 20 nm , we believe these peaks are localized surface plasmon resonance ( lspr ) peaks resulting from interaction of light with electrons at the surface of the silver nano - particles @xcite . considering that the lspr peak position depends on the surrounding media s refractive index along with the shape and size of the metal cluster , the existence of two well resolved peaks in fig 1 may either be due to non - spherical metallic clusters or because they experience two different surrounding media ( possibly air and sns ) .
another interesting feature of fig 1 is that both peaks show red - shift , however it is more pronounced in the @xmath1 peak .
[ xps ] [ xrd ] -0.5 cm before proceeding , we would have to resolve whether silver nano - particles are uniformly distributed throughout the thickness of the film or not .
we have used depth profiling using x - ray photo - electron spectroscopy ( xps ) .
fig 2 shows the xps scan for ag ( ag 3d peak ) @xcite at the film s surface , at 150 and 300 nm below the surface of a 600 nm thick sns : ag nano - composite film . comparing the scans , we find that the xps peaks of silver completely vanishes below 150 nm from the surface of the film .
this clearly implies that the silver nano - particles are only present on and adjacent to the surface of the films .
the deposition of ag on the surface or just below is due to the higher melting point of ag ( 1234 k ) compared to ( 1155 k ) that of sns @xcite .
varying the rate of evaporation does give an uniform distribution of silver along the thickness of the sns film .
however , since this study is a part of a continuous evaluation of sns and their nano - composite films , we have maintained the film fabrication parameters as in our earlier studies @xcite .
the existence of ag at the surface ( in contact with air - sns interface ) and within the sns surface presents a possibility of two lspr peaks in the uv - visible absorption spectra which we shall resolve in our theoretical analysis .
[ xps ] [ saed ] as stated above , the 580 nm absorption peak of fig 1 shows a strong red - shift which increases with increasing film thickness .
an interesting aspect about lspr peaks is that its position strongly depends on the refractive index of the surrounding media .
an increasing refractive index of the surrounding dielectric results in a red - shift @xcite .
hence , clearly , we expect that the sns film s refractive index increases with increasing film thickness . to investigate the possible cause of increase in sns refractive index with film thickness , we investigate the structural properties of samples .
fig 3 shows the diffraction patterns of various thickness sns : ag nano - composite films . without exception
all the films showed polycrystallinity with sns having orthorhombic structure with lattice parameters matching those given in astm card no 83 - 1758 while ag matched those given in astm card no 03 - 0931 .
the grain size of sns and ag nano - particles were evaluated using the scherrers formula @xcite .
we shall be commenting on the grain size of silver nano - particles below , however the sns grains showed a linear increase in grain size with increasing film thickness . in earlier studies @xcite we conducted on thin sns films ( without ag )
grown under identical conditions , we had reported the optical properties of sns films as a function of grain size .
we had reported that the sns film s refractive index increased with increasing grain size , which in turn increased with film thickness ( fig 6 of ref @xcite ) .
we find that the sns grain size ( 12 - 19 nm ) and its variation with film thickness in these nano - composite films matched ( within experimental error ) with previous studies .
as stated initially , we believe that the increasing refractive index with increasing thickness of sns background contributes to the observed red - shift in the 580 nm lspr peak of fig 1 .
as explained above , we use experimentally evaluated refractive index of sns films as a function of film thickness and plot position of the @xmath1 spr peak ( @xmath8 ) with respect to it ( fig 4 ) .
notice that the variation is similar to that shown in fig 1 .
the experimentally evaluated values of refractive index would hence be utilized as input parameters for the gans model calculations we discuss in the next section . also , we notice only a small variation in refractive index with sns grain size for @xmath9 explaining the insignificant or minor red - shift seen in the @xmath0 peak .
[ sem ] [ sem ] other then the xps results , the lone xrd peak of ag also confirms the existence of metallic clusters of silver in the nano - composite films .
further proof of metallic ag clusters in our samples come from the electron diffraction rings ( selected area electron diffraction , saed image ) obtained using the transmission election microscope .
saed pattern ( fig 5 ) shows two sharp rings corresponding to two prominent peaks from ag .
the two ag rings marked ( i ) and ( ii ) in fig 5 correspond to the ( 200 ) and ( 220 ) planes .
the remaining three rings of fig 5 , marked ( iii ) , ( iv ) and ( v ) , correspond to the ( 040 ) , ( 131 ) and ( 200 ) diffraction planes of sns .
an investigation of the diffraction patterns at @xmath10 and @xmath11 , of fig 3 , shows them to be resolvable peaks of sns and ag ( fig 6a ) .
the peaks were de - convoluted and the average grain size of silver nano - particles were evaluated using the scherrer s formula .
fig 6b shows the variation in the silver nano - particle grain size with film thickness .
notice that the average size of the ag nano - particle is around 12 nm for films with thickness less than 600 nm while they are in the range of 32 - 42 nm for higher thickness films . as the size of the metal cluster increases
there is an increase in radiative damping that leads to greater scattering cross - sections and hence larger extinction @xcite .
fig 7 exhibits a sem micrograph of a 650 nm thick sns : ag nano - composite film .
as viewed from the surface , micrograph shows that the film consists of randomly spaced cuboid silver nano - particles @xcite dispersed among nearly spherical sns nano - particles of sns .
the chemical composition of the grains were confirmed using the edax ( energy dispersive analysis of x - ray ) attachment with the scanning electron microscope .
the extinction spectra of fig 1 can be theoretically generated using various computational models , like mie model @xcite , gans model @xcite and discrete dipole approximation ( dda ) model @xcite .
mie theory was developed to study scattering due to spherical particle @xcite .
the model was successfully extended for studying localized surface plasmon resonance due to spherical metal nano - particles dispersed in semiconducting or insulating media .
however , the model needed corrections to include non - spherical shaped metal clusters .
the modifications proposed by gans @xcite . considering that our sem morphology images suggest our silver metal clusters to have near square cross - section on the surface ( @xmath12 ) and xps results suggest the grains to occur only on the film surface
, we expect @xmath13 i.e. the silver nano - particles are oblate shaped with size less than 50 nm .
we have , hence used gans model to theoretically analyze our experimental results .
gans model gives the extinction cross - section/ coefficient which is the sum of light intensity scattered and absorbed by a metal nano - particle whose shape is either oblate or prolate .
the scattering and absorption cross - sections are calculated using @xcite @xmath14 and @xmath15 where , @xmath16\label{first}\end{aligned}\ ] ] where , ` l ' is the depolarization factor that depends on the shape of the scattering/ absorbing metal nano - particle , ` v ' is the volume of the nano - particle , ` @xmath17 ' is the complex dielectric constant of the metal and ` @xmath18 ' is that of the surrounding medium . for non - sphere clusters ,
the depolarization factor ` l ' takes different values along the three axes .
the values completely depend on the shape of the metal nano - particle which is mathematically defined by the aspect ratio ( ratio of minor to major axis of the spheroid ) . for an oblate nano - particle ( @xmath19 ) it is given as @xmath20\nonumber\\ l_x & = & l_y = \left({1-l_z \over 2}\right)\nonumber\end{aligned}\ ] ] where ` e ' is related to the aspect ratio @xmath21 eqn [ first ] is now written as @xmath22\label{sumed}\end{aligned}\ ] ] when an e - m radiation is incident on a spheroid nano - cluster , electrons are set into oscillations along and perpendicular to its symmetry axis , resulting in two plasmon resonance peaks .
each peak is associated with the oscillation direction with respect to the symmetry axis and are named as the longitudinal mode ( lm ) and the transverse mode ( tm ) @xcite . in the case of oblate nano - clusters
the tm excitations appear at higher wavelengths compared to that of the lm excitations @xcite .
expectedly , the separation between the two modes depend on how much deviation the spheroid has from an ideal sphere , or in other words depends on the cluster s aspect ratio and the refractive index of the surrounding medium @xcite . [ scattergs ] our calculations for extinction coefficients were done using the complex dielectric constant for silver listed as a function of wavelength by palik et al @xcite .
we used the data from fig 6 for estimating the grain size of silver . before calculating case of silver dispersed in sns ,
simulation was done for silver in air .
our calculations agreed with literature @xcite and found lspr peak appear at 350 nm .
thus , we can confidently rule out silver - air interface playing a role in this study .
we hence proceed reporting effect of ag - sns interface in this study .
the results of priyal et al @xcite were used for the sns background s dielectric constant .
the studies showed sns film s refractive index follow a trend similar to the cauchy s dispersion relation @xcite with the wavelength , where the coefficients strongly depend on sns grain size .
fig 8 shows the extinction spectra generated theoretically .
the aspect ratio was linearly varied from 0.64 to 0.60 with increasing film thickness .
the values were selected so that peak positions obtained theoretically matched with the experimentally obtained results .
fig 8a shows select calculated extinction coefficient spectra .
these spectra are of course quite dissimilar from those shown in fig 1 since fig 1 also contains the absorption spectra of the background sns film . to recreate fig 1 we have super - imposed data of fig 8a on absorption spectra of sns films of corresponding thicknesses with same / similar grain size used in our previous studies @xcite .
we may hence conclude that the peaks of fig 1 at @xmath23 and @xmath24 correspond to the longitudinal and transverse modes respectively .
fig 8b gives thus theoretically simulated spectra and we find them in fairly good agreement with the experimentally obtained data of fig 1 .
fig 8b also includes the two trend lines of fig 1 to show how well our simulations agree with experimental data .
the above results clearly establishes the formation of lspr peaks in sns : ag nano - composite films .
it is of immense interest that the well established theory easily quantifies the experimental observations .
as stated , sns films have attracted interest in recent years for its application as photovoltaics @xcite . while their properties of being non - toxic and films with good absorption and high band - gap make them suitable candidates , attempts of fabricating photovoltaics show that they do not match the efficiency of silicon @xcite .
a basic issue with thin film solar cell devices is that the absorbing layer being thin does not absorb sufficient amount of light for charge carrier generation and hence leads to low efficiency .
hence , research has been directed towards plasmonic solar cells of sns .
one expects that the metal nano - particles in the medium would at resonant frequency either absorb the incident light or scatter it within the medium @xcite .
if scattering can be made the dominant extinction mechanism , the efficiency of the device can be enhanced @xcite . using the above made calculations ( namely using eqn [ second ] and [ third ] ) , we have calculated the scattering efficiency ( @xmath25 ) using @xmath26 [ scattergs2 ] fig 9a shows scattering efficiency of the lm and tm peaks .
the scattering efficiencies vary as a ( quadratic ) function of the grain size and are marked different for the lm and tm modes .
the scattering contribution is more in the lm peak as compared to that of the tm peak .
fig 9a is misleading since grain size only contributes to extinction s intensity via the volume term of eqn [ first ] .
size indirectly influences the scattering efficiency through the real component of silver s dielectric constant ( @xmath17 ) and sns background s dielectric constant ( @xmath18 ) .
both these factors vary with respective grain size ( silver and sns ) which are in turn dependent on the film thickness . as the film thickness increase , the grain size of silver and sns increases .
this results in increased difference in silver s and sns s dielectric constant , resulting in stronger scattering from the metal nano - particles ( also understood from eqn [ sumed ] ) .
this is possibly due to the fact that the effective size of the metal nano - particle increases as the refractive index of sns increases leading to larger scattering .
the result is promising considering that we now can predict and material manipulate plasmonic layer to obtain high efficiency solar devices .
another aspect of interest for device fabrication would be to bring the two resolved peaks close by .
this would result in an absorption band leading to a range of wavelengths being captured and scattered within the medium encourage easy electron - hole pair generation .
it is the non - spherical shape of the metal nano - clusters ( aspect ratio @xmath27 ) that gives rise to the two peaks .
this is best understood by investigating the peak positions ( @xmath8 ) as a function of aspect ratio ( fig 9b ) .
both , the lm and tm peaks show blue shifts with increasing aspect ratio ( c / a ) , this trend is in agreement to those reported in literature @xcite .
the dotted lines indicate the trends and suggest that they would tend to converge around 0.75 , for which we would have the tm and lm peaks close by resulting in a board absorption band . in conclusion , we can relate all of the above results . for an efficient photo - layer of a solar device , we ideally require sns : ag nano - composite films with 40 nm sized silver metal nano - clusters which are oblate shaped with aspect ratio of 0.75 and large sized sns grains .
this ensures a large difference in sns and silver refractive indices that lead to enhanced scattering of light within the films and the aspect ratio ensure unresolved tm and lm peaks giving broad absorption band .
nano - composite films of sns and ag were grown on glass substrates at room temperature .
the films showed two prominent peaks at @xmath28 and @xmath1 in their uv - visible spectra .
these peaks showed a red - shift as the film thickness increased .
structural studies reveal increase in ag and sns nano - particle size with increasing film thickness .
we were able to generate the observed extinction spectra using gans model .
the results of our calculations suggest that the @xmath0 and @xmath1 corresponds to the longitudinal and transverse mode lspr peaks from oblate silver nano - particles .
the theoretical and experimental results also explains the observed red shift in the spr peak at @xmath1 due to the increasing refractive index of the background sns .
theoretical results also suggests that as the aspect ratio of ag nano - particles decreases with increasing film thickness , the two resonant peaks move away from each other .
the calculations also explain the increase in scattering efficiency with increasing ag nano - particle size as resulting from enhanced difference between the metal and its background s dielectric constant .
both theoretical and experimental results suggest that one can tailor the spr peak positions and scattering efficiency of such nano - composites .
sns : ag nano - composite films hence present itself as an efficient photo - layer in solar voltaics . based on our results we suggest sns :
ag nano - composite films with oblate shaped ( aspect ratio @xmath29 ) ag nano - particles of size around 40 nm which embedded in large grain sized sns background would prove ideal for plasmonic devices .
authors are thankful to the university grants commission ( ugc , india ) for funding this work under their major research project scheme ( f.no .
39 - 531/2010sr ) . 99 ristov m. , sinadinovski g. , mitreski m. , ristova m. , photovoltaic cells based on chemically deposited p - type sns , sol . energy mater
cells , 69 ( 2001 ) 17 .
oladeji i.o , chow l. , viswanathan v. , zhao z. , metal / cdte / cds / cd1xznxs / tco / glass : a new cdte thin film solar cell structure , sol . energy mater .
cells , 61 ( 2000 ) 203 .
frantzis l. , jones e. , little a.d .
, lee c. , wood m. and wormser p. , 16th eur . photovoltaic sol . energy conf .
( proceedings ) , glasgow , scotland , ( 2000 ) 2103 .
reddy k.t.r . ,
reddy n.k .
, miles r.w .
, photovoltaic properties of sns based solar cells .
sol . energy mater .
cells , 90 ( 2006 ) 3041 .
gao c , shen h , sun l. preparation and properties of zinc blende and orthorhombic sns films by chemical bath deposition , appl .
, 257 ( 2011 ) 6750 .
sohila s. , rajalakshmi m. , ghosh c. , arora a.k . and muthamizhchelvan c. , optical and raman scattering studies on sns nanoparticles , j. alloys compd . 509 ( 2011 ) 5843 .
shaaban e.r . ,
el - sadek m.s.a , el - hagary m. and yahia i.s .
, spectroscopic ellipsometry investigations of the optical constants of nanocrystalline sns thin films , phys .
scripta 86 ( 2012 ) 015702 .
jakhar a. , jamdagni a. , bakshi a. , verma t. , shukla v. , jain p. , sinha n. and arun p. , refractive index of sns thin nano - crystalline films , solid state commun . , 168 ( 2013 ) 31 .
catchpole k.r . and polman a. , plasmonic solar cells , optic .
express , 16 ( 2008 ) 21793 .
catchpole k. r. and polman a. , design principles for particle plasmon enhanced solar cells , appl .
93 ( 2008 ) 191113 .
sundararajan s.p . ,
grady n.k . , mirin n. and halas n.j . , nanoparticle - induced enhancement and suppression of photocurrent in a silicon photodiode , nano lett . ,
8 ( 2008 ) 624 .
lee k.s . and el - sayed m.a . , gold and silver nanoparticles in sensing and imaging : sensitivity of plasmon response to size , shape , and metal composition , j. phys .
b , 110 ( 2006 ) 19220 .
pillai s. , catchpole k.r .
, trupke t. , and green m.a .
, surface plasmon enhanced silicon solar cells , j. appl .
( 2207 ) 093105 .
kelly k.l . , coronado e.,l zhao l.l . and
schatz g.c . , the optical properties of metal nanoparticles : the influence of size , shape and dielectric enviornment , j. phys . chem .
b , 107 ( 2003 ) 668 .
link s. and el - sayed m.a . ,
size and temperature dependence of the plasmon absorption of colloidal gold nanoparticles , j. phys .
b , 103 ( 1999 ) 4212 .
kiran p.p . ,
bhakth b.n.s .
, rao d.n . and de g. , nonlinear optical properties and surface - plasmon enhanced optical limiting in ag
cu nanoclusters co - doped in sio2 sol - gel films , j. appl .
phys . , 96 ( 2004 ) 6717 .
islam m.a . and khan i. , zno / ag composite nanoparticles for surface plasmon resonance based sensor application in uv - vis region , international journak of advancements in research and technology , 1 ( 2012 ) 1 .
pustovit , vitaliy n , shahbazyan , tigran v , finite - size effects in surface - enhanced raman scattering in noble - metal nanoparticles : a semiclassical approach , j. opt .
a , 23 ( 2006 ) 1369 .
jia h.j . ,
cheng s.y .
, wu x.k . and yang y.l . , effect of anneal temperature on electrical and optical properties of sns : ag thin films , natural sci . , 2 ( 2010 ) 197 .
yongli y. and shuying c. , preparation of sns : ag thin films by pulse electrodeposition , j. semicond . , 29 ( 2008 ) 2322 .
lu p. , jia h. and cheng s. , optical and elctrical properties of sns : ag films as solar cell absorbers , adv .
res . , 60 ( 2009 ) 11 .
singh y.p . ,
jain a. and kapoor a. , localized surface plasmons enhanced light transmission into c - silicon solar cells , j. sol . energy ( 2013 ) 584283 - 1 .
awazu k. , fujimaki m. , rockstuhl c. , tominaga j. , murakami h. , ohki y. , yoshida n. and watanabe t. , a plasmonic photocatalyst consisting of silver nanoparticles embedded in titanium dioxide , j. am .
( 2008 ) 1676 .
morales j. , sanchez l. , martn f. , ramos - barrado j.r . and sanchez m. , j. electrochem .
soc . , 151 ( 2004 )
prieto p. , nistor v. , nouneh k. , oyama m. , abd - lefdil m. and daz r. , xps study of silver , nickel and bimetallic silver nickel nanoparticles prepared by seed - mediated growth , appl .
, 258 ( 2012 ) 8807 .
lide d.r . , ed .
, crc handbook of chemistry and physics , internet version 2005 . thangaraju b. and kaliannan p. , spray pyrolytic deposition and characterization of sns and sns2 thin films , j. phys .
33 ( 2000 ) 1054 .
jain p. and arun p. , parameters influencing the optical properties of sns thin films , j. semicond . , 34 ( 2013 ) 093004 - 1 .
jain p. and arun p. , influence of grain size on the band - gap of annealed sns thin films , thin solid films , 548 ( 2013 ) 241 .
takeda y. , plaksin o.a . , wang h. , kono k. umeda n. and kishimoto n. , surface plasmon resonance of au nano - particles fabricated by negative ion implantation and grid structure towards plasmonic applications , optical rev .
13 ( 2006 ) 231 .
cullity b d , stock s r. , elements of x - ray diffraction .
nj : prentice - hall inc , 2001 .
wiley b.j .
, i m s.h . , li z.y .
, mc lellan j. , sickkinen a. and xia y. , maneuvering the surface plasmon resonance of silver nanostructures through shape - controlled synthesis , j. phys .
b , 110 ( 2006 ) 15666 .
fu q. , sun w. , mie theory for light scattering by a spherical particle in an absorbing medium , appl .
optics , 40 ( 2001 ) 1354 .
ranjan m. , predicting plasmonic coupling with mie
gans theory in silver nanoparticle arrays , j. nanopart .
res . , 15 ( 2013 ) 1 .
heilmann a. , quinten m. and werner j. , optical response of thin plasma- polymer films with non - spherical silver nano - particles , the european phys .
j. b - cond
. mat . and complex systems , 3 ( 1998 ) 455 .
draine b.t . and flatau p.j . ,
discrete - dipole approximation for scattering calculations , j. opt .
a. , 11 ( 1994 ) 1491 .
gonzalez a.l .
, reyes - esqueda j.a . and
noguez c. , optical properties of elongated noble metal nanoparticles , j. phys .
c , 112 ( 2008 ) 7356 .
noguez c. , surface plasmons on metal nanoparticles : the influence of shape and physical enviornment , j. phys .
c , 111 ( 2007 ) 3806 .
palik e.d . , handbook of optical constants of solids(academic press , new york , 1985 ) .
elkorashy a.m. , optical constants of tin sulphide single crystals measured by the interference method , phys .
( b ) , 159 ( 1990 ) 903 .
noguchi h. , setiyadi a. , tanamora h. , nagatomo t. , omoto o. , characterization of vacuum - evaporated tin sulfide film for solar cell materials .
energy mater .
cells , 35 ( 1994 ) 325 .
ghosh b. , das m. , banerjee p. , das s. , fabrication of vacuum - evaporated sns / cds heterojunction for pv applications .
energy mater .
cells 92 ( 2008 ) 1099 .
avellaneda d. , nair m.t.s .
, nair p.k . , photovolatic structures using chemically deposited tin sulfide thin films . , thin solid films 517 ( 2009 ) 2500 .
, lee k.s . , el - sayed i.h . and el - sayed m.a . , calculated absorption and scattering properties of gold nanoparticles of different size , shape , and composition : applications in biological imaging and biomedicine , j. phys .
b , 110 ( 2006 ) 7238 . | nano - composite films of tin sulfide ( sns ) and silver ( ag ) fabricated by thermal evaporation showed two prominent peaks in the visible region of their extinction spectra .
theoretical modeling of the extinction spectra suggest that these two peaks ( @xmath0 and @xmath1 ) correspond to the longitudinal mode ( lm ) and transverse mode ( tm ) surface plasmon resonance peaks arising from oblate silver nano - particles . using grain size of silver and sns obtained from structural and morphological characterizations of the samples and dielectric constants as per actuals , we have compared the experimental results with those from theory .
the study shows that silver nano - particles efficiently scatters light and can be used for developing plasmonic based sns solar cells with improved efficiencies .
thin films ; chalcogenides ; optical properties |
many problems in quantum physics can be discussed using a model in which one or more mesoscopic or even macroscopic coordinates @xmath4 , etc .
, interact with a background environment ( one coordinate might also represent an experimental probe , or even an observer ) . in such models ( which have a long history @xcite ) all variables , including the environmental ones , are treated _ quantum - mechanically_. the aim is to find the behaviour of @xmath4 , etc . , after _ averaging _ over the environmental variables in some way .
it is certainly not obvious that one can discuss the real world in this way , given the complexity of @xmath1-body systems .
however we now know that many ( but not all ) mesoscopic or macroscopic systems can be described _ at low energies _ by a few `` canonical models '' , where a simple `` central system '' ( eg . , a 2-level system , or an oscillator ) couples to an environment .
remarkably , there appear to be only two types of environment , describable as baths of either oscillators or spins .
one way of trying to justify such models is the `` renormalisation group '' viewpoint @xcite , which maintains that most physical systems fall into a few `` universality classes '' , each scaling to its own `` fixed point '' in the space of possible hamiltonians .
all systems in a given university class use the same canonical hamiltonian- the differences between different systems lying in the different values of the relevant couplings in this hamiltonian .
although this `` hard rg '' philosophy clearly fails in some cases , it is a useful starting point for the present article , in which the quantum environment is modelled by a `` spin bath '' ( usually of 2-level systems , or `` spin-@xmath5 '' systems ) .
the finite hilbert space of each bath spin makes the spin bath appropriate for describing the low energy dynamics of a set of _ localised _ environmental modes .
we concentrate on one particular `` central spin '' model @xcite in which the central system itself reduces to a 2-level system ; but we also discuss cases where the central system is a one - dimensional coordinate ( a `` particle '' ) moving through a field of spins .
another well - known set of canonical models describes the environment as a set of uncoupled oscillators- these include the `` spin- boson '' model @xcite and the `` caldeira - leggett '' model @xcite .
the spin - boson model couples a central 2-level system to the oscillators , and is thus the analogue of the central spin model ; and the caldeira - leggett model couples a tunneling particle to the oscillators .
these oscillator models all derive from a scheme proposed by feynman and vernon @xcite , to describe a central system coupled _ weakly _ to @xmath1 environmental modes ; as they showed , the mapping to an oscillator bath can only be made rigourously _ if the coupling is weak_. oscillator models are thus best adapted to @xmath1 _ delocalised _ environmental modes ( where the coupling is automatically @xmath6 , and thus small for large @xmath1 ) .
however , readers familiar with low - temperature physics will know that at low energies , the entropy and heat capacity of almost all real physical systems are dominated instead by _ local _ modes such as defects , impurity spins , and nuclear spins @xcite .
typically these relax very slowly at low @xmath0 because little phase space is available in their coupling to any delocalised modes ( or to each other )
. however they often couple strongly to any mesoscopic or macroscopic collective coordinate , which then easily perturbs them .
this coupling is of course _ independent _ of @xmath1 .
unfortunately , even though spin bath models have been studied sporadically for many years @xcite , the results have often been misleading , either because they treat some weak - coupling limit ( sometimes made @xcite by arbitrarily multiplying the coupling to each of the @xmath1 bath spins by a factor @xmath7 , for no good physical reason ) , or because they drop some of the important couplings to the bath spins , in order to solve the model . in the weak - coupling limit
, spin bath models can be mapped to oscillator baths @xcite ( in accordance with the original remarks of feynman and vernon @xcite ) .
however one is often nowhere near the weak - coupling limit , and the mapping to the oscillator bath then _ fails _ in general @xcite .
this demands a new approach , which is the subject of this review .
it may be useful to mention _ why _ many physicists are interested in models of this kind . here
are some of the reasons : \(i ) very rapid progress in work on intrinsically quantum processes ( interference , tunneling , etc . ) occurring at the nanoscopic and mesoscopic scales @xcite , plus speculations about the coming `` nanotechnological revolution '' .
perhaps the most exciting idea in this area is that of making `` quantum computers '' using nanoscopic superconductors @xcite , semiconductors @xcite , or nanomagnets . needless to say , the technological repercussions of this work will be enormous , provided the crucial problem of decoherence can be overcome .
\(ii ) physicists need to understand the mechanisms of decoherence and quantum dissipation @xcite in nature , and the crossover to ( or `` emergence '' of ) classical behaviour from quantum physics as either size , temperature @xmath0 , external fields , or couplings to the environment are increased .
these issues are not only relevant to quantum device design , but also to problems in quantum gravity , and to the infamous `` quantum measurement '' problem @xcite .
the existence of low-@xmath0 canonical models , going beyond the phenomenology of stochastic or master equations @xcite to work with closed hamiltonians , is invaluable here .
recent examples include the analysis of quantum spin glass relaxation @xcite , quantum relaxation in nanomagnets @xcite and tunneling superconductors @xcite ) , and the study of quantum chaotic systems , stochastic resonance @xcite and dissipative tunneling in ac fields @xcite .
earlier such models have been used for decades to discuss relaxation in fields like quantum chemistry @xcite or nuclear physics .
\(iii ) both oscillator and spin bath models map to many important models in quantum field theory .
thus the `` spin - boson '' model mentioned above @xcite maps , for specific parameter values , to the kondo model , the thirring and sine - gordon models , and various other 2-dimensional field theories .
although similar mappings have yet to be exploited in great detail for the spin bath , they will obviously be very useful for , eg . , lattice spin models .
most work in these areas has used oscillator bath representations of the environment , with the tacit assumption that delocalised environmental modes dominate the physics .
however , experiments on quantum nanomagnetic systems @xcite , on glasses @xcite , and on mesoscopic conductors @xcite , as well as theoretical debate about the mechanisms of decoherence in nature , clearly require a more general point of view .
in fact we shall see that spin baths behave very differently from oscillator baths .
for example , the oft - discussed connection between dissipation and decoherence which exists for oscillator bath environments @xcite is absent here- one can even have decoherence with no dissipation at all @xcite , because of the quantum phase associated with the spin bath dynamics .
although the existence of a quantum phase associated with spin is obvious ( it is a quantum variable ) , it was not until haldane @xcite and berry @xcite discussed its topological properties that physicists realised its practical importance , in terms of the path traced out on the spin sphere .
just as in the usual aharonov - bohm effect , the `` flux '' enclosed by a path ( given here by @xmath8 , where @xmath9 is the spin and @xmath10 is the enclosed solid angle on the unit spin sphere ) is equal to a dynamical phase- but now these are both in _ spin space _ , not in real space .
these ideas ( and related experiments ) were the centre of enormous interest in the mid - late 1980 s , in almost all fields of physics ( and were extensively reviewed then @xcite ) .
in this article we will be interested in the spin phase of the _ environment_. we stress here that the environmental `` spin '' variables may not necessarily refer to real spins ( they can describe defects , or other such `` 2 level systems '' ) , but they will still have an associated dynamic topological phase , which can be described by an effective spin bath variable . the environmental spin phase interacts with the phase of the central system , causing phase decoherence in its dynamics @xcite . from the point of view of measurement theory , this environmental phase decoherence comes from a `` phase measurement '' made by the spin environment @xcite , in a kind of `` inverse stern - gerlach '' setup ( where the spins , instead of being measured , are themselves doing the measuring ! ) .
such phase decoherence also arises from oscillator baths , in a rather different ( and much less effective ) way @xcite .
in fact , the relevant phases involve both an adiabatic `` berry '' term and a second term coming from transitions between different bath states ( section 3.a ) .
there are also other decoherence mechanisms associated with the bath spins , coming from both the temporal fluctuations of the bias on the central system caused by the spin bath ( section 3.d ) , and from the precession of the spins in the spin bath ( with their associated phase ) _ in between _ transitions of the central system ( section 3.c ) .
thus the question of how the bath spin dynamics influences the central system is not simply a question of looking at berry phases .
practical application of the theory to , eg . ,
squids , or nanomagnets , or `` qubits '' ( section 5 ) , must include all mechanisms properly ( section 3.e ) .
the tactic adopted in this article is to focus on a `` central spin '' model @xcite ( sections 3 and 4 ) , in which the role of each term is exposed rather clearly ; after this one sees how things generalise to other models .
this model is directly relevant to qubits , and to the observation of mesoscopic or `` macroscopic '' quantum coherence- indeed we maintain that any practical design of such devices must involve the elimination , by one means or another , of _ all _ decoherence mechanisms from the relevant spin bath @xcite .
we begin the article ( section 2 ) by showing how both oscillator and spin bath models arise as the low - energy truncated versions of higher energy hamiltonians .
we give several examples , both magnetic and superconducting , to illustrate this .
then , in section 3 we explain how , mathematically , one averages over the spin bath variables to find the behaviour of the central system .
this is done pedagogically- we use the example of the central spin model ( and compare it with the spin - boson model ) .
various simple limits are introduced , and solved , before the general technique is given at the end of section 3 . in section 4
we give some results @xcite for the dynamics of the central spin model , in various regimes , to show how the physics is influenced by the spin bath ; and we also show how the system reduces to the spin - boson system in the weak - coupling limit . for those readers interested in the mathematical details , these are sketched in 2 appendices .
finally , in section 5 we return to physical applications , particularly to quantum magnetic systems and superconductors .
we then discuss decoherence , and show how this should persist even in the @xmath3 limit .
we finish by discussing the important application of the central spin model to decoherence in qubits and in quantum computation .
in studying the low - energy dynamics of a central quantum system coupled to an environment , we begin by truncating out " the unwanted high - energy physics , to produce a low - energy effective hamiltonian .
this is of course a quite general technique in physics , and one way to approach it is illustrated in fig.1 .
typically one has a reasonably accurately known
high - energy " or
bare " hamiltonian ( or lagrangian ) for a quantum system , valid below some ultraviolet " upper energy cut - off energy @xmath11 , and having the form @xmath12 where @xmath13 is an @xmath14-dimensional coordinate describing that part of the system we are interested in ( with @xmath15 the corresponding conjugate momentum ) , and @xmath16 are @xmath17-dimensional coordinates describing all other degrees of freedom which may couple to @xmath18 .
conventionally one refers to @xmath16 as environmental coordinates .
@xmath19 is of course a low - energy form of some other even higher - energy hamiltonian , in a chain extending ultimately back to quarks , leptons , and perhaps strings .
if , however , one is only interested in physics below a much lower energy scale @xmath20 , then the question is - can we find a new effective hamiltonian , of form @xmath21 in the truncated hilbert space of energies below @xmath20 ?
in this @xmath22 , @xmath23 and @xmath24 are generalised @xmath25-dimensional coordinates of interest , and @xmath26 are @xmath1-dimensional environmental coordinates coupled to them .
since we have truncated the total hilbert space , we have in general that @xmath27 and @xmath28 .
why do we make this truncation ( after all , its inevitable effect will be to generate new couplings between the low - energy modes ) ? essentially because in many cases the truncation pushes the new @xmath22 towards some low - energy
fixed point " hamiltonian ; and many different physical systems may flow to the same fixed point .
this allows us to speak of universality classes `` of quantum environment , and of a small number of ' ' canonical `` effective hamiltonians .
all physical systems in the same universality class will be described by the same form for @xmath29 , albeit with different values for the couplings .
as one varies the uv cut - off @xmath20 , the couplings change and any given system moves in the ' ' coupling `` or ' ' effective hamiltonian " space ; but they all move towards the same fixed point ( or fixed line , in some cases ) as @xmath30 is reduced . the various coupling terms in @xmath31 , simply parametrise the path it takes as @xmath32 ( fig.1 ) .
explicit derivations of @xmath29 for particular systems are lengthy ; see eg .
@xcite for general discussions , ref .
@xcite for comparision of spin and oscillator bath systems , and @xcite for specific examples . in this article
we will go directly to the canonical models , giving some examples of each so that readers can see how the high - energy hamiltonians are related to the models for some real systems .
to warm up we recall the basic structure of the oscillator bath effective hamiltonians , and then move on immediately to discuss various canonical models involving spin baths . for models in the general `` universality class '' of oscillator bath environments
, @xmath22 takes the form @xcite : @xmath33 \;\ ; + \;\ ; { 1 \over 2 } \sum_{q=1}^{n } \left ( { p_q^2 \over m_q } + m_q \omega_q^2 x_q^2 \right ) \ ; , \label{2.3}\end{aligned}\ ] ] where the generalised bath coordinates @xmath34 are now oscillator displacements @xmath35 and momenta @xmath36 ; these describe _ delocalised _ modes .
the couplings @xmath37 and @xmath38 are @xmath39 , so that in the `` thermodynamic limit '' @xmath40 , appropriate to a macroscopic environment of delocalised oscillators , these couplings are small @xcite ( a number of studies have also shown how higher - order couplings can be absorbed into linear but @xmath0-dependent couplings @xcite ) .
a special case of ( [ 2.3 ] ) is the feynman - vernon bilinear coupling form @xcite : @xmath41 where the couplings @xmath42 as well . in
recent years great attention has been given to problems where @xmath43 describes a tunneling system ( the `` caldeira - leggett '' model @xcite ) ; there have also been extensive studies of an oscillator coupled to oscillators @xcite , of free particles coupled to oscillators @xcite and of band particles coupled to oscillators @xcite .
suppose now the potential @xmath44 has a 2-well form , with small oscillation frequencies @xmath45 , and a `` bias '' energy difference between the two minima @xmath46 .
then for energies @xmath46 , one further truncates to the celebrated spin - boson `` model @xcite : @xmath47 x_q \;\ ; + \;\ ; { 1 \over 2 } \sum_{q=1}^{n } \left ( { p_q^2 \over m_q } + m_q \omega_q^2 x_q^2 \right ) \ ; , \label{2.4}\end{aligned}\ ] ] where the two - level central system ( with tunneling amplitude @xmath48 and longitudinal bias @xmath49 ) is described by the pauli matrix vector @xmath50 , coupled to background oscillators having energies @xmath51 .
we have introduced a topological phase @xmath52 for the central system , which depends in general on an external field ; the simplest and best - known example is a form @xmath53 , arising from the interference between 2 paths of amplitude @xmath54 in the motion of the central system .
this kind of ' ' aharonov - bohm " interference is well - known in superconductors ( where the phase is just the superconducting order parameter phase ) , and in magnets ( where it is the topological spin phase @xcite ) .
one can have a more complicated dependence of @xmath48 on @xmath52 ( eg .
, using multiple squids , or a nanomagnet with @xmath25-fold rotation symmetry ) , but we will stick with the simple @xmath55 dependence in this article .
for consistency we must assume @xmath56 , otherwise higher levels will be involved .
typically @xmath57 is dropped , because its effects are down on those of @xmath58 by a factor @xmath59 in tunneling rates ; but sometimes @xmath60 ( for reasons of symmetry ) , and then @xmath61 must be retained . the tunneling amplitude @xmath62 , where @xmath63 is the tunneling action .
the fame of the spin - boson model partly arises because many well - known problems in condensed matter physics can be mapped to it- this is a good example of the universality `` mentioned above . because the effect of each oscillator on the central system ( and vice - versa ) is very small , it may be entirely incorporated in second - order perturbation theory ( ie .
, to order @xmath64 ) for the general form ( [ 2.3 ] ) , or to order @xmath65 for the bilinear forms ( [ 2.3a]),([2.4 ] ) . in the latter case
this immediately encapsulates all environmental effects in the spectral function @xcite @xmath66 where @xmath67 . in general @xmath68 also depends on @xmath0 , even in the low-@xmath10 limit @xcite .
the case where @xmath69 is referred to as ' ' ohmic `` @xcite . because the @xmath70 , the @xmath68 are independent of @xmath1 and have the usual ' ' response function " form .
we give just 2 examples here of how a spin - boson model can arise , in the description of mesoscopic systems at low energies .
both truncations ignore the presence of spin bath modes ( for which see sections 2.e and 2.f , where we return to these 2 examples ) .
* ( i ) nanomagnet coupled to phonons or electrons * : the electronic spin dynamics of nanomagnets are often described by a `` giant spin '' hamiltonian @xmath71 , describing a quantum rotator with spin quantum number @xmath72 .
this model @xcite assumes the individual electronic moments are locked together by strong exchange interactions @xmath73 into a monodomain giant spin , with @xmath74 ( summed over local moment sites ) .
this only only works @xcite below a uv cut - off energy @xmath11 considerably less than @xmath73 .
however we are interested in the quantum dynamics for energies @xmath46 , where @xmath20 is controlled mostly by the single - ion magnetic anisotropy ; in real nanomagnets @xmath75 .
any real nanomagnet has couplngs to a spin bath of nuclear and paramagnetic spins , and to oscillator baths of phonons and electrons , which we now describe .
the `` high - energy''coupling between phonons and @xmath76 is described by terms like @xcite : @xmath77 \ ; , \label{1.ph}\ ] ] where m@xmath78 is the electron mass , @xmath79 the mass of the molecule , and @xmath80 is the debye temperature ( with @xmath81 the relevant lattice spacing , and @xmath82 the sound velocity ) .
the interaction @xmath83 and dimensionless ; a typical example is the non - diagonal term @xmath84 , which causes phonon emission or absorption with a change @xmath85 in @xmath86 ( since @xmath87 .
one also has diagonal terms in which @xmath88 is replaced by , eg .
@xmath89 ; and there are also higher couplings to , eg , pairs of phonons .
truncation to the `` quantum regime '' @xcite then gives the spin - boson model ( [ 2.4 ] ) , with a dominant non - diagonal coupling @xmath90 , coming from terms like ( [ 1.ph ] ) . in the absence of external fields in hamiltonian @xmath91 ,
the diagonal coupling @xmath58 is actually _ zero _ ( because of time - reversal symmetry ) .
the caldeira - leggett spectral function for the system has the form @xmath92 where @xmath93 ; here @xmath94 is the density of the medium supporting debye phonons , and @xmath95 . if now we couple _
electrons _ to the giant spin , it is the _ diagonal _ coupling which dominates @xcite .
the electronic coupling to @xmath76 depends on the type of magnetism .
some details have been worked out for kondo interactions with conduction electrons- the coupling to @xmath76 is @xmath96 where @xmath97 is the mean value of the kondo couplings to each individual electronic spin in the nanomagnet , and @xmath98 is a `` form factor '' integrating the localised electron spin density @xmath99 over the nanomagnetic volume @xmath100 . at low energies
the corresponding spin - boson model has an `` ohmic '' diagonal spectral function @xmath101 .
the size of @xmath102 depends on how the electrons permeate the nanomagnet @xcite ; if they permeate freely , @xmath103 , where @xmath104 is the mean dimensionless kondo coupling , and @xmath105 the fermi surface density of states .
typically @xmath106 , so @xmath102 can be big . *
( ii ) rf squid ( flux coupled to electrons ) * : we briefly recall one well - known application of the spin - boson model , to an rf squid coupled to both normal electrons ( in shunts , etc . ) , and bogoliubov quasiparticles @xcite . the flux @xmath107 passing through a superconducting ring with a weak link moves in a multiwell potential , which can be adjusted so that the lowest 2 wells ( each with small oscillation or `` josephson plasma '' frequency @xmath108^{1/2}/\phi_o$ ] , where @xmath109 is the josephson weak link coupling energy ) are almost degenerate , and dominate the low - energy properties .
the high - energy coupling between the flux and the electronic quasiparticles has the form @xcite @xmath110 where @xmath111 labels oscillator states describing a quasiparticle pair @xmath112 with energy @xmath113 , @xmath114 is the relevant junction tunneling matrix element , @xmath115 the symmetric / antisymmetric bcs coherence factor , and @xmath116 is the flux quantum
. thus we have a coupling to both the momenta and coordinates of the oscillators , which can also be written as a coupling to 2 independent oscillator baths @xcite .
the @xmath0-dependence of the coherence factors ( coming from the bcs gap dependence ) as well as the gap structure in their energy dependence , gives a complex structure in @xmath117 .
the reduction to the spin - boson model is now trivial @xcite , the minima in @xmath107-space of the effective potential corresponding to the 2 eigenstates of @xmath118 .
now suppose we have a high - energy hamiltonian of form ( [ 2.1 ] ) , but where the environmental coordinates @xmath119 are a set of @xmath1 spin-@xmath5 variables @xmath120 , ( i.e. , two - level systems ) ; and we assume the interspin couplings to be weak .
then , instead of ( [ 2.3 ] ) , we have @xmath121 @xmath122 \bigg ] \ ; ; \label{2.6}\ ] ] @xmath123 for energy scales @xmath124 .
thus we now have a central system coupled to a `` spin bath '' , described by @xmath125 in ( [ 2.7 ] ) .
the couplings @xmath126 and @xmath127 , between the central system and the bath spins , are usually much greater than the interspin couplings @xmath128 ; this means that the dynamics of each spin is largely `` slaved '' to that of the central system . unlike oscillator baths ( whose modes typically represent delocalised environmental degrees of freedom ) , the @xmath120 represent _ localised _ modes ( whose weak spatial overlap explains why the @xmath128 are small ) . this fact underlies a crucial difference between oscillator and spin bath environments- the couplings @xmath129 and @xmath127 are _ independent of the number _ @xmath1 _ of bath spins_.
thus the larger is @xmath1 , the larger is the total effect of the spin bath on the central system- there is no strict thermodynamic limit in the system , and it is not meaningful to let @xmath130 .
we emphasize also that we see no justification in general for spin bath models in which @xmath131 , or even @xmath132 ( although one can certainly invent artificial models of this kind ) .
thus , if we add more localised environmental modes to our environment , it is clear that the different modes are approximately independent ( as they will be if quasi - localised ) , so that their individual couplings to the central system will be hardly affected , ie .
, will depend only weakly on @xmath1 .
there is nothing to stop generalisation of this model to include bath spins @xmath133 , with @xmath134 ; the @xmath135 states then represent the degrees of freedom of , eg . , a defect , or a spin ( again , localised ) .
this introduces tensor ( eg . , quadrupolar ) couplings to the bath spins @xcite , and thereby complicates the mathematics- but does not alter the basic physics .
we will not discuss this here ( for the relevant formalism , and its application to the @xmath136 - 8 molecular nanomagnet , see refs .
@xcite ) .
a particle moving through a spin bath is described by ( [ 2.5 ] ) , in which @xmath23 and @xmath24 describe the momentum and position of the moving particle .
the diagonal term @xmath129 is analogous to the `` position '' oscillator coupling @xmath37 in ( [ 2.3 ] ) , and likewise @xmath127 to corresponds to @xmath38 .
however both forms can be altered by canonical transformation , corresponding to a rotation between the different coordinates .
the most common problems involve a diagonal coordinate coupling @xmath137 and a non - diagonal momentum coupling @xmath138
. then bath transitions ( spin flips ) are induced by the motion of the particle , whereas a stationary particle sees a `` potential '' @xmath139 , in general time - dependent .
a nice mesoscopic example of this is a large magnetic soliton coupled to background spins @xcite . in many realistic cases
the most important such coupling will be to paramagnetic impurities , but here we consider the simpler case of a hyperfine coupling to a set @xmath140 of @xmath1 spin-@xmath5 nuclear spins . in this case @xmath141 describes the extremely weak internuclear dipolar coupling ; typically @xmath142 ; and @xmath143 is any external field that might unfluence these nuclei .
the `` high - energy '' hamiltonian for such a wall is usually determined as an integral over the magnetisation density @xmath144 and its spacetime gradients @xcite .
> from this one eliminates the details of the wall profile altogether , to produce a `` bare '' hamiltonian ( ie . , neglecting the environment ) for the wall coordinate ; in simple cases where the wall demagnetisation field keeps the wall almost flat , this gives @xcite @xmath145 [ hwall ] for a wall with centre of mass coordinate @xmath24 and surface area @xmath146 .
the `` pinning potential '' @xmath147 , provided the pinning centre is much smaller than the wall width @xmath148 .
the term linear in @xmath24 comes from an external magnetic field @xmath149 .
what now of the environment ? in the literature there is extensive discussion of the effects of magnons ( ie .
, spin waves)@xcite , electrons @xcite , and phonons @xcite on the wall dynamics- these are all oscillator baths .
however at low @xmath0 spin bath effects , coming from nuclear and paramagnetic spins , will completely dominate . even in ni ( where only @xmath150 of the nuclei have spins , with a tiny hyperfine coupling @xmath151 ) , all real samples have an important concentration of paramagnetic spins ( caused by oxygen in the sample ) as well as many defects . in rare earths , the hyperfine coupling @xmath152 ( @xmath153 ) , and hyperfine effects alone are quite massive .
thus we must modify @xmath154 above to @xmath155 in which the electronic spins @xmath156 couple locally to @xmath1 nuclear spins @xmath157 at positions @xmath158 ( @xmath159 ) , via a hyperfine coupling @xmath160 ( and also in general to paramagnetic spins ) .
the internuclear coupling @xmath161 ( @xmath162 ) , ie . ,
@xmath163 , but it gives the spin bath its own dynamics . to write the hamiltonian in the form ( [ 2.6 ] ) , we write the continuum magnetisation @xmath164 , where @xmath165 is the slowly - varying part descibing the wall profile and @xmath166 describes fluctuations around this .
then we rotate the spin quantisation axis to be locally parallel to @xmath165 , and get @xcite @xmath167 \bigg ] \;+\ ; \frac{1}{2 }
\sum_{k } \sum_{k ' } v_{k k'}^{\alpha \beta } i_{k}^{\alpha } i_{k'}^{\beta } \label{inutile } \label{dwh2}\ ] ] displaying explicitly the longitudinal and transverse couplings .
the `` particle '' moves through a slowly fluctuating `` random walk '' potential field @xmath168 ( coming from the sum over couplings @xmath169 to randomly oriented spins ) .
the transverse coupling ( independent of @xmath24 but not of @xmath23 ) causes `` spin flip '' transitions in the spin bath when the wall moves , even if the bath is at @xmath170 .
one may also discuss problems in superconductors and normal metals involving nuclear and paramagnetic spins , and other `` defects '' , which can also be mapped to the same model of a particle moving through a spin bath ( sections 2.f , 5.b ) .
is there a low - energy effective hamiltonian , analogous to the spin - boson model , in which a `` central '' 2-level system couples instead to a spin bath ?
the answer is yes , but the effective hamiltonian does not look quite so simple as the spin - boson one . in the _ absence _ of any external field , the analogue of the spin - boson form in ( [ 2.4 ] ) for a spin bath is actually @xcite @xmath171 + h.c . \right\ } \nonumber \\ & + & { \hat \tau } _ z \sum_{k=1}^n \omega_k^{\parallel } \ : { \vec l}_k \cdot { \hat { \vec \sigma } } _ k + \sum_{k=1}^n \omega_k^{\perp}\ : { \vec m}_k \cdot { \hat { \vec \sigma } } _ k + \sum_{k=1}^n \sum_{k'=1}^n v_{kk'}^{\alpha \beta } \hat{\sigma}_k^\alpha \hat{\sigma}_{k'}^\beta ~ \;. \label{1.24 } \end{aligned}\ ] ] where @xmath172 describes the central spin , and the @xmath173 the spin bath degrees of freedom .
this form is not the most general one- apart from dropping external field effects ( for which see below ) we have also restricted the central spin phase in a simple @xmath174 form ( _ cf_. introduction ) .
as discussed below , both @xmath175 and @xmath176 incorporate spin bath renormalisation effects .
the factor of 2 in front of @xmath175 is somewhat arbitrary ( if the cosine is one , then the actual `` tunnel splitting '' coming from ( [ 1.24 ] ) , in the absence of spin bath effects , will be @xmath177 ) .
the basic form of ( [ 1.24 ] ) is actually fairly easy to understand .
the extra phase in the first `` non - diagonal '' term ( adding to @xmath176 ) comes from the topological phase of the bath spins as they make transitions @xcite .
there are also diagonal terms , and a weak interaction @xmath178 between the bath spins .
this form assumes the diagonal couplings @xmath179 are @xmath180 the uv cutoff @xmath20 , and that @xmath181 .
the ratio @xmath182 is arbitrary . in sections 4 and 5
we shall see there is a weak coupling limit to this model , in which it reduces to a spin - boson system .
before this model was derived in this general form and then solved , a number of special cases had already been looked at @xcite . in particular , shimshoni and
gefen @xcite included only the diagonal terms @xmath183 and @xmath184 , and examined the results in weak coupling in the presence of an ac field ( see also @xcite ) ; they clearly recognised that the problem was different from an oscillator bath one .
let us now discuss the different terms in ( [ 1.24 ] ) , in the order they appear ( cf . fig .
\(i ) _ non - diagonal terms _ : that an extra phase term should exist , coming from the spin bath , is obvious on general grounds ( cf .
introduction ) .
one can understand its algebraic form in the following way .
notice that @xmath185 in ( [ 1.24 ] ) operates on both the central spin _ and _ the spin bath ; and the effect of a single giant spin transition on @xmath186 can always be written as a transition between initial and final nuclear states , in the form @xmath187 , where @xmath188 } \label{chi.2}\ ] ] the integral over @xmath189 is only defined once we know the trajectory of the central system during the tunneling event .
it is in this sense that we say @xcite that the instanton has become an `` operator '' in the space of the spin bath modes .
notice that in general the central spin phase @xmath176 and splitting @xmath190 are renormalised by the bath couplings @xcite : @xmath191 the @xmath192 are `` berry phase '' terms coming from the bath spin dynamics during a central spin flip , and the @xmath193 come from high - frequency modifications of the original high - energy potential ( `` barrier fluctuations '' ) .
one can show that both @xmath194 and @xmath195 are @xmath196 , where @xmath197 is the larger of @xmath179 , and we will ignore these terms from now on .
expanding out the cosine in ( [ 1.24 ] ) gives a series of terms like @xmath198 in which the instanton flip of the central spin couples simultaneously to many _ different _ bath spins- a single central spin transition can cause multiple transitions in the bath ( fig .
2 ) . later we will introduce a parameter @xmath199 which measures the average number of bath spins flipping during each instanton .
\(ii ) _ diagonal terms _ : these act _ between _ transitions of the central spin ( fig . 2 ) , and are also easy to understand . formally ,
one starts by considering the `` initial '' and `` final '' fields ( ie .
, before and after a transition of @xmath172 ) acting on @xmath200 .
calling these fields @xmath201 and @xmath202 respectively ( fig .
3 ) , we define the _ sum _ and the _ difference _ terms as @xmath203 where @xmath204 and @xmath205 are unit vectors . then the truncated diagonal interaction takes the form @xmath206
i.e. , one term which changes during a transition of the central system , and one which does not . the longitudinal coupling @xmath207 determines the gross structure of the bath states in energy space : it also determines an `` internal bias field '' @xmath208 acting on @xmath209 .
the 2 levels of bath spin @xmath210 are split by energy @xmath183 , depending on whether @xmath211 is parallel or antiparallel to @xmath209 .
the effect of this on the @xmath212 fold multiplet of bath states surrounding each central spin state is shown in fig .
4 . suppose we classify these states by their `` polarisation group '' ; all bath states whose total longitudinal polarisation @xmath213 are in polarisation group @xmath214 .
since the @xmath183 vary from one bath spin to another , states in polarisation group @xmath214 are spread over an energy range @xmath215 ; and the entire manifold of states , comprising all polarisation groups , is spread over a larger energy range @xmath216 .
let us define _ normalised _ densities of states @xmath217 and @xmath218 for these 2 distributions , so that @xmath219 where @xmath220 . in almost any physical case one will have strongly overlapping polarisation groups , so that for all but very small values of @xmath1 , or except in the extreme wings of the distributions , one has @xmath221 @xmath222
the simplest case is where the @xmath183 cluster around a single central value @xmath223 with variance @xmath224 for this case we define a parameter @xmath225 , characterising the degree of polarisation group overlap ; overlap is complete when @xmath226 . then @xmath227 and @xmath228 ( so that @xmath229 ) . in the extremely unlikely case
where @xmath230 , @xmath218 can no longer be treated as gaussian- however there is an intrinsic lower limit to the linewidth of each polarisation group set by the interspin interaction @xmath178 .
this `` intrinsic linewidth '' @xmath231 , where @xmath100 is a typical value of @xmath178 ; in any physically realistic case this is usually enough by itself to cause complete overlap of all groups ( the essentially non - interacting case where @xmath232 , ie .
, @xmath233 for all bath spins , and @xmath234 , was actually studied by garg @xcite ) .
the `` transverse '' couplings @xmath184 arise when the fields before/ after a transition , acting on the @xmath235 , are not exactly parallel or antiparallel .
this can happen in many ways , either because of external fields which couple to the bath spins , or because of a lack of symmetry in the underlying dynamics of the central system , or in its coupling to the bath modes .
thus they are non - zero in any realistic situation .
\(iii ) _ internal spin bath dynamics _ : finally , the interaction @xmath178 is usually so weak that it does not change under truncation .
if this term is absent , the spin bath will have no `` intrinsic '' dynamics , and remains inert between transitions of @xmath172 .
thus even if small , @xmath178 is important , since it allows the bath state to evolve during these intervals .
the most important effect of @xmath178 is that it allows the longitudinal bias field @xmath236 to fluctuate in time , between and during transitions of @xmath172 .
notice , however , that with the hamiltonian in the form ( [ 1.24 ] ) , _ only fluctuations within the same polarisation group are allowed_. in nmr language , only @xmath237 processes occur in the intrinsic dynamics of the spin bath- changes in @xmath214 can only occur via the interaction with the central spin .
this will only be true at low @xmath0- at higher @xmath0 longitudinal relaxation ( ie .
, `` @xmath238 processes '' , in nmr language ) between different polarisation groups should be included in a realistic model . such processes ( which arise from the interaction of the spin bath modes with other environmental modes , or with thermally excited higher modes of the central system , above the uv cutoff energy @xmath20 ) are almost always very slow when @xmath239 .
on the other hand @xmath237 fluctuations will persist until @xmath240 ; in the physical examples studied so far this means they persist down to @xmath241 temperatures or below .
we will return briefly to this very low @xmath0 regime at the end of the article ( section 5.c ) .
* external field effects * : in general all of the parameters in ( [ 1.24 ] ) will depend on any external field @xmath242 , because it changes the high - energy dynamics of both central system and bath- however we can make low - field expansions and separate out the most important terms . defining the `` zeeman '' coupling energies @xmath243 and @xmath244 of central and bath spins to this field
, it is easy to see that under the conditions @xmath245 and @xmath246 , the principal changes to ( [ 1.24 ] ) will be ( i ) the addition of an obvious longitudinal coupling @xmath247 to the central spin , and ( ii ) qthe changes @xmath248 , @xmath249 , and @xmath250 , where up to _ linear order _ in @xmath251 one has @xmath252 thus even at low fields one has an important change in all the topological phases in the problem , and also to the _ transverse _ diagonal coupling ( which itself arises from internal fields ) . in general @xmath253
will also change the interspin bath couplings @xmath178 ; in a way which depends on the specific details of the problem . in the next 2 sections
we see how this works for both magnetic and superconducting systems .
if we start from the `` giant spin '' model introduced above for a nanomagnet , then a simple isotropic contact hyperfine coupling to nuclear spins will lead to a hamiltonian ( for @xmath254 ) like : @xmath255 where @xmath71 is the
giant spin " hamiltonian , and @xmath125 is the same as in ( [ 2.7 ] ) .
the generalisation of this simple hamiltonian to include dipolar hyperfine interactions , as well as to higher spin nuclei and to paramagnetic spins ( with tensor and quadrupolar couplings ) can be used if necessary @xcite . here we will assume for simplicity that @xmath256 , and write @xmath257 , i.e. the nuclear spins will be described by spin-@xmath258 pauli matrices .
in fact in many cases even if @xmath259 , the low - energy nuclear spin dynamics is well described by a 2-level system .
the truncation of @xmath260 to a central spin hamiltonian @xmath261 has been discussed in several papers @xcite . as an example we quote the result for a simple easy axis - easy plane nanomagnet ( for which @xmath262 $ ] , and give it a physical interpretation . in this case , assuming @xmath263 and also a weak external field @xmath242 , the effective hamiltonian is @xcite : @xmath264 + h.c .
\right\ } \nonumber \\ & + & { \hat \tau } _ z \bigg [ \xi_h + \sum_{k=1}^n \omega_k^{\parallel } \
: { \hat \sigma } _ k^z \bigg ] \ ; + \ ; \sum_{k=1}^n \sum_{k'=1}^n v_{kk'}^{\alpha \beta } \hat{\sigma}_k^\alpha \hat{\sigma}_{k'}^\beta ~ \;. \label{z.3301}\end{aligned}\ ] ] ie . , a special case of the general form ( [ 1.24 ] ) , with the parameters @xmath265 , @xmath266 , @xmath267 , and @xmath268 .
the vectors @xmath269 and @xmath270 for this easy axis - easy plane case turn out to be ( again assuming @xmath263 , and small @xmath242 ) : @xmath271 in this example there are 2 tunneling trajectories ( clockwise and counterclockwise in the easy plane ) , giving a result @xmath272 for the `` transfer matrix '' @xmath273 ( in zero applied field ) .
the resulting vector @xmath274 is the `` average hyperfine field '' acting on @xmath200 during the tunneling event . to understand its orientation ( and why it is _ complex _ )
we note that the nuclear spin itself exercises a torque on @xmath76 while it is tunneling , and this pushes @xmath76 away from the easy plane .
consequently ( a ) the average field acting on @xmath200 has a component out of the easy plane , in the @xmath275-direction , and ( b ) @xmath76 no longer moves exactly along the easy - plane path , while tunneling , that it would in the absence of @xmath200 ( and so its action increases , via the imaginary part of @xmath274 ) .
the contribution @xmath276 to the topological phase @xcite comes from the area swept out by the giant spin on the spin sphere ( cf .
introduction ) , which changes as the field @xmath242 changes ; it is essentially an `` aharonov - bohm '' contribution to this phase from the external field @xcite , which leads to spectacular oscillations ( fig .
5 ) in the effective tunneling amplitude @xmath277 $ ] for a field perpendicular to @xmath278 @xcite .
very recently oscillations in the tunneling amplitude of @xmath136 - 8 magnetic molecular crystals were seen which are related to this @xcite , although the presence of both nuclear spins and dipolar fields seriously complicates their interpretation @xcite ( see section 5.a ) .
we recall from section ii.b that the giant spin model truncates for nanomagnets to a 2-level system for energies @xmath279 .
contact hyperfine couplings are in the range @xmath280 ( transition metals ) or @xmath281 ( rare earths ) ; on the other hand the internuclear couplings @xmath282 . in the @xmath136 - 8 system just mentioned the hyperfine interactions are actually dominated by dipolar couplings between the 8 @xmath283 ions and the 120 protons in the molecule ; these couplings are in the range @xmath284 ( @xmath285 ) .
when the hyperfine couplings are this weak we must also take into account the effect of external fields on the nuclear dynamics @xcite ( section 5.a ) .
the values of @xmath190 vary over a huge range , but typically @xmath286 ( in @xmath136 - 8 , @xmath287 ) .
we consider again the rf squid , but now concentrate on the coupling of the flux @xmath176 to the spin bath of nuclear and paramagnetic spins which are within a penetration depth of the surface of the superconductor .
this example is very instructive in understanding the weak - coupling limit of the central spin model ( the following discussion is based on refs .
@xcite ) .
suppose to start with we consider a `` cubic geometry '' @xcite , in which a cube of side @xmath288 has a hole of radius @xmath289 through it , with in addition a slit connecting the hole to the exterior , spanned by a cylindrical junction of length @xmath290 and diameter @xmath291 .
the magnetic field inside the hole corresponding to a half - flux quantum is @xmath292 , whereas the magnetic field in the junction is as high as @xmath293 .
there are both nuclear spins and paramagnetic impurities in the spin bath .
consider first the nuclear spins ; assuming all nuclei have spins , we find that in the bulk of the ring , within a penetration depth of the surface , there are @xmath294 nuclear spins coupling to the ring current ; and in the junction itself , a number more like @xmath295 .
thus each ring nuclear spin couples to the squid with a diagonal coupling @xmath296 ; on the other hand for junction spins this coupling is @xmath297 .
notice we have ignored any coupling to substrate spins ( assume , eg . , the ring is in superfluid he-4 ! ) , which might have a much larger coupling to the current . at any temperature such
that @xmath298 , the typical polarisation of these spin baths will be @xmath299 respectively , giving a distribution of longitudinal bias energies with typical values @xmath300 , and @xmath301 acting on the tunneling flux coordinate @xmath176 .
if we now add paramagnetic impurities to the ring , with concentration @xmath302 , and coupling @xmath303 to the current , this gives a typical longitudinal bias energy @xmath304 .
this longitudinal term is obviously bigger than the nuclear contribution , unless the superconductor is very pure indeed !
however , there is another much stronger transverse term , because each spin feels the dipolar fields from the other spins .
this field is @xmath305 ( much higher near to the paramagnetic spins ) , and for the nuclear spins has an associated energy @xmath306 , which is @xmath307 .
physically , when the squid flips , the field on each nuclear spin hardly changes its direction , being dominated by the more slowly varying ( but much stronger ) nuclear dipolar field . for
the paramagnetic spins the analogous coupling @xmath308 is @xmath309 times larger , which in the absence of nuclear fields would give an inter - paramagnetic `` flip - flop '' rate @xmath310 hz , except that in pure samples these processes will themselves be blocked by the local dipolar coupling between the impurity and nearby nuclear spins ( of strength @xmath311 ) ; this will happen once @xmath312 .
we can write down an effective hamiltonian for this system , valid over timescales considerably greater than @xmath313 ; we will use this later to analyse the effect of the spins on the squid dynamics ( section 5.b ) .
we will assume that @xmath314 ( the only case of experimental interest ) ; then we can treat the internuclear dipolar fields as slowly - varying in time .
the effective hamiltonian can then be derived , to give @xcite : @xmath315 \label{cases.3}\ ] ] where @xmath316 , @xmath317 or @xmath318 , depending on the spin , and @xmath184 has just been discussed ; and microscopic analysis shows that @xmath319 , where @xmath20 is the josephson plasma frequency ( section 2.b ) .
notice that @xmath320 , which is the opposite limit considered to that for the giant spin ! notice further that these couplings are far less than @xmath321 ( @xmath322 only because there are so many spins involved ) .
thus the spin bath is no longer `` slaved '' to the central system . in section 4
we see how this allows a mapping to an oscillator bath , coupled to @xmath172 ( ie . , a spin - boson model ) .
finally we note that the external field also acts on the nuclear and paramagnetic spin dynamics , via the zeeman coupling ( which simply adds to @xmath183 ) .
this can help to suppress decoherence effects , by freezing out the spin bath dynamics- for more details see ref.@xcite .
we now recall our assertion that almost any mesoscopic `` central '' system , coupled to its environment , may be described at low energies by ( [ 2.2 ] ) , with the environment being written as a sum of an oscillator bath term ( [ 2.3 ] ) and a spin bath term ( ( [ 2.7 ] ) , or a higher spin generalisation ) .
the simplest example is of a single central spin coupled to both oscillator and spin baths .
such a model seems forbidding but in fact a fairly complete analysis has been given of its dynamics @xcite- we recall some of the results at the end of section 4 .
one can also consider a much more complicated model in which a _ macroscopic array _ of central spins @xmath323 , at positions @xmath324 , couples to both oscillator and spin baths . the effective hamiltonian is then an obvious generalisation of what has gone before : @xmath325 + h.c .
\right\ } \ ; + \ ; \sum_{i < j } v({\bf r}_i - { \bf r}_j )
\hat{\tau}_i^z \hat{\tau}_j^z \nonumber \\ & + & \sum_j\left\ { { \hat \tau } _ j^z \sum_{k=1}^n \omega_{jk}^{\parallel } \ : { \vec l}_{jk } \cdot { \hat { \vec \sigma } } _ k + \sum_{k=1}^n \omega_{jk}^{\perp}\ : { \vec m}_{jk } \cdot { \hat { \vec \sigma } } _ k \right\ } + \sum_{k=1}^n \sum_{k'=1}^n v_{kk'}^{\alpha \beta } \hat{\sigma}_k^\alpha \hat{\sigma}_{k'}^\beta \nonumber \\ &
+ & \sum_j \sum_q \bigg [ c_{jq}^{\parallel } \hat{\tau}_j^z + ( c_{jq}^{\perp } \hat{\tau}_j^{- } + h.c . ) \bigg ] x_q + { 1 \over 2 } \sum_q \left ( { p_q^2 \over m_q } + m_q \omega_q^2 x_q^2 \right ) ~\ ; , \label{qsgl}\end{aligned}\ ] ] where @xmath326 is a `` high - energy '' diagonal coupling between the various `` central spin '' systems . if we throw away the spin bath we get a set of 2-level systems coupling to an oscillator bath , of which the simplest example is the `` pisces '' model ( in which there are only two 2-level systems @xcite ) .
such models seem impossibly complicated , but actually one can solve for their dynamics in many important regimes !
what is crucial is the _ separation _ of the 2 baths .
often ( as with nuclear spins ) their mutual interaction is very weak ( and can be parametrised by a time @xmath327 which may be very long at low @xmath0 ) ; in this case this separation is a good one . if there _ are _ certain spin bath modes that interact strongly with the oscillators , then typically we can simply absorb these modes into an `` augmented '' oscillator bath by a canonical transformation .
an obvious example arises with electronic spins in a metallic host ( the kondo or kondo lattice problems ) ; one rewrites the bath to include the `` kondo resonance '' in the oscillator bath spectrum .
a proof that one may do this in all cases seems rather difficult- in any case the usefulness of these models tends to be established by their application .
models like ( [ qsgl ] ) describe mesoscopic systems like coupled squids or coupled nanomagnets @xcite , quantum spin glasses @xcite and low-@xmath0 dipolar glasses @xcite , as well as coupled anisotropic coupled kondo spins and kondo lattices , coupled nuclear spin systems @xcite , superconducting arrays , or coupled defects in solids .
they are also useful for analysing purely theoretical questions about relaxation , dissipation , decoherence and quantum measurements in quantum systems- many questions remain unanswered , having only been studied thoroughly in restricted models such as the spin - boson model @xcite or the pisces model @xcite .
we return to experimental and theoretical applications in section 5 .
to extract useful information from the low - energy canonical models , we must calculate their dynamical properties .
since we are typically not interested in the environment ( one usually has little control over it ) , one performs a statistical average over the environment .
this procedure is fraught with danger , because of `` memory '' effects in the environment , and because assumptions such as `` self - averaging '' in the environmental correlation functions may not strictly be valid . in this section
we show how the spin bath may be `` integrated out '' by means of 4 different statistical averages , each involving an integration over a particular variable .
the end result is a description of the time evolution of the `` reduced '' density matrix for the central system- _ provided _ we can ignore memory effects in the environment .
the starting point is no different from that involved in functional averaging over oscillators @xcite ; both begin with a path integral form for the propagator of the reduced central system density matrix , written as @xmath328 \ - \ s_o[q^{\prime } ] ) } { \cal f}[q , q ' ] \ ; , \label{3.3}\ ] ] where @xmath329 $ ] is the free central system action , and @xmath330 $ ] is the famous influence functional `` @xcite , defined in general by @xmath331 = \prod_{k } \langle \hat{u}_k(q , t ) \hat{u}_k^{\dag}(q',t ) \rangle \ ; , \label{neto.2}\ ] ] here the unitary operator @xmath332 describes the evolution of the @xmath333-th environmental mode , given that the central system follows the path @xmath334 on its ' ' outward `` voyage , and @xmath335 on its ' ' return " voyage ; and @xmath336 $ ] acts as a weighting function , over different possible paths @xmath337 . for a central 2-level system , the paths @xmath338
are simple ( recall fig .
2 ) : @xmath339 \ ; , \label{neto.1}\ ] ] where @xmath340 is the sign - function , and @xmath341 is the number of transitions of the central system , occuring at times @xmath342 ( for definiteness we assume trajectories starting and ending in the same state , and use the convention that @xmath343 corresponds to @xmath344 ) .
the goal is to find the central spin density matrix ; in this article we give results for the `` return probability '' @xmath345 for the system to be in the same state @xmath346 at time @xmath347 as it was at @xmath348 . using ( [ neto.1 ] ) this
can be written as an `` instanton expansion '' over flips of the central spin ( appendix a ) : @xmath349 \label{neto.22}\ ] ] further simplification arises if the environmental modes are uncoupled- then @xmath330 $ ] factorises , and we can write @xmath336 = \exp(-i\phi [ q , q ' ] ) = \exp(-i \sum_{k=1}^n \phi_k [ q , q'])$ ] , where the _ complex _ phase @xmath350 $ ] contains both real ( reactive ) , and imaginary ( damping ) contributions .
now for an _ oscillator bath _ one simplifying feature is crucial , viz .
, the very weak coupling to each oscillator .
this allows one to evaluate each @xmath350 $ ] up to 2nd order only in these couplings , in terms of a spectral function for the _ unperturbed _ oscillator dynamics ( compare @xmath117 in ( [ 3.6 ] ) ) . even though the paths @xmath334 and @xmath335 may be complicated , the calculation of @xmath330 $ ] is often tractable @xcite .
however because the coupling to each spin bath mode is _ not _ necessarily weak , it will in general strongly alter their dynamics , often _ slaving _ them to the motion of the central system . _ thus we can not start from the unperturbed spin bath dynamics_- the problem is fundamentally non - perturbative in the @xmath351 .
however it is not intractable , because one _ can _ rather easily deal with the dominant longitudinal terms @xmath352 .
the other terms can then be dealt with perturbatively ( and sometimes even non - perturbatively ) .
it is the separation of the effects of the various terms in the hamiltonian which leads to not one , but 4 different averages .
what is quite remarkable is that these averages can be evaluated analytically in most cases ( section 4 ) .
we begin by explaining the 4 different averaging integrals required for a general spin bath .
this is done pedagogically , by solving for 4 different limiting cases of the central spin hamiltonian ( sections iii.a - iii.d ) , each of which requires only one of the 4 averaging integrals .
then the general procedure ( combining the 4 averages ) is given in iii.e .
one reason for going through these averages one by one is that each corresponds to a different physical mechanism of decoherence- we return to this in section 5 ( note that more detailed results for the 3 limiting cases discussed in sections iii.b - iii.d are given in refs .
@xcite ) .
formally the case of pure topological decoherence applies to the following special case of @xmath185 , in which only non - diagonal terms are included : @xmath353 + h.c . \right\ } \ ; , \label{4.1}\ ] ] since @xmath354 are zero , _ all _ the @xmath212 environmental states are degenerate , and there is no exchange of any energy between @xmath355 and the @xmath356 .
the only thing that is exchanged is _ phase _ ; the phase @xmath52 of @xmath355 becomes entangled with that of the @xmath356 , during the transitions of @xmath355 between @xmath357 and @xmath358 , so that the initial and final _ states _ of the spin environment are different .
physically ( [ 4.1 ] ) would arise if the original high - energy hamiltonian contained only `` transverse '' couplings ( to @xmath359 ) , which only act while the central system is making transitions ( in the case of a moving particle coupled to a spin bath , they would be `` velocity couplings '' , only acting when the particle is moving ) .
we wish to determine @xmath345 for this case . for pedagogical purposes
let us begin by assuming that @xmath360 is _ real _ , ie .
, @xmath361 is pure imaginary ; then we have added a pure environmental phase term to the free 2-level hamiltonian .
then , writing @xmath362 , the formal solution to this problem can be written immediately as @xcite @xmath363 \rangle \right\}\ ; , \label{4.4}\ ] ] where the brackets @xmath364 trace over the spin bath . by writing this as an instanton expansion over central spin flips ( see app .
a and refs .
@xcite ) , we transform it to a _ weighted integration over topological phase _ : @xmath365 where @xmath366 is a _ phase - dependent _ tunneling amplitude , @xmath367 is a bessel function , and @xmath368 for small @xmath361 we may approximate the product in ( [ 4.6 ] ) as @xmath369 notice that @xmath199 is just the mean number of environmental spins that are flipped , each time @xmath370 flips .
clearly phase decoherence is important if @xmath371 , in which case @xmath372 .
then , rather surprisingly , we get a _ universal form _ ( shown in fig .
6 ) , in the intermediate coupling limit , for @xmath373 : @xmath374 \equiv \int { d\varphi \over 2 \pi } p_{11}^{(0)}(t,\phi = \varphi ) \;\;\;\;\;\;\;\ ( intermediate \;\ ; coupling ) \label{4.13a}\ ] ] with a phase integration over the _ free _ central spin propagator @xmath375 [ cf . eqtn .
( [ b1.4 ] ) ] . thus _
random phases _ arise because successive flips of @xmath355 cause , in general , a different topological phase to be accumulated by the spin environment .
in fact , the universal behaviour comes from complete phase phase randomisation @xcite , so that all possible phases contribute equally to the answer !
the final form shows decaying oscillations , with an envelope @xmath376 at long times , which can also be understood by noting that the `` zero phase '' trajectories contributing to @xmath377 constitute a fraction @xmath378 of all possible trajectories , where @xmath379 . in the _ strong coupling _ limit of this model ,
where the bath spins rotate adiabatically with the the central spin , one has @xmath380 , so that @xmath381 and @xmath382 $ ] , where @xmath383 , i.e. , the haldane / kramers phase is now @xmath384 , since the @xmath1 bath spins are forced to rotate with @xmath172 .
the results for complex @xmath361 are given in appendix a ; the basic ideas behind them ( and the techniques for their calculation ) are simple elaborations of the above .
we now consider the effective hamiltonian @xmath385 to solve this we will assume the model discussed previously , in which all @xmath352 cluster around a single central value @xmath223 ( cf . eqtns ( [ 4.28]),([gauss ] ) , and fig .
4 ) . since the bath
now just acts as an extra longitudinal field , we are dealing with the trivial case of a biased two - level system , with bias energy @xmath386 , where @xmath387 .
the only question is how to average over the internal bias- this depends on whether we deal with a _
single _ central system , or a statistical ensemble of them ( corresponding to either an average over many measurements on a single system , or a single measurement on a large number of _ non - interacting _ systems- interactions are discussed in section 5.a ) . for a single central spin ,
the dynamics of @xmath172 in this model are completely trivial- one has @xmath388 $ ] , where @xmath389 , ie . , resonant tunneling of an isolated spin in a longitudinal bias feld ( compare ( [ b1.10 ] ) ) . for an _ ensemble _ of central spins
, we must average over the whole bias range . in what follows
let us assume for definiteness a spin bath at some equilibrium temperature @xmath390 ; then the ensemble average is just a _ weighted average over bias _ : @xmath391 where @xmath392 is the appropriate partition function .
@xmath393 \label{4.32}\ ] ] the physical interpretation is obvious @xcite ; only a very small fraction @xmath394 of central spins in the ensemble in the `` resonance window '' , ie .
, , having total bias @xmath395 , can make transitions- this selects states with internal bias around @xmath396 .
all other states lack the near - degeneracy between initial and final energies required for resonant tunneling- they are `` degeneracy blocked '' @xcite .
the resulting correlation function is then @xmath397 where @xmath398 ( except in the unphysical case where the polarisation group linewidths @xmath399 . for the usual case where all polarisation groups overlap , and @xmath218 has the gaussian form ( [ gauss ] )
, one has @xmath400 .
it is not surprising to find that the spectral absorption function @xmath401 , corresponding to ( [ 4.33 ] ) , has the `` bcs '' form @xmath402 finally , let us note that one may imagine a case where one has an ensemble of systems in which , although the bath state is not fixed , the polarisation group is known to be equal to @xmath214 . in this case
we must replace ( [ biasav ] ) by @xmath403 in ( [ 4.32 ] ) for @xmath345 , where @xmath404 is the partition function for the @xmath214-th polarisation group . if we then recalculate @xmath405 in the same way we find almost zero absorption unless @xmath406 , where @xmath407 is again the linewidth of the @xmath214-th polarisation group ( cf .
( [ g_m ] ) . until now we have ignored the `` transverse field '' part @xmath408 of the diagonal term in the effective hamiltonian ( [ 1.24 ] ) . to study this
let us consider again an effective hamiltonian which has no non - diagonal terms apart from the `` bare '' tunneling , but having all diagonal terms : @xmath409 we will assume @xmath410 , ie .
, that the transverse `` orthogonality blocking '' part of the diagonal interaction is much smaller than the longitudinal part . to make things as simple as possible we drop all degeneracy blocking effects , ie . , we assume all @xmath183 are equal ( ie . , @xmath411 ) .
it then follows that the spin bath spectrum is split by @xmath412 into @xmath413 `` polarization groups '' of degenerate lines , with @xmath414 degenerate states in polarisation group @xmath214 ( cf .
( [ density ] ) . at first glance
it seems that the @xmath415 in ( [ 4.14h ] ) simply act on the central spin @xmath355 as an external field .
however this is _
wrong _ ; it ignores their role as dynamic quantum variables .
the dynamics come because the @xmath415 can _ precess _ in the fields @xmath416 acting on them , and these change each time @xmath355 makes a transition ( cf . eqtn .
( [ 1.25 ] ) ) .
quantum mechanically , the precession caused by @xmath417 is equivalent to saying that some bath spins are _ flipped _ when the central spin @xmath172 flips ( in general a different number of them during each transition of @xmath172 ) . to see this formally , recall that @xmath184 exists when the initial and final fields @xmath418 and @xmath419 acting on @xmath420 are not exactly equal and opposite ( cf .
eqtn ( [ 1.25 ] ) ) . defining the small angle @xmath421 by @xmath422
( recall fig .
3 ) , we see that the initial and final states of @xmath420 are related by @xmath423 suppose now the initial spin bath state belongs to polarisation group @xmath214 .
if , when @xmath172 flips , bath spins also flip so that @xmath424 , then since @xmath425 , _ resonance _ is still preserved , a transition is possible- indeed @xmath172 can not flip at all _ unless _ there is a change in polarisation state of magnitude @xmath426 . for this change in polarisation of @xmath426 , _ at least _
@xmath214 spins must flip ; moreover , for resonant transitions to continue ( incoherently ) , the bath polarisation state must change by @xmath427 _ each time _ @xmath355 flips .
let us therefore define @xmath428 as the correlation function @xmath345 restricted to systems for which the bath polarisation is @xmath214 .
for a thermal ensemble , @xmath429 with a weighting @xmath430 , where @xmath431 is the partition function . in appendix a.2
we calculate @xmath432 ( see eq .
( [ b.21b ] ) as a _ weighted average over an orthogonality variable _
@xmath433 : @xmath434 \bigg ) \label{4.21o } \\ & = & 2\int_0^\infty d\!x x\:e^{-x^2 } p_{11}^{(0)}(t,\delta_m(x))\ ; , \label{4.21a}\end{aligned}\ ] ] where @xmath435 is just the usual _ free _
2-level correlator ( eq.([b1.2 ] ) ) , but now with an an @xmath433-dependent tunneling amplitude @xmath436 ; and @xmath437 is the _
`` orthogonality exponent '' _ , defined by @xmath438 the orthogonality blocking term @xmath421 is analogous to the topological decoherence term @xmath361 .
it is important to understand why we must introduce the average over @xmath433 .
mathematically , it comes from the restriction to a single polarisation group ( see appendix a.2 ) .
physically , it corresponds to a phase average just like that in ( [ 4.8 ] ) ( compare the bessel functions ) , but now the phase is that accumulated _ between _ transitions of the central spin ( rather than _ during _ these transitions , as in topological decoherence ) .
this phase accumulates if @xmath439 , because then the field @xmath440 on the @xmath333-th bath spin does not exactly reverse when the central spin flips , and so this spin must start precessing in the new field .
it is random simply because the waiting time between flips is a random variable , in the path integral .
we shall not give full details for the dynamics of this limiting case ( for which see ref @xcite ) , but just enough to understand the physics .
first , note that the terms @xmath441 are easily verified to be incoherent , and so @xmath442 , where @xmath443 .
even the small fraction @xmath444 of systems in an ensemble having @xmath445 will only have coherent dynamics if @xmath446 . the easiest way to see
this is to again calculate the spectral absorbtion function from ( [ 4.21o ] ) , to get * ? ? ?
* : @xmath447 which leads , as @xmath437 increases , to an ever - increasing number of square - root singularities in @xmath448 . for @xmath449
only a single root @xmath450^{1/2}$ ] enters , and @xmath451}$ ] for @xmath452 , ie . , a fairly sharp asymmetric peak at the resonant frequency of the free spin .
for larger @xmath437 the multiple peaks and tails mix , and @xmath448 shows no obvious peak- moreover , its spectral weight is shifted to ever - lower frequencies : @xmath453\;\ ; \\ \delta_{eff } = 2 \delta_\phi e^{-\kappa } \end{array } \right\ } \;\ ; \kappa \ll 1 \label{4.22}\ ] ] @xmath454 notice that this reduction of the transition rate is much slower than the usual polaronic or anderson / hopfield orthogonality catastrophe , relevant to oscillator baths , which gives exponential suppression of @xmath455 for strong coupling .
this is understood as follows . in oscillator bath models ,
band narrowing comes essentially without any bath transitions ( most of the polaron `` cloud '' is in virtual high - frequency modes ) - it is adiabatic . here , however , roughly @xmath437 spins flip each time @xmath76 flips ( the probability of @xmath456 flips is @xmath457 , which peaks at @xmath458 ) , even though we only consider @xmath459 , i.e. , even though @xmath460 ( just as many bath spins flip one way as the other ) .
a further examination of the correlation function shows that the structure of @xmath459 is exceedingly bizarre- it was described in detail in ref .
@xcite . in section 4.a and
8 we will return to the physics of orthogonality blocking , but including other mechanisms as well ( see also fig . 7 below ) .
the previous 3 averages assume no intrinsic spin bath dynamics- the bath acquired its dynamics from the central spin .
consider now a hamiltonian @xmath461 in which we assume @xmath462 , but arbitrary @xmath463 .
the addition of @xmath178 , to what would have been a simple degeneracy blocking hamiltonian , gives the spin bath its dynamics , and causes 2 changes ( as noted previously in our introductory presentation of the central spin model ) .
first , a polarisation group @xmath214 acquires an `` intrinsic linewidth '' @xmath464 , where @xmath100 is a typical value of @xmath465 for the @xmath1 bath spins ( of course normally @xmath466 , unless the @xmath467 happen to be extremely tightly bunched together ) .
second , the transverse part of @xmath468 causes pairwise flipping amongst bath spins ( eg . , transitions @xmath469 ) , at a rate @xmath470 , where @xmath471 .
this `` spin diffusion '' in the bath causes the internal bias @xmath472 to fluctuate in time , inside the energy range of the polarisation group @xmath214 , with a random walk correlation @xmath473 ^ 2 \rangle = \lambda^3 \vert t - t ' \vert$ ] , where @xmath474 , for timescales @xmath475 .
it thus follows that for a given _ single _ central spin , with its surrounding spin bath in polarisation group @xmath214 , the problem reduces to calculating the dynamics of a 2-level system in a longitudinal bias field @xmath476 , where @xmath477 is the applied bias , and the internal field is @xmath478 .
the correlation properties of @xmath479 are those just described- our task is simply to functionally average over these fluctuations in the calculation of @xmath345 . in doing this
we will make the physically sensible assumption of `` fast diffusion '' of @xmath479 , such that the time @xmath480 it takes for the bias to diffuse across the `` resonance window '' , of energy width @xmath190 , satisfies @xmath481 .
then the system has no time to tunnel coherently , but can only make an incoherent `` landau - zener '' transition . since the bias changes by @xmath482 in a time @xmath483 , this formally requires that @xmath484 this problem is solved in appendix a , by performing a _ weighted average over dynamic bias fluctuations _ , with the restriction that these only occur inside polarisation group @xmath214 ; the relevant average is @xmath485 where @xmath486 is the correlator of the dynamic spin bath fluctuations ( see appendix a ) .
one finds that @xmath345 decays as a simple exponential @xmath487 , where @xmath488 this result is easily understood- the bias fluctuations can cause the system to pass briefly through resonance ( allowing a transition of the central spin ) , but only if the net static bias @xmath489 is not greater than the range @xmath490 of the fluctuations . by comparing with the case of pure degeneracy blocking
we see that the important role of the bath dynamics is ( i ) to _ unblock _ the central spin dynamics , by helping it to find resonance , now over an energy window of width @xmath490 ( instead of @xmath321 ) around zero bias ( recall fig .
4 ) , and ( ii ) to change the central spin dynamics from coherent to incoherent tunneling . note that in a model like ( [ hzener ] ) we have eliminated bath fluctuations between different polarisation groups .
the basic assumption is that any `` @xmath238 processes '' in the intrinsic bath dynamics , which could change @xmath214 in the absence of the central spin , are very slow ( at low @xmath0 , @xmath238 for nuclear or paramagnetic spins does become extremely long ) .
however this is not always realistic- we return to this point in section 5.a . if @xmath238 is short , one must make a dynamical average over 2 kinds of fluctuation , usually with quite different time correlation , viz .
, the intra - polarisation group fluctuations , described by ( [ t_2fl ] ) and with correlation time @xmath491 , and the inter - polarisation group fluctuations , occuring on a timescale @xmath238 ( cf .
@xcite ) .
we now turn to the problem of averaging over the spin bath for the general form of the central spin hamiltonian given in eqtn .
( [ 1.24 ] ) .
this can be given in the form of a marvellously simple prescription- one simply applies the 4 averages we have just seen , to the problem of a simple biased 2-level system !
we begin by giving the explicit prescription ( whose proof is given in appendix b ) , and make a few comments on it .
the prescription begins with the following 4 averages ( all of which we have seen in the preceding 4 sub - sections ) : @xmath492 @xmath493 @xmath494 @xmath495 as before , we assume a thermal distribution over spin bath biases , with a corresponding partition function and @xmath496 .
all averages are normalized to unity .
the weighting function @xmath497 in ( [ q.5 ] ) is a generalisation of that in ( [ 4.7 ] ) , to allow for arbitrary directions of the unit vector @xmath498 ; we now define @xmath499 now , suppose we want to calculate @xmath345 .
the prescription is fairly obvious in the light of the results given above for the 4 limiting cases .
one follows the following steps : + ( i ) begin with the quantity @xmath500 which is just the _ free _ central spin correlator ( cf .
( [ b1.10 ] ) ) having tunneling matrix element @xmath501 , and in an `` internal field '' bias @xmath472 .
the energy splitting @xmath502 is given by @xmath503 , and the matrix element @xmath504 is @xmath505 @xmath506 we defined @xmath437 previously ( eqtn .
( [ k.1 ] ) .
we will not give results for the case @xmath507 ; they are extremely complex , do not appear to add new physics , and seem unlikely to be realised in practice .
+ ( ii ) now carry out the averages over topological phase [ eq .
( [ q.5 ] ) ] and orthogonality [ eq . ( [ q.6 ] ) ] , to give an expression @xmath508 describing the central spin dynamics in a bias @xmath472 , coming from a bath in polarisation state @xmath214 : @xmath509 \ ; , \label{q.11}\ ] ] where the weighting function is @xmath497 over winding number @xmath510 ( recall eqtn .
( [ 4.7 ] ) ) ; + ( iii ) then , carry out the bias average [ eq .
( [ q.7 ] ) ] .
we will assume in the following for definiteness an ensemble average over all polarisation groups , thereby ensuring a summation over @xmath214 , to give @xmath511 this result summarizes the central spin dynamics in the case where the spin bath has no dynamics of its own , and only acquires dynamics through its interaction with the central system .
in some cases there will be no intrinsic bath dynamics , and this will be the final answer . if we wish to apply the theory to a single central system , or for some reason we can fix the polarisation group to be a definite value @xmath214 , then we drop the summation over @xmath214 in ( [ 6.25x ] ) , and replace @xmath218 by @xmath217 .
+ ( iv ) when the interaction term @xmath178 plays a role , we apply the 4th average ( [ q.7a ] ) to ( [ 6.25x ] ) , as described in appendix a ( cf . also the discussion in ref .
this gives the completely incoherent form @xmath512 \ ; , \label{diff.2}\ ] ] where the relaxation rate @xmath513 is given by @xmath514 with @xmath515 given by ( [ 5.8a ] ) , and where @xmath516 is a gaussian of width @xmath517 .
since @xmath518 we have also @xmath519 , and so we get @xmath520 this result is the most general one for the dynamics of the central spin , if all 4 bath averages are included- it is generally valid , with only the single restriction that the diffusion of the fluctuating bath bias in energy space be _ fast _ ( cf .
( [ fdiff ] ) . in the absence of such fluctuations
we go back to ( [ 6.25x ] ) . in essentially all physically realistic cases the different polarisation groups strongly overlap .
it is then simpler @xcite to transform the sum over @xmath214 in ( [ 6.25x ] ) or ( [ diff.3 ] ) into an integral over energy bias @xmath477 , using the change of variables @xmath521 , and then integrate over @xmath477 . one way to do this ( using steepest descents )
was detailed in ref .
@xcite ( compare eqtns ( 4.41)-(4.47 ) in that paper ) . for exact answers one can use the identity @xmath522 to evaluate either ( [ 6.25x ] ) or ( [ diff.2 ] ) . in the next section
we will evaluate @xmath345 and its fourier transform for a number of different parameter regimes .
but even before doing the integrals , the qualitative behaviour is obvious .
relaxation is only occurring for central spins which happen to be within a bias @xmath523 of exact resonance .
the width @xmath523 of this `` resonance window '' is coming from the energy which the bath spins can provide to the central spin , by flipping up to @xmath524 bath spins ; hence @xmath525 ( formally this is obvious from the properties of the bessel functions in ( [ 5.8a ] ) and ( [ bessel ] ) , which fall off very fast once @xmath526 ) .
we show graphically the relaxation of different groups in fig .
7(a ) ; again one sees how only groups with @xmath527 relax quickly .
the @xmath237 bath fluctuations help this process by bringing the a central spin in polarisation group @xmath214 to its resonance window ( of width @xmath504 ) .
only transitions of systems having @xmath445 can show ( partial ) coherence ; all transitions with @xmath528 are essentially _
incoherent_. note that the resonance window will _ not _ be visible in a resonant absorption experiment ( fig .
7(b ) ) ; higher @xmath214 groups contribute only a very low frequency contribution to this .
this nicely demontrates that one is very far from any linear - response regime in the present system ( so that , e.g. , the fluctuation - dissipation theorem is somewhat irrelevant here ) . without the bath
, transitions of the central spin would be coherent , but over the far smaller resonance window of width @xmath529 . if we only had @xmath237 bath fluctuations , but @xmath530 ( ie .
, no bath spins flipped during the transitions of the central spin ) , then we would again get incoherent relaxation , but this time with @xmath531 ( the width of the @xmath445 polarisation group ) .
one can also imagine a situation in which @xmath238 is very short , so that all polarisation groups are involved in the relaxation , and the resonance window is just the distribution @xmath532 , with @xmath533 .
these results thus tell us that in the presence of a spin bath , any ensemble of central spins , initially spread over a range of biases , will start relaxing by digging a `` hole '' of width @xmath523 around zero bias @xcite .
this hole reflects the _ intrinsic _ central spin dynamics ( ie .
, it is not being produced by interaction with some external resonant signal- it should not be confused in any way with the `` spectral hole - burning '' done by experimenters working on glasses or in optics , using an external source ) . as discussed in a number of papers @xcite , evaluation of ( [ diff.2 ] ) , using either steepest descents or other means , shows that the system under most conditions relaxes incoherently with a relaxation rate ( after summing over all polarisation groups , and doing the orthogonality and phase integrals ) given approximately as a function of bias @xmath477 by @xmath534 all of this is in complete contrast to how inelastic tunneling works in the presence of an oscillator bath @xcite ; there the relaxation rate typically _ increases _ as one moves away from resonance , usually as a power in bias ( @xmath535 for diagonal coupling to phonons , @xmath536 for non - diagonal coupling to phonons , and @xmath537 for diagonal coupling to ohmic baths like electrons via a dimensionless coupling @xmath538 ) .
thus one does not expect hole - digging for oscillator bath - mediated quantum relaxation , except over a very narrow region of width @xmath539 .
finally , we note that one may also give a formal prescription for the case where some central spin couples simultaneously to an oscillator bath _ and _ a spin bath .
we do not give the details here- they are discussed fairly exhaustively ( along with the results for the central spin dynamics ) in ref .
@xcite ( see also end of section 4 ) .
given the large number of parameters entering into the 4 averages just described , we see little point in an exhaustive description of @xmath345 over the whole parameter domain ( for more extensive results see refs .
instead we concentrate on 3 points .
first , we show how in the strong coupling regime , coherence is destroyed , leaving incoherent quantum relaxation ; this regime applies to almost all mesoscopic or nanoscopic magnetic systems , because of their coupling to nuclear spins and to paramagnetic impurities .
second , we discuss the physics of the weak coupling regime ( applicable to , eg . ,
squids ) , and how in one limit of this regime one may formally map the spin bath to an oscillator bath . finally , and very briefly , we comment on the results obtained when one couples simultaneously to a spin bath and an oscillator bath .
as already explained , the strong - coupling regime is defined by the condition @xmath183 and/or @xmath540 .
this condition applies to virtually all situations in which the couplings are hyperfine ones to nuclear spins , or exchange couplings to paramagnetic spins ; and also when one has dipolar couplings to paramagnetic impurities or defects .
almost all interesting physical examples in this regime fall either into the catagory of strong orthogonality blocking " ( when @xmath541 ) or strong phase decoherence `` ( when @xmath542 )
. in both cases the central system makes transitions accompanied by flips in the bath spins- so that even if the isolated central system is not in resonance , it can ' ' find resonance " by using the flipped bath spins to make up the energy difference . if the couplings are such that roughly @xmath437 bath spins flip , the range of energy bias over which transtions can occur is extended to roughly @xmath543 .
the central system is helped in this task by the fluctuations in bath bias caused by the interspin interactions @xmath178 . in what follows we concentrate on the physics of _ decoherence _ in this regime , with an eye to the physics of `` qubits '' and of `` macroscopic quantum coherence '' .
we also look at the form of the relaxation .
we begin by explaining the results without the bath fluctuations , and then show what happens on adding these . * ( i ) results without bath fluctuations * : in this case we must evaluate ( [ 6.25x ] ) and ( [ q.11 ] ) , suppressing either the orthogonality average or the topological average . in what follows
we look at each case in turn , focussing particularly on the @xmath445 polarisation group contribution . _
( a ) orthogonality blocked regime ( @xmath544)_. in this regime only the orthogonality average @xmath545 , and the average over bias @xmath472 , are relevant- the phase average is approximated by a delta - function .
the presence of the @xmath433-dependent transition matrix element @xmath546 means that polarisation groups with @xmath547 play a dominant role .
suppose however that we are interested in any coherent dynamics of the central spin- what will be found ?
it is obvious that transitions with finite @xmath214 will be essentially incoherent , so we concentrate on central spins for which @xmath445 .
thus we simply integrate @xmath548 over @xmath549 , in the weighted bias average , to get @xmath550 \label{5.18 } \\ & = & 1- 2 \int dx xe^{-x^2 } 2a(x ) \sum_{k=0}^{\infty } j_{2k+1 } \big [ 2 \delta_0 ( x ) t \big ] \ ; , \label{5.18b}\end{aligned}\ ] ] where the @xmath433-dependent spectral weight is @xmath551 .
notice we have just done an `` orthogonality average '' over a `` biased averaged '' expression for the free system with @xmath433-dependent tunneling frequency @xmath552 .
a fourier transform to frequency space ( which is essentially a picture of the relaxation rate , as a function of _ energy bias _
@xmath477 for this system ) gives the absorption spectrum @xmath553^{1/2 } } \eta ( \omega - 2 \mid \delta_o ( x ) \mid ) \ ; , \label{5.19}\ ] ] fig .
8 shows some representative plots for this `` coherent '' part of @xmath554 ; it is in fact _ almost completely incoherent _ , with total spectral weight @xmath555 a result which is very accurate even for @xmath556 .
note that the shape of @xmath557 will change once we include all other ( incoherently relaxing ) polarization sectors @xmath558 , and its total weight increases - in fact the total weight is @xmath559 for large @xmath437 , since @xmath560 different polarization sectors contribute .
the absorption in @xmath554 from these higher @xmath214 groups will be concentrated at frequencies @xmath561 ( compare also fig 7(b ) ) .
note however that relaxation itself ( not described by the linear response function @xmath562 ) will be spread incoherently over a frequency range @xmath563 , ie .
, the `` hole - digging '' in the relaxation occurs over a window of width @xmath564 .
_ ( b ) phase decoherence regime ( @xmath565)_. let us now suppose the transverse field terms @xmath566 are negligible compared to the @xmath567 .
we then deal with an effective hamiltonian @xmath568 + h.c . \right\ } + { \hat \tau } _ z \sum_{k=1}^n \omega_k^{\parallel } \ : { \hat \sigma } _
k^z \;. \label{5n.12}\ ] ] since @xmath569 by assumption , energy conservation requires that environmental spins flip with the central spin , just as in our discussion of pure orthogonality blocking .
thus in this case _ we must also keep the orthogonality average _ , to enforce this constraint , ie .
, we must perform the full average embodied in eqtns .
( [ q.11]),([6.25x ] ) . the full answer , including both the real and imaginary parts of @xmath361 , is rather complicated , and is presented in appendix b. here we will consider the more transparent answer one gets when @xmath361 is purely imaginary and adds to directly as a random variable to the central spin phase .
let us again start with only the @xmath445 contribution to @xmath345 .
then we have @xmath570 with @xmath571 given by ( [ 6.25x ] ) with @xmath445 .
we may now carry out the integration in ( [ 5n.13 ] ) , assuming that @xmath572 is given by the usual gaussian form ( [ gauss ] ) , to get @xmath573 \ ; ; \label{6.25z}\ ] ] with @xmath574 .
the corresponding absorption @xmath554 is @xmath575^{1/2 } } \;. \label{6.27z}\ ] ] it is possible to write analytic expressions starting from ( [ 6.27z ] ) , but in this somewhat pedagogical presentation we simply discuss the case when @xmath576 , i.e. , zero degeneracy blocking , when @xmath577 for all nuclei .
the integration over bias is then absent ( since @xmath572 is now just a set of @xmath578-function peaks , ie . , @xmath579 ) , and we get @xmath580 where @xmath581 now describes the dynamics in zero bias ; it is given by exactly the same weighted average over phase as in ( [ 4.8 ] ) : @xmath582 \right\ } \label{5.6a } \\ & = & \int dx x e^{-x^2 } \left\ { 1 + \sum_{m=-\infty } ^{\infty } ( -1)^{m } f_{\lambda^\prime}(m ) e^{i2m\phi } j_{2 m } [ 4\delta_o t j_m(2x\sqrt{\lambda } ) ] \right\ } \ ; , \label{5.6}\end{aligned}\ ] ] this can be interpreted _ either _ as an orthogonality - blocked expression , with frequency scale @xmath583 which is then _ averaged over @xmath584 _ , to give phase randomisation ; _ or _ as an integration @xmath585 over an _ already topologically decohered _
function having frequency scale @xmath586 .
it is intuitively obvious ( and easily demonstrated ) that only @xmath587 may behave coherently , with a fractional weight @xmath588 in an ensemble .
there are various interesting cases of ( [ 5.6 ] ) for @xmath445 .
if @xmath589 ( i.e. , @xmath590 is parallel to @xmath591 ) , then we go back to pure topological decoherence - the projection operator then commutes with the cosine operator . on the other hand if @xmath592 , we have pure orthogonality blocking as stated earlier , and in fact when @xmath592 , the parameter @xmath593 plays the role of @xmath437 in ( [ 4.21o ] ) .
notice that whereas the case @xmath594 can only occur accidentally , @xmath592 is quite common - indeed it pertains to the model in eqs .
( [ 2.8 ] ) .
we really begin to see the analogy between orthogonality blocking and topological decoherence when @xmath595 ; just as with pure topological decoherence , @xmath596 collapses to a kronecker delta , and we get the _ universal _ projected topological decoherence form : @xmath597 \biggr ] \;\;\;\;(\mu = 0)\ ; ; \label{5.9}\ ] ] @xmath598^{1/2}}\ : \eta ( 4 \delta_o \mid j_0(2x\sqrt{\lambda } ) \mid -\omega ) \ ; , \label{5.10}\ ] ] which generalizes the result of ( [ 4.13a ] ) for pure topological decoherence .
eq.([5.9 ] ) should be compared to ( [ 4.21o ] ) .
we show in fig.9 some results for @xmath599 for selected values of @xmath593 .
the results are startling ; even a very small value of @xmath593 significantly washes out pure topological decoherence ; but for any large value of @xmath600 , we never get back the pure orthogonality blocking spectrum .
the results in the case where @xmath361 is real , and @xmath601 is along the @xmath602-direction ( see eq .
( [ z.29 ] ) ) are obtained by simply converting the bessel functions @xmath603 to bessel functions of imaginary argument @xmath604 . finally , note
again that the above discussion of the @xmath445 polarisation group is irrelevant to the real experimental lineshape- an evaluation of the full expression ( [ 6.25x ] ) , summing over all @xmath214 , for the @xmath605 regime , simply gives incoherent relaxation @xcite , spread over a frequency range @xmath606 . *
( ii ) including bath fluctuations : * the modification of the above results , occasioned by the intrinsic spin bath fluctuations , was given in detail in prokofev and stamp @xcite .
the fluctuations in bias allow the central system to cycle rapidly through the whole range of biases within a given polarisation group ( transitions between _ different _ polarisation groups can occur through @xmath238 processes- usually much slower ) . here
we simply recall the main result , which is obtained by summing the relaxation forms from each polarisation group in an ensemble ( cf .
( [ 6.25x ] ) and ( [ diff.2 ] ) ) , and assuming that the spin bath @xmath238 is longer than all experimental times scales . for a single central system ,
coupled to a spin bath in equilibrium at temperature @xmath607 , one finds that after an initial short - time transient , the relaxation is roughly logarithmic over a very long period ; in fact one finds for the strong coupling regime that @xmath608 } \ ; ; \;\;\;\;\;\;\;\;\;\;\;(t\ll t_c ) \ ; , \label{jltp.446}\ ] ] for times @xmath609 , where @xmath610 , and @xmath611 is the relaxation time of the @xmath445 polarisation group ( compare eqtn .
( [ diff.3 ] ) ) ; thus @xmath612 is extremely long ! for @xmath613
the system settles down to a rather different behaviour @xcite .
this logarithmic behaviour can be _ roughly _ understood @xcite as coming from a distribution of barrier heights , for the different polarisation groups , which are then summed over- as discussed in ref .
@xcite , section 4.3(b ) , the final result looks basically the same as that shown in fig .
the fastest relaxation comes from the those polarisation groups in the `` resonance window '' ( recall the discussion at the end of section 3 ) .
two cautionary notes are in order here .
first , ( [ jltp.446 ] ) applies to a _ single _ relaxing system- but in the case of nanomagnetic systems , all experiments until now
have been done on large numbers of nanomagnets , coupled together via long - range dipolar forces , which drastically changes the relaxation ( see section 5.a below ) .
second , ( [ jltp.446 ] ) should not be applied uncritically to experiments , even on single quantum systems .
this is because in a real experiment there will also be ( i ) couplings to oscillator baths , and ( ii ) the relaxation will change once @xmath614 . in superconductors or metals , electronic oscillator baths will often dominate the relaxation @xcite , even at short times ( for their effect on _ coherence _ , see section 5.c below ) . even in insulating systems ,
one eventually expects the coupling to phonons to take over at long times @xcite , since this causes exponential decay- even if very slow , this will eventually become faster than the spin bath - mediated logarithmic decay in ( [ jltp.446 ] ) .
a question of considerable theoretical ( and practical ) interest is the transition to the weak coupling regime , where the perturbation on the central system dynamics by a single bath spin is small ( even though the net effect of all bath spins , measured by parameters like @xmath199 or @xmath437 , may still be large if @xmath1 is very large ) .
the weak - coupling regime is thus defined by the condition @xmath615 , ie . , both @xmath184 _ and _ @xmath183 are @xmath616 . again , a variety of cases is possible depending on how large are ratios like @xmath617 and @xmath618 , or parameters like @xmath199 and @xmath437 . in the following
we will not be exhaustive , but simply consider two theoretically interesting cases , in which @xmath178 is assumed negligible , and we look at the limiting behaviour arising when either @xmath619 or @xmath620 .
we will also assume @xmath40 , otherwise the problem is trivial ( the bath has little effect at all ) .
there are 2 ways to solve for the dynamics in this regime .
one is to use the averaging already developed above- this simplifies considerably in the weak- coupling regime .
the other is to map the problem onto an oscillator bath one , and then use standard techniques to solve this .
we demonstrate the 2 methods by solving one problem with each .
\(i ) _ longitudinally dominated case _
( @xmath619 ) : we assume the same hamiltonian as in the discussion of the phase decoherence regime ( eqtn .
( [ 5n.12 ] ) ) , but now we can drop the orthogonality average- the projection to a single polarisation group is not required since @xmath621 . again , for simplicity we consider the case when @xmath622 is imaginary .
we shall solve this using the techniques previously developed ; we let @xmath623 in ( [ 5.8a ] ) , and hence use a matrix element @xmath624 ( which is _ independent _ of @xmath214 ) .
all polarisation groups overlap , and so we simply average over bias and topological phase : @xmath625 \bigg ) \right\ } \nonumber \\ & = & 1- \sum_{m=-\infty } ^{\infty } f_\lambda(m ) \int { d\varphi \over 2 \pi } e^{i2m(\phi -\varphi ) } 2 a ( \varphi ) \sum_{k=0}^{\infty } j_{2k+1 } [ 2 \delta_0 ( \varphi ) t ] \ ; , \label{u.2}\end{aligned}\ ] ] with @xmath626 as before , and @xmath627 .
this gives an absorption form @xmath628^{1/2 } } \eta ( \omega /4\delta_o - \mid \cos \varphi
\mid ) \ ; , \label{u.3}\ ] ] which for large @xmath593 simplifies to @xmath629^{1/2 } } \eta ( \omega/ 4 \delta_o - \mid \cos \varphi \mid ) \ ; , \label{u.5}\ ] ] and can be expressed in terms of elliptic functions .
notice that since @xmath630 , the number @xmath1 of environmental spins must be very large to have a noticeable effect , i.e. , for @xmath199 to be appreciable . thus if @xmath631 , since @xmath632 , we have @xmath633 .
this not only implies that @xmath634 ( ie .
, very strongly overlapping polarisation groups ) but also a gaussian half - width @xmath635 , so that the internal bias @xmath636 .
the reason why the two mechanisms ( topological decoherence and degeneracy blocking ) are so easily combined is just because this bias is produced by _ all _ of the environmental spins , whereas only a few of them are actually flipped and their contribution to the bias field is small .
\(ii ) _ transverse dominated case _ ( @xmath620 ) : consider now the case described by @xmath637 with @xmath638 .
one can map this problem to the spin - boson problem , in a way similar to that given in the paper by caldeira et al .
we begin by noting that the time evolution operator for the @xmath333-th bath spin , under the influence of a central spin moving along a path @xmath639 , is just @xmath640 \right\ } \ ; , \label{neto.3}\ ] ] weak coupling means we can expand the time - ordered exponent to second order in @xmath641 ; completing the average in eq .
( [ neto.22 ] ) , and exponentiating the answer , one derives the influence functional for this problem @xcite .
the `` high temperature '' result @xmath642 is readily found : @xmath331 = \exp \bigg\ { -\sum_k { ( \omega^{\parallel}_k)^2 \over 2 } \bigg\vert \int_0^t ds e^{2i\omega^{\perp}_k s } ~[q_{(n)}(s)-q'_{(m)}(s ) ] \bigg\vert ^2 \bigg\ } \ ; , \label{neto.4}\ ] ] the evaluation of this depends on the distribution of couplings . in the `` strong decoherence '' case where the bath states are spread over an energy range @xmath216 ( defined as usual by @xmath643 ) which is large ( ie . , @xmath644 ) , then @xmath336 $ ] is simply approximated by its form for a single kink - anti - kink trajectory .
with kink at @xmath645 and anti - kink at @xmath646 , one only has a contribution if @xmath647 , which gives @xmath336 = e^{- { 1 \over 4}e_o^2(t_2-t_1)^2}$ ] , ie .
, a decay on a time scale @xmath648 .
it then follows that kink/ anti - kink transitions are bound in closed pairs , with phase correlations decaying over a time @xmath649 , and the leading terms sum to give exponential relaxation , ie . , @xmath650 , with a relaxation rate @xmath651 on the other hand if @xmath652 , we will get coherent oscillations of the system over a time scale @xmath653 , ie .
, over roughly @xmath654 oscillation periods .
we will use this result below , in discussing the possibility of seeing true mesoscopic quantum coherence effects in squids and nanomagnets ( section 5.c ) .
it is of course obvious from the original discussion of feynman and vernon that in the weak - coupling regime , where an expansion to 2nd order in all bath couplings is sufficient , a mapping to an oscillator bath model must be possible @xcite .
the method given above is of course not the only one- others are explained in refs .
@xcite . * central spin coupled to oscillator and spin baths * : for reference we also briefly note results for the case where a central spin is coupled simultaneously to a spin bath and an oscillator bath ; the hamiltonian is that in eqtn .
( [ qsgl ] ) , but without the sum over @xmath655 , ( and of course dropping the coupling term @xmath656 between different central spins ) .
some of the dynamic properties of this model were studied in ref .
@xcite- a complete study is still lacking .
quite generally one expects that at very short times the dynamics will be controlled by the spin bath , but at longer times incoherent oscillator bath- mediated transitions will take over .
however these transitions still contain spin bath effects , via an integration over spin bath bias distribution , to give a result @xmath657 e^{-t/\tau(\epsilon , t ) } \ } \label{csosc}\ ] ] where we assume that the ensemble of central spins is in equilibrium with a bath of oscillators at temperature @xmath390 , in external bias @xmath477 ( for more general cases see ref .
@xcite ) . in this case @xmath658 , and @xmath659 is the oscillator bath- mediated relaxation time ; @xmath660 is a `` renormalisation '' constant @xmath661 coming from the spin bath dynamics .
most features of the results are obvious from ( [ csosc ] ) .
thus , the oscillator bath _ unblocks _ transitions for a central spin way off resonance ( ie . , having @xmath662 ) ; these spins relax much as they would without the spin bath . when @xmath663 there is a wide range of relaxation times , and the physical relaxation depends in a complex way on the oscillator bath spectrum as well as the bias . in any experiment
one would see a crossover between purely spin - bath mediated relaxation at short times and this more complex behaviour- this crossover has not yet been studied in any detail .
in this section we briefly review some recent work , mainly experimental , in which interactions with a spin bath play a role .
we begin with magnetic systems where such interactions are strong , and where clear evidence exists for their influence on tunneling phenomena .
we then discuss superconducting and normal systems .
finally , we discuss the difficult and controversial problem of decoherence and the mechanisms which govern it in nature . in this discussion
the results of the central spin model are crucial- they describe the effect of a spin bath on a `` qubit '' , or on a squid , or a magnetic macromolecule , which is trying to show coherent oscillations .
\(i ) _ magnetic solitons _ : magnetically ordered systems support a wide variety of soliton excitations , depending on the symmetry of the order parameter . these couple to various environmental excitations , which strongly affect their dynamics .
these include both linear and non - linear couplings to quasiparticle " excitations such as magnons @xcite , phonons @xcite , electrons @xcite , and photons @xcite , all oscillator baths .
more serious effects come from any localized modes coupling to the soliton , most notably paramagnetic impurities and nuclear spins .
both of these can be understood theoretically using a model in which a moving particle couples to a spin bath @xcite , as in section 2.c .
most experimental work in this area has looked at domain wall tunneling in ferromagnets .
early experiments looked at the dynamics of multi - wall systems @xcite .
more recently attention has focussed on the tunneling of _ single _ domain walls , whose position can be monitored in various ways @xcite .
these walls propagate along magnetic nanowires ( of diameter 300 - 600 )and are thus mesoscopic objects ( the material of choice is usually ni , for which the domain wall thickness @xmath664 ) . although some predictions of the tunneling theory have been verified ( eg . , the square root dependence @xcite of the tunneling exponent on the pinning field @xmath665 ; cf .
@xcite ) , the crossover temperatures to tunneling are usually much higher than calculated values @xcite . a clue to the reason for this
may be found in the microwave resonance experiments of hong & giordano @xcite , which show extremely broad resonances , even down to @xmath666 .
the coupling to phonons or electrons is far too small to give this ; however , oxygen impurities , which will act as paramagnetic impurities , would act as a strong time - varying potential on any wall , and broaden the line @xcite .
it is clear that much more experimental work will be necessary to really understand the quantum dynamics of magnetic domain walls .
\(ii ) _ magnetic macromolecules _ :
on the other hand , a number of spectacular quantum effects have unquestionably been seen in crystalline arrays of magnetic macromolecules , in one of the most important developments in magnetism in recent years .
experiments on tunneling phenomena have had striking success in the `` mn-12 '' and `` fe-8 '' molecules ( each of which behaves at low @xmath0 as a `` giant spin '' of spin 10 ) .
there have also been highly - publicised experiments on the large `` ferritin '' molecule , which claim the observation of `` macroscopic quantum coherence '' ; these are discussed in section 5.c below .
the early molecular work demonstrated resonant tunneling ( coming when spin levels of each molecule are brought into resonance ) , at relatively high temperatures @xcite .
more recent experiments @xcite have gone into the quantum regime , in which only the 2 lowest levels of each giant spin are occupied .
the classical - quantum crossover is clearest in the case of the fe-8 system ( which is very conveniently described by the easy axis / easy plane model discussed in section 2.f ) ; below a temperature @xmath667 , the dynamics is completely independent of @xmath0 ( the energy gap to the next spin level in this system is @xmath668 ) . in mn-12
there are effects arising from `` rogue '' molecules which make things more complicated @xcite . at such low temperatures phonons are utterly irrelevant ( experiments have now been pursued to @xmath669 ) , and the only dynamic environment left is the nuclear spin bath .
the system thus seems at first to be an ideal realisation of the central spin model discussed in sections 3 and 4 .
the main complication is that experiments are done on a crystalline array of molecules , which interact via strong magnetic dipolar interactions .
however , once one knows that the dynamics of a single central spin is going to be incoherent relaxation ( cf . sections 3.e
, 4.a ) , these interactions are rather easy to deal with @xcite .
the system is then described by an obvious generalisation of the central spin hamiltonian , viz .
@xmath670 in zero applied field ( compare eqtn .
( [ qsgl ] ) , without the oscillators ) . here
@xmath671 is the central spin hamiltonian ( [ 1.24 ] ) for the molecule at lattice position @xmath672 , and @xmath673 is the magnetic dipolar coupling between molecules @xmath674 and @xmath655 . to solve for the dynamics of such an interacting array of systems ,
one begins @xcite by defining a distribution function @xmath675 for a molecule at position @xmath676 to be in a longitudinal bias @xmath677 , with polarisation @xmath678 .
it is then trivial to write down a bbgky - like hierarchy of kinetic equations for @xmath679 and its multimolecular generalisations @xmath680 , and @xmath681 , etc .
, of which the first member is @xmath682 \nonumber \\ & - & \sum_{\alpha ' } \int { d\vec{r}\ : ' \over
\omega_0 } \int { d\xi ' \over \tau_n ( \xi ' ) } \bigg [ p_{\alpha \alpha ' } ^{(2 ) } ( \xi , \xi ' ; \vec{r},\vec{r}\ : ' ) - p_{\alpha \alpha ' } ^{(2 ) } ( \xi -\alpha \alpha ' v(\vec{r } -\vec{r}\ : ' ) , \xi ' ; \vec{r},\vec{r}\ : ' ) \bigg ] \ ; , \label{kinetic}\end{aligned}\ ] ] in which relaxation is driven by the nuclear spin - mediated relaxation rate @xmath683 ( cf eqtn ( [ xi_o ] ) , in conjunction with the dipolar interactions . under general conditions we must also solve for @xmath684 in terms of @xmath685 , etc
but if the initial experimental state is either _ polarised _ or _ annealed _
then @xmath684 factorises at @xmath348 ( ie . , @xmath686 ) , and ( [ kinetic ] ) can be solved .
this led to the following predictions @xcite : \i ) relaxation should only occur for molecules having @xmath687 ; consequently a hole " rapidly appears in @xmath688 with time , with initial width @xmath523 determined entirely by the nuclear dynamics ( and of course typically @xmath689 ) .
in fact microscopic calculations of @xmath523 for the @xmath136 - 8 molecule @xcite can be done ( recall fig .
5(b ) ) , since the hyperfine couplings are essentially dipolar ( the relevant nuclei include 120 protons , 8 @xmath690 nuclei , and 18 @xmath1 nuclei ) .
one finds that even the weak molecular dipolar fields strongly distort the hyperfine fields , mixing up the different polarisation groups , and also giving a large value of @xmath437 . the final value @xmath523 will obviously depend sensitively on any nuclear isotopic substitution @xcite .
the subsequent evolution of the hole depends on sample shape ; this has been studied theoretically using monte carlo simulations @xcite .
\(ii ) the short - time relaxation for the total magnetisation @xmath691 should have a `` square - root '' time form @xmath692 \ ; , \label{msqrt}\ ] ] where @xmath693 is the initial magnetization ( or with appropriate modifications for other protocols , such as a zero - field cooling followed by relaxation in a field @xcite ) , and @xmath694 , where @xmath695 is the `` density of states '' of the distribution .
the constant @xmath696 is dimensionless , depending on sample shape , and can be evaluated analytically or numerically .
notice that the existence of the square root does not depend on sample shape , and indeed its persistence over fractional relaxations of @xmath697 is clearly demonstrated in monte carlo simulations for different shapes @xcite .
this result implies that by varying @xmath698 , one can measure @xmath699 , by extracting @xmath700 at successive values of @xmath698 .
if one knows @xmath701 ( as one does in an annealed sample - it should be gaussian ) then one may then extract @xmath702 from measurements of @xmath703 .
the @xmath704 dependence of @xmath705 has since been reported in quite a few experiments , on both fe(8 ) crystals @xcite and mn(12 ) crystals @xcite ( although the situation in mn-12 is seriously complicated by `` impurities '' @xcite ) .
the fe-8 experiments have produced remarkable `` maps '' of @xmath701 , and its time variation @xcite .
the `` hole - digging '' has been found in both fe-8 and mn-12 @xcite , with an `` intrinsic '' short - time intrinsic linewidth which is roughly that expected from the hyperfine interactions , provided one takes account of the effect of internal fields @xcite , which make @xmath238 very short ( so that @xmath437 is effectively @xmath706 ) .
wernsdorfer et al @xcite also used this technique to extract the value of @xmath190 for fe-8 . in a remarkable experiment , wernsdorfer and
sessoli @xcite extended these measurements of @xmath702 to include a _ transverse _
applied field @xmath707 ; as noted in section 2.f , the topological giant spin phase @xmath176 should vary with field @xmath708 @xcite , producing aharonov - bohm oscillations in @xmath709 .
these oscillations were found , both as oscillations in the relaxation rate @xmath710 , _ and _ using a quite different ac absorption ( `` landau - zener '' ) technique ; these independent techniques agreed rather well in the measurement of @xmath190 .
notice , incidentally , that neither method can properly measure @xmath702 near its nodes ( ie . , where @xmath711 ) ; this is because of the distribution of internal transverse fields .
incidentally , we should strongly emphasize that these experiments ( even the aharonov - bohm ones ) do _ not _ demonstrate coherent tunneling- indeed they show exactly the opposite !
this is because the experiments are inherently _ relaxational _
( this is why all rates are @xmath712 , and not @xmath713 ) .
readers puzzled about how an aharonov - bohm effect can occur in a relaxation rate , are encouraged to think about other examples in physics where phase interference shows up in irreversible quantum phenomena .
in the present case we may loosely define a `` decoherence time '' @xmath714 for the molecular spins ( one should not push this too far , given the non - exponential nature of the central spin relaxation ! ) , and one finds that @xmath715 ( no coherent oscillations possible ) but @xmath716 ( ie . , coherence is maintained during a single very rapid tunneling transition ) . in a very recent experiment @xcite , deliberate modification of the nuclear isotopes in an @xmath136 - 8 crystal
has shown a dramatic modification of the bulk magnetic relaxation- this is the most direct evidence so far for the role of the nuclear spins in mediating the quantum relaxation .
it will be interesting to see a quantitative comparison with the calculated results for different isotopes @xcite .
\(iii ) _ quantum spin glasses _ : the work on nanomagnets probes our understanding of collective phenomena in many magnetic systems , ranging from spin chains to quantum spin glasses .
this latter example is very closely related to the molecular nanomagnets just described , since the model hamiltonian is just ( [ hsqrt ] ) , with the dipolar interaction replaced by a set @xmath717 of `` frustrating '' interactions , usually long - ranged .
a typical experimental example is provided by the disordered dipolar spin system liho@xmath718y@xmath719f@xmath720 ( where the ho moments interact via dipolar interactions @xcite , and the @xmath702 are induced by a transverse field ) . until now most theory has ignored the environment in this problem , but from our discussion above , this is clearly a mistake if one wishes to discuss dynamics .
this is also the view expressed in recent papers of cugliandolo et al .
@xcite , who have included coupling to an oscillator bath environment .
this is presumably correct in metallic glasses , where the coupling to electrons dominates ( compare the discussion following ( [ 1.35 ] ) ) , but in insulators the coupling to nuclear spins dominates ( in li@xmath719ho@xmath718f@xmath720 this is clear- the hyperfine coupling even modifies the phase diagram ! ) .
an observation of strong hole - digging in @xmath721 in a spin glass would be consistent with this , since oscillator baths typically give most rapid relaxation _
away _ from resonance ( @xmath722 discussion after eqtn .
( [ xi_o ] ) .
note also that one of the main theoretical questions concerning quantum spin glasses , viz .
, what happens when @xmath723 , could be examined experimentally using molecular magnets like fe-8 , or the liho@xmath718y@xmath719f@xmath720 system , in a strong transverse field , where one might reasonably expect to see coherent propagation of tunneling events from one site to another if @xmath190 is sufficiently large . in mesoscopic conductors the standard weak localisation theory @xcite evaluates a `` decoherence time '' @xmath724 in terms of the electron - electron scattering rate , which is itself strongly influenced at low @xmath0 by elastic impurity scattering .
this is essentially an oscillator bath problem- the relevant oscillators being diffusons , cooperons , and phonons .
curiously , given the known importance of scattering off dynamic `` 2-level '' fluctuators in these systems @xcite , there has been little theory on the effect of these on @xmath714 , apart from the pioneering work of altshuler and spivak , and feng et al .
this is of course a spin bath problem , with the spins representing defects , paramagnetic impurities , etc .
, in the environment . it will be interesting to see if the `` saturation '' in @xmath724 reported at low @xmath0 @xcite may at least be partially explained by such scattering .
the physics of this saturation depends on the more fundamental question of how decoherence behaves in the low-@xmath0 limit in conductors , and has caused considerable debate in the recent theoretical literature @xcite . in the case of superconductors
the situation is similar , in that almost all theoretical work on dissipation and other environmental effects has looked at the effects of electronic quasiparticle modes or photons , ie .
, delocalised modes which can be mapped directly to oscillator modes .
the pioneering papers @xcite led to a massive subsequent literature , both experimental @xcite and theoretical @xcite . in most work on tunneling
there is no question that theory and experiment correspond very well @xcite .
however the situation is more delicate for coherence , discussed below . in spite of the extensive theory of spin impurity effects in superconductors we are aware of only 2 theoretical papers @xcite ( and no experiments ) examining their effect on the flux dynamics ( in particular tunneling ) of squids .
an understanding of decoherence mechanisms is central to the exploitation of mesoscopic systems in quantum devices , as well as to general questions about how quantum mechanics applies on the large scale , @xcite , and the quantum measurement problem @xcite .
it has become particularly important now that efforts are being made to construct `` qubit '' devices , with a view to making quantum computers .
the last 15 years have seen a total transformation in how such questions are discussed- instead of vague analyses in terms of `` measurements '' by the environment , we now have precise and generally applicable models , which can be tested in many experiments .
the basic issues are ( i ) whether phase coherence can be preserved in the reduced density matrix of the system of interest and ( ii ) what are the decoherence mechanisms destroying it . here
we briefly review studies of superconducting and magnetic systems , and then examine things from a more general theoretical standpoint .
\(i ) _ decoherence in superconductors _ : this has been discussed intensively ever since the theoretical predictions of leggett et al .
@xcite concerning `` macroscopic quantum coherence '' in squids , and subsequent proposals for experimental searches @xcite .
to date no experimental success has been reported ( although there is good evidence for resonant `` one passage '' tunneling transitions between near degenerate levels in 2 wells @xcite ) .
almost all microscopic analyses of this problem have assumed environments of electronic excitations which can be mapped onto oscillator baths ( see , eg . , @xcite ) .
in our opinion , as discussed in section 4.b , the basic problem is simply that the main source of decoherence in most systems ( including squids ) at low @xmath0 will not be any oscillator bath , but the spin bath of paramagnetic and nuclear spins . as discussed in section 2.g and 4.b , the low - energy scale of this spin bath means it will not usually have a big effect on squid tunneling , but its effect on macroscopic coherence or on superconducting qubits will be rather large .
although a superconducting qubit has not yet been built , experiments may be getting rather close @xcite . to see how big spin bath effects on coherence might be , let us recall that the effect of paramagnetic impurities is to create a gaussian multiplet of spin bath states of width @xmath725 in energy for a squid containing a total of @xmath726 paramagnetic impurities interacting with the supercurrent , where @xmath727 is the change in field on each paramagnetic impurity caused by the change in flux state of the squid . to see coherence it is necessary , from the discussion of section 4.b , that @xmath728 , because the decoherence time coming from the spin bath is @xmath729 ; this essentially sets a lower bound for @xmath321 .
as discussed in some detail in a recent paper @xcite , this turns out to be a rather stringent requirement on real squids ; in fact in the experiments of the lukens group @xcite one infers a value @xmath730 from their resonant linewidths .
obviously this value could be reduced a great deal by careful attention to the nuclear and paramgnetic spin impurity composition in the system ( as well as to the sample geometry ) .
\(ii ) _ decoherence in magnets _ : the most dramatic claims for the observation of macroscopic coherence have been made by awschalom et al .
@xcite , working on randomly oriented dilute ensembles of ferritin macromolecules ( which order antiferromagnetically , but carry an excess moment of somewhat random size ; the antiferromagnetic `` nel '' moment is @xmath731 ) . in an effort to make the molecule size as uniform as possible , these authors filtered them magnetically .
they also artificially engineered molecules of smaller size .
the essential result was the observation of an absorption peak at @xmath732 frequencies , whose frequency varies approximately exponentially with the size of the molecules .
this was interpreted as a signature of coherent tunneling between `` up '' and `` down '' states of the neel vector .
there have been widespread objections to this interpretation , both on theoretical and experimental grounds @xcite , and so far no other group has succeeded in confirming the experiments .
note that the awschalom group saw similar resonances ( also with an exponential dependence of resonant frequency on size ) in large feco@xmath733 particles , but did _ not _ attribute this to tunneling @xcite . as we discussed in the previous sub - section
, there is now very extensive evidence that nanomagnetic molecules in macroscopically ordered crystals tunnel _
incoherently _ in the low-@xmath0 quantum regime @xcite .
there is thus an apparent contradiction between the ferritin work and that done in the mn-12 and fe-8 systems ( particularly since the ferritin molecules are much larger and certainly contain a lot of spin disorder ) .
thus in the very well - characterised @xmath136 - 8 molecular crystals used by the florence and grenoble groups , the parameter @xmath437 characterising decoherence from the orthogonality blocking mechanism varies @xcite , even in an ideal sample , between @xmath734 in zero applied field ( depending on the annealing - dependent spread @xmath735 in the intermolecular dipolar fields ) , to @xmath736 when @xmath737 ( where the first zero in @xmath190 is supposed to occur @xcite ) .
recalling from sections 3.c and 4.a that coherence is practically eliminated unless @xmath446 , it is hard to see how experiments on these particular molecules in low fields will stand much chance of seeing it .
on the other hand it is clear that future experiments on _ single _ nanomagnets in the quantum regime might have a chance of seeing coherence _
_ one could raise @xmath321 to values @xmath738 the hyperfine couplings @xmath197 ( presumably using a large external transverse field ) , thereby making @xmath446 and so removing decoherence ( and also reducing the problem to a straightforward spin - boson model- see section 4.b ) .
another possibility , which could be realised in , eg .
, the liho@xmath718y@xmath719f@xmath720 system at high transverse fields , would be to see coherent propagation of spin flips ( ie .
, spin waves ) in a lattice of spins , by making @xmath739 ( here one could also make @xmath740 , thereby _ freezing _ the nuclear dynamics ! ) .
conceivably the same could be done in an @xmath136 - 8 crystal ( now with the inequality @xmath741 operating ) .
we see no reason why such experiments could not be done in the next few years .
\(iii ) _ decoherence as _ @xmath3 : let us now consider the general question of decoherence effects at low @xmath0 .
decoherence is often ( particularly in conductors ) characterised by a `` decoherence time '' @xmath714 , for the phase dynamics of the degree of freedom of interest . if and when @xmath714 is meaningful , it may be much shorter than the energy relaxation time @xmath742 ( _ cf . _ the example of a single oscillator coupled to an oscillator bath @xcite , or the examples of topological decoherence given for the spin bath in sections 3.a and 4.a above ) ) .
coherence exists if @xmath743 , where @xmath321 is the characteristic frequency of the system s phase dynamics .
notice that what allows us to discuss this problem with any generality at all is the assumption , discussed in sections 1 and 2 , that a few canonical models describe the low-@xmath0 behaviour of most physical systems .
extensive study of the relevant canonical oscillator bath models ( in particular , the spin - boson model @xcite and the `` oscillator on oscillators '' model @xcite ) show that with a power - law form @xmath744 , decoherence disappears as @xmath170 for @xmath745 ; for the ohmic form @xmath746 decoherence is finite at @xmath747 , but can be made small if @xmath748 .
if the electronic spectrum is gapped the ohmic dissipation falls off exponentially in the low @xmath0 limit ( thus for superconductors @xcite one has @xmath749 for @xmath750 , and for magnetic solitons @xcite one has @xmath751 for @xmath752 , where @xmath753 and @xmath754 are the bcs and magnon gaps respectively ) .
thus , if one believes the oscillator bath models , coherence ought to be easily observable at low @xmath0 ; the condition @xmath755 is clearly satisfied for temperatures well below the gap energy . if we examine the canonical spin bath models we find a very different story @xcite .
consider first the central spin model as @xmath3 ; we will go to such a low temperature that all of the spins in the spin bath order in the field of the central spin ( ie .
, @xmath756 in some cases ) , ie .
, all intrinsic fluctuational dynamics of the bath is frozen out .
does decoherence disappear ?
no , because the mechanisms of topological decoherence ( induced bath spin flip ) and orthogonality blocking ( precession of the bath spins in between central spin flips ) still exist- the bath can still acquire dynamics from the central spin . we emphasize here that this physics can not be described by an oscillator bath model . from the results in section 4 we see
there is a residual constant decoherence as @xmath3 , coming from the spin bath , which in an experiment would be signalled by a saturation of @xmath714 once oscillator bath effects had disappeared .
the extent of this decoherence is characterised by the behaviour of @xmath459 , ( cf . section 4.a ) , and we saw that unless both @xmath437 and @xmath199 were @xmath757 , decoherence was strong . in the case where all bath spins are polarised by the central spin ,
transitions are blocked anyway- there are no spins in the @xmath445 polarisation group ! we conclude that for any system described by the central spin model ( ie . , where the central system reduces to a 2-level system at low energies ) , a general consequence of the coupling to a spin bath will be a loss of coherence , via either the topological decoherence or orthogonality blocking mechanisms , even in the @xmath3 limit . the residual coherence ( if any ) will depend on the strength of the couplings to the spin bath , in a way discussed quantitatively in sections 4 and 5 .
we may generalise these considerations to models in which a `` particle '' moves through a spin bath ( the same model also describes a network of spins , or of mesoscopic superconductors , etc .
, coupled to a spin bath ) , and get the same result .
consider , eg .
, a particle hopping from site to site on a @xmath758-dimensional hypercubic lattice , whilst coupled to a spin bath @xcite , and described by a hamiltonian @xmath759 + h.c .
\bigg ] \nonumber \\ & + & \sum_k \bigg [ \;^{\parallel}\omega_k^{ij } ( c_i^{\dagger}c_i - c_j^{\dagger}c_j ) \vec{\hat{\sigma}}_k^z + \omega_k^{\perp } \hat{\sigma}_k^x \bigg ] \bigg\ } + \sum_{k , k ' } v_{kk'}^{\alpha \beta } \hat{\sigma}_k^{\alpha } \hat{\sigma}_{k'}^{\beta } \ ; ; \label{sblatt}\end{aligned}\ ] ] where @xmath760 sums over nearest neighbour sites .
the couplings @xmath761 and @xmath762 to the bath spin @xmath200 usually depend upon which `` lattice rung '' @xmath763 the particle happens to be , because the coupling normally has a finite range and the spin @xmath200 has some position with respect to the lattice ( one can also add a longitudinal coupling to the @xmath764 on each site , as in eqtn .
( [ qsgl ] ) ) . in the continuum limit ( [ sblatt ] ) reduces to ( [ 2.6 ] ) and ( [ 2.7 ] ) , after dropping the external field @xmath143 and the dependence of @xmath127 on @xmath24 . to isolate the decoherence effects let us assume @xmath765 , ie .
, we now study the analogue of the strong - coupling regime in section 4.a . without loss of generality
we may then concentrate on the `` phase decoherence '' regime , ie .
, on an effective hamiltonian @xmath766 , where @xmath767 \label{hproj}\ ] ] this is just a generalisation of ( [ 5n.12 ] ) to the lattice ; the dependence of the coupling on the lattice position is dropped because it is inessential to what follows .
coherence , if it exists , will appear in the function @xmath768 ( the probability to start at site @xmath769 , and be at site @xmath341 a time @xmath347 later ) .
perfect coherence ( ie . ,
no bath ) yields @xmath770 } \;=\ ; \prod_{\mu=1}^d j_{n_{\mu}}^2(2 \delta_o t ) \label{ballistic}\ ] ] for which @xmath771 at long times ; moreover , the 2nd moment @xmath772 at long times . both are characteristic of coherent `` ballistic '' band motion .
if we now go to the interacting case one finds , by a straightforward generalisation of the calculations in sections 3 and 4 , some rather interesting results @xcite .
consider first a bath with all spin states equally populated .
then at long times @xmath773 but @xmath774 , _ independent _ of @xmath758 ! moreover ,
if we start with an initial gaussian wave - packet of width @xmath775 , one finds @xmath776 as @xmath777 .
these results show that the naive inference of ballistic propagation from the second moment result is wrong- in reality one has strongly anomalous diffusion ( with an energy - dependent diffusion coefficient , demonstrated by the @xmath778-dependence of the results ) .
in fact the probability function @xmath779 decays like @xmath780 at long times ( ie . , @xmath781 ) and distances @xmath782 , out to a `` ballistic '' distance @xmath783 , for dimension @xmath784 .
thus there is an advance `` ballistic front '' which decays in amplitude with increasing distance / time from the origin , but only as a power law ; and it is followed by a much larger anomalously diffusive ( and of course incoherent ) contribution . for sufficiently long times ( in fact for @xmath785 )
this result is _ independent of _ @xmath199 .
similar results apply for the orthogonality- blocked case as a function of @xmath437 .
now let s take the limit @xmath3 , meaning that we allow the bath spins to order in the field of the particle .
it is clear that if @xmath569 for all spins , _ and _ provided there are no non - diagonal momentum couplings to the bath , we can not get decoherence by the same mechanism as above , since there is only one state in the relevant polarisation group- the particle will then move freely without disturbing the spins in any way .
however in any other case phase will be exchanged with the bath in the same way as above , with or without dissipation ( which will certainly arise in the weak - coupling limit @xmath786 ) .
we therefore conclude that a finite decoherence in the @xmath3 limit is a generic consequence of the existence of spin bath environments .
\(iv ) _ qubits and quantum computation _ : what is the impact of these results on hopes for quantum computation ?
a quantum computer is an information processing device which can be imagined as an assembly of 2-state qubits , these being none other than the spin - boson and central spin models discussed in this article @xcite .
such a computer has yet to be built , and papers and books on this topic tend to divide into 2 classes .
the first simply ignores decoherence ( apart from occasionally referring to it as the main stumbling block preventing the construction of a quantum computer ! ) , whereas the second regards decoherence as the crucial problem , and either tries to treat it theoretically ( eg . , @xcite ) or maintains that a quantum computer will never be built because of it ( eg . , @xcite ) .
the basic problem here is the lack of serious theory on the effects of decoherence , starting from realistic models which can be tested quantitatively by experiment .
the analysis of the central spin model in sections 3 and 4 can be used for a single qubit- it will be clear that the main task is to reduce diagonal couplings to the spin bath ( and also any oscillator bath ) as far as possible .
however this is only the beginning of the problem- the operation of a quantum computer involves multi - qubit wave - function entanglement , and thus one wants to understand the behaviour of a decoherence time @xmath787 , governing loss of @xmath214-spin phase correlations , in the presence of coupling to spin and oscillator baths .
we are aware of no studies of this problem ( or even recognition that it is a problem ) in the literature
. studies of mutual coherence and decoherence in the problem of 2 spins coupled to an oscillator bath @xcite give some feeling for what might happen , as do the studies of lattice systems .
however we still have no answer to , for example , the question of how @xmath788 behaves for large @xmath214 .
if it ends up having a generic decrease @xmath789 then the whole quantum computing enterprise will be in very serious trouble !
it is clear that in the near future research on quantum computers will have to proceed on both practical designs ( with serious attention given to the decoherence characteristics of the relevant materials ) , and also on general studies of decoherence for systems of coupled qubits , themselves coupled to spin and oscillator baths .
this promises to be one of the great challenges in condensed matter physics during the next decade .
we would like to thank nserc and the ciar in canada , and the russian foundation for basic research ( grant 97 - 02 - 16548 ) and european community ( grant intas-2124 ) , for financial support of our research .
we also thank i. affleck , y. aharonov , b. barbara , j. clarke , m. dub , w. hardy , a.j.leggett , t. ohm , c. paulsen , r. penrose , g. rose , i.tupitsyn , w.g .
unruh , i.d .
vagner , and w. wernsdorfer for helpful discussions about the spin bath , going back to 1987 .
a number of papers touching upon the present subject ( particularly on decoherence ) have appeared since this article was written . on the highly controversial question of zero - temperature decoherence in mesoscopic conductors @xcite ( which according to some authors throws the entire conventional theory of metals into doubt ) , a large number of suggestions have appeared .
a list of them appears in a short experimental review by mohanty @xcite , which concludes that only the original suggestion , of electronic coupling to zero - point fluctuations of the em field @xcite , can explain the experimental saturation of @xmath714 .
this conclusion is hotly disputed by other authors , both on experimental and theoretical grounds @xcite .
the idea that the decoherence might be coming from `` two - level systems '' in the sample ( ie .
, from a spin bath ) has been explored by imry et al .
@xcite and zawadowski et al .
@xcite , but rejected by mohanty , mainly because it would imply very large low - frequency noise levels in the sample . in our opinion this requires further work- some of the decoherence mechanisms discussed in the present article do not appear in these papers , and would not necessarily show up in the low - frequency noise . in any case
this controversy shows clearly how the theory of transport , including weak localisation theory , depends crucially on a correct understanding of decoherence mechanisms .
the most interesting recent experimental progress concerning the mechanisms of decoherence may be work in magnetic systems .
the most recent work by wernsdorfer et al .
@xcite , which looked at the dependence of tunneling rates and `` hole widths '' in the resonant quantum relaxation of crystals of @xmath136 - 8 molecules , has been supplemented by further work on the same system @xcite .
taken together these experiments give rather strong evidence that the tunneling is mediated and controlled by the nuclear spins in the system .
no direct test has yet appeared of the theory of the decoherence coming from these same nuclear spins , however ( which depends on calculations of the parameter @xmath437 ; cf .
section 2.6 , including fig . 5 , and sections 3.3 , 4.1 , 5.1 , and 5.3 , including fig
. 8 , and refs .
experiments looking for coherence in @xmath136 - 8 have also been done at higher fields @xcite ; however we are unable to see why the reported results give evidence either for or against coherence , since they simply show very broad and rather weak peaks in the esr spectrum as a function of field .
we emphasize a point here which has often been made , viz .
, that any demonstration of coherence requires direct observation of combinations of multi - time correlation functions- even a very sharp peak in , say , the ac absorption is not enough to demonstrate coherence . in this connection the reader
is referred to figs . 7 - 9 in the present paper , which show peaks in @xmath557 even when there is no coherence at all .
several reviews of the many different experiments in tunneling nanomagnets have also recently appeared @xcite .
one topic not covered in these is the liho@xmath718y@xmath719f@xmath720 system ( section 5.1.3 ) ; some interesting new results on this compare the thermal and quantum annealing , showing the efficiency of the latter in quantum optimisation@xcite .
more general discussions of coherence and decoherence appearing recently include several theoretical reviews @xcite ( although these do not really discuss spin bath environments )
. there has also been interesting experimental work on systems other than superconductors and magnets ; see , eg .
, wiseman et al .
@xcite for coupled quantum dots , and myatt et al .
@xcite for decoherence in quantum optical systems .
in this appendix we give the derivations of some key formulae in sections 3 and 4 .
instanton methods are used to handle the spin environment . in appendix
a.1 we give details of the fairly trivial calculations required to deal with averaging over bias , topological phase , and fluctuations in bias .
then in appendix a.2
we discuss the more lengthy derivation of the expression ( [ 4.21o ] ) involved in orthogonality blocking .
we wish to evaluate @xmath790 , the probability for the central spin to return to an initial state @xmath346 after time @xmath347 , in the presence of a static bias @xmath477 and a noisy bias @xmath791 .
we use ssh physics.ubc.ca standard instanton techniques @xcite , because they easily generalise to include the spin bath .
we begin by ignoring the topological phase of the central spin since its effects are trivial to add . then the amplitude for a 2-level system to flip in a time @xmath792 is @xmath793 , and the return probability
is given by summing over even numbers of flips : @xmath794\ ; , \label{b1.2}\ ] ] now consider the modification introduced by a _
static bias @xmath477 . in instanton
language we begin with the return _ amplitude _ @xmath795 , and laplace transform it ; using the action @xmath796 in bias @xmath477 , over time @xmath792 , we get @xmath797 @xmath798 where @xmath799^{1/2}$ ] ; then the standard answer for @xmath800 is just : @xmath801 \ ; = \ ; [ 1- { \delta_o^2 \over e^2 } \sin^2 et ] \label{b1.10}\end{aligned}\ ] ] this representation avoids the difficulty in the usual representation coming from the square root @xmath802^{1/2}$ ] in the cosine .
we now add a fluctuating bias @xmath791 to @xmath477 , with the correlation @xmath473 ^ 2 \rangle = \lambda^3 \vert t - t ' \vert$ ] for short times .
noise averages are then given by the gaussian average @xmath803 \rangle \;=\ ; \int { \cal d}\epsilon ( t ) \ ; f [ \epsilon ] e^{-{1 \over 2 } \int dt_1 \int dt_2 \epsilon ( t_1 ) k ( t_1 - t_2 ) \epsilon ( t_2)}\ ] ] with a noise correlator @xmath486 .
averaging over a phase function @xmath804 then gives the standard result @xmath805 in applying this to ( [ b1.6 ] ) we assume fast diffusion ( see text ) , and thus only expand to @xmath806 .
this gives @xmath807 , with @xmath808 where @xmath809 is the probability @xmath791 takes value @xmath472 ( ie .
, it is the lineshape of the polarisation group ) . adding the the topological phase @xmath810 to these calculations changes the flip amplitude to @xmath811
; one then sums over all paths with an even number of flips _ and _ over all combinations of @xmath812 ( clockwise and counterclockwise ) flips ; thus ( [ b1.2 ] ) becomes @xmath813\ ; , \label{b1.4}\ ] ] with similar obvious modifications to include bias .
the generalization to include the phase from the bath spins ( topological decoherence ) is now obvious for both imaginary and full complex @xmath361 ( see text , section 3.a ) .
we consider the situation described in section [ sec:4].b .
, where the `` initial '' and `` final '' fields on the bath spin @xmath420 are @xmath418 and @xmath814 , related by an angle @xmath421 , which is assumed small , and is defined by @xmath815 .
we choose axes in spin space such that the initial and final spin bath wave - functions are related by @xmath816 where @xmath817 ( compare eqtn .
( [ b.2 ] ) .
in general the initial spin bath state will belong to some polarisation group @xmath818 ( not necessarily @xmath819 ) , where the polarisation is defined along some direction defined by the central system ( for example , in a nanomagnetic problem , one could define it as the direction of initial orientation of the nanomagnetic spin ) . as explained in the text , during a central system transition energy conservation
requires the polarisation to change from @xmath818 to @xmath820 ( and back , for further transitions ) ; for `` pure '' orthogonality blocking ( ie .
, when no other terms are involved in the hamiltonian ) , @xmath821 . in what follows
we calculate the correlation function @xmath822 , the central spin correlator defined under the restriction that the spin bath transitions are between subspaces defined by @xmath823 and @xmath824 subspaces , which are supposed to be in resonance .
the statistical weight of states with @xmath825 is negligible , so we will assume that @xmath826 .
we enforce the restriction to a polarisation group @xmath214 using the projection operator @xmath827 we can now write down an expression for the _ amplitude _ ( not the probability ! ) @xmath828 for the central spin @xmath355 to stay in state @xmath346 during a time @xmath347 : @xmath829 where @xmath830 is @xmath831 \;. \label{b.6}\ ] ] > from ( [ b.5 ] ) we can now write the full correlation function @xmath832 as @xmath833 we now use the assumption that the @xmath421 are small ; more precisely we assume that the orthogonality exponent @xmath834 , defined previously by @xmath835 ( cf . eqtn ( [ k.1 ] ) ) , can be approximated by the perturbative expansion @xmath836 .
this assumption makes it much easier to calculate the average in ( [ b.7 ] ) .
we consider first the problem with only one environmental spin @xmath420 , and calculate the average @xmath837 in this case ; since @xmath838 is a product of operators acting separately on each @xmath200 , the average over all spins is also the product of single spin results .
we only need consider processes with @xmath839 , or @xmath840 flips of the environmental spin , i.e. , we expand in powers of @xmath421 , and stop at @xmath841 .
then it is clear that , if the initial state of @xmath420 is @xmath842 @xmath843 \ ; , \label{b.10}\end{aligned}\ ] ] where the first term arises from the sequence @xmath844 $ ] , the second from the sequence @xmath845 $ ] , with a flip when @xmath846 ; and so on .
in the same way we find @xmath847 \bigg ] \ ; , \label{b.11}\end{aligned}\ ] ] to order @xmath841 .
the sequence @xmath848 will have a similar expression , but with reversed signs coming from the @xmath849 factors .
we now observe that the state with polarisation @xmath850 consists of @xmath851 spins up and @xmath852 spins down .
consequently , for each @xmath420 , we add @xmath853 or @xmath854 averages like ( [ b.11 ] ) , and then take the product @xmath855 substituting ( [ b.11 ] ) into this expression we get @xmath856 where the `` effective action '' @xmath857 has two contributions @xmath858 : @xmath859 + \sum_{p^\prime > p}(-1)^{p^\prime -p } \cos [ 2\sum_{j = p}^{p^\prime -1}\xi_j^\prime ] \nonumber \\ & - & \sum_{p = 1}^{2 m } \sum_{l=1}^{2n } ( -1)^{l+p } \cos [ 2\sum_{i = l}^{2n } \xi_i - \sum_{j=1}^{p-1 } \xi_j^\prime ] \bigg\ } \ ; , \label{b.12c}\end{aligned}\ ] ] @xmath860 + \sum_{p^\prime > p}(-1)^{p^\prime -p } \exp [ 2i \sum_{j = p}^{p^\prime -1}\xi_j^\prime ] \nonumber \\ & - & \sum_{p = 1}^{2 m } \sum_{l=1}^{2n } ( -1)^{l+p } \exp [ 2i \sum_{j=1}^{p-1 } \xi_j^\prime -2i \sum_{i = l}^{2n } \xi_i ] \bigg\ } \ ; , \label{b.12d}\end{aligned}\ ] ] we recall now that @xmath861 , which allows us to neglect the contribution due to @xmath862 and drop the correction @xmath863 to the coefficient @xmath834 in @xmath864 .
notice also that the phase factor in front of @xmath865 in ( [ b.12 ] ) cancels exactly the phase proportional to @xmath818 in the formula ( [ b.7 ] ) for @xmath866 .
thus , quite surprisingly , we find the correlation function to be independent of @xmath818 in this limit : @xmath867 we can render this expression more useful by changing variables ; first we consider the whole sequence @xmath868 together , and then define new angular variables @xmath869 so that now @xmath870 \bigg\ } \;.
\label{b.16}\end{aligned}\ ] ] thus we have mapped our problem onto the partition function of a rather peculiar system of spins , interacting via infinite range forces , with interaction strength @xmath871 . to deal with this partition function , we define `` pseudo - spins '' @xmath872 and @xmath873 , such that @xmath874 we can think of @xmath875 as rotating in our fictitious angular space defined by the projection operator ( [ b.4 ] ) .
now consider the term @xmath876 in ( [ b.16 ] ) defined by @xmath877 this is easily calculated , viz .
, @xmath878 where @xmath879 is the @xmath214th - order bessel function . using @xmath880 to reorganize the sum over @xmath341 and @xmath25 in ( [ b.16 ] ) and changing the integration variable @xmath881 , we then find @xmath882^{2s } \over ( 2s ) ! }
\right ) \nonumber \\ & = & p_{m}(t)= \int_0^\infty dx x\:e^{-x^2 } \big ( 1 + \cos [ 2 \delta_o(\phi ) j_m(2x\sqrt{\kappa } ) t ] \big ) \equiv 2\int_0^\infty dx x\:e^{-x^2 } p^{(0)}_{11}(t,\delta_m(x ) ) \;. \label{b.21b}\end{aligned}\ ] ] @xmath883 here we come to the crucial point in our derivation .
eq.([b.21b ] ) gives the final answer as a _ superposition of non - interacting _ correlation functions for effective tunneling rates @xmath884 with the proper weighting . for @xmath445
this is the form quoted in eq.([4.21o ] ) of the text .
it is worth noting that non - zero @xmath214 enters this calculation as the overall phase factor which we can follow from ( [ b.7 ] ) up to ( [ b.20 ] ) , where we finally integrate over @xmath885 to produce the bessel function of order @xmath214 .
this observation allows one to generalise any calculation done for @xmath445 to finite @xmath214 by simply replacing @xmath886 in the final answer - the prescription which we use in other appendices .
we outline here the derivations for section 4 , in which topological decoherence , degeneracy blocking , and orthogonality blocking are all simultaneously incorporated ( the average over bias fluctuations being essentially trivial- see appendix a.1 ) .
we have demonstrated in section 3 and appendix a how each different term in the effective hamiltonian ( [ 1.24 ] ) influence the central spin dynamics . from these limiting cases
we learned that static ( or diagonal ) terms in the hamiltonian can be partly absorbed into a redefinition of the transition amplitude between states with equal initial and final energies .
if we now deal with the full central spin hamiltonian , we can still write the instanton expansion in central spin transitions in the form ( see ( [ b1.10 ] ) ) : @xmath887 @xmath888 where the sum over @xmath889 with @xmath890 describes all possible clockwise and anticlockwise transitions , and the operator product is defined as @xmath891 @xmath892 here @xmath893 as before projects on the polarisation state @xmath894 , and @xmath421 describes the mismatch between the initial and final nuclear spin states .
if all the couplings were equal ( @xmath895 ) then the above set of equations would be the complete solution of the @xmath214 polarisation group dynamics .
one may then further average over different grains in the ensemble by summing over different polarisation groups with proper weigthing .
if there is a small spread in the nuclear hyperfine couplings , it will produce an internal bias field acting on the grain , as described in the text ( section [ sec:4].c .
the final answer for the ensemble of grains is obtained then by averaging ( [ y.1 ] ) over the bias field .
this bias field is due to _ all _ environemntal spins interacting with @xmath76 , and hardly changes when a few nuclei ( of order @xmath896 ) flip with @xmath76 .
for this reason there is no back influence of the induced nuclear spin flips on the bias field , at least during the time scale set by the damping of coherent oscillations ( when many environemntal spins are flipped the coherence is obviously already lost ) .
the crucial observation is that if the sum over the clockwise and anticlockwise trajectories and the average of the operator product in ( [ y.2 ] ) can be presented as some weighted average and/or sum , of form @xmath897 with fixed integer values @xmath81 and @xmath898 , then the problem may be considered as solved because the instanton summation then reduces to that of a coherent ( non - interacting ) dynamics with the renormalized tunneling amplitude @xmath899 and the final answer acquires a form @xmath900 where @xmath901 is described by eq .
( [ b1.10 ] ) .
we have already seen that eq .
( [ y.5 ] ) is indeed valid for the cases of pure topological decoherence and pure orthogonality blocking- we now prove that it also holds when we combine the effects of topological decoherence with the projection on a given polarisation state . here
we evaluate the @xmath445 contribution ; the result for @xmath441 then follows from the generalisation explained at the end of the previous appendix . with the usual assumption that the individual @xmath361 are small ( but not necessarily @xmath593 ) ,
the `` effective action '' @xmath904 can be written as ( compare eq.([b.16 ] ) ) @xmath905 which generalizes from orthogonality blocking ; the @xmath906 are defined as in ( [ b.15 ] ) , and @xmath907 as before .
we use the same trick of introducing pseudo - vectors " @xmath908 , and @xmath909 , to decouple integrals over the new variables @xmath910 . after some lengthy , but straightforward algebra we get @xmath790 in the form @xmath911 where @xmath912 as before . the case of complex @xmath361 ( even assuming @xmath913 in the effective hamiltonian ) is more subtle technically , but goes through in exactly the same way . here
we just outline the key steps ; a more detailed derivation may be found in @xcite .
the effective action now has the form @xmath914 \nonumber \\ & + & \cos ( \chi_\rho - \chi_{\rho^\prime } ) \big [ ( \lambda - \lambda^\prime ) -(-1)^{\rho + \rho^\prime } ( \eta - \eta^\prime ) -i(\gamma - \gamma^\prime ) ( ( -1)^\rho + ( -1)^{\rho^\prime } ) \big ] \bigg\ } \ ; , \label{c.25}\end{aligned}\ ] ] where the constants are defined by : @xmath915 @xmath916 @xmath917 as before we change variables according to @xmath918 when @xmath919 , to introduce odd and even spin fields @xmath920 which are used to decouple integrations over @xmath906 with the final goal to get the answer in the form of eq .
( [ y.5 ] ) .
this indeed can be done , and the final answer reads @xmath921 \ ; , \label{5.29}\end{aligned}\ ] ] and ( [ 5.29 ] ) has an obvious generalisation to include the bias integration @xmath922 .
the weight is given by @xmath923 } e^{4(\eta^\prime m_2 ^ 2 - \lambda^\prime m_1 ^ 2 ) } \nonumber \\ & \times & { x_1x_2 \over 8(ab - c^2 ) } i_0\left ( { ( a+b ) x_1x_2
\over 8(ab - c^2 ) } \right ) \exp \left\ { { ( a - b+2ic)x_1 ^ 2 + ( a - b-2ic)x_2 ^ 2 \over 16(ab - c^2 ) } \right\ } \ ; , \label{5.34}\end{aligned}\ ] ] and the renormalized tunneling splitting equals @xmath924 where @xmath925 * figure 1 * the flow of a class of effective hamiltonians describing a central system coupled to a background environment , in coupling constant space , as the uv cutoff in the joint hilbert space is reduced from @xmath11 to @xmath20 . here
we show flow to a fixed point fp , in a simplified 2-dimensional space of couplings @xmath926 , but one may also have fixed lines or more complex topologies .
* figure 2 * a typical path for a 2-level central system ( solid line ) coupled to environmental modes ( wavy lines ) as a function of time , showing the couplings which exist in both the spin - boson and central spin models .
we show both diagonal couplings d to @xmath209 and non - diagonal couplings nd to @xmath927 ( in the central spin model these are strong enough to lead to multiple excitation of environmental modes ) .
* figure 3 * definition of the longitudinal and transverse parts of the diagonal coupling to a bath spin in the central spin hamiltonian , in terms of the initial and final fields @xmath928 and @xmath929 acting on this spin- this also defines the angle @xmath930 , and the mutually perpendicular unit vectors @xmath931 and @xmath932 ( see text ) .
* figure 4 * classifying the states of the central spin hamiltonian .
each level of @xmath355 is associated with a @xmath212-fold multiplet of bath states ( fig 6(a ) ) .
these are classified into polarisation groups @xmath933 ( where @xmath214 is the total polarisation along @xmath934 ) , separated by energy @xmath223 and with width @xmath935 ; fig 6(b ) shows the density of states @xmath936 of the separate groups , and fig .
6(c ) their sum @xmath532 .
we show @xmath532 for 2 different values of the parameter @xmath937 ; in realistic cases @xmath938 ( ie .
, the polarisation groups strongly overlap ) , and @xmath532 is gaussian .
longitudinal transitions between 2 different polarisation groups @xmath939 and @xmath940 go at a rate @xmath941 ; transitions within a polarisation group at a rate @xmath942 .
* figure 5 * example of the application of the central spin model to a magnetic macromolecule ( the @xmath136 - 8 molecule , further descibed in section 5 ) . in ( a )
we show the effective tunneling matrix element @xmath943 , for this easy axis / easy plane nanomagnet in the presence of a field @xmath944 in the @xmath433-direction ( transverse to the easy axis ) , assuming an angle @xmath945 between @xmath946 and the magnetic `` hard axis '' ( perpendicular to the easy plane ) .
aharonov - bohm oscillations appear when @xmath945 is small , so that the action of the 2 relevant paths on the spin sphere have similar _ magnitudes _ , but almost opposite phase . for larger @xmath945 ,
one path dominates over the other and oscillations are suppressed . in ( b ) we show a histogram of the @xmath183 for this system- the main figure shows the protons and the lower inset the @xmath1 and @xmath947 contributions .
the upper inset in ( b ) shows the variation of @xmath216 and @xmath523 with @xmath948 ( the parameter @xmath523 is discussed in sections 4 and 5 ) .
these figures are adapted from ref .
[ 69 ] .
* figure 7 * the effect of relaxation on a statistical ensemble of central spins , each interacting with a spin bath . in ( a )
we assume that @xmath950 and @xmath951 , and show the normalised time dependence of 3 different contributions @xmath428 to the total relaxation function @xmath345 ; they sum to give a roughly logarithmic time dependence for the total function @xmath345 .
the small @xmath214 contributions relax quickly ( up to @xmath547 ) , so the effect on an initial ensemble distributed over bias @xmath477 , at short times , is to dig a hole around zero bias , of width @xmath952 . in ( b )
we show the spectral absorption function @xmath557 for @xmath953 , dividing this into the @xmath445 contribution and the contributions from higher @xmath214 groups ( which relax more slowly and thus peak at lower @xmath10 ) .
* figure 8 * the spectral absorption function @xmath954 for several values of @xmath437 , for an ensemble of central spins in the @xmath445 polarisation group , in the case where orthogonality blocking dominates , and degeneracy blocking effects ( ie . , a bias average ) are also incorporated .
contributions from higher polarisation groups @xmath528 are not shown ; they are spread over a range @xmath955 , up to @xmath952 .
contributions from groups with @xmath956 are negligible .
* figure 9 * graphs of @xmath954 for `` projected topological decoherence '' ( ie .
, including a bias average over an ensemble in which topological decoherence dominates ) , for several different values of the parameter @xmath957 .
contributions from higher polarisation groups , which are spread over an energy range @xmath958 , are not shown .
classical oscillator models of environments go back to rayleigh ( at least ) .
quantum models of bosonic environments began with the development of quantum electrodynamics ( see , eg .
, p.a.m.dirac , proc .
a114 * , 243 ( 1927 ) , or e. fermi , rev .
phys . * 4 * , 87 ( 1932 ) ) , and later became an integral part of quantum optics and condensed matter physics ( in the form of `` quasiparticles '' ) .
the theory of quantum `` brownian motion '' ( for which see , eg . , i.r .
senitzky , phys . rev . * 119 * , 670 ( 1960 ) ; j. schwinger , j. math . phys . * 2 * , 407 ( 1961 ) , g.w .
ford , m. kac , p. mazur , j. math . phys .
* 6 * , 504 ( 1965 ) , and a.o.caldeira , a.j .
leggett , physica * 121a * , 587 ( 1983 ) ) has also exercised an important influence on such models .
the development of oscillator models for fermionic baths belongs mostly to the history of condensed matter physics , beginning with f. bloch , zeit .
fur physik * 81 * , 363 ( 1933 ) , for spin excitations , or l.d .
landau , jetp * 16 * , 574 ( 1946 ) , d. bohm and e.p .
gross , phys .
rev . * 75 * , 1851 ( 1949 ) , and s. tomonaga , prog .
. phys . * 5 * , 349 ( 1950 ) , for electronic charge excitations . the first attempt to put oscillator models on a general footing was by feynman and vernon @xcite , who recognised their perturbative nature ;
more recently their development in many fields in physics has followed the ideas of leggett et al .
@xcite , particularly in the application to tunneling phenonena in complex systems @xcite , and to quantum dissipation theory @xcite .
`` random level '' models of quantum environments begin with the work of e.p .
wigner , ann .
math * 53 * , 36 ( 1953 ) ( see also d.l hill , j.a .
wheeler , phys .
rev . * 89 * , 1102 ( 1953 ) ) and f.j .
dyson , j. math .
phys . * 4 * , 140 , 157 , 166 ( 1962 ) . in recent years
they have been formulated field - theoretically ( see , eg .
efetov , adv .
phys . * 32 * , 53 ( 1983 ) ) , and applied in condensed matter physics ( particularly mesoscopic physics- see k.b efetov , _ `` supersymmetry in disorder and chaos '' _ , c.u.p .
( 1997 ) , c.w.j .
beenaker , rev .
* 69 * , 731 ( 1997 ) , or b.l .
altshuler , b.d.simons , in _ `` mesoscopic quantum physics '' _ , les houches session lxi ( ed .
e.akkermans et al . , ( 1994 ) ) , pp . 1 - 98 ) ; and in nuclear physics . for general studies of environmental dissipation with such models ,
see m. wilkinson , j. phys . * a21 * , 4021 ( 1988 ) , m. wilkinson , e.j .
austin , phys .
rev * a46 * , 64 ( 1992 ) , and j. phys . * a28 * , 2277 ( 1995 ) ; and a. bulgac , g. do dang , d. kusnezov , ann . phys . *
242 * , 191 ( 1995 ) , and phys .
rev * e54 * , 3468 ( 1996 ) .
`` spin bath '' models in which the environment is explicitly modelled as a set of `` spins '' , or 2-level systems , probably began in nmr and quantum optics ( see eg .
, y.r.shen , phys .
rev * 155 * , 921 ( 1967 ) , or a.s .
davydov , a.a .
serikov , phys .
( b ) * 51 * , 57 ( 1972 ) ) , but these studies were not only in the weak coupling limit but also used phenomenological damping equations , in which all `` memory '' and `` back - reaction '' effects ( including phase exchange with the environment ) are lost .
more recent studies have appeared in the context of @xmath959 noise @xcite , squid tunneling @xcite , in landau - zener dynamics @xcite , and in magnetic systems @xcite , as well as in a more general context @xcite .
attempts to map spin bath models to an oscillator bath description ( which are only generally valid in the weak - coupling limit ) include refs .
@xcite . some papers @xcite acheive a weak - coupling limit by multiplying all couplings by @xmath7 , where @xmath1 is the number of bath spins- we do not believe this procedure is physically correct .
the solution to the central spin " model ( the analogue for spin baths of the spin - boson model ) , including all relevant couplings and going beyond the weak coupling limit , was first given in refs .
@xcite , with earlier partial solutions in ref .
@xcite . for the original derivation of the dynamics of the central spin model ,
prokofev , p.c.e .
stamp , /cond - mat 9511011 ( ubc preprint , april 20 , 1995 ) , and ref.@xcite .
these papers contain analytic results for most parameter regimes of the model ( some of which are quoted in section 4 of the present article ) .
recent theoretical work on how to make qubits in superconductors includes y. makhlin et al .
, nature * 398 * , 305 ( 1999 ) , l.b .
ioffe et al . ,
nature * 398 * , 679 ( 1999 ) , and j.e .
mooij et al . , science * 285 * , 1036 ( 1999 ) .
a recent experiment finds coherence of a single cooper pair in mesoscopic superconductors ; see y. nakamura , yu .
a. pashkin , j.s .
tsai , nature * 398 * , 786 ( 1999 ) recent ideas on how to make qubits for quantum computers using nuclear spins in quantum hall systems have attracted considerable interest- see v. privman , i.d.vagner , g. kventsel , phys lett .
a*236 * , 141 ( 1998 ) , and b. kane , nature * 393 * , 133 ( 1998 )
. for work on quantum measurement theory from @xmath961 , see m. jammer , _ `` the philosophy of quantum mechanics '' _ ( wiley , 1974 ) , and _
`` quantum theory and measurement '' _ , ed .
wheeler , w.h zurek ( princeton , 1983 ) ; see also j.s .
`` speakable and unspeakable in quantum mechanics '' _ , ( c.u.p . ,
the idea of using squids to test quantum theory , including measurement theory , on the macroscopic scale , comes from a.j .
leggett , prog .
phys ( japan ) supp .
* 69 * , 80 ( 1980 ) [ see also ref .
@xcite ] .
d. b. schwartz , b. sen , c. n. archie , j. e. lukens , phys .
* 55 * , 1547 ( 1985 ) ; r. rouse , s. han , j. e. lukens , phys .
lett . * 75 * , 1614 ( 1995 ) ; s. han , r. rouse , j. e. lukens , phys .
lett . * 76 * ,
3404 ( 1996 ) ; and d.j . flees , s. han , j.e .
lukens , phys .
lett . * 78 * , 4817 ( 1997 )
. for recent experiments on dipolar glasses , see d. salvino et al .
lett . * 73 * , 268 ( 1994 ) ; s. rogge , d. natelson , d.d .
osheroff , phys .
lett . * 76 * , 3136 ( 1996 ) ; d. natelson , d. rosenberg , d.d .
osheroff , phys .
lett . * 80 * , 4689 ( 1998 ) ; c. enss , s. hunklinger , phys . rev . lett . * 79 * , 2831 ( 1997 ) ; p. strehlow , c. enss , s. hunklinger , phys . rev . lett . * 80 * , 5361 ( 1998 ) .
experiments showing @xmath959 and telegraph noise in mesoscopic conductors include k. rahls , r. buhrman , phys .
lett . * 60 * , 2434 ( 1988 ) ; n. birge , b. golding , w. haemmerle , phys .
lett . * 62 * , 195 ( 1989 ) ; and n. zimmerman , b. golding , w. haemmerle , phys .
lett . * 67 * , 1322 ( 1991 ) .
the application of the spin phase to magnetic systems is reviewed in i.k .
affleck , j. phys cm * 1 * , 3047 ( 1989 ) ; e. manousakis , rev .
phys . * 63 * , 1 ( 1991 ) ; e.s .
fradkin , _
`` field theories on condensed matter physics '' _ ( addison - wesley , 1991 ) ; and a. auerbach , _ `` interacting electrons and quantum magnetism '' _ ( springer - verlag , 1994 ) .
more general discussion is in a. shapere , f. wilczek , _ `` geometric phases in physics '' _ ( world scientific , 1989 ) . for the mapping from non - linear couplings ( to oscillator multiplets ) to @xmath0-dependent linear couplings to a different oscillator bath ,
m. kagan , n.v .
prokofev , jetp * 69 * , 1250 ( 1989 ) , a.h .
castro - neto , a.o .
caldeira , phys .
rev . e*48 * , 4037 ( 1993 ) , and m. dub , p.c.e .
stamp @xcite for multiple phonons ; p.c.e .
stamp @xcite for magnons ; and n.v .
prokofev , phys .
rev . b*49 * , 2148 ( 1994 ) for luttinger liquids .
m. uehara , b. barbara , j. physique bf 47 , 235 ( 1986 ) ; m. uehara , b. barbara , b. dieny , p. c. e. stamp , phys . lett .
* 114a * , 23 ( 1986 ) ; b. barbara , p .c .
e. stamp , m. uehara , j de physique * 49 * , c8 - 529 ( 1988 ) ; ll .
balcells et al . , z. phys .
* b89 * , 209 ( 1992 ) ; x. x. zhang et al .
, phys . lett .
* a163 * , 130 ( 1992 ) ; x. x. zhang et al . , j. phys .
cm * 4 * , l163 ( 1992 ) , and phys . rev .
* b53 * , 3336 ( 1996 ) ; w. wernsdorfer et al .
, phys . rev .
* b53 * , 3341 ( 1996 ) , and _ ibid . _ * b 55 * , 11552 ( 1997 ) .
k. hong , n. giordano , phys . rev . * b 51*,9855 ( 1995 ) ; k. hong , n. giordano , j. magn .
magn . mat . *
151 * , 396 ( 1995 ) ; k. hong and n. giordano , j. phys .
cm 8 , l301 ( 1996 ) ; and k. hong , n. giordano , europhys .
lett . 36,147 ( 1996 ) . m. novak and r. sessoli , in ref .
@xcite , pp 171 - 188 ( 1995 ) ; b. barbara _ et al .
_ , j. magn .
magn . mat .
, * 140 - 144 * , 1825 ( 1995 ) ; j. r. friedman _ et al . _ ,
lett . , * 76 * , 3830 - 3833 ( 1996 ) ; l. thomas _ et al .
_ , nature * 383 * , 145 - 147 ( 1996 )
. j. m. hernandez _
et al . _ ,
* 35 * , 301 - 306 ( 1996 ) .
altshuler , b. spivak , jetp lett .
* 42 * , 447 ( 1985 ) ; s. feng , p.a .
lee , a. stone , phys .
56 * , 1960 ( 1986 ) ; s. feng , in _
`` mesoscopic phenomena in solids '' _ , ed .
altshuler et al . , ( elsevier , 1991 ) , pp .
107 - 128 .
golubev , a.d .
zaikin , phys .
lett . * 81 * , 1074 ( 1998 ) ; and /cond - mat 9804156 ; and for an opposing point of view , i.l .
aleiner , b.l .
altshuler , m.e .
gershenson , phys .
lett . * 82 * , 3190 ( 1999 ) some papers in the debate concerning the coherence experiments in ferritin @xcite include refs . @xcite ; a. garg , phys .
* 70 * , c2198 ( 1993 ) , and reply of d. d. awschalom et al .
_ , c2199 ( 1993 ) ; a. garg , phys .
lett . * 71 * , 4241 ( 1993 ) , and d. d. awschalom et al . ,
_ , c4276 ( 1993 ) ; or a. garg , science * 272 * , 425 ( 1996 ) and j. tejada , science * 272 * , 424 ( 1996 ) , and reply by d. d. awschalom et al , _ ibid .
_ , 425 ( 1996 ) ) ; n. v. prokofev , p. c. e. stamp @xcite ;
j. tejada et al .
lett . * 79 * , 1754 ( 1997 ) . the instanton calculus ( which goes back to work by langer in 1967 ) is nicely reviewed by s. coleman _ `` aspects of symmetry '' _ ( c.u.p . , 1985 ) , ch . 7 , and by l.s .
schulman , _ `` techniques and applications of path integration '' _ ( wiley , 1981 ) , ch . 29 . for a general discussion of computation , including a reprint of feynman s famous paper on quantum computation , see `` _ feynman s lectures on computation _ '' , ed .
hey , r.w .
allen ( addison - wesley , 1996 ) . | the quantum dynamics of mesoscopic or macroscopic systems is always complicated by their coupling to many `` environmental '' modes . at low @xmath0
these environmental effects are dominated by
_ localised _
modes , such as nuclear and paramagnetic spins , and defects ( which also dominate the entropy and specific heat ) .
this environment , at low energies , maps onto a `` spin bath '' model .
this contrasts with `` oscillator bath '' models ( originated by feynman and vernon ) which describe _ delocalised _ environmental modes such as electrons , phonons , photons , magnons , etc .
the couplings to @xmath1 spin bath modes are _ independent _ of @xmath1 ( rather than the @xmath2 dependence typical of oscillator baths ) , and often strong .
one can not in generalmap a spin bath to an oscillator bath ( or vice - versa ) ; they constitute distinct `` universality classes '' of quantum environment .
we show how the mapping to spin bath models is made , and then discuss several examples in detail , including moving particles , magnetic solitons , nanomagnets , and squids , coupled to nuclear and paramagnetic spin environments .
we then focus on the `` central spin '' model , which couples a central 2-level system to a background spin bath .
it is the spin bath analogue of the famous `` spin - boson '' oscillator model , and describes , eg . , the tunneling dynamics of nanoscopic and mesoscopic magnets and superconductors .
we show how to average over ( or `` integrate out '' ) spin bath modes , using an operator instanton technique , to find the central spin dynamics .
the formal manouevres involve 4 separate averages- each average corresponds physically to a different `` decoherence '' mechanism acting on the central spin dynamics .
each environmental spin has its own topological `` spin phase '' , which by interacting with the phase of the central system , decoheres it- this can happen even without dissipation .
we give analytic results for the central spin correlation functions , under various conditions .
we then describe the application of this theory to magnetic and superconducting systems .
particular attention is given to recent work on tunneling magnetic macromolecules , where the role of the nuclear spin bath in controlling the tunneling is very clear ; we also discuss other magnetic systems in the quantum regime , and the influence of nuclear and paramagnetic spins on flux dynamics in squids .
finally , we discuss decoherence mechanisms and coherence experiments in superconductors and magnets .
we show that a spin bath environment causes decoherence even in the @xmath3 limit .
control of this decoherence will be essential in the effort to construct `` qubits '' for quantum computers . |
quantum monte carlo ( qmc ) represents a promising methodology for solving electronic structure and quantum many - body problems in general .
the common thread of qmc methods is the use of stochastic sampling both to evaluate expectation values for many - body wave functions as well as to solve the stationary schrdinger equation @xcite .
in addition to being a useful tool for treating interactions in a many - body framework , qmc methods can handle many different systems ranging from atoms to molecules to extended crystals .
the obtained results for energy differences are typically within a few percent of experimental values . even on absolute scale qmc correlation energies
are remarkably accurate , reaching 90 - 95% of the exact values . the favorable scaling in the number of particles
makes qmc very promising both in applications and also in further theoretical developments .
the key approximation which hampers further increase in accuracy is the so - called fixed - node approximation which is used to overcome the notorious fermion sign problem .
the fermionic antisymmetry implies that wave functions will have both positive and negative negative values ; however , the statistical approaches such as qmc typically require non - negative distributions otherwise they become inefficient .
one means of overcoming this fundamental difficulty is to use the node ( wave function zero locus ) between the positive and the negative regions as a boundary condition in solving the schrdinger equation .
this avoids the sign problems since one finds the absolute value of the wave function with prescribed boundaries . in the case that one would know the exact nodal hypersurface , the method would provide the exact solution .
unfortunately one seldom possesses such knowledge .
therefore we have to use the best available nodes which come from optimized trial ( or variational ) wave functions .
we fix the nodal surface of the solution to be identical to the nodes of the trial function and solve the schrdinger equation with approximate nodes .
the fixed - node approximation provides an upper bound to the true energy , and in practice it provides very remarkable accuracy as mentioned above .
the goal of this work is to study the impact of increasing electronic density on the fixed - node bias .
this requires a system which shows significant error in the mean - field ( hartree - fock ) nodes but for which we know an excellent approximation to the exact node @xcite which is not too difficult to construct across the range of densities .
in particular , we focus on studying a `` be - like '' series of four electron systems in a coulombic potential with varying nuclear charge @xmath0 to @xmath1 .
we contrast the fixed - node bias of the hartree - fock vs. two - configuration trial wave functions and analyze the corresponding trends as @xmath2 increases .
let us briefly mention the key facts about the fixed - node diffusion monte carlo method ( fn - dmc ) .
consider an operator @xmath3 where @xmath4 is the hamiltonian and @xmath5 is a real parameter ( imaginary time ) .
it is straightforward to show that applying this operator to an initial trial wave function @xmath6 @xmath7 projects out the ground state @xmath8 within the given symmetry class . this projection can be formulated as a solution of the imaginary - time schrdinger equation @xmath9 where @xmath10 is the green function and @xmath11 is the vector of electron coordinates .
we have introduced the importance sampling by @xmath12 so that the desired ground state @xmath13 is obtained from the solution @xmath14 for large @xmath5 .
note that the fixed - node condition implies that @xmath13 vanishes at the nodes of @xmath6 and also that @xmath15 for any @xmath16 .
further details about the fn - dmc method can be found in ref .
@xcite .
let us consider the lowest state of four electrons in the coulomb potential with charge z ( we assume atomic units throughout ) . to aid in understanding the nature of fixed - node errors for this system , we recall the properties of the nodal domain topology of the two different trial wave functions .
the hartree - fock ( hf ) trial wave function for be is given by a slater determinant which is block diagonal in spin so that it can be broken into a product of the spin channels @xmath18 \times \det[\phi_{1s}(r_3),\phi_{2s}(r_4)]\end{aligned}\ ] ] where @xmath19 .
the nodes are defined implicitly for any trial wave function as @xmath20 let us analyze the determinant for the non - interacting orbitals first .
we can write analytically the hydrogenic functions as @xmath21 and @xmath22e^{-zr/2}\ ] ] so that we get @xmath23=0\ ] ] where @xmath24 is a function which does not vanish for any finite @xmath25 while @xmath26 it is then easy to show that the function @xmath27 is monotonous since @xmath28 for any @xmath29 . using numerically accurate hf orbitals for be , it is possible to carry out similar arrangements and to demonstrate qualitatively the same behavior . the monotonicity is important since it immediately implies that equation @xmath30=0 $ ] is equivalent to @xmath31 and also that this is the only solution .
the node of the hf / non - interacting trial function is then given by @xmath32 which clearly shows that there are @xmath33 nodal pockets @xcite .
however , it has been found some time ago that the correct number of nodal domains is two , see , for example , ref .
( for our purposes it is very instructive that the corresponding fixed - node bias in the correlation energy is significant , about 10% of the correlation energy @xcite . )
the accurate nodal surface for the @xmath34 ground state is actually remarkably well described by two configurations wave function where hf is augmented by adding @xmath35 double excitation which corresponds to a near - degeneracy effect .
this configuration state function is a sum of three slater determinant products @xmath36\det[\phi_{1s},\phi_{2p_i}]$ ] for @xmath37 as required by the @xmath34 symmetry .
the 2-configuration wave function is thus given as @xmath38 where the @xmath39 and @xmath40 are the variational coefficients .
we write the many - body wave - function as a product of an antisymmetric determinantal part , @xmath41 , times a jastrow correlation factor @xmath42 our jastrow factor includes 1- , 2- , and 3-body terms and has the form @xmath43 where the @xmath44 are one - body basis terms , the @xmath45 are two - body basis terms , the @xmath46 , @xmath47 , @xmath48 are coefficients , and the @xmath49 label electrons while capital @xmath50 labels ions .
the sums in the jastrow factor can be understood as taking into account electron - ion , electron - electron , and electron - electron - ion correlation , respectively @xcite .
further details about the correlation functions can be found in ref.@xcite . as explained above ,
we employ two types of trial functions .
the first one has the hf nodes so that @xmath51 while the second state @xmath52 our one - particle orbitals were generated using a numerical hf code .
the variational parameters to be optimized in variational monte carlo are the jastrow coefficients , @xmath53 and @xmath48 , a parameter in the basis terms @xmath44 and @xmath45 , and the determinantal weight , @xmath40 .
these were optimized by algorithms which minimize the combination of energy and its variance as described elsewhere ref.@xcite .
all qmc calculations were done using the software package qwalk @xcite .
the correlation energy is defined as customary @xmath56 where @xmath57 is the total non - relativistic energy of an atom with nuclear charge @xmath2 and @xmath58 electrons , while @xmath59 is its hartree - fock counterpart .
@xmath60 can be expanded in a laurent series as in @xcite @xmath61 we can use mller - plesset perturbation theory to analyze the first few terms . in particular , one finds that @xmath62 . the first term is a sum of the energies of occupied orbitals , @xmath63 where each orbital energy @xmath64 is proportional to @xmath65 .
the next term is equal to first order correction to this energy with the perturbing potential being the difference between the exact potential and the effective hartree - fock potential , @xmath66 , @xmath67 which will be linear in @xmath2 .
for the four electron system , @xmath68 will increase linearly with @xmath2 @xcite @xcite .
chakravorty et .
explain that this is due to the near - degeneracy effect since the single configuration hartree - fock wave function is not the correct zeroth order wave function in the non - interacting limit @xmath69 .
this deficiency means @xmath70 is not equal to @xmath71 . as such , expansions of the hf energy and the actual total energy in z@xmath72 for be and be - like atoms differ in the linear coefficient .
this results in a leading term proportional to @xmath2 in the expression for @xmath55 @xcite .
we carried out the fixed - node calculations for trial functions with @xmath73 and @xmath74 nodes for @xmath75 .
the results are listed in table [ table : corrvsdensity ] and plotted in fig .
[ fig : error ] .
the linear increase with growing @xmath2 is very clear and is quite striking considering that it covers the range from li@xmath76 through be to very highly ionized cases .
interestingly , this linear increase matches the linear increase in the correlation energy based on perturbation analysis above .
in fact , the observed slope of -0.111(1 ) is very close the analytically derived result of -0.0117 for the linear correction of the total energy in the mp2 theory @xcite . .
the squares correspond to the hf nodes while the circles correspond to the 2-configuration nodes .
the linear fit to the error from the hf nodal structure has a slope of 0.0111(1 ) .
the error bars are much smaller than the plot symbols . ]
.fn - dmc ground state energies for @xmath73 and @xmath74 wave functions compared to the estimated exact energies estimated from experiments @xcite for z=4 through 28 and extrapolation @xcite to infinite basis set for z=3 [ cols= " < , > , > , > " , ] [ table : expectr ] this is confirmed by the small fixed - node error of the @xmath0 case , less than 4% of @xmath55 , which is much smaller than the error for be which is larger than 10% of @xmath55 .
this is in line with the fact that the li@xmath72 @xmath77 electrons are highly diffuse and jastrow - type correlations play much more important role overall . one way to look at the nodes is to scan the wave function with a `` probe '' electron or pair of electrons for a given position for the rest of the particles .
one gets a view of the subspace of the 11-dimensional node which corresponds to our system .
we therefore fixed one of the spin - up electrons at @xmath78 and similarly , one of the spin - down electrons at @xmath79 .
we then scanned the space by the remaining pair of electrons and we draw the zero isosurface of the @xmath74 , see fig . [ fig : beciwf ] .
the scan shows clearly the effect of two configurations which enable the pair of the electrons to `` visit '' both inside and outside region of the two almost concentric spheres which are fused on one side creating thus a nodal `` opening . ''
this can be contrasted with the hf node which is given exactly by two ideal , concentric spheres . during the qmc walker evolution process the pair of particles
thus can not get from outside to inside
it will be forever locked in one of the ( four ) equivalent domains .
these results clearly indicate that the impact of the density on the fixed - node errors is very significant .
in particular , comparison of @xmath1 with @xmath0 is perhaps the most revealing . for @xmath1
the electron density is highly localized around the nucleus , and the hf orbitals are very close to noninteracting ones since the electron - electron interaction is much smaller than the electron - ion contribution . however , the nodal error is significant both in absolute terms ( 283 mha ) and in relative terms ( 28.6 % of @xmath55 ) .
interestingly , this system is considered `` easy '' in the sense that the one - particle contributions dominate overall . on the other hand ,
li@xmath80 exhibits nontrivial many - body correlations related to the weak anionic bonding which is usually more demanding to describe by traditional approaches based on basis sets and excitations .
in fact , the fndmc result is surprisingly good even with the hf nodes , which , as explained above , have incorrect topology .
this seems counterintuitive but can be understood once we realize that significantly lower density of li@xmath80 make the nodal errors less important .
the second interesting observation is that the two - configuration wave function is very accurate and captures nodes correctly in the range of systems , from the weakly bonded anion to the very highly ionized cations .
although it is easy to understand that the near - degeneracy effect is crucial it is interesting to see that higher excitations appear to have only marginal effect for all studied systems .
we studied the dependence of the fixed - node error on electron density . for this purpose
we have chosen the system of four electrons in a coulomb potential with varying nuclear charge @xmath2 .
this system has somewhat unique properties : the hf wave function s fixed - node bias is rather large , the hf node is invariant to varying @xmath2 , and for the be atom it was known that very accurate nodes were not too difficult to construct . expanding on these insights , our calculations have convincingly shown : \a ) the fixed - node bias increases with increasing density and appears to be predominantly linear with @xmath2 and for the case of four - electron be - like coulomb system is very closely related to the first - order correction in the mp2 theory ; \b ) the 2-configuration wave function produces highly accurate nodes across the values of @xmath2 we tested , i.e. it appears to be very consistent and basically universal for this system .
this has important implications for the further analysis of the fixed - node errors for larger and more complicated systems since the results suggest that even small fixed - node errors in the region with high electronic density could have much stronger influence on the total correlation energy recovered than large fixed - node errors in low electronic density regions .
we would like to thank the center for high performance simulations ( chips ) at north carolina state university for computational resources , and also both michal bajdich and jindrich kolorenc for their insightful suggestions and comments .
i would like to thank the physics department and graduate school of north carolina state university for continuing support .
we are supported by dmr-0804549 and oci-0904794 grants and by dod / aro and doe / lanl doe - de - ac52 - 06na25396 grants .
we acknowledge also allocations at ornl through incite and cnms initiatives as well as allocations at nsf , ncsa , and tacc centers . | we analyze the effect of increasing charge density on the fixed node errors in diffusion monte carlo by comparing fn - dmc calculations of the total ground state energy on a 4 electron system done with a hartree - fock based trial wave function to calculations by the same method on the same system using a configuration interaction based trial wave function .
we do this for several different values of nuclear charge , z. the fixed node error of a hartree - fock trial wave function for a 4 electron system increases linearly with increasing nuclear charge .
= 1 |
exploring the mechanism for the breaking of the @xmath10 electroweak gauge symmetry is a priority in high energy physics . not only are this symmetry and its breaking @xcite necessary components for the consistency of the successful standard model ( sm ) @xcite , but measurable properties of the breaking mechanism are also very sensitive to possible phenomena that have not yet been observed at collider experiments .
measuring these properties , or setting limits on them , can constrain broad classes of extensions to the sm .
a natural extension to the sm that can be tested with higgs boson search results at the fermilab tevatron collider is the addition of a fourth generation of fermions with masses much larger than those of the three known generations @xcite .
while fits to precision electroweak data favor a low - mass higgs boson in the sm , the addition of a fourth generation of fermions to the sm modifies the fit parameters such that a much heavier higgs boson is allowed @xcite .
currently available data are consistent with a higgs boson of mass 900 gev with suitably chosen fourth generation fermion masses @xcite
. precision measurements of the @xmath11 boson decay width @xcite exclude models in which the fourth neutrino mass eigenstate has a mass less than 45 gev .
if the neutrino masses are very large , however , a fourth generation of fermions is not excluded .
one consequence of the extra fermions is that the @xmath12 coupling is enhanced by a factor of roughly three relative to the sm coupling @xcite .
the reason for this is that the lowest - order @xmath12 coupling arises from a quark loop .
the top quark contribution is the largest due to its large coupling with the higgs boson . in the limit @xmath13 where @xmath14 is the fourth - generation quark mass
, the higgs boson coupling cancels the mass dependence for each of the three propagators in the loop , and the contribution to the @xmath12 coupling becomes asymptotically independent of the masses of the two fourth - generation quarks .
each additional fourth - generation quark then contributes as much as the top quark , and the @xmath12 coupling is thus enhanced by a factor @xmath15 of approximately three .
the production cross section will be enhanced by a factor of @xmath16 . for @xmath17 near the low end of our search range , @xmath18 gev , the @xmath19 production cross section is enhanced by roughly a factor of nine relative to the sm prediction .
this factor drops to approximately 7.5 near the upper end of the search range , @xmath20 gev , assuming asymptotically large masses for the fourth - generation quarks .
the reason for this drop is that the denominator of the enhancement factor , the sm cross section , has a larger contribution from the sm top quark as @xmath17 nears @xmath21 .
the partial decay width for @xmath22 is enhanced by the same factor as the production cross section . however , because the decay @xmath22 is loop - mediated , the @xmath23 decay continues to dominate for higgs boson masses @xmath24 gev .
we consider two scenarios for the masses of the fourth - generation fermions . in the first scenario , the `` low - mass '' scenario , we set the mass of the fourth - generation neutrino @xmath25 gev , and the mass of the fourth - generation charged lepton @xmath26 gev , in order to evade experimental constraints @xcite and to have the maximum impact on the higgs boson decay branching ratios . in the second scenario , the `` high - mass '' scenario , we set @xmath27 tev , so that the fourth - generation leptons do not affect the decay branching ratios of the higgs boson . in both scenarios ,
we choose the masses of the quarks to be those of the second scenario in ref .
@xcite , that is , we set the mass of the fourth - generation down - type quark to be @xmath28 gev and the mass of the fourth - generation up - type quark to be @xmath29
. the other mass spectrum of ref .
@xcite chooses @xmath30 gev , resulting in slightly larger predictions for @xmath31 and nearly identical decay branching ratios .
we use the next - to - next - to - leading order ( nnlo ) production cross section calculation of ref .
@xcite , which builds on the nnlo sm calculations of refs .
@xcite .
the cdf and d0 collaborations have searched for the sm higgs boson in the decays @xmath23 and @xmath32 using all sm production processes : @xmath19 , @xmath33 , @xmath34 , and vector - boson fusion ( vbf ) @xcite .
the results of these searches for the sm higgs boson can not be used directly to constrain fourth - generation models , as the @xmath12 coupling is enhanced but the @xmath35 and @xmath36 couplings are not , and the signal acceptances and the backgrounds in the multiple analysis channels differ for the various production modes . previously , the cdf and d0 collaborations have published searches for the process @xmath1 , also neglecting the @xmath37 , @xmath38 , and vbf signal contributions @xcite .
this paper is an update of ref .
@xcite , which provides a fourth - generation interpretation of the @xmath1 searches , and a model - independent limit on @xmath39 . here
we update the searches detailed in ref .
@xcite with those using 8.2 fb@xmath5 from cdf @xcite and 8.1 fb@xmath5 from d0 @xcite and include additional channels @xcite .
interpretations of sm higgs boson searches within the context of a fourth generation of fermions have been performed by atlas @xcite and cms @xcite , which exclude @xmath40 gev and @xmath41 gev , respectively .
we present new limits on @xmath42 in which the @xmath19 production mechanism is considered as the unique signal source .
these limits are compared to models for higgs boson production in which the @xmath12 coupling is enhanced by the presence of a single additional generation of fermions . in this comparison ,
the decay branching ratios of the higgs boson are also modified to reflect changes due to the fourth generation relative to the sm prediction . while the decays of the heavy quarks and leptons may include @xmath43 bosons in the final state
, we do not include these as additional sources of signal .
the branching ratios for @xmath23 are calculated using hdecay @xcite modified to include fourth - generation fermions @xcite .
the modified higgs branching ratio to @xmath44 is multiplied by the cross section @xcite to predict the fourth - generation enhanced @xmath45 production rate .
we include as an approximation to the uncertainties on the decay branching ratios the relative uncertainties predicted for the sm branching ratios due to uncertainties in @xmath46 , @xmath47 , and @xmath48 presented in ref .
@xcite .
the branching ratio for the decay @xmath32 is also considerable at large masses , having a peak at around @xmath49 gev and rising above @xmath50 gev . in order to include a search for @xmath2
, we introduce the assumption that @xmath51 remains as predicted by the sm , even though the production cross section is enhanced .
this assumption holds true in fourth - generation models as predicted by the fourth - generation version of hdecay mentioned above . in setting limits on @xmath42
, we include the process @xmath2 assuming that its signal yield scales in the same way as the @xmath44 channel .
the event selections are similar for the corresponding cdf and d0 analyses .
both collaborations select events with large and two oppositely charged , isolated leptons , targeting the @xmath52 signal in which both @xmath43 bosons decay leptonically .
d0 selects events containing electrons and/or muons , dividing the data sample into three final states : @xmath53 , @xmath54 , and @xmath55 .
each final state is further subdivided according to the number of jets in the event : 0 , 1 , or 2 or more ( `` 2 + '' ) jets . however , in the fourth - generation interpretation and the model - independent @xmath19 limits , due to the low s / b in the 2-jet multiplicity bin , only the 0 and 1-jet multiplicity bins were taken into account
. decays involving tau leptons are included in two orthogonal ways .
a dedicated analysis ( @xmath56 ) using 7.3 fb@xmath5 of data studying the final state involving a muon and a hadronic tau decay , is included @xcite .
final states involving other tau decays and mis - identified hadronic tau decays are included in the @xmath53 , @xmath54 , and @xmath55 final state analyses .
the cdf analysis separates opposite - sign candidate events into five non - overlapping channels .
events are classified by their jet multiplicity ( 0 , 1 , or @xmath57 2 ) , and the 0 and 1 jet channels are further divided according to whether both leptons are in the central part of the detector or if either lepton is in the forward part of the detector .
the presence of neutrinos in the final state prevents event - by - event reconstruction of the higgs boson mass and thus other variables are used for separating the signal from the background .
for example , the angle @xmath58 in signal events is smaller on average than that in background events , the missing transverse momentum is larger , and the total transverse energy of the jets is lower .
the cdf analyses use neural - network ( nn ) outputs as the final discriminants ; the list of network inputs includes likelihood ratio discriminant variables constructed from matrix - element probabilities @xcite .
the d0 @xmath53 , @xmath54 , and @xmath55 final state channels use random forest boosted decision tree ( bdt ) outputs as the final discriminants while the @xmath56 channel uses neural networks .
both experiments use separate nns and bdts for the different jet multiplicities to distinguish the @xmath19 signal from the backgrounds for each of the test masses , which are separated by increments of 5 or 10 gev .
cdf introduces a search for @xmath59 using 8.2 fb@xmath5 of data @xcite .
the discriminant variable is the reconstructed mass of the four leptons @xmath60 .
the version of this search used for the standard model interpretation also includes the expected yield for @xmath37 , @xmath38 , and @xmath61 production , but these production modes are assumed not to contribute when interpreting the results in this paper .
d0 also includes for the first time channels in which one of the @xmath43 bosons in the @xmath62 process decays leptonically and the other decays hadronically @xcite .
electron and muon final states are studied separately , each with 5.3 fb@xmath5 of data .
random forests are used for the final discriminants . both cdf and
d0 conduct all searches included in this paper in the range @xmath63 gev .
the details of the signal and background estimations and the systematic uncertainties are provided in refs .
we set limits on @xmath6 as a function of @xmath17 .
we use the same two statistical methods employed in ref .
@xcite , namely the modified frequentist ( cl@xmath64 ) and bayesian techniques , in order to study the consistency of the results .
each method is applied at each test mass to calculate an observed upper limit on @xmath42 .
pseudo - experiments drawn from systematically varied background - only predictions are used to compute the limits we expect to obtain in the absence of a signal . correlated systematic uncertainties are treated in the same way as they are in ref .
the sources of correlated uncertainty between cdf and d0 are the total inelastic @xmath65 cross section used in the luminosity measurement , the sm diboson background production cross sections ( @xmath66 , @xmath67 , and @xmath68 ) , and the @xmath69 and single top quark production cross sections .
instrumental effects such as trigger efficiencies , lepton identification efficiencies and misidentification rates , and the jet energy scales used by cdf and d0 remain uncorrelated .
to minimize the degrading effects of systematics on the search sensitivity , the signal and background contributions are fit to the data observations by maximizing a likelihood function over the systematic uncertainties for both the background - only and signal+background hypotheses @xcite .
when setting limits on @xmath6 , we do not include the theoretical uncertainty on the prediction of @xmath6 in the fourth - generation models since these limits are independent of the predictions .
when setting limits on @xmath17 in the context of fourth - generation models , however , we include the uncertainties on the theoretical predictions as described below . since both the cdf and the d0 analyses categorize events into separate channels based on the number of reconstructed jets , we include the systematic uncertainties on each jet category according to the procedure recommended by ref .
this procedure treats the factorization and renormalization scale uncertainties as independent in the inclusive @xmath19 , @xmath70 or more jets , and @xmath71 or more jets calculations .
we propagate these uncertainties through to the calculations of the exclusive @xmath72 jets , @xmath73 jet , and @xmath74 or more jets calculations required to predict the yields in the fourth - generation models .
when setting a limit on @xmath39 , we do not include a theoretical uncertainty on the prediction of the quantity that we are constraining experimentally .
nonetheless , we do require theoretical input for the differential distributions of the number of jets in @xmath19 production .
we therefore use only the factorization and renormalization scale uncertainties on the @xmath73 or more jets and @xmath74 or more jets inclusive calculations when setting a limit on the inclusive process .
we follow the pdf4lhc prescription @xcite for evaluating the parton distribution function ( pdf ) uncertainties on the signal predictions .
qcd scale uncertainties that affect the pdf predictions are included as part of the total scale uncertainty , added linearly to the other components of the scale uncertainty .
the remaining component of the pdf uncertainty is treated as uncorrelated with the qcd scale uncertainty .
the scale choice affects the @xmath75 spectrum of the higgs boson when produced in gluon - gluon fusion , and this effect changes the acceptance of the selection requirements and also the shapes of the distributions of the final discriminants .
the effect of the acceptance change is included in the calculations of ref .
@xcite and ref .
@xcite , as the experimental requirements are simulated in these calculations .
the effects on the final discriminant shapes are obtained by reweighting the @xmath75 spectrum of the higgs boson production in our monte carlo simulation to higher - order calculations . the monte carlo signal simulation used by cdf and d0
is provided by the lo generator pythia @xcite which includes a parton shower and fragmentation and hadronization models .
we reweight the higgs boson @xmath75 spectra in our pythia monte carlo samples to that predicted by hqt @xcite when making predictions of differential distributions of @xmath19 signal events . to evaluate the impact of the scale uncertainty on our differential spectra
, we use the resbos @xcite generator , and apply the scale - dependent differences in the higgs boson @xmath75 spectrum to the hqt prediction , and propagate these to our final discriminants as a systematic uncertainty on the shape , which is included in the calculation of the limits .
the combined limits on @xmath42 are listed in table [ tab : limits ] for both the cl@xmath64 and the bayesian methods , and are shown in fig .
[ fig : xslimits ] along with the fourth - generation theory predictions for the high - mass and low - mass scenarios .
the uncertainty bands shown on the low - mass theoretical prediction are the quadrature sum of the mstw 2008 @xcite 90% c.l .
pdf uncertainties and the factorization and renormalization scale uncertainties from table 1 of ref .
the limits calculated with the two statistical methods agree within 10% for all higgs boson mass hypotheses .
an excess is seen in the neighborhood of @xmath76 gev .
this excess is localized to the d0 @xmath77 search ( including charge conjugate final states ) @xcite .
the excess is not as pronounced in the sm - optimized search from d0 @xcite .
combination with cdf , which sees no significant excess at @xmath76 gev @xcite results in a total local significance somewhat in excess of two standard deviations .
the trials factor , or look - elsewhere effect , further dilutes the significance , in this case by a large factor due to the large range of @xmath17 explored .
we do not claim that this excess is statistically significant . in order to set limits on @xmath17 in these two scenarios ,
we perform a second combination , including the uncertainties on the theoretical predictions of @xmath78 due to scale and pdf uncertainties at each value of @xmath17 tested .
the resulting limits are computed relative to the model prediction , and are shown in fig .
[ fig:4glimits ] for the low - mass scenario , which gives the smaller excluded range of @xmath17 . in this scenario ,
we exclude at the 95% c.l . a sm - like higgs boson with a mass in the range 124 286 gev . using the median limits on @xmath6 , expected in the absence of a signal to quantify the sensitivity , we expect to exclude the mass range 120 267 gev . in the high - mass scenario , which predicts a larger @xmath79 at high @xmath17 than that predicted in the low - mass scenario , we exclude at the 95% c.l . the mass range 124 300 gev and expect to exclude the mass range 120 290 gev . the upper edge of the observed exclusion range in the high - mass scenario is determined by the fact that we did not perform the search for higgs bosons with mass exceeding 300 gev .
the tevatron combined observed ( solid black lines ) and median expected ( dashed black lines ) 95% c.l .
upper limits on @xmath6 .
the shaded bands indicate the @xmath80 standard deviation ( s.d . ) and @xmath81 s.d .
intervals on the distribution of the limits that are expected if a higgs boson signal is not present . also shown the prediction for a fourth - generation model in the low - mass and high - mass scenarios , 4 g ( low mass ) and 4 g ( high mass ) respectively .
the hatched areas indicate the theoretical uncertainty from pdf and scale uncertainties .
the lighter curves show the high - mass theoretical prediction .
, scaledwidth=80.0% ] the tevatron combined observed ( solid black lines ) and median expected ( dashed black lines ) 95% c.l .
upper limits on the ratio of the higgs boson production cross section to that predicted by the 4 g ( low mass ) model ; the uncertainties in the signal prediction are included in the limit .
also shown is the prediction of the signal rate in the high - mass scenario , divided by that of the low - mass scenario .
, scaledwidth=80.0% ] in summary , we present a combination of cdf and d0 searches for the @xmath1 and @xmath2 processes and set an upper limit on @xmath6 as a function of @xmath17 , assuming the standard model ratio of @xmath51 .
we compare these limits with the prediction of the minimal sm with a sequential fourth generation of heavy fermions added on , and exclude at the 95% c.l . the higgs boson mass range @xmath82 gev , with an expected exclusion of 120 267 gev .
we thank the fermilab staff and the technical staffs of the participating institutions for their vital contributions .
this work was supported by doe and nsf ( usa ) , conicet and ubacyt ( argentina ) , cnpq , faperj , fapesp and fundunesp ( brazil ) , crc program , cfi , nserc and westgrid project ( canada ) , cas and cnsf ( china ) , colciencias ( colombia ) , msmt and gacr ( czech republic ) , academy of finland ( finland ) , cea and cnrs / in2p3 ( france ) , bmbf and dfg ( germany ) , ministry of education , culture , sports , science and technology ( japan ) , world class university program , national research foundation ( korea ) , krf and kosef ( korea ) , dae and dst ( india ) , sfi ( ireland ) , infn ( italy ) , conacyt ( mexico ) , nsc(republic of china ) , fasi , rosatom and rfbr ( russia ) , slovak r&d agency ( slovakia ) , ministerio de ciencia e innovacin , and programa consolider - ingenio 2010 ( spain ) , the swedish research council ( sweden ) , swiss national science foundation ( switzerland ) , fom ( the netherlands ) , stfc and the royal society ( uk ) , and the a.p .
sloan foundation ( usa ) .
higgs , phys .
* 12 * ( 1964 ) 132 ; _ idem _ , phys .
* 13 * ( 1964 ) 508 ; _ idem _ , phys . rev . * 145 * ( 1966 ) 1156 ; f. englert and r. brout , phys .
* 13 * ( 1964 ) 321 ; g. s. guralnik , c. r. hagen and t. w. b. kibble , phys .
* 13 * ( 1964 ) 585 .
b. holdom , _ et al . _ , pmc phys .
a * 3 * , 4 ( 2009 ) .
h. j. he , n. polonsky and s. f. su , phys .
d * 64 * , 053004 ( 2001 ) .
g. d. kribs , t. plehn , m. spannowsky , and t. m. p. tait , phys .
d * 76 * , 075016 ( 2007 ) .
m. baak _ et al .
_ , arxiv:1107.0975 [ hep - ph ] ( 2011 ) .
the aleph , delphi , l3 , opal , sld collaborations , the lep electroweak working group , and the sld electroweak and heavy flavor groups , phys . rept . * 427 * , 257 ( 2006 ) .
e. arik , o. cakir , s. a. cetin , and s. sultansoy , acta phys .
b * 37 * , 2839 ( 2006 ) . c. anastasiou , r. boughezal , and e. furlan , arxiv:1003.4677 [ hep - ph ] ( 2010 ) .
achard _ et al . _
[ l3 collaboration ] , phys .
b * 517 * , 75 ( 2001 ) . c. balazs
, j. huston , i. puljak , phys .
d * 63 * 014021 ( 2001 ) .
+ c. balazs and c .- p .
yuan , phys .
b * 478 * 192 - 198 ( 2000 ) .
+ qing - hong cao and chuan - ren chen , phys .
d * 76 * 073006 ( 2007 ) .
a. d. martin , w. j. stirling , r. s. thorne , and g. watt , eur .
j. c * 64 * , 653 ( 2009 ) . | 0.3 in we combine results from searches by the cdf and d0 collaborations for a standard model higgs boson ( @xmath0 ) in the processes @xmath1 and @xmath2 in @xmath3 collisions at the fermilab tevatron collider at @xmath4 tev . with 8.2 fb@xmath5 of integrated luminosity analyzed at cdf and 8.1 fb@xmath5 at d0 , the 95% c.l .
upper limit on @xmath6 is 1.01 pb at @xmath7 gev , 0.40 pb at @xmath8 gev , and 0.47 pb at @xmath9 gev . assuming the presence of a fourth sequential generation of fermions with large masses , we exclude at the 95% confidence level a standard - model - like higgs boson with a mass between 124 and 286 gev .
+ _ preliminary results _ 0.5 in |
our understanding of the local universe and the local galaxy population originates primarily from the all - sky photographic schmidt surveys and the established catalogues of bright galaxies derived from them , such as the catalogue of galaxies and clusters of galaxies @xcite , the morphological catalogue of galaxies @xcite , the uppsala general catalogue of galaxies @xcite , the eso / uppsala catalogue @xcite , the southern galaxy catalogue @xcite , the catalogue of principal galaxies @xcite , the edinburgh / durham southern galaxy catalogue @xcite , the apm catalogue @xcite , the third reference catalogue of bright galaxies @xcite and the supercosmos sky survey @xcite .
while these catalogues have provided invaluable information and insight , uncertainty remains as to their completeness , particularly for low surface brightness and compact galaxies @xcite .
in addition there are concerns as to the photometric accuracy ( e.g. * ? ? ?
* ) , the susceptibility to scale errors @xcite , plate - to - plate variations @xcite and dynamic range .
these photographic - based catalogues have been the starting point for numerous spectroscopic surveys aimed at measuring the local space density of galaxies ( i.e. the local galaxy luminosity function ) .
the space density of galaxies is our fundamental census of the local contents of space and therefore a crucial constraint for models of galaxy formation ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
if the imaging catalogues are in omission and/or photometrically inaccurate then regardless of the completeness of the spectroscopic surveys our insight into the galaxy population will be incomplete and most likely biased against specific galaxy types . over the past two decades
there have been numerous estimates of the local galaxy luminosity function ( e.g. eep , @xcite ; mt stromlo / apm , @xcite ; autofib , @xcite ; esp , @xcite ; ssrs2 , @xcite ; durham / ukst , @xcite ; sdss - cd , @xcite ; 2dfgrs , @xcite ) and of the three - parameter schechter function used to represent it @xcite . typically the surveys agree broadly on the faint end slope ( @xmath18 , @xmath19 ) but show a marked variation in the characteristic luminosity ( @xmath20 , @xmath21 per cent ) and normalization ( @xmath6 , @xmath22 per cent ) .
the uncertainties in the schechter parameters result in an uncertainty of @xmath23 per cent in the local luminosity density , @xmath24 .
this uncertainty is usually expressed as the normalization problem which has been somewhat overshadowed by the more notorious faint blue galaxy problem @xcite .
the latter describes the inability of basic galaxy number count models to predict the numbers of galaxies seen at faint magnitudes ( @xmath25 mag ) in the deep pencil beam ccd - based surveys ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the lesser known normalization problem describes the inability of number count models to explain the galaxy counts even at bright magnitudes ( @xmath26 mag ) by as much as a factor of @xmath27 ( see discussions in * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in many ways the normalization problem is the more fundamental : while luminosity evolution , cosmology and/or dwarf galaxies can be , and have been , invoked in varying mixtures to explain the faint blue galaxy problem ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , none of these can be used to resolve the normalization problem . in the past
the problem was typically circumvented by renormalizing the number count models to the range @xmath28 mag ( e.g. * ? ? ?
* ; * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the justification was that the bright galaxy catalogues , on which the luminosity function measurements are based , are shallow and therefore susceptible to local clustering .
however the crucial normalization range typically occurs at the faint limit of the photographic surveys ( where the photometry and completeness are more likely to be a problem ) and at the bright end of the pencil beam ccd surveys ( where statistics are poor ) . while convenient , the clustering explanation overlooks two more worrisome possibilities : gross photometric errors and/or gross incompleteness in the local catalogues .
if either of these two latter explanations play a part this will have important consequences for the new - generation spectroscopic surveys , namely the 2dfgrs @xcite and the sdss @xcite .
the input catalogue of the 2dfgrs is an extensively revised version of the photographic apm survey ( which is known to show a peculiar steepening in its galaxy counts at bright magnitudes , @xcite ) , with zero - point and scale - error corrections from a variety of sources including the 2mass @xmath29-band survey and the data presented in this paper ( see @xcite for details ) . in the case of the sdss which leaps forward in terms of dynamic range , uniformity and wavelength coverage the effective exposure time is relatively short ( @xmath30 s ) and the isophotal detection limit is comparable to that of the photographic surveys .
hence while issues of photometric accuracy should be resolved the question mark of completeness may remain . to address the above problems within a single , well - defined dataset we require a survey that is reasonably deep and yet has a large enough solid angle to provide accurate and statistically significant galaxy counts over the crucial normalization range .
furthermore , the survey s photometry must be accurate and its completeness high , i.e. it must probe to low surface brightnesses .
[ surveys ] shows a number of imaging surveys in terms of their sky coverage and magnitude range . dashed and solid lines indicate photographic and ccd - based surveys , respectively . typically the faint surveys are ccd - based while the local surveys are photographic ( with the notable and recent exception of the sdss - edr , @xcite ) .
while the ccd surveys make significant improvements in surface brightness and magnitude limits their sky coverage is small .
it is only very recently that large ccd mosaics such as the wide field camera ( wfc , * ? ? ?
* ) and the sdss instrument @xcite have been constructed that now allow a large area of sky to be surveyed within a realistic time frame . in this paper
we present the millennium galaxy catalogue ( mgc , sections [ data][sellimits ] ) .
the mgc represents a new medium - deep , wide - angle galaxy resource , which firmly connects the local and distant universe within a single dataset ( cf .
fig [ surveys ] ) . in section
[ counts ] we produce the galaxy number counts spanning the range @xmath5 mag .
we then focus on the normalization problem by comparing our counts over the range @xmath31 mag to the predictions of a number of local luminosity function estimates in section [ lumfs ] .
our counts provide stringent constraints on the normalization of the luminosity function and hence on the local luminosity density .
our conclusions are given in section [ conclusions ] .
the 2dfgrs and sdss will essentially supersede all previous redshift surveys and therefore it is important to verify their photometric accuracy and completeness on as large a scale as possible .
we will provide a detailed comparison of the 2dfgrs and sdss - edr imaging catalogues with the mgc in a companion paper @xcite .
a more long - term aim of the mgc project is to provide structural information on the galaxy population around the crucial normalization point ( @xmath32 mag ) .
it is ironic that since the advent of the hubble space telescope we have a greater understanding of the morphological mix of galaxies at faint magnitudes than at bright magnitudes .
for example , @xcite and @xcite published morphological galaxy counts spanning the range @xmath33 mag , yet no reliable morphological galaxy counts at brighter magnitudes exist .
consequently , no accurate local morphological luminosity functions exist ( compare , for example , the conflicting results of @xcite and @xcite ) and the evolution of the different morphological types can not be accurately constrained .
the mgc will enable us to remedy this situation as it allows morphological classification and the extraction of structural parameters to @xmath34 mag .
the data and catalogues presented in this paper are publically available at http://www.roe.ac.uk/@xmath35jol / mgc/.
all data frames where taken using the wide field camera ( wfc ) .
the wfc is mounted at prime focus on the @xmath36 m isaac newton telescope ( int ) situated at la palma .
the wfc is a mosaic of four 4k@xmath372k thinned eev ccds with a smaller 2k@xmath372k loral ccd which is used for auto - guiding .
each of the science ccds measures @xmath38 pixels with a pixel scale of @xmath39 arcsec / pixel this gives a total sky coverage of @xmath40 deg@xmath1 per pointing .
the four science chips are arranged as shown in fig .
[ outline ] .
full details of the wfc are provided at http://www.ast.cam.ac.uk/@xmath35wfcsur/technical.html ( see also * ? ? ?
the data constituting the mgc comprise @xmath41 overlapping fields forming a @xmath42 arcmin wide equatorial strip from @xmath43 to @xmath44 ( j2000 ) .
the observations were taken during @xmath45 observing runs , 1999 march 1516 , 1999 april 1617 , 1999 june 613 and 2000 march 26april 4 . each field
was observed for a single @xmath46 s exposure through a kitt peak national observatory @xmath2 filter .
field @xmath47 is centered on ra @xmath48 , dec @xmath49 ( j2000 ) and field @xmath41 is centered on ra @xmath50 , dec @xmath49 ( j2000 ) .
hence each field is offset from the previous by @xmath51 arcmin along the equatorial great circle .
[ outline ] shows the survey outline for the first three pointings .
note the substantial overlap between neighbouring fields .
the survey region was chosen because it is contained within both the 2dfgrs @xcite and the sdss - edr @xcite regions , thus providing redshifts and colours for the brighter galaxies and allowing a detailed check of the photometry and completeness of these surveys .
all observations were taken during dark or grey time through variable conditions .
the seeing ranged from @xmath52 to @xmath53 arcsec with the median seeing at @xmath54 arcsec .
the air masses ranged from @xmath55 to @xmath56 .
[ obsstat ] shows a summary of the general observing conditions across the survey .
as much of the data were collected during clear but non - photometric nights it was necessary to dedicate a single pristine photometric night ( 2000 march 30 ) to obtaining suitable calibration data at various stages along the survey strip . in total @xmath57 mgc fields were observed during the photometric night . of these ,
six were only @xmath58 s exposures as they had already been observed previously .
these observations were interspersed with @xmath59 s observations of standard stars spanning a wide range in airmass .
the standard stars where taken from the @xcite standard areas sa98 , sa101 , sa104 and sa107 .
some science frames were later found to be of too poor a quality to be useful and these were re - observed : the first eight fields were replaced with two @xmath60 s exposures each and field @xmath61 is a single @xmath62 s exposure .
all the preliminary data reduction flat - fielding , bias correction and astrometric calibration was done by the cambridge astronomy survey unit ( casu ) and full details of this process are provided by @xcite . briefly , a number of bias frames are collected each night and the median is subtracted from the data . all data
( including flat - fields ) are corrected for a known non - linearity .
a twilight flat - field is taken during evening and morning twilight ( when possible ) and a median flat - field derived for that particular run is divided into each data frame . after this process
an initial astrometric calibration is made to the hst guide star catalogue .
finally the frames are matched to the apm catalogue which itself is calibrated onto the tycho-2 astrometric system .
[ mgc036 ] shows the final reduced image for one of our pointings , field 36 .
it illustrates problems with satellite trails , ccd defects , gaps between ccds , etc . to assess the final astrometric accuracy we compared the positions of doubly detected objects in regions where neighbouring fields overlap ( cf .
[ outline ] ) .
the overlap regions are of size @xmath63 deg@xmath1 and each contains @xmath64 objects in the range @xmath65 mag , , @xmath66 refers to the kron magnitude , uncorrected for galactic extinction ( cf .
section [ photometry ] ) . ] where detections can be easily and confidently matched .
[ radecoff ] shows the median positional differences for these objects for each overlap region as well as the overall @xmath67ra and @xmath67dec distributions .
we find that both of these distributions have an rms of @xmath68 arcsec .
note , however , that they are slightly but significantly offset from zero which is most likely due to residual radial distortions .
as mentioned previously , four @xcite standard star fields were observed at a range of air masses throughout the course of the photometric night . for each observation of each standard star
we computed a zero - point @xmath69 where @xmath2 was taken from @xcite and @xmath70 is the flux of the star as measured from the data .
we then fitted these zero - points with a double linear function in airmass ( @xmath71 , where @xmath72 is the zenith distance ) , and colour , @xmath73 where @xmath74 is again taken from @xcite . in fig .
[ stdzp ] we show the data and the fit .
the residuals have an rms of @xmath75 mag and show no obvious trend with airmass , colour , @xmath2 or time of observation .
note , however , that a systematic error of @xmath76 mag was needed to achieve an acceptable fit . from the lower panel of fig .
[ stdzp ] we can see that for several stars multiple observations of the same star give consistently high or low results .
this may indicate that these stars are slightly variable or that the errors on the @xcite photometry have been underestimated . for each mgc field
we then computed a theoretical zero - point @xmath77 , which is expected to be correct only for those fields observed during the photometric night .
we extracted objects as described in section [ objex ] and identified duplicate detections in regions where any two images overlapped .
using only objects with @xmath78 mag and stellaricity @xmath79 ( cf .
section [ class ] ) we computed , for each overlap region , the median of the magnitude differences of the double detections , @xmath80 , as well as an error on the median , @xmath81 . in fig .
[ dm_fn](a ) we show @xmath82 for all the overlap regions using the initial , theoretical zero - points . a linear least - squares routine was then used to adjust the zero - points of the non - photometric fields in order to minimize the quantity @xmath83 where the sum goes over all overlap regions .
the process of object extraction , matching and zero - point adjustment was then repeated until a stable solution was reached ( four iterations ) .
since the zero - points of the photometric fields are held fixed at their theoretical values , the above procedure assumes that the observing conditions were perfectly stable throughout the photometric night .
one can derive an estimate of the real - life error on the photometric zero - points , @xmath84 , by comparing the scatter of @xmath82 with @xmath81 for those overlap regions that only involve photometric fields .
we found @xmath85 .
thus we modified the above calibration procedure by including the zero - points of the photometric fields in the parameters to be fitted and adding the additional constraints that they must lie close to their theoretical values .
in other words we now minimize the quantity @xmath86 where the first sum again runs over all overlap regions and the second sum runs over all photometric fields . in fig .
[ dm_fn](c ) we show @xmath82 for all the overlap regions using the final zero - points .
[ dm_fn](d ) shows the histogram of the individual @xmath87 values .
the width of this distribution indicates an internal photometric accuracy of @xmath88 mag for objects in the range @xmath78 mag . due to the paucity of photometric fields beyond mgc field @xmath89 ( cf .
top panel of fig .
[ obsstat ] ) the absolute calibration in the second half of the survey is less reliable than in the first : up to field @xmath89 the median of the @xmath87 distribution is @xmath90 mag , which is consistent with zero , but beyond field @xmath89 it is @xmath91 mag .
this is needed to reconcile the photometric fields @xmath92 and @xmath93 with the photometric fields at field numbers @xmath94 . given the above calibration process the relationship between an object s @xmath66 magnitude and its landolt @xmath2 magnitude is given by : @xmath95 where the error on the colour term is @xmath96 .
object extraction was performed using extractor , which is the starlink adapted version of sextractor developed by @xcite .
sextractor initially derives a background map by first defining a grid over the image and then passing a median filter ( set to a size of @xmath97 pixel ) over each pixel within the grid cell .
the local sky within the grid cell is then taken as the mode or @xmath98-clipped mean of the pixel distribution within the cell .
finally , a background map is constructed via a bicubic spline interpolation over these points .
we opted for the largest possible mesh size ( @xmath99 pixel ) to minimize the smoothing out of any extended low surface brightness features .
after convolving the image with a filter , sextractor detects objects as groups of connected pixels above the uniform surface brightness threshold of @xmath100 mag arcsec@xmath4 .
once an object has been detected sextractor redetects the object using @xmath51 different detection thresholds , which are exponentially spaced between the peak flux value of the object and the detection threshold .
if at any level the object breaks up into two or more disconnected subcomponents , each containing at least @xmath59 per cent of the total flux of the object , then the object is deblended .
this multithresholding is ideally suited for galaxy extraction because no assumptions concerning the shape of the object are being made @xcite .
the photometry was performed using sextractor with a constant analysis isophote of @xmath3 mag arcsec@xmath4 to provide a uniformly processed catalogue .
four of the various types of magnitudes provided by sextractor are included in the mgc : an isophotal magnitude ( iso ) , a corrected isophotal magnitude ( isocor ) , an adaptive aperture magnitude ( kron ) and a best magnitude ( best ) .
the isocor magnitude is calculated by correcting the iso magnitude for the fraction of the flux outside the limiting isophote assuming a gaussian profile ( see * ? ? ? * and references therein for details ) .
kron magnitudes ( i.e. elliptical apertures of 2.5 kron radii , @xcite ) are known to underestimate the fluxes in perfect exponential profiles by @xmath101 mag and in de vaucouleur profiles by @xmath102 mag .
nevertheless , they have been shown to be the most robust to variations in redshift , bulge - to - disc ratio , isophotal limit and seeing @xcite .
the best magnitude is taken to be the kron magnitude except in crowded regions where the isocor magnitude is used instead .
all objects are individually corrected for galactic extinction using the dust extinction maps provided by @xcite and adopting @xmath103 . from now on we refer to the best magnitudes before and after extinction correction as @xmath66 and @xmath104 , respectively .
[ plate ] shows a random selection of galaxies in the range @xmath105 to @xmath106 mag . as a result of the substantial overlap regions ( cf .
[ outline ] ) the catalogue contains many duplicate objects .
these were used in previous sections to verify the astrometry and to calibrate the photometry .
we now remove duplicate detections by imposing ra limits for each field , effectively splitting each overlap region in half .
( note that the ra limits do not apply to objects detected on ccd 3 . ) for a small number of objects , all lying very close to an ra limit , this procedure did not remove one of the duplicate detections , because the two detections happened to lie on either side of the limit .
these cases were fixed by hand . at this stage
the catalogue contains a total of @xmath107 objects to @xmath108 mag .
the catalogue comprises galaxies , stars and various unwanted objects and artefacts such as satellite trails , ccd defects , cosmic rays , diffraction spikes , asteroids and spurious noise detections . as a starting point for classification
we used the stellaricity parameter provided by sextractor , which is produced for each object by an artificial neural network ( ann ) that has been extensively trained to differentiate between stars and galaxies @xcite .
the input of the ann consists of nine object parameters ( eight isophotal areas and the peak intensity ) and the seeing .
the output consists of a single number , called stellaricity , which takes a value of @xmath47 for stars , @xmath109 for galaxies and intermediate values for more dubious objects . fig .
[ stell ] shows the number of objects as a function of stellaricity and @xmath66 . at @xmath110
mag the stellaricity distribution is clearly bimodal with almost all values at the extremes and so star galaxy separation is trivial . at fainter magnitudes the star
galaxy separation requires more effort .
hence we now define two catalogues : mgc - bright , which contains all objects with @xmath111 mag , and mgc - faint which contains the rest . for mgc - bright
we adopted the following classification strategy .
brighter than @xmath112 mag we classified all objects with stellaricity @xmath113 as stars .
this low stellaricity cut was used because the objects in the upper left - hand corner of fig .
[ stell ] are , in fact , flooded stars . for the rest of mgc - bright
we classified all objects with stellaricity @xmath114 as stars .
this is a ` natural ' value to adopt because the stellaricity distribution rises sharply from @xmath115 to @xmath116 .
all objects so far classified as non - stellar were then inspected visually and classified into one of the following categories : galaxy , star , asteroid , satellite trail , cosmic ray , ccd defect , diffraction spike or spurious noise detection .
incorrectly deblended galaxies were then repaired by hand , but their original catalogue entries were retained and classified as obsolete .
asteroids were verified using images from the supercosmos sky survey ( sss , * ? ? ? * ) .
we also inspected all objects with fwhm , semimajor or semiminor axis less than the seeing ( these turned out to be primarily cosmic rays and ccd defects ) . during this
process galaxies were also assigned one of three quality classes , @xmath117 , depending on the level of : ( i ) contamination by ccd defects , satellite trails , cosmic rays and diffraction spikes ; ( ii ) blending with a similarly bright object ; ( iii ) missing light due to a ccd edge and ( iv ) failed background estimation due to nearby bright objects .
the breakdown of the final mgc - bright catalogue is shown in table [ brighttab ] and is a good indication of the level of contamination in purely automated galaxy catalogues .
.breakdown of mgc - bright ( @xmath111 mag ) . [ cols="<,>,>",options="header " , ] @xmath118 the three quoted errors are due to : ( i ) statistics and lss , ( ii ) the correlated errors on @xmath119 and @xmath18 and ( iii ) the uncertainty in @xmath120-corrections .
@xmath121 we have reduced @xcite s zero - point uncertainty from @xmath101 to @xmath122 mag because we have found that the 2dfgrs and mgc photometries agree to this level once the zero - point offset has been applied .
we have also excluded the error induced by @xmath120-corrections as we treat this error separately and consistently for all surveys in section [ fitphi ] . in table
[ lumftab ] we list the lf parameters from the 2dfgrs @xcite , sdss - cd @xcite , esp @xcite , cs @xcite , durham / ukst @xcite , mt stromlo / apm @xcite , ssrs2 @xcite and nog @xcite surveys .
all @xmath119 values have been converted to the @xmath104 system using equations ( [ survcoleq ] ) below .
these were derived from equation ( [ coloureq ] ) and the colour equations of @xcite , @xcite ( sdss - cd ) , @xcite ( cs ) , @xcite ( ssrs2 ) , @xcite and @xcite ( nog ) .
we also correct for a known zero - point offset in the 2dfgrs photometry : mgc@xmath1232dfgrs@xmath124 mag .
similarly , combining @xcite s @xcite @xmath125mgc@xmath126sdss - edr@xmath127 mag with @xcite s @xmath128sdss - edr@xmath1232dfgrs@xmath129 mag gives @xmath125mgc@xmath1262dfgrs@xmath130 mag .
] @xmath131 we assume @xmath132 @xcite and @xmath133 @xcite .
where available we have used the parameters for the currently favoured @xmath134 cosmology ( 2dfgrs , sdss - cd and cs ) .
the nog parameters do not depend on the cosmological model @xcite , the ssrs2 used @xmath135 and the rest used @xmath136 .
we have attempted to correct the @xmath119 values of these latter surveys to a @xmath137 cosmology using @xmath138,\ ] ] where @xmath139 and @xmath140 are the luminosity distance and median redshift of the survey , respectively . since all of these surveys fixed @xmath6 by using one of the @xcite estimators of the mean galaxy density we also corrected their @xmath6 values using @xmath141 where d@xmath142/d@xmath143 is the comoving volume element . since the 2dfgrs , sdss - cd and cs have all derived their lf parameters for several different cosmological models we can use these to test this simple correction procedure . in all cases we find good agreement ( well within the quoted error bars ) between the transformed and measured values .
the final corrected @xmath104 lfs are shown in fig .
[ lumfmod](a ) . to model the counts we use ( @xmath144 , a @xmath145-correction of @xmath146 and an @xmath147-correction of @xmath148 $ ] . at @xmath149 this combination matches well the @xmath120-correction given by @xcite ( their fig .
8) which was derived using @xcite models to match the colour redshift trend seen in the 2dfgrs ( using sdss - edr colours ) . in principle
, for each of the surveys in table [ lumftab ] we should use the same @xmath150-correction as was used in the derivation of the schechter parameters of that survey or else correct for the use of a different @xmath120 .
however , in several cases different @xmath145-corrections were used for different galaxy types , sometimes interpolating between types , and hence a mean correction is not readily available .
[ lumfmod](b ) shows how the counts predicted by the various lfs compare with the mgc - bright counts ( corrected for lss ) .
clearly , the data prefer some models over others .
we have performed a goodness - of - fit test for the predicted counts in the range @xmath151 mag ( using the full covariance matrix of the counts ) and list the resulting probabilities in column 7 of table [ lumftab ] .
the model counts based on the durham / ukst , mt stromlo / apm and ssrs2 lfs are all significantly too low , while the sdss - cd model counts are too high .
the other models provide good fits .
these conclusions do not change if we compare the models to the uncorrected counts . from fig .
[ lumfmod](b ) it is clear that the main difference between the various predicted counts and the data lies in their normalization , @xmath152 , and not their shape . as discussed in section [ introduction ] it is precisely this uncertainty in @xmath152 that is commonly referred to as the ` normalization problem ' . for a given set of schechter parameters what is the uncertainty in @xmath152 induced by the error of each of the parameters ?
in the general case there is no analytic formula relating @xmath152 with the schechter parameters .
hence the ` euclidean ' case must serve as a guideline . ignoring binning effects we have @xmath153 . using the intersurvey rms of each parameter as an indication of its uncertainty we find @xmath154 and similar or worse values for the individual surveys .
hence we identify the errors in both @xmath6 and @xmath119 as the main causes of the uncertainty in @xmath152 .
what are the sources of these uncertainties ?
in addition to the difficulty of reliably calibrating the photometry of large photographic surveys @xcite showed that the derived lf parameters depend on the limiting isophote and the photometric method ( e.g. isophotal , corrected , total , etc . ) .
they concluded that when using isophotal magnitudes for a limiting isophote of @xmath155 mag arcsec@xmath4one might expect errors in @xmath119 of up to @xmath156 mag , in @xmath6 of up to @xmath157 per cent and in @xmath18 of up to @xmath158 .
hence photometric uncertainties contribute substantially to the normalization problem .
although the underestimation of the magnitudes due to the limiting isophote will cause some galaxies to fall below the limiting magnitude of a survey , it also causes the derived volume over which such galaxies can be seen to be _
underestimated_. in addition , fig .
[ appbbd ] ( together with visibility theory ) shows that at @xmath159 mag apparently no significant galaxy population below @xmath155 mag arcsec@xmath4 exists . in other words , at these surface brightness limits it is typically a case of missing light from the outer isophotes rather than missing galaxies : @xcite found only @xmath160 per cent of galaxies missing from the 2dfgrs due to their low surface brightness . however , blended and unresolved objects take the total incompleteness of the 2dfgrs to @xmath161 per cent , which may be indicative of photographic surveys in general .
lss will clearly affect all but the very largest surveys .
@xcite found that the @xmath162 deg@xmath1 overlap region between the sdss - edr and the 2dfgrs was overdense by @xmath163 per cent relative to the full 2dfgrs .
similarly they found a @xmath164 per cent difference ( to @xmath165 ) between their @xmath166 deg@xmath1 ngp and @xmath167 deg@xmath1 sgp regions .
hence it is clear that much shallower surveys like the ssrs2 ( @xmath168 mag ) can not accurately measure @xmath6 even if they have a very large angular size . on the other hand , due to the filamentary structure of the universe even deep surveys are susceptible to lss if they extend only over a small solid angle .
finally , we remark that each of the surveys listed in table [ lumftab ] uses different @xmath145-corrections , and only the 2dfgrs use @xmath147-corrections
. indeed , evolutionary effects , which were ignored by @xcite , are the reason why the sdss - cd normalization is so high . when evolution is included ( or when normalizing to the sdss - edr counts instead of using the method of @xcite ) the sdss - cd normalization agrees much better with the 2dfgrs results @xcite .
although both @xmath119 and @xmath6 appear to contribute equally to the uncertainty in @xmath152 it is customary ( and sensible ) to fix @xmath119 and use @xmath152 to constrain @xmath6 alone instead of a combination like @xmath169 . over the range @xmath31 mag
the mgc arguably provides the most reliable number count data in existence and hence it provides very reliable constraints on @xmath6 .
we now determine @xmath170 as a function of @xmath119 and @xmath18 by fitting the model counts for a given combination of @xmath119 and @xmath18 to the corrected mgc - bright counts over the range @xmath171 mag ( i.e. the mgc-2dfgrs counts at @xmath172 mag ) , where @xmath6 is the free parameter .
we use the full covariance matrix for the fits .
[ phistar ] shows the result as a contour plot in the @xmath119@xmath18 plane .
in particular , we derive the @xmath170 values appropriate for the lfs listed in table [ lumftab ] .
the revised values and their associated probabilities are listed in columns 7 and 8 of that table .
[ lumfmod](c ) and ( d ) show the resulting lfs and their predicted counts after revision .
all surveys now give comparable and reasonable probabilities .
note , however , that the variation of @xmath6 among the different surveys after renormalization is no less than before .
this is due to the fact that the counts constrain the combination @xmath169 and that the variation in @xmath173 is comparable to that of @xmath6 before renormalization ( as seen from equation [ nerror ] ) . in table
[ lumftab ] we also list the errors on @xmath170 induced by various sources .
first , we list the statistical error derived from the @xmath174 fit .
this error includes remaining uncertainties due to lss because we have used the full covariance matrix of the counts in the fit .
secondly , we list the errors due to the uncertainties in @xmath119 and @xmath18 , which can be essentially read off from fig .
[ phistar ] . here
we take into account that the errors on @xmath119 and @xmath18 are generally highly correlated by assuming a correlation coefficient of @xmath175 @xcite .
finally , we have estimated the errors resulting from uncertainties in the @xmath120-correction .
this error is slightly more subtle than the previous ones .
for example , increasing @xmath120 has the effect of moving the contours in fig .
[ phistar ] to the left and up but it also moves the lfs in a similar direction , thus again decreasing the original change in @xmath170 for a given survey .
@xcite estimated errors on their @xmath120-correction by demanding statistical consistency between the lfs derived from their high and low-@xmath143 samples .
since our @xmath120 is essentially identical to theirs we also adopt their error of @xmath176 per cent .
we estimate the resulting @xmath170 error for a given survey by first adding @xmath177 to its @xmath119 value and then re - deriving @xmath170 while using the increased ( or decreased ) @xmath120 . as in section [ modcounts ]
we thus approximate the effect on @xmath119 by evaluating @xmath178 at the median redshift of the survey .
comparison with @xcite s @xmath120 induced error on @xmath119 indicates that this procedure underestimates the effect on @xmath119 , and hence we overestimate the error on @xmath170 .
we also conservatively ignore any effect on @xmath18 , which is difficult to gauge .
taking @xmath179 mag and using the revised lf normalizations we calculate the luminosity density , @xmath180 .
note that we convert @xmath181 to @xmath182 .
the calculated values of @xmath183 are listed in the last column of table [ lumftab ] , where the errors include all the @xmath119 , @xmath18 and @xmath170 uncertainties while taking into account the correlation between the errors of @xmath119 and @xmath18 ( with a correlation coefficient 0.75 as above ) as well as the virtually perfect correlations between the second error component of @xmath170 and the errors of @xmath119 and @xmath18 and between the @xmath120 errors on @xmath119 and @xmath170 .
we find a weighted mean @xmath15 luminosity density of @xmath184 mpc@xmath17 .
note that the 2dfgrs and sdss - cd , which must be considered the most reliable lfs , give a slightly higher value of @xmath185 mpc@xmath17 . in the previous section we minimized @xmath174 as a function of @xmath6 only while treating @xmath119 and @xmath18 as fixed parameters .
@xmath6 only affects the normalization of the predicted counts but not their shape which , in principle , also contains information .
however , for the lfs listed in table [ lumftab ] the fits are already statistically acceptable ( cf . column 9 of table [ lumftab ] )
and hence we can not expect to derive useful limits on any other parameters . nevertheless , in fig .
[ phistar ] we mark as grey - shaded regions those combinations of @xmath119 and @xmath18 where the adjustment of @xmath6 does not result in an acceptable fit ( at the @xmath186 and @xmath187 per cent confidence levels ) , i.e. where the shape of the predicted counts disagrees with the data . as expected these limits can not exclude any of the measured values .
here we have presented a detailed description of the millennium galaxy catalogue ( mgc ) , a deep ( @xmath188 mag arcsec@xmath4 ) wide ( @xmath0 deg@xmath1 ) survey along the equatorial strip from @xmath189 to @xmath190 .
we have demonstrated that the internal photometric accuracy of the mgc is @xmath191 mag and that the astrometric accuracy is @xmath192 arcsec in both ra and dec .
using sextractor we have derived a source catalogue containing over 1 million objects spanning the range @xmath193 mag .
all non - stellar detections brighter than @xmath34 mag have been visually inspected and the objects repaired where necessary .
we have taken care to exclude objects from regions where the photometry is likely to be erroneous , resulting in a robust and clean estimation of the galaxy number counts over the range @xmath194 mag .
these data finally connect the faint pencil beam ccd surveys of the past decade to the local universe .
the selection boundaries of the mgc are well defined and to @xmath34 mag we are demonstrably robust to star galaxy separation and low- and high - surface brightness concerns .
we contest that the mgc galaxy number counts in this range are the state of the art , superseding all previous intermediate number count data .
we use the counts to test various estimates of the galaxy luminosity function and find that many of them predict counts where the normalizations are inconsistent with our observations . in fig .
[ phistar ] we present the best - fitting value of @xmath6 as a function of @xmath119 and @xmath18 . in table
[ lumftab ] we list the appropriate @xmath6 values for a number of popular @xmath2-band luminosity function estimates . from these revised values
we constrain the @xmath15-band luminosity density of the local universe for each of these luminosity functions .
we find @xmath195 mpc@xmath17 .
the 2dfgrs and sdss - edr consistently give a slightly higher value of @xmath196 mpc@xmath17 .
we wish to thank all those who participated in observing the mgc fields .
the data were obtained through the isaac newton group s wide field camera survey programme .
the isaac newton telescope is operated on the island of la palma by the isaac newton group in the spanish observatorio del roque de los muchachos of the instituto de astrofsica de canarias .
we also thank casu for their data reduction and astrometric calibration . | the millennium galaxy catalogue ( mgc ) is a @xmath0 deg@xmath1 , medium - deep , @xmath2-band imaging survey along the celestial equator , taken with the wide field camera on the isaac newton telescope .
the survey region is contained within the regions of both the two degree field galaxy redshift survey ( 2dfgrs ) and the sloan digital sky survey early data release ( sdss - edr ) .
the survey has a uniform isophotal detection limit of @xmath3 mag arcsec@xmath4 and it provides a robust , well - defined catalogue of stars and galaxies in the range @xmath5 mag . here
we describe the survey strategy , the photometric and astrometric calibration , source detection and analysis , and present the galaxy number counts that connect the bright and faint galaxy populations within a single survey .
we argue that these counts represent the state of the art and use them to constrain the normalizations ( @xmath6 ) of a number of recent estimates of the local galaxy luminosity function .
we find that the 2dfgrs , sdss commissioning data ( cd ) , eso slice project , century survey , durham / ukst , mt stromlo / apm , ssrs2 , and nog luminosity functions require a revision of their published @xmath6 values by factors of @xmath7 , @xmath8 , @xmath9 , @xmath10 , @xmath11 , @xmath12 , @xmath13 and @xmath14 , respectively .
after renormalizing the galaxy luminosity functions we find a mean local @xmath15 luminosity density of @xmath16 mpc@xmath17 .
[ firstpage ] catalogues galaxies : general galaxies : luminosity function , mass function galaxies : statistics cosmology : observations . |
characterization of the initial distribution of density perturbations and understanding their subsequent evolution into present - day structures are among the central aims of cosmology .
particular attention has been focused on the low - redshift distribution of matter over large scales , which can probe the linear physics of the early universe and therefore test models for the origin of large - scale structure , such as inflation .
the @xmath3-point correlation functions - point cfs are valid statistics only if the universe is homogeneous on large scales .
we invoke the assumptions of large - scale homogeneity and isotropy , so that the cfs depend only relative distances and are independent of orientation . ]
( cfs ) have become the most widely used statistical tools to quantify the observed distribution of matter , in part because they can be easily related to the predictions of such models and thus discriminate between them . in principle , recovering the current distribution of density perturbations is straightforward ; one simply maps the three - dimensional distribution of mass with a sufficiently large redshift survey .
existing redshift surveys , however , typically contain relatively small numbers of objects , have inconvenient shapes , and necessarily suffer from redshift - space distortions , all of which make it difficult to obtain precise cf measurements , particularly for the higher - order functions .
another possibility is to map the angular distribution of galaxies as projected onto the surface of the sky .
angular surveys can sample larger volumes of space and contain greater numbers of objects .
furthermore , the effects of redshift distortions need not be considered in an angular survey .
thus , if the redshift distribution is well constrained , as is often the case for large angular surveys , angular cfs can in practice provide far greater statistical power and more insight into the evolution of clustering , than sparser , redshift - survey data ( groth & peebles 1977 ; fry & seldner 1982 ; jing , mo , & brner 1991 ; baugh & efstathiou 1993 ; cappi & maurogordato 1995 ; cress 1996 ; maddox , efstathiou & sutherland 1996 ; gaztaaga & baugh 1998 ; cress & kamionkowski 1998 ) . to date , angular - clustering studies have focussed primarily on the angular two - point correlation function ( 2pcf ) , @xmath4 , but with advent of new , deep surveys at various wavelengths ( radio , optical , ir , x - ray ) and more powerful computing resources , increasing attention is being focused on higher - order cfs .
these are particularly important in assessing primordial gaussianity ( fry 1984 ; goroff 1986 ; bernardeau 1994 ; fry & scherrer 1994 ; gaztaaga & bernardeau 1998 ; white 1998 ; scoccimarro 1998 ; scoccimarro , couchman , & frieman 1998 ) , since to leading order they , unlike the 2pcf , reflect the inherently nonlinear gravitational amplification of the initial fluctuations .
furthermore , they contain information about the cosmological parameters ( bouchet 1992 ; juszkiewicz , bouchet , & colombi 1993 ; gaztaaga , croft , & dalton 1995 ; bouchet 1995 ; catelan 1995 ; martel 1995 ; jing & brner 1997 ; szapudi 1998 ; kamionkowski & buchalter 1998 ) and about the nonlinear bias between the observed and underlying density distributions ( jensen & szalay 1986 ; szalay 1988 ; fry & gaztaaga 1993 ; frieman & gaztaaga 1994 ; fry 1994 ; jing 1997 ; matarrese , verde , & heavens 1997 ) , which can not be obtained from @xmath4 .
in particular , the angular three - point correlation function ( 3pcf ) , @xmath5 , and relatedly , the normalized angular skewness , @xmath6 ( proportional to @xmath7 evaluated at zero lag , for a distribution smoothed over angular scale @xmath8 ) , are the lowest - order intrinsically nonlinear statistics and can therefore yield important constraints on models of structure formation .
moreover , if different populations are differently biased relative to the underlying mass distribution then measuring the angular 3pcfs of these populations can provide multiple , complementary constraints .
measurements of @xmath1 , as well as the normalized angular 3pcf , @xmath9 $ ] , over quasilinear ( ql ) scales have yielded results which appear to agree loosely with the predictions of the so - called hierarchical model ( peebles 1975 ; peebles 1980 ) for gravitational evolution of gaussian initial conditions ( peebles & groth 1975 ; fry & seldner 1982 ; jing & zhang 1989 ; tth , hollsi , & szalay 1989 ; borgani , jing , & plionis 1992 ; gaztaaga 1994 ; bernardeau 1995 ; cappi & maurogordato 1995 ; gaztaaga , croft , & dalton 1995 ; gaztaaga & bernardeau 1998 ; jing & brner 1998 ; magliocchetti 1998 ) .
it is obvious , however , that improved measurements [ from surveys such as the sloan digital sky survey ( sdss ; loveday 1998 ) ] and more sophisticated theoretical modeling of angular cfs are required . in particular , there has been little theoretical work to date on the detailed behavior of angular cfs beyond the two - point function , despite the practical advantages they can offer . in this paper , we present a calculation of the full angular 3pcf .
we have compared our results with theirs and find that they agree for some of the models we have considered .
] , @xmath7 , as well as @xmath1 , for an arbitrary , biased , tracer - mass distribution in flat and open cold - dark - matter ( cdm ) models , assuming gaussian initial conditions .
the calculation , based on leading - order pt results , is restricted to the ql regime , where @xmath10 , and further invokes the small - angle approximation .
we take into account such factors as the explicit dependence on the cosmological parameters , the shape and linear evolution of the cdm power spectrum , and the form of redshift distribution function .
we consider both the redshift distribution of the apm galaxy survey , as well as that of a survey with a mean redshift of @xmath2 .
we also examine the effect of linear and nonlinear bias , which , through an extension of the fry ( 1996 ) bias - evolution model to the case of an arbitrary expansion history , are allowed to evolve in time .
following bkj , we define the bias parameters , @xmath11 , by an expansion of the unsmoothed tracer - mass field in terms of the unsmoothed dark - matter field , rather than by relating the smoothed fields , as is often done ( e.g. , fry & gaztaaga 1993 ) . while many of the results we obtain are similar to those found in a related paper ( buchalter & kamionkowski 1999 , hereafter bk99 ) dealing with the spatial 3pcf and skewness , understanding the detailed behavior of angular statistics in practice can be equally , if not more , illuminating .
we extend previous derivations of @xmath1 ( bernardeau 1995 ) to include a scale - dependent , leading - order correction which arises in the case of nonlinear bias , as defined for the unsmoothed fields , and which becomes large for positive effective power - spectrum indices .
this behavior , in principle , may allow better constraints on the linear- and nonlinear - bias parameters on the basis of large - scale skewness measurements alone , or at least differentiate between the smoothed and unsmoothed biasing scenarios .
for plausible models , we find that @xmath1 is relatively insensitive to the adopted cosmology , as compared with its stronger dependence on the biasing scheme . in general , the presence of linear bias tends to flatten and reduce the scale dependence of @xmath1 , while a nonlinear bias tends to produce a relative increase .
the full angular 3pcf , @xmath7 , shows similar sensitivities to the bias parameters , but can also depend significantly on the cosmological model , and in particular on the shape of the power spectrum .
we find that the configuration dependence of @xmath7 is in general more complex than the simple hierarchical model suggests , and , since angular statistics probe clustering over a range of redshifts , is also sensitive to evolutionary effects .
these properties can be used to discriminate between various models which can yield nearly degenerate predictions for @xmath1 , or for the spatial 3pcf at @xmath12 , such as evolving versus non - evolving bias models , or open versus flat cosmological models .
neglecting the full geometric and evolutionary variation of @xmath7 throws away valuable information which might , in practice , be harder to obtain from three - dimensional data .
in several instances , we illustrate how accounting for time dependences in , e.g. , the power spectrum or the bias , can impact the resulting calculations , especially in the case of high - redshift surveys . our theoretical results , in general , should be comparable to current and future measurements of angular clustering .
in 2 we outline the calculations used and present results for the normalized angular skewness . in
3 we derive predictions for the full angular 3pcf and in 4 we summarize our conclusions . since the calculations are greatly simplified by working initially in fourier space ,
we first obtain the angular bispectrum the two - dimensional projection of the spatial bispectrum and then derive the angular 3pcf from this .
a derivation of the equation that relates the spatial and angular bispectra is presented in the appendix .
this result is essentially a fourier - space generalization of limber s equation for the 3pcf .
since angular cfs involve projections of their full spatial counterparts , our derivations must involve integrals over the line of sight , which will in general depend on the adopted cosmological model .
we shall consider open and flat friedmann - robertson - walker ( frw ) cosmologies with a possible cosmological constant , so that the scale factor , @xmath13 , satisfies the friedmann equations , @xmath14,\end{aligned}\ ] ] where @xmath15 is the present nonrelativistic matter density in units of the critical density , @xmath16 is the contribution of the cosmological constant to the total present energy density , @xmath17 is the present value of the hubble parameter , and a dot denotes a derivative with respect to time .
the scale factor is chosen such that @xmath18 , where @xmath19 @xmath20 km s@xmath21 mpc@xmath21 .
if we take our position as the origin , @xmath22 , then the angular - diameter distance to an object at redshift @xmath23 is given by @xmath24 where @xmath25 , @xmath26 , @xmath27 , respectively , for open , flat , and closed geometries .
for an einstein - de sitter universe , @xmath28 .
the distance to the horizon is given by @xmath29 .
we wish ultimately to derive an expression for the angular 3pcf of a distribution of tracer masses ( e.g. , galaxies , quasars , clusters , radio sources , etc . ) , @xmath30 , where @xmath31 represents two - dimensional coordinates on the sky , @xmath32/ \overline{\sigma}$ ] is the fractional perturbation to the unsmoothed tracer - mass density field , @xmath33 , and angular brackets denote an average over direction in the sky , with fixed values for the distances @xmath34 .
though statistical homogeneity and isotropoy ensure that the 3pcf will only depend on the relative angular separations between the three points , measuring @xmath7 for all possible geometric configurations can nonetheless be a daunting task . for this reason
, many studies have focussed instead on the normalized angular skewness , @xmath35 , obtained from the moments of counts in cells smoothed with an effective angular scale half - width @xmath8 ( it is implicitly assumed that @xmath36 , where @xmath3 is the number of discrete counts ) .
this one - point statistic is more readily calculable than the full angular 3pcf , while still preserving information about the overall scale - dependence of @xmath7 . in deriving the following results , we implicitly make use of the small - angle approximation in assuming that a small patch of sky can be treated using a fourier expansion ( peebles 1980 ) . in this approximation ,
the component of the three - dimensional wave vector along the line of sight is taken to be negligible compared to the orthogonal components , so that the angular cfs depend only on the latter .
we later comment on this assumption .
we define @xmath37 using a two - dimensional spherical top hat window function , so that in terms of the fourier components , @xmath38 , of the unsmoothed two - dimensional density field , we have @xmath39 where @xmath40 is a two - dimensional wave vector with magnitude @xmath41 , @xmath42 is the fourier transform of the two - dimensional spherical top - hat window function , and @xmath8 is an angular smoothing radius .
we can then obtain the angular 2pcf , as measured using counts in cells , @xmath43 we define @xmath44 where the angular , or projected , power spectrum of the tracer mass , @xmath45 ( usually written in terms of @xmath46 s in papers on the cosmic microwave background ) , can be related to the full tracer - mass power spectrum , @xmath47 , by a fourier - space analog of limber s equation ( kaiser 1992 ) , @xmath48 here , @xmath49 is a @xmath50-space selection function for the tracer mass , normalized such that @xmath51 . we assume the fractional perturbation to the _ unsmoothed _ three - dimensional tracer mass density field , @xmath52 ,
may be expanded in terms of the local perturbation to the unsmoothed , underlying matter field , @xmath53 , via @xmath54 where @xmath55 is the linear bias term , @xmath56 the first nonlinear term , etc . , and @xmath57 is itself written as @xmath58 where @xmath59 , @xmath60 is the linear solution , and @xmath61 characterizes the leading - order departure from the gaussian initial conditions . the linear solution for the underlying spatial density contrast has the separable form @xmath62 so that fluctuations evolve simply as the linear growth factor , @xmath63 .
this in turn implies that , to leading order , the spatial and time dependence of the power spectrum can likewise be factorized . for frw cosmologies ,
the growth factor is given ( as a function of redshift ) by @xmath64 ^ 3}. \label{d}\ ] ] for an einstein de - sitter universe @xmath63 is simply the scale factor , @xmath65 .
assuming the unsmoothed three - dimensional tracer - mass density contrast to be related to the underlying distribution via equation ( [ dexp ] ) , we can write the leading - order result for the full linear power spectrum of the tracer mass , @xmath66 where @xmath67 is the primordial power spectrum , @xmath68 is the overall amplitude , and @xmath69 is a model - dependent transfer function . substituting equations ( [ pkt ] ) , ( [ pp2 ] ) , and ( [ pp ] ) into ( [ varpi ] ) , taking @xmath70 , and @xmath71 , we arrive at the angular variance ( the 2pcf at zero lag ) , @xmath72 where the @xmath55 term is left inside the @xmath50 integral to allow for the possibility of bias evolution .
the counts - in - cells angular 3pcf is given by @xmath73 the angular , or projected , bispectrum , @xmath74 , is defined via @xmath75 and is related to the full three - dimensional bispectrum , @xmath76 , by @xmath77 ( see appendix ) .
the full bispectrum is given , to leading order in pt , by @xmath78 \nonumber \\ & & + \frac{b_2}{{b_1}^2 } \biggr\ } \;\ ; + \;\ ; \mbox{(cyc . ) } , \label{bf2}\end{aligned}\ ] ] ( fry 1984 ; goroff 1986 ; matarrese , verde , & heavens 1997 ) where @xmath79 is the angle between @xmath80 and @xmath81 , and @xmath82 is a function of the expansion history , equal to @xmath83 for an einstein - de sitter universe ( peebles 1980 ) and differing from this value only slightly for other reasonable choices of the density parameters ( bouchet 1992 ; bernardeau 1994 ; bouchet 1995 ; catelan 1995 ; martel 1995 ; scoccimarro 1998 ; kamionkowski & buchalter 1998 ) .
note that the expression for @xmath76 includes the dependence on @xmath56 , since the leading - order result for the 3pcf includes second - order terms in the expansion of equation ( [ dexp ] ) .
evaluating equation ( [ askew ] ) with @xmath84 , @xmath85 , taking @xmath86 to lie in the @xmath87 direction , and substituting equations ( [ bp1 ] ) , ( [ bp2 ] ) , and ( [ bf2 ] ) we obtain , after a little algebra , an expression for the area - averaged skewness , @xmath88 + \frac{b_2}{{b_1}^2 } \right ) , \label{askewness}\end{aligned}\ ] ] where one integral vanishes under the requirement @xmath89 , and a factor of 3 arises from symmetry considerations applied to the 2 cyclic permutations in equation ( [ bf2 ] ) . noting that @xmath90 , we can evaluate the @xmath79 integrals by using the summation theorems for bessel functions ( gradshteyn & ryzhik 1980 ; bernardeau 1995 ) which yield : @xmath91 where @xmath92 , so that @xmath93
. using the above results gives @xmath94 - \frac{1}{2}\frac{(1-\mu)}{b_1 } \w2(x_1)\w2(x_2 ) \nonumber \\
& & + \frac{b_2}{b_1 ^ 2 } \sum_{j=0}^{\infty } ( 2j+1)\w2_{2j+1}(x_1)\w2_{2j+1}(x_2 ) \biggr\ } .
\label{askewness2}\end{aligned}\ ] ] we re - emphasize that our results in general are based on leading - order pt and are thus restricted to the ql regime , i.e. , scales large enough so that the rms density contrast fluctuations are small compared with unity .
numerical simulations and observations both confirm that higher - order nonlinear corrections have little impact on spatial statistics on ql scales ( szalay 1988 ; tth , hollsi , & szalay 1989 ; gott , gao , & park 1991 ; fry , melott , & shandarin 1993 ; jain & bertschinger 1994 ; scoccimarro 1998 ) . by comparing with @xmath95-body results , gaztaaga & bernardeau ( 1998 ) test the validity of pt as applied to angular statistics and find good agreement on scales @xmath96 , with the details depending on the shape of the power spectrum .
in addition , simulations show that ql scales can still obey leading - order pt even when smaller scales have become fully nonlinear ( bouchet & hernquist 1992 ) , and further that the predictions of ql pt may hold even on scales where the rms fluctuation exceeds unity ( bernardeau 1994 ; baugh , gaztaaga , & efstathiou 1995 ; fry , melott , & shandarin 1995 ; szapudi 1998 ) .
our results have also relied on the use of the small - angle approximation .
the fourier integrals used to calculate @xmath97 and @xmath7 , however , extend over all values of @xmath41 , including small values for which this approximation breaks down , but which nonetheless may contribute significant weights to the integrands on degree scales .
moreover , the small - angle approximation might appear to go against the assumption of quasilinearity ; in practice , the degree to which these two assumptions are mutually plausible will in general depend on the selection function in question . despite these considerations , our results for @xmath1 and @xmath0 on scales near @xmath98
do agree well with data from observations ( e.g. , gaztaaga 1994 ; see figure [ fig : s3_apm ] ) and @xmath95-body simulations ( frieman & gaztaaga 1999 ) on these scales . bernardeau ( 1995 ) investigates the validity of the small - angle approximation for models with different redshift selection functions , @xmath99 , and finds good agreement with numerical results for @xmath1 on scales near and below @xmath98 .
gaztaaga & bernardeau ( 1998 ) find that the small - angle approximation can yield reasonable agreement with numerical results for @xmath1 out to the @xmath100 scale .
verde ( 1999 ) perform an exact calculation for the angular 3pcf using a spherical - harmonic decomposition , and a more complete comparison and assessment of the small - angle results against the exact results are presented therein .
equations ( [ varpizero ] ) and ( [ askewness2 ] ) are one - point results which , for a given choice of @xmath15 , @xmath16 , @xmath55 , and @xmath56 , depend only on the smoothing radius , @xmath8 , and the form of the power spectrum .
if we assume , for computational simplicity , a scale - free power spectrum , @xmath101 [ @xmath102 , then we can calculate an expression for @xmath1 where the @xmath26 integrals can be separated from the @xmath50 integration and evaluated analytically ; combining equations ( [ varpizero ] ) and ( [ askewness2 ] ) , we obtain @xmath103 + 3\frac{b_2}{b_1 ^ 2 } \left [ 1 + \delta(n ) \right ] \right\ } , \label{s3}\ ] ] where @xmath104 ^ 2},\ ] ] and @xmath105 ^ 2}{\left [ \int_0^{\infty } dx \ : x^{n-1 } j_1 ^ 2(x ) \right]^2}. \label{delta}\ ] ] note that while the term in curly brackets in equation ( [ s3 ] ) depends only very weakly on @xmath15 through @xmath82 , the @xmath106 term depends on the cosmological model both through the presence of the linear growth factor , @xmath107 ( which is often neglected ) , as well as through the explicit dependence of @xmath50 on the expansion history [ see equation ( [ w ] ) ] . equations ( [ s3])([delta ] ) are valid for @xmath3 in the range @xmath108 ; the individual terms in the series for @xmath109 can be evaluated explicitly ( watson 1966 ) and we find that @xmath110 for @xmath111 , but diverges for @xmath112 , as illustrated in figure [ fig : delta ] .
the bracketed @xmath3-dependent terms in equation ( [ s3 ] ) arise from the effects of smoothing the contiguous field , @xmath52 , defined in equation ( [ dexp ] ) . in particular , since smoothing at a fixed physical mixes different mass scales , the higher - order moments for smoothed fields will generally contain additional terms reflecting the scale dependence of the rms fluctuation ( bernardeau 1994 , 1995 ) .
the @xmath113 case thus recovers the no - smoothing result for @xmath1 ( bernardeau 1995 ) , since the rms fluctuation in this case is independent of scale [ see equation ( [ varpizero ] ) ] . though the filter is designed to separate out the nonlinear , small - scale fluctuations , an unbounded scale - free power spectrum with @xmath112 produces so much small - scale power that the fluctuations induced by smoothing over nonlinear scales become comparable to or greater than the large - scale linear perturbations , leading to the divergence in @xmath109 seen in the figure .
equation ( [ s3 ] ) differs from the known result for @xmath1 assuming constant , nonlinear bias ( bernardeau 1995 ) only in that it contains the @xmath109 term .
the difference is attributable to the fact that previous authors usually define the bias parameters using the form of equation ( [ dexp ] ) , but relating the _ smoothed _ , rather than the unsmoothed , observed and underlying density fields .
such a definition is technically nonlocal , only fixing the linear- and nonlinear - bias parameters at the chosen smoothing scale . while it can be shown that the `` smoothed '' and `` unsmoothed '' linear - bias parameters are identical
, the value of the smoothed nonlinear - bias parameter will , unlike the unsmoothed @xmath56 defined in equation ( [ dexp ] ) in the limit of continuous fields , depend on the smoothing scale . comparing the result for @xmath1 with those of fry & gaztaaga ( 1993 ) and bernardeau ( 1995 )
, we can infer that the smoothed nonlinear - bias parameter will vary as @xmath114 $ ] . in the limit of no smoothing ( @xmath113 ) ,
the two parameters are identical , as expected , but can differ dramatically on large scales ( see 2.2 ) .
similar correction terms , arising from nonlinear bias , are expected for higher - order angular moments as well .
such terms might be used to discriminate between the smoothed and unsmoothed bias pictures , and possibly distinguish between linear and nonlinear bias based on the skewness alone , as discussed in 2.2 and in bk99 . unless stated otherwise , it will be hereafter understood that `` bias '' shall refer to the unsmoothed bias prescription defined by equation ( [ dexp ] ) .
while the power spectrum for any viable cosmological model will not be given by a simple power law , it can be shown , using the results of bernardeau ( 1995 ) and gaztaaga & bernardeau ( 1998 ) , that an exact result for @xmath1 can be obtained for arbitrary power spectra , using a properly redefined _ effective _ index ( scoccimarro 1998 ) .
this result , analogous to that in the spatial case ( bernardeau 1994 ) , is strictly valid , however , only for linear biasing , since it fails to account for the above - mentioned large - scale variation associated with @xmath56 . since we wish to consider nonlinear biasing , and because , unlike @xmath1 , no corresponding exact result can be obtained for the full angular 3pcf in the case of arbitrary power spectra , even for linear biasing , the results of this paper are based entirely on numerical integration of the appropriate equations .
many of these will involve highly oscillatory integrands , making them difficult to evaluate ; we have performed checks on the numerical accuracy of our results , and found them to be good to within a few percent . for realistic cdm models ,
we employ a transfer function given by @xmath115 ^{1/4 } } \label{tk}\ ] ] ( bardeen 1986 ) , where @xmath116 , @xmath117 , and @xmath118 is the physical wavenumber in units of @xmath20 mpc@xmath21 , related to the comoving wavenumber , @xmath119 , through our adopted normalizations by @xmath120 @xmath20 mpc@xmath21 . the result for @xmath1 is then obtained by using equation ( [ tk ] ) together with ( [ varpizero ] ) and ( [ askewness2 ] ) . note that in the case of a scale - dependent transfer function the @xmath26 and @xmath50 integrals can not be separated .
our calculations also require the assumption of a selection function ( sf ) along the line of sight , and we consider two functional forms chosen to characterize low - redshift ( @xmath121 ) and high - redshift surveys ( @xmath122 ) .
for the former , we consider the sf of the automatic plate measuring ( apm ) galaxy survey , @xmath123 ; \:\:\:\:\:\:\:\ : z_m = 1.412 z_c,\ ] ] where @xmath124 is the median reshift of the survey , taken to be 0.12 .
this functional form provides very good fit to the apm redshift distribution over the entire apm magnitude range ( baugh & efstathiou 1993 ; maddox , efstathiou , & sutherland 1996 ; gaztaaga & baugh 1998 ) , and should also approximate comparable optical surveys , such as the sdss . to compare the expected results for a survey such as the apm , which reaches only modest redshifts ( @xmath125 ) , with those from a much deeper angular survey , we also employ the @xmath50-space sf of kaiser ( 1992 ) ( intended to mimic a magnitude - limited survey ) , @xmath126 } { w_*^{1+\alpha } \gamma\left[(1+\alpha)/\beta\right ] } ; \:\:\:\:\:\:\:\ : w _ * = \overline{w}\left(\gamma\left[(2+\alpha)/\beta\right]\gamma \left[(1+\alpha)/\beta\right]\right ) , \label{kaiser}\ ] ] where @xmath127 is the mean conformal lookback time .
we will hereafter choose @xmath128 , @xmath129 , and @xmath130 , which yields a peak redshift of @xmath131 , so that this model , which we will refer to as the `` high-@xmath23 '' sf , might characterize , for example , a radio survey such as the vla first survey ( becker , white , & helfand 1995 ) , with a median redshift of @xmath132 ( cress & kamionkowski 1998 ) .
figure [ fig : dndz ] shows the above apm sf together with the high-@xmath23 sf , as well as the apm sf obtained using @xmath133 , for comparison .
baugh & efstathiou ( 1993 ) and maddox , efstathiou & sutherland ( 1996 ) find that angular statistics in the apm survey depend much more sensitively on the peak redshift of the sf , rather than on its precise shape ; we find this conclusion to hold in general for sfs we consider .
bernardeau ( 1995 ) finds that the small - angle approximation works reasonably well for an apm - like sf ; for deeper surveys in which a fixed angle can probe larger scales , this approximation is expected to provide still better results . in figure
[ fig : s3_a12 ] , we plot the variation of @xmath1 with @xmath8 , assuming the apm sf , for an unbiased ( @xmath134 , @xmath135 ) tracer - mass population in four flat ( @xmath136 ) cosmological models : standard cdm ( scdm ; @xmath137 , @xmath138 , @xmath139 ) , a tilted cdm model ( tcdm ; @xmath137 , @xmath138 , @xmath140 ) , cdm with a cosmological constant ( @xmath141cdm ; @xmath142 , @xmath138 , @xmath139 ) , and cdm with a high hubble parameter ( hcdm ; @xmath137 , @xmath143 , @xmath139 ) , as well as an open cdm model ( ocdm ; @xmath142 , @xmath138 , @xmath139 ) .
the thin horizontal line in the figure shows the ( virtually @xmath15-independent ) semi - analytic results obtained for a scale - free power spectrum with a canonical index value of @xmath144 ( peebles 1980 ) .
the results for @xmath1 in general are very similar to those found by bk99 for the normalized spatial skewness .
the scale - dependent cdm transfer function naturally induces a dependence on the smoothing scale , @xmath8 , and compared to equation ( [ s3 ] ) , a more substantial dependence on the combination of cosmological parameters @xmath145 . yet , although the five cosmological models shown in the figure span a broad range of parameter values , @xmath146 , they yield fairly similar results , and are certainly all consistent within the limits of current observations ( see figure [ fig : s3_apm ] ) .
the slight variation seen is mostly due to the dependence on @xmath147 ; there is little difference between the scdm and tcdm models , which differ only in the value of @xmath3 , and virtually no difference between the ocdm and @xmath141cdm models .
if the analysis is restricted to the range @xmath148 , as suggested by current data ( bartlett 1998 ) , one can infer from the figure that the predicted values for @xmath1 for a survey such as the apm are relatively insensitive to the adopted cosmological model .
figure [ fig : s3_k ] shows the results for @xmath6 for the same five cosmological models , assuming the high-@xmath23 sf .
the primary dependence is again seen to be with @xmath147 , and while there is a somewhat greater spread between the different models , the differences between realistically viable models are still relatively small compared to those induced by the bias dependences which we explore in the next section .
overall , we find that the high-@xmath23 sf curves for these models show greater variation of @xmath1 with @xmath8 than their apm sf counterparts . this is simply due to the fact that in a deeper angular survey , a fixed angle probes clustering over a broader range of physical scales , including very large scales , while conversely , in a shallower survey , a fixed angle is sensitive to clustering only over a relatively narrow range of smaller physical scales .
thus , at larger values of @xmath8 , the high-@xmath23 result incorporates physically large scales and tends toward smaller values for @xmath1 .
these can be roughly approximated using equation ( [ s3 ] ) by @xmath149 \approx r_3(n_{k } ) \times ( 9/14)$ ] , where @xmath150 , the value of the effective spectral index of the ( untilted ) linear - theory cdm power spectrum , tends towards unity at large scales ( recall that @xmath151 is of order unity and independent of @xmath8 ) .
for the apm case , however , @xmath1 remains higher at larger @xmath8 , since even large angular scales in a shallower survey still correspond only to smaller physical scales , where the effective power - spectrum index is negative . at smaller @xmath8 ,
both sets of curves tend toward the expected larger values of @xmath152 obtained with @xmath153the effective power - spectrum index at small scales ; for tilted models , @xmath1 tends toward a value @xmath154 smaller . note that in both figures , the semi - analytic results for a scale - free power spectrum with a canonical index value of @xmath144 ( thin , horizontal lines ) provide poor fits to the cdm predictions .
other authors have investigated ql results for the normalized angular skewness using cdm models ( gaztaaga & frieman 1994 ; frieman & gaztaaga 1994 ; gaztaaga & baugh 1998 ; gaztaaga & bernardeau 1998 ) , and find similar results .
the results of the previous section already allow for the existence of a constant , nonlinear bias between the tracer - mass and the underlying matter distributions , but can be generalized by allowing the bias parameters to evolve with time , as is suggested both by theory and observations ( fry 1996 ; peacock 1997 ; matarrese 1997 ; steidel 1998 ; cress & kamionkowski 1998 ; catelan 1998 ; taruya , koyama , & soda 1998 ; tegmark & peebles 1998 ; coln 1998 ; baugh 1998 ; magliocchetti & maddox 1998 ) . fry ( 1996 )
proposes a bias - evolution model , which assumes that objects in an einstein - de sitter universe form at a fixed formation redshift , @xmath155 , by some arbitrary local process which induces a bias at that epoch , and are subsequently governed purely by gravity .
we note that this model fails to account for merging , and can not produce an anti - bias ( @xmath156 ) .
bk99 discuss a generalization of the fry model to the case of arbitrary expansion history .
other , more general models have been proposed ( e.g. , tegmark & peebles 1998 ) , but we employ the generalized fry model as a first approximation to investigate the effects of bias evolution , and merely quote the relevant results below .
the bispectrum in this model is given by @xmath157 \nonumber \\ & & \;\ ; + \;\ ; \mbox{(cyc . ) } , \label{bfe}\end{aligned}\ ] ] where @xmath158 + b_{2 _ { * } } } { [ d(w)+b_{1_{*}}-1]^2 } , \\
c_2(w ) & = & \frac{d(w)}{d(w)+b_{1_{*}}-1 } , \\
c_3(w ) & = & \frac{(4/7)[d^2(w ) + b_{1 _ { * } } -1]}{[d(w)+b_{1_{*}}-1]^2 } , \label{cs}\end{aligned}\ ] ] @xmath159 , and a subscripted asterisk denotes the value of that parameter at the epoch of formation .
note that we have now ignored the very weak dependence of the bispectrum on @xmath82 ; the dependence on the expansion history , i.e. , on the species contributing to the total energy density , is contained in the growth factor , @xmath63 .
the result for @xmath1 can then be extended to the case of evolving bias by using the bispectrum of equation ( [ bfe ] ) in equation ( [ askewness ] ) , which amounts to making the substitutions @xmath160 inside the brackets in equation ( [ askewness2 ] ) , and replacing @xmath55 by @xmath161 . for the apm and high-@xmath23 sfs , respectively , figures [ fig : s3_a12_bias ] and [ fig : s3_k_bias ] display the results for @xmath6 in five different bias scenarios , each employing the scdm , @xmath141cdm , and ocdm models above : an unbiased scenario ( @xmath162 , @xmath163 , taken from figures [ fig : s3_a12 ] and [ fig : s3_k ] ) , a non - evolving , linear bias ( @xmath164 , @xmath163 ) , a non - evolving , nonlinear bias ( @xmath164 , @xmath165 ) , an evolving , linear bias ( @xmath166 , @xmath167 ) , and an evolving , nonlinear bias ( @xmath166 , @xmath168 ) . for the non - evolving bias scenarios , the @xmath141cdm and ocdm models are taken to be identical , as suggested by figures [ fig : s3_a12 ] and [ fig : s3_k ] , and therefore only the @xmath141cdm models are plotted in these cases .
these two models do , however , differ slightly more in evolving bias scenarios , due to the different time dependences of their linear growth factors , and the differences between them would increase with decreasing @xmath15 . for
the evolving cases we hereafter assume a formation redshift of @xmath169 .
the qualitative dependences of @xmath1 on the bias scenario are similar to those of its spatial counterpart ( bk99 ) .
specifically , it is clear from these figures that the dependence of the angular skewness on the biasing scheme can potentially be far more significant than that upon the cosmological parameters within a given scheme .
for example , adding a very small linear bias to the low-@xmath147 models would yield a result very similar to the unbiased scdm predictions .
generally , the presence of any significant linear bias , @xmath170 , tends to reduce the dependence of @xmath1 on @xmath8 , as compared with the unbiased cases ( solid lines ) .
furthermore , an observed @xmath1 curve which is far below the predicted unbiased result can only be achieved by a large linear bias term ( assuming @xmath171 ) , while one well above this value can only arise from the existence of a non - zero @xmath56 term or from anti - biasing ( @xmath172 ) .
the curves for an evolving bias can produce less drastic shifts than their non - evolving counterparts , despite the fact that the respective bias terms are initially larger .
this is because the evolution towards an unbiased state can effectively mimic a smaller ( especially at late times ) , constant bias , as demonstrated by bk99 .
again we find that the apm curves in figure [ fig : s3_a12_bias ] are flatter than their high-@xmath23 counterparts in figure [ fig : s3_k_bias ] , for reasons explained in 2.1 .
note also that models with nonlinear bias exhibit an upturn at large @xmath8 .
the upturn arises from the contribution of the higher - order terms in the sum in equation ( [ askewness2 ] ) , which , in the presence of a non - zero @xmath56 term , can become significant on large scales , where the effective power - spectrum index becomes positive ( see figure 1 ) .
bk99 perform the analogous three - dimensional calculation and find similar large - scale behavior of the normalized spatial skewness , @xmath173 .
depending on the cosmological model , this transition here occurs at angular scales of about a few degrees for the apm sf , and slightly higher values in the case of the high-@xmath23 sf .
thus , the scdm model , whose power spectrum peaks at larger wavenumber than those of the low-@xmath147 models , will exhibit the upturn at smaller values of @xmath8 , while the low-@xmath147 models with nonlinear bias are expected to show an upturn only at larger angular scales , where they begin to sample positively - sloped regions of the power spectrum .
however , unlike the case for @xmath173 ( bk99 ) , the exact location and severity of this upturn in @xmath1 for the different cosmological models is influenced not only by the shape of the power spectrum , but also by the weightings arising from the sf ( see figure [ fig : dndz ] ) and from the terms proportional to powers of @xmath50 in equations ( [ askewness2 ] ) and ( [ varpizero ] ) .
for example , the apm sf , which resembles a low-@xmath50 spike , exacerbates the upturn by effectively sampling the @xmath50 integrals only around the low peak value of @xmath174 , where the @xmath175 factor is extremely large . in the case of the high-@xmath23
sf , while large values of @xmath8 still sample a positively - sloped region of @xmath176 , the sf itself , which is everywhere much shallower than in the apm case , is effectively zero where the @xmath175 factor is large , and peaks at a high value of @xmath50 , where this factor is very small .
as seen in figure [ fig : s3_k_bias ] , this behavior tends to lend less support to the divergence of the sum in equation ( [ askewness2 ] ) at these scales , even in the case of a large , constant value of @xmath56 .
note that for both sfs , the non - evolving , nonlinear cases exhibit more of an upturn , since in these cases the large value of @xmath56 is maintained for all time , whereas in the evolving cases , the divergence is less severe due to the fact that the tracer population evolves towards an unbiased state with time .
the upturn in @xmath1 at large @xmath8 in nonlinear - bias models differs from the decreasing behavior expected if the higher - order terms in the sum in equation ( [ askewness2 ] ) are neglected , i.e. , if the bias parameters , as defined for _ smoothed _ density fields in equation ( [ dexp ] ) , are are assumed to be constant , as in peak - biasing or halo - biasing models ( mo & white 1996 ) .
bk99 show in the case of the normalized spatial skewness that this different large - scale behavior might be used to obtain better constraints on @xmath55 and @xmath56 on the basis of skewness measurements alone , since combinations of linear and nonlinear bias which yield effectively degenerate results for @xmath173 on smaller smoothing scales , can , at least in principle , be distinguished by considering smoothing scales where the relevant _ effective _ index ( bernardeau 1994 ) is positive , i.e. , scales greater than @xmath177 @xmath178 mpc in typical cdm models ( see figures 2 and 5 in bk99 and related discussions ) . similarly , we can infer from figures [ fig : s3_a12_bias ] and [ fig : s3_k_bias ] that combinations of these parameters which are degenerate at small @xmath8 become more distinct at large @xmath8 ( due to the upturn ) , and that comparing ql pt predictions with large - scale measurements ( @xmath179 ) of @xmath1 might also resolve this degeneracy . if this large - scale behavior is not seen , this would argue in favor of smoothed - bias prescriptions , rather than the unsmoothed formulation we employ . at such scales , however , one might be limited by the variance induced from the smaller number of independently sampled cells ( seto & yokoyama 1998 ) , and may require a more accurate derivation which does not assume the small - angle approximation , but instead employs a full spherical harmonic decomposition ( verde 1999 ) . figures [ fig : s3_a12_bias ] and [ fig : s3_k_bias ] also provide illustrations of one of the potential degeneracies that can arise when considering the normalized skewness .
it is known that the one - point statistics @xmath173 and @xmath1 can be nearly degenerate with respect to combinations of cosmological and biasing models , with respect to constant @xmath55 and @xmath56 ( barring large - scale measurements , as mentioned above ) , and also with respect to evolving vs. non - evolving bias models ( bk99 ) .
this latter degeneracy is illustrated in figure [ fig : s3_a12_bias ] by the almost identical predictions of the ocdm model with evolving , linear bias and the ocdm/@xmath141cdm model with constant , linear bias , and in figure [ fig : s3_k_bias ] by the similarity between the scdm models with constant , linear bias and with evolving , linear bias .
it would clearly be impossible to distinguish between such similar predictions given present data of the type shown in figure [ fig : s3_apm ] .
as an example of what can be learned from current measurements , we compare in figure [ fig : s3_apm ] our predictions for @xmath1 using the apm sf with data from the apm galaxy survey , containing over @xmath180 galaxies ( see gaztaaga 1994 and references therein ) .
the data points ( gaztaaga 1997 ) are not corrected for the effects of source fragmentation , and the error bars shown are derived from the scatter in four sub - zones , plus shot - noise corrections , and are thus extremely conservative , especially at large @xmath8 .
the curves plotted are for the scdm and @xmath141cdm models , with `` best - fit '' values for @xmath55 of 1.14 and 1.22 , respectively ( assuming only constant , linear biasing ) .
it should be noted that discrepancies between the data and our ql pt predictions are expected at small scales , due to : 1 ) shot - noise fluctuations arising from the discreteness of the observed distribution [ i.e. , the @xmath181 are no longer reliable estimators of the area - averaged cfs ( peebles 1980 ; gaztaaga 1994 , 1997 ) ] , 2 ) estimation biases arising from the fact that the ratio of two unbiased estimators is not itself an unbiased estimator ( hui & gaztaaga 1998 ) , and 3 ) nonlinear effects not accounted for in our models ( boucher , schaeffer , & davis 1991 ; jain & bertschinger 1994 ; scoccimarro 1998 ; scoccimarro & frieman 1998 ; munshi & melott 1998 ) . at large scales
we expect discrepancies to arise from the breakdown of the small - angle approximation and the large scatter in the data induced by the smaller number of independently sampled cells .
if we therefore exclude the data points at both the lower and upper ends of the range of @xmath8 , and restrict the comparison to ql scales of @xmath182 , where our predictions are expected to be valid , we find reasonable agreement with plausible cdm models having a small amount of linear bias .
of course , this conclusion is merely qualitative , as these data are clearly unable to place tight constraints on the models .
the 2-@xmath183 error on @xmath55 in both models is roughly @xmath184 , depending on exactly which points are excluded , in agreement with previous determinations of @xmath185 for the apm survey ( baugh & efstathiou 1993 ; gaztaaga 1994 ; bernardeau 1995 ) .
that the data for optically selected galaxies are roughly fit in the ql regime by the predicted values of @xmath1 in cdm models , requiring little or no bias ( gaztaaga 1994 ; bernardeau 1995 ; cappi & maurogordato 1995 ; roche & eales 1998 ; magliocchetti 1998 ) , lends some support to the scenario of gravitational evolution ensuing from gaussian initial conditions , but better data are clearly needed . note that the thin solid line , derived from equation ( [ s3 ] ) with the apm sf , no bias , and a constant power - law index of @xmath144 , yields a poorer fit in the ql range
the normalized angular skewness , while containing some information about the overall dependence of clustering strength with scale , discards detailed and valuable information about the configuration dependence which is contained in the full angular 3pcf .
it is known that the shape dependence of the spatial bispectrum and 3pcf , in leading order pt , can be used to distinguish between the effects of gravitational clustering and bias , as well as to obtain independent constraints on the linear- and nonlinear - bias parameters and on cosmological parameters ( jing & zhang 1989 ; fry 1994 ; fry 1996 ; jing & brner 1997 ; matarrese , verde , & heavens 1997 ; bk99 ) .
it stands to reason that the full angular 3pcf should also break these degeneracies and yield similar constraints . with this in mind , many authors have investigated the normalized angular 3pcf , @xmath186 where the angular distances @xmath187 form a triangle on the sky .
this quantity , which like @xmath1 is independent of the amplitude of the power spectrum , was introduced based on empirical arguments for the so - called hierarchical model ( peebles 1975 ; peebles & groth 1975 ; peebles 1980 ) , and motivated by data which appeared to indicate that @xmath0 was simply a constant .
while equation ( [ qdef ] ) may be taken as a definition , the result for @xmath0 in ql pt is , in general , _ not _ merely a constant ; we demonstrate that the apparent constancy of @xmath0 , as measured from past data , may simply be an artifact of coarse averaging over configuration shapes ( peebles & groth 1975 ; fry & seldner 1982 ) , combined with the effects of the shallow sfs which have characterized past surveys .
there are , however , other problems associated with the definition of the normalized angular 3pcf in equation ( [ qdef ] ) . in particular , @xmath0 is apt to exhibit rapid variation and divergence where the denominator in equation ( [ qdef ] ) happens to acquire values of or near zero .
this behavior is artificial in the sense that it does not reflect the true behavior of the 3pcf itself . bk99
demonstrate the analogous problems associated with the definition of the normalized spatial 3pcf , @xmath188 , and speculate that practical limitations in measuring this behavior may , in part , be responsible for observed discrepancies between ql pt predictions and those of @xmath95-body simulations .
the skewness is less susceptible to this problem since it is scaled by the square of the variance , a positive - definite quantity . in places where @xmath0 exhibits this rapidly - varying or divergent behavior ,
we will instead consider @xmath189 , defined as @xmath190 , where @xmath191 is the rms angular fluctuation on a smoothing scale of @xmath98 , obtained for each model using equation ( [ varpizero ] ) .
the chosen smoothing scale is , of course , arbitrary ; we adopt a value of @xmath98 as a convenient choice for the sake of graphical representation . like @xmath0
, @xmath189 is independent of the overall normalization of the power spectrum , but it is not susceptible to the above mentioned problem associated with the definition of @xmath0 .
observational measurements of both @xmath0 and @xmath1 , however , will necessarily be ratios of estimators , and thus subject to associated errors discussed by hui & gaztaaga ( 1998 ) . taking equation ( [ varpi ] ) with @xmath192 and omitting the window function , since we presume that the full cfs are now measured by direct counting , rather than by smoothed counts in cells , we obtain the direct - counting angular 2pcf , @xmath193 the full angular 3pcf for projected triangles with sides of angular measures @xmath194 , @xmath195 , and @xmath196 can be evaluated using equation ( [ askew ] ) , @xmath197 \nonumber \\ & & \;\ ; + \;\ ; \mbox{(cyc.)},\end{aligned}\ ] ] where @xmath198 is the angle between @xmath199 and @xmath200 , @xmath79 is the angle between @xmath86 and @xmath201 , and similar conventions are used in each permutation term . taking @xmath202 , and @xmath203 to be the angle between @xmath204 and @xmath205 , so that @xmath206 and performing the @xmath207 and @xmath208 integrations , we have @xmath209 \biggr\ } \;\ ; + \;\ ; \mbox{(cyc.)}. \label{z}\end{aligned}\ ] ] for evolving bias we simply substitute @xmath210 inside the brackets , and replace @xmath55 by @xmath161 .
equation ( [ z ] ) does not contain terms analogous to the higher - order terms in the sum in equation ( [ askewness2 ] ) [ which gave rise to the @xmath109 term in equation ( [ s3 ] ) ] , since evaluation of the 3pcf by direct counting requires no smoothing . using our model for unsmoothed bias ,
such terms _ would _ be induced in the calculation of the 3pcf using counts in cells , arising from factors of the form @xmath211 in the appropriate integrands .
one must be careful to realize that if the unsmoothed - bias prescription is correct , the nonlinear - bias parameter measured from the direct - counting angular 3pcf in equation ( [ z ] ) will be different from the smoothed value which could be inferred from @xmath1 or from the counts - in - cells 3pcf , though the two can be related , e.g. , via the @xmath109 term in equation ( [ s3 ] ) . for a scale - free power spectrum , equation ( [ 2pcf ] ) yields the familiar result , @xmath212 . for realistic cdm power spectra
, there is a further dependence on @xmath213 which arises from the scale - dependence of the transfer function .
likewise , had we ignored the shape - dependent terms in the bispectrum in equation ( [ bf2 ] ) , we would find for power - law spectra the hierarchical approximation @xmath214 .
these terms , however , introduce into equation ( [ z ] ) an additional , non - trivial dependence on the shape of the triangular configuration , even for the case of scale - free power spectra .
realistic models must account for this variation , as well as the scale dependence induced by the transfer function . in order to interpret the results of the various cdm models
, it will be instructive to first investigate the results obtained for @xmath7 for scale - free power spectra . in this case
, we have @xmath215 \varpi(\theta_{13})\varpi(\theta_{23 } ) + \mbox{(cyc . ) } , \label{zpowerlaw}\end{aligned}\ ] ] where @xmath216 and @xmath217 .
figure [ fig : q_powerlaw ] displays the results for @xmath218 for @xmath219 [ recall @xmath106 is of order unity , equal to \{1.26 , 1.19 , 1.15 } and \{1.21 , 1.18 , 1.17 } for these values of @xmath3 , for the apm and high-@xmath23 sfs , respectively ] .
since @xmath0 in this case is independent of overall scale , we specify the triangle configuration by fixing @xmath220 to have length 1 , @xmath195 to have a relative length @xmath221 , and examine the variation of @xmath0 with @xmath203 , for @xmath222 , @xmath223 , and the `` limit '' of @xmath224 .
the figure reveals shape dependence on both @xmath225 and @xmath226 , the extent of which can ( particularly in the case of the @xmath226 dependence ) depend strongly on the index @xmath3 .
specifically , we note that the amplitude of the shape variation tends to increase with increasing @xmath3
. an `` average '' value for @xmath0 can be obtained by taking @xmath227 and @xmath228 = 1/2 in equation ( [ zpowerlaw ] ) , yielding @xmath229 ; while this is not generally valid , since the three angles appearing in the cyclic terms can not be varied independently , this result is exactly true for all configurations when @xmath230 ( the no - smoothing case ) , as demonstrated by the thin , solid , horizontal line . in the left panels of figure [ fig:2d ] we illustrate , for the apm
sf , the behavior of @xmath231 for several configurations falling in the quasilinear regime , for an unbiased tracer mass in the scdm , tcdm , @xmath141cdm and ocdm models . in this case
, the results for the latter two models differ only by about the thickness of the curves , and can barely be distinguished . other models , however , yield quite distinct predictions and might be differentiated on the basis of angular 3pcf measurements from large surveys . in particular , the scdm and tcdm can be clearly distinguished from each other ( bkj show that this is not the case when considering the spatial 3pcf at @xmath12 ) , and from the low-@xmath147 models , which show comparatively little variation with @xmath203 .
generally , we now find a dependence of @xmath0 on scale ( compare , e.g. , the top and bottom panels ) arising from the shape of the cdm transfer function . the observed shape dependence on @xmath195 ( compare , e.g , the top and middle panels ) and @xmath203 of these models can be understood in terms of the results for scale - free power spectra : over the scales considered here , the scdm model has the largest effective power - spectrum index , while the low-@xmath147 models have the smallest .
the results for the apm sf shown in figure [ fig:2d ] roughly agree with the predictions and @xmath95-body results of frieman and gaztaaga ( 1999 ) .
we thank them for pointing out an error in an earlier calculation of ours . for a shallow sf such as that describing the apm galaxy survey ,
three points describing an elongated triangle in projection are more likely to correspond to a configuration which is elongated in real space .
thus , our results for @xmath0 using the apm sf are similar to corresponding results for the normalized spatial 3pcf , @xmath188 ( fry 1984 ; jing & brner 1997 ) , which predict that clustering in the mildly nonlinear regime favors elongated structures . for low-@xmath147 models , however , the apm sf happens to yield predictions which do not vary strongly with either the shape or size of the triangular configuration . since it is known that measurements of the power spectrum from surveys such as the apm favor models with low values of @xmath147 ( efstathiou , sutherland , & maddox 1990 ; baugh & efstathiou 1993 ) , it is not surprising that measurements of the angular 3pcf in shallow surveys have indicated it to be configuration independent , particularly given the present observational errors ( peebles and groth 1975 ; groth & peebles 1977 ; jing & zhang 1989 ; tth , hollsi , & szalay 1989 ; jing , mo , & brner 1991 ; borgani , jing , & plionis 1991 ; jing , & brner 1998 ) .
while this has no doubt has lent support to the empirically - motivated hierarchical model which assumes @xmath7 to be , in fact , shape independent ( i.e. , @xmath232constant ) , we demonstrate that the angular 3pcf , in general , can depend more significantly on the configuration and analysis of deeper angular surveys must take this variation into account .
the results for high-@xmath23 sf for the same configurations ( assuming @xmath233 are shown in the right panels of figure [ fig:2d ] .
we now plot @xmath189 , since for this choice of sf the combination of 2pcfs appearing in the denominator of @xmath0 can acquire values of or near zero in our models . furthermore , we now take the defining parameters for triangles projected on the sky to be @xmath234 where @xmath235 , so that @xmath236 and @xmath237 ( peebles & groth 1975 ) .
for this choice of parameters , @xmath238 effectively fixes the overall size of the triangle , while @xmath225 and @xmath239 determine the exact size and shape .
figure 6 in bk99 provides a graphical illustration of the variation of triangle geometry with @xmath238 , @xmath225 , and @xmath239 .
this parameterization is convenient since each choice of \{@xmath238 , @xmath225 , @xmath239 } involves a minimum scale set by @xmath238 .
the \{@xmath220 , @xmath195 , @xmath203 } parameter set , with @xmath240 and @xmath241 is equivalent for @xmath203 ranging down to @xmath242 , but for @xmath226 less than @xmath242 this parameter set can sample scales ( given by @xmath194 ) all the way down to zero .
since we have chosen to normalize @xmath7 by the square of the variance at a fixed angular scale , the \{@xmath220 , @xmath195 , @xmath203 } set can yield dramatically different amplitudes for @xmath189 if @xmath243 as opposed to those for @xmath244 , and is therefore less tenable .
even once their different normalizations have been accounted for , we find that the high-@xmath23 curves have significantly lower amplitudes than their apm - sf counterparts in the left panels .
this is simply because projected triangles in a deeper survey can correspond to much larger physical triangles , for which the clustering amplitude will be significantly smaller .
since @xmath189 is normalized by the square of the variance at a fixed scale , these plots now reveal clearly the expected decrease in clustering strength with increasing scale ; this behavior , however , is not described simply by the @xmath245 prediction of the hierarchical model . unlike the apm case , we now find that _ all _ models show a significant variation with shape , as well as scale , with the low-@xmath147 models in some cases now showing the strongest shape dependence .
the scdm and tcdm models appear less differentiated than in the apm case , but we find that the @xmath141cdm and ocdm models can now differ somewhat more substantially , particularly on larger scales , since deeper surveys are more sensitive to cosmological effects .
most importantly , however , we still find overall that the high-@xmath147 and low-@xmath147 models can be fairly well resolved on @xmath246 scales , and the distinction increases for larger scales .
naturally , a thorough comparison with measurements spanning the full configuration space in a more complete fashion would allow for better discrimination among models .
the results for the different bias scenarios investigated in 2.2 are demonstrated in figure [ fig : z_bias ] , which shows @xmath189 , assuming @xmath247 and @xmath222 , for the @xmath141cdm ( thick lines ) and ocdm ( thin lines ) models , for the high-@xmath23 sf . though this is only one slice of the parameter space we are exploring , we use it to illustrate the similar dependences of the full angular 3pcf on the bias scheme as were seen for @xmath1 . in particular , we see a general flattening and reduction of @xmath189 with increasing linear bias , and a relative increase with increasing nonlinear bias , as compared with the corresponding linear scheme .
note the slightly different predictions between these two models , arising from their different expansion histories ; these differences would be greater for lower values of @xmath15 . as with @xmath1 ( see figures [ fig : s3_a12_bias ] and [ fig : s3_k_bias ] ) , it is clear that measurements of @xmath189 should be able distinguish between the effects of a significant bias and those arising simply from unbiased gravitational evolution .
consideration of the full configuration dependence of the angular 3pcf , however , allows for much better discrimination between different cosmological and biasing schemes than could be obtained from the skewness alone .
it is known that measurements of the bispectrum and 3pcf can be used to obtain independent constraints on the constant - valued linear- and nonlinear - bias parameters ( fry & gaztaaga 1993 ; fry 1994 , 1996 ; jing 1997 ; matarrese , verde , & heavens 1997 , bk99 ) .
this can be inferred here from the different manners in which @xmath55 and @xmath56 appear in equation ( [ z ] ) ; the different weightings of these two parameters is sampled for every triangular configuration , producing a graph such as figure [ fig : z_bias ] for every pair of values @xmath248 [ as opposed to @xmath1 , which yields only one constraint on the combination of @xmath249 and @xmath250 . as another example of the added information which can be gleaned from the full geometric dependence of the full 3pcf , we plot in figure [ fig:2dfull_bias_deg ] @xmath251 , assuming the high-@xmath23 sf , for two of our bias scenarios : the scdm constant , linear - bias model ( with @xmath164 and @xmath252 ) , and the scdm evolving linear - bias model ( with @xmath166 and @xmath253 ) .
these two cases yielded nearly identical results for @xmath1 in figure [ fig : s3_k_bias ] ( where evolving , linear bias effectively approximated a smaller , constant , linear bias ) , particularly on scales near @xmath98 . here , we see that by considering the variation of @xmath189 with triangle geometry , the degeneracy between constant and evolving bias can be alleviated , to a degree , because the angular 3pcf probes a broad range of redshifts and allows us to see the effects of evolution in projection . bk99 show that these evolving and non - evolving bias models might also be distinguished using measurements of @xmath254 as a function of redshift in a deep ( @xmath255 ) survey , or , to a lesser extent , by large - scale ( @xmath256 mpc ) measurements of @xmath254 at @xmath12 . in practice , however , measuring the angular 3pcf over scales of @xmath257 in a deep optical survey , or a radio survey such as the vla first survey , may be more feasible and would likely yield stronger statistical constraints .
the parameter values we have investigated here are only intended as illustrations . the full statistical power which could be gained by exploring both the configuration and evolutionary dependences of @xmath7 in a complete fashion could be useful in distinguishing between the signatures of gravitational evolution versus bias , independently measuring the linear- and nonlinear - bias parameters , and constraining the shape of the power spectrum . with better data , as might be obtained from emerging surveys , one could further determine the linear evolution of the power spectrum and of the bias , which in turn provide direct constraints on @xmath15 , @xmath258 , and possibly the epoch of galaxy formation
. however , it is already clear from figures [ fig:2d][fig:2dfull_bias_deg ] that the behavior of the angular 3pcf is not simply that of the hierarchical model which assumes @xmath0 to be a shape - independent constant , and further that , in addition to the full geometric dependence , one must consider the detailed dependences on the forms and evolutionary behaviors of the power spectrum , the biasing mechanism , and the selection function , as these can have a substantial impact on the results , particularly for high - redshift surveys .
we have calculated the leading - order results for the normalized angular skewness and 3pcf , assuming gaussian initial conditions , for an arbitrary , biased tracer - mass distribution in flat and open universes .
we have considered two selection functions , intended to represent low - redshift ( @xmath121 ) and high - redshift ( @xmath259 ) surveys .
our derivations incorporate such features as the scale dependence and linear evolution of the cdm power spectrum and the presence of a possibly evolving linear or nonlinear bias .
our results for @xmath1 for the case of an unbiased tracer mass are in agreement with previous work ( fry & gaztaaga 1993 ; bernardeau 1995 ) and we confirm that pt predictions for @xmath1 in typical cdm models , with a small amount of linear biasing ( @xmath260 ) , agree in the ql regime with data from the apm galaxy survey .
similar to bk99 , we extend the result for @xmath1 in the case of nonlinear bias , to include a scale - dependent , leading - order correction which becomes significant for positive effective spectral indices , corresponding to scales @xmath179 for cdm models .
this correction term can be used to relate the unsmoothed nonlinear - bias parameter defined in the limit of continuous fields , with the traditionally - used , scale - dependent , nonlinear - bias parameter defined for smoothed density fields . for a given cdm model , this large - scale behavior of @xmath1 might alone allow a more accurate determination of the linear- and nonlinear - bias parameters ( or at least discriminate between the smoothed and unsmoothed bias formulations ) , if other factors , such as sampling variance and the breakdown of the small - angle approximation , can be addressed .
we also derive predictions for the normalized angular 3pcf , considering @xmath189 rather than @xmath0 in cases where the latter is ill - behaved .
we show that @xmath7 , as predicted by ql pt , is not given simply by the form of the `` empirical '' hierarchical model , i.e. , with @xmath0 simply a constant of order unity , but has a more complex dependence on scale , shape , and expansion history . while for shallow sfs the weak shape dependence of @xmath0 , combined with measurements coarsely averaged over configuration shapes , can yield results consistent with this hierarchical model , ignoring the detailed geometric and evolutionary variations of the angular 3pcf in general can destroy much valuable information . in several cases ,
we illustrate explicitly the importance of the various evolutionary dependences which factor into our calculations of angular statistics , such as the linear evolution of the power spectrum , the evolution of the bias parameters , and the @xmath15 dependence of the sfs .
the proper treatment of these detailed dependences will be important in interpreting data from the emerging generation of deep , high - precision surveys , particularly when considering the high - redshift universe .
we find that biasing plays a major role in predictions for @xmath1 and @xmath7 .
in particular , the scale dependence of @xmath1 and the configuration dependence of @xmath0 ( or @xmath189 ) bear characteristic imprints of bias , such as a relative flattening and decrease with increasing @xmath55 and a relative increase with increasing @xmath56 . compared with the dependence on the biasing scenario ,
the dependence of @xmath1 on the cosmological model is fairly weak .
this is not the case , however , for @xmath0 , which can vary appreciably with the values of @xmath147 and @xmath3 . though @xmath1 preserves information about the overall scale dependence of the 3pcf , it provides only one constraint on the combination of the cosmological parameters , @xmath249 , and @xmath261 , leaving various possible degeneracies between these quantities , including the degeneracy between constant- and evolving - bias models .
these can be partially alleviated by considering the dependence of the normalized angular 3pcf on the projected - triangle geometry .
in particular , the variation of @xmath0 can be used to distinguish between the effects of gravitational evolution and bias , place constraints on the value of @xmath147 and the shape of the power spectrum , measure the bias terms @xmath55 and @xmath56 and their time evolution , and perhaps constrain @xmath15 directly . comparing predictions for @xmath1 and @xmath0 with several data sets characterizing different tracer populations could allow multiple , complementary constraints on these important quantities . while many of the results obtained here for @xmath1 and @xmath7 are qualitatively similar to the corresponding results obtained by bk99 for their spatial counterparts , @xmath173 and @xmath254 , the use of angular statistics offers many practical advantages , such as greater statistical power and the absence of redshift distortions , which warrant their separate consideration .
our derivations have also relied on specific assumptions , such as the chosen bias - evolution model and the validity of both the small - angle approximation and the assumption of quasilinearity .
based on analogous results from three - dimensional calculations , we expect that qualitative features such as the upturn in @xmath1 at large smoothing angles in the case of nonlinear bias , the reduced configuration ( scale ) dependence of @xmath0 ( @xmath1 ) with increasing linear bias of any form , and the sensitivity of @xmath0 to the shape of the power spectrum will be retained in more general calculations .
the utility of the predictions for @xmath1 and @xmath0 will of course rely on the precision with which these statistics can be measured .
while current data , of the type shown in figure [ fig : s3_apm ] , is clearly insufficient to obtain strong constraints on the shape of the power spectrum or on the bias parameters , it is hoped that future surveys , such as the sdss , will greatly improve upon this situation .
in particular , if we optimistically assume the relevant uncertainties to be predominantly statistical , we can infer that the errors in measurements of @xmath1 from the sdss will be roughly a factor of ten smaller than those in figure [ fig : s3_apm ] .
of course , the comparison of predictions with two - dimensional survey data will require the additional consideration of systematic effects . among these
are finite - volume and boundary effects , which can have a significant impact on degree scales and above ( baugh , gaztaaga , & efstathiou 1995 ; szapudi & colombi 1996 ; colombi , szapudi , & szalay 1998 ; gaztaaga & bernardeau 1998 ; munshi & melott 1998 ) , estimation biases ( hui & gaztaaga 1998 ; kerscher 1998 ) , poisson noise ( peebles 1980 ; gaztaaga 1994 ) , and sampling variance ( seto & yokoyama 1998 ) .
gaztaaga & bernardeau ( 1998 ) also find that ql pt predictions for angular statistics may underestimate projection effects , particularly for apm - type surveys .
in addition , the above calculations can be extended in several ways .
for example , one might consider stochastic - biasing models which may yield different results , particularly on smaller scales ( fry & gaztaaga 1993 ; catelan 1998 ; catelan , matarrese , & porciani 1998 ; taruya , koyama & soda 1998 ; blanton 1998 ; coln 1998 ; taruya & soda 1998 ; narayanan , berlind , & weinberg 1998 ) .
several authors have also investigated the impact of higher - order nonlinear terms in pt calculations ( jain & bertschinger 1994 ; scoccimarro & frieman 1996a , b ; scoccimarro 1998 ; heavens , matarrese , & verde 1998 ; scoccimarro & frieman 1998 ) , also expected to become important on smaller scales , where @xmath262 .
similar calculations can also be performed for higher - order angular moments and @xmath3-point cfs , such as the kurtosis and four - point correlation function , and so on ( suto & matsubara 1994 ; bernardeau 1994 ; lokas 1995 ; gaztaaga & bernardeau 1998 ) , which , in principle , provide additional , independent constraints .
the formalism can also be applied to structure - formation models with non - gaussian initial conditions , such as topological - defect or isocurvature models , to investigate how these might be distinguished from inflationary models ( luo & schramm 1993 ; jaffe 1994 ; fry & scherrer 1994 ; chodorowski & bouchet 1996 ; white 1998 ) .
we are indebted to j. frieman and e. gaztaaga for helpful discussions and suggestions , and in particular for their help in clarifying an earlier problem in the calculation .
we also wish to thank r. scoccimarro , as well as licia verde , for numerous helpful comments and discussions .
this work was supported at columbia by d.o.e .
contract defg02 - 92-er 40699 , nasa nag5 - 3091 , nsf ast94 - 19906 , and the alfred p. sloan foundation , and at berkeley by nag5 - 6552 .
in this appendix , we derive a simple fourier - space analogue of limber s equation for the bispectrum of the two - dimensional projection of a three - dimensional scalar field .
our derivation is similar to that of kaiser ( 1992 ) , who calculated the projected ( two - point ) power spectrum .
we first consider the case of an einstein - de sitter universe and at the end we generalize the results to open or closed universes .
jaffe & kamionkowski ( 1998 ) discuss a different application of this formalism to isotropic vector fields .
consider a three - dimensional field , @xmath263 , at a conformal lookback time @xmath50 , and its fourier transform , @xmath264 .
we will take @xmath265 to have zero mean ( or subtract the mean explicitly ) . to project onto the two - dimensional sky ,
we apply the `` selection function , '' @xmath266 , and denote the resulting field by @xmath267 where @xmath31 is a two - dimensional vector on the sky pointed at by the three - dimensional unit vector @xmath268 .
we consider only small patches , and thus can perform a two - dimensional fourier transform on essentially flat areas of sky , giving @xmath269 , where @xmath270 denotes two - dimensional fourier - space vectors . as in kaiser
( 1992 ) , the contribution from a thin shell of width @xmath271 centered at @xmath272 to @xmath269 is @xmath273 where the @xmath274 factor arises from the integral of @xmath275 over the shell .
we define the three - dimensional bispectrum , the fourier - space analogue of the three - point function , by @xmath276 the two - dimensional analogue is @xmath277 where the dimensionality of the dirac delta function is obvious from its arguments .
we can form the contribution to the two - dimensional bispectrum from the thin shell , @xmath278 ^ 3\over w_0 ^ 4 } \int { dk_3\over2\pi } { dk_3'\over2\pi } \nonumber\\ & \times & j_0\left(k_3\delta w\over2\right ) j_0\left(k'_3\delta w\over2\right ) j_0\left(k''_3\delta w\over2\right ) b(r , r',r''),\end{aligned}\ ] ] where @xmath279 now , we make the same simplifying assumptions as in the original kaiser derivation : the shells have width @xmath280 but are thick compared to wavelengths contributing to @xmath265 of interest , so @xmath281 . since the @xmath282 factors have width @xmath283 , we can pull the @xmath76 out of the integral and set @xmath284 ; only modes perpendicular to the line of sight contribute in the small - angle approximation . then using the identity , @xmath285 we can write the contribution from the shell as @xmath286 the bispectrum is a `` cumulant '' ; that is , contributions to the total bispectrum simply add .
thus , we can sum up the contributions from different shells , and convert the sum to an integral : @xmath287 with the full definition of the bispectrum , this just gives the desired formula , @xmath288 ( allowing the three - dimensional bispectrum to vary with lookback time , @xmath50 ) .
note that the integrand in the formula is weighted by @xmath289 ; this is in contrast to the ( two - point ) power spectrum version of the calculation which weights by @xmath290 [ see equation ( [ pp2 ] ) ] .
hence , the bispectrum is weighted considerably more heavily to late times ( small @xmath50 ) . to generalize to an open or closed universe
, we must replace the distances @xmath50 in the derivation above with the angular - diameter distance .
thus in an open or closed universe one replaces @xmath291 in the first three arguments of @xmath292 and in the @xmath293 in the denominator in equation ( [ eq : bispecformula ] ) , where @xmath294 in an open universe and @xmath295 in a closed universe . | we calculate the normalized angular three - point correlation function ( 3pcf ) , @xmath0 , as well as the normalized angular skewness , @xmath1 , assuming the small - angle approximation , for a biased mass distribution in flat and open cold - dark - matter ( cdm ) models with gaussian initial conditions .
the leading - order perturbative results incorporate the explicit dependence on the cosmological parameters , the shape of the cdm transfer function , the linear evolution of the power spectrum , the form of redshift distribution function , and linear and nonlinear biasing , which may be evolving .
results are presented for different redshift distributions , including that appropriate for the apm galaxy survey , as well as for a survey with a mean redshift of @xmath2 ( such as the vla first survey ) .
qualitatively , many of the results found for @xmath1 and @xmath0 are similar to those obtained in a related treatment of the spatial skewness and 3pcf ( buchalter & kamionkowski 1999 ) , such as a leading - order correction to the standard result for @xmath1 in the case of nonlinear bias ( as defined for unsmoothed density fields ) , and the sensitivity of the configuration dependence of @xmath0 to both cosmological and biasing models .
we show that since angular cfs are sensitive to clustering over a range of redshifts , the various evolutionary dependences included in our predictions imply that measurements of @xmath0 in a deep survey might better discriminate between models with different histories , such as evolving vs. non - evolving bias , that can have similar spatial cfs at low redshift .
our calculations employ a derived equation
valid for open , closed , and flat models for obtaining the angular bispectrum from the spatial bispectrum in the small - angle approximation .
# 1#1 # 1 # 1 2w |
the dwarf galaxies of the local group ( for reviews see mateo 1998 , tolstoy et al .
2009 ) display a variety of shapes . while this is particularly true for dwarf irregular ( dirr ) galaxies , it also applies to dwarf spheroidals ( dsph ) . unlike globular clusters
, they tend to exhibit significantly non - spherical images of stellar content .
the origin of these shapes is a subject of ongoing debate but one of the interesting possibilities involves the evolutionary connection between dirrs and dsphs .
this scenario proposes that dsphs evolved from the primordial dirrs via tidal stirring caused by the presence of a large host galaxy , such as the milky way or andromeda ( mayer et al . 2001 ) . the general picture painted by the tidal stirring model studied by the means of high resolution @xmath0-body simulations ( mayer et al .
2001 ; klimentowski et al . 2009 ; kazantzidis et al . 2011 ) involves strong mass loss and morphological transformation , as well as a rebuilding of the orbital structure of the initial disk . in such simulations , a primordial disk galaxy embedded in an extended dark matter halo
is placed on different orbits around a milky way - sized galaxy and evolved for a significant fraction of the hubble time .
a thorough study of different aspects of the tidal stirring scenario , for different initial dwarf structures and different orbits , was recently performed by kazantzidis et al .
( 2011 ) and okas et al .
an important and general feature of the morphological evolution of the dwarf established in these studies is the formation and longevity of a triaxial , tidally induced bar .
the bar , usually formed out of the disk at first pericenter passage , tumbles with a period much shorter than the orbital period of the dwarf , which leads to a number of observational biases in the characteristic parameters of the dwarfs ( okas et al .
. as the tidal evolution progresses , the bar shortens and the stellar component of the dwarf becomes more and more spherical . if the tidal force is strong enough ( on tight orbits or orbits with small pericenters ) , the final shape of the stellar component can be almost exactly spherical . in this paper
we discuss selected observational characteristics of the tidal stirring model that may help to distinguish it from other scenarios for the formation of dsph galaxies , e.g. , those in which they form in isolation by purely baryonic processes such as cooling , star formation , feedback from supernovae and uv background radiation ( ricotti & gnedin 2005 ; tassis et al .
2008 ; sawala et al .
if the overall picture suggested by the tidal stirring model is correct , some of the less evolved dwarfs in the local group may still be in the bar - like phase .
such elongated shapes are indeed observed at least in a few cases , such as ursa minor ( e.g. , irwin & hatzidimitriou 1995 ) , sagittarius ( majewski et al .
2003 ) and the more recently discovered hercules dwarf ( coleman et al .
2007 ) .
[ cols="<,<,^,^,^,^,^,^,^,^,^,^ , < " , ] let us now focus on the 2d shapes of the simulated dwarfs , as they would be seen by a distant observer .
we illustrate the evolution of such shapes using our default simulation o1 . in figure [ viewsrun1apo ]
we show how the dwarf s stellar component would appear when viewed along the principal axes of the stellar distribution @xmath1 , @xmath2 and @xmath3 ( from the upper to the lower row ) at subsequent apocenters of its orbit around the milky way ( from the left to the right column ) .
the images were produced by selecting stars within @xmath4 kpc and rotating their positions to align them with the principal axes determined for stars with @xmath5 , as described above .
the plots show the surface density distribution of the stars with contours equally spaced in log density . in agreement with the conclusions from figure [ shapesrun1 ] , at second apocenter
the dwarf already has a pronounced bar and its shape changes from the oblate ( initial disk ) to the prolate ( a bar ) . at the third apocenter
the shape is again oblate ( as shown by the green line in the upper panel of figure [ cbbaapo ] which crosses the diagonal line back again from the prolate to the oblate part of the diagram at this time ) .
later on , the shape gradually becomes more and more spherical with all axis ratios approaching unity .
the final image of the dwarf , at the end of the simulation , at its most non - spherical appearance ( i.e. , viewed along the @xmath2 axis ) is shown in figure [ shapealongyrun1 ] . in figure [ shapesalongy ]
we show analogous images for the remaining simulations o2-o7 and s6-s17 . in all cases stars within @xmath4 kpc
were extracted from the last output of the simulation and rotated to align the distribution with the principal axes .
only the views along the @xmath2 axis are shown , i.e. we see the dwarfs along the line of sight in which they appear most non - spherical .
the dwarfs are shown at the end of their evolution when most are far from their pericenters , and therefore weakly affected by tidal forces at this time , except for the case of o4 which has just passed its 6th pericenter .
although this dwarf has already reached an almost spherical shape during its earlier evolution , it is strongly affected by tidal forces from the milky way at this particular instant ( note that its pericentric distance was 12 kpc at this passage ) which manifests itself by strong distortion of the outer contours .
going back to our single - number measures of the shapes , @xmath6 and @xmath7 , we summarize our findings by comparing in figure [ losshapes ] the values of these parameters measured along the three lines of sight at the end of each simulation and listed in columns 7 - 12 of table [ properties ] . for each run
we plot @xmath6 with filled symbols and @xmath7 with open symbols .
the triangles , squares and circles correspond to measurements along the @xmath2 , @xmath1 and @xmath3 axis of the stellar component , respectively .
as discussed above , the observation along the intermediate axis @xmath2 always gives the most non - spherical appearance , i.e. , triangles always mark the highest values of @xmath6 and @xmath7 . in addition , the next highest value of ellipticity is almost always the one seen along the @xmath1 axis ( so that the points are ordered triangle - square - circle from top to bottom ) .
the exceptions occur for runs o4 , s7 , s11 and s12 .
since these cases correspond to @xmath8 this means that only in these four cases do the dwarfs end up more prolate than oblate ( o4 should be excluded because its shape is due to strong tidal forces at pericenter ) .
it is interesting , therefore , that although almost all dwarfs go through a bar - like phase during their evolution , most of them end up more oblate than prolate .
we also note that the values of @xmath6 and @xmath7 are quite similar , although not exactly equal , and they preserve the relative triangle - square - circle hierarchy of the measurements . to illustrate this similarity further , in figure [ a2baapo ] we plot @xmath6 as a function of @xmath7 for the line of sight where the dwarfs appear most non - spherical ( observation along @xmath2 ) .
the lines join the values measured at subsequent apocenters , from the most non - spherical at the upper right corner of each panel to the almost spherical in the lower left .
the upper panel shows the results for simulations of dwarfs on different orbits ( o1-o7 ) and the lower one those with different structure of the dwarf ( s6-s17 ) .
the lines corresponding to each simulation can be identified by referring to the last column of table [ properties ] .
we can see that there is a one - to - one correspondence between the two measures of shape , and thus conclude that both measures , although defined differently from a mathematical point of view , essentially measure the departures from sphericity in a similar way and can be used interchangeably to quantify the shapes of galaxies with resolved stellar populations .
in this section we discuss the shape parameters of the real dwarfs , satellites of the milky way using photometric data .
our dwarf galaxy sample includes objects within heliocentric distance of about 260 kpc ( close to the largest apocenter considered in the simulations ) , brighter than @xmath9 mag and with good photometric data available .
these requirements restrict the sample to the eight classical dsph galaxies , the sagittarius dwarf and the large and small magellanic clouds ( lmc and smc ) .
the data sets we used comprise the positions and magnitudes of stars in at least two bands to allow for the selection of probable members of dwarf galaxies from the color - magnitude diagrams ( cmd ) . in the case of the fornax dwarf we used the data from coleman et al .
( 2005 ) , for sextans and draco the data were retrieved from the sloan digital sky survey database . for these dwarfs we selected member stars by iterative fitting of the position of the red giant branch ( rgb ) and choosing stars within a strip of variable width along the branch whose shape was adjusted to match the equal - density contours of the stars in the cmd . for the lmc and smc
, we extracted the stars from the 2mass survey and also included stars from the asymptotic giant branch ( see van de marel & cioni 2001 ; gonidakis et al .
2009 ) . for the remaining dwarfs ( sculptor , carina , ursa minor , leo i , leo ii and sagittarius ) we used samples obtained previously by s.r.m .
and collaborators with more precise photometric membership determination using color - magnitude as well as color - color diagrams to extract only the giant stars . the technique , described in detail in majewski et al .
( 2000a ) , relies on the sensitivity of the @xmath10 filter to magnesium features that reflect stellar surface gravity , temperature and abundance in later type stars .
when combined with the wide - band @xmath11 and @xmath12 filters of the washington system , the @xmath10 filter is especially useful for discriminating giant stars from foreground dwarfs on the basis of differences in their respective @xmath13 colors at a given @xmath14 color . the data ,
although not publicly available , are described in the references provided in table [ dwarfs ] and we refer the reader there for details concerning each dwarf galaxy .
table [ dwarfs ] lists the properties of the dwarfs included in this analysis : the dwarf name ( column 1 ) , the position of the adopted center ( columns 2 and 3 ) , the heliocentric distance of the dwarf ( column 4 ) , the 2d half - light radius @xmath15 in arcmin and kpc ( columns 5 and 6 ) , the number of member stars within @xmath16 ( column 7 ) , the method used for the membership determination ( column 8 , cm = color - magnitude , cm+cc = color - magnitude plus color - color ) , and the source of data ( column 9 , the three numbers give respectively the reference for the dwarf center , distance and the photometric data used for the measurements of its shape ) .
although the values of the half - light radii are available in the literature , for consistency we determined them from the surface density distribution of the selected member stars by fitting a plummer profile . for this purpose , the surface density profiles were measured in logarithmic bins of projected radii and fitted with data points weighted by density adjusting two parameters , the half - light radius and normalization .
the obtained values of @xmath15 , listed in table [ dwarfs ] , are in good agreement with published scales ( see , e.g. , table 2 in okas et al .
2011 ) .
we measured the shape parameters @xmath17 and @xmath7 first using member stars in the whole range @xmath18 .
the values of these parameters are listed in the second and fifth column of table [ shapeparameters ] respectively and marked as ` whole ' . in order to distinguish between intrinsic bars and external tidal extensions
it is also interesting to look at the parameters calculated separately in the inner part within @xmath19 and the outer part within @xmath20 . those are shown in the columns of table [ shapeparameters ] marked respectively as ` inner ' and ` outer ' . in the following we will focus on @xmath7 , which is a more customary quantity used to describe bars , and adopt @xmath21 as a threshold signifying the presence of a bar - like shape .
values above this threshold were typed in boldface in table [ shapeparameters ] .
the motivation behind this division into inner and outer sample is that the tidal effects on the shape of dwarf galaxies can take two forms .
the shapes can be affected by tides so that the whole orbital structure of the dwarf is rebuilt and a bar forms , or only the outer contours of the stellar distribution are elongated signifying the transition to the tidal tails .
the latter situation is illustrated very well by the special case of simulation o4 which ends at the pericenter of the orbit .
this moment of the evolution is shown in the upper right panel of figure [ shapesalongy ] .
while the inner contours of the stellar density distribution are circular , the outer ones are strongly elongated due to strong tidal force at the pericenter of the orbit . on this orbit with a very small pericenter ( as well as for the very tight orbit o2 ) the dwarf transforms into an almost perfectly spherical shape early on in the evolution , as demonstrated by the time - dependence of the shape parameters in figures [ axisratios2 - 19]-[a2xyz2 - 19 ] .
the shape remains spherical for most of the time at later stages , except at pericenters , where strong elongation takes place . in figure [ shapessmoothed ]
we show the contour maps of the surface density of the stars in real dwarfs up to projected radii of the order of @xmath16 .
the maps were obtained by smoothing the discrete distribution of stars with gaussians of scale @xmath22 and binning it into @xmath23 grids . as discussed in section 3 ,
such smoothing scales are small enough not to obscure subtle morphological features .
the contours joining the regions of equal surface density of the stars are equally spaced in log surface density . as verified in table [ shapeparameters ] ,
only in the case of ursa minor and sagittarius are all three values of the bar mode parameter for the whole , inner and outer samples found to have @xmath24 .
this strong elongation , determined by the discrete shape measures , is confirmed in a more qualitative way by the images of these dwarfs in figure [ shapessmoothed ] .
we conclude that these dwarfs possess the tidally induced bar and that their outer parts are affected by tides . in the case of the sagittarius dwarf , the outer parts are likely stretched by tidal forces acting _ now _ , since the dwarf is very close to the pericenter of the orbit and the pericenter distance is very small ( @xmath25 kpc ) ; ( like the dwarf in simulation o4 at the end of the evolution ) . in the case of ursa minor , which is much farther away from the center of the milky way at present
, the outer contours probably signify a transition to the tidal tails . for the remaining dwarfs
, the inner bar is detected only in the lmc and possibly in carina ( @xmath26 in the inner part ) .
in addition , fornax , carina , leo i and the smc have @xmath24 in the outer parts , signifying mild tidal extensions .
we have studied the shapes of dwarf galaxies orbiting the milky way using new discrete measures appropriate for systems where individual stars are resolved .
first , we traced the evolution of shape parameters in a set of collisionless @xmath0-body simulations of late type progenitors of present day dsph galaxies .
the initially disky dwarfs were placed on different orbits around a milky way - type galaxy and the inner structure of the satellites was also varied .
in addition to the strong mass loss and randomization of stellar orbits , tidal interactions inherent to such configurations lead to significant evolution of the shape . usually at the first pericenter passage
the dwarfs undergo bar instability and their overall distribution of stars transforms from the initial disks to triaxial bars .
the bars are preserved for a few gyrs of evolution but become shorter in time and produce , in the most tidally affected cases , almost spherical stellar components .
the effect of tidal forces on the shape of the dwarfs can be twofold .
in addition to the tidally induced bar , which can be easily identified with the full 3d information available in the simulations , the tidal forces can stretch the outer contours of the stellar distribution even in dwarfs that have already evolved into spheres .
this is particularly well visible at pericenters but can also occur at other parts of the orbit where the dwarf is not being strongly shocked , but is already surrounded by the material stripped from its main body in the form of tidal tails .
the two effects can be distinguished by measuring shapes separately in the inner and outer parts of the dwarfs .
we found the bar phase of the evolution to be quite common among the simulated dwarfs .
given that the dwarf models were not susceptible to bar instability in isolation , the presence of the bars can be considered as a signature of tidal stirring .
one has to keep in mind however , that the tidal forces are extremely effective not only in inducing the bars but also in destroying them .
if the tidal force is very strong , the bar can transform into a spherical shape over a rather short timescale .
it would therefore be very difficult to find a strong correlation between the formation of a bar and the strength of the tidal force or a dwarf s distance from the milky way .
nevertheless , given the variety of shapes within the dwarf galaxy population in the vicinity of the milky way we attempted to check if the tidal stirring scenario is broadly consistent with the properties of real dwarfs .
to this end we measured the shapes of brighter dwarf galaxies having good enough available photometry for large samples of stars . out of eleven dwarfs studied , we found clear signatures of inner bars in ursa minor , sagittarius , lmc and possibly carina . in six out of eleven dwarfs
we found indications of tidal extensions in the outer parts which likely signify transitions to the tidal tails .
the elongated shapes of ursa minor , sagittarius , carina and the inner bar of the lmc have been pointed out by many previous studies .
the results are however particularly interesting for sagittarius because its elongated shape is usually interpreted as due to tidal forces acting at the present time when the galaxy is at the pericenter of its orbit .
we find instead that the shape is elongated down to the very center of the dwarf thus lending support to the model proposed by okas et al .
( 2010a ) . in this scenario ,
the progenitor of sagittarius was a disk galaxy similar to the present lmc , which transformed into a bar at the first pericenter passage .
the bar is preserved until the present and the dwarf is currently at the second pericenter on its orbit around the milky way .
the presence of the inner bar in the lmc also seems consistent with the tidal stirring scenario which predicts that bars are typically formed at the first pericenter . since the lmc is probably on its first passage around the milky way ( besla et al .
2007 ) , it should be forming its bar right now .
it may seem surprising that among the eight classical dsph galaxies in the vicinity of the milky way a clear inner bar is detected only in ursa minor which , according to the scenario presented here , would imply that it has not passed many pericenters on its orbit around the milky way . on the other hand ,
ursa minor is dominated by the most metal - poor stellar population among the classical milky way dsphs ( kirby et al .
this means that it lost its gas rather early , probably via ram pressure stripping on its first passage around the milky way ( mayer et al .
2007 ) , a mechanism acting most effectively for dwarfs accreted early .
however , the gas may also have been lost via a strong blow - out in the formation of the initial stellar population while the system was still evolving in isolation .
given the stochasticity in the star formation history outcomes , as well as the wide distribution of orbital and structural parameters of the progenitor dwarfs , it would be difficult to exactly reproduce the history of any given dwarf just from its present shape .
the case of ursa minor however proves that bars are present even among the eight classical dwarfs and may be present also in other systems but not detectable because of their orientation along our line of sight .
the present study was based on photometric data alone .
however , considerable progress towards discriminating between different scenarios of dsphs formation is expected if these data are combined with kinematic measurements .
although rich samples of radial velocity measurements are presently available only for a few dsphs ( e.g. , walker et al . 2009 ) , such samples are bound to grow substantially in the near future thanks to a new generation of multi - object spectrographs ( such as the prime focus spectrograph on the subaru telescope ) .
kinematical samples for a few thousand stars could help detect remnant rotation that should be present in at least some dsphs if they formed from disky progenitors , as envisioned by the tidal stirring model .
such a rotation signal has until now only been tentatively detected for some dwarfs .
in addition , such data will also help distinguish bars from other structures .
bars seen perpendicular to the longest axis should show little rotation in the inner parts , contrary to disks seen edge on . on the other hand , rotation detected in systems that appear
spherical may indicate a bar seen along the longest axis ( see okas et al .
2010b ) .
we end with a note of caution that the picture established here with the help of collisionless simulations using standard assumptions concerning , e.g. , the initial dark matter profiles of the dwarfs , may well be modified if hydrodynamical processes are included or the assumptions changed .
although few hydrodynamical simulations of tidal stirring have been performed to date , they indicate that the shapes may be significantly affected by the presence of the gas and star formation ( mayer et al .
the details of the morphological transformation will also be different if the initial dark matter profile has a core rather than a cusp ( governato et al .
2010 ; okas et al .
2012 ) and the initial stellar and gaseous disk of the dwarf is thicker .
we plan to address these issues in our forthcoming papers .
this research was partially supported by the polish national science centre under grant n n203 580940 .
we wish to thank gary da costa for providing photometric data for the fornax dwarf in electronic form .
s.r.m . is grateful for financial support from nsf grants ast 97 - 02521 , ast 03 - 07851 , ast 03 - 07417 and ast 08 - 07945 , a fellowship from the david and lucile packard foundation , and a cottrell scholarship from the research corporation .
he also thanks william e. kunkel , james c. ostheimer , christopher palma , richard j. patterson , michael h. siegel , sangmo tony sohn , and kyle b. westfall for their help in producing the photometric catalogs used in this analysis .
is funded by the center for cosmology and astro - particle physics ( ccapp ) at the ohio state university . j.l.c .
acknowledges support from national science foundation grant ast 09 - 37523 . the work of l.a.m .
was carried out at the jet propulsion laboratory , under a contract with nasa .
l.a.m . acknowledges support from the nasa atfp program .
the numerical simulations were performed on the cosmos cluster at the jet propulsion laboratory .
this work also benefited from an allocation of computing time from the ohio supercomputer center ( http://www.osc.edu ) .
aihara , h. , et al .
2011 , apjs , 193 , 29 aparicio , a. , carrera , r. , & martinez - delgado , d. 2001 , aj , 122 , 2524 bellazzini , m. , ferraro , f. r. , origlia , l. , pancino , e. , monaco , l. , & oliva , e. 2002 , aj , 124 , 3222 bellazzini , m. , gennari , n. , & ferraro , f. r. 2005 , mnras , 360 , 185 besla , g. , kallivayalil , n. , hernquist , l. , robertson , b. , cox , t. j. , van der marel , r. p. , & alcock , c. 2007 , apj , 668 , 949 coleman , m. g. , da costa , g. s. , bland - hawthorn , j. , & freeman , k. c. 2005 , aj , 129 , 1443 coleman , m. g. , jordi , k. , rix , h .- w . , grebel , e. k. , & koch , a. 2007 , aj , 134 , 1938 gonidakis , i. , livanou , e. , kontizas , e. , klein , u. , kontizas , m. , belcheva , m. , tsalmantza , p. , & karampelas , a. 2009 , a&a , 496 , 375 governato , f. , et al .
2010 , nature , 463 , 203 hernquist , l. 1990 , apj , 356 , 359 irwin , m. , & hatzidimitriou , d. 1995 , mnras , 277 , 1354 kazantzidis , s. , okas , e. l. , callegari , s. , mayer , l. , & moustakas , l. a. 2011 , apj , 726 , 98 kirby , e. n. , lanfranchi , g. a. , simon , j. d. , cohen , j. g. , & guhathakurta , p. 2011 , apj , 727 , 78 kleyna , j. t. , geller , m. j. , kenyon , s. j. , kurtz , m. j. , & thorstensen , j. r. 1998 , aj , 115 , 2359 klimentowski , j. , okas , e. l. , kazantzidis , s. , mayer , l. , & mamon , g. a. 2009 , mnras , 397 , 2015 kunder , a. , & chaboyer , b. 2009 , aj , 137 , 4478 lee , m. g. , et al .
2003 , aj , 126 , 2840 okas , e. l. , kazantzidis , s. , majewski , s. r. , law , d. r. , mayer , l. , & frinchaboy , p. m. 2010a , apj , 725 , 1516 okas , e. l. , kazantzidis , s. , klimentowski , j. , mayer , l. , & callegari , s. 2010b , apj , 708 , 1032 okas , e. l. , kazantzidis , s. , & mayer , l. 2011 , apj , 739 , 46 okas , e. l. , kazantzidis , s. , & mayer , l. 2012 , submitted to apj letters , arxiv:1201.5784 majewski , s. r. , ostheimer , j. c. , kunkel , w. e. , & patterson , r. j. 2000a , aj , 120 , 2550 majewski , s. r. , ostheimer , j. c. , patterson , r. j. , kunkel , w. e. , johnston , k. v. , & geisler , d. 2000b , aj , 119 , 760 majewski , s. r. , skrutskie , m. f. , weinberg , m. d. , & ostheimer , j. c. 2003 , apj , 599 , 1082 mateo , m. l. 1998 , ara&a , 36 , 435 mateo , m. , olszewski , e. w. , & walker , m. g. 2008 , apj , 675 , 201 mayer , l. , governato f. , colpi , m. , moore , b. , quinn , t. , wadsley , j. , stadel , j. , & lake , g. 2001 , apj , 559 , 754 mayer , l. , kazantzidis , s. , mastropietro , c. , & wadsley , j. 2007 , nature , 445 , 738 mo , h. j. , mao , s. , & white , s. d. m. 1998 , mnras , 295 , 319 navarro , j. f. , frenk , c. s. , & white , s. d. m. 1996 , apj , 462 , 563 ( nfw ) palma , c. , majewski , s. r. , siegel , m. h. , patterson , r. j. , ostheimer , j. c. , & link , r. 2003 , aj , 125 , 1352 pietrzyski , g. , et al .
2008 , aj , 135 , 1993 pietrzyski , g. , grski , m. , gieren , w. , ivanov , v. d. , bresolin , f. , & kudritzki , r. p. 2009 , aj , 138 , 459 ricotti , m. , & gnedin , n. y. 2005 , apj , 629 , 259 sawala , t. , scannapieco , c. , maio , u. , & white , s. d. m. 2010 , mnras , 402 , 1599 saviane , i. , held , e. v. , & bertelli , g. 2000 , a&a , 355 , 56 schaefer , b. e. 2008 , aj , 135 , 112 siegel , m. h. , & majewski , s. r. 2000 , aj , 120 , 284 skrutskie , m. f. , et al .
2006 , aj , 131 , 1163 sohn , s. t. , et al .
2007 , apj , 663 , 960 stadel , j. g. 2001 , phd thesis , univ . of washington szewczyk , o. , pietrzyski , g. , gieren , w. , ciechanowska , a. , bresolin , f. , & kudritzki , r .-
2009 , aj , 138 , 1661 tassis , k. , kravtsov , a. v. , & gnedin , n. y. 2008 , apj , 672 , 888 tolstoy , e. , hill , v. , & tosi , m. 2009 , ara&a , 47 , 371 van der marel , r. p. , & cioni , m .-
l. 2001 , aj , 122 , 1807 walker , m. g. , mateo , m. , & olszewski , e. w. 2009 , aj , 137 , 3100 westfall , k. b. , majewski , s. r. , ostheimer , j. c. , frinchaboy , p. m. , kunkel , w. e. , patterson , r. j. , & link , r. 2006 , aj , 131 , 375 widrow , l. m. , & dubinski , j. 2005 , apj , 631 , 838 | we study the shapes of milky way satellites in the context of the tidal stirring scenario for the formation of dwarf spheroidal galaxies . the standard procedures used to measure shapes involve smoothing and binning of data and
thus may not be sufficient to detect structural properties like bars , which are usually subtle in low surface brightness systems .
taking advantage of the fact that in nearby dwarfs photometry of individual stars is available we introduce discrete measures of shape based on the two - dimensional inertia tensor and the fourier bar mode .
we apply these measures of shape first to a variety of simulated dwarf galaxies formed via tidal stirring of disks embedded in dark matter halos and orbiting the milky way .
in addition to strong mass loss and randomization of stellar orbits , the disks undergo morphological transformation that typically involves the formation of a triaxial bar after the first pericenter passage .
these tidally induced bars persist for a few gyr before being shortened towards a more spherical shape if the tidal force is strong enough .
we test this prediction by measuring in a similar way the shape of nearby dwarf galaxies , satellites of the milky way .
we detect inner bars in ursa minor , sagittarius , lmc and possibly carina .
in addition , six out of eleven studied dwarfs show elongated stellar distributions in the outer parts that may signify transition to tidal tails .
we thus find the shapes of milky way satellites to be consistent with the predictions of the tidal stirring model . |
the discovery that most elliptical galaxies and spiral bulges host a black hole at their centers , plus the tight relations observed between black - hole mass @xmath3 and galaxy luminosity @xmath9 or stellar velocity dispersion @xmath10 @xcite , suggest that the formation and growth of central black holes is deeply intertwined with that of their host galaxies .
recent theoretical work ( e.g. , @xcite ) supports this view , and also predicts how the relations between the properties of black holes and their host galaxies have changed over time .
direct observation of the evolution of the @xmath0 and @xmath1 relations over cosmological time would offer unique insight into galaxy and black - hole formation .
there have been many attempts to measure the evolution of the @xmath3 relations ; we cite a partial list of these : * @xcite examine a sample of quasars at redshifts up to 3.3 .
the luminosity @xmath9 and velocity dispersion @xmath10 of the host galaxy in such objects can not be measured reliably because the light from the galaxy is overwhelmed by the quasar flux ; instead , they use the width of the narrow [ o iii ] emission line as a surrogate for @xmath10 .
they estimate the black - hole mass using the `` photoionization '' method , which is based on an empirically calibrated relation involving the continuum luminosity of the active galactic nucleus ( agn ) and the width of the h@xmath11 or other broad emission lines .
the theoretical assumptions that underlie this relation are that the bulk velocities of the emitting clouds are determined by their orbital motion in the gravitational field of the black hole , and that the emitting region is photoionized by an ultraviolet continuum spectrum of fixed shape .
they find an @xmath0 relation that is consistent with the local one , suggesting that this relation is independent of redshift . * using similar methods on a larger but lower - redshift quasar sample from the sloan digital sky survey
, @xcite find that galaxies of a given dispersion at @xmath12 have black - hole masses that are larger by @xmath13 than at @xmath14 .
* @xcite and @xcite measured the velocity dispersion of 14 seyfert i galaxies at redshift @xmath15 ; they used stellar absorption lines , which should provide a more direct measure of the dispersion than emission lines .
they measured the black - hole mass using the photoionization method .
they found that galaxies of a given dispersion at @xmath15 host black holes having @xmath16 .
* @xcite measured the bulge luminosities of 11 quasar hosts in the redshift range @xmath17 using the hubble space telescope ( hst ) , and estimated their black - hole masses using the photoionization method .
they conclude that the @xmath1 relation at @xmath18 is close to the relation at @xmath14 ; @xmath19 .
this result is remarkable since black holes can only grow with time , while elliptical galaxies fade with time as their stars die : even if the black holes do not grow at all , passive stellar evolution models would predict @xmath20 between @xmath21 and @xmath22 at @xmath23 .
if their result is correct , then either black holes must be ejected from the galaxy centers and replaced with smaller ones , or the galaxy luminosity must grow substantially through mergers .
a common thread among all these investigations and others is that the high - redshift sample is selected by some measure of agn visibility .
black - hole masses and galaxy properties are then derived and an @xmath0 or @xmath1 relation is fitted to the data . since the low - redshift relations are linear in @xmath24 , @xmath6 , and @xmath7 ( eqs . [ eqn : msig ] or [ eqn : ml_hr ] below ) , the high - redshift data are often analyzed by assuming that the slope of the relation is the same as at low redshift and estimating the offset in the intercept or zero - point , which may be expressed as @xmath20 at fixed @xmath9 or @xmath10 .
any such offset is then interpreted as evidence that the ratio of @xmath3 to @xmath10 or @xmath9 has changed over time .
of course , the offset can be ( and sometimes is ) equally well expressed as @xmath25 or @xmath26 at fixed black - hole mass .
proof of evolution in the @xmath3 relations requires demonstrating that the high- and low - redshift galaxy samples were assembled with no selection effects that would bias the relation between the typical galaxy properties and black - hole masses being measured or , at least , that the selection effects in the high- and low - redshift samples were the same .
this may be a more difficult task than has commonly been assumed .
the local black - hole sample has mostly been drawn from `` normal '' galaxies with quiescent or low levels of nuclear activity . despite the proximity of these galaxies ,
detecting and weighing their central black holes requires exquisite spatial resolution , very high signal - to - noise observations , and elaborate modeling of the observations .
investigating the properties of black holes in nearby galaxies remains a frontier problem ; to date , dynamically determined black - hole masses are available for only slightly more than three dozen galaxies . since there was little evidence of a black hole in most of these galaxies before hst observations were taken ( except sometimes for a modest rise in the velocity dispersion in spectra taken at ground - based resolution ) , and since black holes are found in almost all nearby galaxies that have been examined carefully with hst , the black holes in local quiescent galaxies are selected mainly on the basis of galaxy properties such as @xmath9 or @xmath10 .
the techniques used to measure black - hole masses in quiescent galaxies at low redshift can not be used at high redshift , both because the radius of influence of the black hole can not be resolved beyond a few tens of mpc , and because the galaxies have substantially lower surface brightness , by the factor @xmath27 .
black holes at high redshift are identified instead by their association with agn .
the observer thus _ first _ locates a black hole that is accreting matter and weighs it by the properties of the agn emission lines ( a difficult task , but one that we shall not examine in detail here ) , and then _ secondly _ attempts the ( still difficult ) task of measuring the properties of the host galaxy .
the existence and properties of the agn depend both on the properties of the black hole ( mass , spin , orientation ) and on the properties of the galaxy ( mass inflow rate , orientation , etc . ) . in short ,
selection is done at low redshift by galaxy properties , and at high redshift by a combination of black - hole and galaxy properties that depends on the method used .
galaxy samples obtained with different selection techniques will generally satisfy different @xmath3 relations .
this effect , analogous to the @xcite bias that is familiar in studies of galactic structure , arises because of the cosmic scatter in the @xmath3 relations there is not ( so far as we know ) an exact 11 relation between black - hole mass and any single measurable property of galaxies such as @xmath9 or @xmath10 . in this paper
we argue that determinations of the redshift evolution of the @xmath3 relations may be strongly biased by selection effects unless the same sample selection techniques are used at all redshifts .
correcting for the selection bias introduced by the use of different sample selection criteria at different redshifts is extremely difficult : to make accurate corrections it is necessary to know both how the selection depends on the galaxy and black - hole properties and the cosmic scatter in the @xmath3 relations . at present
we have only crude upper limits to the latter quantity .
we shall show that even if the cosmic scatter were known , the bias in the samples described above usually can only be corrected with additional information , such as how galaxy properties and black - hole mass determine the probability distribution of agn luminosity
. the selection bias can be especially large at large @xmath3 , as the following argument shows .
consider the @xmath1 relation . at high
@xmath9 the number density of galaxies falls off rapidly , as illustrated by the steep cutoff in the @xcite luminosity function .
cosmic scatter in the @xmath1 relation implies that there is a distribution of black - hole masses at a given @xmath9 .
the rare high - mass black holes can arise from either the peak of this distribution in the rare galaxies with large @xmath9 , or from the high - mass tail of the distribution in the more numerous galaxies of modest @xmath9 .
if the number density of galaxies is falling off rapidly with @xmath9 , the contribution from galaxies with modest @xmath9 may actually overwhelm the population of black holes of similar mass associated with galaxies of higher @xmath9 .
this problem was first explicitly identified by @xcite in the context of the correlation between black - hole mass and dark - matter halo mass , and by @xcite in the context of the @xmath0 relation , but apparently has not been appreciated by most observers studying the evolution of the @xmath3 relations .
in this paper we present a more general exposition of the biases incurred when studying evolution of the @xmath3 relations by comparing samples obtained by different selection criteria and at different redshifts .
we start by considering the extraction of samples from hypothetical joint distributions of @xmath3 and @xmath9 or @xmath5 we then consider the selection bias that occurs if the high - redshift samples are selected by agn flux or luminosity .
we discuss the prospects of correcting for selection bias , and argue that selection bias may place fundamental limits on the determination of the evolution of the @xmath3 relations .
we conclude with a brief review of attempts to measure the evolution of the @xmath3 relations and how they may have been affected by selection bias .
understanding object selection bias requires knowledge of the joint probability distribution of black - hole mass @xmath3 and a galaxy property @xmath28 which for the present discussion is either @xmath6 , where @xmath9 is the rest - frame @xmath29-band galaxy luminosity , or @xmath7 , where @xmath10 is the line - of - sight velocity dispersion in the main body of the galaxy ( the precise definition of @xmath9 or @xmath10 does not concern us , so long as it is defined consistently in all samples ) .
the true form of the joint distribution unfortunately is unknown ; however , we can construct a hypothetical form of this distribution that accurately represents the present state of the observations , and in any case suffices to show how sample selection bias can occur .
we write the probability of finding a galaxy in the interval @xmath30 and @xmath31 as @xmath32 where @xmath33 . from the definition of conditional probability
this can be rewritten as @xmath34 where @xmath35 is the probability that a randomly chosen galaxy in a given volume lies in the interval @xmath31 and @xmath36 is the probability that the black - hole mass lies in the range @xmath30 given that the galaxy property is @xmath37 .
the function @xmath38 is then given by either the volume - limited luminosity function more properly , the luminosity function of early - type galaxy components ( ellipticals and spiral bulges ) , since these are the components that correlate with black - hole mass or the velocity - dispersion distribution of early - type components . without loss of generality ,
we may write @xmath39 $ ] where @xmath40 and @xmath41 , so @xmath42 is the mean value of @xmath43 at a given value of the galaxy property @xmath37that is , @xmath42 is either the @xmath1 or @xmath0 relation .
the limited observational data on black - hole masses in local galaxies is consistent with the hypothesis that @xmath44 is independent of @xmath45 adopting this hypothesis for simplicity we have @xmath46g(s ) .
\label{eqn : joint}\ ] ] the intrinsic variance or `` cosmic scatter '' in black - hole mass at a given value of galaxy property @xmath37 is @xmath47 observational errors may make @xmath44 appear to be yet broader , but for this discussion , we assume that observational errors are negligible ; the selection effects that we are concerned with are solely due to the fact that there is an intrinsic and irreducible range of @xmath3 at any @xmath48 it should be stressed that although the functional form ( [ eqn : joint ] ) is consistent with the available data , it is far from unique . as a foil , we may consider the form @xmath49\phi_\bullet(\mu ) .
\label{eqn : jointa}\ ] ] a physical model that motivates ( [ eqn : joint ] ) is one in which the galaxy property @xmath37 determines the black - hole mass @xmath43 with a cosmic scatter described by the function @xmath50 , while ( [ eqn : jointa ] ) is motivated by models in which the black - hole mass @xmath43 determines the galaxy property @xmath37 , with cosmic scatter described by the function @xmath51 .
if @xmath52 , then @xmath53 is the number of black holes per unit volume with log mass in the range @xmath54 we do not know which of ( [ eqn : joint ] ) or ( [ eqn : jointa ] ) is correct ( possibly neither , or both ) ; our motivation for choosing the former is that it provides a simple representation of what local observations of black holes in inactive galaxies measure . of the three functions that contribute to @xmath55
, @xmath38 is probably the best determined .
we discuss the galaxy - property functions in detail in @xcite , but summarize them briefly here . for the galaxy luminosity function
we use the @xcite sloan digital sky survey ( sdss ) luminosity function , transformed to the @xmath29-band and redshift @xmath56 , augmented with the @xcite luminosity function of brightest cluster galaxies , which appear to be undercounted in the sdss .
we note that the blanton et al .
function refers to total rather than bulge luminosity in s0 and spiral galaxies ; however , this difference is less important at the bright end of the luminosity function , where the selection effects are strongest . for the velocity - dispersion function
we use the @xcite sdss results , augmented at the high-@xmath10 end as prescribed by @xcite to account for an artificial high-@xmath10 cut - off in sheth et al .
both functions are shown in figure [ fig : g ] .
the function @xmath42 should be the least - squares fit of @xmath24 to @xmath6 or @xmath7 in a sample selected by galaxy properties .
the @xmath0 relation @xmath42 is based on the galaxy sample from @xcite augmented by a few galaxies with more recent @xmath3 determinations ( see @xcite ) .
in contrast to the treatment in @xcite , which treated @xmath3 and @xmath10 symmetrically , the appropriate treatment for our purposes is a least - squares fit of @xmath24 on @xmath7 , which gives : @xmath57 for @xmath58 ( which we will use throughout this paper ) .
we discuss the @xmath1 relation in detail in @xcite ; in brief we use the @xcite relation between @xmath3 and galaxy _ mass _ , transformed back to luminosity by adopting the mass - to - light ratio @xmath59 based on the @xmath60 estimates given in @xcite
. a least - squares fit of @xmath3 on @xmath61 for galaxies with @xmath62 gives : @xmath63 equations ( [ eqn : msig ] ) and ( [ eqn : ml_hr ] ) have the form @xmath64 where @xmath65 and @xmath66 are constants .
@xcite show that the @xmath0 and @xmath1 relations make different predictions for @xmath3 in luminous or high - dispersion galaxies , which have lower @xmath10 values for a given luminosity than is implied by eliminating @xmath3 from equations ( [ eqn : msig ] ) and ( [ eqn : ml_hr ] ) .
thus at least one of the @xmath0 or @xmath1 relations must curve away from the linear fit ( [ eqn : msrel ] ) in the most luminous galaxies .
we are not concerned with this issue here , and simply show the different selection biases that can result from using @xmath10 or @xmath9 as the galaxy property that correlates with @xmath3 , _ assuming _ that the linear relation ( [ eqn : msrel ] ) holds at all masses .
however , the reader should bear in mind that neither the @xmath1 or @xmath0 relation is well determined at high galaxy masses , where the potential selection bias is most important .
the scatter function , @xmath67,$ ] is poorly known at best . for this analysis
, we assume that at any galaxy @xmath9 or @xmath10 , @xmath68 is described by a normal distribution about @xmath42 , with cosmic scatter @xmath69 .
thus @xmath70\,d\mu= { d\mu\over \sqrt{2 \pi{\sigma_\mu}^2 } } \exp\left [ - { [ \mu - f(s)]^2 \over 2{\sigma_\mu}^2}\right ] .
\label{eqn : prob}\ ] ] again , we emphasize that @xmath69 does not embody any observational errors in the determination of @xmath71 it represents the intrinsic spread in @xmath3 at any galaxy property .
@xcite conclude that only an upper limit to @xmath69 is presently known for either the @xmath0 or @xmath1 relation , both because of the small sample of reliable @xmath3 determinations , and because of uncertainties in the observational errors in @xmath3 , which must be accurately measured to isolate the contribution of @xmath69 to the total residuals .
for the present analysis we will explore the sensitivity of the selection biases to @xmath69 on the assumption that @xmath72 for dispersion @xmath10 and @xmath73 for luminosity @xmath9 .
apart from the poor knowledge of @xmath69 , there are at least two other major uncertainties in the function @xmath50 : ( i ) a more general treatment would allow for the possibility that it varies with @xmath37 ; ( ii ) there is little or no justification of the assumed normal form .
as will be shown below , for the most massive black holes or galaxies , the selection bias may depend on the form that @xmath67 $ ] takes at several standard deviations away from the mean , yet an observational determination of the form of @xmath50 in this region would require a sample of well - determined black - hole masses several orders of magnitude larger than is presently available .
thus at the highest black - hole masses , the selection bias is likely to depend sensitively on knowledge that is not presently at hand , so it is not possible to apply reliable corrections for this bias . with these caveats , in figure [ fig:2df ]
we show estimates of @xmath55 as modeled by equation ( [ eqn : joint ] ) for each of the @xmath0 and @xmath1 relations , and two assumed values of @xmath74 projection of @xmath55 onto the @xmath43-axis produces the mass function of black holes , @xmath75g(s)\,ds.\end{aligned}\ ] ] the importance of cosmic scatter for converting the luminosity or dispersion function @xmath38 to a black - hole mass function has been discussed several times @xcite .
what may be less appreciated , however , is that cosmic scatter implies that the most massive bhs are often hosted by modest galaxies that _ a priori _ would not be expected to harbor bhs of high mass . to illustrate this point , figure [ fig : bh_df ] shows @xmath76 for four different versions of @xmath77
the salient feature is that as the cosmic scatter @xmath69 increases , the contribution to the density of the most massive black holes from the wings of the scatter function @xmath50 overwhelms the `` native '' population of massive black holes harbored by the galaxies with the largest values of property @xmath37 .
a second illustration of this point occurs in the top two panels in figure [ fig : bh_cuts ] , which show the probability distribution of @xmath9 and @xmath10 for @xmath78 : note that the solid curves , corresponding to @xmath79 for @xmath9 and 0.3 for @xmath10 , have a prominent shoulder to the right of the peak , the peak arising from low - luminosity or low - dispersion galaxies , and the shoulder from high luminosities and dispersions .
the source of the selection bias can be seen more directly in figure [ fig:2df ] , which shows @xmath55 for each of the @xmath0 and @xmath1 relations , and two assumed values of @xmath69 . only for @xmath80
is there an exact relation between @xmath3 and @xmath10 or @xmath81 our model for @xmath55which may not be correct , but is consistent with the data implies that the density contours in figure [ fig:2df ] are symmetric about the mean relationship ridge - lines in the vertical direction ; in other words , the conditional probability of @xmath43 given @xmath37 is symmetric about @xmath82 . on the other hand , the conditional probability of @xmath37 given @xmath43 ,
@xmath83 is not symmetric about @xmath84 , as seen in figure [ fig : bh_cuts ] , which shows the distribution of @xmath37 at selected values of @xmath8 as @xmath3 and @xmath69 increase , the distribution of @xmath37 for a given @xmath3 moves further and further away from the value implied by the inverse of the @xmath3 relations , @xmath85 , where @xmath65 and @xmath66 are the coefficients given in equation ( [ eqn : msrel ] ) .
for @xmath86 @xmath87 generally falls well out in the wings of the distribution of @xmath37 for both @xmath9 and @xmath88 particularly for the larger values of @xmath74 the galaxies with @xmath84 for the highest mass black holes are so far down in the step cutoffs of the @xmath9 and @xmath10 distribution functions that they are completely overwhelmed by the population of galaxies of modest mass that harbor high mass black holes as statistical outliers . the mean value of @xmath37 at a given @xmath3 is @xmath89\,ds\over \int g(s ) h\left[\mu - f(s)\right]\,ds}. \label{eqn : means}\ ] ] this is shown as the red lines in figure [ fig:2df ] .
the mean of @xmath6 or @xmath7 at a given @xmath3 is offset from @xmath90 which is just the @xmath37 location of the @xmath3 relation ridgelines shown in the figure . in the presence of cosmic scatter
, the mean @xmath37 of a sample of galaxies that host a black hole of given @xmath43 is different from the mean @xmath43 hosted by a sample of galaxies of given @xmath37 .
this difference is the source of the selection bias .
selection bias typically occurs because the galaxy samples used to probe the @xmath3 relations at cosmological distances are not selected by galaxy @xmath9 or @xmath10 , but by the visibility of their agns . for most of the discussion in this section
we shall assume that agn luminosity does not depend directly on @xmath4 @xmath88 or any other galaxy property unrelated to @xmath8 this model is appropriate if , for example , the probability that the agn associated with a black hole of mass @xmath3 has luminosity @xmath91 is given by @xmath92 and @xmath93 and @xmath94 .
the physical content of this assumption is that the luminosity history of an agn is determined by the black - hole mass and scales with the eddington luminosity , which is proportional to @xmath3 .
we first consider the bias that occurs in a sample in which ( i ) all of the objects are in a narrow redshift range ; ( ii ) the survey contains all agns brighter than a given flux . in this case
the probability that an agn in the relevant redshift range is accepted in the survey depends only on its luminosity @xmath91 , and by equation ( [ eq : probone ] ) this in turn depends only on its black - hole mass @xmath43 .
thus the probability distribution of a galaxy property @xmath37 at a given value of @xmath43 is not biased by the selection , and is simply @xmath95 ( eq .
[ eq : scond ] ) .
some examples are shown in figure [ fig : bh_cuts ] , which plots the probability distribution of @xmath96 for several values of black - hole mass @xmath3 and cosmic scatter @xmath69 . here
@xmath97 is the inverse of the @xmath1 or @xmath0 relation ( eq .
[ eqn : joint ] ) .
the mean values of the distributions @xmath98 are displaced to values lower than @xmath87 in most of the examples shown .
as expected , the width and offset of the distributions are larger for larger values of @xmath69 . also as expected , the offsets are generally larger for more massive black holes .
@xmath99 is shown as the red lines in figure [ fig:2df ] ; the offset of these lines from the relation @xmath82 demonstrates the selection bias . in this model
, the selection bias can be corrected : an assumed form for the probability distribution @xmath55 such as ( [ eqn : joint ] ) with redshift - dependent parameters can be used to compute @xmath95 , which can then be fitted to the distribution of galaxy property @xmath37 in the sample at a given value of black - hole mass @xmath43 ( separate issues , discussed in [ sec : difficult ] , are whether this model for the selection effects is realistic and whether the assumed form of @xmath55 is correct ) .
however , most papers in the literature have not taken this approach , so it is worthwhile to estimate the biases that might be introduced by using simpler statistics to estimate the evolution in the @xmath3 relations .
figure [ fig : bh_bias ] shows the object selection bias @xmath100 as a function of @xmath3 and @xmath69 . in this example we have expressed the bias in terms of galaxy properties , since we have selected by @xmath71 however , it may also be represented as @xmath101 where @xmath66 is the slope given in equation ( [ eqn : msrel ] ) .
note that the bias is not always a monotonic function of black - hole mass , nor does it always have the same sign . in this simple model ,
the bias at black - hole masses of @xmath102 is @xmath103 in the @xmath0 relation for @xmath104 , and @xmath105 in the @xmath1 relation for @xmath79and even larger at larger black - hole masses .
the bias is smaller at lower @xmath9 or @xmath10 , but we caution that the near - zero bias in the @xmath1 relation at small @xmath3 is an artifact of our assumption that the luminosity function has the @xcite form at low luminosities ( see below ) , and this may not be true for the early - type galaxy components that are believed to host the black holes .
figure [ fig : bh_bias ] also shows that a significant portion of the bias comes from galaxies with @xmath106 for the larger values of @xmath74 indeed for @xmath107 nearly half of the bias comes from such galaxies for all but the smallest @xmath69 shown .
thus , not only is the bias sensitive to @xmath108 it also depends on the shape of the wings of the error distribution .
if , as is likely , the wings are more extended than in our assumed log - normal distribution , the bias will be even larger .
many of the features in these plots can be understood analytically . from the definition of @xmath98 ( equation [ eqn : means ] ) and equation ( [ eqn : msrel ] )
, we can write @xmath109\,dx\over \int h(x)g[b^{-1}(\mu - a - x)]\,dx}. \label{eq : meanss}\ ] ] if the cosmic scatter @xmath110 is not too large , we can evaluate this by expanding @xmath38 in a taylor series , @xmath111_{s=(\mu - a)/b}+\hbox{o}(\sigma_\mu^4 ) .
\label{eq : meansss}\ ] ] this result shows that ( i ) the bias is proportional to @xmath112 if @xmath69 is not too large ; ( ii ) the bias in @xmath98 is positive if @xmath38 declines with @xmath37 and proportional to @xmath113this is why the bias is large and negative for the most luminous or high - dispersion galaxies ; ( iii ) the bias is near zero if @xmath38 is constant , which corresponds to a luminosity function @xmath114 when @xmath115 .
the bias in the left panel of figure [ fig : bh_bias ] is small for low - luminosity galaxies because the assumed luminosity function is close to this form .
a different statistical method , which more closely parallels the approach used in most observational papers , is simply to estimate the average value of the difference between @xmath43 and the prediction of the @xmath116 relation ( [ eqn : msrel ] ) for all the galaxies in the sample , @xmath117 let us assume that the survey flux limit at the given redshift corresponds to a luminosity @xmath118 with @xmath119 .
then @xmath120 where @xmath121 , and @xmath122 the luminosity function ( number per unit volume ) of agns is @xmath123 , where @xmath124 which is just the inner integral of the denominator of ( [ eq : xxxx ] ) . to evaluate the inner integral in the numerator of equation ( [ eq : xxxx ] ) , note that the functional form of @xmath125 ( eq .
[ eqn : prob ] ) implies that @xmath126 , so @xmath127 integrating the right side of the equation by parts , and noting that @xmath128 gives @xmath129 thus @xmath130 remarkably , the result depends only on the directly observable luminosity function @xmath131 for the type of agn targeted in the survey , and is independent of assumptions about the luminosity history @xmath132 or the distribution of host galaxy properties @xmath38 .
also , in contrast to the analogous result ( [ eq : meansss ] ) , equation ( [ eq : biasone ] ) is not just the first term in a taylor series in @xmath112 but valid for all values of @xmath112 , no matter how large . to illustrate the application of this result
, we adopt the quasar luminosity function from @xcite , @xmath133 with @xmath134 and @xmath135 , where @xmath136 is the `` break '' luminosity .
the corresponding bias @xmath20 is shown as a function of the lower luminosity limit @xmath137 in figure [ fig : bh_biasst ] .
note that the bias in figure [ fig : bh_biasst ] becomes smaller but does not vanish as the survey goes to fainter and fainter luminosity limits : asymptotically , @xmath138 which equals @xmath139 . as a second example
, we examine the bias in a flux - limited sample of agn .
we assume that the survey galaxies are distributed uniformly in euclidean space , that the probability distribution of agn luminosities is given by equation ( [ eq : probone ] ) , and that the survey contains all galaxies with flux @xmath140 exceeding some limiting flux @xmath141 .
the bias is then given by @xmath142 where @xmath143 is the volume within which an agn of luminosity @xmath144 can be detected , @xmath145 is the solid angle covered by the survey , and @xmath146 is the number density of galaxies .
we have @xmath147 where @xmath148 is defined by equation ( [ eq : defphi ] ) . to evaluate the inner integral in the numerator of equation ( [ eq : xxx ] ) we use equation ( [ eq : parts ] ) , which yields @xmath149 where the last line follows from an integration of the numerator by parts .
the results of the previous subsections depend on the assumption that the probability distribution of agn luminosity is determined by the black - hole mass and not the galaxy properties ( eq . [ eq : probone ] ) .
a foil to this is to assume that the luminosity is determined by the galaxy properties and not the black - hole mass .
thus equation ( [ eq : probone ] ) is replaced by @xmath150 where @xmath151 is a constant . in this case @xmath152 in any flux - limited or luminosity - limited sample @xcite .
given the example bias calculations shown in the previous section , it may be tempting to conclude that similar bias corrections may be estimated and applied after the fact to existing surveys to determine the @xmath0 or @xmath1 relations .
we believe that in general such corrections are difficult or impossible to apply reliably to current surveys , for the following reasons . in the examples above , we assumed that at any galaxy property @xmath28 @xmath94 followed a normal distribution about the mean @xmath153 relation @xmath82 , with standard deviation characterized by a cosmic scatter @xmath74 we saw that the bias is very sensitive to @xmath69 ( in most examples @xmath154 ) yet this parameter is poorly known for either the @xmath0 or @xmath1 relations , in part because an accurate estimate of @xmath69 requires knowing the observational errors in the @xmath3 determinations . @xcite
have studied this problem and conclude only that @xmath155 for the @xmath0 relation and @xmath156 for the @xmath1 relation . moreover , the @xcite analysis assumes , as we do , that @xmath69 is constant over the parameter range of the @xmath3 relations ; the sample of galaxies with well - determined @xmath3 is simply too small to explore the possibility that it is not .
even if an accurate estimate of @xmath69 were available , the functional form of the distribution of @xmath157 is unknown .
the assumption of a normal distribution is an obvious first step , yet the steep falloff of the @xmath9 or @xmath10 distribution at large values means that the selection bias for large black - hole masses is sensitive to the wings of this distribution , where @xmath158 .
the assumed normal distribution is likely to _ under_estimate the selection bias since more realistic distributions have fatter tails . given that the local black - hole sample is so small that @xmath69 is poorly determined , the sample of galaxies with good @xmath3 determinations would have to be several orders of magnitudes larger before an accurate form for the scatter function could be determined .
the @xmath3 relations are imperfectly known at @xmath159 and in particular the range of @xmath9 or @xmath10 that is used to determine them is rather limited .
for example , in the sample of 31 local galaxies with measured black - hole masses used by @xcite to determine the @xmath0 relation , the interquartile ranges in @xmath9 and @xmath10 are only factors of 5.3 and 1.6 respectively .
thus , the calculation of the selection biases for @xmath160 must contend with the fact that there are only a few black holes observed in this mass range in the local sample ( 4 in the @xcite sample ) .
@xcite show that the @xmath161 relation must be curved ( in logarithmic coordinates ) in the sense that @xmath10 appears to increase only slowly ( if at all ) with @xmath9 for the most luminous galaxies .
this implies that the present log - linear @xmath0 and @xmath1 relation can not be consistent at high galaxy luminosity .
the joint @xmath153 distributions shown in figure [ fig:2df ] are thus based on extrapolated estimates at the highest @xmath3 values that are likely to change if more and better determinations of black - hole masses @xmath160 become available .
the heart of our analysis of selection bias in [ sec : bhmass ] and [ sec : bhmasstwo ] is the assumption that the black hole determines the agn luminosity , or more precisely that the galaxy properties do not .
this is clearly true if , for example , ( i ) black holes radiate either at the eddington luminosity @xmath162 or not at all , and ( ii ) the bolometric correction is independent of galaxy properties .
this assumption is also true in more general circumstances , for example , ( i ) can be replaced by the weaker assumption that the probability that a black hole is radiating at some @xmath91 depends only on the ratio @xmath163 ( eq . [ eq : probone ] ) .
this assumption may be approximately correct : for example , @xcite find that their simulations are well - fit by a model in which the probability distribution of agn luminosities depends only on @xmath164 where @xmath165 is the peak luminosity of the agn and @xmath166 .
nevertheless , if galaxy properties do affect the agn luminosity most likely by determining the feeding rate of matter onto the black hole at luminosities @xmath167then the biases will be different from those presented in [ sec : bhmass ] and [ sec : bhmasstwo ] . in the limiting case that the agn luminosity is determined entirely by galaxy properties rather than black - hole mass , as might be plausible for low - luminosity agns , there is no selection bias in a flux- or luminosity - limited survey ( [ sec : prop ] ) .
we have assumed that the joint probability distribution of black - hole mass @xmath43 and galaxy property @xmath37 can be modeled by equation ( [ eqn : joint ] ) .
this corresponds to a physical model in which the galaxy property is the independent variable , and the black - hole mass is determined by the properties of the galaxy through the cosmic - scatter function @xmath168 and the @xmath3 relation @xmath42 .
this assumption is convenient for modeling the local sample of galaxies with @xmath3 determinations , which was selected by galaxy properties , plus the ready knowledge of @xmath169 the volume distribution of @xmath48 the alternative model provided by equation ( [ eqn : jointa ] ) , in which the black - hole mass is the independent variable and the galaxy property is determined by the black hole through the function @xmath170 , is much more difficult to fit to the local observations as we have little direct knowledge of @xmath171 the black - hole mass function .
however , our present understanding of black - hole and galaxy formation does not allow us to say which model , if either , is correct .
most of our estimates of selection bias have been based on the assumption that surveys are complete to a given agn luminosity or flux , but this is an oversimplification .
any survey that is based on measurements of the luminosity of the host galaxy requires that the host is bright enough and large enough to be separated from the agn .
any survey that uses the velocity dispersion measured from stellar absorption lines requires that the absorption lines are strong enough , and that agn emission lines in the vicinity of the absorption lines are weak enough , to allow a reliable dispersion measurement . at the opposite extreme , surveys of low - luminosity agns require that the emission lines are strong enough to be detected against the continuum flux from the galaxy .
in such studies the selection depends in a complex way on properties such as the ratio of agn to galaxy luminosity , and on galaxy properties other than @xmath9 or @xmath10 , such as the effective radius or central surface brightness .
we have intended this paper mainly as a planning guide for future surveys to explore the evolution of the @xmath3 relation , rather than as a critique of existing surveys .
the following brief comments on existing surveys are mostly intended to illustrate the application and impact of the estimates of selection bias that we have made in earlier sections .
several authors have determined the @xmath0 relation for nearby agns , and compared this to the local relation for inactive galaxies . using seven seyfert 1 galaxies with black - hole masses measured by reverberation mapping , @xcite find @xmath172 . for 16 agns with reverberation - based black - hole masses @xcite find @xmath173 ( although onken et al .
interpret their result in terms of calibrating the reverberation - mapping method rather than an offset in the @xmath0 relation ) .
@xcite have compared the black - hole masses in dwarf seyfert 1 galaxies to an extrapolation of the @xmath0 relation to lower dispersions ; using black - hole masses based on the photoionization method calibrated by the standard `` isotropic '' formula , they find @xmath174 ( although they interpret their results with onken et al.s calibration and thus find @xmath175 ) . @xcite have measured the @xmath0 relation in 88 nearby agn using the photoionization method to determine @xmath3the sample is much larger than that of @xcite but the photoionization mass estimates are less direct than those from reverberation mapping .
they find @xmath176 .
similarly , @xcite have determined the @xmath1 relation from a sample of 29 quasars with @xmath177 ; they determine black - hole masses using the photoionization method and luminosities using hst photometry .
they find @xmath178 , although they interpret their result in terms of recalibrating the photoionization method .
a common feature of the @xcite , @xcite , @xcite , and @xcite samples is that their black - hole masses are relatively modest : generally @xmath179 and often more than an order of magnitude smaller .
selection bias occurs at all @xmath2 depending on the local slopes of the @xmath9 and @xmath10 distributions , but these samples are safely removed from the very strong biases seen at large values of @xmath9 , @xmath10 , and @xmath3 , where the distribution functions are falling rapidly . although we have emphasized selection bias in the context of low- and high - redshift samples with similar black - hole masses , one selected by galaxy properties and one by black - hole mass , selection bias can also arise if the two samples are selected in the same way but are centered on different black - hole mass ranges .
we further note that selection effects in these samples are difficult to model for example , ( i ) as @xcite point out , the agn variability timescale depends on luminosity , and if the timescale is too long , reverberation mapping is impractical ; ( ii ) to be able to measure the velocity dispersion requires some minimum ratio of the luminosity of the host galaxy to the luminosity of the agn . if we make the simplest possible assumption , that the samples of nearby galaxies are complete but flux - limited , then equation ( [ eq : fluxlim ] ) implies @xmath180 .
@xcite , building on the initial work of @xcite , examine a sample of 14 seyfert 1 galaxies at @xmath181 , measuring the velocity dispersions from stellar absorption lines and the black - hole masses by the photoionization method .
compared to the local sample of inactive galaxies with measured black - hole masses , they find @xmath16 ; note that this result is based on a calibration of the photoionization masses that _ assumes _ that the @xcite sample should have exactly the same @xmath0 relation as inactive galaxies , which need not be the case if selection bias is accounted for .
@xcite argue that their offset @xmath20 is robust because the black - hole masses are measured by the same technique , with the same calibration , in the local and high - redshift samples , but the selection bias in the two samples is likely to be quite different ; the @xcite sample is more nearly flux - limited , while the @xcite sample is more nearly luminosity - limited , and the selection bias in these two cases is quite different ( eqs . [ eq : fluxlim ] and [ eq : biasone ] ) .
a useful next step in assessing the bias would be to determine the luminosity function for agn selected by the criteria used by @xcite , for use in equation ( [ eq : biasone ] )
. it may also be significant that the typical @xmath182 in the @xcite sample is roughly an order of magnitude less massive than the typical @xmath183 in the @xcite sample .
as is evident in figure [ fig : bh_bias ] , the bias is a strong function of @xmath2 and indeed the bias in the @xmath0 relation changes sign at @xmath184 .
thus we would expect that selection bias might affect the results even if the @xcite sample were chosen in precisely the same way as the @xcite sample .
figure [ fig : bh_bias ] predicts a bias of @xmath185 at @xmath183 for @xmath186 .
this is only 40% of the offset claimed by @xcite .
thus , it appears that selection bias can not explain all or even most of the @xcite @xmath187 ; however , given the uncertainties discussed in the previous section , the possibility that the correction may be as large as the claimed offset can not be ruled out .
@xcite find @xmath19 compared to the local @xmath1 relation , using a sample of 11 quasars at @xmath23 ; the black - hole masses are determined with the photoionization method .
if we assume that the velocity dispersion of the host galaxy does not change between @xmath188 and @xmath14 , and that the luminosity of the host fades by 12 magnitudes as predicted by passive stellar evolution , then this result corresponds to @xmath1890.8 as measured by the @xmath0 relation . however , the potential selection bias is rather large for this sample , partly because the cosmic scatter in the @xmath1 relation may be larger than in the @xmath0 relation .
if the sample is luminosity limited , the model shown in the left panel of figure [ fig : bh_bias ] predicts @xmath190 for @xmath79 for the range of black - hole masses in the sample , and the flux - limited model predicts @xmath191 for @xmath79 ( note that our assumption of a euclidean metric is not valid at @xmath188 , and should be generalized to a realistic cosmological model ) .
a further complication is that their method requires measuring the bulge luminosity in the presence of the bright agn , so imposes a selection on the ratio of agn to galaxy luminosity . without an accurate estimate of the selection bias for the @xcite sample
, their measurement of @xmath192 does not provide clear evidence for rapid evolution in the ratio of black - hole mass to stellar mass in the galaxy .
@xcite find that galaxies of a given dispersion at @xmath12 have black - hole masses that are larger by @xmath193 than at @xmath194 .
in contrast to most other studies , they carefully consider the effects of selection bias using monte - carlo models and conclude that for @xmath104 selection bias increases @xmath24 by @xmath195 at low redshift and @xmath196 at @xmath12 , contributing a net offset of @xmath197 .
they also argue that scatter in the relationship between the width of the [ o iii ] emission line and the true stellar dispersion contributes an additional offset of @xmath198 , so after correcting for these biases @xmath13 .
their estimates of selection bias are similar to , but somewhat smaller than , the bias that we estimate for their sample from equations ( [ eq : biasone ] ) and ( [ eq : fluxlim ] ) , assuming that the low - redshift sample is flux - limited and the high - redshift sample is luminosity - limited ; we believe our estimates are more robust because they do not depend on an assumed form for the distribution @xmath132 of agn luminosities at a given black - hole mass .
finally , we note that theoretical models of the evolution of the @xmath0 relation @xcite predict @xmath199 at @xmath200 .
strong selection bias could reduce or eliminate the difference between this prediction and the positive values of @xmath20 found at high redshift in most of the above studies .
we have described a selection bias that affects investigations into the cosmological evolution of the @xmath0 or @xmath1 relations .
we summarize the main points as follows : * cosmic scatter in the @xmath0 or @xmath1 relations means that a galaxy of given @xmath9 or @xmath10 will host black holes with a range of @xmath3 values . *
the steep decline in the luminosity function for the most luminous galaxies means that the rare occurrence of high - mass black holes in numerous `` modest '' galaxies overwhelms the frequent occurrence of black holes of similar mass in the rare galaxies with very high luminosity .
a similar conclusion applies to dispersion instead of luminosity . *
selection of a sample of black holes hosted by inactive galaxies at low redshift explores the distribution of black - hole masses @xmath3 that a galaxy will host , given its luminosity @xmath9 or dispersion @xmath5 selection of a sample by agn luminosity at high redshift explores the distribution of @xmath9 or @xmath10 at a given @xmath3 ( if , as is likely , the distribution of agn luminosity is determined mainly by the black - hole mass rather than the galaxy properties ) .
confusing the two distributions may create a false signature of evolution .
* the selection bias is substantial for the current estimates of the cosmic scatter in the @xmath1 or @xmath0 relations , particularly for galaxies containing the most massive black holes .
neither the root - mean - square cosmic scatter nor the shape of the scatter function is known accurately and hence the bias can not be reliably corrected for , given our present and even foreseeable state of knowledge . * our rough estimates of selection bias as applied to a number of evolutionary studies show that the potential bias may be larger than many of the @xmath201 or @xmath202 values observed . with the notable exceptions of @xcite and @xcite
, it appears that most studies of the evolution of the @xmath3 relations have not considered the effects of selection bias . in most cases ,
accounting properly for selection bias is likely to reduce the the evolution in the @xmath3 relations that has been observed in recent surveys . *
the only way to avoid selection bias is to choose high- and low - redshift galaxy samples using precisely defined , objective criteria that are precisely the same for the two samples . to our knowledge , this has not yet been done .
the work was initiated during a lorentz center workshop attended by the authors .
we thank the staff of the lorentz center and dr .
tim de zeeuw for hosting us .
we thank drs .
karl gebhardt , richard green , and nadia zakamska for useful conversations .
barth , a. j. , greene , j. e. , & ho , l. c. 2005 , , 619 , l151 bernardi , m. , et al.2003 , , 125 , 1849 bernardi , m. , et al.2006 , , 131 , 2018 blanton , m. r. , et al .
2003 , , 592 , 819 boyle , b. j. , shanks , t. , croom , s. m. , smith , r. j. , miller , l. , loaring , n. , & heymans , c. 2000 , , 317 , 1014 dressler , a. 1989 , in iau symposium 134 , active galactic nuclei , ed .
d. osterbrock and j. s. miller ( dordrecht : kluwer ) , 217 ferrarese , l. , & merritt , d. 2000 , , 539 , l9 fine , s. , et al .
2006 , , 373 , 613 gebhardt , k. , et al . 2000a , , 539 , l13 gebhardt , k. , et al .
2000b , , 543 , l5 gebhardt , k. , et al .
2003 , , 583 , 92 greene , j. e. , & ho , l. c. 2006 , , 641 , l21 hring , n. & rix , h. 2004 , , 604 , l89 hopkins , p. f. , hernquist , l. , cox , t. j. , di matteo , t. , robertson , b. , & springel , v. 2005 , , 163 , 1 kormendy , j. 1993 , in the nearest active galaxies , ed .
j. beckman , l. colina and h. netzer ( madrid : consejo superior de investigaciones cientificas ) , 197 kormendy , j. , & richstone , d. 1995 , , 33 , 581 labita , m. , treves , a. , falomo , r. , & uslenghi , m. 2006 , , 373 , 551 lauer , t. r. , et al .
2007 , , 662 , 808 magorrian , j. , et al .
1998 , , 115 , 2285 malmquist , k. g. 1924 , medd .
lund astron .
obs . ser .
ii , 32 , 64 novak , g. s. , faber , s. m. , & dekel , a. 2006 , , 637 , 96 onken , c. a. , ferrarese , l. , merritt , d. , peterson , b. m. , pogge , r. w. , vestergaard , m. , & wandel , a. 2004 , , 615 , 645 peng , c. y. , impey , c. d. , rix , h .- w .
, kochanek , c. s. , keeton , c. r. , falco , e. e. , lehr , j. , & mcleod , b. a. 2006 , , 649 , 616 postman , m. , & lauer , t. r. 1995 , , 440 , 28 robertson , b. , hernquist , l. , cox , t. j. , di matteo , t. , hopkins , p. f. , martini , p. , & springel , v. 2006 , , 641 , 90 salviander , s. , shields , g. a. , gebhardt , k. , & bonning , e. w. 2006 , astro - ph/0612568 schechter , p. 1976
, , 203 , 297 sheth , r. k. , et al .
2003 , , 594 , 225 shields , g. a. , gebhardt , k. , salviander , s. , wills , b. j. , xie , b. , brotherton , m. s. , yuan , j. , & dietrich , m. 2003 , , 583 , 124 tremaine , s. et al .
2002 , , 574 , 740 treu , t. , malkan , m. a. , & blandford , r. d. 2004 , , 615 , l97 tundo , e. , bernardi , m. , hyde , j. b. , sheth , r. k. , & pizzella , a. 2006 , astro - ph/0609297 yu , q. , & lu , y. 2004 , , 602 , 603 yu , q. , & tremaine , s. 2002 , , 335 , 965 woo , j .- h . ,
treu , t. , malkan , m. a. , & blandford , r. d. 2006 , , 645 , 900 or velocity dispersion @xmath7 at selected values of @xmath3 ( @xmath203 is given in the upper right of each panel ) .
these probability distributions are obtained by taking horizontal cuts through the joint distributions shown in figure [ fig:2df ] , and are appropriate when the sample selection is on black - hole mass .
two distributions are shown with dashed and solid lines , corresponding respectively to @xmath204 for @xmath61 on the left or @xmath205 for @xmath7 on the right .
the horizontal coordinate is the difference between either @xmath61 or @xmath7 and the nominal values @xmath97 determined from the @xmath1 or @xmath0 relations ( [ eqn : msig ] ) and ( [ eqn : ml_hr]).,title="fig : " ] or velocity dispersion @xmath7 at selected values of @xmath3 ( @xmath203 is given in the upper right of each panel ) .
these probability distributions are obtained by taking horizontal cuts through the joint distributions shown in figure [ fig:2df ] , and are appropriate when the sample selection is on black - hole mass .
two distributions are shown with dashed and solid lines , corresponding respectively to @xmath204 for @xmath61 on the left or @xmath205 for @xmath7 on the right .
the horizontal coordinate is the difference between either @xmath61 or @xmath7 and the nominal values @xmath97 determined from the @xmath1 or @xmath0 relations ( [ eqn : msig ] ) and ( [ eqn : ml_hr]).,title="fig : " ] | programs to observe evolution in the @xmath0 or @xmath1 relations typically compare black - hole masses , @xmath2 in high - redshift galaxies selected by nuclear activity to @xmath3 in local galaxies selected by luminosity @xmath4 or stellar velocity dispersion @xmath5 because agn luminosity is likely to depend on @xmath3 , selection effects are different for high - redshift and local samples , potentially producing a false signal of evolution .
this bias arises because cosmic scatter in the @xmath0 and @xmath1 relations means that the mean @xmath6 or @xmath7 among galaxies that host a black hole of given @xmath3 , may be substantially different than the @xmath6 or @xmath7 obtained from inverting the @xmath1 or @xmath0 relations for the same nominal @xmath8 the bias is particularly strong at high @xmath2 where the luminosity and dispersion functions of galaxies are falling rapidly .
the most massive black holes occur more often as rare outliers in galaxies of modest mass than in the even rarer high - mass galaxies , which would otherwise be the sole location of such black holes in the absence of cosmic scatter . because of this bias
, @xmath3 will typically appear to be too large in the distant sample for a given @xmath9 or @xmath5 for the largest black holes and the largest plausible cosmic scatter , the bias can reach a factor of 3 in @xmath3 for the @xmath0 relation and a factor of 9 for the @xmath1 relation .
unfortunately , the actual cosmic scatter is not known well enough to correct for the bias .
measuring evolution of the @xmath3 and galaxy property relations requires object selection to be precisely defined and exactly the same at all redshifts . |
percolation @xcite is one of the most - fundamental problems in statistical mechanics .
many phase transitions in physical systems can be explained in terms of a percolation transition .
the pivotal property of percolation is connectivity .
one can describe this in terms of weighted graphs ( often also called networks ) with nonzero or zero weights , corresponding to occupied / connected or unoccupied / disconnected edges . here , we extend the problem to negative weights , in particular we study bond percolation . as an example
, one can imagine an agent traveling on a graph , who has , when traversing an edge , either to pay some resource ( positive weight ) or he is able ( once ) to harvest some resource ( negative weight ) .
paths including negative edge weights also appear in the context of domain walls in random - bond ising systems @xcite .
one percolation problem is whether there exists a path ( or loop ) spanning the full system with negative total weight , such that each edge is traversed at most once .
hence , the percolating objects are paths and it is sufficient to look for minimum - weighted ( or `` shortest '' ) paths ( mwps ) .
percolation properties of string - like objects have been studied occasionally @xcite , but to our knowledge never allowing for negative weights .
one realizes immediately that the negative weights lead to properties which are fundamentally different from conventional percolation .
e.g. , negative - weight percolation ( nwp ) lacks transitivity : if there is a valid ( negative - weight ) path @xmath0 from @xmath1 to @xmath2 via @xmath3 , it is possible that the path @xmath4 is not valid , as for the paths @xmath5/@xmath6 in figure [ fig : example ] .
hence , there is no simple definition of percolation clusters in this case .
note that , since strings are thus considered , several percolating strings or loops might coexist in the same sample .
this we observe indeed , see below . here
, we study nwp numerically .
first we introduce the model .
then we outline the algorithm to obtain mwps , since no standard shortest - path algorithm can be used .
the results for different types of two- ( @xmath7 ) and three - dimensional ( @xmath8 ) systems , in particular the critical exponents describing the percolation transitions , show that nwp is fundamentally different compared to conventional percolation and from previous models with string - like percolation .
thus , the study of percolation models exhibiting negative weights might lead to many new insights concerning the behavior of disordered systems . ,
6 with positive and negative edge weights ( numbers at lines ) .
numbers in black boxes denote the total weight of the minimum - weighted path from node 0 to the corresponding node .
[ fig : example ] ]
we consider @xmath7 and @xmath8 lattice graphs @xmath9 with side length @xmath10 , where adjacent sites @xmath11 are joined by undirected edges @xmath12 .
weights @xmath13 are associated with the edges , representing quenched random variables .
we consider either bimodal ( @xmath14j ) or `` gaussian - like '' ( gl ) distribution of the edge weights , where @xmath15 denotes the fraction of negative or gaussian - distributed edge weights , respectively , among edges with unit weight ( fraction @xmath16 ) .
the bimodal weights are taken to be @xmath17 and the gaussian weights have zero mean and unit width .
these weight distributions explicitly allow for loops @xmath18 with negative weight , given by @xmath19 . for any non - zero value of @xmath15
, a sufficiently large systems will exhibit at least small loops with negative total weight .
we investigate ( i ) mwps in the presence of negative weighted loops on square lattices with periodic boundary conditions ( bcs ) in one and free bcs in the other direction .
the path ends are allowed to terminate on the free boundaries .
further , we study ( ii ) minimum - weighted configurations of negative - weighted loops on square , @xmath7 hexagonal and cubic lattices with fully periodic bcs , see figure [ fig : samples ] . in either case we find critical values of @xmath15 above which ( i ) negative - weight paths appear that span the lattice across the direction with periodic bcs or ( ii ) percolating loops emerge , that span the lattice along some direction . ,
middle ( right ) : @xmath7 ( @xmath8 ) loop configuration for @xmath20 .
spanning loops are colored red . each sample is taken slightly above the respective critical point @xmath21 for @xmath14j disorder .
dashed lines denote free bcs , solid lines indicate periodic bcs .
[ fig : samples],title="fig : " ] , middle ( right ) : @xmath7 ( @xmath8 ) loop configuration for @xmath20 . spanning loops
are colored red .
each sample is taken slightly above the respective critical point @xmath21 for @xmath14j disorder .
dashed lines denote free bcs , solid lines indicate periodic bcs . [
fig : samples],title="fig : " ] , middle ( right ) : @xmath7 ( @xmath8 ) loop configuration for @xmath20 . spanning loops
are colored red .
each sample is taken slightly above the respective critical point @xmath21 for @xmath14j disorder .
dashed lines denote free bcs , solid lines indicate periodic bcs .
[ fig : samples],title="fig : " ] since we are looking for mwps on undirected graphs in the presence of negative weights , traditional `` shortest path '' algorithms can not be applied .
the reason is that for applying these traditional algorithms a special condition must hold : for the distance @xmath22 of any shortest path from a source @xmath23 to @xmath24 , @xmath25 holds , where @xmath26 denotes the set of neighbors of @xmath27 .
this equation is not fulfilled in our case , as can be seen from figure [ fig : example ] for node @xmath28 .
it has minimum distance @xmath29 to node @xmath23 , but it is connected to node 5 via an edge of weight @xmath30 and @xmath31 .
hence , a different approach has to be applied .
we determine mwps and loop configurations through an appropriate transformation of the original graph , detailed in @xcite , and obtaining a minimum - weighted perfect matching ( mwpm ) @xcite , by using exact combinatorial optimization algorithms . here
we give a concise description of the mapping , pictured as 3-step procedure illustrated in figure [ fig : alg ] : \(1 ) each edge , joining 2 sites on @xmath32 , is replaced by a path of 3 edges .
therefore , 2 `` additional '' sites have to be introduced for each edge of @xmath32 .
the original sites are then `` duplicated '' along with their incident edges .
for each of these pairs of of duplicated sites , one additional edge is added that connect the two sites of a pair .
the resulting auxiliary graph @xmath33 is depicted in figure [ fig : alg](b ) , where additional sites appear as squares and duplicated sites as circles .
figure [ fig : alg](b ) also illustrates the weight assignment on @xmath33 .
a more extensive description can be found in @xcite .
\(2 ) a mwpm on the auxiliary graph is determined @xcite .
a mwpm is a minimum - weighted subset @xmath34 of the edges contained in @xmath35 , such that each site of @xmath35 is met by one edge in @xmath34 .
this is illustrated in figure [ fig : alg](c ) , where the bold edges represent @xmath34 for the given weight assignment .
the dashed edges are unmatched .
due to construction , the auxiliary graph consists of an even number of sites and since there are no isolated sites , it is guaranteed that a perfect matching exists .
\(3 ) finally it is possible to find a relation between the matched edges @xmath34 on @xmath35 and a configuration of negative - weighted loops @xmath36 on @xmath32 by tracing the steps of the transformation ( 1 ) back .
note that each edge contained in @xmath34 that connects an additional site ( square ) to a duplicated site ( circle ) corresponds to an edge on @xmath32 that is part of a loop , see figure [ fig : alg](d ) .
more precise , there are always two such edges in @xmath34 that correspond to one edge on @xmath32 .
all the edges in @xmath34 that connect like sites ( i.e. duplicated - duplicated , or additional - additional ) carry zero weight and do not contribute to a loop on @xmath32 . afterwards a depht - first search can be used to explore @xmath36 and to determine the properties of the individual loops . for the weight assignment in figure [ fig : alg](a ) , there is only one loop with weight @xmath37 and length @xmath38 .
note that the result of the calculation is a collection @xmath36 of loops ( and one path for ( i ) , see below ) , such that the total loop weight is minimized .
hence , one obtains a global collective optimum of the system .
clearly , all loops that contribute to @xmath36 possess a negative weight .
note that @xmath36 can be empty and that sub paths are neither allowed to intersect nor to terminate at some site within the lattice .
so as to induce a mwp , as in problem ( i ) , special care is needed .
we have to allow the paths explicitly to terminate at a certain node .
therefore , two extra sites are introduced to the graph .
one extra site is connected to each of the free boundaries by adding edges with weight zero to the transformed graph that join the site with the sites of the respecting boundary .
any subsequent mwpm will contain a path of minimal weight , joining the extra sites .
this does not necessary coincide with the `` shortest '' path , as explained above , but yields the minimal - weighted path in the presence of negative weighted loops . in contrast to the loops , mwps can carry a positive weight in principle , but , as we will see , this will not be the case close to and beyond the percolation transition we are interested in . with edge weights ,
( b ) auxiliary graph @xmath35 with proper weight assignment : black edges carry the same weight as the respective edge in the original graph and gray edges carry zero weight .
( c ) minimum - weight perfect matching : bold edges are matched and dashed edges are unmatched .
( d ) loop configuration ( bold edges ) that corresponds to the mwpm .
[ fig : alg ] ] @llllllll type & geom . &
@xmath21 & @xmath39 & @xmath40 & @xmath41 & @xmath42 & @xmath43 + p@xmath14j & @xmath7 sq & 0.1032(5 ) & 1.43(6 ) & 1.03(3 ) & 0.76(5 ) & 2.52(8 ) & 1.268(1 ) + l@xmath14j & @xmath7 sq & 0.1028(3 ) & 1.49(9 ) & 1.09(8 ) & 0.75(8 ) & 2.58(6 ) & 1.260(2 ) + l@xmath14j & @xmath7 hex & 0.1583(6 ) & 1.47(9 ) & 1.07(9 ) & 0.76(8 ) & 2.59(2 ) & 1.264(3 ) + l - gl & @xmath7 sq & 0.340(1 ) & 1.49(7 ) & 1.07(6 ) & 0.77(7 ) & 2.59(3 ) & 1.266(2 ) + l@xmath14j & @xmath8 cu & 0.0286(1 ) & 1.02(3 ) & 1.80(8 ) & & 3.5(3 ) & 1.30(1 ) + @xmath14j paths : ( a ) rescaled probability @xmath44 that the path weight is negative for different system sizes @xmath10 , ( b ) illustrates the quality @xmath1 of the critical exponent @xmath45 and ( c ) shows the unscaled data near @xmath46 .
[ fig : path_nu ] ]
within this problem , the aforementioned collection @xmath36 consists of a set of loops , which might be empty , and a mwp spanning the lattice between the free boundaries .
since the mwp is anyway of @xmath47 in the vertical direction by construction , we call the path percolating , if its projection on the horizontal axis ( i.e. its _ roughness _
@xmath48 ) covers the full systems .
we studied @xmath7 lattices with @xmath14j disorder and sampled over up to @xmath49 realizations of the disorder to perform averages , subsequently denoted by @xmath50 .
firstly , we investigate the percolation probability . for a very small fraction @xmath15 of negative edge weights ,
the path will cross the lattice in a rather direct fashion , since overhangs are likely to increase the weight of the path and hence the weight of the whole configuration @xmath36 .
increasing @xmath15 also increases the spanning probability @xmath51 , i.e. the probability of the horizontal extension to be @xmath47 .
it is expected to scale as @xmath52 $ ] , where the critical exponent @xmath39 describes the divergence of the correlation length at the critical point @xmath21 and @xmath53 is a universal function .
a data collapse , obtained using the method described in @xcite , yields @xmath21 and @xmath39 as listed in table [ tab : crit_exp ] with a `` quality '' @xmath54 of the scaling assumption .
we want to point out that the value of @xmath39 found here is clearly distinct from the value @xmath55 of conventional bond percolation , that would give a much worse quality of @xmath56 ( see also figure [ fig : path_nu](b ) ) .
note that many negative weights are needed to allow the mwp to percolate . and
indeed , the path weight becomes negative above @xmath57 with a probability that approaches unity quickly .
therein , the finite - size behavior is described by an exponent @xmath58 consistent with the value @xmath39 stated above , see figure [ fig : path_nu ] . further note that @xmath59 even for large @xmath15 , where the spanning probability seems to saturate slightly above @xmath60 .
the reason is , as already mentioned , that we actually do not optimize the length of a single path but the weight of the whole configuration @xmath36 .
however , this behavior is clearly distinct from conventional percolation theory .
next , we consider the probability @xmath61 that a bond belongs to the path , where @xmath62 is the mean path length .
it exhibits the finite - size scaling behavior @xmath63 $ ] , where @xmath40 signifies the percolation strength @xcite . here , we fixed @xmath21 and @xmath39 as obtained above and determined the value of @xmath40 ( @xmath64 ) listed in table [ tab : crit_exp ] .
adjusting all three parameters in the above scaling assumption yields @xmath65 , @xmath66 and @xmath67 with a quality @xmath68 .
note that the values agree within the errorbars .
we also determined some quantities ( see also table [ tab : crit_exp ] ) just at @xmath69 with additional simulations up to @xmath70 ( @xmath71 samples ) .
we studied the associated finite - size susceptibility @xmath72 .
its finite size - scaling at @xmath21 can be described using the exponent @xmath41 via @xmath73 .
furthermore , the mean path length shows the critical behavior @xmath74 , where @xmath43 denotes the fractal exponent of the paths .
here , calculations at @xmath65 would result in a slightly smaller value of @xmath43 .
we further find the roughness exponent @xmath75 from @xmath76 to be compatible with unity .
note that the obtained exponents satisfy the scaling relations @xmath77 and @xmath78 .
we also measured the mean path weight @xmath79 at the percolation point @xmath21 and found that @xmath80 for @xmath81 seems to hold . we can further probe @xmath36 to not only investigate the mwp but to yield exponents that describe the small loops .
a detailed study of the scaling behavior ( length @xmath82 as function of the spanning length @xmath83 , not shown here ) at @xmath21 shows that they seem within error bars to exhibit _ the same _ fractal dimension @xmath84 as the mwps .
the loops also exhibit a mean ( negative ) weight that increases linearly with the loop length @xmath82 .
further , one expects the distribution @xmath85 of loop lengths @xmath82 to exhibit an algebraic scaling @xmath86 , where @xmath42 signifies the fisher exponent . here , finite - size effects and the presence of the path lead to a suppression of large loops and thus to a deviation from the expected scaling behavior already for rather small values of @xmath82 . in this question
, we find a most reliable value @xmath42 if we account for corrections to scaling via @xmath87 , see table [ tab : crit_exp ] and inset of figure [ fig : len_distrib ] . in principle , at criticality , the fisher exponent and fractal dimension are related through the scaling relation @xmath88 .
this would lead us to expect @xmath89 . finally , in @xmath7 we can associate a cluster with each loop .
the respective volume @xmath90 is measured as the number of enclosed plaquettes .
it scales as @xmath91 with @xmath92 , revealing the compact nature of the loops interior .
here , we studied @xmath7 ( hex , @xmath8 ) lattices with size up to @xmath93 ( @xmath94 , @xmath95 ) and @xmath96 ( @xmath97 , @xmath98 ) samples .
first , we analyze the largest loop for a given realization of the disorder .
we find the linear extensions of the loops by projecting it onto all perpendicular axes .
a loop is said to span the lattice , if its projection completely covers at least one axis . from the probability @xmath99 that a loop spans the lattice we estimate @xmath21 and @xmath39 listed in table [ tab : crit_exp ] .
the qualities of the scaling law were found to be @xmath100 , @xmath101 , @xmath102 and @xmath103 .
further , it is interesting to note that the mean number of spanning loops @xmath104 satisfies a similar scaling relation as the percolation probability , see figure [ fig : loops_2d](a ) , governed by the same values for @xmath21 and @xmath39 as @xmath99 . in either case
we find @xmath105 for large values of @xmath15 , as mentioned already in the introduction , see also @xcite .
as above , the probability @xmath106 that an edge belongs to the percolating loop can be used to determine the exponent @xmath40 , see figure [ fig : loops_2d](b ) and table [ tab : crit_exp ] . herein , the qualities of the scaling law were @xmath107 , @xmath108 , @xmath109 and @xmath110 .
interestingly , the values of @xmath21 , @xmath39 and @xmath40 for the @xmath7 loops with @xmath14j disorder are reasonable close to those of the @xmath7 random - bond ising model with @xmath111 , @xmath112 and @xmath113 , see @xcite and references therein .
this probably means that the @xmath7 ferromagnet to spin glass transition at @xmath114 can be explained in terms of a percolation transition in the following way : for a spin glass , one starts with a ferromagnetic configuration and searches for loops in the dual lattice with negative weight .
these loops correspond to clusters of spins , which can be flipped to decrease the energy . if these loops are small , i.e. not percolating , the ground state is ferromagnetic , otherwise the ground state exhibits spin - glass order .
a different argument relating this transition to a percolation transition was also briefly mentioned in a study , which focuses on the critical slowing down of polynomial - time algorithms at @xmath114 phase transitions @xcite .
right at @xmath21 we studied @xmath7 ( hex , @xmath8 ) systems up to @xmath115 ( @xmath116 , @xmath117 ) with @xmath118 ( @xmath119 , @xmath120 ) samples . for the @xmath7 systems we found @xmath41 listed in table [ tab : crit_exp ] with qualities @xmath121 , @xmath122 and @xmath123 . in @xmath8
, due to the very small percolation probability @xmath124 at @xmath21 our results are less clear .
hence , we can draw no final conclusions for that case .
regarding the fractal dimension , spanning loops and the non - spanning loops exhibit within error bars the same exponents @xmath125 in @xmath7 ( values of @xmath43 listed in table [ tab : crit_exp ] ) . as in the mwp case ,
the non - spanning loops are compact ( @xmath126 ) . on the other hand , in @xmath8 , we find @xmath127 different from @xmath128 . also regarding the mean weight @xmath129 for given length @xmath82 , we find in @xmath7 that spanning and non - spanning loops exhibit both @xmath130 ( i.e. @xmath131 for the spanning loops ) , while in @xmath8 the quality of the data is again not sufficient to observe a clear power law .
gl loops : the main plot shows the average number of dual adjustments per site for different system sizes @xmath10 and the inset illustrates the associated finite - size susceptibility .
[ fig : performance ] ] the distribution of loop - lengths , excluding the truly spanning loops , is in @xmath7 in good agreement with a power - law decay @xmath86 , see figure [ fig : len_distrib ] and table [ tab : crit_exp ] . in @xmath8
we find again strong finite size effects and hence a most reliable estimate of @xmath42 using a scaling form @xmath87 .
finally , we address the performance of the algorithm .
we doe not want to measure the running time in terms of cpu minutes , since this is machine dependent and is also influenced by external factors like which other processes are running .
hence , we have to look at the algorithm more closely . so as to determine a minimum - weight perfect matching , the algorithm attempts to find an optimal solution to an associated _ dual _ problem @xcite .
therein , the basic task of the algorithm is to find _ augmenting _ paths that improve the solution to the latter .
while executing , the solution to the dual problem is adjusted several times . as a measure for the algorithm performance we consider the number @xmath132 of such adjustment operations , until the algorithm terminates .
figure [ fig : performance ] shows the average number of these operations per lattice site @xmath133 for the case of a gaussian - like distribution of the edge weights .
the value of @xmath134 increases with increasing @xmath15 and curves for different system sizes deviate from each other as the critical point is approached from below . at @xmath21
it scales like @xmath135 .
moreover , the corresponding susceptibility @xmath136 , i.e.the fluctuations of the @xmath137 , diverges at the critical point where is scales as @xmath138 , as can be seen from the inset of figure [ fig : performance ] . for the case of a @xmath14j distribution of the edge weights
, we found similar but slightly different results ( not depicted ) : here , data sets that describe the scaling of @xmath134 for different system sizes , fall onto one universal curve everywhere .
this holds also for @xmath136 , which is peaked slightly below the critical point at @xmath139 .
we have introduced a percolation problem , where edge weights with possibly negative values are attached to the edges .
a system is called percolating if spanning loops or paths of negative weight exist . hence , this model might be fundamentally different from classical percolation problems .
we studied systems in @xmath7 and @xmath8 up to large system sizes . in all cases ,
the universality class is indeed clearly different from standard bond / site percolation .
further , the values of the fractal exponents obtained for the paths / loops differ from those reported in the context of shortest paths or hulls of percolation clusters , so far . nevertheless , the numerical values for the exponents found here are close to those of disorder - induced single defects ( @xmath140 ) and multiple defects ( @xmath141 ) in a @xmath7 elastic medium at zero temperature @xcite .
we observe universality in @xmath7 , hence the properties are independent of the type of weight distribution , lattice geometry and the same for loops or paths
. we also studied ( not shown here ) a related problem , where the loops are allowed to intersect .
the corresponding mapping was recently used in a different context to determine extended ground states of ising spin glasses @xcite .
we found again the same critical exponents . on the other hand
, we observe in @xmath7 the same behavior as for the @xmath114 ferromagnet to spin glass transition for the random - bond ising model , hence this physical transition can be probably explained by a percolation transition in the spirit of the model we have introduced here .
hence , studying percolation problems with negative weights and similar generalizations might be a key approach to describe many yet not well - understood phase transitions in terms of percolation transitions .
we also anticipate applications to other fields like social problems ( see introduction ) or other types of networks . | we describe a percolation problem on lattices ( graphs , networks ) , with edge weights drawn from disorder distributions that allow for weights ( or distances ) of either sign , i.e. including negative weights .
we are interested whether there are spanning paths or loops of total negative weight .
this kind of percolation problem is fundamentally different from conventional percolation problems , e.g. it does not exhibit transitivity , hence no simple definition of clusters , and several spanning paths / loops might coexist in the percolation regime at the same time .
furthermore , to study this percolation problem numerically , one has to perform a non - trivial transformation of the original graph and apply sophisticated matching algorithms . using this approach , we study the corresponding percolation transitions on large square , hexagonal and cubic lattices for two types of disorder distributions and determine the critical exponents .
the results show that negative - weight percolation is in a different universality class compared to conventional bond / site percolation . on the other hand
, negative - weight percolation seems to be related to the ferromagnet / spin - glass transition of random - bond ising systems , at least in two dimensions . |
a fascinating fact of cuprate research is that despite three decades of extensive research , experiments still uncover new and unexpected results which surprise our hitherto reached consensus .
the latest addition to this investigation comes from various spectroscopic findings of a fully gapped or ` insulating'-like electronic structure in deeply underdoped region,@xcite where nodal @xmath1-wave pairing symmetry is expected .
this feature is ubiquitously exhibited in many hole - doped cuprates studied so far , including la - based,@xcite bi - based,@xcite cl - based,@xcite and also yb - based compounds.@xcite this unexpected result led to essential questions such as : is the fully gapped state related to superconductivity other than @xmath1-wave , or is it derived from the exotic normal state , or does it arise from the interplay of superconductivity with any normal state phase , or others ?
nodeless @xmath1-wave pairing state is demonstrated earlier to exist in electron - doped cuprates,@xcite and also in 122 iron - chalcognenides.@xcite however , in these cases either interaction or band structure topology removes the nodal states in the non - sc state itself , which thus yields a fully gapped electronic structure despite @xmath1-wave being the underlying pairing symmetry . but this description does not apply to the nodeless state in hole - doped systems since as temperature is raised above the sc state , a truncated fermi surface ( fs ) is experimentally verified to exist in the nodal region.@xcite motivated by this unexpected behavior , several possible explanations have been put forward in recent times which include coulomb disorder effect,@xcite polaron effect,@xcite @xmath4 pairing,@xcite topological superconductivity,@xcite , dynamical mean - field theory calculations@xcite .
however , no consensus is yet reached . -0.6
cm our theory stems from the fact that the antiferromagnetic ( afm ) phase of the parent compound persists up to the underdoped region and couples to superconductivity.@xcite the evidence of the coexistence of afm and sc in the underdoped region is well documented in numerous nmr,@xcite neutron scattering@xcite data in multilayered bi - based cuprates@xcite upto doping @xmath58 - 12% , tl- and hg - based cuprates@xcite upto doping @xmath510% , and yb - based cuprates@xcite upto doping @xmath58 - 10% .
interestingly , the fully gapped state observed in these materials lies indeed within this doping range .
based on these data , we draw a generalized cuprate phase diagram as shown in fig .
[ fig1](a ) .
the interplay between the two competing interactions generically yields a dynamically generated new sc order parameter with finite center - of - mass .
earlier calculations predicted that for the competition of afm and @xmath1-wave sc , the third sc order parameter is a nodal @xmath1-wave sc if no magnetic field is applied,@xcite or a time - reversal symmetry breaking @xmath6-triplet sc if a magnetic field is applied in the heavy - fermion systems@xcite .
here we find that constrained by the symmetry consideration of the @xmath1-wave sc pairing symmetry and a single valued afm ordering vector @xmath2 in the cuprates , the third order parameter is an odd - parity superconductor which does not break time - reversal symmetry .
most interestingly , the dynamically generated triplet superconductivity is also nearest neighbor , as in the @xmath1-wave case , but it gives a fully gapped electronic structure which removes the nodes of the @xmath1-wave pairing .
we develop a generic mean - field hamiltonian based on the short range coulomb repulsion and attractive potential to self - consistently evaluate all three coupled order parameters at a given doping . for experimental verification of our theory , we compute the temperature dependence of the superfluid density ( inversely proportional to the square of the magnetic penetration depth ) in the uniform coexistence state of all three interactions .
we show that the exponential temperature dependence of the superfluid density will be a definitive test of the triplet sc origin of the apparent ` insulating ' state in lightly doped cuprates .
furthermore , the amplitude modulation of the second sc gap can be directly visualized by scanning tunneling microscopy or spectroscopy ( stm / sts ) .
in the afm state , the fs is reconstructed owing to the reduced brillouin zone ( bz ) defined by the modulation vector @xmath2 and the residual fs consists of small hole - pockets centering at the nodal points .
the corresponding fs topology is shown in fig .
[ fig1](b ) for a realistic value of the afm gap at doping @xmath70.08 ( see below ) .
it is interesting to notice that when superconductivity appears in the new magnetic bz , two types of cooper pairs form .
the typical @xmath1-wave cooper pairs form between two electrons with opposite momenta and spin lying on different quadrant of the bz as @xmath8 , where @xmath9 is the fermionic creation operator at the crystal momentum @xmath10 with up spin ( say ) .
additionally , two afm quasiparticles @xmath9 and @xmath11 lying inside and outside the magnetic bz boundary , respectively , now become decoupled , and naturally form a second cooper pairs with finite but single - valued center - of - mass momentum @xmath12 as @xmath13 .
such cooper pairs give fulde - ferrell - larkin - ovchinnikov ( fflo ) type unconventional superconductivity originally proposed for pauli limiting systems.@xcite we lay out below a ginzburg - landau criterion to deduce the symmetry of the fflo state in cuprates .
we now write down a mean - field based hamiltonian for the single band case , relevant for most cuprates,@xcite in which all aforementioned order parameters are treated on equal footing .
the mean - field hamiltonian is derived in the apendix a from the full interacting hamiltonian , and here we start with an effective hamiltonian expressed in the nambu - gorkov basis in the magnetic bz @xmath14 : @xmath15 here @xmath16 is the non - interacting dispersion of the effective cuo@xmath17 band expressed by tight - binding formalism in the appendix a and the parameters are fitted to the la@xmath18sr@xmath19cuo@xmath20 ( lsco ) dispersion.@xcite @xmath21 is the onsite coulomb repulsion . the other three symbols @xmath22 , and @xmath23 are the three mean - field order parameters for the afm , @xmath1-wave and @xmath24-wave fflo sc phase , respectively , as defined by @xmath25 , @xmath26 , and @xmath27 .
here @xmath28 is the number of lattice sites , and @xmath29 are the attractive interactions originate from the momentum dependence of the pairing interaction @xmath30 , and @xmath31 .
@xmath32 is the structure factor for @xmath0-wave pairing symmetry , while that for the @xmath33-pairing is @xmath34 .
it is interesting to notice that both pairing symmetries transform in the same way under the particular choice of afm wavevector @xmath12 as @xmath35 and @xmath36 .
the last identity also makes the triplet pairing respects time - reversal symmetry under this particular magnetic order . to see that we write down the ginzburg - landau type free energy as : @xmath37 + ... , \end{aligned}\ ] ] where @xmath38 is a constant and ` ... ' represents all the other quadratic , quartic and dipole coupling terms .
it is interesting to notice that the staggered in - plane magnetic moment @xmath39 only allows the creation of a vector form of the fflo pairing @xmath40 for a singlet sc order parameter @xmath41 .
furthermore , as mentioned before , since the time - reversed or the complex conjugate of @xmath40 lies outside the magnetic zone boundary with an opposite sign , the above coupling remains invariant under time - reversal symmetry .
finally , since superconductivity breaks guage symmetry , while magnetism does not , the third order parameter must also break the guage symmetry , i.e. , the third order parameter must be a sc phase which allows the hamiltonian to be guage invariant . in this way
, the commutator of any two order parameters generates the third one . returning back to our mean - field hamiltonian in eq .
[ ham ] , we diagonalize it in three steps by using bogolyubov transformation method .
first we diagonalize the triplet sc gap with the change of basis as @xmath42 , @xmath43 , where @xmath44 is the bogolyubov operator induced by @xmath45-term , and the corresponding coherence factors are @xmath46 , and @xmath47 .
the corresponding dispersion energies are @xmath48 , with @xmath49 . in this new basis ,
the effective complex afm gap and @xmath1-wave sc gap can be expressed by @xmath50 , and @xmath51 . with the effective gaps ,
the above hamiltonian reduces to a typical afm+@xmath1-sc phase which can now be diagonalized in the same fashion as done previously in refs . , also see appendix b. the corresponding effective afm and sc coherence factors are @xmath52 , @xmath53 , @xmath54 , and @xmath55 .
the quasiparticle energies are @xmath56 , where @xmath57 .
we evaluate all three order parameters and the chemical potential self - consistently for given values of constant interactions @xmath21 ,
@xmath58 and @xmath59 at a given doping by solving the following coupled equations : @xmath60,\\ \delta^1_{0}&=&v^1_{0}\sum^{\prime}_{{\bf k}}\sum_{\nu=\pm}s_{{\bf k}}\big[{\rm re}[f^2_{{\bf k}}-g^2_{{\bf k}}]u_{{\bf k}}^{\nu}v_{{\bf k}}^{\nu}\nonumber\\ & & + 2{\rm re}[f_{{\bf k}}g_{{\bf k}}]\alpha_{{\bf k}}\beta_{{\bf k}}((u^{\nu}_{{\bf k}})^2-(v^{\nu}_{{\bf k}})^2)\big]\tanh{(\beta e_{{\bf k}}^{{\nu}}/2)},\\ \delta^2_{0}&=&v^2_{0}\sum^{\prime}_{{\bf k}}p_{{\bf k}}\sum_{\nu=\pm}\big[\big[{\rm re}[f_{{\bf k}}g_{{\bf k}}]((u^{\nu}_{{\bf k}})^2-(v^{\nu}_{{\bf k}})^2)\nonumber\\ & & + 2\alpha_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^{\nu}v_{{\bf k}}^{\nu}(|f_{{\bf k}}|^2-|g_{{\bf k}}|^2)\big]\tanh{(\beta e_{{\bf k}}^{\nu}/2)})\big ] .
\ ] ] here the prime over a summation indicates that the summation is restricted within the magnetic bz .
@xmath61 is the fermi - dirac function . for @xmath62
( @xmath63 is the nearest neighbor tight - binding hopping parameter ) , @xmath64=-87 , -60 mev , we get @xmath22=0.1 , @xmath65=-20,-10 mev , respectively at doping @xmath66=0.08 in lsco which are realistic gap values for this material.@xcite the resulting normal state fs is shown in fig . [ fig1](b ) .
however , to facilitate the visualization of the nature of gap openings in the electronic structure , we use artificially large value of @xmath41=100 mev and @xmath40=50 mev in fig .
[ fig2](a ) . in the presence of three competing interactions ,
the cuo@xmath17 antibonding band splits into four quasiparticle states with finite gap at everywhere in the bz .
the bands are colored with corresponding filling factor which adds to 1 at each momentum . at the nodal point ,
the band gap is purely determined by the fflo pairing , since a afm+@xmath1-wave pairing gives nodal states there .
and the gap at the antinodal region is a mixture of all three gaps determined by @xmath67 . to experimentally verify the sc origin of the fully gapped state
, we calculate the penetration depth ( @xmath68 ) or superfluid density via linear response theory by accounting for both the diamagnetic current ( proportional to number of sc quasiparticles ) and residual paramagnetic current ( carried by normal state electrons ) due to external magnetic field .
the detailed derivation is given in the appendix c , and we here write down the final expression : @xmath69.\nonumber\\\end{aligned}\ ] ] here @xmath70 , @xmath71 are bare electronic charge and speed of light respectively , and @xmath72 is the unit cell volume .
@xmath73 , @xmath74 , and @xmath75 are different band masses , and velocity defined in appendix c. due to @xmath76 symmetry in the system , the cross - terms are zero and the penetration depth matrix is diagonal .
therefore , we plot @xmath77 component only in fig .
[ fig4 ] and normalized to its zero-@xmath78 value .
a spectroscopic method to differentiate between the sc and magnetic origins of the nodal gap is to observe whether the gapped quasiparicle state has a band - folding or band - bending at the magnetic zone boundary or at the normal state fermi momentum . to directly compare our result with the angle - resolved photoemission spectroscopy ( arpes ) data , we compute the single - particle spectral weight along the nodal direction as shown in fig . [ fig2](b ) , while fig .
[ fig2](c ) is the arpes data@xcite for the same sample . in both cases
, we see that the band folding ( indicated by vertical arrows ) occurs at a momentum which is far below the magnetic bz boundary at @xmath79 point ( red line ) .
the additional band appears in the theory is a shadow band induced by the sc order , and possess much weak intensity when the sc gap is reduced to the realistic value of @xmath8010 mev . in fig . [
fig3](a ) , we plot the extracted band gap ( not the individual order parameters ) as a function of fs angle ( @xmath81 ) measured with respect to the @xmath82 axis centering at m@xmath83 point . we see that in the non - superconducting afm semimetallic state ( @xmath84 , but @xmath22=finite ) , gapless fermi pocket exists in the region of @xmath85 .
the corresponding hole - pocket fs is also seen in the arpes data for the same sample at much above the sc transition temperature @xmath86 [ reproduced from ref . in fig .
[ fig3](b ) ] . in case of a pure @xmath1-wave sc without any triplet component (
@xmath87 and the other two order parameters are finite ) , the electronic state is gapped throughout the bz expect at @xmath88 .
the non - monotonic nature of the band gap is reproduced in such case and has been observed in arpes measurements ( see , e.g. , ref . ) . however ,
when the triplet component is turned on , a finite gap opens everywhere as described by @xmath67 .
it is interesting to notice that the odd - parity triplet sc of present form is fully gapped throughout the bz [ see _ inset _ to fig .
[ fig3](a ) ] , regardless of the topology of the underlying fs and is argued to be topologically non - trivial if time - reversal symmetry is retained,@xcite which is also applicable in the present case .
the obtained gap structure agrees remarkably well with that observed in the arpes measurement in the same sample as shown in fig .
[ fig3](b ) at low temperatures .
finally , we compute the penetration depth for different cases of the competition of the phases under study to verify their existences in the underlying spectrum by using eq .
[ lambda1 ] .
it is well known that for a fully gapped superconductor , the inverse squared penetration depth @xmath89 ( proportional to the superfluid density ) shows exponential @xmath78-dependence in the low-@xmath78 region , where nodal @xmath1-wave sc exhibits a linear - in-@xmath78 behavior , which sometimes modifies to the @xmath90-dependence due to disorder . to get a quantitative understanding ,
we calculate the temperature dependence of all gap values self - consistently at fixed values of the interaction parameters and at a fixed doping .
expectedly , we see that for a pure nodal @xmath1-wave ( @xmath91 ) , the linear - in-@xmath78 dependence of @xmath92 is evident even in the presence of afm ground state . with slowly turning on the fully gapped triplet pairing
, the exponential dependence of @xmath92 appears whose @xmath78-range of flat or exponential region depends on the relative strength of @xmath93 . for @xmath94 , the computed penetration depth shows a fully exponential behavior at all temperatures , which is a hallmark feature of @xmath95-wave like sc .
the transition from a pure @xmath1-wave like @xmath96 behavior to nodeless @xmath1-wave behavior is reported in theory and experimental data of electron doped cuprates.@xcite since exponential behavior only appears in the very low - temperature ( 10 - 15
in conclusion , the present paper provides a generalized framework for the phase stability of the afm and @xmath1-wave superconductivity in hole - doped cuprates .
we show that the uniform coexistence of these two phases is always accompanied by a third and robust triplet order parameter .
the symmetry of the @xmath2 value afm state and @xmath1-wave pairing restricts the symmetry of the triplet order parameter in cuprates to be odd - parity and fflo type , which commences a fully gapped quasiparticle spectrum .
the odd - parity order parameter breaks spin - rotational symmetry , but via coupling to the afm state , it respects time - reversal symmetry .
interestingly , such a fully gapped odd parity pairing symmetry in a time - reversal invariant condition is called topological superconductor.@xcite in general , such fflo term can be expected to be negligibly small , however , the critical fluctuations of the parent phases can significantly enhance the strength of the triplet pairing interaction . while fflo type sc has been proposed in many pauli limited superconductors , such as nodal @xmath1-wave in cecoin@xmath97 at high magnetic field,@xcite and @xmath24-wave in the rashba - type spin - orbit coupling systems without any realization to date , underdoped cuprates provide a clean and zero field platform to discover this exotic phase , if exists .
the author acknowledges valuable discussions with x .- j .
zhou , c. batista , and b. altunkaynak .
i also thank c. panagopolous for suggesting several references for the the work is supported by the u.s .
doe through the office of science ( bes ) and the ldrd program and faciliated by nersc computing allocation .
from the study of the group algebra@xcite , it is obvious that the coexistence of the any two non - commutating order parameter produce a third order parameter . in case of afm and @xmath1-superconductivity
, there should be a third , dynamically generated , order parameter.@xcite the strength of this third order parameter is subjects to the corresponding coupling constant . in the present work ,
we develop the theory for how the coexistence of afm and @xmath1-wave superconductivity can generate a triplet and non - zero center of mass superconducting order parameter .
our starting hamiltonian is @xmath98 here @xmath99 is the bare dispersion , @xmath21 is the onsite coulomb interaction , @xmath100 are the pairing strengths for @xmath1-wave and @xmath24-wave superconductivity , and @xmath28 is the number of sites .
we model the bare dispersion in the tight - binding form as @xmath101 , where @xmath102=(250,-25 , 12 , 35 , -155 ) in ev which is obtained by fitting to the dispersion obtained by arpes in lsco.@xcite @xmath103 is the fermion annihilation operator at crystal momentum @xmath10 with spin @xmath104 .
@xmath2 is the afm nesting vector in 2d .
we assume here a commensurate afm so that @xmath105 .
the staggered spin magnetization is defined as @xmath25 . for three order parameters ,
the crystal symmetry of them should be such that the commutator of any two of them should give the third one@xcite .
for this reason , if @xmath106 is assumed to be singlet @xmath1-wave symmetry , the afm state greentrees that @xmath107 should be triplet ( see main text ) .
so we get , the singlet interaction @xmath108 and @xmath109 , where @xmath32 and @xmath110 .
we assume that @xmath100 are attractive .
then the sc order parameters are defined as , for singlet state , @xmath111 . on the other hand ,
the triplet order is decoupled in the main band and in the magnetic band as @xmath112 , while @xmath113 .
this is the reason why the time - reversal symmetry remains invariant in this triplet sc state . after substituting these mean - field orders ,
the total hamiltonian reads , @xmath114 the prime over the summation indicate that the sum is defined in the magnetic bz . defining a nambu operator @xmath115 , one can define a @xmath116 hamiltonian matrix @xmath117 as follows @xmath118 @xmath119 we diagonalize the above hamiltonian in three steps by bogolyubov transformation method .
we first diagonalize the triplet part , by the following unitary matrix , @xmath120 where @xmath121 and @xmath122 are defined in the main text .
sc gap become , the hamiltonian then transforms to @xmath123 @xmath124 all quantities here are defined in the main text .
the above hamiltonian looks same as in a afm-@xmath1-sc state with effective afm- and sc - gaps which is diagonalized by the following unitary matrix@xcite @xmath125 the afm cnherences factors @xmath38 and @xmath126 and the sc coherence factors @xmath127 and @xmath128 are also defined in the main text .
therefore , the total unitary matrix that diagonalizes the full hamiltonian is @xmath129 .
now we derive the expression for the self - consistent order parameters in the eigenbasis .
@xmath130\nonumber\\ & = & v^1_{0}\sum^{\prime}_{{\bf k}}s_{{\bf k } } \big[\big((f_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^+-g_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^+)(f_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^++g_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^+ ) \big\langle b^{\dag}_{{\bf k},\uparrow}b_{{\bf k},\uparrow}\big\rangle\nonumber\\ & & ~~+(f_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^-+g_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^-)(f_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^--g_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^- ) \big\langle b^{\dag}_{{\bf k}+{{\bf q}},\uparrow}b_{{\bf k}+{{\bf q}},\uparrow}\big\rangle\nonumber\\ & & ~~+(-f_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^+-g_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^+)(f_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^+-g_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^+ ) \big\langle b^{\dag}_{-{\bf k},\uparrow}b_{-{\bf k},\uparrow}\big\rangle\nonumber\\ & & ~~+(f_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^--g_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^-)(-f_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^--g_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^- ) \big\langle b^{\dag}_{-{\bf k}+{{\bf q}},\uparrow}b_{-{\bf k}+{{\bf q}},\uparrow}\big\rangle\big)\nonumber\\ & & -(-f^*_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^++g^*_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^+)(f^*_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^++g^*_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^+ ) \big\langle b^{\dag}_{{\bf k},\uparrow}b_{{\bf k},\uparrow}\big\rangle\nonumber\\ & & ~~-(f^*_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^-+g^*_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^-)(-f^*_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^-+g^*_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^- )
\big\langle b^{\dag}_{{\bf k}+{{\bf q}},\uparrow}b_{{\bf k}+{{\bf q}},\uparrow}\big\rangle\nonumber\\ & & ~~-(f^*_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^++g^*_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^+)(f^*_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^+-g^*_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^+ ) \big\langle
b^{\dag}_{-{\bf k},\uparrow}b_{-{\bf k},\uparrow}\big\rangle\nonumber\\ & & ~~-(f^*_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^--g^*_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^-)(f^*_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^-+g^*_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^-)\big\langle b^{\dag}_{-{\bf k}+{{\bf q}},\uparrow}b_{-{\bf k}+{{\bf q}},\uparrow}\big\rangle\big]\nonumber\\ & = & v^1_{0}\sum^{\prime}_{{\bf k}}\sum_{\nu=\pm}s_{{\bf k}}\big[{\rm re}[f^2_{{\bf k}}-g^2_{{\bf k}}]u_{{\bf k}}^{\nu}v_{{\bf k}}^{\nu}+2{\rm re}[f_{{\bf k}}g_{{\bf k}}]\alpha_{{\bf k}}\beta_{{\bf k}}((u^{\nu}_{{\bf k}})^2-(v^{\nu}_{{\bf k}})^2)\big]\tanh{(\beta e_{{\bf k}}^{{\nu}}/2 ) } \ ] ] in the first line and right hand side of the above equation we have taken into account of the fact that the @xmath1-wave structure factor @xmath131 for this particular commensurate wavevector . since we have identified @xmath132 as the excitation energies of the fermion quasi - particles , the probability of its excitation in thermal equilibrium is the usual fermi function , @xmath133 with @xmath134 and so on .
therefore @xmath135 the triplet sc order is @xmath136.\nonumber\\\end{aligned}\ ] ] as mentioned before , the difference between the triplet and singlet pairings in the magnetic brilloun zone is that in the former case , the pairing symmetry picks up the complex conjugate as we go from the main band to the afm subbands shifted by @xmath137 vector , and also changes sign .
writing in bogolyubov quasiparticle form as before , we get @xmath138\nonumber\\ & = & v^2_{0}\sum^{\prime}_{{\bf k}}p_{{\bf k}}\sum_{\nu}\big[\big[{\rm re}[f_{{\bf k}}g_{{\bf k}}]((u^{\nu}_{{\bf k}})^2-(v^{\nu}_{{\bf k}})^2)+2\alpha_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^{\nu}v_{{\bf k}}^{\nu}(|f_{{\bf k}}|^2-|g_{{\bf k}}|^2)\big]\tanh{(\beta e_{{\bf k}}^{\nu}/2)})\big ] . \ ] ]
the quantum mechanical electric current can be written as @xmath140 now in a magnetic field the momentum operator becomes @xmath141 , so that the current is @xmath142^{\dag}\psi)\nonumber\\ & = & -\frac{e}{2m}(\psi^{\dag}{\bf p}\psi-({\bf p}\psi)^{\dag}\psi)-\frac{e^2{\bf a}}{mc}\psi^{\dag}\psi\end{aligned}\ ] ] the first term represents the current due to the normal electrons ( paramagnetic current ) and the second term is identified as the diamagnetic current which is contributed by the superconducting electrons @xmath143 .
the paramagnetic current ( @xmath144 ) has the tendency to cancel the diamagnetic current ( @xmath145 ) .
therefore the velocity of the superconducting electrons are identified as @xmath146 where @xmath147 , the band mass , is a @xmath148 tensor in two dimensional space defined below . which gives @xmath150 where @xmath151 is called the response function .
now in an anisotropic system the response function becomes a @xmath148 tensor as defined below @xmath152 where at @xmath153 , the penetration depth is related to @xmath154s as @xmath155 the first term @xmath157 arises due to normal electrons , which is often called the paramagnetic current " term because it tends to cancel the diamagnetic current @xmath158 , which aries because of superconducting electrons . in the presence of an electromagnetic field ( the field includes the effects of screening supercurrents ) , the canonical momentum is modified and the kinetic energy is @xmath159 , where @xmath160 is the crystal momentum or the band momentum . thus the resulting perturbation hamiltonian term is @xmath161 the band momentum is defined as @xmath162 , where @xmath163 is the band velocity for the band @xmath164 .
note that the symbol @xmath165 without a superscript of @xmath166 is the band velocity , while with the superscript it gives the afm conherence factors .
@xmath167\nonumber\\ & = & -\frac{e}{c\omega}{\bf a}(0)\sum_{{\bf k}}^{\prime}\big [ f_{{\bf k}}^2\big({\bf v}_{{\bf k}}\alpha_{{\bf k}}^2+{\bf v}_{{\bf k}+{\bf q}}\beta_{{\bf k}}^2\big ) + g_{{\bf k}}^2\big({\bf v}_{{\bf k}}\beta_{{\bf k}}^2+{\bf v}_{{\bf k}+{\bf q}}\alpha_{{\bf k}}^2\big)\big ] \big(\big\langle b_{{\bf k},\uparrow}^{\dag}b_{{\bf k},\uparrow}\big\rangle-\big\langle b_{-{\bf k},\downarrow}^{\dag}b_{-{\bf k},\downarrow}\big\rangle\big)\nonumber\\ & & ~~~+\big [ f_{{\bf k}}^2\big({\bf v}_{{\bf k}}\beta_{{\bf k}}^2+{\bf v}_{{\bf k}+{\bf q}}\alpha_{{\bf k}}^2\big ) + g_{{\bf k}}^2\big({\bf v}_{{\bf k}}\alpha_{{\bf k}}^2+{\bf v}_{{\bf k}+{\bf q}}\beta_{{\bf k}}^2\big)\big ] \big(\big\langle b_{{\bf k}+{\bf q},\uparrow}^{\dag}b_{{\bf k}+{\bf q},\uparrow}\big\rangle-\big\langle b_{-{\bf k}-{\bf q},\downarrow}^{\dag}b_{-{\bf k}-{\bf q},\downarrow}\big\rangle\big)\nonumber\\ & = & -\frac{e}{c\omega}{\bf a}(0)\sum_{{\bf k}}^{\prime}{\bf v}_{{\bf k}}^+ \big(\big\langle b_{{\bf k},\uparrow}^{\dag}b_{{\bf k},\uparrow}\big\rangle-\big\langle b_{-{\bf k},\downarrow}^{\dag}b_{-{\bf k},\downarrow}\big\rangle\big ) + { \bf v}_{{\bf k}}^- \big(\big\langle b_{{\bf k}+{\bf q},\uparrow}^{\dag}b_{{\bf k}+{\bf q},\uparrow}\big\rangle-\big\langle b_{-{\bf k}-{\bf q},\downarrow}^{\dag}b_{-{\bf k}-{\bf q},\downarrow}\big\rangle\big).\end{aligned}\ ] ] here @xmath72 is the unit cell volume . @xmath168 and @xmath169 , and @xmath170
. now the probability of excitation of the fermion quasiparticles in an electromagnetic field in the thermal equilibrium is the usual fermi function in the corresponding energy level sifted by the electromagnetic field . in the limit of small @xmath171 , we can expand the fermi function in taylor s series and keeping only the first term we get , @xmath172 @xmath179\nonumber\\ & = & -\frac{e^2}{c\omega}\sum_{j=1}^2a_j(0)\sum_{{\bf k}}^{\prime}\left[\left(\frac{1}{m^*_{i , j}}\right)\big\ { ( f_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^+-g_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^+)^2 -(f_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^++g_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^+)^2\big\}\right.\nonumber\\ & & ~~~~~~~~~~~~+\left.\left(\frac{1}{m^*_{{\bf k}+{\bf q}i , j}}\right)\big\ { ( f_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^+-g_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^+)^2 -(f_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^++g_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^+)^2\big\}\right ] \left(\big\langle b_{{\bf k},\uparrow}^{\dag}b_{{\bf k},\uparrow}\big\rangle + \big\langle b_{-{\bf k},\downarrow}^{\dag}b_{-{\bf k},\downarrow}\big\rangle\right ) \nonumber\\ & & ~~~~~~~~~+2\left(\frac{1}{m^*_{i , j}}\right)(f_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^++g_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^+)^2 + 2\left(\frac{1}{m^*_{{\bf k}+{\bf q}i , j}}\right){\bf v}_{{\bf k}+{\bf q}}(f_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^++g_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^+)^2\nonumber\\ & & ~~~~~~~~~~~~+\left[\left(\frac{1}{m^*_{i , j}}\right)\big\ { ( f_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^-+g_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^-)^2 -\left(\frac{1}{m^*_{i , j}}\right)(f_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^--g_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^-)^2\big\}\right.\nonumber\\ & & ~~~~~~~~~~~~\left.+\left(\frac{1}{m^*_{{\bf k}+{\bf q}i , j}}\right)\big\ { ( f_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^-+g_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^-)^2 -(f_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^--g_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^-)^2\big\}\right ] \left(\big\langle b_{{\bf k}+{\bf q},\uparrow}^{\dag}b_{{\bf
k},\uparrow}\big\rangle + \big\langle b_{-{\bf k}-{\bf q},\downarrow}^{\dag}b_{-{\bf k},\downarrow}\big\rangle\right)\nonumber\\ & & ~~~~~~~~~+2\left(\frac{1}{m^*_{i , j}}\right)(f_{{\bf k}}\beta_{{\bf k}}v_{{\bf k}}^--g_{{\bf k}}\alpha_{{\bf k}}u_{{\bf k}}^-)^2 + 2\left(\frac{1}{m^*_{{\bf k}+{\bf q}i , j}}\right)(f_{{\bf k}}\alpha_{{\bf k}}v_{{\bf k}}^--g_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^-)^2\nonumber\\ & = & -\frac{e^2}{c\omega}\sum_{j=1}^2a_j(0)\sum_{{\bf k}}^{\prime}\nonumber\\ & & ~~~f_{{\bf k}}^2\left(\left(\frac{1}{m^*_{i , j}}\right)\alpha_{{\bf k}}^2 + \left(\frac{1}{m^*_{{\bf k}+{\bf q}i , j}}\right)\beta_{{\bf k}}^2\right)\left[2(v_{{\bf k}}^+)^2+\big((u_{{\bf k}}^+)^2-(v_{{\bf k}}^+)^2\big ) \big(1-\big\langle
b_{{\bf k},\uparrow}^{\dag}b_{{\bf k},\uparrow}\big\rangle -\big\langle b_{-{\bf k},\downarrow}^{\dag}b_{-{\bf k},\downarrow}\big\rangle\big)\right]\nonumber\\ & & + g_{{\bf k}}^2\left(\left(\frac{1}{m^*_{i , j}}\right)\beta_{{\bf k}}^2 + \left(\frac{1}{m^*_{{\bf k}+{\bf q}i , j}}\right)\alpha_{{\bf k}}^2\right ) \left[2(u_{{\bf k}}^+)^2-\big((u_{{\bf k}}^+)^2-(v_{{\bf k}}^+)^2\big ) \big(1-\big\langle b_{{\bf k},\uparrow}^{\dag}b_{{\bf k},\uparrow}\big\rangle -\big\langle b_{-{\bf k},\downarrow}^{\dag}b_{-{\bf k},\downarrow}\big\rangle\big)\right]\nonumber\\ & & + 4f_{{\bf k}}g_{{\bf k}}\alpha_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^+v_{{\bf k}}^+\left(\frac{1}{m^*_{{\bf k}i , j}}+\frac{1}{m^*_{{\bf k}+{\bf q}i , j}}\right ) \big(1-\big\langle b_{{\bf k},\uparrow}^{\dag}b_{{\bf k},\uparrow}\big\rangle -\big\langle b_{-{\bf k},\downarrow}^{\dag}b_{-{\bf k},\downarrow}\big\rangle\big)\nonumber\\ & & + f_{{\bf k}}^2\left(\left(\frac{1}{m^*_{i , j}}\right)\beta_{{\bf k}}^2 + \left(\frac{1}{m^*_{{\bf k}+{\bf q}i , j}}\right)\alpha_{{\bf k}}^2\right ) \left[2(v_{{\bf k}}^+)^2+\big((u_{{\bf k}}^+)^2-(v_{{\bf k}}^+)^2\big)\right.\nonumber\\ & & ~~~\left.\times\big(1-\big\langle b_{{\bf k}+{\bf q},\uparrow}^{\dag}b_{{\bf k}+{\bf q},\uparrow}\big\rangle -\big\langle b_{-{\bf k}-{\bf q},\downarrow}^{\dag}b_{-{\bf k}-{\bf q},\downarrow}\big\rangle\big)\right]\nonumber\\ & & + g_{{\bf k}}^2\left(\left(\frac{1}{m^*_{i , j}}\right)\alpha_{{\bf k}}^2 + \left(\frac{1}{m^*_{{\bf k}+{\bf q}i , j}}\right)\beta_{{\bf k}}^2\right ) \left[2(u_{{\bf k}}^+)^2-\big((u_{{\bf k}}^+)^2-(v_{{\bf k}}^+)^2\big)\right .
\nonumber\\ & & ~~~\left.\times \big(1-\big\langle b_{{\bf k}+{\bf
q},\uparrow}^{\dag}b_{{\bf k}+{\bf q},\uparrow}\big\rangle -\big\langle b_{-{\bf k}-{\bf q},\downarrow}^{\dag}b_{-{\bf k}-{\bf q},\downarrow}\big\rangle\big)\right]\nonumber\\ & & + 4f_{{\bf k}}g_{{\bf k}}\alpha_{{\bf k}}\beta_{{\bf k}}u_{{\bf k}}^-v_{{\bf k}}^-\left(\frac{1}{m^*_{{\bf k}i , j}}+\frac{1}{m^*_{{\bf k}+{\bf q}i , j}}\right ) \big(1-\big\langle b_{{\bf k}+{\bf q},\uparrow}^{\dag}b_{{\bf k}+{\bf q},\uparrow}\big\rangle -\big\langle b_{-{\bf k}-{\bf q},\downarrow}^{\dag}b_{-{\bf k}-{\bf q},\downarrow}\big\rangle\big)\nonumber\\ & = & -\frac{e^2}{c\omega}\sum_{j=1}^2a_j(0)\sum_{{\bf k}}^{\prime}\sum_{\nu=\pm } \left(\frac{1}{m^{\nu}_{{\bf k}i , j}}\right)\left(1-\frac{e^{\pi}_{{\bf k}}+e^{m}_{{\bf k}}}{e^{\nu}_{{\bf k}}}\tanh{(\beta e^{\nu}_{{\bf k}}/2)}\right ) + \left(\frac{1}{m^{\nu}_{{\bf k}i , j}}\right)\tanh{(\beta e^{\nu}_{{\bf k}}/2)}. \ ] ] a. ino , c. kim , m. nakamura , t. yoshida , t. mizokawa , z .- x .
shen , a. fujimori , t. kakeshita , h. eisaki , and s. uchida , electronic structure of la@xmath19sr@xmath18cuo@xmath20 in the vicinity of the superconductor - insulator transition .
b * 62 * , 4137 ( 2000 ) .
k. tanaka , w. s. lee , d. h. lu , a. fujimori , t. fujii , risdiana , i. terasaki , d. j. scalapino , t. p. devereaux , z. hussain , and z .- x .
shen , distinct fermi - momentum - dependent energy gaps in deeply underdoped bi2212 .
science * 314 * , 1910 ( 2006 ) .
e. razzoli , g. drachuck , a. keren , m. radovic , n. c. plumb , j. chang , y .- b .
huang , h. ding , j. mesot , and m. shi , evolution from a nodeless gap to @xmath180-wave in underdoped la@xmath19sr@xmath18cuo@xmath20 .
lett . * 110 * , 047004 ( 2013 ) .
y. peng , j. meng , d.mou , j. he , l. zhao , y.wu , g. liu , x. dong , s. he , j. zhang , x. wang , q. peng , z. wang , s. zhang , f. yang , c. chen , z. xu , t. k. lee , x. j. zhou , disappearance of nodal gap across the insulator superconductor transition in a copper - oxide superconductor .
* 4 * , 2459 ( 2013 ) .
d. gustafsson , d. golubev , m. fogelstrm , t. claeson , s. kubatkin , t. bauch , and f. lombardi , fully gapped superconductivity in a nanometre - size yba@xmath17cu@xmath181o@xmath182 island enhanced by a magnetic field .
nanotechnology * 8 * , 25 - 30 ( 2013 ) .
t. das , r. s. markiewicz , and a. bansil , nodeless @xmath1-wave superconducting pairing due to residual antiferromagnetism in underdoped pr@xmath183ce@xmath18cuo@xmath184 .
lett * 98 * , 197004 ( 2007 ) .
s. shimizu , s .-
tabata , s. iwai , h. mukuda , and y. kitaoka , high - temperature superconductivity and antiferromagnetism in multilayer cuprates : cu and f nmr on five - layer bacacuo(f , o ) phys . rev .
b * 85 * , 024528 ( 2012 ) .
s. shimizu , s .-
tabata , h. mukuda , y. kitaoka , p. m. shirage , h.i kito , and a. iyo , antiferromagnetism , superconductivity , and pseudogap in three - layered high - tc cuprates ba2ca2cu3o6(f , o)2 probed by cu - nmr , phys . rev .
b * 83 * , 214514 ( 2011 ) .
h. mukuda , s. shimizu , a. iyo , and y. kitaoka , high - tc superconductivity and antiferromagnetism in multilayered copper oxides a new paradigm of superconducting mechanism .
soc . jpn .
* 81 * , 011008 ( 2012 ) . s .-
baek , t. loew , v. h i n k o v , c. t. lin , b. keimer , b. bchner , and h .- j .
grafe , evidence of a critical hole concentration in underdoped ybacuo single crystals revealed by cu nmr , phys .
rev b * 86 * , 220504(r ) ( 2012 ) .
d. haug , v. hinkov , y. sidis , p. bourges , n. b. christensen , a. ivanov , t. keller , c. t. lin and b. keimer , new j. phys .
* 12 * , 105006 ( 2010 ) .
v. baldent , d. haug , y. sidis , v. hinkov , c. t. lin , and p. bourges , phys .
b * 83 * , 104504 ( 2011 ) g. c. psaltakis and e. w. fenton , j. phys .
c * 16 * , 3913 ( 1983 ) ; b. kyung , phys .
b * 62 * , 9083 ( 2000 ) .
p. fulde , and r. a. ferrell , superconductivity in a strong spin - exchange field .
phys . rev .
* 135 * , a550 ( 1964 ) ; a. i. larkin , and yu n. ovchinnikov , zh .
teor . fiz . *
47 * , 1136 ( 1964 ) ; a. i. larkin , and yu n. ovchinnikov , inhomogeneous state of superconductors .
jetp * 20 * , 762 ( 1965 ) .
while a single band model is valid for the low - energy physics of most of the cuorates , other bands such as materials specific cuo chain state@xcite or hgo like bands@xcite are required to be included in ybco and hgcco cuprates , respectvely .
r. s. markiewicz , s. sahrakorpi , m. lindroos , hsin lin , and a. bansil , one - band tight - binding model parametrization of the high - tc cuprates including the effect of @xmath187 dispersion .
b * 72 * , 054519 ( 2005 ) . | we develop a microscopic theory to show that the transition from an antiferromagnetic ( afm ) phase to @xmath0-wave superconductivity occurs through the robust companion of a triplet superconductivity .
more interestingly , the symmetry combination of afm and @xmath1-wave superconducting ( sc ) phase prescribes the triplet state to be odd parity and fulde - ferrell - larkin - ovchinnikov ( fflo ) type with its center - of - mass momentum being single valued at the afm wavevector @xmath2 .
furthermore , the fflo state produces a fully gapped quasiparticle spectrum which explains the recent observations of the ` insulating'-like electronic structure of lightly doped cuprates in hole doping side .
we have studied the stability of all three phases within the coupled self - consistent mean - field theory .
finally , we calculate the superfluid density to propose its exponential temperature dependence as a test to the sc origin of this ` insulating ' state , while the phase modulation of the fflo state can be visualized by scanning probes .
the results are generic to all uncoventional sc pairing symmetries including @xmath3 in pnictide and others . |
in a recent paper chevalier ( 1999 ) discusses physical conditions in supernova remnants ( snrs ) evolving inside molecular clouds .
he considers a clumpy gas distribution in the clouds , where the dominant part of the cloud mass is contained in the compact dense clumps filling only 10% or less of the cloud volume and the rest of the cloud is filled with tenuous gas with a number density @xmath0 @xmath1 .
thus a snr exploding inside a cloud evolves mostly in such low density medium .
chevalier discusses a number of consequences of such a model and compares it to observations of three snrs : w44 , ic 433 and 3c391 . in his discussion
the observed , very flat , cosmic ray electron distributions responsible for the radio synchrotron spectra are interpreted as shock - compressed ambient distributions radiating downstream of the radiative shock .
however , the existence of a very flat ambient electron distribution in a molecular cloud is a matter for debate .
the uniform magnetic field structures observed in several clouds and efficient damping of short alfvn waves in a partly neutral medium ( hartquist & morfill 1984 ) suggests a possibility of an efficient cosmic ray exchange between clouds and a steep - spectrum galactic cosmic ray population . in the present note
we point out the possibility of an acceleration of the flat spectrum electrons at a snr shock wave , if the second - order fermi acceleration in the vicinity of the shock is taken into account . in the next section
we summarize results of an approximate analytic theory for the particle spectral indices ( ostrowski & schlickeiser 1993 , @xmath2 ` os93 ' ) , with a few typing errors of the original paper corrected .
then , in section 3 , we demonstrate that the physical conditions considered by chevalier ( 199 ) for shock waves with the alfvn velocity , @xmath3 , non - negligible in comparison to the shock velocity allow for generation of the required flat particle distributions ( c.f .
, also , drury 1983 , hartquist & morfill 1983 , drge et al .
1987 ; schlickeiser & frst 1989 ) . in the final section ( section 4 ) we briefly discuss an application of the present model to actual conditions in astrophysical objects .
os93 derived a simplified kinetic equation for the particle distribution formed at the parallel shock wave due to action of the first - order acceleration at the shock and the second - order acceleration in the shock turbulent vicinity . with the phase - space distribution function at the shock , @xmath4
, an ` integral ' distribution is defined as @xmath5},\eqno(2.1)\ ] ] where @xmath6 and @xmath7 are the plasma flow velocity in the shock frame and the spatial diffusion coefficient of cosmic ray particles , respectively ( @xmath8 upstream and @xmath9 downstream of the shock ) .
analogously , the momentum diffusion coefficient is indicated as @xmath10 ( @xmath8 , @xmath9 ) . for the simple case of @xmath11 ( @xmath12 ) , and @xmath13 the approximate transport equation for the function ( 2.1 ) takes the closed form : @xmath14 @xmath15 where we have included the source term @xmath16 . the mean acceleration speed due to the first order process at the shock @xmath17 and the mean escape time due to advection downstream of the shock @xmath18 can be derived from the spatial diffusion equation ( cf .
os93 ) as : @xmath19 @xmath20 where the shock compression @xmath21 . in more general conditions with @xmath22
the function @xmath23 must be explicitly present in the kinetic equation : @xmath24 \right\}\ ] ] @xmath25 @xmath26 let us note in the above equation the asymmetry between the first two terms .
if only the first - order acceleration takes place ( @xmath27 ) , the known solution @xmath28 , with @xmath29 , is reproduced by eq .
obtaining a solution for a more general situation may be a difficult task .
however , if we consider momenta much above ( or below ) the injection momentum , and the power - law form for the diffusion coefficient @xmath30 , the solution is also a power - law . in such conditions , from eq .
( 2.5 ) one can derive the spectral index @xmath31 .
the skilling ( 1975 ) formula is used for the momentum diffusion coefficient , relating it to the spatial diffusion , @xmath32 . for a given jump condition at the shock for the alfvn velocity and a given value of @xmath33 , the resulting spectral index depends on only two parameters , the shock compression ratio @xmath34 and the velocity ratio @xmath35 .
os93 checked the validity range of the approximate equation ( 2.5 ) with the use of numerical simulations . for @xmath36
the equation can be used for the range of parameters preserving @xmath37 , while , for @xmath38 , it provides a quite reasonable description of the particle spectrum at all @xmath39 . in order to find a power - law solution of eq .
( 2.5 ) let us assume the following forms for the diffusion coefficients ( @xmath40 = @xmath41 , @xmath9 ) : @xmath42 , @xmath43 , where the constants @xmath44 and @xmath45 are related according to the formula @xmath46 . with these formulae and the power - law form for the distribution function , @xmath47 , equation ( 2.5 ) yields a quadratic equation for @xmath31 : @xmath48 with coefficients : @xmath49 @xmath50 @xmath51 equation ( 2.6 ) has two solutions valid for , respectively , particle momentum much below ( @xmath52 ) and much above ( @xmath53 ) the injection momentum .
only the later one is of interest in the present considerations .
let us also note that the energy spectral index often appearing in the literature is @xmath54 and the synchrotron spectral index @xmath55 .
in the snr model discussed by chevalier ( 1999 ) the shock wave propagates inside an inhomogeneous cloud . in the cloud , the dense clumps occupy less than 10% of the cloud volume and the shock evolution proceeds mostly in the tenuous inter - clump medium with a number density @xmath56 @xmath1 and a magnetic field @xmath57 g. much stronger magnetic fields ( @xmath58 g ) may occur further downstream in the radiative shock .
the measured shock velocities are in the range @xmath59 @xmath60 km / s . with the notation @xmath61 , @xmath62 and @xmath63
we find the ratio @xmath64 it can be of order @xmath65 for the parameters considered above . thus , the second order acceleration process can substantially modify the energy spectrum of particles accelerated at the shock . in fig - s
1,2 we present spectral indices derived from the formulae of the previous section for the ` canonical ' kolmogorov value @xmath66 .
the shock compression ratio @xmath67 for the high mach number adiabatic shock is assumed .
an efficient particle acceleration requires a high amplitude turbulence near the shock ( cf .
ostrowski 1994 ) and oblique magnetic field configurations may occur there .
thus , the mean magnetic field downstream of the shock is @xmath68 due to shock compression of both the uniform and the turbulent field components .
below we use an effective jump condition for the alfvn velocity @xmath69 ( in general @xmath70 and , respectively , @xmath71 ) .
the results for @xmath72 are considered , where the analytic formulae provide an accurate approximation for the actual spectra . for mean models for the snrs discussed by chevalier ( 1999 ) one
has : a. ) for w44 : @xmath73 , @xmath74 , @xmath75 ( @xmath76 ) b. ) for ic 443 ( shell a ) : @xmath77 , @xmath78 , @xmath79 ( @xmath80 ) c. ) for 3c391 : @xmath81 , @xmath82 , @xmath83 ( @xmath84 ) with these parameters one can easily explain flat spectral indices for the cases ( a ) and ( b ) if reasonable values of the magnetic field are involved .
2 shows the spectral index @xmath85 versus the magnetic field @xmath86 for the above listed choices of particle density @xmath87 and the shock velocity @xmath88 .
the measured indices are indicated on the respective curves .
the steep spectrum of 3c391 can not be exactly reproduced with the use of the simple model considered in this paper , as even neglecting the second order acceleration leads , for the assumed shock compression @xmath67 , to a somewhat flatter spectrum with @xmath89 .
however , the observed trend in the spectral index variation is well reproduced by the model .
a derivation of the particle spectral index at a shock front is presented for a situation involving the second - order acceleration process in the shock vicinity .
we prove that even very weak shocks may produce very flat cosmic ray particle spectra in the presence of momentum diffusion . as a consequence ,
the dependence of the particle spectral index of the shock compression ratio can be weaker than that predicted for the case of pure first - order acceleration .
one should note an important feature of the acceleration process in the presence of the second - order fermi process acting near the shock . because the same alfvn waves scatter particle momentum in direction and in magnitude , there exists a strict link between the first- and the second - order acceleration processes .
the particle spectrum is shaped depending on @xmath35 , by the compression @xmath34 , the momentum dependence of @xmath90 and , possibly , by the anisotropy of alfvn waves determining the ratio of @xmath90 to @xmath91 .
the last factor is not discussed here since under the present considerations we restrict ourselves to isotropic wave fields , but the role of anisotropy consists in decreasing the importance of momentum diffusion . only through the above parameters the spectrum can be influenced by other physical characteristics of the shock and the medium in which it propagates .
it is important that the alfvn waves amplitude , or the magnitude of the spatial diffusion coefficient related to it , do not play any substantial role in determining the spectral inclination as both , first- and second - order , processes scale with it in the same way .
one could argue that an energy transfer from waves to particles can cause quick damping of the waves and will leave us with the pure first - order fermi acceleration at the shock . from the above discussion one can infer that such an objection is not valid for the isotropic wave field .
only the presence of high amplitude one - directional wave field enables the efficient first - order shock acceleration in the absence of momentum diffusion , but a detailed calculation is not possible without a detailed model for the upstream and downstream wave fields . in comparison to parallel shocks , the magnetic field inclined to the shock normal leads to a higher mean energy gain of particles interacting with the shock and a higher escape probability for downstream particles .
also , due to small cross - field diffusion the normal diffusive length scale near the shock decreases . as long as the shock velocity along the field is non - relativistic ,
the particle spectrum produced in the first - order process is not influenced by a field inclination . because of smaller diffusive zones near the shock , the role of momentum diffusion may be of lesser importance as compared to the parallel case .
this effect can be partly weakened by the presence of the magnetic field compression at the shock , which leads to higher alfvn velocity downstream of the shock .
also the presence of high amplitude alfvn waves ( cf . michaek & ostrowski 1996 ) and an admixture of fast mode waves propagating obliquely with respect to the magnetic field ( schlickeiser & miller 1998 , michaek et al .
1998 ) may lead to a more efficient acceleration than the one considered by os93 .
finaly , we would like to note a novel approach to the shock acceleration by vainio & schlickeiser ( 1999 ) , who discuss conditions allowing for the shock generated flat spectra without action of the second order process .
i am grateful to the anonymous referee and to horst fichtner for valuable remarks and corrections .
the present work was supported by the kbn grant pb 179/p03/96/11 .
chevalier r.a . , 1999 , apj , in press ( astro - ph/ \9805315 ) + drge w. , lerche i. , schlickeiser r. , 1987 , a&a . , 178 , 252 + drury l.oc . , 1983 , space sci.rev . , 36 , 57 + hartquist t.w . , morfill g.e . , 1983 ,
apj , 266 , 271 + hartquist t.w . ,
morfill g.e . , 1984 , apj , 287 , 194 + michaek g. , ostrowski m. , 1996 , nonlinear proc .
geo- os93 ) + ostrowski m. , 1994 , a&a . , 283 , 344 + schlickeiser r. , frst e. , 1989 , a&a , 219 , 192 + schlickeiser r. , miller j.a . , 1998 , apj , 492 , 352 + skilling j. , 1975 , mnras , 172 , 557 + vainio r. , schlickeiser r. , 1999 , a&a , ( in press ) + | the particle acceleration process at a shock wave , in the presence of the second - order fermi acceleration in the turbulent medium near the shock , is discussed as an alternative explanation for the observed flat synchrotron spectra of supernova remnants ( snrs ) in molecular clouds .
we argue that medium alfvn mach number shocks considered by chevalier ( 1999 ) for such snrs can naturally lead to the observed spectral indices . |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.